Bad Year Economics: Cultural Responses to Risk and Uncertainty (New Directions in Archaeology)

NEW DIRECTIONS IN ARCHAEOLOGY Editors Francoise Audouze Director, Centre de Recherche Arche'ologiques, Meudon, France Richard Bradley Professor of Archaeology, University of Reading Timothy Earle Professor of An thropology, University of California, Los Angeles Joan Gero Assistant Professor of Archaeology, University of South Carolina Ian Hodder Lecturer in Archaeology, University of Cambridge Patrick Kirch Professor ofArchaeology, University of California, Berkeley Colin Renfrew Disney Professor of Archaeology, University of Cambridge, and Master of Jesus College Jeremy Sabloff University Professor of Anthropology and the History and Philosophy of Science, University of Pittsburgh Andrew Sherratt Department of Antiquities, Ashmolean Museum, Oxford Norman Yoffee Professor of Anthropology, University of Arizona

BAD YEAR ECONOMICS

BAD YEAR ECONOMICS: CULTURAL RESPONSES TO RISK AND UNCERTAINTY EDITED BY PAUL HALSTEAD AND JOHN O'SHEA

CAMBRIDGE

UNIVERSITY PRESS

PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE The Pitt Building, Trumpington Street, Cambridge, United Kingdom CAMBRIDGE UNIVERSITY PRESS The Edinburgh Building, Cambridge CB2 2RU, UK 40 West 20th Street, New York NY 10011-4211, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia Ruiz de Alarcon 13, 28014 Madrid, Spain Dock House, The Waterfront, Cape Town 8001, South Africa http://www.cambridge.org © Cambridge University Press 1989 This book is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1989 Reprinted 1995 by The Ipswich Book Company, Ipswich, Suffolk. First paperback edition 2004 A catalogue record for this book is available from the British Library Library of Congress cataloguing in publication data Bad year economics: cultural responses to risk and uncertainty/ edited by Paul Halstead and John O'Shea. p. cm.—(New directions in archaeology) Bibliography: p. Includes index. ISBN 0 521 33021 1 hardback 1. Economic anthropology. 2. Food supply—Cross-cultural studies. I. Halstead, Paul. II. O'Shea, John. III. Series. GN448.2.B33 1989 306' .3-dc 19 88-30004 CIP ISBN 0 521 33021 1 hardback ISBN 0 521 61192 X paperback

CONTENTS

List of contributors List of List of tables Preface

figures

1 Introduction: cultural responses to risk and uncertainty P. Halstead and J. O 'Shea

vi vii viii ix

1

2 The spirit of survival: cultural responses to resource variability in North Alaska 8 L. Mine and K. Smith 3 Saving it for later: storage by prehistoric hunter-gatherers in Europe P. Rowley-Conwy and M. Zvelebil

40

4 The role of wild resources in small-scale agricultural systems: tales from the Lakes and the Plains J.O'Shea 57 5 The economy has a normal surplus: economic stability and social change among early farming communities of Thessaly, Greece P. Halstead 68 6 Changing responses to drought among the Wodaabe of Niger K.Legge

81

Contents

7 Of grandfathers and grand theories: the hierarchised ordering of responses to hazard in a Greek rural community H. Forbes 87 8 Risk and the polis: the evolution of institutionalised responses to food supply problems in the ancient Greek state P. Garnsey and I. Morris 98 9 Monitoring interannual variability: an example from the period of early state development in southwestern Iran H. Wright, R. Redding and S. Pollock 106 10 Public intervention in the food supply in pre-industrial Europe W. Jongman and R. Dekker

114

11 Conclusion: bad year economics / . O 'Shea and P. Halstead

123

References Index

127 140

CONTRIBUTORS

Rudolf Dekker, Sub-faculty of Economic History, University of Rotterdam Hamish Forbes, School of Archaeology and Oriental Studies, University of Liverpool Peter Garnsey, Faculty of Classics, University of Cambridge Paul Halstead, Department of Archaeology and Prehistory, University of Sheffield Willem Jongman, Sub-faculty of Economic History, University of Rotterdam Karen Legge, Centre for African Studies, University of Liverpool Leah Mine, Museum of Anthropology, University of Michigan, Ann Arbor Ian Morris, Department of History, University of Chicago John O'Shear Museum of Anthropology, University of Michigan, Ann Arbor Susan Pollock, Department of Anthropology, S.U.N.Y. Binghampton, New York Richard Redding, Cranbrook Institute of Science Peter Rowley-Conwy, Clare Hall, University of Cambridge Kevin Smith, Museum of Anthropology, University of Michigan, Ann Arbor Henry Wright, Museum of Anthropology, University of Michigan, Ann Arbor Marek Zvelebil, Department of Archaeology and Prehistory, University of Sheffield

Contents

7 Of grandfathers and grand theories: the hierarchised ordering of responses to hazard in a Greek rural community H. Forbes 87 8 Risk and the polis: the evolution of institutionalised responses to food supply problems in the ancient Greek state 98 P. Garnsey and I. Morris 9 Monitoring interannual variability: an example from the period of early state development in southwestern Iran H. Wright, R. Redding and S. Pollock 106 10 Public intervention in the food supply in pre-industrial Europe W. Jongman and R. Dekker

114

11 Conclusion: bad year economics / . O 'Shea and P. Halstead References Index

123 127 140

CONTRIBUTORS

Rudolf Dekker, Sub-faculty of Economic History, University of Rotterdam Hamish Forbes, School of Archaeology and Oriental Studies, University of Liverpool Peter Garnsey, Faculty of Classics, University of Cambridge Paul Halstead, Department of Archaeology and Prehistory, University of Sheffield Willem Jongman, Sub-faculty of Economic History, University of Rotterdam Karen Legge, Centre for African Studies, University of Liverpool Leah Mine, Museum of Anthropology, University of Michigan, Ann Arbor Ian Morris, Department of History, University of Chicago John O'Shear Museum of Anthropology, University of Michigan, Ann Arbor Susan Pollock, Department of Anthropology, S.U.N.Y. Binghampton, New York Richard Redding, Cranbrook Institute of Science Peter Rowley-Conwy, Clare Hall, University of Cambridge Kevin Smith, Museum of Anthropology, University of Michigan, Ann Arbor Henry Wright, Museum of Anthropology, University of Michigan, Ann Arbor Marek Zvelebil, Department of Archaeology and Prehistory, University of Sheffield

Vll

FIGURES

Page 1.1 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12 2.13 2.14 2.15 2.16 3.1 3.2

Maize yields from southeastern Iowa 1925-45 5 Physiographic features of North Alaska and location of sites discussed 9 Tree-ring chronology of climatic variability 14 Parallel fluctuations in climatic patterns and caribou herd size 16 Aboriginal whaling success at Barrow, Alaska 17 Measures of resource productivity/scarcity and resource variability/predictability from tree-ring chronology 17 Distribution of aboriginal societies in North Alaska, 1850-80 18 19 Representations of inua (species-spirit) Representation of ilitkusiq (breath-spirit) 19 Representations of taktok (spirit doubles) 20 Archaeological phases for North Alaska 21 Distribution of published archaeological surveys in North Alaska 25 Distribution of reported archaeological sites for Period I 25 Distribution of reported archaeological sites for Period IIa27 Distribution of reported archaeological sites for Period IIb29 Distribution of reported archaeological sites for Period III 33 Distribution of reported archaeological sites for Period IV 36 Map of the Rivers Dee and Clwyd, North Wales 41 Average monthly catches on the upper stretches of the Dee41

Figures

3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 5.1 5.2 6.1 6.2 7.1 7.2 7.3 9.1 9.2 9.3 10.1 10.2

Macro-variation in salmon catches on the Dee and Clwyd 42 Intermediate variation in salmon catches on the Dee 42 Intermediate variation in salmon catches, Dee vs Clwyd 43 Micro-variation in salmon catches on the Dee 43 Micro-variation in salmon catches on the upper Dee 44 Very local variation in salmon catches on the upper Dee 44 Long-term variation in salmon catches on the Dee 46 Long-term variation in numbers of caribou 47 Model of settlement variability 48 Principal alluvial plains and location of sites in Thessaly 71 Yields of three field crops in coastal Thessaly 73 Map of southwest and central Niger 82 Rainfall in southwest and central Niger 82 Peninsula of Methana 88 Distribution of total plots and trees owned by Household 2, Kosona 91 Projected wheat production of households, compared with household requirements 93 Uruk sites mentioned in the text 107 Section of pit at Tappeh Sharafabad 108 Survivorship curves for sheep and goat at Sharafabad 109 Price fluctuations in the Antwerp region in the sixteenth century 117 Relationship between income and necessary expenditure on food in Antwerp in the fifteenth and sixteenth centuries 117

TABLES

Page 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3.1 7.1 7.2 7.3 9.1 9.2

Coping strategies vs. spatio-temporal scale of subsistence stress 10 Expectations of climatic variability and resource availability from tree-ring chronologies 15 Values of ring-width growth indices by climatic period 16 Archaeological sites used in regional analyses 22-4 Period I: expectations for resource availability and subsistence strategies 26 Period Ila: expectations for resource availability and subsistence strategies 28 Period lib: expectations for resource availability and subsistence strategies 30 Period III: expectations for resource availability and subsistence strategies 32 Period IV: expectations for resource availability and subsistence strategies 35 Various aspects of salmon catches on the Dee and Clwyd 45 Annual productivity of wheat, Household 30, Kosona (evenyear rotation) 92 Annual productivity of wheat, Household 30, Kosona (oddyear rotation) 93 Annual wheat production compared with minimum household requirements, Household 30, Kosona 94 Bone counts and densities by taxa for two years 109 Sheep and goat fusion data by element for two years 110

Figures

3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 5.1 5.2 6.1 6.2 7.1 7.2 7.3 9.1 9.2 9.3 10.1 10.2

Macro-variation in salmon catches on the Dee and Clwyd 42 Intermediate variation in salmon catches on the Dee 42 Intermediate variation in salmon catches, Dee vs Clwyd 43 Micro-variation in salmon catches on the Dee 43 Micro-variation in salmon catches on the upper Dee 44 Very local variation in salmon catches on the upper Dee 44 Long-term variation in salmon catches on the Dee 46 Long-term variation in numbers of caribou 47 Model of settlement variability 48 Principal alluvial plains and location of sites in Thessaly 71 Yields of three field crops in coastal Thessaly 73 Map of southwest and central Niger 82 Rainfall in southwest and central Niger 82 Peninsula of Methana 88 Distribution of total plots and trees owned by Household 2, Kosona 91 Projected wheat production of households, compared with household requirements 93 Uruk sites mentioned in the text 107 Section of pit at Tappeh Sharafabad 108 Survivorship curves for sheep and goat at Sharafabad 109 Price fluctuations in the Antwerp region in the sixteenth century 117 Relationship between income and necessary expenditure on food in Antwerp in the fifteenth and sixteenth centuries 117

TABLES

Page 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3.1 7.1 7.2 7.3 9.1 9.2

Coping strategies vs. spatio-temporal scale of subsistence stress 10 Expectations of climatic variability and resource availability from tree-ring chronologies 15 Values of ring-width growth indices by climatic period 16 Archaeological sites used in regional analyses 22-4 Period I: expectations for resource availability and subsistence strategies 26 Period Ila: expectations for resource availability and subsistence strategies 28 Period lib: expectations for resource availability and subsistence strategies 30 Period III: expectations for resource availability and subsistence strategies 32 Period IV: expectations for resource availability and subsistence strategies 35 Various aspects of salmon catches on the Dee and Clwyd 45 Annual productivity of wheat, Household 30, Kosona (evenyear rotation) 92 Annual productivity of wheat, Household 30, Kosona (oddyear rotation) 93 Annual wheat production compared with minimum household requirements, Household 30, Kosona 94 Bone counts and densities by taxa for two years 109 Sheep and goat fusion data by element for two years 110

Tables 9.3 9.4 9.5 10.1

Domestic artefact counts and densities by year Chert artefact counts and densities by year Administrative artefact counts and densities by year Distribution of grain reserves in Pavia in 1555

IX

110 111 112 120

PREFACE

The idea of a volume looking at cultural responses to uncertainty and scarcity across a range of social scales and economic systems had its origin in discussions between the editors and Peter RowleyConwy during an extended car journey in 1982. The idea was developed in more concrete form in a symposium entitled 'Cultural responses to risk and uncertainty' at the 1984 Theoretical Archaeology Group (TAG) meeting in Cambridge. The session brought together many of the present contributors and established an organisation that has been retained in this volume. Following the TAG session, the participants were asked to produce expanded versions of their papers, and a number of other researchers with similar interests were approached in an effort to broaden the scope of the volume. The present volume consists of nine studies, ranging in their focus from simple hunter-gatherers to modern states. We have added to these an introductory and concluding chapter, which attempt to draw the various studies together and to consider at a more general level the potential value and limitations of the study of risk and uncertainty. The editors would like to acknowledge a number of individuals who helped make this volume possible. Todd Whitelaw bore much of the burden for organising the TAG session and we gratefully acknowledge his contribution to its success. Thanks are due to Glynis Jones, who compiled the unified bibliography, Amy Steffian, who created the index, and Kay Clahassey, who produced many of the illustrations for the volume. We would also like to thank the contributors for enduring our editorial interven-

Tables 9.3 9.4 9.5 10.1

Domestic artefact counts and densities by year Chert artefact counts and densities by year Administrative artefact counts and densities by year Distribution of grain reserves in Pavia in 1555

IX

110 111 112 120

PREFACE

The idea of a volume looking at cultural responses to uncertainty and scarcity across a range of social scales and economic systems had its origin in discussions between the editors and Peter RowleyConwy during an extended car journey in 1982. The idea was developed in more concrete form in a symposium entitled 'Cultural responses to risk and uncertainty' at the 1984 Theoretical Archaeology Group (TAG) meeting in Cambridge. The session brought together many of the present contributors and established an organisation that has been retained in this volume. Following the TAG session, the participants were asked to produce expanded versions of their papers, and a number of other researchers with similar interests were approached in an effort to broaden the scope of the volume. The present volume consists of nine studies, ranging in their focus from simple hunter-gatherers to modern states. We have added to these an introductory and concluding chapter, which attempt to draw the various studies together and to consider at a more general level the potential value and limitations of the study of risk and uncertainty. The editors would like to acknowledge a number of individuals who helped make this volume possible. Todd Whitelaw bore much of the burden for organising the TAG session and we gratefully acknowledge his contribution to its success. Thanks are due to Glynis Jones, who compiled the unified bibliography, Amy Steffian, who created the index, and Kay Clahassey, who produced many of the illustrations for the volume. We would also like to thank the contributors for enduring our editorial interven-

Preface

tions with remarkable patience and good humour and for producing a series of papers with which we are very proud to be associated. Finally, we must thank Glynis Jones and Sue O'Shea who, by their encouragement, persuasion and threats, curbed our more glaring excesses and ensured our completion of this task.

Chapter 1 Introduction: cultural responses to risk and uncertainty Paul Halstead and John O'Shea

The world about us is in a constant state of flux. The nomad in the Kalahari Desert and the Western city-dweller in front of his television set are both repeatedly confronted by changes in their natural surroundings and in the behaviour of other members of their own species. In this respect, at least, their experience is no different from that of all other organisms. Culture endows man with exceptional flexibility in coping with his surroundings and, in consequence, human beings regularly cope with an unusual diversity of natural and social environments. As a result, the normal lifestyle of people in different parts of the world can be radically different. Yet the inherent temporal instability of these environments still poses problems and unusually severe perturbations frequently claim human lives. Shortage of food, one of the most basic and yet least reliable of the requirements for human survival, remains a common cause of loss of life. The means by which human beings secure their food supply in the face of such uncertainty are thus as central to society as the consequences of shortage are drastic and they have far-reaching ramifications throughout cultural behaviour and social life. Human communities have developed an impressive array of cultural mechanisms for buffering variability. The diversity of these mechanisms, however, should not mask the fact that an effective strategy must match, in both capacity and scale, the variability with which it is to cope. The structure of environmental variability will be discussed more fully below, as will that of the corresponding cultural coping mechanisms. An issue of particular

importance, however, is that of predictability. Individual hazards are often quite unpredictable, in terms of their precise timing and so on, and from the perspective of a potential victim this element of uncertainty significantly exacerbates the risk from perturbations. From the more sheltered perspective of the social scientist, however, the basic structure of variability - its frequency, duration, spatial scale, severity and regularity - can often be predicted quite accurately for any given situation. It follows that the basic structure of an effective coping mechanism in the same context can also be specified. This emphasis on variability and on cultural responses to it has important implications for both synchronic and diachronic studies in the social sciences. Synchronically this approach provides a powerful analytical tool for investigating a range of cultural phenomena attested ethnographically. Bitter controversy surrounds the interpretation of a number of these phenomena, with rival camps favouring 'functionalist' or non-materialist explanations. Both sides proceed in essentially the same, flawed manner, picking a puzzling aspect of human behaviour and then proceeding to explain it on an ad hoc basis in terms of some preferred cultural or environmental factor. The power of the approach presented here is that, since the basic structure of variability (and hence the basic structure of a successful coping mechanism) can be independently specified, the potential of individual cultural forms as buffering mechanisms can be investigated with rigour. This approach offers similar analytical potential for the syn-

Paul Halstead and John O 'Shea chronic investigation of ancient societies. Again, the basic characteristics of the necessary buffering mechanisms can be independently predicted and the incomplete historical or archaeological record of past cultural practices and institutions can then be evaluated in this light. This analytical potential is exploited in the following chapters in the investigation of a number of ethnographically, historically or archaeologically documented societies. Diachronically this framework has unusual heuristic value for the investigation of long-term processes of social and cultural change. Variability, particularly that which results in severe and unpredictable scarcity of vital resources, exercises a powerful selective pressure on human behaviour. Moreover, most of the buffering mechanisms deployed by man are cultural, and of these the most powerful tend to be social mechanisms. Environmental variability is thus a powerful force for long-term social change and in this respect the approach adopted here addresses one of the most critical areas in the social sciences. Given the universal prevalence and potentially drastic consequences of variability, it is perhaps surprising that social scientists have in the past concentrated on the 'normal' or 'average' conditions of human existence. In part this simply mirrors a practical problem common to all the social sciences variability is harder to grasp and to present than some measure of central tendency. Moreover, different aspects of variability are relevant at different scales of analysis - this issue of analytical scale is one to which we must return later. There are also other reasons, relevant to individual disciplines within the social sciences. For example, ethnographic observations rarely cover a sufficient time span to encompass the full range of variability affecting the community under study. Anthropologists are frequently more interested in cultural 'norms' - perceptions of what should happen - than in what actually happens. Similarly, some of the early agronomic treatises central to ancient economic history are as much prescriptive statements of agricultural theory as descriptive accounts of agricultural practice. Finally, archaeologists, faced with the need to model past economic systems for which they have no direct statistical data, have usually preferred the simplicity of averages to the complexity of variability. In each case the traditional emphasis on norms or averages is readily comprehensible in its disciplinary context, but these figures are an abstraction of the variable reality with which mankind must of necessity cope. This variability exercises an important influence on the ways in which human beings behave and, when it extends to vital resources, on their very survival. This volume, then, poses three basic questions: (1) How do societies buffer themselves against periodic variation in food availability? (2) How do these coping activities influence other aspects of cultural organisation? (3) To what extent can these coping strategies provide the impetus for social change? Culture and environment Human beings do not exist in a vacuum. Rather, their behaviour is constrained by their surroundings or environment. Their environment has three major components: abiotic (physicalchemical surroundings), biotic (organisms of other species) and

social (organisms of the same species). Needless to say, human beings mould all three components of their environment to a greater or lesser extent, so we are not concerned here with rival deterministic claims, but rather with the interaction between man and his environment. From a human perspective, this interaction can usefully be construed as a game in which the environment poses problems which man must solve. The particular problems posed by the environment in any individual case are of course a product both of local environmental conditions and of the specific nature of human behaviour in that context. For example, the specific problems posed today by the natural environment of southern England would be quite different for human beings trying to ensure their survival by hunting and gathering, farming or stock-broking. The boundary between environment and culture is thus difficult to draw. For the sake of simplicity, however, two conventional distinctions are drawn here. Firstly, the behaviour of human beings outside the social unit of analysis will be considered an aspect of the environment and the behaviour of those inside the unit of analysis will be considered part of culture. From the perspective of the individual household, for example, taxation is as much a part of the environment as is a bolt of lightning, while from the perspective of the state it is an integral component of the cultural repertoire. The boundary between environment and culture depends, therefore, on the scale at which analysis is conducted. To a great extent this convention also distinguishes between circumstances which can be controlled and those which cannot. Secondly, problems posed by the interaction of culture and environment are treated as aspects of environment. From the perspective of the peasant farmer, the distinction between the environmental problem of drought and the cultural problem of a bad harvest (i.e. of cultural failure to mitigate the effects of drought) is largely semantic. Similarly, personal accidents of injury or illness are treated here as environmental problems, regardless of the role of human error in their causation. These conventions perhaps blur some of the traditional dichotomies beloved of social scientists, but we are primarily concerned here with the problems faced by human beings in the real world, and not with the polarised abstractions of academics. A rather different problem is raised by the complexity of the natural and social environment and hence by the multiplicity of problems which the environment poses, whether practical problems for the inhabitant of the real world or analytical problems for the inmate of the ivory tower. A useful analytical tool in this respect, which is widely used to reduce complex ecological problems to manageable proportions, is the concept of the 'limiting factor': although organisms must cope with a multiplicity of environmental problems, their survival is usually limited in any given context by just one or two critical resources. The limiting factors on human survival obviously vary from case to case, depending on the particular environmental and cultural context and on the temporal, spatial and social scale of analysis. Certain general observations are, however, possible. Respiration is a requirement of human survival which, in terms of its immediacy, takes logical precedence over virtually all others. Pre-

Introduction: cultural response to risk and uncertainty cisely because of its immediacy, however, this requirement is met by normal biological mechanisms and so rarely sets the limits on survival. At the other extreme, hazards such as dormant volcanoes may operate as limiting factors so rarely that they can be ignored by human populations. Such hazards can be, and regrettably are, 'coped' with by the ability of human fertility to compensate for occasional catastrophic mortality. Between these two extremes are other, more critical variables operating on a timescale closer to that of the human lifespan, which can be coped with neither by normal biological mechanisms nor by drastic demographic fluctuations. Culture is particularly well suited to coping with these problems on an intermediate temporal scale. The availability of staple foods is of outstanding, though by no means unique, importance in this respect. Because food is needed very regularly, and yet tends to be both irregular in its availability and unstable once acquired, it has long been a basic limiting factor on human survival. The structure of variability in food supplies is discussed in some detail in the next section. Variability In practice, variability may be conceptualised in two differing ways: as the actual pattern of variation in the food supply, or as the operation of those factors, ranging from climate to microorganisms to human judgement, that influence the availability of a particular food resource. Regardless of the focus, the crucial aspect of the analysis is the timing, frequency and severity of shortages. The influence of variability on human societies can be understood with reference to several key aspects of variability itself. Any source of variability will exhibit three diagnostic characteristics, temporal structure, spatial structure and relative intensity. These characteristics determine the scope and severity of the problem with which a society must contend. Knowledge of these features of the environment also provides the analyst with a powerful tool for predicting the kinds of strategies that would be successful in a given situation. By temporal structure, we refer both to the timescale over which a given risk operates (i.e. how often it occurs) and to the duration of the resultant scarcity (i.e. how long it lasts). Spatial structure relates to the size of the affected area. A consideration of spatial structure also concerns the relative homogeneity of effect within an area, that is, the degree to which scarcity is evenly (or patchily) spread over the landscape. Hail damage to agricultural fields may occur over a wide area in association with a given storm cell and yet destroy fields within this area seemingly at random, with one field devastated and the next field untouched. Finally, intensity refers to the severity of shortages and to the degree of variation in severity that can occur. A particular cause of crop failure may be consistent or variable in the severity of its effects. Intensity may, therefore, be seen as having both a temporal and a spatial component. In addition to these characteristics, variability can also be classified in terms of its relative predictability. Sources of variation that are predictable may represent seasonal or annual phenomena, such as the production of vegetal foods in the temperate zone, or

they may be cyclical over a variety of longer time scales (see Mine and Smith, this volume). What sets variation of this kind apart from other forms is its certainty. The effects are repeatedly felt at regular intervals. Likewise, the cultural responses to such scarcity can be regular and unambiguous. Indeed, within a given environment, a society's ability to cope successfully with such regular and predictable kinds of variation may be viewed as the minimum necessary conditions for survival and, as such, integral to normal existence. Unpredictable variation may result from common factors, such as climate, isolated and sporadic occurrences, such as some forms of pest infestation, or may be cyclical phenomena on a timescale too long to be recognised. The lack of predictability associated with variability of this kind is often more difficult to buffer for the very reason that one never knows when the buffering mechanism will be called into play. So, where buffering mechanisms for times of expected scarcity are direct and unambiguous, the institutions employed to buffer unanticipated shortages may well perform many other functions within the society, particularly in good years. The necessary compromise with the institution's varied social functions may also sacrifice efficiency in risk buffering. Furthermore, the risk-buffering function of the institution may not be readily apparent either to an outside investigator or to a native. By evaluating the factors that produce scarcity in terms of these three basic characteristics, therefore, one can obtain a relatively detailed profile of those elements with which a buffering strategy must cope. At the same time, the detailed evaluation of these characteristics may highlight unexpected solutions to problems of scarcity, made possible by exploiting particular aspects of the structure of variability. The nature of such strategies will be addressed in the next section. Cultural responses to variability To counteract scarcity, societies employ a wide range of practices which we term collectively 'buffering mechanisms'. As this term implies, such practices are designed to lessen the impact of variability by dampening its effects. Buffering may be realised through very different kinds of activity, including everything from myth to alternative cultivation practices and from storage to exchange. The usefulness of a given practice depends, of course, on the social and environmental context, including both the structural characteristics of the society at large and the structure of resource failure the society is likely to experience. Despite the wide range of practices that may operate to lessen the effects of resource variability, such responses can conveniently be grouped into four basic categories: mobility, diversification, physical storage and exchange. Mobility is the simplest of these responses and works by taking advantage of the spatial and temporal structure of resource failure in effect to move away from scarcity. The common huntergatherer system of flexible territorial boundaries and extensive kin networks that permits bands to move great distances in years of drought is a good example. One corollary of this use of mobility is to place a high premium, among both hunter-gatherers and pas-

Paul Halstead and John O 'Shea toralists, on the gathering of information to monitor environmental variability over a large area beyond their normal annual range. Diversification is nearly as simple as mobility and includes a broad range of passive to active practices. The underlying principle of diversification is that broadening the base of the subsistence system, either by exploiting a wider range of plant and animal species or by exploiting broader and more varied areas, reduces the risk of catastrophic shortages - in effect it avoids placing all the eggs in one basket. At the passive end of the spectrum is the designation of reserved or emergency foods, that is, foods that under ordinary circumstances are not or cannot be eaten, but which are consumed in the face of extreme hunger. This represents almost a passive form of banking or savings. At the more active end of the spectrum is the use by farmers of dispersed fields (see Forbes, this volume), which are employed to lessen the risk of total loss due to micro-climatic factors, pests and the like, even at the expense of less efficient cultivation. By the term 'physical storage', we refer to that range of activities that is directed toward stabilising available food so that it may be consumed at some later time. This is principally a means of dealing with the temporal structure of food availability. Although storage is of critical concern to agricultural societies, it may also play an important stabilising role in non-agricultural economies (see Rowley-Conwy and Zvelebil, this volume). The final category of buffering strategy is exchange. The concepts of sharing and of reciprocity are virtually universal as social values and have been suggested by some investigators to be at the very core of the origins of human culture. As a device for buffering scarcity, exchange functions in a fashion similar to storage, in that present abundance is converted, this time via social transactions, into a future obligation in time of need. If I help my neighbour out of a lean season this year, I have the right to expect the aid to be reciprocated when the situation is reversed. The capacity of such relations, and indeed the certainty of reciprocation, can vary widely depending on the scale of the social units involved and the character of the exchange networks. So, for example, in small-scale societies, food may be given freely in time of need with little formal recognition of the anticipated reciprocity, while in more complex circumstances, it may be necessary to denote obligation through the counter-movement of valuable tokens. The process of conversion, through which a token becomes imbued with the latent equivalent value of food, finds its logical extension in modern cash economies. Interactions that are characterised by 'negative reciprocity', such as raiding, theft and appropriation, can also be placed under the broad heading of exchange. All four categories of buffering mechanism exploit favourable aspects of the temporal and spatial structure of variability to mitigate the risk of scarcity. Mobility and diversification use local abundance to counter local scarcity, while storage balances seasons of plenty against lean seasons and good years against bad. Exchange secures a stable food supply by playing off temporal variability against spatial variability. As a result, different types of response are suited to buffering different sorts of risk.

Two categories of response may be mutually incompatible: storage, for example, seriously prejudices mobility and vice versa. On the other hand, different categories of response may also be combined in a single buffering strategy. The 'hxaro' system of guest friendship among the Bushmen, for example, whereby those in need visit distant exchange partners, neatly combines mobility and exchange. More important still, most societies deploy an array of different strategies in a hierarchy of responses, which are equated with both the scale of the producing and consuming units (individuals, households, villages, states) and with the magnitude of the resource failure encountered. In this sense, societies are faced with a trade-off between security, as reflected in the ability to withstand increasingly severe and rare fluctuations, and efficiency in terms of normal subsistence activities and of avoiding investment in precautions that will only rarely be used. Relatively minor shortages may be handled by a single mechanism at a local level, while serious or generalised resource failure may require mobilisation of the entire population and progressively more drastic counter-measures. In the light of this difference in the scale of response, a distinction may be drawn between low-level and highlevel mechanisms. Low-level mechanisms are the most efficient and the most reliable, but are of strictly limited scope. High-level mechanisms are more powerful, in terms of the scale of shortage which they can buffer. By virtue of their size and the relative rarity with which they may be activated, however, they can be both costly, in the energy invested in their maintenance, and unreliable, in that they may depend on distant social relations, may entail reversal of cultural norms and, more generally, may fail through long periods of disuse. In a way, the fate of high-level coping mechanisms is shaped as much by the frequency and regularity with which they are not used as by that with which they are activated. In the long term, therefore, there is strong selective pressure for communities to embed these 'emergency' mechanisms as regular aspects of the cultural repertoire (see Garnsey and Morris, this volume). For example, surpluses may be earmarked for feasting at a particular annual ceremony, thus ensuring their continued production through a long run of good years. Whilst this embedding may reinforce the survival of a cultural form used only rarely as a coping mechanism, it may also undermine this latent function if the surplus cannot be redirected in time of need. In this way, high-level mechanisms may be embedded to the point where they are irreversibly transformed, with radical consequences for the articulation and survival of that society (see Jongman and Dekker, this volume). Predicting uncertainty

In any society, therefore, we can expect an essential balance between those factors that produce scarcity and the mechanisms that are employed to ameliorate these effects. Yet, for any given environmental setting, there are numerous alternative strategies that could successfully be pursued. The particular strategy that is likely to be employed is strongly influenced and constrained by several features of the society in question, including the size and density of the local (and regional) population, the intensity of subsistence exploitation and the level of technological sophistication.

Introduction: cultural response to risk and uncertainty

1935

Year

1935

Farmington Township Des Moines Township

1940

Year . . .

1945

19!

Lee County, Iowa Appanoose County, Iowa

Fig. 1.1. Maize yields from southeastern Iowa, 1925-45, for (a) the two most dissimilar townships within Van Buren county and (b) the two most dissimilar counties within southeastern Iowa.

These aspects of the cultural ecology still do not determine the particular strategy a society will practise, but they do limit quite substantially the number of successful options. Because of the limited range of choices open to a society, given these factors, it is possible to use data on variability to predict both the kinds of buffering strategies that will be employed and to estimate the parameters the various mechanisms must maintain in order to be effective. This predictability is of particular use in the study of prehistoric societies, where data on social relations and activities are necessarily indirect, but where the structure of resource variability can often be determined independently. The nature of resource variability and of its potential fit with cultural buffering mechanisms can be illustrated by reference to maize yields from southeastern Iowa (USA) during the years 1925 to 1945. Figure 1.1a plots yields in terms of bushels per acre for the two most dissimilar townships within Van Buren county (a township is a square region, six miles [9.7 km] to a side, while Van Buren county covers an area of roughly 500 square miles). Figure 1.1b plots yields over the same period of time but for the two most dissimilar counties within southeastern Iowa, a territory encompassing roughly 6000 square miles. Maize yields in the region are characterised by relatively high temporal variability (i.e. considerable variation from year to

year) and at the same time by a very low level of spatial variability (i.e. little variation from place to place). Given such a pattern of variability, the most likely mechanisms for accommodating years of bad harvest would be local storage of grain in good years and diversification, either in the crops grown or through the exploitation of non-agricultural food sources, such as livestock. In the absence of a sophisticated transportation network, however, the lack of spatial variability over very large areas would seem to rule out exchange as an important buffering strategy. Although this example highlights the relationship between the structure of variability and the selection of effective countermeasures, it can also be used to illustrate how cultural variables, such as population density and subsistence technology, exert a strong influence on the definition of the effective environment. Modern-day farming in the region is extensive and largely rainfed, meaning that summer precipitation is a crucial factor affecting crop yield over the entire area. Yet, when the same basic crop, maize, was cultivated by native Americans in the region, using small, optimally located and widely separated fields, the factors producing variation in yield were dramatically different. Although the fields were still subject to the same precipitation regime, smaller-scale effects, such as soil type, micro-climate, insects, fire and hail increased in importance as factors producing significant variability. As a result, spatial variability was probably greater than that represented for modern farmers, and may have been sufficient to make some level of exchange a useful option. Risk buffering and social change

The importance of variability and buffering lies not only in the immediate realm of provisioning and economic activities, but extends beyond these to exert a strong influence on culture at large, shaping societal organisation and providing the crucial conditions that give rise to social change and transformation. It is these wider ramifications that make the investigation of variability and buffering so significant. Although human beings are biologically and culturally adapted to a wide range of environments, the hallmark of culture is its flexibility and the strength of human culture is most fully realised in coping with recurrent but unpredictable deviations from normality. Yet, within any society, the delicate balance between variability and cultural response is matched by a similar balancing between increased security, represented by progressively more drastic and expensive buffering mechanisms, and increased efficiency, allowing only for common eventualities. The most powerful mechanisms, and the most costly, are those that cope with problems of unusual severity or exceptional scale. Although prone to falling into disuse because of the infrequency with which they are activated, these high-level coping mechanisms may serve a critical function in cases of extreme shortage. As a result there is strong selective pressure for them to become increasingly embedded within more regular cultural practices and so, potentially, to develop widespread ramifications throughout the social system In this way, the critical energy or information required for the operation of the buffering mechanism is maintained, but often at the cost of considerably reduced ef-

Paul Halstead and John O 'Shea ficiency. At the same time, this process may serve to promote the social institution(s) within which the coping mechanism is embedded. It thereby creates in these multi-layered social institutions both the potential and the means for social change. One likely outcome of this process is the radical transformation of society, as perhaps in the appropriation of risk-related surplus for the maintenance of an elite. In this case associated changes in other aspects of social or economic behaviour are likely to unleash new sources of risk, leading either to the development of new coping mechanisms, and potentially to the start of another cycle of cultural change, or to the undermining of existing coping mechanisms, resulting in catastrophic rather than transformational change. Either way, high-level coping mechanisms are an unstable element at the core of human culture, which can trigger a chain reaction of changes throughout society. Prospect The conclusion of the previous discussions has been that cultural mechanisms for coping with scarcity play a central role in the maintenance and transformation of human societies. This approach also has rather broader implications, which may be considered at three levels - empirical, heuristic and theoretical. At an empirical level, recognition of the importance of variability may help to resolve some apparent contradictions in ethnographic or historical records. Such records are usually assumed to represent normal or average conditions and contradictory information is thus explained in terms of differing local circumstances or in terms of the unreliability of certain sources. In fact many ethnographic and historical reports are based on too few observations to make a reliable estimate of average conditions and so these 'contradictions' may be accurate reflections of actual variability, rather than inaccurate estimates of a hypothetical norm. Of course the quality and quantity of information needed to document variability directly is far greater than that required to produce a simple average figure and such data are often unavailable. The basic structure of variability in a particular situation, however, may often be extrapolated from better-documented cases - indeed, the structure of variability can often be established even in cases where the average is unknown. For example, the relative structure of variability in the ancient crop yields from a particular area may be established by extrapolation from modern data for the same area without estimating any absolute mean value. This point is of particular importance for prehistorians, who rarely have direct evidence for such variables as average crop yields and, as has already been argued, the structure of variability may be more significant than average conditions for understanding human society. At a heuristic level, an emphasis on variability is of considerable value in the analysis of cultural practices and institutions. Environmental variability is a fact of life and extreme variability may threaten life. The consequences of variability must be evaded or buffered and man employs a wide array of cultural mechanisms to this end. It must be stressed that not all cultural behaviour serves to buffer risk, nor can any individual cultural form

be understood solely in such terms. Nonetheless, some practices and institutions serve, literally, a vital role in mitigating the effects of variability and some of these buffering mechanisms occupy a central position in human behaviour and in the articulation of human communities. In this volume, a wide range of cultural practices and institutions is examined from this perspective, ultimately with a view to evaluating the heuristic value of this approach for understanding social change. This approach may also be useful at a more theoretical level. Although the study of economy-of how societies satisfy their (material) wants-has long been a central concern of social science, this area of enquiry is the subject of considerable controversy. Economic and social determinists dispute the importance of this concern to the study of human society; environmental and cultural determinists dispute the definition of wants; and substantivists and formalists dispute the means by which these wants are satisfied. No polar stance is required. The environment is understood to include both a natural and a social component. Moreover, the environmental problems facing man are specific to particular cultural contexts and as such are the product of both (external) environmental and (internal) cultural factors. Finally, the indisputable anthropological observation that economic activity is normally socially embedded in no way invalidates the basic tenet of economics that this activity is frequently concerned with the allocation of resources which are scarce, and in some cases fatally scarce. On the contrary, human communities use a variety of cultural mechanisms to cope with risks such as the extreme scarcity of resources, and the most powerful of these mechanisms entail the mobilisation of social relationships. Viewed from this perspective, these traditional polarisations are as obstructive to the study of man as they are non-sensical. Organisation of this volume The contributions in this volume examine the problems of variability and risk buffering over a wide range of scales and in strikingly different cultural and historical contexts. The diversity is intentional and is designed to emphasise both the wide applicability of the approach adopted here and the wealth of different ways in which human communities cope with environmental variability. The papers that follow are ordered along a continuum of increasing organisational complexity, starting with simple huntergatherers and ending with modern states. This organisation was selected for two basic reasons. First, buffering strategies tend to be additive, that is, as societies become larger and more complex, new higher-order mechanisms are added or superimposed onto more basic practices. The present organisation of chapters therefore allows the more fundamental kinds of mechanisms to be examined first, in the contexts of hunter-gatherers or simple farmers, and then for a variety of higher-order mechanisms to be examined which operate concurrently in more complex societies. The second value of the present organisation is the emphasis it places on the increasing scale and costs of buffering strategies as the focus shifts from simpler to more complex social forms. It should be stressed that these contributions are all case

Introduction: cultural response to risk and uncertainty

7

studies, in which the authors have selected a particular aspect of variability and risk buffering on which to concentrate. No author has attempted to analyse all potential sources of risk or all the possible cultural practices that might play some role in risk buffering. This process of selection, exemplified in the individual chapters, highlights the value of the approach as an analytical tool, en-

abling the investigator to use risk buffering at a given scale as a means to focus on particular kinds of social institutions or practices for either synchronic or diachronic study. Equally, it stresses the practical necessity of clearly defining the nature and scale of the problem, given the nested and all-pervasive character of both risk and risk-buffering activities.

Chapter 2 The spirit of survival: cultural responses to resource variability in North Alaska

Leah D. Mine and Kevin P. Smith

While the basic structure of responses to scarcity is constrained by the nature of those stresses which coping mechanisms must mediate to be effective, the implementation of coping strategies is predicated on the sociocultural context, which defines the range of organisational and technological options for mediating periods of subsistence stress. In this chapter, we reconstruct the spatio-temporal scales of variability in the major faunal resources of interior and coastal Alaska for the late prehistoric and protohistoric periods from variability in relevant climatic and ecological factors. From the structure of resource variability, we predict the basic structure of coping responses, and examine how specific coping strategies were modified over the past 1000 years to adjust to changes in resource structure and sociocultural context.

While environmental changes of certain magnitudes require adaptive adjustments in subsistence behaviour, the nature of the response is determined in large measure by sociocultural rather than by environmental variables (Euler, Gumerman, Karlstrom, Dean and Hevly 1979: 1089). At the time of European contact in the early 1800s, Inupiat Eskimos inhabiting coastal and inland North Alaska pursued ecologically distinct ways of life. The tareumiut, or 'people of the sea', practised a subsistence economy based on sea-mammal hunting, with an emphasis on whaling. Tareumiut settlement was accordingly distributed along the coast in relatively permanent villages. The nunamiut, or 'inland people', in contrast, focused primarily on

hunting the migratory caribou and ranged widely over the interior regions from the Brooks Range north to the Arctic Coastal Plain (Figure 2.1). For traditional hunting societies, the interior and coast constituted complementary resource zones which experienced similar short-term and long-term patterns of scarcity in their primary subsistence resources, but differed significantly in the timing of those fluctuations, as well as in the density, distribution and variety of alternative resources. By the historic period, cultural responses to subsistence stresses of the coast and interior had given rise to a survival strategy in which the Tareumiut and Nunamiut were economically interdependent societies, linked by inter-regional alliances which converted short-term resource abundance into longterm social insurance. The particular configuration of responses to scarcity developed by the Nunamiut and Tareumiut represents the end point of an extended history of survival tactics, in which the changing structure of resource variability and the emergence of new institutions and technology continually redefined the constraints on, and options for, survival. In this chapter, we examine the interplay of natural and cultural factors constraining hunter-gatherer responses to risk in North Alaska. We first outline a model linking the basic structure of survival strategies to the basic spatial and temporal structure of subsistence stress for hunter-gatherers. We then evaluate the sociocultural context of survival strategies, including the social, technological and cognitive systems which sup-

The spirit of survival

6 Ekseavik

11 Pingok Island

2 Onion Portage

7 Walakpa

12 Tigara/Tikera

3 Nunagiak

8 Utkiavik/Utquiagvik

13 Kangiguksuk

9 Sikoruk/Tukuto Lake

14 Kavik

1 Cape Krusenstern

4 Thetis Island 5 Kotzebue

10 Ivisahpat

15 Niglik 16 Nuwuk

Fig. 2.1. Major physiographic features of North Alaska and sites specifically discussed within the text

port their maintenance and utilisation. Finally, we examine the operation and development of coping strategies in North Alaska over the past millennium, through consideration of the changing structure of subsistence resources and the changing sociocultural context for mediating subsistence stress. The structure of resource variability and coping responses

Halstead and O'Shea (introduction to this volume) have identified four basic categories of response to risk which can be considered general coping strategies. These are (1) diversification, which attempts to counteract scarcity of one resource through recourse to others; (2) mobility, which attempts to even out spatial discrepancies in resource availability by movement between areas of localised resource abundance; (3) storage, which attempts to even out temporal discrepancies in resource availability, by 'saving it for later'; and (4) exchange, which attempts to play off temporal variability in resource availability against spatial variability in resource abundance (cf. Colson 1979; Minnis 1985: 3242). As general coping strategies, diversification, mobility, storage and exchange potentially operate over a broad range of spatial and temporal scales. For example, although a particular form of storage may be limited in its efficacy to mediating crises of relatively short duration (Rowley-Conwy and Zvelebil, this volume), the strategy of storage (that of evening out temporal discrepancies in resource availability by 'saving it for later') is not

necessarily so limited. It is the adaptation of these general strategies to the specific structure of subsistence stress which generates the variety of cultural phenomena collectively known as 'buffering mechanisms'. Responses to subsistence stress are effective only if they transcend, spatially and temporally, the adverse conditions. Accordingly, an analysis of resource variability provides a basis for characterising the basic structure of mediating responses by specifying those contingencies which a given strategy must meet to be effective. Resource variability can be measured along a number of axes, reflecting differences in resource abundance, predictability, and duration and extent of fluctuations, relative to the availability of alternative resources and the coping capabilities of the group under consideration. For example, the productivity of a resource determines its potential for utilisation, while greater predictability or stability permits a greater reliance on a resource, ceteris paribus. However, productivity is a function of both physical characteristics, such as body size and abundance, and such cultural factors as the organisational and technological capabilities of a society to be successful in procuring that resource (Jochim 1976, 1981). Similarly, the reliability of a resource results both from spatiotemporal fluctuations in its availability and from the success rate of a particular method of procurement. The productive potential and reliability of a resource will therefore vary through time and space in response to differences in resource structure as well as to differences in hunting/storage tactics and technology.

Leah D. Mine and Kevin P. Smith

10

Table 2.1. Coping strategies vs. spatio-temporal scale of subsistence stress

increases, the size of (a) the spatial area over which a given strategy operates and (b) the social unit integrated by that response must correspondingly increase. For instance, in the event of primary resource failure, procurement may be diversified to include local secondary resources and famine foods. However, if the duration of the failure extends beyond the limited support capacity of local secondary resources, diversification must be expanded to include non-local resources. Similarly, as the size of the area monitored or exploited increases, so does the degree of mobility. More frequent residential moves or longer-distance forays may be adequate responses to reduced local resources, but effective utilisation of non-local resources may involve extra-regional migration. Exchange networks can also expand to dilute the impact of subsistence stress. Intra-societal exchange or sharing evens out short-term, localised differences in hunting success through the pooling of resources. In a long-term or regional resource failure, however, where survival depends on access to resources and goods outside the normal network of reciprocal obligations, social access will be maintained through increasingly formalised means of exchange, such as trading partnerships, or through delayed obligations inherent in 'social storage'. In the absence of accepted, mutualistic avenues of exchange, necessary resources may be acquired through negative reciprocity, in the form of raiding and theft. While specific tactical uses of the various coping strategies may be mutually exclusive, at other levels of analysis, and in particular structures of response, they represent nested levels of response within the overarching survival strategy utilised by a society. For example, where high interannual variability makes the amount of a resource which is needed unpredictable (such as when resources are stored for use over the winter, but winter conditions vary greatly), one effective strategy may be physically to store enough of the resource to overcome the 'worse-case' scenario (Rowley-Conwy and Zvelebil, this volume). In good years, unused surpluses from such a tactical response may then be converted into social alliances and debts, through gift giving and the hosting of feasts, thereby creating a form of social storage against long-term variability. Examination of adaptive strategies must therefore concurrently consider solutions to long-term fluctuations and shortterm variability in resource abundance and predictability, as these form the matrix upon which both immediate decision-making and long-term processual change are operative.

Continuum of stress conditions General coping strategies

Low severity (localised, short-term)


i

xV '

\

/

\

/

A

A

A

\/ \ V * / \\ * / \\ v y \ ,. ---.. / "".--. \ "* * •

.•

•-.

10,000 -

5,000 -

i,

5000 '

/ v

j

/ /

Clwyd total (including sea trout in estuary nets)

/

A 25,000 •

Dee and Clwyd total Dee total (salmon)

\-

1960

1970

1980 1970 Dee net catch as % of total 90 T

Dee catch as % of total

80-

80

70-

60-

60 50 -

1960

1980

Fig. 3.3. Macro-variation in salmon catches: the Dee and Clwyd 1953-83. a: total catch; b: Dee proportion. Source: Annual Reports.

To some extent, such fluctuations may cancel each other out. In Arctic terms, a bad caribou year will not always be a bad seal year. Not all resources vary independently of each other, however, particularly if they belong to the same food chain. Even where this is not the case (as with seal and caribou), a bad caribou year is bound to coincide with a bad seal year sooner or later (but see Mine and Smith, this volume), with the result that overall carrying capacity will vary dramatically from year to year. In particular, troughs in resource availability will vary in severity. The salmon catch on the Dee and Clwyd, for example, plummeted in 1968 after 15 years of variable but rising catches (Figure 3.3a). Spatial variability again adds a further dimension. At a continental level, most high-latitude areas have markedly uneven resource distribution, with concentrations along watercourses and coasts (Perlman 1980; Zvelebil 1986). At a more local level, resource availability can also vary greatly between years. Favourable resource points are often not fixed and regular - for example, one river may have a rich salmon run in a year when the next river

1960

1970

1980

Fig. 3.4. Intermediate variation in salmon catches: River Dee nets (stretches 1-3) against rods (stretches 4-8), 1953-83. a: total catch; b: net proportion. Source: Annual Reports.

has a poor one. One pool within a river may yield richly one year, while other normally more productive ones may not. Again, salmon catches from North Wales, from the rivers Dee and Clwyd (see Figure 3.1), are instructive. At the level of macro-variation, the overall figures for the last 30 years show major interannual fluctuations, and to some extent (e.g. the crash of 1968) the pattern of variation is similar in the two rivers (Figure 3.3a). In any given year, however, the catch may increase in one river, yet decline in the other, as in 1956, 1960, 1962, 1967, 1969, 1974, 1975, 1977, 1978 i.e. 30 percent of the years considered (Figure 3.3b). Intermediate level variation can also be examined. The salmon in each river are exploited both by nets in the estuaries and by rod fishermen higher up. Figure 3.4a shows the rod and net catches on the Dee, and Figure 3.5a the net catches in the estuaries of the Dee and Clwyd. Once again, in both cases there are several years in the last 30 where the two sets of catches vary independently of each other (Figures 3.4b and 3.5b).

43

Saving it for later

Micro-variation

Intermediate variation Fish (kg) 20,000 -

15,000 -

10,000 -

5,000 -

J

J

Salmon (kg)

\

Stretch 1

8000 - . Clwyd nets (salmon and sea trout) Clwyd nets (sea trout only, 1974-83)

Stretch 2 7000 -

Stretch 3

6000 -

5000 4000 -

3000 -

2000 1 t

1960

-

1000 -

1970

1960

1980

Dee net catch as % of total

1980

1970

% of total I\J

—

/"

60 -

70-

60-

50 -

50-

40 -

40-

30 -

/

** \

/ N»

//

/

""\ v

*\

\ \

/

20 -

1960

1970

1980

Fig. 3.5. Intermediate variation in salmon catches: River Dee nets (stretches 1-3) against River Clwyd nets, 1953-83. a: total catch; b: Dee proportion. Note: sea trout are insignificant in the Dee, but common in the Clwyd. The Clwyd net total includes both salmon and sea trout; separatefiguresfor sea trout are available from 1974-83 only. Source: Annual Reports.

Micro-level variation is shown in Figures 3.6 and 3.7. There is marked interannual variability within (Figure 3.6a) and between (Figure 3.6b) the three net groups in the Dee estuary. Figure 3.7a breaks the rod-caught total down into five short stretches of the river. Even at this level, there is marked interannual variation between the stretches (Figure 3.7b) in terms of (a) the absolute quantity of salmon (all the lines are markedly uneven, but some stretches are consistently more productive than others), and (b) the timing of good and bad years for each stretch (e.g. stretches 4, 5 and 8 peak in 1955, when 6 and 7 decline). It is important to bear in mind the short distances over which this variation is occurring (see Figure 3.1). Very local variation is also considerable. This may be examined by means of the records concerning the Llangollen Fechan substretch of the Dee, about one km long on the south bank, and falling within stretch 6 (Figure 3.1). This includes the pool of Dwy

A/ \ K

< >

V"

V

V

10 " • - . . • • • •

1960

1980

Fig. 3.6. Micro-variation in salmon catches: comparison between the Dee nets (stretches 1-3), 1953-83. a: total catch; b: proportions. See Figure 3.1 for locations. Source: Annual Reports.

Craig, 'the most famous pool on the whole of the Dee' (HughesParry 1949:173). Figures from 1961 to 1983 (Figure 3.8) show that the contribution of Llangollen Fechan to the stretch 6 total varies considerably from year to year. There is thus variability at all levels. Moreover, as the geographical scale decreases, the coefficient of variability increases, i.e. the more local the level of consideration, the more marked is the interannual variability (Table 3.1). The Dee and Clwyd salmon thus make clear that interannual variation viewed over a large area is itself the product of more intense variations on a local scale. Because these local variations do not harmonise with each other, the regional curve is smoother than the local ones. In the absence of detailed figures, it is not clear how far this can be extrapolated to other resources, but most migratory resource species

44

Peter Rowley-Conwy and Marek Zvelebil

Very local variation

Micro-variation Salmon (kg)

Stretch 4 Stretch 5 Stretch 6 *-+-

3000 -

Stretch 7 Stretch 8

2000

1000 -

% of total

^ \V

40-

\ A

/

"X

/

V

V \

20N

10 -

,

•••

/

1975

1980

Llangollen Fechan as % of Stretch 6 total 30 - ,

50-

30-

1980

1980

1970

1960

1975

*•

/

Iy•4 .."\ /*

\

A /

' *\

4

\ I ,.x V

*'

N"

\

V

V

\ /

* - * _ ,

1960

\ /X 1970

/ \

< I

A

/ v v/.

'•••.

V

••

V —

1980

Fig. 3.7. Micro-variation in salmon catches: comparison between the upper Dee stretches (4-8), 1953-83. a: total catch; b: proportions. See Figure 3.1 for locations. The totals for stretches 7 and 8 were lumped in 1957 and 1960-2, and cannot therefore be shown. The total is included in the overall total from which the percentages in Figure 3.7b are calculated. Source: Annual Reports.

will probably show a similar pattern. This has very significant implications for hunter-gatherer groups exploiting such species. Longer-term fluctuations Our third scale of fluctuation is the long-term - when an environmental change presents hunter-gatherers with situations they have not faced before. On the Dee between 1928 and 1983, the interannual fluctuations discussed earlier are superimposed on longer-term cycles of salmon availability (Figure 3.9). There has been an overall decline from around 1940 to the present. The causes of this are beyond the scope of this discussion but, in hunter-gatherer terms, it amounts to a collapse in this resource. Long-term changes are also evident at more local levels of consideration. For example, the productivity of the Dee increased greatly up to 1968, while the Clwyd changed less (Figure 3.3). At an even more local level, the nets of stretch 3 on the Dee improved

Fig. 3.8. Very local variation in salmon catches: comparison between the Llangollen Fechan substretch and stretch 6, 1961-83. a: total catch; b: Llangollen Fechan proportion. See Figure 3.1 for location. Source: Annual Reports and unpublished records of Capt. T. Hughes-Parry.

their catches considerably compared to stretches 1 and 2 (Figure 3.6b); similarly, the rods of stretches 4 and 7 (and recently also 6) increased their proportion of the total rod catch (Figure 3.7b). The implications of such local changes will be examined below. Responses to risk

The variability characteristic of the high latitudes introduces elements of risk and uncertainty, as dietary requirements have to be continuously met from resources that are unevenly distributed in time and space. Responses to this risk may be considered in terms of the four basic options distinguished in the introduction to this volume by Halstead and O'Shea, namely mobility to exploit resources elsewhere; diversification of the resource base to include other resources; storage to extend the period of utility of a resource; and exchange, comprising an array of possible social mechanisms for evening out variation. The potential of each category will be considered in relation to seasonal, interannual and long-term variation, before we proceed to a discussion of the reasons for the selection of the storage option. Responses to seasonal and interannual variation are closely linked. Diversification is widely practised by hunter-gatherers and provides a cushion against variability in individual resources. In

Saving it for later

45

Table 3.1. Various aspects of salmon catches on the Dee and Clwyd

predict how many deer will be available in the coming winter, so the amount of salmon it is necessary to store cannot be precisely determined. Secondly, the salmon runs are themselves variable, so the quantity available for storage fluctuates. In this context it is clearly insufficient to store enough to get through an average expected resource trough. Hunter-gatherers must clearly aim to overcome not the average but the worst likely case. This sets a definite target for the amount of food to be stored. As the resources to be stored also vary, it may on occasion be difficult to achieve this target; but the survival of hunter-gatherers who practise storage does testify to the fact that the set-up usually works. Occasionally, of course, a resource for storage may be so unproductive, and alternative resources in the subsequent resource trough so scarce, that things go wrong. In the case once again of the Ainu, 200 deaths in the year 1725 did indeed result from the failure of the stored salmon/winter deer hunting combination that year (Watanabe 1972). Most resource troughs are, however, not as bad as the worst foreseeable case, and most seasonal resource concentrations will produce sufficient food for storage. The normal aim, therefore, results in an excess of stored food in the subsequent resource trough - in other words, the basic operation of the system creates a surplus in most years. This is of the greatest importance. Surplus is a feature more usually associated with agricultural societies, and its role in hunter-gatherer society is rarely stressed. Arguably, it is this existence of surplus which makes exchange a particularly important response to interannual variability in storage-based economies (see p. 50). Long-term fluctuations are quite distinct from seasonal and interannual variability, because, as we will argue, hunter-gatherer groups respond to them in very different ways. Clearly, storage is no answer to long-term fluctuations; flexibility rather than forward planning is the answer. Hunter-gatherers practise various combinations of responses to long-term changes. A good documented example concerns responses to fluctuations in caribou availability. Caribou herds oscillate wildly in numbers (for a well-documented example, see Meldgaard 1986); two examples are plotted in Figure 3.10, and the different human responses in the two cases are most illuminating. As Burch (1972) points out, in times of a long-term decline in prey numbers, predators must either (a) decrease correspondingly, through starvation and/or emigration, or (b) intensify previously less important resources or begin to exploit new ones (i.e. diversify), or (c) a combination of these. Burch describes the response of the Nuataqmiut of the Upper Noatak Basin to the decline of the Western Brooks caribou herd:

mean

standard deviation

71.9

±6.1

8.5

B Intermediate variation 1 Dee nets as % of Dee total 65.6 66.9 2 Dee nets as % of Dee and Clwyd net total

±7.8 ±7.9

11.9 11.8

22.5 35.9 41.4

±8.6 ±8.0 ±12.6

38.2 22.3 30.4

25.5 12.9 26.2 25.6 10.4

±8.2 ±4.9 ±8.3 ±8.5 ±5.0

32.2 38.0 31.7 33.2 48.0

13.0

±6.2

47.7

A Macro-variation 1 Dee catch as % of total Dee and Clwyd catch

C Micro-variation 1 nets stretch 1 as % of net total stretch 2 as % of net total stretch 3 as % of net total 2 rods stretch 4 as % of rod total stretch 5 as % of rod total stretch 6 as % of rod total stretch 7 as % of rod total stretch 8 as % of rod total D Very local variation Llangollen Fechan as % of stretch 6 total

coefficient of variability (SD as % of mean)

The figures are based on the period 1953-83, except for rod stretches 7 and 8 (figures for 1957 and 1960-2 were lumped, so are not included except in the overall total from which the averages are calculated), and Llangollen Fechan (1961-83 only). the higher latitudes, however, the effects of bad years tend to pervade much of the ecosystem, because of its relative simplicity. The potential of diversification is greatly enhanced when advantage is taken of spatial variability in resources, and mobility is well documented among high-latitude hunter-gatherers as a response to both seasonal (Binford 1980; and see below) and interannual variation (Binford 1983:119-20). Indeed, much of the movement entailed in the seasonal round of high-latitude hunter-gatherers may be intended primarily to monitor the availability of resources in parts of the landscape used only in bad years. Interannual fluctuations also complicate (for example) storage in summer to overcome winter scarcity. They do so in two connected ways. Firstly, it is impossible to predict how deep a winter trough will be, but the amount to be stored has to be decided in the previous summer and autumn. Secondly, the summers themselves are variable in the quantity of resources available for storage. This is shown by the case of the Ainu of northern Japan, who survived the winter mainly on deer hunted during the winter and salmon stored from the previous summer and autumn (Watanabe 1972). Firstly, it is not possible in the previous summer to

During the population peak of the mid-nineteenth century, the Nuataqmiut lived in a world of more or less continuous abundance ... By 1890, extreme annual fluctuations in caribou abundance had become part of the ordinary circumstances of life of these people, something they had not known 30 years previously . . . they realized that they would have to leave the country or they would die out. Unlike

Peter Rowley-Conwy and Marek Zvelebil

46

River Dee (kg) 50,000 - .

40,000 -

30,000 -

Llangollen Fechan (kg) I - 2000

20,000 -

10,000 -

1920 1924

- 1000

1928

1932 1936

1940

1944

1948

Dee catch, 5yr running average Dee catch, annual total

1952

1956

1960

1964

1968 1972

1976

1980 1984

Llangollen Fechan total

Fig. 3.9. Long-term variation in salmon catches on the Dee: macro-level, 1928-83 (thin line: total catch; thick line: 5-year running average). Very local level, Llangollen Fechan 1920-84 (no records available 1953-60). Note that the Llangollen Fechan line is vertically exaggerated 10 times relative to the Dee total. Source: Annual Reports and unpublished records of Capt. T. Hughes-Parry.

most of the other groups in the area, they did not have sufficient alternative resources in their home area even to consider remaining there. Unfortunately, too, all of the neighbouring groups were in similar or worse positions than they were as far as caribou were concerned, so they could no longer be relied on for help in time of need... the majority of Nuataqmiut decided to move about 600 km to the northeast [where caribou were abundant]. Other families, a much smaller number, decided instead to move to the coast... where they had to switch from an emphasis on caribou to one on sea mammals . . . By 1900, not a single family was wintering in the Upper Noatak Basin (Burch 1972:357-8 [added emphases]) The main responses to interannual fluctuations (cf the introduction by Halstead and O'Shea) were clearly not effective regarding longer-term changes: storage was not practical given the time-scale involved; diversification was not possible because of the absence of alternative resources; and exchange was not possible because other groups were as badly situated. Only a new dimension of mobility (emigration) prevented starvation. This is quite distinct from the West Greenland response described by Meldgaard (1983). In this case the hunters lived in coastal villages, moving to communal summer hunting camps in the interior; cari-

bou provided both an important food source and many raw materials. In periods when only a few caribou are to be found in the inland area, people have to find substitutes for caribou skin, sinew, antler, fat, tallow and meat. This is possible, but most of the substitutes, such as dog skin, whale sinew and walrus tusk, do not possess the same good qualities as the raw materials obtained from caribou . . . No substitutes can be found in the inland region as caribou is the only resource of any importance in this area. During a caribou minimum the part-time 'inland hunters' must therefore become full-time 'sea hunters' (Meldgaard 1983:267 [added emphasis]). In this case diversification was possible, because of the availability of alternative (albeit inferior) sources of all the products formerly supplied by the caribou. This represents a quantitative response within the existing system, rather than a qualitative transformation of the whole system as in the Nuataqmiut case. Long-term fluctuations are directly visible in the archaeological record, while at the same time falling outside the reach of the social anthropologist. Meldgaard's caribou study, for example, was directly linked to archaeological study of the inland cari-

47

Saving it for later 350,000

300,000 -

250,000 -

200,000 -

150,000 -

100,000 -

50,000 -

1650

1700

1750

1800

1850

1900

1950 Year

West Greenland Western Brooks

Fig. 3.10. Long-term variation in numbers of caribou. Western Brooks herd from the estimate by Burch (1972, Figure 6); West Greenland herd from Meldgaard (1983, Figure 2). Dashed line: from estimates and censuses; dotted line: from historical sources. The maxima and minima before 1950 are assumed to be of the same magnitude as the most recently documented ones (Meldgaard op. cit.).

bou camps. These showed periodic abandonments, coinciding with caribou minima (Gronnow, Meldgaard and Nielsen 1983). Selection of the storage option Why storage? The foregoing discussion has emphasised the general usefulness of storage as a risk-buffering method in the context of seasonal variation. Our purpose now is to attempt to predict the specific situations when storage will and will not be employed as a major risk-buffering method. Not all combinations of the four basic risk-buffering methods are equally easy to achieve. Most importantly, mobility and storage are to a large extent mutually incompatible. The main problem is access to the stored reserves. The difficulties involved in transporting major stored reserves to a new base camp (except by water in favourable circumstances) would normally be so great as to make dependence on such reserves unviable.

The resources must of course be suitable for storage. Nevertheless, such resources do not have to be stored; there are many examples of apparently suitable high-latitude resources being exploited only for immediate subsistence needs, by mobile groups. We will now propose a model accounting for the decision between storage and mobility. Within the general precondition of the suitability of resources, we believe that the spatial and temporal characteristics of the local environment are the crucial factors. We may envisage two equally accessible and predictable resources (termed resources 1 and 2), exploited successively in the annual cycle of a human group, and we assume for the sake of the argument that the largescale transport of stores is impractical. If the two resources appear at points too far apart to be exploited from a single location, a choice must be made: either to store resource 1 and stay put, ignoring resource 2; or to move to resource 2 and not to store resource 1. The mobility option is more

Peter Rowley-Conwy and Marek Zvelebil

48

likely for a number of reasons. Firstly, movement to resource 2 means that less reliance is placed on resource 1, so less has to be taken than if it was to be stored. Interannual fluctuations in resource 1 will thus reach crisis level less frequently. Secondly, there is always the risk that stored reserves will spoil or be expropriated - something that cannot happen to a resource not yet harvested. Thirdly, in the event of crisis the diversification response will be easier to sustain over the shorter settlement periods associated with the mobile strategy. In this spatial context, mobility is thus a lower-risk strategy from a number of points of view, although the risk of a major interannual fluctuation in either resource remains a danger. If, on the other hand, resources 1 and 2 appear close enough together to be exploited from a single centre, then risk takes on a very different aspect. Storage of resource 1, for consumption during the procurement of resource 2, diminishes risk by providing two major food sources during the time resource 2 is available. There is thus less risk of crisis, because the human group is spreading the risk over two resources by extending the availability of one of them through storage. Interannual fluctuations in resource 2 are thus less dangerous. Given this second spatial context, storage is clearly the lower-risk option. This, it will be remembered, is the strategy employed by the Ainu, using the combination of salmon and deer, as described above (p. 45).

Some examples These two alternatives may fit into an annual cycle in a variety of ways. Some of these are modelled in Figure 3.11. Model A takes six hypothetical resources, each assumed to be available for two months of the year, appearing so far apart that the transport of stored reserves from one resource point to the next is impractical. The group thus moves around the six resources, exploiting each in turn, for immediate subsistence needs only, with minimal storage. These are what Binford (1980:17) terms 'serial specialists', specialising on a series of sequentially available, but widely separated, resources. The Netsilik (Balikci 1968, 1970) approximate to this strategy, frequently occupying almost as many camps as there are months in the year. They do practise some storage, but this does not appear to be a major activity; it mainly comprises caches left behind near particular settlements, and revisited when the settlement is again within striking distance, or short-term caches for use within the period of occupation of one camp. The diversification strategy appears to be more important - Balikci stresses 'the endless search for alternatives, the practices of substitution and ad hoc inventiveness . . . Each seasonal hunt knows some alternative' (1968:80). In model B of Figure 3.11, three of the hypothetical resources are close enough together to be exploited from one base camp, and resources for storage transported back to the base camp from the satellite locations. In this model, stored reserves of resource 4 can help to buffer risk throughout the occupation of the base camp (the availability periods of resources 4, 5 and 6); and those of resource 5 for the periods of availability of resources 5 and 6. Risks inherent in the interannual variability of resources 5 and 6

2

• •6 •a

V

•

5 Resource point Residential mobility Transport of stores

Fig. 3.11. Model of settlement variability. Residential mobility covers distances too great to permit the transport of stored reserves. Logistic mobility is over short distances, so resources may be transported (see text).

are thus diminished. This category corresponds to Binford's (1980:13) semi-nomadic hunter-gatherers. In model C, the situation prevailing for part of the year in model B is extended to cover the full year, with two base camps being modelled. The discussion of the storage of resources 4, 5 and 6 by type B hunter-gatherers is relevant to the whole of the annual cycle of type C groups. Variations within this category form the classic iogistic' or semi-sedentary hunter-gatherers of Binford, some of the most well-known being the Nunamiut, whom Binford studied in this connection. Model D shows the position when the two base camps in model C are combined into one, because of the availability of resources within striking distance for enough of the year to make permanent settlement possible. One advantage of this is that there are no breaks in the risk-buffering capacity of storage of the kind there are in model C - storage does not help type C huntergatherers in the early availability periods of resources 1 or 4, because they cannot rely on the storage of the previous resource in the event of a crisis because of the transport problem. At these times, type C groups are as vulnerable to interannual fluctuations as type B groups during resource periods 1 to 4, or type A groups

Saving it for later

49

throughout the year. Type D groups to some extent bypass this problem, although type D adaptations will of course never be as regular as the model implies; the role of stored resources may dwindle to nothing during the more productive parts of the year, while being of overwhelming importance when few resources are immediately available. Type D groups correspond to the sedentary groups discussed by Rowley-Conwy (1983), and to Binford's type 4 (compact and relatively permanent) groups, although some of Binford's type 3 (semi-sedentary) groups are also included here, namely those 'who occupy more or less permanently a single settlement from which a substantial proportion of the population departs seasonally to occupy shifting camps' (1980:13). Specialpurpose camps are a feature of even the most sedentary huntergatherers; the important thing for the present model is the degree of all-year risk-buffering implied by the existence of the single base camp of the type D groups. An ethnographic example of a type D group is the North Alaskan Eskimo, a coastal group living in large permanent villages, relying very heavily on storage for much of the year (Spencer 1959). These four examples have taken the question of mobility or storage outlined earlier in this section and have shown how the answer we propose, based on the spacing at which successive resources appear, influences settlement under a variety of circumstances. The model clearly views storage as a solution to an environmental problem, namely the seasonal and interannual variability of resources. It must be admitted that others have not taken this view. Some explicitly reject the notion that the environment plays much part in the decision whether or not to store, and maintain that the storage and non-storage options are both equally feasible in almost any environment (Woodburn 1980:101, 111). That is not the view taken here. Among many high-latitude foragers, the seasonal scarcity of resources means that the storage of resources available in concentrations at other times of the year is quite definitely a prerequisite for survival. There is, of course, a choice involved. Devotees of Joseph Heller's Catch 22 will remember the case of Milo Minderbinder, the black-market king who monopolised food supplies to the United States Air Force mess halls throughout the Mediterranean theatre of operations during the Second World War. Prices became intolerable; the highaltitude fliers were faced with the same choice as high-latitude foragers, since 'their alternative - there was an alternative, of course, since Milo detested coercion and was a vocal champion of freedom of choice - was to starve' (Heller 1964:391).

availability, as well as by the requirement to stay put to guard the stored produce. At the most general level, therefore, it may be expected that the longer the period of the year a hunter-gatherer site is occupied in the high latitudes, the more likely this is to involve storage. There will of course be many individual variations on this, and each site and its immediate environment will differ and need individual attention. Nevertheless, this is a useful starting point from which to examine the question with regard to the archaeological record.

Some corollaries of storage In this section we examine some of the other aspects of hunter-gatherer societies associated with storage. We consider three major categories. Period of settlement occupation As will have emerged from the foregoing discussion, storage plays a decisive role in the length of time for which a huntergatherer settlement is occupied. The occupation period is lengthened by extending the use of a resource beyond its period of

Technological factors Technological strategies for coping with resource variations by storage involve firstly the development of capture technology designed to maximise the harvest of a resource that is available for a limited period of time, and secondly the development of processing, preservation and storage facilities designed to extend the period of consumption of the harvest. Technology forms a necessary accompaniment to storage: without effective technology capable of producing a surplus, there would be nothing to store; and without effective preservation and storage techniques, large catches of food become useless in a matter of days. The ethnographic literature abounds with illustrations of the failure of one or other aspect of this technological system (for the northern circumpolar area, see Eidlitz 1969). The harvest can be maximised by increasing the rate of capture. With the appropriate technology used on suitable resources, large numbers of the prey species can be taken in a short time. Torrence (1983) has developed the idea of'time-stressed' environments in the higher latitudes, where a great deal of work is concentrated into a short period, and where the consequences of the failure to capture mobile resources are particularly serious. A response in these cases is the development of specialised, curated technology, and time-saving capture and processing facilities, such as weirs, traps, snares and processing gear. Bamforth (1986) argues that tool curation increases technological efficiency. Bleed (1986) makes a distinction between maintainable and reliable tool kits; reliable equipment is designed to minimise failure, and is likely to be applied to discontinuous but predictable resources. Such tool kits would be appropriate for the time-stressed exploitation of seasonal resources in higher latitudes (Zvelebil 1986). In effect, the development of such technology is a form of storage: it stores labour, in that it saves time during the short periods of harvest, and in doing so it reduces the rate of failure and increases productivity. Most specialised items need to be made or curated in advance, and so redeploy labour from outside the harvest period to within it. Curated capture and processing technology may be divided into items which are easily transportable (much processing equipment, such as items made of ground stone, or complex individual hunting equipment), and those which are not (large or fixed capture installations such as traps, weirs, large nets, etc.). The former might be used at a variety of sites, and could thus be made and curated elsewhere. The latter, however, require construction and maintenance at the point of use, in advance of the harvest. This factor thus links with the discussion on sedentism: not only does

Peter Rowley-Conwy andMarek Zvelebil

50

storage extend the occupation period of a site after the period of the harvest, but the associated technology requires some human presence before it. Longer occupation of sites diminishes the utility of the diversification response: the longer the secondary resources have to be used if the main resource fails, the less likely they are to ensure the human group's survival through the crisis. As the length of occupation lengthens towards sedentism, however, another aspect of technology may counterbalance the problem. If a site exploits and stores several successive resources, so that there are several peaks and slack periods in work effort, then there may be time to develop mass-capture technology to use on the secondary resources as well as the primary ones (Rowley-Conwy 1983). An example of this is the 30 m long nets used by the Bering Strait Eskimo to catch ptarmigan: this is a clear case of diversification, as this otherwise minor practice was intensified if the salmon runs were poor (Nelson 1899). The more sedentary a group is, the easier this option becomes, because of (a) the availability of more slack periods between resources, and (b) the possibility of storing the technology on the settlement, rather than caching it at some point on the migratory cycle (Rowley-Conwy 1983).

notion of social storage rests on the use of prestige items, usually craft goods, as tokens of value which can be exchanged for food in times of need (O'Shea 1981). Possession of such tokens, linked to social prestige, would therefore serve as insurance against localised shortages of basic foods. At a more general level, prestige objects can also serve as gifts employed to maintain the network of social relationships and obligations which also operate to redistribute food in situations of localised resource scarcity. It must be emphasised that neither of these can cope with widespread, regional failures in food supply, unless more complex social mechanisms also exist to organise surplus production, exchange and redistribution on a larger scale. It could in fact be argued that this is where hunter-gatherers are most vulnerable: in the absence of such complex levels of social integration, the economy is not able to endure a resource failure on a regional scale, so that a shift to an alternative strategy is necessary. Within the context of localised resource variability, therefore, social storage will be an important mechanism for reducing risk. Its importance will be greatly restricted, however, if storage is not widespread. If storage is not practised, and resources are for immediate consumption only, then the only place that food will be available in a crisis is at another point in the natural environment. Unless social storage can be used to gain immediate access to a more productive resource point, therefore, the mechanism will be of limited use to mobile groups. Among groups practising storage, on the other hand, the mechanism may play a much greater role. In the context of unequal quantities of dwindling stores being available to different groups, the mechanism comes into its own, with a wide variety of this and other redistributive mechanisms being available to many groups. Other social mechanisms will be more relevant to mobile groups, but are likely to operate more on a day-to-day than a season-to-season scale because of the absence of stored resources. The group with sufficient stored reserves in any given year could therefore gain prestige via the tokens exchanged in the social storage system, or by hosting communal feasts, etc. Such mechanisms would thus help to maintain equilibrium in the food supply but they would not maintain equilibrium in social relations in the long run. This is because, despite interannual fluctuations, some localities are on average basically more productive than others. Some groups would thus more often be suppliers of food and accumulators of prestige than the reverse. This potential can clearly be seen in the North Welsh salmonfigures.The Dee estuary is, despite relative fluctuations, generally more productive than the upper stretches of the river (Figure 3.4) (although different technologies are now used in the two areas, so this need not always have been the case). At the micro-level, during thefirstpart of the period considered, stretch 6 was more often more productive than the other stretches (Figure 3.7). With a resource such as the Atlantic salmon, available in North Wales over several months, other factors are also relevant. Figure 3.2 shows that monthly catches 'roll' up theriverduring

Socialfactors It was mentioned above (p. 45) that hunter-gatherers must aim to get through not just an average resource trough, but the worst likely one. A surplus of stored food is therefore often available, because troughs are frequently not so serious. We believe that the availability of this surplus is what makes possible some of the developments among more sedentary hunter-gatherers practising storage, which parallel developments in farming societies. After all, why does it matter whether the surplus is the result of fishing or farming? The consideration of salmon in the North Welshrivers(Figures 3.3-3.8) illustrates a number of important points in this regard. At all the levels considered, there was spatial as well as interannual variation. At the intermediate scale, Figure 3.4b shows the major interannual differences between the productivity of the estuary and the upper stretches of the River Dee, while Figure 3.5b shows the same for the two estuaries. Figures 3.6b and 3.7b make the same point at the micro-level, and Figure 3.8b at the very local level. Fluctuations are most marked at the most local level of consideration (Table 3.1). If each of the stretches on the Dee is (for the sake of argument) taken to be the exploitation area of a local group using storage, then it is easy to envisage a situation in which (as a result of interannual fluctuations) one group has a sufficiency of stored reserves, while another has been unable to store enough and faces the possibility of running out. Within one community, some families may find themselves better placed than others, either because they have worked harder, or because they have preferential access to more productive resource points. This is the sort of situation in which social strategies would be expected to play a vital role. We will here confine ourselves to one particular mechanism, which we believe is of direct relevance to high-latitude hunter-gatherers. This is social storage. The

Saving it for later

51

the summer. Despite their high productivity, the lower stretches would not be in such a good position to capitalise on this. Storage in the spring would be less important, as the summer would be less likely to see resource crises - some salmon are available everywhere, for immediate subsistence needs. Stretches 7 and 8, on the other hand, have maximum salmon availability in the autumn, just before salmon availability declines, and as the winter resource trough approaches. These upper stretches would therefore be in a much better position regarding salmon storage, and would therefore be more consistently able to convert this into prestige by supplying less well-endowed groups. Under these circumstances, it is not difficult to envisage the prestige and status of well-endowed groups, families or individuals becoming more permanent. All the required elements are thus present to permit the development of a degree of status-linked ranking. This is confirmed by the anthropological record. The hereditary, ranked aristocracies of the Northwest Coast Indians are the most developed and well-known examples of their kind among foragers. Population and local group rank among the Southern Kwakiutl are directly linked to salmon availability and reliability (Donald and Mitchell 1975); and it was frequently the case that high-status individuals were those with preferential access to particular well-endowed resource points - 'men of a local group whose chief owned an inlet suitable for duck netting engaged in this pursuit; the retainer of an owner of halibut and sea mammal hunting grounds repaired to these places as early as the weather permitted' (Drucker 1951:43, added emphases). Longer-term fluctuations mean that such acquired status need not be eternal. In Figure 3.6, the nets of stretch 2 are replaced as the most productive by those of stretch 3 over the period of a generation; if each were the area of a local group, we would have to envisage an increased 'flow' of prestige from the stretch 2 group to the stretch 3 group, in particular the leading members of the groups. In Figure 3.7, the rods of stretch 7 have increased their proportion of the catch over a similar period, and in a huntergatherer situation would thus be expected to increase their prestige and status during this period. The reasons for such variations may be many - in the case of the upper Dee (Figure 3.7), better regulation of the river, and the destruction of a weir downstream in 1966, have permitted the salmon to ascend the river more quickly, thus reducing their accessibility on the lower stretches. This mimics the natural hydrographic changes occurring in all river systems, and shows the part that such changes can play in huntergatherer subsistence. There are indications of still longer-term changes - at the end of the last century, for example, the summer and autumn runs of salmon were much stronger, and the early spring run very much weaker, than in recent years (T. HughesParry, personal communication). This section has linked a number of consequences to the fact that hunter-gatherers must plan for the worst likely case, not the average case. We have argued (a) that coping with interannual fluctuations in this way is a prerequisite for survival in many higher-latitude environments, and (b) that a number of technological and social factors follow logically from this.

Storage by prehistoric European hunter-gatherers Initial considerations In the previous sections we have put forward a model for the appearance of storage among hunter-gatherers, and pointed to some of the corollaries of such a strategy. This section will examine the later Palaeolithic and Mesolithic of Europe for evidence of storage and its corollaries. Clearly we cannot expect to find much trace of the stored reserves themselves. Such reserves are, after all, accumulated in order that they may be used up later, and so such reserves will not survive under the normal operation of the system. Abandoned reserves would hardly ever survive: the most we could hope for would be the survival of animal bones from stored reserves, but in this case either the removal of the decayed stores by people in order to reuse the facility, or their destruction by wild carnivores and carrion eaters, would be likely to destroy any recognisable pattern. Our best hope would seem to be the charring of stores during the destruction of a building by fire. Even here, there would be problems of interpretation, however: Are charred hazel nuts in a pit evidence that nuts were stored in the pit? Or merely that the last filling of the pit was burnt refuse from elsewhere on the site? In the absence of very lucky finds, therefore, our inferences must be indirect. As a starting point for a survey of the European archaeological record, we must consider which areas of Europe are most likely to have seen the development of a storage economy. We have stressed two essential aspects: firstly, that seasonal resources must be available; and secondly, that the local resource make-up within areas of seasonal resource availability is the crucial factor. The first factor by itself is insufficient (see also the effective criticism by Schalk 1977). In general, Europe north of 40° latitude can be characterised as quite strongly seasonal (Zvelebil 1986). Coastlines and rivers with migratory (particularly anadromous and catadromous) fish form the basic areas where local resource make-ups will sometimes favour the storage option and its corollaries (Rowley-Conwy 1983; Zvelebil 1986). This is because these are among the most productive and concentrated resources that would have been available, and the ethnographic records of both Europe and North America testify to their role in storage systems. Zvelebil (1986) has distinguished three areas of Europe where local factors may often combine to favour the storage option: (a) northern Europe, (b) the Atlantic and Baltic coasts and their rivers, and (c) the Black Sea basin and its rivers (cf. also Clarke 1976). These are the areas we will examine. We will seek evidence for the various corollaries of storage mentioned above, namely (a) specialisation on various categories of seasonally available resources in systems with increasingly sedentary base camps, with possible evidence for storage facilities, and (b) mass-capture technology. Evidence of resource specialisation Anadromous and catadromousfish.The two major salmonids of Western Europe are the Atlantic salmon (Salmo salar) and the smaller sea trout (S. trutta). Their remains are found in

Peter Rowley-Conwy andMarek Zvelebil

52

many of the classic Upper Palaeolithic sites. The occurrence of the resource is, as stressed above, no demonstration that it formed part of a more complex storage system with increased sedentism. Recent discussions disagree about the importance of salmon (cf. Jochim 1983; Mellars 1985), and our emphasis of the importance of the local resource spectrum would lead us to expect a range of combinations, varying through time and space. Archaeological evidence either way is scarce from the Upper Palaeolithic; fish bones have rarely been collected and analysed systematically enough to provide clear indications. The most telling statement comes from one of the earliest pieces of work on the subject:

find we may expect. Also on the Bann River is the site of Newferry 1, dated to the later Mesolithic (Woodman 1978). A large ash layer, 12 cm deep and at least 40 m2 in extent, may be a location where salmon were smoked for storage (Movius 1937). Salmon are a less important resource in the Baltic, and this is reflected in the archaeological record. The only definite mesolithic finds of salmon in Denmark, for example, are two salmonid vertebrae from Meilgaard (Rosenlund 1976) and an unspecified number at Vedbaek (Aaris-S0rensen 1980). Other fish were much more important (see below). The sturgeons (Acipenseridae) are the other major anadromous fish family of Europe. The Black, Azov and Caspian seas, together with the vast area pervaded by the rivers draining into these waters, should be regarded as the original and true home of the sturgeon' (Rohan-Csermak 1963:5), although they are also known from the North Sea and Baltic. The common sturgeon is present in all coastal European waters, and penetrates the larger rivers of western Europe although in these areas it is now very rare. The Black and Caspian Seas and the rivers draining into them, however, support six species of sturgeon (Rohan-Csermak 1963). The most important is the beluga or great sturgeon, Huso huso, reputed to be the largest freshwater fish in the world. Specimens 9 m in length and weighing 1400 kg have been caught on the middle Danube in Hungary, while an eighteenth-century record mentions a specimen caught in Russia that weighed 2000 kg. Sturgeon occur in greatest concentrations along the lower and middle reaches of large rivers; in the Danube, for instance, the most intensive fishing was carried on up to the confluence of the Danube and Morava rivers near Bratislava. Sturgeon migrations take place twice a year, in spring and autumn. Similar methods are employed for their capture as for salmon. Modern productivity studies and historical evidence point to sturgeon as an important resource, comparable to the salmon in other regions (Rohan-Csermak 1963). Unfortunately for archaeologists, the sturgeon skeleton does not preserve well. However, the bony scutes that run the length of the body of these fish are sometimes encountered. They have been found at Soroki I and II on the River Dniester (Markevitch 1974:59, 74). Although this falls far short of demonstrating the existence of an intensive fishery, it is significant that these sites are regarded as occupied all year round (Tringham 1971;478). The same is true of the Iron Gates region of the Danube: beluga {Huso huso), stellate sturgeon (Acipenser stellatus) and Russian sturgeon {A. gueldenstaedti) have all been recovered from the site of Padina (Brinkhuizen 1986). The site contains late mesolithic and early neolithic levels, although even in the latter domestic animals were of low importance. Fishing was very important, and the site appears to have been permanently occupied (Clason 1980). That sturgeon were one resource contributing to the permanent occupation of Soroki I and II, Padina and possibly other sites in the Iron Gates gorge seems likely. Evidence for sturgeon elsewhere is extremely sparse; Benecke (1986) quotes much medieval evidence from the southern Baltic, and believes that the present rarity of the sturgeon in this area is due to over-fishing during the Middle Ages. This raises the possibility that sturgeon could have

It is an interesting fact that among the numerous salmon remains from the caves, which we have examined, we have not met with an entire skeleton, having seen only portions of the vertebral column, as if only the edible portions were taken home to the caves. The bones of the head of the salmon, had they been there, would have been as well preserved as those of the small cyprinoids [chub, bream] which we find in the same deposits. These cyprinoids, on the contrary, which constituted what we may call the every-day fishing of the Aborigines, are recognised in all parts of their skeleton. They were evidently caught near the abode, and furnished fresh food; whilst the salmon went to form a food reserve (Sauvage 1875:223) Analysis of human bone from French sites, however, suggests only a minor role for salmon (Hayden, Chisholm and Schwarcz 1987). Remarkably little evidence of salmonids is available from the Mesolithic. It is sometimes said that salmon bones are so soft that they do not preserve well archaeologically, but Casteel (1976) has contested this. Certainly, salmon bones do survive at least sometimes, but little information is available concerning relative hardness of bones of the various fish species. Processing methods may pulverise salmon bones into fragments so small that they will be missed by all but the finest sieving (Casteel 1976). More important regarding the European.Mesolithic may be the fact that the coasts of Norway and the British Isles (where salmon must have been exploited, whether or not they were stored) are areas where animal bone of all kinds is often not preserved on archaeological sites. One good example is available, however, namely the site of Mount Sandel, on the Bann River in Ireland (Woodman 1985a). Many bones of both salmonid species were recovered, all in a charred state because bone is otherwise not preserved (van Wijngaarden-Bakker 1985). A number of factors favour salmon and sea trout storage. The site is only 800 m from a point on the river where the fish could easily be procured in large numbers, and their bones are common. Of great significance is the fact that the site was occupied for longer than just the period of salmon availability (van Wijngaarden-Bakker 1985; Woodman 1985b); in this context, storage would have been a definite risk-buffering mechanism (see above). Large pits on the site could have been used for storage, and some posthole arrangements are interpreted as possible fish drying or storage racks (Woodman 1985a). This remarkable excavation draws attention to the sorts of

Saving it for later

53

been more important in the Mesolithic than its present numbers suggest. The main catadromous fish in Europe is the eel (Anguilla anguilla). Little direct evidence relevant to the present argument is available, although small numbers of eel bones are common on many sites. It has been suggested that the late mesolithic sites of Dyrholm and Kolind in Denmark may have been eel procurement sites. Although no eel bones were found (they could have been removed if the fish were processed elsewhere), the season of occupation is correct, the location suitable, and the artefactual evidence suggestive (Rowley-Conwy 1983:122). Other fish. A variety of other marine and freshwater fish is seasonally available in European waters, and some of these appear to have been of major importance in the Mesolithic. Clark (1952: Figure 42) drew attention to the importance offish along the rich west coast of Norway; at some of these sites, particular fish species are so important as to indicate specialised procurement, for example cod at Viste (Brogger 1908). Cod appear to have been very important at other sites, such as Norsminde in Denmark (Rowley-Conwy in preparation a), and at a number of sites in Varangerfjord, northern Norway. Some structures at Nyelv Nedre Vest contain large numbers of cod bones; they indicate presence in the spring, while other seasonal evidence suggests all-year occupation. Storage in this markedly seasonal environment would probably have been essential. The site of Gropbakkeengen, on the other hand, while also yielding many cod bones, may have been occupied only in spring, and could thus have been a specialised extraction camp (Renouf 1981). Specialised exploitation of other species also took place. At Nyelv Nedre Vest, various structures contained many bones of saithe {Pollachius virens) and haddock (Melanogrammus aeglefinus), available in summer (Renouf 1981). Saithe were also very common in the middens on the Hebridean island of Oronsay (Mellars 1978). Whatever the details of the total economy to which these sites belong, the importance of saithe does raise the question of storage. Further south, large numbers of bones of mackerel (Scomber scombrus) were found at the late mesolithic site of Ordrup Naes in Denmark (Becker 1939), and of meagre (Argyrosomus regius) at Arapouca in Portugal (Morais Arnaud in press). Arapouca is dated to the sixth millennium be; the presence of only the headbones of Argyrosomus, despite the presence of numerous vertebrae of more cartilagenous fish, raises the possibility that the bodies of these fish were not eaten on the spot, but taken elsewhere for storage (Morais Arnaud in press). Other marine fish are also likely to have been stored at times in the European Mesolithic. Possible candidates include such species as herring (Clupea harengus), garfish (Belone belone) and flatfish such as plaice and flounder (Pleuronectes platessa and Platichthysflesus). All these have been recovered in quantity alongside other fish from the late mesolithic sites at Vedbaek in Denmark (one site contains some 40 000 fish bones per square metre of deposit), although further information is not yet available (AarisSorensen 1980). Evidence from Eastern Europe, particularly the Black Sea basin, suggests that non-migratory river fish may also achieve the

importance associated with stored resources. Some are certainly more available in some seasons than others, despite being nonmigratory, and suitable mass-capture technology can produce abundant returns. Fish remains, mostly of cyprinids (the carp family) and catfish, occur in large quantities at all the major sites of the Lepenski Vir culture on the Danube (Srejovic 1967). At Lepenski Vir itself, fish bones made up 57 percent of all faunal remains (Bokonyi 1970), 77 percent at Vlasac (Srejovic and Letica 1978), 80 percent at Padina III (Clason 1980), and they were also common at sites on the Rumanian bank of the river, such as Icoana and Veterani (Bolomey 1973). Whether sieving contributed to these figures is unclear. These extensive fish remains, plus such material items as harpoons and fishhooks, all point to the importance of fishing. Regarding storage of these fish, it is important to remember that sedentism has been demonstrated on the basis of faunal remains at Lepenski Vir and probably Vlasac (Bokonyi 1972), at Padina (Clason 1980), and at Icoana (Bolomey 1973). Some storage is thus probable; and the fish must be likely candidates. The possible role of sturgeon was discussed above. A similar picture emerges from other settlements on rivers in the Black Sea basin. Fishing was very important at settlements of the Bug-Dniester culture (Markevitch 1974); Tringham (1971:47-8) mentions hearths at Soroki II, surrounded by dense deposits of bones of cyprinids (carp and roach), fish which are available in summer. It would be interesting to know whether these could be places where fish was dried, or whether the pits interpreted as dwellings might in fact have been for storage. Fishing was also an important activity among hunter-gatherers living along the middle Dnieper and its tributaries (Dolukhanov 1979, 1986). Sites along the lower Don, such as Rakushetchnyi Yar, also yielded large numbers offish bones, predominantly of catfish and roach (Belanovskaya 1975). Marine mammals. The various species of seal and small whale form resources which are frequently concentrated and seasonal, and also highly productive in calorific terms. There is considerable evidence that they were of importance in the European Mesolithic. Several sites in the eastern Baltic region have such an overwhelming predominance of seal bones that they must have been highly specialised extraction camps. At Nakamae, on the island of Saaremaa, seal formed 92 percent of the animal bone; at Loona, Sventoji and Silinupe, 73 percent, 76 percent and 65 percent respectively (Paaver 1965). Further north in the Varangerfjord region, the site of Gressbakken contained many bones of juvenile seal (probably Phoca hispida), while many harp seal (P. groenlandicd) were found at Angsnes (Renouf 1981). Significantly, both these areas were characterised by permanent settlements: at Varanger, Nyelv Nedre Vest (Renouf 1981), and in the eastern Baltic at sites such as Sarnate in Latvia, where finds of ski runners and immature seals indicate occupation in winter, the presence of boats and nets point to warm-season occupation, and finds of water chestnuts indicate occupation in August (Vankina 1970; Dolukhanov 1979). In this context, the sealing sites are most logically interpreted as logistic procurement camps from which seal meat and blubber was transported back to

Peter Rowley-Conwy and Marek Zvelebil

54

the main base camp for storage. Possible evidence of storage has been found at Mynamaki in southern Finland, where seal bones and pottery were found as the sole contents of pits sealed with a layer of stones. A ritual explanation has been suggested (Vakevainen 1978), but the possibility remains that this represents a primary storage context. There is less evidence of sealing further south. 01by Lyng in Denmark contains some 13 percent of marine mammals, particularly harp seal (Phoca groenlandica) (Mohl 1970a), although specialised exploitation of rookeries of juvenile grey seal (Halichoerus grypus) on offshore islands such as Hesselo does not appear to have started until the Neolithic (M0I111970b). However, the flat elongated ceramic bowls common in Ertebolle culture contexts have long been interpreted as lamps in which blubber was burnt (Mathiassen 1935). This interpretation obviously demands blubber storage. The other Ertebolle pot type, the large pointed-based jar, suggests one way in which this blubber (and other commodities) could have been stored; analysis of food encrustation on a few sherds from Tybrind Vig has, however, revealed traces of fish and plant foods (Andersen and Malmros 1984). Analysis of a number of pots from Otterbole/Jetbole on the Aland islands revealed traces of seal blubber (Meinander 1954) and Welinder (1976) has suggested that seal blubber and meat were stored in similar vessels by Pitted Ware groups in Sweden. The casual exploitation of beached whales is likely to have taken place throughout the Mesolithic (Clark 1952). The organised driving of small whales has been suggested in Varangerfjord in northern Norway: the sites of Angsnes and Gressbakken contain many individuals of white-beaked dolphin {Lagenorhynchus albirostris) and are in areas which would have been suitable for driving groups of the animals ashore (Renouf 1981). Such drives inevitably mean the availability of a large quantity of meat and blubber; this activity must most probably have been part of a system involving storage. Birds. Various sites are known from the Mesolithic, which are best interpreted as special-purpose camps procuring birds of various types. Perhaps the best example is Aggersund in Denmark. This small site (c. 8 x 12 m) contained large numbers of bones of whooper swan (Cygnus cygnus). This species of swan is present in Denmark only in winter, and all other seasonal evidence from the midden fell within the period of its availability (M0I1I 1978). Another site in Denmark specialising on birds was Solager. Most of the bones from this site are late mesolithic, although a few come from the upper (neolithic) levels (Skaarup 1973). 560 bones come from wild land mammals - although the site was on an islet too small to support populations of red deer, roe deer and wild pig. A further 50 bones come from marine mammals; while no fewer than 2197 are from birds. 759 (35 percent of the bird bones) are from whooper swan, the same species as at Aggersund; 424 (19 percent) are of eider (Somateria mollissima) and 290 (13 percent) of velvet scoter (Melanitta fused) (Winge 1903). Permanent settlements have been suggested for the Danish Mesolithic of this period (Rowley-Conwy 1983); if so, these sites would fit into the system as specialised procurement sites from which birds were removed to the main settlements for storage.

Further evidence of specialisation on birds comes from a variety of sites. At Narva-Riigikula, on the southern shore of the Gulf of Finland, 90 percent of the bones were of birds, mainly ducks (Gurina 1967). House 3 at Gressbakken on Varangerfjord contained 1110 bones of guillemot (Uria aalge or U. lomvia), from at least 92 individuals, and 141 bones of other birds (Renouf 1981). The eastern Baltic and Varangerfjord are (like late mesolithic Denmark) areas where permanent settlements have been documented (see above). Other sites are less easy to evaluate. The site of Viste, in southwestern Norway, contained many bones of great auk (Brogger 1908), as did Klinteso in Denmark (Winge, quoted in Brogger 1908), possibly indicating specialisation on this now-extinct flightless bird. The famous site of Teviec off the Breton coast contained many bones of mallard (Anas platyrhynchos) and widgeon (A. penelope), although quantitative information is lacking (Pequart and Pequart 1937). This site, like Solager, is on an offshore island too small to support populations of wild pig and deer, bones of which were also found on the site. One may speculate that many other sites along the western rim of the continent which have no bones preserved may also have been camps specialising on birds. Land mammals. Storage of the meat of land mammals is obviously difficult to prove; bones from these animals will end up in the deposits of most sites whether or not the meat they carried went through a period of storage before consumption. One or two suggestions can be offered, however. It has been suggested that body-part representation of the animals at the famous site of Star Carr may mean that the most high-yielding parts of the animals were removed from the settlement (Legge and Rowley-Conwy 1988). If this is correct, the parts removed are likely to have been preserved in some way. This may also be the case for wild pigs at Ringkloster (Rowley-Conwy in preparation b). A different line of evidence comes from the age structures of red deer killed at certain sites. It is unusual for the age structure to mimic that in a live herd; this is, however, the case at the northern Spanish palaeolithic sites of La Riera, occupied from 19 000 to 7000 be (Clark and Strauss 1983), and El Juyo, of Magdalenian III date (Klein, Allwarden and Wolf 1983). For both these sites the authors argue that only the driving of entire deer herds into killing enclosures would produce such an age structure; individual stalking would produce a different pattern, so mass kills must have been the method used. In the relatively open landscape of lateglacial northern Spain, red deer may have formed quite large groups (cf. Legge and Rowley-Conwy 1988). The slaughtering of entire herds, to which the age structures at La Riera and El Juyo testify (also Mithen 1987), would thus have involved more animals than could be consumed in a short time. This hunting method must therefore have involved some storage of the meat. In glacial Europe storage would have been relatively easy, as meat could be frozen. We cannot demonstrate this, but there is some circumstantial evidence. The terminal palaeolithic sites of Stellmoor and Meiendorf in northern Germany show a major killing of reindeer in late autumn, a gap in midwinter, and a minor kill in the spring (Sturdy 1975). A recent reanalysis has convincingly

Saving it for later

55

suggested that body-part representation and butchery practices (at least in the ahrensburgian layer at Stellmoor) indicate the processing of meat for storage by drying (Gronnow in press). Large pits on many upper palaeolithic sites on the Central Russian Plain are interpreted as facilities for storing meat, and it is difficult to see how human groups could have survived the winters without extensive stored reserves (Soffer 1985). Evidence available for the storage of the meat of land mammals is thus very limited. However, by analogy with recent hunter-gatherers, we would expect many prehistoric groups to have stored land mammal meat. The scarcity of evidence reflects, we believe, not a scarcity of storage in prehistory, but rather our inability to recognise the practice in the archaeological record. Plant foods. Temperate European woodlands contain both oak (Quercus spp.) and hazel (Corylus spp.), and both acorns and hazel nuts would have been important, potentially storable resources, being extremely productive, rich in calories, and seasonally available. The hazel is documented in varying quantities from many mesolithic sites in East and West Europe, and acorns have been recorded at mesolithic sites with good preservation of organic materials (Dolukhanov 1979). Storage, however, is hard to demonstrate. Nuts stored in their fresh state would of course not survive, so we cannot expect to find untouched caches except under the most exceptional circumstances. Their potential role in storage systems may be illustrated by reference to the ethnographic records of, for example, California (Baumhoff 1981) and Japan (Matsuyama 1981). The best evidence available in Europe comes from the far north. Nuts do not just grow on trees: water chestnut (Trapa natans) is likely to have been an important storable resource, replicating to some extent the nutritional and caloric contents of cereals. Water chestnuts were more widespread in prehistoric Europe than they are today, as finds in Sweden, Finland and Estonia show that its range extended further north during the climatic optimum (Auer 1925; Zvelebil 1981). Archaeological remains testify to its importance. At Sarnate, for example, large lenses of water chestnut shells, up to 40 cm thick and extending over several square metres, were deposited around hearths inside dwellings. Burnt remains of water chestnut were also found in the ashes. Some dwellings contained wooden mallets with fragments of water chestnut embedded in their working surfaces. Such mallets, as well as remains of water chestnut in sometimes large quantities, have also been found at other stone age sites in the eastern Baltic (Vankina 1970; Dolukhanov 1979; Rimantene 1979) and Finland (Zvelebil 1981; Vuorela and Aalto 1982). At Orimattila in southern Finland, a broken pot was found associated with an accumulation of water chestnuts, suggesting storage (Zvelebil 1981:37). The use of other plant materials has been little documented in the European Palaeolithic and Mesolithic. However, attention must be drawn to the results of two recent analyses of charred food residues adhering to late mesolithic pot sherds from south Scandinavia. From the late mesolithic/early neolithic site of Loddesborg in Scania, analysis revealed the presence of seeds which (because of the absence of the amino acid alanin) definitely did not come from any cultivated cereals (Arrhenius 1984). Similar ma-

terial from the Ertebolle site of Tybrind Vig revealed traces of a monocotyledon, probably of the grass family (Andersen and Malmros 1984). Evidence for the use of such plants is unexpected, and suggests that the use of such eminently storable resources may have been more widespread in the European Mesolithic than hitherto suspected. Mass-capture technology Our final area of consideration is mass-capture technology. Such technology is necessary if the resource is to be stored, but is an 'overkill' strategy otherwise, and is therefore good evidence for the likelihood of storage. Remains of large nets are known from a number of sites. The best known is that from Antrea, now in the USSR, dated to the boreal period, and probably measuring around 27 m in length (Palsi 1920; Clark 1975). Another is known from Tuorsniemi in Finland, radiocarbon dated to 1900 be; the original length is unknown, but no fewer than 802 bark floats were associated with it. It may have been for catching seals, as several seal bones were associated with it (Luho 1954; Zvelebil 1981). A remarkable find from Lundfors, on the coast of northern Sweden, is dated to the later fourth millennium be. Netting was not preserved, but several lines of stone weights were recovered from the bed of a former inlet, indicating the positions of several nets. This too was probably for sealing. Lundfors was occupied for longer than just the period of seal availability, and may indeed have been a permanent settlement (Broadbent 1979: Figure 80). Further south, there is a late Atlantic net find from Scania (Klindt-Jensen 1957:43). Stone net sinkers are widely preserved in areas where the nets themselves have not survived, such as Varangerfjord (Renouf 1981), the eastern Baltic (Indreko 1948; Vankina 1970; Oshibkina 1978,1983; Dolukhanov 1979; Loze 1979; Rimantene 1979, etc.) and the Black Sea basin (Tringham 1973; Markevitch 1974; Belanovskaya 1975; Dolukhanov 1979). These are precisely the areas we have argued would see the development of storage. Net sinkers need not always be easily identifiable, as any stone about the right shape will do - for example, an unshaped stone was found at Niitjoki in Finland, with a twist of netting still attached to it (Topelius 1912: Figure 4). Nevertheless, definite evidence of nets in the form of recognisable stone weights appears to be lacking from much of the interior of Europe away from the areas we have discussed. Fish traps and larger fixed installations made of wooden stakes form the other category of mass-capture equipment for which we have evidence. Basket traps are known, for example, from Sarnate (Vankina 1970), Sventoji (Rimantene 1979), Abora I (Loze 1979) and Marmuginski peat bog (Dolukhanov et al. 1970) and several sites in Denmark (Becker 1941). A larger installation is known from Vis 2 in northern Russia (Oshibkina 1978), and another from Finland, where the Niitjoki river connects two lakes (Topelius 1912). Evidence is also available from Tybrind Vig in Denmark (Trolle-Lassen 1984; Andersen 1985); in the same region a possible find is reported from Svinninge Vejle, and another from Magleby Long; the latter example consisted of lines of close-set stakes extending 50 m either side of a wickerwork fish trap (Becker 1941). The best preserved example dates from the Bronze Age, and

Peter Rowley-Conwy and Marek Zvelebil

56

was found in Skedemosse on the Baltic island of Gotland. This consisted of a triple catching chamber and a 25 m arm connecting these to the dry land, all constructed of close-spaced stakes (Hagberg 1967). Little evidence is available from further south and west. Jacobi (1987) has suggested that the many axe-resharpening flakes found in the Thames might derive from the construction and maintenance of salmon traps, but no concrete examples are known.

towards richer areas and/or individuals with access to superior resource points. The environmental model thus accounts for both the growth of exchange, and for the emergence of unequal wealth - and so of social stratification - in certain specific environmental contexts. Once these developments are triggered off, the system becomes progressively more locked into these specific solutions to the problems posed by environmental risks. For example, by evening out the consumption of fluctuating resources, storage permits the growth of larger and more sedentary settlements. This has a number of benefits within the context of a storage-based system, such as (a) mutual protection of stored reserves, (b) increased potential for more specialised division of labour to cope with the periodic intensification of secondary resources, and (c) more scope for the exchange of prestige items within a system of social storage. Conversely, it becomes more difficult to use the mobility option as population rises - which in turn favours yet more resource and social storage. This, we believe, provides a more complete picture of the development of social complexity among hunter-gatherers than other models. Environmental variability is the key factor, one which demands considerable skills on the part of human exploiters. Each of our three levels of resource fluctuation (the seasonal, the interannual and the long-term) emphasises this. Thus storage each year is based not just on the experience and problems of the previous year, but on consideration of the worst likely case - a process demanding considerable planning depth. In its turn, planning for the worst likely case cannot lead to the adoption of a rigid and immutable strategy, because longer-term fluctuations are always likely to be at work outflanking any particular strategy. To planning skills must therefore be added a flexible socioeconomic behaviour, and a versatile, adaptable technology. Finally, in stressing planning, technology and behavioural flexibility, we have surely emphasised three of the most important factors enabling humans (originally tropical animals) to survive in the uncertain and variable environments of the north without major physiological change.

Conclusions

In the foregoing we have (a) considered the nature of seasonal, interannual and longer-term resource variability, and the problems faced by human groups dependent on such resources, (b) put forward a model predicting the circumstances under which the storage solution would be chosen, (c) suggested that certain social and technological features are likely to appear as a result of storage, and (d) examined the archaeological record of Europe for evidence of hunter-gatherer storage. We have found evidence for storage in the general areas of Europe we predicted. We thus conclude that the general statement that storage must be based on concentrated, seasonally available resources is supported by the evidence. Our more detailed suggestion was that selection of either storage or mobility will depend on particular aspects of the local resource spectrum, in particular the successive availability of several resources in a restricted area. This suggestion can be only partially tested with the evidence at present available in the archaeological record. Estuaries, coasts and rivers are most likely to have this resource pattern, and this is indeed where the evidence is most common. Our model has a number of implications for the way we view prehistoric hunter-gatherers. Our argument is (a) that storage is a response to a particular set of environmental conditions, and (b) that a series of other developments follow from this (see the section entitled 'Some corollaries of storage' on pp. 49-51). This is no less true regarding the social sphere: our model stresses that social developments are among these corollaries of storage. This adds a dimension to the view put forward by Bender (1978), that storage appears in response to developing alliance networks placing extra demands on production. This view operates at a general level only, and has never satisfactorily been able to explain why such developments in alliance networks (and in consequence storage and sedentism) should occur at particular places and times. The model put forward above, on the other hand, provides a specific context for these developments: firstly, social storage can operate much more effectively in the context of storage, and secondly, prestige items will more frequently flow

Notes We would like to thank Mr. B.P. Hodgson of Welsh Water, Northern District, for access to the Annual Reports of the Fisheries, Recreation and Amenities Officer, and Capt. T. Hughes-Parry, owner of the Llangollen Fechan beat on the River Dee, for access to his records, and both for many interesting conversations about salmon stocks and catches. We thank Todd Whitelaw and Paul Halstead for comments on an earlier draft of the paper.

Chapter 4 The role of wild resources in small-scale agricultural systems: tales from the Lakes and the Plains

J.M. O'Shea

This chapter examines the effect of interannual variability on smallscale agricultural systems, and the use of wild resources to stabilise fluctuations in agricultural productivity. Societies based on small-scale agriculture are frequently vulnerable to severe fluctuations in food availability. For such systems to approximate self-sufficiency, it is crucial that other, highly productive food resources be available, whose structure of interannual variability is largely independent of that governing agriculture. In the Old World, this problem is neatly solved by the coupling of agriculture with animal husbandry. In the New World, where this was not an option, a similar result was achieved by the hunting of large mammals or the harvesting of anadromous fish. Two ethnographic examples of this strategy from North America are briefly examined: the hunting-farming pattern of the Pawnee in the Central Plains region, and the fishing-farming complex of the Huron of the Upper Great Lakes region. It is concluded that the coupling of highly productive wild resources with simple agriculture represents a common coping strategy in cases where large domestic animals are not available, particularly in agriculturally marginal environments. Such buffering mechanisms vary considerably in their organisation, however, ranging from multi-community or multi-ethnic systems with local subsistence specialisation and regularly functioning exchange networks, to very generalised subsistence strategies, in which the members of a given group may divide or switch their efforts from one subsistence pursuit to another, depending on current local conditions.

The continued survival of small-scale agricultural communities is dependent on balancing those factors which produce variation in food availability from year to year with the strategies available to the community to even out such fluctuations. It will be useful to consider these factors and strategies in general terms, before moving to a more detailed discussion of specific cases.

The kinds of variability relevant to small, self-sufficient farming communities can be classified on the basis of their relative frequency and predictability. Frequency can be thought of in terms of seasonal, interannual and long-term, while high and low levels of predictability can be distinguished. Seasonal variation has the most immediate impact on farming communities. The cycle of planting, growth and harvest is fundamental to any simple farming community. The fact that there is a delay between planting and harvest imposes a limitation on the organisation of community life and consumption patterns. In the temperate zone, there is the further problem of winter, a season during which crops can be neither grown nor harvested. Clearly, if a farming economy is to be successful, it must have the means to provision itself during these non-productive periods. Yet, perhaps the most significant aspect of seasonal variation in food availability, from the perspective of the subsistence farmer, is that it is predictable. That is, it can be anticipated and, as such, can be accommodated. As was discussed in Chapter 1 of this volume, predictability allows various buffering mechanisms, such as storage, to be employed effectively. Since seasonal change occurs with perfect regularity, the value or necessity of the buffering mechanisms are constantly reinforced, and they become part of the 'normal' programme of activities. Interannual variation is the complement to seasonal variation. Whereas seasonal variation is completely predictable (and monocausal), interannual variation arises from a wide range of differing factors or agents and in most circumstances is totally unpredictable. Interannual variation in food production may result

J.M.O'Shea

58

from climatological factors, animal or insect pests, plant diseases and human activities. Likewise the impact of such factors varies drought may strike an entire region, while a hail storm may devastate one field and leave the next untouched. Therefore, in considering the factors that may produce fluctuations in food resources from year to year, it is not sufficient to identify the agent(s) responsible, but the structure of variability itself must be determined. Indeed, each agent that contributes to this variability must be assessed in terms of its spatial and temporal scale and severity (see Chapter 1). At yet a further level of complexity, the synergistic effects of such agents might also be considered. Not all interannual variability is unpredictable, of course. Some plant species, such as olives (cf. Forbes 1982;234-5), exhibit strong cyclical behaviour in yield. Yet, for most agricultural communities, the most critical feature of interannual variability is its unpredictability. Farmers (like hunter-gatherers; see Mine and Smith, this volume) have a sense of the causes of crop destruction and often have a sense of how frequently differing sources of failure may occur (cf. Hillman 1973:230-1). Yet this knowledge is always imperfect and, until the harvest is safely in store, the cultivator is left in doubt as to the relative success or failure of any given year's production. Unlike seasonal variation, which is inevitable, the factors that result in agricultural failure are probabilistic, so that the measures taken by the community to protect itself against crop failure may or may not be needed in any given year. In bad years, such precautions may save the community from starvation, but in good years, such measures may appear as little more than wasted effort. The balance struck between efficiency and security is determined by the perceived likelihood of failure, its expected severity and its expected duration. Such expectations must be drawn, in some manner, from the community's collective experience (cf. Colson 1979; Flannery 1986). Each major category of coping strategy - storage, diversification, mobility and exchange - can similarly be evaluated relative to the constraints imposed by the timing and territorial requirements of subsistence agriculture. Storage is the cornerstone of an agricultural economy, because of the inevitable lag between planting and harvest (Morgan and Munton 1971:43; O'Shea 1981). The physical storage of foodstuffs provides the primary means of levelling the predictable, seasonal variation in food availability. Its potential to buffer interannual variation, which requires that a crop be successfully stored over two or more years, is determined by the character of the crop, the climate, and the technology for storage. An alternative to physical storage on an interannual scale is exchange (which might be termed indirect or 'social' storage), in which food is exchanged against a future obligation of return. Exchange relations of this kind may be used in conjunction with physical storage so that the stored resources of a network of communities are available to buffer local shortage. These are the most common strategies employed by agricultural communities (beyond the level of crop and field diversification; see Forbes, this volume). They also entail the least disruption of the community and maximise local control over access to alternative food resources. Mobility provides a range of options in which the redistribution of population within a region is employed to lessen the ef-

fects of local shortage. For agriculturalists, mobility may take the form of the temporary movement of the community into an adjacent territory (e.g. the Bemba; Richards 1939) or it may represent a permanent abandonment of a region in the face of repeated failure. In either event, such movement constitutes a very serious decision for small-scale farmers, given the strong territorial basis of such economies. An extremely useful form of buffering, from the perspective of agricultural communities, can be realised when mobility is coupled with organisational fluidity. This may simply involve a redistribution of population within the pre-existing territory, perhaps to facilitate the harvesting of wild resources, or it may entail a shift in location and subsistence activity of some portion of the community. Diversification in agricultural economies may involve the use of a mixture of domesticated plants and livestock (cf. Halstead, this volume), the use of spatial strategies in the location of fields, or may entail the use of non-domesticated secondary or fallback resources. In poor agricultural years, farmers may resemble hunter-gatherers and may utilise a range of wild resources in the immediate environment. Of course, given the more sedentary character of most agricultural systems, such wild resources would rapidly be depleted by a population of any size unless sanctions existed to discourage use under less than emergency conditions (cf. Minnis 1985). Likewise, the simple operation of an agricultural system, particularly if coupled with stock raising, may of itself tend to degrade local wild foods (cf. Flannery 1969). A second potential limitation to the use of wild plant foods is that they may be affected at the same time as agricultural losses. Drought, for example, may well reduce both crop yields and, at the same time, reduce the availability of many wild plant foods. The usefulness of fall-back wild resources is thus determined by their relative independence of factors responsible for agricultural loss and by the balance between the density of the resource and the density of the population that may attempt to exploit it. Domestic animals represent a particularly effective form of diversification for small-scale farming systems. Large domestic animals are capable of converting non-usable plant material and agricultural waste into usable food. They can also convert and effectively store surplus agricultural produce. In addition, in all but the most extreme cases, the causes of agricultural failure are independent of animal loss and vice versa. Animals are also mobile, so that they can accompany the community if movement is dictated by extreme conditions. In effect, domestic animals function both as food converters and as a bank for surplus food value. They also offer the potential for the production of what Sherratt has termed 'secondary products' (1981), in that they may be exploited for milk, wool and traction even while they serve as a reservoir of stored food. Given these features, it is easy to see why domestic animals are so attractive to small-scale agricultural societies. Indeed, the linking of subsistence farming and stock raising produces a strong synergistic relationship. They are mutually stabilising, in that they have complementary patterns of interannual variability. The surplus agricultural produce may be 'banked' by use as animal feed, while the animals may increase agricultural

The role of wild resources in small-scale agricultural systems

59

efficiency by clearing fields and enhancing soil fertility via manure. Furthermore, both cereals and livestock have good storage characteristics; grain is typically dried, and preserves well, while animals store food value 'on the hoof and regularly produce interest in the form of increase in herd or flock size. If properly managed, the combination of livestock and farming may also reduce the strain on local secondary resources, allowing wild foods to remain viable as potential 'emergency food' to be exploited at infrequent intervals. Of course to realise this advantage, livestock must be managed to prevent overgrazing and substantial tracts of land must be left uncultivated. In most simple agricultural systems this is not a problem. The coupling of agriculture and domestic animals also permits a diet with an advantageous protein/carbohydrate balance (cf. Speth and Spielmann 1983) to be maintained year-round. Finally, surplus animal production can easily be used as a medium or commodity for exchange (cf. Rappaport 1968). It is not surprising, therefore, that animal husbandry and agriculture formed a well-integrated package that spread over much of the Old World. Although the initial development of agriculture may have been relatively similar in the Old and New Worlds, the characteristics of its spread, once a 'neolithic' package had been assembled, were very different. The stability and resilience that the combination of plant cultivation and animal husbandry produces may explain, in part, the divergence in pattern of agricultural dispersal: gradual and tentative in the New World, explosive in the Old. New World agricultural systems had the same need to buffer both regular and unpredicted scarcity. Although a number of domesticated animals were utilised to varying degrees throughout the New World, including dog, turkey and guinea pig, the absence of large meat-bearing domestic animals was significant not only for the immediate pattern of consumption, but also for the lack of secondary products and of the synergistic effects of livestock. For example, the absence of large traction animals placed severe limits on the range that bulk goods could be exchanged under most circumstances, and also hindered the development of more extensive forms of cultivation. The most serious effect, however, was the lack of a large, readily available food reserve to draw upon in times of agricultural shortfall. It seems likely that this lack was responsible, in part, for the very slow spread of agricultural economies in temperate North America, which witnessed the gradual incorporation of cultivation into what remained a diversified, hunter-gatherer type of economy, and for the relatively slow development of selfsufficient agricultural systems (cf. Ford 1977). Wild resources and social mechanisms must, therefore, have been crucial to the long-term survival of subsistence agricultural systems in the New World, particularly in the temperate zone. As long as agriculture remained one small part of a broad-spectrum economy, few adjustments to the pre-existing hunter-gatherer system would be required. As larger, more agriculturally based, systems formed, however, buffering activities focused directly on the agricultural system would be required. In effect, these systems required a substitute for the large domestic animals of the Old World.

Of course, even in the Old World, food production did not always spread in the form of the complete neolithic package. The early emphasis on sheep in the initial spread of Cardial Impressed Ware in the western Mediterranean (cf. Lewthwaite 1982) or of pigs and cattle in the Bug-Dniester culture north of the Black Sea (cf. Tringham 1971:98), are both instances in which specific elements of the so-called neolithic economy were incorporated into an essentially broad-spectrum economy. These instances appear more similar to the New World pattern of dispersal, but in both cases the cultures rapidly shifted into fully food-producing economies employing both domestic animals and agriculture. The critical point, therefore, is when societies come to derive a major portion of their diet (perhaps 40 percent or more) from agriculture and animal husbandry. Societies that incorporate food production on this scale are termed here 'agriculturally dependent'. Economies of this kind are qualitatively different from those that incorporate some agriculture into what is essentially a broadspectrum economy. Crop failure in an agriculturally dependent system results in major food deficiency and accordingly requires the intense harvesting of wild foods to make up the shortfall and avert starvation. The harvesting of wild resources, therefore, must also be relatively specialised and intensive, and may entail major structural accommodations within the society. Given the major labour and technological reorganisation such a system might require, it also seems reasonable to expect that buffering of this type will be most pronounced near the geographical and climatological margins of viable agriculture - areas of the most marked interannual variation in crop yields. Conversely, the greater the reliability of agricultural harvest, the less pronounced will be the formalised integration of a major wild component into the economy (although the emergency use of wild resources will always be present). If a wild resource is to buffer a subsistence agricultural system in a manner similar to that achieved through the use of domestic animals, it must fulfil four basic requirements. (1) It must occur in dense patches and be suitable for intensive harvesting. (2) It must have a high potential for storage. (3) It must exhibit a pattern of interannual variability that is largely independent of agricultural production. (4) It must have a pattern of seasonal availability compatible with the agricultural cycle and with buffering agricultural shortages. The large herd animals of North America, such as bison, and a variety of anadromous fish species fit all of these requirements. In some situations, gathered plant foods, such as wild rice or nuts, might also fulfil these requirements, although such resources are often affected by the same environmental conditions that produce agricultural shortfall. Although the time and labour requirements of an agricultural system place serious constraints on the range of wild resources that might potentially be used for intensive harvesting, there is still considerable latitude in the way communities may organise themselves to accommodate the divergent requirements of agriculture and intensive hunting-fishing-collecting. Indeed, it is possible to predict a series of organisational forms that would accommodate such a system of wild buffering. These model systems divide into two basic types, simple and complex.

J.M.O'Shea

60

Simple systems Simple systems, those which involve only a single cultural entity, can themselves be divided on the basis of the periodicity of wild resource harvesting. Regular systems encompass those adaptations in which both agriculture and intensive wild-animal harvesting occur every year. In such systems farming and hunting (for example) are both accommodated within the community's annual economic cycle, although the community as a whole may or may not be equally involved in both activities or in the seasonal movements required for them. By contrast, in episodic systems the harvesting of wild resources varies in intensity from year to year according to current agricultural success. Such systems may rely on early estimates of crop yields to determine whether intensive wild harvesting is needed in a given year or the community may wait until the harvest itself before deciding on the requirements for wild procurement. As with a regular system, episodic systems may or may not require the involvement of the entire community.

In so far as these model systems represent idealised forms, they create somewhat artificial distinctions that may not be fully separable in life. For example, most communities that participate in a complex, specialised system will, at the same time, be involved in a non-specialised system. Indeed, given the limitations on bulk transport that often characterise non-state societies, a purely specialised system would often be hard pressed to buffer a community against agricultural failure. Ford's (1972) study of the Tewa pueblos and their multi-layered exchange relations is a good case in point. Similarly, there is nothing to preclude a community involved in a simple system from also participating in a complex, non-specialised system. The distinctions are not completely arbitrary, however, since certain combinations of systems produce conditions in which the divergent goals of exchange participants may undermine the stability of the system. While the four models may not exhaust the range of possible buffering systems, they do begin to circumscribe the range of options that are available to subsistence agriculturalists. To illustrate the varied operation of 'wild' buffering systems, two ethnographic cases from North America will briefly be presented, exemplifying simple and complex systems respectively. For each a brief sketch of subsistence practices and relevant environmental features is provided, followed by a description of the buffering strategies that were practised.

Complex systems Complex systems, unlike simple systems, entail the interaction of more than a single social or cultural group. Two varieties of complex system can be distinguished, based on the degree to which the participant groups are economically specialised. In specialised systems, the principal buffering effect derives from specialised subsistence production (or overproduction) coupled with exchange across ecological boundaries. A common form of complex system develops when a series of communities tends to specialise in agricultural production while maintaining regular exchange relations with nomadic hunter-gatherers (cf. Spielmann 1982). Such a system is complementary, in that the buffering needs of the agriculturalists are met while the carbohydrate requirements of the foragers are equally satisfied. A wide variety of locational models can be envisioned for exchange-based systems of this type. Non-specialised systems represent cases in which networks of simple farming communities are linked via the maintenance of extensive reciprocal exchange relations. These networks provide a vehicle by which food resources in one locality may be made available to communities experiencing scarcity, or the linkage that allows communities to join forces for collective hunts or fish harvesting in time of need. Although there may be a degree of overproduction of subsistence goods for use in trade or ritual, the episodic nature of this form of buffering network limits the value of high levels of overproduction. Covering ceremonies and ritual may also obscure the latent importance of risk buffering to the exchange transactions. In both varieties, the movement of goods and the maintenance of obligation, the critical elements of an exchange network, are the factors that produce a buffering effect. The essential difference between specialised and non-specialised systems, therefore, is the role played by the individual groups of communities that participate in the exchange network. In non-specialised systems, all participant communities pursue essentially similar subsistence activities. Indeed, at the level of a single community, a non-specialised complex system may resemble a simple type of subsistence strategy.

The Pawnee The Plains region of North America, stretching from the northern limits of aboriginal maize agriculture in southern Canada (Moodie and Kaye 1969) to the margins of the agricultural Pueblo region in the American Southwest (Wedel 1961), is the first region in which the use of wild resources in agricultural buffering strategies will be considered. The economies of the sedentary Plains Village Tribes, that mixed in varying degrees maize agriculture and bison hunting, epitomise a simple/regular subsistence strategy. Although plant cultivation is known in the region during the Woodland period (100 BC to AD 900), a stable agricultural economy that provided a significant proportion of the total diet probably did not appear until the formation of what Lehmer (1954) has termed the Plains Village pattern or tradition. In the Central Plains region, this tradition is represented, prehistorically, by taxonomic units such as Upper Republican, Loup River, St. Helena and Nebraska phases. These early Plains Village systems are replaced after about AD 1500 by Lower Loup Phase and the Pawnee of historic times. The Pawnee lived in large, earthlodge villages located on the Platte, Loup and Republican Rivers in east central Nebraska and hunted over a broad region of Kansas and Nebraska. During the nineteenth century the Pawnee tribe existed as four band villages, although there is strong evidence to suggest that these were, themselves, the product of a recent coalescing of smaller autonomous villages (cf. Parks 1979). The Pawnee, along with their close relatives the Arikara, are speakers of a northern Caddoan language and probably are the direct descendants of the original Plains Village groups of the Central Plains tradition (Strong 1935:277). The Pawnee subsistence cycle began in mid-spring, when

The role of wild resources in small-scale agricultural systems

61

fields were prepared for planting. Each woman tended her own plot (or plots), which averaged between 0.5 and 1.5 acres per cultivator (Will and Hyde 1917:65), while a typical family maintained a total of 1 to 3 acres (Dunbar 1880:276). Planting took place in late April or early May, and the women continued to work in the vicinity of their fields until the second hoeing, typically in midJune. During this portion of the summer the men were engaged with local hunting and maintenance activities. Once the second hoeing was completed, the crop was left until harvest in the autumn, as the Pawnee departed from the permanent village and fields for the summer hunt. The entire community, barring only the very old and sick, would leave the earthlodge village for the summer bison hunt. Each band had its own habitual hunting area, and it was not uncommon for a band to travel 150 miles or more away from their permanent settlement (cf. Mora in Nasatir 1952:119-27; Holen 1983). The bison hunt, as observed in historic times, was a wellregulated affair, with specially designated tribal 'police' assigned to maintain order and prevent premature frightening of the bison herd. The hunt could be very productive. Stuart (1953:157) reported that the South Band Pawnee produced 1200 hides and the Skidi band 800 hides for trade in 1811, and that in a good year they might produce half as many again. These numbers are staggering, particularly when one considers that this is only that portion offered for trade and excludes the additional hides consumed for the numerous domestic uses. Stuart's figures also suggest a minimum quantity of usable meat (following Wheat 1972:108) in the range of 600-900 tons. Of the meat, a portion would be consumed immediately and the rest cut into thin strips and dried for storage and later use. The hunt continued until early September when the bands returned to their agricultural fields. On their return, a portion of the crop was harvested 'in the milk', roughly two weeks before ripening. This green corn harvest represents a classic trade-off between risk avoidance and maximisation. Corn harvested in the milk stage was food in hand, a factor of some concern, given the high potential for loss from small animals and grasshoppers during the corn's final maturation. By harvesting prior to the completion of maturation, however, roughly 20 percent of the expected caloric value of the grain is lost (Morrison 1954:569). As a result, only a portion of the maturing corn was harvested at this time, ensuring that at least some of the crop would be recovered. Following the final harvest of the crop and its storage in subterranean caches, the bands set out on their winter hunt. By all accounts, these were less joyful outings and the bands kept mainly to the sheltered river valleys (as did the bison, deer and elk). Smaller quantities of meat were obtained during the winter hunt, although in normal years there was still enough to warrant its storage (cf. Dunbar 1918:605). Likewise, a portion of the autumn maize harvest was carried along during the winter hunt and was consumed during this time (cf. Stuart 1953:154). The bands returned to their earthlodge villages in the spring and subsisted on stored corn and dried bison meat until the prairie greened and the cycle began again. For the Plains Village tribes, maize was the principal agri-

cultural crop, although beans, squash, pumpkin and sunflower were also cultivated (cf. Gilmore 1977). Maize yields under native cultivation typically varied around an average figure of 20 bushels per acre (Will and Hyde 1917:142). The pattern of maize cultivation was surprisingly similar throughout the Plains region and beyond. The soil was not broken in its entirety, but rather was mounded into small hills that measured approximately 1.0-1.5 metres in diameter and 0.2-0.4 metres in height. The spacing between the mounds was on the order of a metre. Beans were often planted between the corn hills and were allowed to use the maturing maize stock as a support. Fields were typically located along stream and creek bottoms, locations that contained soils of high fertility that could easily be worked by hand. The same corn hills would often be used over several years, although each spring the women would hoe up additional soil and humus to the mound. In other years the same field might be used but with new hills built in alternate locations. There is little evidence to suggest that depletion of soil fertility was a serious problem among the Plains Villagers. The principal causes of crop loss on the Plains were drought and grasshoppers (Will and Hyde 1917:141). Additional losses were frequently experienced due to hail, birds, plant disease and, increasingly during the historic period, raiding. The length of growing season was rarely a concern, although occasionally a very late spring frost might force the maize crop to be replanted. Horticulture in the grasslands offers different problems and possibilities from forested environments. Although the grasslands are notoriously difficult to break and work, particularly in the absence of iron tools, they do not require the massive commitment of labour and long delays associated with forest clearance. Plains farmers, therefore, were able to scatter their fields, both to exploit micro-environmental advantages and to decrease the risk of catastrophic crop loss resulting from insect infestation, hail, crop disease or raiding (cf. Wilhelm 1973:335). Likewise, since the length of growing season was rarely a constraint, the Plains Villagers could plant in low, drought-resistant locations without serious fear of killing frost. The occurrence of adequate precipitation was, however, a major limiting factor for reliable maize cultivation on the Plains, particularly west of the 100th meridian. Finally, during the historic period, raiding of agricultural fields and stores became a new and highly significant threat to agricultural subsistence. The Plains Villagers employed a wide range of coping mechanisms. The most commonly employed buffering mechanism was storage in subterranean caches. These bell-shaped pits were large and frequently held the produce of two or more extended families (cf. Weltfish 1965:268). It is also apparent that special care was taken in the storage of seed corn. Among the Hidatsa, for example, two years' worth of seed corn was kept in sealed ceramic containers. Native cultivators believed that, after two years, too little of the seed would germinate to be of value (Wilson 1917:48). Such a strategy ensured that they could survive a year of total loss and still have seed to plant in a second year. Two years of seed corn was not always sufficient, however. On several occasions, for example, the Ponca were forced to trade for new seed either after several consecutive bad years or when, in time of particularly

J.M.O'Shea

62

severe shortage, they were forced to eat the seed corn itself (Will and Hyde 1917:141). Several forms of diversification are documented, of which the dispersal of agricultural fields is the most obvious example. Another common form of diversification was the cultivation of a variety of maize strains, each with its own particularly desired characteristic, such as a shorter growing season or drought resistance. The combination of dispersed fields and differing maize strains tended to ensure that at least some agricultural return would be realised. These practices could be supplemented by an array of 'fall-back' or emergency plant foods that could be gathered locally. Diversification and storage were supplemented by strong social obligations for sharing within Pawnee villages. The earthlodge, which was typically inhabited by two (not necessarily related) extended families, was the primary productive unit in Pawnee society. Although each woman maintained the produce of her fields separately, it was often stored in separate bags within a single shared cache pit. The women of a given earthlodge also shared in the duties of food preparation, with each woman drawing on her own stores in turn (Weltfish 1965:15). Thus a significant degree of effective pooling or sharing would have been realised within each lodge and, since the holdings of each cultivator were scattered, a degree of low-level buffering would have been derived from such practices. Similarly, social prescriptions concerning feasting and generosity encouraged similar sharing between households within a Pawnee village. Although there is little direct evidence concerning the sharing or trade of foodstuffs between Pawnee villages, the ubiquitous movement of individuals and families between villages, and the occasional movement of whole villages (as in 1832-6 when the Arikara lived with the Skidi band of the Pawnee) suggests that some succour from serious shortages could have been obtained in this manner. Inter-community exchange beyond the bounds of the individual tribe was highly developed among the Plains Villagers and presented a final defence against serious shortage. This intercommunity trade was organised under the ritual cover of the wellknown calumet ceremony (Blakeslee 1975). So significant was the calumet that tribes in open conflict would maintain a truce (or even declare peace) so that the ritual and the exchange associated with it could be conducted. Although the ostensible focus of exchange associated with the calumet ceremony was reciprocal gift giving between the host and his fictive kinsman, it provided an opportunity for the extensive exchange of foodstuffs, skins and craft goods. Through the mechanism of the calumet, therefore, autonomous and ethnically dissimilar groups of Plains Village farmers were able to articulate on an 'as needed' basis for the exchange of basic subsistence goods. This network provides a limited type of complex/unspecialised buffering system. The major counterbalance to agriculture production, however, was large mammal hunting, with an overwhelming emphasis on bison but also including elk, white-tailed deer and pronghorn antelope. In historic times, aided by the acquisition of the horse, bison hunting became almost an industrial activity in scope and intensity. Two alternative seasonal patterns of big game hunting

were employed during the eighteenth and nineteenth centuries. The first, practised by most Plains Villagers, employed major communal hunts in both summer and winter, sandwiching the season of agricultural activity. The second pattern, practised by tribes living near the Missouri River trench, such as the Omaha, utilised only a single community hunt which occurred in the summer. During the winter, hunting was conducted by small groups of men for deer, elk and beaver in sheltered areas (cf. Fletcher and La Flesche 1911:271). It seems likely that these alternative forms relate to the primary setting of the groups, with the latter being more characteristic of a prairie/woodland settingand the former more typical of open grasslands. Bison and other large mammals" also provided important raw materials for the manufacture of clothing, tools and ornaments. These supplied not only the tribe's own needs, but also provided an outlet for craft manufacture that, in some cases, played an important role in maintaining inter-tribal trade relations. The Pawnee provide an example of a simple/regular system, in that both agriculture and intensive hunting were practised every year as part of the normal course of events. Hunts were not cancelled in years of good harvest, although they were intensified and lengthened in years of poor harvest. In addition, the redundant exchange of foodstuffs between Pawnee villages and with other village tribes via the calumet ceremony, constitutes a second, complex/non-specialised buffering network. An interesting variation on the simple system described for the Pawnee is found among the contemporary nineteenth-century Hidatsa of North Dakota. In addition to the buffering activities discussed for the Pawnee, the Hidatsa participated in a system of complementary exchange with a number of nomadic Plains groups. The nomads would transport products of the hunt to the agricultural villages of the Hidatsa to exchange for corn (cf. Wood 1974), and later for European manufactured goods. In these transactions, surplus corn was exchanged in good years for meat, skins and crafted items. From the perspective of the nomads, however, this exchange relationship had a serious weakness: the Hidatsa themselves ventured out on at least one community hunt each year. So in addition to the scalar limits on the buffering value of meat carried to the villages by the nomads, and the universal difficulties of continuing trade relations in years when either the farmers or hunters had little food to exchange, there was the added difficulty that the Hidatsa had direct access to the alternative resources the nomads sought to bring in trade. During the historic period, the Teton Sioux solved this problem by forcing the Hidatsa into the role of specialised agriculturalists by preventing them from leaving their villages to hunt. Lewis and Clark (Coues 1893:231) noted that the Sioux burned the prairies in the vicinity of the Hidatsa villages, that they attempted to drive game away from the villages, and through persistent raiding attempted to prevent hunting parties from leaving the villages. The actions of the Sioux, therefore, can be viewed as attempting to transform the Hidatsa subsistence economy from essentially a simple/regular system with occasional trade relations into a complex/specialised system. The continued development of this exchange relationship was abruptly truncated

The role of wild resources in small-scale agricultural systems

63

by the great smallpox outbreak of 1837 that decimated the Hidatsa(Chardonl932). The system of agricultural buffering employed by the Plains Village farmers, such as the Pawnee and the Hidatsa, represents an extreme variety of a simple system in which relatively large local populations were supported via the intensive exploitation of both bison and agriculture. The intensity of this exploitation in turn increased the vulnerability of the systems both to perturbations in the key resources themselves and to the influence of outside factors that might disrupt the intensity with which the resources could be exploited or the security which could be assumed for stored produce. Such intensive exploitation appears to have been a product of recent times, the proto-historic and historic periods, and represents a marked change from the low-intensity pattern of resource exploitation that characterised the earlier, dispersed settlements of the Central Plains tradition (cf. Wedel 1986). A measure of the increased vulnerability of the system during the historic period is well illustrated in the reports of the Pawnee Indian agents during the period 1860 to 1874, during which time the Pawnee experienced six poor harvests and 12 inadequate hunts, representing a serious shortfall in subsistence goods nearly every year (Wishart 1979:396). Yet, there is the strong suggestion that, despite the large populations and the intensity of resource exploitation, the Plains Village system probably could have remained viable in the absence of the additional stresses of warfare. Of course, the large villages of the historic period are, themselves, the product of warfare and the needs of mutual defence, so such an assertion is completely hypothetical. Yet, the fact that the Plains Village system persisted as long as it did in the face of such disruptions must highlight the resilience inherent in simple types of buffering systems when suitable wild resources are available.

grown in large village fields. Isotopic analysis of Huron skeletal material suggests that a more realistic figure might be 50 percent of the diet (Schwarcz, Melbye, Katzenberg and Knyf 1985). Although maize was the principal agricultural crop of the region during the late prehistoric and historic era, beans, squash and sunflower were also grown (Yarnell 1964). The pattern of maize cultivation was surprisingly similar to that described for the Plains region. The differences in practices between the two regions relate, primarily, to the problems posed by cultivation in a forested, as opposed to a grassland, environment. The first step in field preparation was forest clearance, an activity of considerable effort, given the heavily forested character of the region. The soil surface was not entirely broken, but rather was mounded into small hills. Fields tended to be concentrated to take advantage of good agricultural soils (cf. Heidenreich 1971). Although the fields cultivated by each individual were of a similar size to those in other regions, the concentration of these small holdings into relatively limited areas made them effectively large village fields. The tendency for villages to maintain a small number of large fields, rather than a series of small dispersed fields, reflects the heavy labour costs associated with field clearance, the limited distribution of good agricultural soils, and the deleterious crop effects inherent in small fields within a forested region. Maintaining large village fields would minimise the small-field effects and would make the crop easier to protect from bird and small animal predation. The 'cost' of the large-field strategy is increased subsistence vulnerability. Localised pest infestation, meteorological conditions such as hail, and crop disease might be sufficient to destroy a community's entire corn crop. Likewise, a single raid by hostile neighbours might be sufficient to reduce the community to destitution (cf. Thwaites 1896-1901, vol. 35, p. 191). In this sense, the large-field strategy commits the community to an 'all or nothing' return from their agriculture, whereas the small, dispersed field approach often returns at least partial harvests. The largefield strategy may also limit the society's ability to exploit small but particularly advantageous micro-environmental settings for agriculture. This more intensive use of fields also meant that villages were forced to relocate on a 10-20-year cycle to maintain soil fertility.

The Huron The upper Great Lakes region, as used in the present discussion, includes the land areas bordering Lakes Superior, Huron and Michigan in Canada and the United States. Maize agriculture became significant in this region c. AD 1000 and continued up to the time of contact. The Huron were a confederacy of four related Iroquoisspeaking tribes that inhabited a region centring on the narrow isthmus separating Georgian Bay and Lake Simcoe in central Ontario. During the historic period, the Huron lived in large, palisaded villages of 1,500 to 2,000 people, or in smaller satellite settlements that tended to be tied to the larger villages for rituals and defence (Trigger 1976:32; Warrick 1984). Each village was composed of one or more longhouses, multi-family dwellings that typically ranged from 40 to 75 metres in length and might house as many as 24 nuclear families (Dodd 1984:207-8). Although farming was the responsibility of each individual family, significant pooling of produce occurred within the longhouse, making it effectively the primary productive unit within Huron society (cf. Hayden 1977). The Jesuits estimated that three quarters of all the food eaten by the Huron was maize (Heidenreich 1971:162), which was

Estimates of typical yields for agriculture in the region vary widely. A minimum figure for the upper Great Lakes region as a whole, offered by Hinsdale, is 15 bushels/acre (1928:45), reflecting a certain conservatism relating to the actual productive potentials of maize in the upper Great Lakes region. On the other hand, the Jesuit missionaries living with the Huron during the first half of the seventeenth century estimated typical yields in the range of 27 bushels/acre (cf. Thwaites 1896-1901, vol. 15, p. 157). A principal natural cause for crop failure was the occurrence of killing frosts. Few areas in the upper Great Lakes have modern growing seasons as long as 140 days, and much of the region where agriculture was practised has a modern season length that averages nearly 120 frost-free days, a frequently cited minimum value for successful native cultivation (Yarnell 1964). A second, and often overlooked, source of crop loss was insufficient precipitation and soil moisture. The particular importance of this

J.M.O'Shea

64

perhaps obvious factor is that the field settings and soil types a native cultivator would select for frost avoidance, light soils in upland settings (Heidenreich 1978:375), are precisely the settings and conditions that are the most vulnerable to drought (Trigger 1976:36). Cultivators, therefore, were forced to make a series of complex compromises in the selection of field location. In addition to these very generalised sources of crop failure, Huron agriculture was also vulnerable to a variety of local pests, including blackbirds, grackles, jays, raccoons, sandhill cranes, grasshoppers and caterpillars (Heidenreich 1978:380). In addition, produce, once harvested and stored, was vulnerable to rodents within the living structures and to fire (Sagard 1939:95). Another source of agricultural loss that became increasingly serious for the Huron during the seventeenth century was the destruction of standing crops by raiding parties (cf. Thwaites 1896-1901, vol. 35, p. 191). The degree of risk inherent in agriculture in the upper Great Lakes region is well illustrated in historical reports of famine among the Huron. During the period AD 1628-50 (excluding the years 1630-3, a time for which no reports were made) the Jesuit Relations report that the Hurons experienced severe famine in 1628, 1635, 1638, 1643 and 1649. In other words, during this period the Huron could expect a severe crop failure one in every four to five years! One aspect of crop failure within Huronia that is not certain is the spatial scale of major crop failures or the frequency of failure at differing spatial scales. The Jesuit Relations imply at least three levels of scale: nearly universal, affecting not only Huronia but the adjacent territory of the Neutral Iroquois (cf. Thwaites 1896-1901, vol. 27, p. 65); an intermediate level affecting Huronia and Neutral country differentially (cf. Thwaites 1896-1901, vol. 15, p. 157); and a yet smaller scale, in which areas within Huronia itself were differentially affected (cf. Thwaites 1896-1901, vol. 17, p. 119). Clearly, differing buffering mechanisms will be required for each level. As would be expected, the buffering activities employed by the Huron were both extensive and multi-layered. The primary local buffering strategy for a Huron community was physical storage. Champlain noted that the Huron, for fear of famine, planted enough corn each year to last them through several years (Trigger 1969:27). The processing of a proportion of the harvest while in the milk stage (green corn) was also customary (Sagard 1939:106). Maize was dried and stored in large bark containers within each longhouse. The use of large village fields probably minimised the differences between longhouse production within a single village, limiting the utility of very localised exchange. Although storage in the longhouses probably offered some additional security from rot and pests, the every-present danger of fire in these highly flammable structures would also have put the stored crops in a degree of peril. Resource diversification played a major role in the Huron subsistence system. In addition to the cultivation of a variety of domestic plants, the Huron recognised a range of wild plants that were utilised as famine or emergency foods. These included acorns, blueberries, tree bark and lichens (Aller 1954) and were typically exploited only in times of dire need. Given the large size

of the Huron villages during the historic period, the exploitation of such plant foods in times of emergency necessitated that the population disperse into small, wide-ranging collecting groups. Whereas the use of wild plant foods was largely in the context of emergency food, the exploitation of animal resources was a more central and regular aspect of Huron subsistence diversification. Among the agricultural communities of the upper Great Lakes, generally, hunting was pursued year-round, but with particular emphasis during the late autumn and winter. Deer hunting in the Great Lakes was not on the scale of western bison kills or antelope drives, but under proper conditions large numbers could be taken in surrounds and collective hunts (cf. Champlain 1907:288-9), and during months of snow cover when the animals tend to yard. The food value of these animals, however, deteriorated as winter ran its course and their fat was depleted (see Speth 1983). In years of good harvests, meat probably served as a welcome dietary supplement. In years of poor harvest, hunting may have provided one crucial fall-back resource with which the inhabitants could survive the winter. In addition to larger game, small mammals, particularly beaver, were a common source of winter food in the western Great Lakes region (cf. Henry 1921:131 -2). When the role of hunting in these economies is being considered, it is important to remember that deer, elk, moose, bear and beaver supplied not only meat, but also skins for clothing and trade. For the Huron, however, hunting was insufficient to buffer serious agricultural loss (Heidenreich 1978:382). It has even been suggested that the resident deer population in Huronia was already greatly diminished by historic times. For this reason the Hurons did only limited hunting themselves, and instead relied on regularly operating exchange relations for the acquisition of meat and hides. In regions of lower population density, hunting was a more viable option. By contrast, the upper Great Lakes region was in prehistoric times (and still is) incredibly rich in fish. Lake trout, whitefish and sturgeon comprised the most important food species. The density offish available in some prime lake locations, such as the Straits of Mackinac or Sault Ste Marie is legendary (cf. Henry 1921:55—65). Fishing was an important activity for the Huron year-round, but particularly during the spring and autumn, with spring fishing tending to be in shallow waters and winter fishing tending toward deeper waters (Heidenreich 1978:382). The catch from autumn fishing was typically smoked and stored for consumption over the winter (Sagard 1939:95). In suitable locations, fishing continued through the winter, as the natives devised elaborate systems of string nets beneath the ice cover (cf. Champlain 1907:331-2). Ice break-up on the lakes in the spring was eagerly awaited, often as a means of alleviating the lean times that resulted as winter stores were depleted (Thwaites 1896-1901, vol. 35 pp. 183-5). Since village location was determined primarily by agricultural requirements and since the large village fields required constant vigilance to protect the crop from pests, fishing tended to be conducted by task groups that journeyed to the fisheries and, in most cases, processed the catch on the spot for transport back to

The role of wild resources in small-scale agricultural systems

65

the village (Heidenreich 1978:382). In a sense, the problem for the Huron was similar to that of the peoples of the Northwest Coast region of North America, where the limiting factor was not the availability of the food resource but of the technology and manpower for the processing and storage of the catch (cf. Oberg 1973). Yet, despite the abundance, the fish runs too occasionally failed, often with catastrophic results, as the Jesuits reported in the year 1650 (Thwaites 1896-1901, vol. 35, pp. 175-7). The harvesting of anadromous fish satisfies the theoretical requirements described for a wild resource to counterbalance the expected year-to-year variability in agricultural yields. The fish of the Great Lakes represent a seasonally very dense resource, one with a relatively high suitability for storage, and one with a pattern of seasonal and interannual variation that is in large part independent of the factors influencing agriculture. Indeed, Cleland (1982) suggests a shift in the primary season offish harvesting in the Great Lakes from spring-running fish in the Archaic to autumn runs in the Woodland period. Such a shift would be particularly crucial if fishing was to serve as a back-up for agriculture. Exchange, as a buffering device, operated on a series of levels among the Huron. Although there were strong prescriptions relating to reciprocity and generosity within Huron villages, the value of such exchange for buffering agricultural shortages was probably limited, given the practice of cultivating large village fields and the resulting similarity in harvests between households within a village. Exchange between Huron villages had the potential of levelling local crop losses. As was noted previously, such exchange does not appear to have been achieved through the movement of goods, but rather by the relocation or dispersal of population from villages suffering serious shortages to the vicinity of villages with surplus produce (cf. Thwaites 1896-1901, vol. 35, pp. 177, 183-5; Heidenreich 1978:378). Even this kind of buffering would be of limited value, however, in the face of larger-scale failure. Exchange beyond the borders of Huronia was along two major axes, to the agriculturally based Neutral Iroquois villages to the south and to the Algonquin hunter-gatherer-fishers to the north and west (Tooker 1964:12-20). The trade with the Neutrals was principally for deer meat and hides, although tobacco was another commodity that moved northward in this trade. Huron sources claim that this trade was necessary, since deer were scarce in Huronia (cf. Kinietz 1965:20). Although the southern trade for deer was important to the Huron, it seems to have constituted primarily a 'normal year' activity. The Huron also sought to obtain corn from the Neutrals in years of poor harvest (Trigger 1969:27), although the magnitude of such trade and its reliability is unclear. The Jesuits similarly mention attempts on the part of the Neutrals to obtain maize from the Huron. During a particularly bad year they offered to exchange their children for corn (Thwaites 18961901, vol. 15p. 157). The trade with the Algonquins involved the complementary exchange of dried fish and meat for maize. For the Algonquins this provided an important carbohydrate source in the diet and a degree of security against the eventuality of a bad autumn hunt. For the Hurons in good years, the exchange similarly produced

dietary diversity. During years of poor harvest, however, the dried fish and meat might have made the difference between life and starvation. The trade, importantly, also involved a variety of manufactured items, including skins and furs from the Algonquins and fish nets and wampum beads from the Huron. These non-food items might have played a crucial role in the continued operation of the exchange network in years when one side was experiencing serious food shortage. The system that developed among the Huron and the Algonquins fits a complex/specialised model. Both the Huron and the Algonquins effectively specialised in their own subsistence activities. This is not to say that there was no overlap - several of the more southerly Algonquin groups did maintain small maize fields, while the Huron themselves were active fish harvesters - but this does not lessen the effective specialisation that had developed. Both societies regularly overproduced \n anticipation of exchange. The regularity of this overproduction played an important role in the risk-buffering value of the system, not simply by ensuring that excess food was on hand, but in providing momentum for the continuation of trade, even in those years when one side had little or no food to exchange. The maintenance of trade when one side experiences a bad year is the critical test for a complex system. One mechanism which enhanced the likelihood of continued trade in poor years was the practice of Algonquin bands regularly wintering at particular Huron villages (Trigger 1976:63). This tying of specific agricultural villages with particular hunter-gatherer bands must have helped regularise exchange and bind trade partners, supplementing the more typical devices of fictive kinship and the exchange of wives or children (Trigger 1969:39). The fact that both sides produced non-subsistence items for exchange is also significant, ensuring that a two-way movement of goods continued even in years when one side had little or no food. Despite the complementarity of this system, its actual capacity to buffer catastrophic crop losses was limited. One limit was the quantity offish or meat that could be processed and transported to the Huron villages. A second was the size of the resident population in the villages themselves, particularly since the scarce provisions had to supply not only the Huron, but also their Algonquin visitors, A first line of defence would be the stored crops retained from a previous year, if they had not been dissipated through trade. In years when famine was not regional in scale, food exchange between villages would also have been a possibility, along with population dispersal among villages to decrease the number of dependants in areas of severe shortage. In all, it seems clear that the mechanisms available were sufficient to lessen the impact of agricultural shortfall, but they certainly were not adequate to compensate for major crop losses. Hunger would still have been rampant. It also seems likely that a succession of bad years would completely overwhelm such a system. A buffering strategy of the type employed by the Huron would have provided a much greater degree of security for villages with smaller resident populations. This apparently was not an option available to the Huron during historic times, because of an almost constant state of warfare with the Iroquois confederacy.

J.M.O'Shea

66

There is, however, strong evidence to suggest that smaller villages were the norm among the prehistoric Huron (cf. Warrick 1984), a situation that would not only make failure easier to buffer, employing the strategies already described, but would also have enhanced the potential of local hunting and collecting as additional stabilising activities. The shift through time, therefore, seems to be toward a less stable subsistence economy in response to the pressing needs of warfare and defence. Such a shift is virtually identical to the historical trajectory described in the Pawnee example.

tems are inherently less flexible and afford less control to individual participant communities. The lack of flexibility and control is the direct result of the necessary coordination of differing groups through time in a regular pattern of exchange. Such a system cannot tolerate deviation from the expected in response to special local circumstances without jeopardising the entire system. Complex systems seem most likely in situations of high population density, in instances where the territorial security required to maintain a simple system is absent, or in cases of natural (or artificial) limitations on access to suitable wild resources. This latter circumstance was well illustrated by the interaction of the Teton Sioux and the Hidatsa, in which the Sioux coerced the Hidatsa into a complex system by forcibly preventing them from conducting their own hunting forays. To the extent that complex systems involve the cooperative interaction of numerous distinct social entities, they are inherently less stable than simple systems. Yet, they may persist over long periods of time under certain conditions, as in the well-established systems of the American Southwest. Unfortunately, neither case study provides much insight into the conditions under which complex systems might persist. One factor which may be important is the degree to which the resource complementarity that characterises complex specialised systems is dictated by environmental conditions as opposed to those with a solely cultural basis. Similarly, systems that are only required to operate at relatively lower intensities and which support lower densities of population probably have a better chance of remaining viable through time. Beyond these general observations, the consideration of risk buffering and the needs of fully agricultural systems may also suggest an explanation for some of the differences observed in the spread of agricultural economies within the New World. The potential of wild resources to buffer agricultural shortfall depends, fundamentally, on the existence of a wild staple that can be harvested intensively and stored. The existence of such resources may significantly influence the rate at which agricultural dependence develops within a given region. Just as the absence of large domestic animals tended to confine earlier plant cultivation within an essentially broad-spectrum hunter-gatherer system in the New World, so the availability of suitable wild resources to buffer agriculture may have conditioned the rate at which agriculturally dependent systems developed and spread throughout the New World. It is interesting to note in this regard that both ethnographic examples presented in this chapter derive from regions that are marginal from the perspective of maize cultivation, yet where agriculturally based economies could nonetheless be maintained because of the presence of dense and abundant wild resources. In regions more ideally suited to maize agriculture but that lacked suitable wild resources, such as much of the central and non-coastal eastern United States, investment in food production would have been extremely risky, despite the better growing conditions. In the absence of wild resources, the main buffering strategies for cultivators in these areas would be limited to storage and exchange. Major dependence on food production could therefore only become viable once the network of exchange relations had developed beyond a critical size and regularity.

Discussion The case studies suggest, in at least a preliminary way, certain regular features that might be associated with the differing forms of wild buffering systems, and also some contrasts between the differing systems. In particular, simple and complex systems differ in the nature of their operation and their likely stability through time. Simple systems, in which wild and domestic resources are managed within the confines of a single community (via a shifting adaptive posture), afford the greatest degree of control for simple societies (although uncertainty remains with respect to the natural environment). Such systems also offer the greatest flexibility, since decisions can be made on a year-to-year (or season-to-season) basis in response to immediate local conditions. Furthermore, when relatively low population densities are supported by simple systems, there is apt to be a greater range of viable options for coping with catastrophic shortage, including dispersal, exploitation of emergency foods, and individual trade partnerships. Although greater flexibility and control make simple systems relatively reliable, the perpetuation of such systems seems, in the absence of major climatic or environmental change, to be dependent on the maintenance of ample and secure territories. Within such territories, and in the absence of significant warfare, simple systems appear capable of withstanding all but the most severe environmental perturbations. Yet, when territorial integrity begins to break down, simple systems become extremely vulnerable and are often either transformed into complex systems or collapse altogether. The reason why territorial integrity is so crucial to societies practising a simple-system strategy relates to the periodic need to shift adaptive posture, which means that agricultural fields, settlements and stored produce must be left virtually unattended for a portion of the year while wild resources are pursued. Similarly, since the community does not maintain direct control over the wild animals that are harvested, it is forced to accommodate to the seasonal patterns of movement of the game. The hunters or fishers are thus vulnerable during travel and during the course of their harvesting operations. They are likewise vulnerable to incursions by other groups into their hunting and fishing territories and to raiding of the stores and agricultural facilities that were left behind. Complex systems, involving two or more cultural entities, provide an almost perfect contrast to the strengths and weaknesses of simple systems. Although they are potentially more efficient and are capable of supporting larger populations, complex sys-

The role of wild resources in small-scale agricultural systems

67

Such an expectation does fit with known developmental sequences in regions such as Tehuacan in central Mexico and the American Midwest. At Tehuacan, although domestication has a long history, major dependence on food production appears rather late in the Santa Maria phase (900-450 BC) (MacNeish 1967), and is contemporaneous with the appearance of nuclear villages with central plazas or small mound groups, suggestive of enhanced local integration (MacNeish 1972). This pattern of development mirrors in some important respects the neolithic developments proposed for Crete that led to the formation of the Minoan palace system (cf. Halstead 1981b; Halstead and O'Shea 1982). Similarly in the Midwest, an area that by late prehistoric times was totally dependent on intensive maize cultivation (cf. Ford 1977), serious reliance on food production was relatively late (perhaps by AD 800), even though imported tropical cultigens were known in the region by at least 300 BC, and native domesticates had been cultivated at least since late Archaic times (Asch and Asch 1985). Well before the advent of agricultural dependence, however, significant regional integrative systems had come into being (Adena, Hopewell, etc.). It may be a curious irony that the spectacular Woodland period ceremonial complexes, that were once thought to represent a florescence based on maize agriculture (cf. Griffin 1960), may, in fact, have provided the necessary preconditions for serious reliance on food production.

Although this paper has focused on the use of wild food resources such as fish and bison as alternatives to domestic animals, it has also suggested that perhaps two distinct routes to agricultural dependence may have been followed in the New World; one, resembling the European neolithic pattern of dispersed settlement, that occurred in those regions where dense harvestable wild food resources were available, and a second, slower, pattern characteristic of regions that lacked sufficiently dense wild resources and in which storage and exchange represented the only viable means for covering food shortages. This latter pattern would appear to invert the old idea of food production providing the basis for more complex social forms, since the present model requires the existence of regional integration as a precondition for agricultural dependence. The study concludes on a slightly paradoxical note. On the one hand, it illustrates how little real difference there may be between hunters and farmers, which in some cases may be little more than a storage pit in which hunting tools and agricultural implements are alternatively stored. Indeed, it could be argued that the hunting and fishing described in the present study were pursued according to what was effectively an agricultural logic of intense harvesting and storage, just as early experiments with plant cultivation were conducted according to the logic and organisation of the hunter-gatherer (cf. Flannery 1969). Yet it also stresses the magnitude of the difference between economies that incorporate some agriculture and those that place a major reliance on food production. This may suggest a need to refocus attention, away from a concentration on the origins of agriculture (in which cultivation is easily understood within the logical paradigm of the hunter-gatherer), and toward an understanding of the crucial shift from agriculture as an element in a broad-spectrum economy to food production as the core of the subsistence economy.

Conclusions This study has employed subsistence reliability as a basis for considering the organisation of simple agricultural economies in the New World that developed in the absence of large domestic animals. Based on this consideration, it appears that much of the developmental divergence between the Old and New World, in terms of both its rate and its form, can be traced to the alternative means that were available for buffering variability in agricultural production. Indeed, from this perspective, the diverse array of New World practices and developments can be seen as variations on a single theme.

68

Chapter 5 The economy has a normal surplus: economic stability and social change among early farming communities of Thessaly, Greece Paul Halstead

The subject of surplus may seem out of place in a volume concerned with scarcity, but the two issues are integrally linked. The importance of surplus lies in its problematic relationship with the emergence and maintenance of elites and other non-producing specialists. It is proposed that surplus be viewed as a normal response to the risk of scarcity in certain types of economy and certain types of environment. The changing role of surplus, and the way in which it may be appropriated by an emerging elite, are explored in the particular context of early farming communities in Thessaly, Greece. In conclusion, the Thessalian case is reviewed in a broader context. Surplus: problems of definition

The traditional view of surplus, and of its relationship to the development of social complexity, is that: 'Society persuaded or compelled the farmers to produce a surplus of food over and above their domestic requirements, and by concentrating this surplus used it to support a new urban population of specialised craftsmen, merchants, priests, officials, and clerks' (Childe 1954:30-1). An agricultural surplus was needed to support the full-time specialists who created 'civilisation', and in this sense surplus was a necessary precondition of the development of social complexity. The early civilisations occurred in rather similar natural environments. This suggested that the breakthrough which permitted the creation of surplus, and so of civilisation, was the coincidence in the Nile, Euphrates and Indus valleys of (a) the technological expertise and (b) the environmental opportunity for ir-

rigation agriculture. In Mesopotamia, for example: 'If once the flood waters could be controlled and canalized, the swamps drained, and the arid banks watered, it could be made a Garden of Eden. The soil was so fertile that a hundred-fold return was not impossible ... Here, then, farmers could easily produce a surplus above their domestic needs' (Childe 1954:98-9). Given Childe's view of civilisation as 'progress' (e.g. 1954:292; cf. Dunnell 1980), it was perhaps natural for him to stress the conditions under which farmers were able to produce a surplus. In so doing, however, he came close to treating the ability to produce a surplus as a sufficient precondition of the development of civilisation. A major reappraisal of the relationship between surplus and the development of social complexity was triggered by Pearson in a paper somewhat misleadingly entitled 'The economy has no surplus' (Pearson 1957). He offered a two-fold critique of the traditional view. Firstly he rejected the idea of an absolute surplus - of redundant production above subsistence which, as it were, invites use to support specialists. The minimum biological requirements of food for human beings are rather variable, as are culturally acceptable levels of mortality. Moreover, the notion of absolute surplus assumes 'primacy ... of eating over thinking, socializing, governing, crafting, trading, playing' (Pearson 1957:325). But man does not live by bread alone and cultural perceptions of minimum subsistence levels incorporate varying amounts of other resources. If a minimum subsistence level could not be defined, absolute sur-

The economy has a normal surplus

69

plus could not exist. In this sense, Pearson argued, the economy has no surplus. The second part of Pearson's critique followed from the first. If an absolute surplus could not be defined, either biologically or culturally, surplus had to be seen in a relative sense: 'goods or services would be surplus only if the society ... declared them to be available for a specific purpose'. And whether this purpose was trade, feasting, the support of an elite or whatever, T h e essential point is that relative surpluses are initiated by the society in question'. Moreover, relative surpluses might result from increased production, but 'they may also be created with no change whatsoever in the quantity of subsistence means by re-allocating goods or services from one use to another' (Pearson 1957:323). In short: There are always and everywhere potential surpluses available. What counts is the institutional means for bringing them to life' (Pearson 1957:339). In arguing that the institutional means for mobilising surplus are more critical than the environmental opportunity or technological capacity for producing it, Pearson effectively rejected surplus as a sufficient precondition for the development of social complexity - a stance which has won widespread acceptance as perception of civilisation in terms of progress has receded. Pearson's argument that absolute surplus must be rejected, because of the lack of a biologically defined minimum subsistence level, has rightly been criticised by Harris (1959): problems in the calculation of such a minimum arise, mercifully, from a shortage of data from controlled starvation experiments, rather than because human beings do not starve if fed too little. Paradoxically, however, it is Pearson's contention that the economy has no (absolute) surplus which seems to have made more impact than his observation that the economy always has a (relative) surplus. Sahlins, for example, advances a broadly similar argument (Sahlins 1974). He attacks the false dichotomy between production/or and production beyond subsistence: even in the smallestscale societies, individual households are not autonomously viable - they are dependent on others for mates, peace, food in emergencies, etc. The economy provisions not just individual households, but also society (Sahlins 1974:187n2) and in this sense, again, the economy always has a (relative) surplus. Though Sahlins notes the existence (and indeed necessity) of production beyond immediate, domestic, subsistence needs, however, his emphasis is on underproduction in small-scale societies. The domestic mode of production 'harbors an antisurplus principle. Geared to the production of livelihood, it is endowed with the tendency to come to a halt at that point. Hence if "surplus" is defined as output above the producers' requirements, the household system is not organized for it. Nothing within the structure of production for use pushes it to transcend itself (Sahlins 1974:86). Thus the economy, or at any rate the domestic mode of production, has no surplus, relative or absolute, and so the mobilisation of surplus also requires institutionalised means for ensuring increased production. Appropriately, the onerous task of stimulating intensification and greater production is assigned to emerging leaders. Tribal powers ... encroach upon the domestic sys-

tem to undermine its autonomy, curb its anarchy, and unleash its productivity' (Sahlins 1974:130,140). Sahlins' emphasis on the catalytic effect of emerging leaders has been widely adopted to account for the development of social complexity - and of the 'surplus' which underpinned it (e.g. Friedman and Rowlands 1977; Redman 1978:233; Bender 1978; Friedman 1984). This approach has been elaborated for bronze age Greece by Gamble (1981,1982) and Renfrew (1982). Renfrew's account is the more explicit. He reaffirms the traditional role of surplus as a necessary precondition of social complexity: 'no complex society can function unless the level of subsistence production is sufficient to feed a range of specialists, including the leaders and organisers, in addition to those engaged in food production' (Renfrew 1982:265). Like Pearson, Renfrew apparently opts for a relative rather than absolute interpretation of surplus. The term "surplus" is a dangerous one ... We shall use "production beyond subsistence". .. production may be disaggregated into subsistence production . . . and into social production and trade or cash production, and it is for these latter purposes that production beyond subsistence level... is available' (Renfrew 1982:267-8). Tacitly following Pearson and Sahlins, however, in assuming that the (domestic) economy has no (absolute) surplus, Renfrew concludes that the appearance of social complexity required substantially increased production and, therefore, economic intensification. Such intensification could not be understood in terms of technological or environmental opportunities, as these had remained more or less unchanged in Greece from the beginning of the Bronze Age until this century (Renfrew 1982:275). Rather T h e crux of the matter is incentive, whether voluntary or coercive - the carrot or the stick' (Renfrew 1982:269). The carrot (in this case) was waved by an emerging elite, offering the benefits, material and symbolic, of interaction with other emerging states or 'peer polities' (Renfrew 1982:289). There are thus two strands to Renfrew's argument: (1)

(2)

Production beyond subsistence is a necessary precondition of an elite (the basis of the traditional view as elaborated by Childe); Production beyond subsistence is stimulated by the elite (a variant of the arguments advanced by Pearson and Sahlins).

Although each strand, viewed independently, is plausible, together they make up a circular argument, in which the elite is a precondition of its own existence. The task of the emerging elite would be easier if they could appropriate an existing surplus, rather than having to stimulate its creation, and some have escaped the circular argument by assuming precisely that (Friedman and Rowlands 1977; Giddens 1981:22; Friedman 1984). In fact this assumption is consistent with Pearson's assertion that the economy always has a relative surplus. Sahlins in effect argues the same, though the point is concealed by his emphasis on underproduction in the domestic economy. The key to the problem is present in Sahlins, and to varying degrees in the works of Pearson, Harris and Friedman, but to find

Paul Halstead

70

it we must cut the Gordian knot of Pearson's substantivist approach to surplus. Pearson at once goes too far, and yet not far enough, in his desire to weld the economy indissolubly to society. In denying the possibility of establishing a minimum subsistence level, and so of identifying an absolute surplus, he throws away the analytical tool for recognising that individual social units at times have more or less food than they need just to survive. Then in stressing the importance of relative surplus in, to use Sahlins' term, 'provisioning society', he obscures the vital role of relative surplus, and of the social relations which it maintains, in ensuring bare subsistence in times of shortage. For, as Sahlins makes abundantly clear, a substantial degree of domestic economic failure is characteristic of primitive economy (1974:69) and overproduction within the domestic economy is a prerequisite of the survival of society as a whole (1974:86).

documented archaeologically (Figure 5.1). The small coastal plains, in the southeast of Thessaly, have a Mediterranean climate with mild, wet winters and hot, dry summers. The land-locked lowland interior is dominated by the western Kardhitsa basin (2200 km2) and eastern Larisa basin (1000 km 2 ), where a modified Mediterranean climate prevails with rather colder winters and more rainfall. The lowland soils are mostly formed on fertile alluvium or marl and the early farmers faced a well-wooded landscape (Bottema 1979). The western basin is well watered, with an average rainfall of 600-1000 mm and a dense network of more or less perennial watercourses. The eastern and coastal plains are 'semi-arid', with an average rainfall below 600 mm and perennial water sources restricted to a handful of major rivers, lakes and springs: in particular, much of the southern part of the Larisa basin has less than 500 mm rainfall and only ephemeral seasonal torrents (Sivignon 1975). The distinctive mound settlements of the early farmers, arising from construction in mudbrick or pise and repeated occupation in the same spot, are widely distributed in lowland Thessaly, with the glaring exception of the poorly watered southern Larisa plain. Conditions for the survival and recovery of settlement remains are good in this part of the plain and the absence of early sites must be treated as significant (Halstead 1977, 1984). The early settlements, covering c. 0.5-1.0 ha, were permanent villages of somewhere between 50 and 300 persons. There is no evidence of institutionalised inequality, so the upper limit on the size of these settlements may reflect the inability of egalitarian communities to maintain order and cohesion if they grew any larger (Forge 1972; Chagnon 1979). Dependence on farming is apparent from the overwhelming predominance of remains of domesticates over those of wild plants and animals and from the fact that, in this environment, aggregations of 50-300 people effectively precluded any other form of subsistence (Halstead 1981a). Cereal and pulse crops evidently provided the bulk of the diet (doubtless supplemented by fruits, greens and so on for vitamins, minerals and relishes), and livestock played a subsidiary role: cattle, pigs and goats, which are all well suited to woodland, were far less common than sheep, which are better suited to grazing sprouting crops, stubble and fallow fields. This suggests that stock rearing was largely restricted to the part of the landscape cleared for cultivation and so not very extensive. Moreover, the mortality patterns of early livestock in Greece conform to a 'meat' production strategy (cf. Payne 1973), with peak mortality among juveniles as opposed to infants or adults (Halstead 1987a). This does not mean that meat only was produced, but that management was not geared to maximising the yield of milk (the most productive strategy in terms of energy [Leggel981]). The manure of livestock doubtless helped to maintain the fertility of the arable sector and so contributed to the remarkable longevity of the early villages, some of which were occupied more or less without interruption for centuries or even millennia. Given the modest size of the villages, cultivation was probably on too small a scale to warrant the considerable expense of keeping work animals (Halstead 1987b) and available archaeozoological ev-

Differences in production within any given village are even more critical than output differences between villages. At least no Moalan village seems to be starving, whereas it is apparent that some men do not produce enough food for family needs . . . At the same time, no village . . . appears to have much surfeit, whereas some families are producing more food than they can consume. (Sahlins 1974:69-70) The same point is made even more forcefully by Allan's proposition that: subsistence cultivators, dependent entirely or almost entirely on the produce of their gardens, tend to cultivate an area large enough to ensure the food supply in a season of poor yields. Otherwise the community would be exposed to frequent privation and grave risk of extermination or dispersal by famine, more especially in regions of uncertain and fluctuating rainfall. One would, therefore, expect the production of a 'normal surplus' of food in the average year. (Allan 1965:38) Allan's argument was based on experience of traditional cultivators in East Africa, where normal surplus was used in a range of social, ritual and economic contexts. And the size of the normal surplus could be considerable - Allan estimated that amongst the Tonga between 1936 and 1949 it varied between about 10 percent and 150 percent of subsistence needs and averaged 40 percent (1965:39). Production for and production beyond subsistence are evidently inseparable in more ways than Pearson realised, and any discussion of surplus, relative or absolute, is meaningless unless the temporal and social scales of analysis are specified. The following section evaluates the probable role of normal surplus among early farmers in Thessaly and suggests how it could have been appropriated by an emerging elite. Early farmers in Thessaly c. 6000-4500 be (the Early and Middle Neolithic)1 Farming communities proliferated in Greece from c. 6000 be. The remains of their settlements are most numerous in the northeast mainland and the region of Thessaly is particularly well

The economy has a normal surplus

71

Fig. 5.1. Map of Thessaly, showing the principal alluvial plains and the location of sites mentioned. K: Kardhitsa basin; L: Larisa basin; 1: Prodhromos; 2: Filia; 3: Akhillion; 4: Tsangli; 5: Pirasos; 6: Sesklo; 7: Dhimini; 8: Iolkos; 9: Pevkakia; 10: Visviki; 11: Argissa; 12: Nessonis; 13: Rakhmani; 14: Ay Sofia; 15: Otzaki; 16: Servia

idence suggests that traction or pack animals were not a priority in Greece until the Bronze Age at the earliest (Halstead 1987a). A small-scale horticultural regime probably prevailed, therefore, in which human labour rather than land or plough animals was the limiting factor on crop production. Thessaly has long been renowned as the granary of Greece (e.g. Garnsey, Gallant and Rathbone 1984) and the density of early village sites reflects not only the intensity of archaeological reconnaissance, but also the suitability of the region for a mixed farming economy. Nonetheless, the natural and social environments of early Thessalian farmers posed problems on a variety of time scales, which may conveniently be considered in terms of seasonal, interannual and long-term variability. Seasonal variability The climate of the Thessalian plains, particularly inland, is markedly seasonal, and winter cold and summer heat are amongst the most severe in lowland Greece. On the other hand, seasonal variability is very predictable and many of the mechanisms for coping with it were permanent, integral features of early farming culture. Mudbrick houses, for example, provide remarkably effective insulation against extremes of temperature and, though surface water is fairly scarce in summer, the early farmers colonized those parts of the plains endowed with more or less permanent watercourses.

The seasonal harvest of cereal and pulse crops could be stored for the rest of the year. Storage took place indoors, probably in pots at Sesklo (Theokharis 1976) and in a clay- or dunglined basket in House T at Tsangli (Wace and Thompson 1912), while at Servia in neighbouring Macedonia the grain itself has survived through carbonisation in a house fire (Ridley and Wardle 1979:202). In the case of the glume wheats, emmer (Triticum dicoccum) and einkorn (7". monococcum), the reliability of storage may have been enhanced by storing the crop as spikelets, because in this partly threshed state the enclosing glumes provide the grain with some protection against fungal and insect attack (Hillman 1981). This practice (now widely documented in prehistoric Greece - Jones 1987; also Housley 1981) may help to account for the prevalence of the glume wheats over free threshing bread (or macaroni) wheat (T. aestivum (durum)), which is easier to process for consumption but more vulnerable in storage. The seasonal climate makes it difficult to grow staple crops for storage in sufficient quantity. To some extent the time available during winter and spring, for preparing the ground and sowing, and in early summer, for harvesting, can be extended by growing a range of different crops, and early farmers evidently took advantage of this. Emmer, pea (Pisum sativum), grass pea (Lathyrus sativus) and bitter vetch (Vicia ervilia) dominate individual samples from Sesklo (Tsountas 1908:359) and Prodhromos (Halstead and Jones 1980) and were probably cultivated as crops in their

Paul Halstead own right. Einkorn, bread/macaroni wheat, six-row barley {Hordeum vulgare), common millet (Panicum miliaceum), lentil {Lens culinaris) and chick pea (Cicer arietinum) have also been found (e.g. Kroll 1983) and, to judge from finds elsewhere in Greece, several of these may also have been cultivated. Tasks such as weeding and herding could be undertaken by children and old people not fit for heavier work (e.g. du Boulay 1974), but inevitably the seasonal nature of agriculture created a heavy demand for labour at certain critical periods (e.g. Delille 1977:120) and of course the labour of tilling and sowing produced no return until the end of the growing season. Such acute and highly seasonal demands on labour place the stable family household at a substantial advantage as the basic unit of production and consumption over the looser groupings commonly found among hunter-gatherers and tropical cultivators (Flannery 1972). The rectangular dwellings of early farmers in Thessaly are of a size suggesting that the family household was the normal residential unit (Halstead 1984) and at fifth-millennium be Sesklo the ubiquitous distribution of coarse pottery suggests that storage did indeed take place at the domestic level (Kotsakis 1981). Early farming households were well suited to meeting their considerable seasonal labour requirements under normal or average conditions, but significant increases in production targets, loss of labour or decreases in yields could not easily be absorbed. Interannual variability

In fact production targets, labour supply and crop yields must all have varied considerably from year to year. Some of the variability in production targets and labour supply will have been fairly predictable and so relatively easy to cope with. Periodically production requirements might be raised in anticipation of feasts celebrating major rites of passage (van Gennep 1909), and houses would need rebuilding every generation or so. Such predictable demands on production or labour can often be scheduled to conform with the year-to-year vicissitudes of farming - a bumper harvest is an opportunity to celebrate a rite of passage (MacCormack 1978) or organise a 'beer party' for friends and neighbours helping in a major project like house building (Allan 1965:44-5). Also, with the passage of a generation, the ratio of consumers to workers within a household changes (and with it the output required of each worker), as growing children expand first the level of consumption and then the size of the labour force (Chayanov 1966; Sahlins 1974:87-92). Superimposed on these gradual changes, unpredictable losses of labour, through illness or injury, and of time available for work in the fields, because of unusually severe winter conditions, could further exacerbate the seasonal labour crisis. Given the range of tasks which can be carried out by young children in a small-scale horticultural regime, parents who raise a large family are somewhat insured against shortage of labour in the future (cf. Athens 1977:366). In fact the close spacing and longevity of early farming villages implies that, whether farming was introduced by a (presumably) modest number of immigrants or adopted by a sparse indigenous population of hunter-gatherers, substantial demographic growth did take place from an early stage.

72 Of course the main source of uncertainty on the interannual scale would have been variation in crop yields. Sown crops could, at various stages in their growth cycle, be damaged by late frosts, drought, hail storms, desiccating 'livas' winds, insect pests, blight or marauding deer (e.g. Khalikiopoulos 1905:452). The impact of such hazards would have been lessened by the documented practice of growing a range of crops with different susceptibilities and differently timed growth cycles. In parts of Greece today, mixed 'maslin' crops are also grown as an insurance against weather inimical to only one of the species and there is some evidence for this practice from bronze age Macedonia (Jones, Wardle, Halstead and Wardle 1986). Suggested early maslin samples from Thessaly (Renfrew 1972), however, may simply reflect postdepositional mixing (Renfrew 1966:26; Dennell 1974). The dispersal of individual garden plots could have provided some protection against crop diseases, pests and perhaps local storm damage (cf. Christodoulou 1959:86), but rather less protection against climatic hazards like drought than in the topographically heterogeneous Methana peninsula in southern Greece (Forbes, this volume). The timing and volume of water flow in the surface drainage network of Thessaly is too unpredictable to support regular irrigation without modern technology, but in drought years the water level recedes earlier than usual and late crops have been sown on this naturally watered ground in recent times (Leake 1967:424). Moreover, a few perennial springs offer a more manageable, if modest, supply of water, which is nowadays used almost exclusively to irrigate fruit trees and vegetables, but has been diverted within living memory to save cereal crops from drought (cf. Arnon 1972:53; Hillman 1973:231). Given the small scale of neolithic cultivation, such emergency irrigation could have played an important role in localities with a suitable water source. Such preventive measures would have provided protection against many potential sources of crop loss, but from time to time a shortage of cereal and pulse staples must have arisen because of events like the destruction of stored crops by fire at Servia or because of unusually severe weather conditions. During the 1970s, the coastal plains of Thessaly experienced three winters with growing-season rainfall (October-May) at or below the 300 mm regarded as the minimum for a successful wheat crop (Arnon 1972:4, 7, 55), followed by a severe drought {c. 150 mm rainfall). In the years of low rainfall, poor yields of wheat and barley were partly offset by a reasonable lentil crop (Figure 5.2). In the drought year, however, all three crops performed badly and diversification offered no protection against shortage. Early Thessalian farmers could have used a variety of further mechanisms to avoid such shortages. At sites such as Argissa (Boessneck 1962), Sesklo (Renfrew 1966; Kroll 1983) and Prodhromos (Halstead and Jones 1980), there are remains of a variety of wild mammals (including deer, boar, aurochs and hare), birds, fish, molluscs, nuts and fruits, but collectively these make up only a very minor part of the archaeozoological and archaeobotanical assemblages. On a small scale, farmers could switch to hunting and gathering and this doubtless saw individual households through brief periods of hardship. For the most part, however, wild resources in Thessaly were neither

73

The economy has a normal surplus

•

40-i

-40-1 1969/70 1970/1*

1971/2

1972/3* 1973/4

1974/5* 1975/6

1976/7** 1977/8

Fig. 5.2. Yields of three field crops in coastal Thessaly, 1969/701977/8 (controlling for long-term improvement by regression). wheat; barley; lentil; *low rainfall (growingseason precipitation 250-300 mm); **drought (growing-season precipitation 150 mm). Source: crop yields from the annual Agricultural Statistics of Greece of the National Statistical Service of Greece and from the Greek Ministry of Agriculture; precipitation data from the Agrarian Bank of Greece and Meteorological Service of Greece.

abundant nor concentrated and, with the possible exception of fishing and fowling from sites near the River Pinios and Lake Karla, their pursuit would have entailed a level of mobility which conflicted with cultivation of future crops. Another source of diversity was livestock and three aspects of husbandry enhanced the reliability of domestic animals as a source of food in the event of crop failure. Firstly a mixture of cattle, sheep, goats and pigs was maintained, which reduced the risk of losing the entire herd through disease or shortage of fodder. Secondly all four species were reared to a meat production strategy, which creates a more resilient herd structure than dairying (Ellis, Jennings and Swift 1979; Redding 1981), because breeding females, young livestock and milk yields are very vulnerable to disease and to the sort of climatic extremes which cause crop failure (e.g. Haresign 1983). Thirdly sheep, the commonest animal, are a particularly valuable 'walking bank' of food, because of the way they store fat (Halstead 1987a). Their dominance in this wooded environment, however, suggests that livestock were few in number (above, p. 70). Arguably the most effective household-level mechanism for coping with crop failure was to store surplus from good harvests for use in bad years and in theory a household which aimed merely to break even in an average year would find that good and bad harvests balanced out in the long run. In practice, of course, the storage life of grain with traditional techniques is too short (Will

and Hyde 1917:138; White 1970:189) for farmers to secure themselves against repeated crop failure simply by direct, long-term storage, so the prudent farmer sows enough to break even in the event of at least the more usual and expectable levels of crop loss. Given the paucity of alternative food sources, the remarkable longevity of many early farming settlements is inexplicable without such overproduction and direct storage of the resulting 'normal surplus'. This strategy would secure a household against most yearto-year fluctuations in the size of the harvest, but its potential must still have been limited, because of the unreliability of direct storage and inevitable drudgery of producing surplus food which the household was normally unable to consume. Some surplus grain could be used to fatten up domestic livestock, which could be slaughtered in time of dearth - in effect a form of indirect storage (cf. Flannery 1969:87) - but sooner or later shortages would be faced which could not be coped with at the level of the individual household. A key member of the work force might be permanently lost through premature death, a failed harvest might occur at a stage in the domestic cycle when the ratio of producers to consumers was particularly unfavourable, or a succession of poor crops might deplete all stores. It is for this reason that, as Sahlins stresses (1974:101), the self-sufficient household of the 'domestic mode of production' is dividual households can only be economically self-sufficient in the short term (Sahlins 1974:69-74). In situations where shortage is not universal, a potentially powerful coping mechanism is exchange of surplus food between households. There are abundant ethnographic examples of such exchange, which takes a variety of cultural guises ranging from simple sharing to the provision of feasts in formal social contexts, and from hospitality in the context of'casual' visiting by needy neighbours or relatives to more or less blatant exchanges of food in return for goods or services (e.g. Richards 1939:142-7; Allan 1965:44-5; Sahlins 1974:123-48, 219; O'Shea 1981). Sahlins has argued that such social obligations are the driving force behind overproduction in 'primitive' economies (1974:101-48), but normal surplus produced by households aiming only for self-sufficiency would offer considerable scope for exchange among households. The potential for such exchange among early farmers in Thessaly is not dependent on the assumption of a socially prescribed goal of overproduction. In Thessaly, food must have been exchanged between neighbouring households to maintain the cohesion and survival of the early farming villages. In the absence of institutionalised authority, kinship ties form the basis of village solidarity and kinship is widely affirmed by the sharing or giving of food (Sahlins 1974:125-6). Given the long-term non-viability of individual households, the persistence of the village as the normal form of early farming settlement surely indicates that neighbouring households rendered mutual assistance in times of subsistence crisis. Village settlement was maintained in spite of its considerable disadvantages to individual households for both normal and emergency food production, but it did facilitate cooperation between households in adversity. Early farming households were also dependent on links with

Paul Halstead

74

those outside their own village. Among the Bemba,

munities today in southeast Asia (Forge 1972) and South America (Chagnon 1968), where raiding is endemic. In the latter, tropical areas the natural environment is quite stable from year to year, whereas in Thessaly climatic hazards to agriculture were a serious threat and a powerful incentive to the suppression of hostility. Again, the proven ability of many early farming communities in Thessaly to survive the rare but inevitable incidences in which severe crop failure afflicted the whole village indicates that, whatever their overt motivation, exchange relationships between inhabitants of distant villages must on occasion have served as a vehicle for the provision of much-needed subsistence relief. Thus exchanges of food must have taken place over a range of social and geographical scales and the nature of the exchanges is likely to have changed with increasing social distance (Sahlins 1974:185-230). In early Thessalian villages, cooking facilities were usually located in open yards between the houses (Theokharis 1980; e.g. Otzaki - Milojcic 1955:168, 1971; Sesklo-Theokharis 1971; Akhillion - Gimbutas 1974; also Servia - Ridley and Wardle 1979), an arrangement which would have facilitated and, arguably, forced sharing of cooked food at least among neighbouring households (cf. Richards 1939:121; also Whitelaw 1983). Cooked food is widely subject to particularly strong obligations to share, where uncooked food remains uncontested private property (e.g. Sahlins 1974:125), and it is surely significant that the most elegant part of the early ceramic repertoire was that suited to the serving and consumption of food and drink (e.g. Theokharis 1973 Plate 33). Between close kin or neighbours, food might be given without particular concern for reciprocation or in the confident expectation that help would eventually be reciprocated. Over greater social distances, however, food might be exchanged for more immediate reciprocation in labour or in the form of valuable items which could themselves ultimately be converted back to food. The provision of food in return for labour is widely documented in the ethnographic literature - the example of the African 'beer party' has already been noted - and is likely to have been the norm where distance forced the recipient household to reside with the donor, rather than taking food home (cf. Lightfoot 1979). Given the drudgery of agriculture in a seasonal environment, such an arrangement has obvious attraction to the donor household, but threatens the recipients with permanent destitution if they are unable to fulfil their obligations to their hosts and cultivate their own gardens. Alternatively, hospitality could be repaid with livestock, provided the 'exchange rate' operated to the nutritional advantage of the hungry household, but again the loss of livestock needed to fertilise the fields could compromise the long-term viability of the household. A third solution, particularly attractive where the distance between donor and recipient household was not so great as to rule out the movement of food rather than consumers, would be to avoid immediate reciprocation in labour or kind. O'Shea (1981; also Flannery 1974) has argued that such an arrangement may be facilitated between socially or geographically distant partners by the exchange of food for valuables: valued 'tokens' are received in the expectation that in the future the donor or a third party will be prepared to accept them in return for the provision of food (or

kinship obligations result in quite a considerable distribution of food. If a man's crops are destroyed . . . relatives in his own village may be able to help him by giving him baskets of grain or offering him a share in their meals. But if the whole community has been visited by the same affliction . . . the householder will move himself and his family to live with other kinsmen in an area where food is less scarce. (Richards 1939:108-9) In eastern Thessaly, drought years often come in clusters: on the coast precipitation fell below 300 mm in three out of four growing seasons between 1931-2 and 1934-5 and in the Larisa plain failed to reach this level 11 times (including twice below 150 mm) in the 14 growing seasons between 1964-5 and 1977-8. Under such circumstances, even if complete crop failure is averted, stores in the worst-hit villages are likely to drop to the point where neighbouring households cease to help each other and retreat instead into isolation (Sahlins 1974:128-9). Such is the topographic uniformity of lowland Thessaly, however, that farmers forced to look outside their own village for help might well have to look very far indeed over distances of 50-75 km (Halstead 1981b; Garnsey et al. 1984). Given the density of settlement in Thessaly, marriage partners could have been secured within the immediate vicinity of most villages (Goody 1976:31), but marriage alliances may have been sought further afield for other reasons. Obsidian from southern Greece reached Thessaly, but local stone seems to have been used for basic necessities such as cutting and grinding equipment (e.g. Wijnen 1982) and so the acquisition of exotic raw materials should perhaps be seen more as a reflection than a sufficient cause of contact between distant villages. Such contact is also documented in the widespread distribution of fine pottery styles, arising largely from local manufacture of similar vessels, though a few pieces were apparently exchanged (e.g. Wace and Thompson 1912:241). The contrast between the ubiquitous distribution of some fine wares and the more restricted distribution of others (Rondiris 1981; Halstead 1984) suggests that the latter at least were actively signalling long-distance (up to 50-75 km) interactions, or chains of interactions, rather than passively reflecting opportunities for acculturation. Similarly, the small corpus of published stamp seals and sealings from Thessaly includes four examples with an identical design, from Nessonis in northeast, Pirasos in southeast, Tsangli in central and Filia in southwest Thessaly (Yiannopoulos 1913; Matz 1933; Theokharis 1959:64, Fig. 28, 1973:311, Fig. 211a, 333 Plate 272b-c). Stamp seals were often pierced, as if for wearing about the person, and so may symbolise the maintenance of distant social ties on an individual basis. Trade or marriage partnerships between distant households and villages are literally a vital element in the primitive economy, in that such exchanges are a safeguard against, and indeed the alternative to, open hostilities (Mauss 1970:79-81; Sahlins 1974:302). The longevity and close spacing of early farming settlements in Thessaly suggest that such a peaceful accommodation was normally maintained (cf. Vayda 1979:210-12) - in contrast with the mobility and dispersal of similarly sized farming com-

The economy has a normal surplus

75

labour). In effect this cultural mechanism allows food to be committed to 'social storage' (O'Shea 1981; Halstead and O'Shea 1982) and later recouped on a time scale longer than that ensured by direct storage. Fine pots, some of which were exchanged over substantial distances, may have served as such social storage tokens: considerable skill and time (of the order of 10-20 hours for experimental reproduction of an unexceptional piece - M. Nikolarakis personal communication) were invested in their manufacture. The exchange of food for tokens is a contentious issue, both for anthropologists and for those they study (O'Shea 1981), and frequent cultural proscriptions against such transactions underline the vital importance of food in times of dearth: 'Food has too much social value - ultimately because it has too much use value to have exchange value' (Sahlins 1974:218). Such normative proscriptions lose their weight, however, with increasing social distance (Sahlins 1974:219) and also tend to be 'abandoned' in times of stress (O'Shea 1981:177; cf. Sahlins 1974:129). The importance of such material tokens is that, in times of extreme need, they allow exchanges of food to be made over greater temporal, social and spatial distances than would otherwise be possible. Early Thessalian farmers mustered an impressive array of mechanisms to cope with interannual variability in their food supply. At the level of the individual household, a diversified strategy of crop husbandry reduced the risk of crop failure, storage of normal surplus took care of bad years which followed good ones and diversification to make use of livestock and wild resources would cater for short-term shortages of staple crops. If the buffering potential of such mechanisms was exhausted, exchanges between households might be resorted to. The sharing of normal surplus among closely related or neighbouring households would facilitate future reciprocation, but more distant partners offered the only hope of surviving those rare failures which were both severe and widespread. Distant partnerships would be inherently unstable, however, because they would not be easy to maintain through inevitable periods of redundancy, because hospitality in a distant village posed a serious obstacle to raising future crops at home and because such distant relationships would be very prone to imbalance. Western Thessaly is far less vulnerable to widespread crop failure through drought than the semi-arid eastern plains, so reciprocation of west Thessalian hospitality would entail a net flow of mobile or portable resources from the east - either of livestock and labour, which would undermine food production in the east, or of material tokens, which would gradually lose their value through inflation. Moreover, in the long term the growth of population, itself perhaps a short-term solution to periodic labour crises, threatened to undermine the effectiveness of lower-level coping mechanisms.

were changes in the social environment - most obviously in the size, distribution and organisation of the human population. During the late fifth and early fourth millennia be new villages were established between the earlier foundations, while the southern Larisa plain was belatedly colonised by a rash of shortlived hamlets. At the same time there are indications at the east Thessalian villages of Ayia Sofia (Milojcic 1976a), Sesklo, Dhimi ni and perhaps Visviki (Theokharis 1973) of the emergence of institutionalised inequality. In each case this is projected in a large 'megaron' house set in a central courtyard and at Ayia Sofia associated with a unique minority burial mound. The megaron elite e\ idently made it possible for communities to grow beyond the limii imposed by egalitarian organisation because, during the later fourth millennium, most of the hamlets and many of the villages were abandoned and the human population aggregated into a smaller number of settlements. The largest of these, covering up U seven ha during the third millennium and up to 25 ha by the end o the second millennium be, served as central places for their smalle neighbours (Halstead 1984). Further consideration of the problems of food production, in normal and abnormal years, sheds light on both the causes and consequences of these radical change in the social environment.

Long-term variability In addition to seasonal and interannual variability, early Thessalian farmers were faced with longer-term shifts in their environment. Secular climatic changes doubtless occurred, but will not have posed such problems as in more marginal areas of human habitation (e.g. Mine and Smith, this volume). Far more drastic

Change in the social environment c. 4500-3750 be (the Late Neolithic) The infilling of the landscape fuelled by continuing population growth undermined earlier household-level mechanisms for coping with interannual variability. Crop diversification may no\* have been enhanced by cultivation of bread wheat, barley and len tils (Tsountas 1908:359; Renfrew 1966; Kroll 1979), in addition to earlier known crops, and some evidence for increased importance of cattle, pigs and goats, at the expense of sheep, may indicate greater use of woodland (Halstead 1981a; but cf. Payne 1985:222 • The increasing density of settlement, however, will have further rt duced the subsistence potential of sparse wild resources and the colonisation of the southern Larisa plain brought farming to an area largely devoid of watercourses with potential for fishing, fowling, emergency irrigation or summer pasture in years of drought. The potential for direct storage of normal surplus was unchanged, but the evidence for storage is far richer: coarse pottery suitable for bulk storage is now much more abundant (Hourmouziadhis 1979:116; Kotsakis 1983:214n33), reinvestigation of the settlement at Dhimini has produced a plethora of built indoor storage facilities (Hourmouziadhis 1979) and elsewhere numerous deep pits (e.g. Milojcic 1976a) attest to outdoor storage as well. Most of these pits seem excessively large (where dimensions are re ported) for the storage of just seed corn (cf. Reynolds 1974) and such pits are not a practicable means for a household to store its annual supply of grain for consumption. Subterranean storage is effective, provided the pit is tightly sealed, because carbon dioxide given off during spoilage of the outermost 'skin' of grain protects the remainder of the contents. As a result, small pits (holding, say a week's or a month's rations for a household) entail disproportionately high losses from spoilage of the ouier skin, while large pits cannot be opened at intervals through the year without jeo-

Paul Halstead pardising their entire contents (Reynolds 1974). It follows that these pits represent storage of grain surplus to immediate requirements and additional to that noted in earlier periods. This development may be related to contemporary changes in the pattern of exchange. In the parts of lowland Thessaly favoured by the earliest farmers, individual households faced a variety of hazards which could be overcome by seeking help from more fortunate neighbours, but in the arid southern Larisa plain the main hazard was drought, against which sharing between co-villagers offered little protection. Partly for this reason perhaps, this area was colonised by small hamlets. These offered less scope than villages for internal sharing, but did allow easier cultivation closer to home and greater opportunities for reliance in emergency on the sparse wild resources available. Even within the village settlements sited in more favoured areas, however, sharing may now have been subject to greater constraints. Again Hourmouziadhis' (1979) reinvestigation of Dhimini is instructive: most cooking facilities are located indoors and the settlement is partitioned by stone walls so that, architecturally, sharing was only encouraged within small clusters of buildings corresponding perhaps to an extended family household or a 'courtyard group' (Flannery 1976:75). A similarly subdivided settlement plan is suggested by more limited excavations at other east Thessalian sites. The declining effectiveness of household-level coping mechanisms and of mutual assistance between near neighbours would have enhanced the importance of exchanges with more distant households. Conversely, the scope for such exchanges would have been increased by the progressive infilling of the east Thessalian landscape, which brought larger numbers of potential producers and, equally important, consumers of normal surplus within easy reach of one another. In effect this meant more opportunities for 'banking' normal surplus and, as long as surplus banked in this way could ultimately be recouped in kind or in services, this would favour increased overproduction and so further enhance the effectiveness of the social storage system. The increased evidence for storage, and in particular for subterranean storage, suggests that substantially higher levels of overproduction did take place - at least in east Thessaly. With the increasing capacity of the social storage system, medium-distance exchanges within eastern Thessaly would have become more viable alternatives to imbalanced, long-distance exchanges between east and west and by 4000 be west Thessaly had lost touch with the rapidly changing ceramic fashions of the east and the distribution of the more elaborate fine wares was strongly restricted to the latter area (Halstead 1984). Exchanges within east Thessaly must also have been subject to some imbalance. Some households, by virtue of misfortune or mismanagement, and some villages, by virtue of location, would have been short of food more often than others. The use of material tokens would absorb short-term imbalance in exchanges, but in the medium term tokens would be distributed increasingly unevenly, both between villages and - if tenure of fields was vested in particular descent groups - within villages. Fixed tenure of plots at least inside the settlement is suggested (Kotsakis 1986) at fifth-

76 millennium Otzaki (Milojcic 1960:12-13) and Sesklo (Kotsakis 1986) and fourth-millennium Dhimini (House 18 - Hourmouziadhis 1979:149), where houses were repeatedly built on the same spot. The same may well have applied to the ownership of fields and markedly unequal access to fine pottery is apparent by the mid-fifth millennium be within the settlement at Sesklo (Kotsakis 1981, 1983, 1986). As more and more tokens were manufactured by households in need of food, their value would decrease (Sahlins 1974:295-7, 311-12;O'Shea 1981). Households with a particularly unreliable food supply could counter this trend towards 'depreciation' by investing more labour and using scarcer raw materials in the manufacture of craft goods and both tactics were employed in east Thessaly - more elaborately decorated pottery vessels were produced and fine shell bracelets were traded inland from the coast (Theokharis 1973; Halstead 1984). In the long term, however, the greater value of the tokens in circulation simply enhanced the ability of the more successful farmers to accumulate wealth and, if the social storage system was not to collapse from an inflationary spiral, these farmers ultimately had to be able to redeem tokens for food, labour or personnel (O'Shea 1981:178-9). Two consequences of this development are apparent in the archaeological record. Firstly, villages in unfavourable locations risked crippling temporary losses of labour or permanent losses of personnel and many later foundations, especially in the marginal southern Larisa plain, were short-lived. Secondly, by acquiring rights to additional labour, the most successful households could further enhance their productivity and would be better able to maintain exchange partnerships in other villages. In consequence, they would be better placed both to consign normal surplus to social storage and to recoup food when needed and so could act as 'social storage agents' for their less successful neighbours. At Sesklo (Theokharis 1973) and Dhimini (Hourmouziadhis 1979), the number of buildings in the central court suggests preferential access to labour and/or produce, and the fact that the development of institutionalised social inequality was materially projected at different sites in the same way (a central megaron) underlines the importance for the emerging elite of contacts with similar households in other villages. The well-being of ordinary households was now highly dependent on access to a successful elite household and this, coupled with the ability of elite authority to suppress conflict in large communities, accounts for the subsequent aggregation of the population of east Thessaly into fewer and larger settlements. (In west Thessaly there is as yet no evidence of a megaron elite and only modest settlement aggregation is documented.) This shift in the pattern of settlement in turn had radical implications for the scale and nature of land use, for the range of mechanisms available for coping with subsistence crises and for the organisation of society. Change in the social environment c. 3750 be-1200 BC (the Final Neolithic and Bronze Age) The growth of large settlements which began in the fourth millennium bc would have forced many households to farm further from home and so to adopt more extensive techniques of husbandry. Any consequent reduction in the diversity of crops plan-

The economy has a normal surplus

77

ted, in the intensity of tillage, manuring and weeding (Halstead 1981a) or in the opportunities for emergency irrigation would have increased both the difficulty of producing a storable reserve in good years and the risk of crop failure in bad years. Similarly, the increasing difficulty under such an extensive regime of herding small numbers of livestock would have favoured communal herding (such as was customary in the region in recent times - Sivignon 1975:181-2) and so perhaps weakened household control of the livestock 'bank'. In short, settlement nucleation must have undermined existing household-level coping mechanisms and so caused yet further dependence on surplus controlled by the emergent elite. Increasing dependence on centrally controlled surplus is reflected in the apparent collapse of earlier mechanisms for exchange between ordinary households. From the later fourth and third millennia be onwards, cooking facilities were either placed indoors, in the 'kitchen' which formed a distinct part of many dwellings, or were set outdoors in a closed yard (e.g. Argissa Milojcic 1976b; cf. Halstead 1980; Pevkakia - Milojcic 1972 Plan 2). The preparation and consumption of food were now entirely private, thus actively discouraging sharing between neighbours. At the same time, the production of fine, decorated pottery ceased, while ornaments and simple tools appeared made of copper - a raw material which was at once rare and, probably, imported and so far more amenable to elite control than potting clay (cf. Hourmouziadhis 1979:95-6, 1980:120). The place of decorated pottery may also partly have been taken by woollen textiles. A cache of 22 spindlewhorls in a final neolithic house at Rakhmani (Wace and Thompson 1912) hints at large-scale, centralised production of yarn (J. Carrington-Smith personal communication), while at bronze age Pevkakia a substantial improvement in the survivorship of male sheep suggests a greater emphasis on wool. Moreover, this change in male survivorship is apparently not matched by a change in overall age structure, implying that more female sheep were killed young - perhaps because of the need for a smaller proportion of breeding stock in larger, communally herded flocks (Halstead 1987a). Like copper, therefore, woollen textiles may have been amenable to centralised control, so the demise of decorated pottery may mean that the elite took control of the production of the fine craft goods used as tokens in social storage transactions.

In fact these indications of diversification at the regional level may have conferred little advantage at the household level. Firstly, artistic evidence from southern Greece hints that hunting may have been primarily an elite sport (Vickery 1936) and, for ordinary households, residence in large settlements would have made the pursuit of wild resources even less reliable than previously as an emergency source of food. Secondly, households farming distant fields would have been ill placed to exploit the full range of available annual crops - particularly the labour-intensive pulses (cf. Halstead 1981a, 1987b) and the more demanding of the cereals. Thirdly, although walnut and chestnut have very different growth requirements from annual crops and so should be subject to quite independent fluctuations in yields, in recent times they have flourished only on some of the mountain slopes encircling the Thessalian lowlands. Tree crops may have been accessible, therefore, to only a few settlements on the margins of Thessaly. The use of horses and donkeys as pack animals could partly have offset the difficulties of increasingly extensive agriculture and may have been accompanied by other developments in transport and cultivation technology. During the third and second millennia be, oxen may have been raised at Pevkakia (Halstead 1987a, after Jordan 1975; Amberger 1979) and there is widespread evidence from other parts of Europe for ard ploughs and wheeled carts (Sherratt 1981). Work animals are costly to maintain, however, and carts at least are the product of skilled craftsmanship, so these technological advances may only have been available to a minority with access to abundant fodder and the services of craft specialists (Halstead 1987b; also Goody 1976; Sherratt 1981). With the development of extensive agriculture, therefore, access to expensive capital items and to scarce areas of good land near the settlement would increasingly have replaced human labour as the limiting factor on household productivity and so led to further disparity between elite and ordinary households. The pressures favouring aggregation were thus reinforced and indeed, in eastern Thessaly, a number of nucleated settlements continued to grow in size through the third and second millennia be (Halstead 1977, 1984), so forcing ordinary households into a vicious circle of increasing dependence on the elite. By the late second millennium BC, there are signs of a three-tier settlement hierarchy, though not of regional polities as extensive as those of southern Greece. At the same time, elite control of specialised craft production is reflected in the association of workshops with a 'palatial' complex at the coastal centre of Iolkos (Theokharis 1961) - an association which, on a grander scale, is characteristic of the contemporary palaces of southern Greece (Keramopoullos 1930; Finley 1957; Alexiou 1961; Iakovidhis 1977; Killen 1984, 1985; Shelmerdine 1985). From the southern palaces, there is archival evidence that many craftsmen were full-time specialists (e.g. Killen 1984:52), collectively constituting a substantial drain on food amassed by the elite. At least some of this food, however, may have been normal surplus acquired in exchange for palace craft products (Halstead 1988), rather than additional surplus extracted by taxation. In Thessaly too, the erosion of household autonomy during the Bronze Age was presumably matched by opportunities for the

Settlement growth is paralleled by indications of continuing diversification in the subsistence economy: the remains of wild animals are far more abundant than previously (Boessneck 1962; Jordan 1975; Amberger 1979; Hinz 1979); einkorn, broomcorn millet and Celtic bean are added to the range of annual crops attested (Tsountas 1908:360; Renfrew 1966; Kroll 1983); there is widespread palynological evidence from other parts of northern Greece that the chestnut and walnut were introduced or brought into cultivation (e.g. Bottema 1980); and the range of domestic livestock was enlarged to include the horse and donkey (Boessneck 1962; Jordan 1975). The increase in hunting does not take place until well into the third millennium be (later Early Bronze Age), however, and the evidence for all the new domesticates come from the second millennium be (Middle and Late Bronze Ages) too late to reverse the erosion of household independence.

Paul Halstead elite to extract surplus by coercion, but the costs of maintaining an elite could have been offset at least in part by the economies of scale possible with centralised redistribution of normal surplus. The principal burden imposed by the elite on the labour of subordinate households, therefore, may not have been taxation, but the need to live in large settlements which reduced the efficiency of farming. Economic stability and social change in Thessaly, 6000 DC-1200BC

During the period 6000 be-1200 BC, farming communities in Thessaly experienced radical social and economic change. In attempting to identify and explain these changes, the preceding discussion illustrates three general points raised in the introduction to this volume: (1)

(2)

(3)

at an empirical level: the major problems of survival and reproduction faced by past human communities in a given context can be identified; at a heuristic level: a range of viable solutions to these problems can be specified, allowing a number of past cultural traits to be recognised as buffering mechanisms, the effective use of which is demonstrated by the longevity of many of the early farming settlements; at a theoretical level: a choice between current polarised stances is neither necessary nor helpful - the problems facing prehistoric Thessalians arose from the interaction between culture and nature and were countered by a combination of economic and social behaviour.

More specifically, settlement in permanent villages ensured that cultivated cereals and pulses were the only viable staple foods and so enforced dependence on storage as a response to the highly seasonal climate of Thessaly. This in turn created seasonal labour crises which favoured the family household as the basic unit of production and consumption. The agricultural economy was also vulnerable to variation from year to year in demand, labour and yields. The central role of storage in coping with seasonal variability, coupled with delayed returns from the cereal and pulse crops, strictly limited the potential of mobility in buffering against interannual scarcity. Diversification is well documented throughout the period: the cultivation of a range of species reduced the risk of crop failure, while livestock and wild resources were alternative foods in the case of short-term or very local shortage of staples. Total crop failure or repeated poor harvests, however, required the use of'normal surplus' stored up during good years - either within the household or, in more severe cases, through exchange with other households. The barriers to such exchanges posed by the emphasis on the family as the basic domestic unit were transcended by the use of craft goods as valuable tokens. Finally, the need to ensure a viable household labour force in the face of uncertain life expectancy may have encouraged substantial population growth. Whatever the reason, the early farming population did grow and colonised parts of east Thessaly increasingly marginal for agriculture and so subject to greater risk of crop failure. More-

78 over, as losses afflicting individual households were outweighed by more pervasive losses due to drought, the potential benefit of sharing with near neighbours was reduced and the importance of more distant exchanges enhanced. Many of the new foundations were small hamlets, while increasing investment in fine craft goods and widespread provision for bulk storage in pits suggest that the isolating effects of dispersed settlement were offset by greater exchange. Although the expansion of settlement enhanced the opportunities to commit and recoup surplus through social storage, it also heightened the vulnerability of some farmers to agricultural failure. At the same time, the growing value of craft tokens in circulation allowed successful farmers to accumulate durable wealth and then to gain rights to the labour and/or produce of those less fortunate. By c. 4000 bc, the central 'megaron' signals the emergence in east Thessaly of an elite able to dominate social storage transactions with other communities. Thereafter the attraction of population to successful elites led to the continuing growth of larger settlements, which in turn necessitated more extensive agriculture and so undermined existing household-level buffering mechanisms. The use of exotic or scarce raw materials (metal, wool) in craft production and the growing importance of'capital' items (oxen, pack animals, carts) in extensive agriculture reinforced the grip of the elite. Thus once imbalance in social storage transactions in east Thessaly had initiated the process of settlement aggregation, ordinary households were trapped into increasing dependence on the elite. An initially ranked society, in which elite status was related to success in agriculture and social storage, was rapidly transformed into a stratified society, in which the elite had preferential access both to the land, labour and capital equipment necessary for agriculture and to the craft goods used in social and economic intercourse. Moreover, the internal colonisation of east Thessaly, which triggered these developments, was itself made necessary by population growth and made possible by a regional social storage system which enabled vulnerable communities to survive temporary dearth. Population growth and social storage arose from attempts by early farming households to buffer themselves against the inevitable, but unpredictable, year-to-year vicissitudes of the agricultural economy. In sum, cultural mechanisms for ensuring economic stability in the face of interannual variability repeatedly posed new problems for early farmers and so, in the long term, promoted radical social and economic change. Ultimately elite control of the surplus necessary for its own maintenance was a consequence of short-term attempts by all households to ensure a reliable food supply. Conclusion: risk, surplus and social complexity

Most of the preceding discussion has been devoted to early farmers in Thessaly, so as to explore in some detail the role of surplus in the maintenance and, ultimately, transformation of that society. Obviously, a host of factors conditions the creation and appropriation of surplus in different historical and geographical contexts, but some tentative generalisations may be offered by way of conclusion. In the case of prehistoric Thessaly, the appropria-

The economy has a normal surplus tion of existing surplus by an emergent elite entailed three essential steps: (1) (2) (3)

the production of a normal surplus was a basic household response to the risk of crop failure; some of this normal surplus was 'banked' through social storage in return for material valuables; sustained imbalance between farmers in the likelihood of subsistence success or failure allowed some to accumulate valuables and rights to labour or produce, and so to transform predominance in social storage transactions into economic and political dominance.

Which sorts of societies, in which sorts of environments, are likely to have taken these steps? 1 Periodic discrepancies between food acquired and food needed are universal. Surplus, therefore, in the sense of food which a foraging or producing group cannot expect to eat before it spoils, is probably a feature of all human economies - ranging from the meat exceeding immediate requirements after a successful kill to the enduring 'butter mountains' and 'wine lakes' stockpiled by the EEC. Among hunter-gatherers who are not geared up for storage, day-to-day surpluses are rapidly absorbed by temporary suspension of work or by sharing within a larger social unit (e.g. Sahlins 1974:17-18; Lee 1979:256-7; Whitelaw 1983). Equally such groups tend, by virtue of their small size and mobility, to have access in the event of scarcity to a variety of alternative food sources. Among farmers and hunter-gatherers dependent on seasonal storage, the availability of food must be reckoned over rather longer units of time and quite substantial discrepancies may accumulate - particularly in environments subject to significant interannual variability. In the event of scarcity, such groups tend, by virtue of their large size (e.g. Fletcher 1981:100, Fig. 4.1) and because they are tethered to fields or facilities (cf. Binford 1980), to have access to few alternative food sources. In such situations, a strategy of overproduction and a substantial, recurrent surplus may be anticipated (cf. Rowley-Conwy and Zvelebil, this volume). Such a 'normal surplus' is widely documented, for example among cereal farmers in Africa (Allan 1965), Europe (Forbes, this volume) and the Near East (Hillman 1973:231-2), among root crop cultivators in Melanesia (Vayda, Leeds and Smith 1961), among African and Near Eastern pastoralists (Legge, this volume; Dahl and Hjort 1976:133-7) and among sedentary hunter-fishers of the American northwest coast (Suttles 1968). 2 Overproduction is only an effective buffer against risky environments if normal surplus from good years can in some way be 'banked' for use in the worst years. Direct storage of surplus is unreliable over a period of years, given the severely limited storage life of most foodstuffs, while indirect storage, by feeding surplus to livestock, entails massive energetic losses in conversion (e.g. Rappaport 1968:66-7). Alternatively, normal surplus may be banked through social storage - giving it to others and so establishing some sort of entitlement to future reciprocation (O'Shea 1981; also Sahlins 1974, passim). The nature of this entitlement is rather variable.

79 Among non-storing hunter-gatherers, strong social pressure to share is reinforced by a crowded settlement pattern (Whitelaw 1983) and, with no provision for efficiently storing any surplus, there is little concern to ensure balanced reciprocity. The settlements of farmers and storing hunter-gatherers are far less crowded (Fletcher 1981:100 Fig. 4.1), in part reflecting a greater reluctance to share and a greater concern to ensure reciprocation in food, labour or valuables. A contrast should be drawn between tropical farmers, however, for whom agricultural work and the ripening of crops may be relatively continuous (e.g. Forde 1963:380; Grigg 1974:72; Ellen 1982:159 Fig. 7.4), and those of higher latitudes often facing sharply seasonal environments. In the tropics, production and consumption are often undertaken by rather large social units (e.g. compounds - Flannery 1972; collective work groups - Grigg 1974:58) or by several overlapping small units (wife + husband, wife + brother, etc. - Rappaport 1968:44). In environments which are not markedly seasonal, such collaborative arrangements spread the drudgery and risks of agricultural work and also provide numerous channels for sharing food. Moreover, sharing is often encouraged by the difficulty of direct storage, because of the prevalence of root crops and fruits (e.g. Norman, Pearson and Searle 1984:13 Fig. 1.2) and of warm, wet conditions conducive to spoilage (e.g. Tempany and Grist 1958:246-7), and by the relatively short lapse between harvests. In higher latitudes, by contrast, farmers tend to grow intrinsically more storable seed crops, in environments where there is often a full year between harvests but storage is facilitated by seasonal cold and/or aridity (e.g. Duckham and Masefield 1970:520). Also the strongly seasonal nature of agricultural work militates against widespread cooperation and favours the restricted family household as the basic unit of production and consumption. In spite of the obstacles to sharing posed by such households, social storage may still be vital to survival and it is in such societies that the use of valuable tokens to facilitate exchange of food should be most developed. The family household was the norm among early cereal/pulse farmers in the Near East and Mesoamerica (Flannery 1972), as well as in Greece, and this may account for the precociousflorescenceof craft production in those societies. 3 Even where 'levelling mechanisms', such as social pressures to share surplus, are only weakly developed, institutionalised inequality is unlikely to develop unless imbalances in production are sustained over many years, if not generations. Such imbalance is relatively unlikely among tropical cultivators, who tend to shift their fields frequently and often own land communally (Grigg 1974:57-8). Similarly, among pastoralists the ownership of grazing is often communal (Lefebure 1979; Tapper 1979) and the widespread practice of loaning or giving surplus livestock to other households (to reduce the labour costs of herding and the risk of loss) operates as a powerful constraint on the accumulation of wealth (Dahl 1979). Among mid-latitude farmers fixed-plot farming is more usual and suggestions of an initial phase of shifting cultivation in temperate Europe are ill-founded (Rowley-Conwy 1981). Sustained imbalance is still unlikely, however, where the

Paul Halstead main hazards to farming are equally likely to afflict all producers (e.g. damage by birds, loss of labour through illness). Such conditions may have obtained for early farmers in the less arid parts of southeast Europe, for instance in west Thessaly (above, p. 75). Where the principal threat is from the physical environment (drought, flooding, frost, etc.), even minor differences between plots in height, aspect, soil conditions and so on may make the difference between success and failure. Such conditions are widespread, especially in the semi-arid regions of the world where early cereal/pulse farming developed and where successful harvests commonly depended on uncertain rainfall or on irrigation from rivers which could not be relied upon to supply the right amount of water in the right place at the right time (e.g. Adams 1981). It is not intended to deny that social complexity can arise under different conditions or for different reasons, nor that the development of complexity in the semi-arid regions often followed quite divergent paths - a point developed elsewhere for the contrasting trajectories of Thessaly and southern Greece (Halstead 1981b; Halstead and O'Shea 1982). As regards the initial appropriation of surplus by an emergent elite, however, it is argued that the model developed here for prehistoric Thessaly should be widely applicable to early cereal/pulse farmers. In seeking to ensure a stable food supply, early farmers paved the way for the unequal access to resources which is the hallmark of social stratification. Childe (1954:75-6) recognised the problem facing early farmers: A second defect in the neolithic economy was the very selfsufficiency the barbarian village prized so highly. Such a community ... could reasonably plan ahead to meet future

80 eventualities. But all its labours and plans might be frustrated by events still beyond its control: droughts or floods, tempests or frosts, blights, or hailstorms might annihilate crops and herds. And even a local failure might spell famine and annihilation for the self-contained and isolated community. Its reserves were too small to tide over any prolonged succession of disasters or to let it take preventive measures on an effective scale. But he misunderstood the solution: The worst contradictions in the neolithic economy were transcended when farmers were persuaded or compelled to wring from the soil a surplus above their own domestic requirements, and when this surplus was made available to support new economic classes not directly engaged in producing their own food' (Childe 1954:77). Normal surplus probably did, eventually, support new economic classes, but it was first wrung from the soil as a strictly domestic initiative. And the growth of large settlements and centralised economic institutions did not so much solve, as reformulate, the problems facing early farmers. Notes The Thessalian case study draws heavily on unpublished settlement reconnaissance by David and Lisa French and on reinvestigation of the classic sites of Sesklo and Dhimini by George Hourmouziadhis, Kostas Kotsakis and the late Dhimitris Theokharis. I am also indebted to Andrew Sherratt, who first introduced me to Allan's work on 'normal surplus', and to Glynis Jones, Kostas Kotsakis and John O'Shea, for comments on an earlier draft of this paper. 1 The Thessalian case study is based on data presented and evaluated in greater detail, together with more complete bibliographical information, in Halstead 1984.

Chapter 6 Changing responses to drought among the Wodaabe of Niger Karen Legge

The Wodaabe of Niger are pastoral 'nomads' inhabiting the West African Sahel, a region characterised by low and uncertain rainfall. Drought poses a recurrent threat to the well-being of livestock and adaptations to seasonal, year-to-year and long-term cycles of drought permeate Wodaabe economics and social organisation. The first part of this chapter explores traditional Wodaabe mechanisms for coping with drought. Responses to individual years of low rainfall are often an extension of responses to the more predictable seasonal cycle of aridity but, in the long term, more severe cycles of drought are encountered which necessitate radical disruption of Wodaabe behaviour. The second part of the chapter explores further sources of long-term disruption-the fundamental economic changes which have taken place since colonisation of Niger by the French. The impact of these changes has widely undermined traditional mechanisms for coping with drought.

The Wodaabe of Niger are one of the groups of nomadic pastoralists inhabiting the Sahelian region of West Africa. This region is characterised by low and uncertain rainfall, and drought exercises a profound influence on the well-being of livestock and on the behaviour of the Wodaabe. Strategies for alleviating the effects of seasonal and interannual variability in rainfall permeate many aspects of Wodaabe economic and social organisation, and the potential risks of longer-term cycles of drought are vividly illustrated by the prolonged famine of 1968-74. In the long term, however, drought is by no means the only source of variability in the environment of the Wodaabe and it will be argued below that the crisis of 1968-74 was exacerbated, if not precipitated, by some

of the radical changes which have taken place in the economy of Niger during and since the colonial period. This chapter begins with a review of Wodaabe cultural ecology. Particular emphasis is placed on the problems posed on a seasonal and interannual timescale by drought, and on the ways in which the Wodaabe have coped with these problems in the past. The second part of the chapter describes the long-term effects of colonisation and economic change on the social and economic organisation of the Wodaabe, and examines the ways in which these changes have undermined traditional mechanisms for coping with drought and shortage. The regional setting The Wodaabe of central and southwest Niger (Figure 6.1) make up a small percentage of the 'Fulani' pastoral nomads who are widely spread across West Africa. They are pure pastoralists in the sense that they do not normally cultivate, and rely almost solely on their animals and animal products for income - though not for food consumption. The diet of the Wodaabe relies heavily on millet, acquired in return for animal produce. This part of Niger is also inhabited by Tuareg nomadic pastoralists (once heavily involved, with their camels, in the caravan trade), by Fulfude and Tamasheq speaking agropastoralists, and Hausa sedentary farmers. An important distinction should be noted here between agropastoral systems and mixed farming. In an agropastoral system, the more people depend on cattle the less

Karen Legge

82

Libya 500 km

0

X

Algeria

y

\

Mali

\

Niger )

1 mey

f

Chad

y\

C ^ ^ ^

Nigeria

Fig. 6.1. Map showing the departments of southwest and central Niger discussed in this chapter

they depend on agriculture (in other words, the more cattle they have, the less effort they put into agriculture). A mixed farming system is an integrated system where the more cattle a person has, the more land they can buy or utilise, one where wealth in land and cattle is built up together. This will be discussed below. The basic social and economic unit of the household is the wuro - a man, his wife or wives, and children. Ideally, this unit is self-sufficient and provides most of the labour needed for the care of the herd, which should comprise the equivalent of at least four camels, six cattle or 40 sheep and goats per person. Animals within this herd, however, are owned individually (White 1984), although more than one person may have rights to an animal (Maliki, White, Loutan and Swift 1984). The climate of the Sahel is markedly seasonal. In the pastoral zone of Niger, annual rainfall generally ranges between 200 and 350 mm (Figure 6.2) and the rainy season lasts from two to four months, some time between June and September/October. It is only during this season that pasture and water are abundant.

Fig. 6.2. Rainfall in southwest and central Niger. Annual rainfall (mm) after Swift 1984

There are very few year-round rivers in the pastoral zone and during the dry season the only surface water is to be found in lowlying areas, where pools collect. These provide water that is easily available, but they tend to evaporate and to become polluted from use as the dry season advances. Most water tends to be drawn from traditional wells that are privately owned, either long-term deep wells that can last up to 25 years, or temporary shallow wells that are dug for one dry season (Knight and Starr 1984). From year to year, the duration of the rainy season varies, and the distribution of rainfall is changeable both in timing and place. In years when rainfall is particularly low or scattered at the beginning and end of the dry season, the Wodaabe suffer since, in order to provide a reasonable amount of grazing, rainfall is needed throughout the period. In the longer term, there are also periods when there are stretches of consecutive years of very low rainfall, or no rainfall at all. There is evidence of these cycles of severe drought occurring throughout the history of the region. In dealing with the problems of long-term variation, however, the colonisation of Niger by the French must also be considered. The disruption caused by colonisation, and by the ways in which the indigenous population had to adapt to it, undermined the balance of the mechanisms the Wodaabe use for coping with drought. There is also a spatial dimension to these problems. The Wodaabe have to cope not only with seasonal, interannual and long-term shortage, but also with variation in the location of available resources. Annual rainfall and the duration of the rainy season increase from north to south, as do quality and abundance of pasture. In the south of Niger, however, pastoralists must compete for land with agriculturalists. On a more local level, there is also year to year variation in the distribution of rainfall, pasture and watering points. Finally the location of markets and trading posts may alter in response to long-term shifts in the natural and political environments. Traditional responses to drought 4 In a near subsistence pastoral economy, loss of a substantial portion of the family herd means immediate destitution. This is a risk that every herdsman faces several times in his life, and is a major determinant of his behaviour' (Swift 1973:73). Traditional Wodaabe strategies for dealing with the uncertainty of their environment may be reviewed in terms of four categories of response - mobility, diversification, storage and exchange. Three timescales can also be distinguished - seasonal, interannual and long-term. With increasing timescale, hazards also become more severe and responses to these three levels of hazard are quite different. Mobility Mobility is the most obvious response to the temporally and spatially variable resources of the Sahel. Although 'nomadic' is not necessarily a useful term to describe herders, it does focus on movement as the key strategy crucial for the best use of grassland and water (Brokensha, Horowitz and Scudder 1977). Movement is used by the Wodaabe to offset difficulties such as disease and drought, to maximise the effective use of limited resources and to

Changing responses to drought among the Wodaabe of Niger

83

avoid overgrazing. It is important to consider this movement not as random wandering, but rather as highly informed and flexible exploitation of the environment. A herder's primary concern is to ensure the economic viability of his herd, and access to food, pasture and water is essential; but so too is a knowledge of other human and animal populations, and also the marketing possibilities of the area (Stenning 1959). Stenning, whose work was based in neighbouring Northern Nigeria, identifies three forms of movement: transhumance or kodol- 'regular seasonal movements'; migratory drift or eggol- 'a gradual displacement of transhumance orbits'; and migration or perol- 'the assumption of new transhumance orbits by a sudden and often lengthy movement' (Stenning 1957:59). For most Wodaabe groups, patterns of seasonal transhumance are well established and can cover several hundred kilometres. As the rains end, in September or October, the Wodaabe herds retreat south from their rainy-season pastures in the northern Sahel. They move into the agricultural areas of central and southern Niger that have been used for cultivation during the rainy season. The timing of this is crucial since herders have to be careful not to arrive before the harvest. The most critical period in this cyclical movement is the very end of the dry season. At this time, the grazing is very scarce, especially after the crop residues have been finished (Campbell 1977), and fadama lands, which are seasonally waterlogged and retain green grass well into the dry season, are very important. Just as movement is important in the dry season in order to search out scarce resources, so it is in the rainy season to maintain the quality and quantity of pasture and water. At the beginning of the rainy season in June, the Wodaabe move often, in search of new grass, and it is this that takes them so far northward into the drier parts of the Sahel. Wodaabe water their animals as much as possible at this time (several times a day if they can) not only for them to drink, but also to increase lactation (Maliki et al. 1984). Residence patterns also vary with the seasons. Members of the same kinnidal (migratory group) come together at the beginning of the rainy season, when the abundance of pasture and water allows a greater concentration of animals and people, and decide on the cycle of movement for the whole group. In the dry season, migratory group members share pasture, but individual households decide singly about movement strategies and tend to scatter to maintain economic viability (Maliki et al. 1984). The impact of these residential adjustments on the exchange of labour and information is considered below. Migratory drift involves adjustment of the cycle of transhumance in response to interannual changes in the availability of grassland and water and to competition with other pastoralists and agriculturalists (Stenning 1959). There are also political factors involved in decisions about movement, such as changing alliances and conflicts between lineages, clans and households. On a seasonal and interannual timescale, movement is a carefully regulated strategy to offset the effects of a harsh environment. Stenning (1959) argues that it is only when conditions change very suddenly, for example in the case of disease, war, famine or political pressure, that sudden shifts in the pastoral population occur.

As a response to long-term risk, migration is very important: 'the possibility of moving away remains the most important protection against disaster' (Swift 1973:74). Scott and Gormley (1980) argue that the Wodaabe move more readily than some other herders and that it is this that results in their high survival rate in times of hardship. Diversification On a seasonal scale, diversity in the composition of the herd is one of the main ways in which an individual herder is able to exploit the potential of his environment (Campbell 1977; Scott and Gormley 1980). Cattle are usually the main component of the herd, with varying numbers of sheep and goats, and occasional camels and horses. In this way, with some animals browsing and others grazing, the environment is used more evenly. Moreover, 'species diversification provides a more regular milk supply', but 'because of different grazing and watering requirements, more people are needed to tend a herd made up of different species' (Maliki et al. 1984:496). Breeding is carefully controlled to avoid the loss of any animals, so the young are born in the rains at a time of relative plenty. On an interannual scale, 'species diversification ... reduces the danger of loss of an entire herd from one disease' (Maliki et al. 1984:496). Diversification of the herd is also advantageous in times of grain shortage when prices are higher, since small stock can be sold to buy food. Furthermore, in the event of the loss of some animals, numbers can be made up again since small stock breed more quickly and can be sold to buy cattle. Gathering is an important fallback activity. It is often a subsidiary activity in times of abundance, but is taken up on a larger scale when grain and milk become scarce. Colson (1979) argues that in times of shortage there is a shift away from 'preferred foods'. For example, when milk yields are drastically reduced in times of drought, the Fulani pastoralists of Senegal replace milk with purchased rice and oil (Benefice et al. 1983, cited by Fleuret 1986). Food preparation may also alter so that less gets thrown away (Colson 1979). Responses to long-term variability often involve a more marked change of strategy. For example, some Wodaabe resort to agropastoralism, growing crops to feed themselves when they are forced to abandon the nomadic pastoral economy. Farming is a strategy that it is hoped will be no more than temporary, although inevitably some people settle and take up farming when losses of livestock are so severe that the herd owner cannot recoup his losses (Stenning 1959). Storage The Wodaabe largely store 'on the hoof, since the nature of their seasonal movement does not allow them to carry other stored resources. As one Wodaabe herder put it: The cow is like the granary of the Hausa. A Hausa farmer, having worked all year to fill his grain store, is not going to just sell his harvest, his food for the year... Anyway, a cow is so much better than a granary because ants and lizards

Karen Legge can eat the grain and a fire can destroy the granary. Instead, the cow is our insurance, our certitude, and it can be milked morning and night, and it will always be with us, our certitude is to have a good, growing, healthy flock of animals. (Scott and Gormley 1980) As a response to interannual variability, herd size is increased beyond the needs of normal subsistence. This is a safeguard against drought or disease, since any stock that survive a crisis will be used to rebuild the herd. Traditionally, therefore, it is advantageous for individual households to build up their herds, regardless of the dangers of overstocking, since this improves the chances that some of their stock will survive a severe drought. There are, however, inevitable constraints on herd size - the availability of pasture and water and also the availability of labour, as herding is intensive. These limitations on storage in turn influence exchange relationships among the Wodaabe. Exchange In addition to the three timescales - seasonal, interannual and long-term - three principal categories of exchange may be distinguished: animal loans, exchanges of labour and exchanges of animals for agricultural produce. During the rainy season, milk and milk products are widely shared, but on an interannual scale loans of livestock are more significant. On this scale, there are two types of traditional loans among the Wodaabe. Dialle animals are loaned to a herder to use their milk and are later returned with any offspring. Such loans are for short periods of temporary hardship, providing milk as an immediate source of food. Habbanae animals are loaned until they have produced three calves, which are kept by the borrower; the milk from these animals can also be used during the loan. Habbanae loans serve to rebuild a herd, or to assist a young man whose father cannot give him enough stock to start a herd of his own. As social relationships are articulated through stock, even a wealthy herder may have habbanae animals both on loan and at loan. A third means of acquiring the use of livestock is to herd jokeere stock belonging to sedentary farmers. Herders can increase their milk supply by doing this and so stay within the pastoral economy, but they face enormous disadvantages, which make it a very last resort. Traditionally, this arrangement was rarely used, and then only in times of extreme hardship (Dupire 1962a: 127), but it is of considerable relevance in the long term (see below, p. 85). The exchange of labour between households is used in the dry season to split the herd and so make the best use of sparse grazing (Dupire 1962a; Colson 1979; Dyson Hudson 1984). Similarly, on an interannual scale, when drought or disease threatens, kin or stock partners are useful if livestock need to be relocated elsewhere: 'social relations, the most important safeguard in a risky environment, must be maintained through regular contact with other Wodaabe in scattered camps' (Maliki et al. 1984:317). Indeed, 'the Fulani are widely believed to have lost fewer animals than the Tuareg, in part because of their wide network of kinsmen scattered throughout West Africa1 (Berg 1976:24; emphasis added). In response to variability in the longer term, people extend the ex-

84 change of labour outside their own community, and some members of the household leave in search of migrant work as labourers. There are two main contexts in which seasonal interaction in the exchange of animals for agricultural produce takes place, and both are vital for the survival of the pastoral population. The first is the use of harvested fields for pasture at the beginning of the dry season. This relationship works to the mutual benefit of pastoralists and farmers. The animals graze the harvested fields and provide manure, which is highly valued by the farmers; they also trample in residues, break up ridges and strip stalks to be used later as building materials. To gain benefit from the manure, a farmer must keep the herder on his land for several days. Some farmers plant trees, the pods of which are very nutritious, or dig wells to attract herders to their fields (Horowitz 1972; McCown, Haalan and deHaan 1979). This relationship of interdependence takes many forms, and often the Wodaabe are not paid in cash but in grain, thatching grass or simply with the rights to use the land - especially in areas where dry-season grazing is particularly scarce. There is, however, inevitable cause for antagonism in relationships that are articulated around close proximity. If herders arrive early with their cattle, the animals can damage unharvested crops. Moreover, in order to reach a farmer with whom the herder has established a relationship, he may need to pass through the land of other agriculturalists who may not yet have harvested. The second form of interaction which is very necessary to the Wodaabe also takes place at the time when herders and settled agriculturalists are in close proximity. At the beginning of the dry season, when milk supplies start to decrease, Wodaabe women take sour milk and butter to the market in order to supplement their diet with purchased millet. Animals are also occasionally sold - ideally only those that do not contribute to the productivity of the herd. In this way, the Wodaabe acquire the millet that makes up the bulk of their diet, along with other household products. Farmers provide the Wodaabe with a market for their milk products, while the Wodaabe are a ready market for Hausa grain. The Wodaabe also maintain contact at other times (Dupire 1962b). If they are far from markets, Wodaabe trade with diallo (itinerant traders), but by doing this they often obtain a low price for their produce and may pay increased grain prices. Markets, however, are the preferred venue for sales and in northern Nigeria 'market places may primarily exist to enable Hausa farmers to sell grain etc. to the pastoralist who sells his milk and butter' (Hill 1972:293). On an interannual scale, the Wodaabe tend to sell cattle in bad years and for this they use the market place, selling through a 'broker' or 'middleman' (Dupire 1962b). The broker, who receives a percentage of the sale price, obviously wants to reach the maximum price possible. Using a broker also enables a sale to be kept anonymous, since it is against Wodaabe concepts of co-operation and mutual help for one herdsman to profit from another (Reisman 1977). The exchange of information (about the availability of grazing and water) is also an important response to interannual fluctuation in resources, and the market place provides a useful venue for this. In the long term, in cases of extreme hardship,

Changing responses to drought among the Wodaabe of Niger

85

when Wodaabe are driven out of the pastoral economy, they may use the market to sell traditional medicines and charms or to sell skills such as the repairing of calabashes or the braiding of hair (Malikie/fl/. 1984).

Secondly, access to harvested fields has also been drastically reduced. A growing number of farmers have bought cattle and so use their own stubble fields, denying access to the Wodaabe. Thirdly, improved water supplies have cut off the escape routes the Wodaabe traditionally used; some have stayed too long in the dry-season pastures because of the water supply, and cattle have starved to death while standing in water (Scott and Gormley 1980). One Tuareg group in Niger actually asked the government to close the boreholes in their region, since they were so disruptive (Bernus 1974). Diversification as a risk-avoidance strategy has been detrimentally affected by the growth of migrant labour. Inadequate labour has obliged many Wodaabe to abandon the traditional maintenance of a mixed herd, as this is very labour-intensive particularly in the dry season (White 1984). Storage on the hoof has also been undermined by the need of the Wodaabe to sell livestock to pay taxes and by the increasing commercialisation of their relationship with farmers. Reciprocal exchange between Wodaabe and Hausa continued, but in a modified form. Exchanges of millet for animal products were more often on a commercial rather than reciprocal basis (Campbell 1977), forcing Wodaabe to sell stock to traders or farmers in order to buy food as well as to pay taxes. This reorientation of economic activity meant that the Wodaabe faced problems of the instability of the market (Swift 1977). Pastoralists generally have no control over the prices of what they buy, and as a result are vulnerable both to short-term fluctuations in prices and to such long-term changes as the shift towards cash cropping (Swift 1977). At times of shortage during a drought the situation is particularly severe. 'Because of their increased dependence on the market, the shift in terms of trade that takes place in a drought, with livestock prices plummeting and cereal prices rising sharply, meant that most Wodaabe had to sell the animals that had survived in order to buy cereals' (White 1984:4). If the market is flooded with animals selling at greatly reduced prices and the cost of grain is high after a poor harvest, pastoralists may face destitution or, at best, temporary or permanent sedentarisation (Diarra 1975). During the 1968-74 drought, not only were cattle sold off at reduced prices, but reproductive females were amongst those sold. At Niamey slaughterhouse, 70 percent of the cattle slaughtered in 1973 were female, compared with only 30 percent in 1968 at the beginning of the drought (Berg 1976). Sales of livestock undermined the traditional loaning network, since cattle were quite simply not available. Unable to rebuild their herds by traditional means, many Wodaabe have had to resort to herding jokeere stock after a drought in order to remain in the pastoral economy. This system creates problems for the individual Wodaabe who is paid only with the milk of the cattle he herds. Jokeere herds produce little milk, however, because farmers tend to retain lactating cows and to loan young male stock. Thus, although poor herders have become increasingly dependent on jokeere animals, they may still have to sell their own stock to buy grain and so the problem of maintaining a viable herd remains unsolved (White 1984).

The long-term: the impact of economic change on responses to drought Although French colonisation of Niger began in the 1890s, the establishment of a civilian administration and transformation of the traditional economy has been a gradual process spanning the twentieth century. First, to 'internalise' the cost of colonisation, the French placed a head tax on the sedentary population and an animal tax on the nomadic. A further export tax was placed on cattle, since the colonial administration wanted to benefit from such trade between Niger and Nigeria (Ball 1978). Secondly, farmers and pastoralists alike were drawn into the cash economy, because they had to pay taxes in cash and were encouraged to supply cash crops and meat for developing urban markets (Campbell 1977; Watts 1983). Thirdly, there was a notion that herders should be settled, or at least their movements regulated, for reasons of health and education (Scott and Gormley 1980). There were large-scale veterinary campaigns against rinderpest (Brokensha et al. 1977) and boreholes were constructed as reliable sources of water. Together these three interventions by the colonial administration 'created a whole new dimension of uncertainty in what was already a demanding environment' (Baier 1980:140). To understand how the French undermined traditional mechanisms for coping with variability, it is first necessary to examine the ways in which the population adapted to the changing regional economy. To pay their taxes, the Wodaabe had to increase sales of animals and labour to the urban centres. Similarly agriculturalists increasingly grew cash crops at the expense of food crops in order to pay their taxes. This directly affected pastoralists, since access to staple foodstuffs was greatly reduced in the markets. Cash cropping also caused an increase in the demand for productive land (Campbell 1977), and growing numbers of farmers were able to buy their own cattle from the sale of their crops. The result of increased water and medical supplies was a considerable expansion in herd size, leading to problems of overstocking. Moreover, instead of small groups spread out in search of water and pasture, huge groups now accumulated around boreholes particularly in the dry season, so exacerbating the problem of overgrazing. These changes in economic organisation have seriously undermined Wodaabe risk-buffering strategies. Traditional patterns of mobility have been severely disrupted by the expansion of cultivation and by the development of water supplies. First, agricultural expansion northwards has reduced land available for grazing and cut off traditional transhumance routes (Diarra 1975). Fadama land, for example, has been taken up with the production of cotton, which matures at exactly the time the Wodaabe would use the land (Raynaut 1971). The effect of agricultural expansion is worst in the dry season, which some Wodaabe have had to spend in more northerly areas of very limited rainfall and pasture.

Karen Legge

86

Jokeere stock create further problems. Sedentary farmers often want stock kept relatively near their own land so that they can use them for manure and milk. The herder's mobility is thus restricted and he may be forced to graze cattle near wells with large numbers of other stock, so increasing the risk of disease. If stock dies, the herder is blamed and often has to pay compensation. Relationships with other Wodaabe are also jeopardised, since a herder who has to take on jokeere animals is not always trusted to be a competent herder and Wodaabe would rather not give habbanae loans to those who are not able to manage stock {MdAMetal. 1984). The Wodaabe are ill placed to benefit from working in the towns, because they have few relevant skills, and the detrimental effect of migrant labour on the manpower available for herding has already been noted. Nonetheless, in spite of its disadvantages, migrant labour has become increasingly widespread as other survival strategies have been undermined.

ever, responses to long-term variability are likely to be radically different from those operating in the shorter term - as for example in the switch to agropastoralism by herders forced out of the pastoral economy. All of the four basic categories of response distinguished above are integral to the traditional lifestyle of the Wodaabe, but mobility and exchange (of information, labour and resources) are the most fundamental. Within the category of exchange, social relationships with other pastoralists allowed the rebuilding of herds after a subsistence failure but the more economic relationship with sedentary farmers was arguably even more vital. The Wodaabe were dependent on farmers for the millet which made up the bulk of their diet. It is highly probable that, without this interaction, the Wodaabe would have been forced out of their traditional pastoral economy into agropastoralism. The long-term consequence of French intervention in pastoralism was that the traditional strategy of increasing herd size beyond the level of subsistence (storage on the hoof) was seen as irrational and subsequent problems of overgrazing as the fault of the herder. The provision of better water supplies and veterinary care was intended to improve the life of the Wodaabe. In reality, however, the highly integrated system of traditional responses to variability was severely undermined by these measures and by the development of a market economy. While in the short term these measures may have seemed to facilitate the herding economy, in the long term they made the Wodaabe more vulnerable to severe episodes of drought. Significantly, current attempts to rebuild the Wodaabe pastoral economy, following the disastrous drought of 1968-74, are increasingly articulated around the reestablishment of traditional practices such as the loaning of livestock.

Conclusion

Traditional Wodaabe responses to variability were diverse and flexible. The combination of mobility, diversified herd structure and extensive networks of social relationships allowed the Wodaabe to cope with great variability in resources. In general, problems of greater scale and severity elicited responses of greater scope and intensity. This is exemplified by the distinction between transhumance, migratory drift and migration, which operate respectively on a seasonal, interannual and long-term basis. Responses to interannual variability are often extensions of responses to seasonal variability. Long-term responses may be 'embedded' in more regular aspects of behaviour. Loans of livestock, for example, which served to rebuild herds after pastoral households had succumbed to a particularly severe hazard, were also central to the articulation of social relationships and so to the manipulation of resources, labour and information on a day-to-day basis. How-

Notes I am grateful to Dr. Paul Baxter, Department of Social Anthropology, University of Manchester, for his advice in writing this paper.

87

Chapter 7 Of grandfathers and grand theories: the hierarchised ordering of responses to hazard in a Greek rural community H. Forbes

This chapter is based on ethnographic data from a modern, predominantly subsistence-orientated, agricultural community in southern Greece. The community's environmental background is summarised with special reference to the inherent variability of factors affecting crop production, and several superficially 'inefficient' behaviours are identified as buffers against environmental variability. It is argued that these mechanisms do not all operate simultaneously or at the same level, but represent a hierarchised set of culturally specific responses to hazard. The interconnectedness of social and economic factors in these buffering mechanisms is discussed, and it is argued that several of these primarily economic behaviours are forces which discourage change in the social sphere.

It goes without saying that climatic - and, more broadly, environmental - factors play a crucial role in affecting agricultural communities and how they make a living. In the past, most writers on the subject of traditional or 'primitive' agrarian economies have tended to treat discussions of the environment in general, and climatic factors in particular, in a normative fashion, frequently not looking further than averages of rainfall, temperature and the like. This equilibrium-centred thinking frequently assumed that 'the climate' of a locality is essentially static from year to year, though accepting that it is punctuated by occasional hiccoughs or 'crises'. Recently, however, it has become apparent that any location's 'climate' (which is in any case a complex interrelationship of a multiplicity of different factors) can be expected to fluctuate widely from year to year as a matter of course. This

variability is quite exclusive of the more occasional major fluctuations which can be guaranteed to occur, although the exact timing of such occurrences is usually not predictable. The natural corollary of this realisation is that agrarian households are not adapted to some average climatic pattern: such an adaptation would mean starvation as often as sufficiency. Rather, their economies are adjusted to a wide range of expectable, though not precisely predictable, climatic conditions. Furthermore, it will be argued here that the very success of such adjustments has made it remarkably difficult for either outside observers or native actors to identify just how effective some of these mechanisms are at providing a stable food supply in the face of the constant variability of environmental parameters. Frequently, practices which may, after careful investigation, appear to the field researcher as survival mechanisms, are legitimised by farmers simply by the statement 'that's the way our grandfathers did it'. This occurrence is so widespread that it has been called the 'grandfather response' (Alland 1975:70). Besides the normal interannual fluctuations there will inevitably be more serious, though more infrequent, perturbations which will be beyond the capacity of the group's normal coping mechanisms. It is these rarer but more severe perturbations which many studies of environmental variability have defined as 'hazards' (Vayda and McCay 1975:294). The use of such a term is perhaps unfortunate inasmuch as it suggests that these fluctuations are in some way qualitatively different from other forms of en-

H. Forbes

Fig. 7.1. The peninsula of Methana (inset: map of southern mainland Greece showing the location of Athens and Methana). Key: • commune head; • village; O hamlet; road; track

vironmental variability. Nevertheless, the term does indicate, even if in a rather sloppy way, that the largely invisible year-to-year mechanisms for coping with fluctuations in environmental parameters have distinct limitations beyond which they are not effective. It is the intention here to identify sound principles rooted in ecological and evolutionary theory, which explain: (1) why the normal coping mechanisms cannot be expected to cope with the full range of environmental variability, but (2) why communities or individuals do not automatically face imminent demise as a result of this fact. These arguments will be based on the results of the study of a specific community in modern Greece. The data presented here result from a two-year study in the peninsula of Methana, undertaken by the author from 1972 to 1974. A number of further short summer visits have been made during the 1980s. Methana is located on the southern edge of the Saronic Gulf, in relatively close proximity to Greece's major urban centre, Athens and Peiraieus. The existence of hot springs and its proximity to the capital have made Methana a popular spa and holiday resort for Greeks (rather than foreign nationals) from all over Greece, though particularly from the Athens-Peiraieus area, since the 1870s (Miliarakis 1886:207-8; Akritas 1957:20). Despite these facts, the economy of the peninsula is still predominantly agrarian, and although rural-urban migration has affected the population, it has not led to results comparable to some of the extreme situations known elsewhere in Greece where agriculture has been largely abandoned. Especially in the pseudonymously named village of Kosona, where the study was focused, the primary goal of by far the greatest number of households is agricul-

tural subsistence. By this term I mean that the value of the crops grown specifically for cash sale is estimated to be less than the value of the crops grown primarily for household consumption. Virtually all Kosona households are self-sufficient in all three elements of the Mediterranean staple dietary triad of cereals (here = wheat), olive oil and wine. A wide variety of other arable and tree crops is also grown. The list of arable crops includes (besides wheat) barley, oats, millet, vetch, peas, broad beans, chick peas, sesame and cotton, as well as many different vegetable crops. The range of important tree crops apart from olives includes figs, almonds, carobs and pears. Sources of income other than agriculture are known, but of those Kosonites who indulge in such cashproducing occupations, almost all do so only in relatively slack periods in the agricultural cycle. Employment connected with summer tourism is particularly important in this respect, since the period from the end of the cereal harvest (late June) until the vintage (end of September) sees a relative lull in agricultural activity. Generally, however, Kosona households have a cash-poor, subsistence-orientated economy of a kind frequently considered to typify pre-industrial peasantries (Forbes 1982:158-99). This is not to say that Methanites have entirely opted out of a cash economy, since cash is needed for the replacement of equipment, and even minor food items may be bought, but the primary aim of household economies tends to be production by their members for their own direct consumption. Contrary to the definition of Wolf (1966:2-4, 60), however, the peasants of Methana are not obliged to transfer their surplus production to a dominant group for further redistribution to the non-agrarian sector. Agricultural households on Methana have very largely, if not entirely, supported themselves on the cultivation of their own land from as far back as Methanites can remember. Furthermore, during the period of fieldwork the only tax levied by the government on agricultural production has been on produce for sale in the wholesale market in the capital. The proceeds of this tax go to the farmers' own social security fund and to capital improvements in villages. Thus productivity in excess of households' needs is not siphoned off to support the status of a dominant group, nor are peasants forced to produce in excess of subsistence needs owing to the coercion of an elite. The term 'peasantry' is not introduced here as a means of embarking on an extended semantic argument about social typologies, or where to place the agrarian economy of Methana within a range of social, economic or political 'types'. Readers who consider such an indulgence necessary or enjoyable are referred elsewhere (Forbes 1982:17-19). In this article the term 'peasant' will be used interchangeably with 'farmer', in keeping with recent usage in the anthropological literature on agricultural change and the adaptive strategies of agriculturalists in developing countries (Barlett 1980a:546). The abandonment of a focus on peasantry as a social type allows us a more fruitful focus: the family farm as a basic unit of study. The emphasis on the individual household is important not simply because it eliminates an excessive concentration on terminology to the detriment of substance. More importantly, much as older views of environmental systems were limited by a normative approach, so a normative approach to human be-

Of grandfathers and grand theories

89

haviour obscures the range of variability and the responses behind it. Since the fundamental decision-making unit in most non-industrialised agriculture is the farm household, a focus on decisions at this level leads away from erroneous generalisations about group behaviour based on inadequate quantified data (Barlett 1980a:547; Netting, Wilk and Arnould 1984: esp. xiii-xxxviii). The farmers of Methana exploit a landscape that is unusually rugged even by Greek standards. The cause is the largely volcanic origin of the peninsula, which has produced a complex landscape composed of volcanic craters, lava ridges and steep slopes, in which even gently sloping land is at a premium. Because of the extremely rugged nature of the terrain, mechanisation is restricted to static installations such as olive oil extraction establishments, irrigation pumps and threshing machines. The day-to-day running of the small family farms is dependent on the energy input of family members and their domestic animals. The soil, however, seems unusually productive for arable crops compared with that found in most areas surrounding Methana, and there are large amounts of cultivable land at all altitudes from sea level virtually to the highest point on the peninsula (743 m). Because of the settlement pattern this means that there is a full range of arable land at all altitudes from sea level to well above 700 m within one to one and a half hours' walking distance from every village on Methana (Figure 7.1). The soils themselves are highly variable, particularly as regards water-retentiveness. This is largely a factor of soil-particle size, with some soils being sandy and very free draining, while others are stiff heavy clays. The inherent 'dampness' or 'dryness' of soils is also mediated, however, by the amount of exposure a locality presents to sunlight and drying winds: one result of the extreme ruggedness of the landscape is that certain craters, deep pockets or steep slopes may be much more sheltered from evaporation than flatter or more open areas.

events and of the onset of periods of hot, dry weather, particularly on wheat but also on olives. Rainfall events on Methana are least dependable, and amounts of rain tend to be lowest, at the beginning and end of the rainfall period; temperatures also tend to rise rapidly in the two months prior to the cereal harvest (Forbes 1982:453-4). Wheat produces poor yields if soil-moisture is low at seeding time, and also if rainfall in the two months before harvest is below 50-75 mm. High temperatures in the pre-harvest period also result in poor yields (Leonard and Martin 1963:285-6). A substantial fall of rain towards the end of September or in October is also important for the olive crop. As already noted, however, a rainfall at this point in the year is very uncertain: frequently the first significant rainfall may not occur until November (Forbes 1982:453-4). Under these circumstances, poor or even non-existent crop returns are a fairly common occurrence. The subsistence farmers of Methana therefore make a living in a natural environment characterised by a semi-arid climate with relatively high temperatures and a low and unpredictable rainfall regime. But they must also be seen as integrated into a wider social, economic and political system. This introduces further problems of unpredictability in the social environment: unforeseen lurches in government agrarian policy, fluctuations in national and international commodity prices, taxation, and occasionally war, or at least the economic and social disruptions caused by the periodic sabre-rattling and brinkmanship of nationalist governments in the Eastern Mediterranean area. All these factors are potential or actual hazards with which peasantries must cope effectively if they are to survive (Vayda and McCay 1975:294). The primary response of Methana fanners to the array of potential socio-economic and political hazards has been one of withdrawal: the lack of emphasis on cash crops, plus the low cash inputs into household economies generally, means that few mistakes can be made in gauging markets, and there is little spare cash for external agencies (e.g. governments; 'liberating' partisans in wartime) to get their hands on. Partly because of this withdrawal I shall concentrate my discussion on defence mechanisms associated with the natural environment - particularly climatic perturbations. The other reason for this approach is that the threat of serious climatic perturbation is ever-present, while potential hazards in the sociopolitical sphere tend to erupt more sporadically.

Climatically, Methana is not well favoured. With a mean of under 400 mm annual rainfall (Kayser and Thompson 1964:1:03), concentrated almost exclusively in the winter months from the end of October to the beginning of April, it is located within the driest zone of Greece. The low annual rainfall figure, combined with a mean temperature of 18.5°C (Kayser and Thompson 1964:1:04), places Methana squarely in the band of environments defined as semi-arid, with all the special agrarian problems such environments present (Hall, Cannell and Lawton 1979; Bailey 1979:79, 93). These factors by themselves are important in considerations of mechanisms for ensuring adequate food supplies, but there are two further variables which are every bit as important. The first is the range of variability in the amount of rainfall from year to year. On a global scale, interannual rainfall levels are generally most variable where rainfall is least abundant (Ricklefs 1980:847-8). Thus on Methana not only is rainfall low, but amounts vary widely from year to year, and extreme drought conditions can be expected relatively frequently. Although variability in other aspects of the climate can have adverse consequences for crop production (e.g. high temperatures, drying winds), it is primarily variability in rainfall - both in amount and distribution - which has the most profound effects on agriculture on Methana. The second factor is the effect of the timing both of rainfall

One result of the Methana fieldwork has been the identification of a considerable number of practices which allow the population an inbuilt resilience against environmental shock. Many of these are practices which at one time or another have been singled out by western agricultural economists as 'irrational', inefficient or uneconomic. They include: 1

2

3

A quite remarkable level of fragmentation of households' land-holdings, resulting in increased time and energy expenditure involved in travelling between plots; The practice of polycropping trees and arable crops on the same plot, which has the potential for reduced production in either or both crops; The deliberate aim of sowing certain crops in such large

H. Forbes

4

amounts that overproduction will occur under 'average' climatic conditions, even though no market sale is intended; The practice of keeping excess agricultural produce in storage over relatively long periods. This leads to inevitable problems of deterioration in quality and loss of cash value when they are sold.

Methanites also talked of people having survived serious food shortages in the past by breaking the normal social rules: begging for food; getting rid of excess dependents by forcing them into unwanted marriages; eating foods outside the culturally acceptable range. All of these behaviours can be defined as hazard-response mechanisms (HRMs). In their different ways they allow (or have allowed) Methanites to survive various levels of fluctuation in climatic and social conditions from the statistical norm. But it is also apparent that not all of these HRMs work with the same frequency or at the same level of mobilisation. Furthermore, it is evident that, although differing in detail, a comparable array of HRMs can be identified in other parts of the world, and that a hierarchisation of responses fits these comparative data as well as those from Methana. I am therefore suggesting not only that HRMs are an essential feature cross-culturally, but that they will always be manifested as a hierarchised or nested system of responses. Such a hierarchy allows an energetically efficient mobilisation of successive responses to any particular level of threat. To understand the rationale for the hierarchisation of HRMs it is necessary to turn to basic evolutionary and ecological theory. This is not simply to develop a biologically derived model, but because human populations and natural systems alike are subject to the same sorts of environmental threats. If they are to be successful, all populations, whether human or non-human, must therefore develop mechanisms to cope with such threats. In evolutionary theory it is accepted that populations of specialised organisms (in an evolutionary sense) which are finely adapted to a particular niche are prone to fail if that restricted niche is disturbed. Generalised organisms, on the other hand, while less energetically efficient, normally exploit a broader niche and a more varied habitat. They are thus less at risk from perturbations. By the same token, if a human population were to fine-tune the efficiency of its food-getting behaviour to a narrow range of climatic factors (such as average rainfall), it too would risk extinction in the face of periodic fluctuations. The same situation applies to whole natural ecological systems. It is now understood that they are not in some finely balanced state, but must withstand a range of natural traumas and shocks. They have therefore evolved tactics to absorb the consequences of change. Resilience itself is thus now seen by some ecologists as a system property which is maintained by various evolutionary processes. It is the very success of this resilience that causes the appearance of stasis (Vayda and McCay 1975). In human terms, populations build resistance into their survival systems by employing various HRMs to suit their particular conditions. However, as with most adaptive features in the biological world, HRMs have inbuilt costs: in this case, social or economic/energetic costs. The most effective way for a human popula-

90 tion to balance the opposed needs of resilience and efficiency is to organise HRMs at different levels, to deal with increasing levels of hazard severity. If lower-level HRMs were to be so generalised as to cover virtually all eventualities, they would be unacceptably inefficient. To ensure moderate efficiency without losing overall resilience, those HRMs which are constantly in operation can be expected to cope with most of the environmental variability (we may hazard a guess of perhaps 80 percent based on tentative evidence from Methana) but to rely on a back-up system or systems to cope with the rest. In this way the demands of resilience and efficiency can be balanced. For the Methana data I divide these HRMs into three response levels, progressing from low to high levels of threat: 1 2 3

First-defence mechanisms Safety-net mechanisms Emergency mechanisms

On logical grounds I would suggest that a fourth level should exist on Methana - catastrophe mechanisms. But no hazards of a monumentality sufficient to necessitate their appearance can be documented. Several of the categories of apparently irrational farming practice previously noted can be recognised as first-defence mechanisms: (a) Poly cropping. The practice of polycropping - growing arable crops under tree crops - is characteristic of the Mediterranean zone, but it has been criticised as something to be discouraged in modern agriculture (e.g. Pansiot and Rebour 1961:100-2). Methanites also note that the practice has disadvantages. If rainfall is low, one or both categories of crop will suffer poor returns due to soil-moisture competition. The exact results depend on details of the rainfall pattern. The full adaptive significance of polycropping on Methana, despite the risk of reduced yields, is complex, involving factors of labour economies, shortage of good land, restricted nitrogen supplies, and climatic conditions (Forbes 1982:312-23). But two elements in the equation are the cultivation and manuring of both crops in a single operation, and the reduction of risk. Trees and arable crops exploit two different, though overlapping, niches on the same plot, arable crops being relatively shallowrooted, while tree roots tend to be found at greater depths. Depending on the patterning of rainfall, one category of crop on a plot may produce a satisfactory harvest while the other may fail almost entirely. Inopportune climatic conditions at flowering time may also seriously affect the productivity of any of the tree crops while leaving arable crops largely unaffected. Given these potential problems, polycropping increases security by allowing a household two chances to extract a crop from a single plot of land while maximising the return from a given labour input (Forbes 1982:312-23). As Norman (1977:86) has stressed in consideration of an African example, mixed cropping under indigenous technological, sociological and economic conditions is consistent with the desire for both security and profit maximisation. (b) Landfragmentation. There is a remarkable degree of fragmentation in households' landholdings on Methana, which inevit-

Of grandfathers and grand theories

91

Fig. 7.2. Distribution of total plots and trees owned by Household 2, Kosona. Key: A: almonds; C: carobs; F:figs;O: olives; P: pears; V: vines; T: arable land. Contours in metres. ably leads to increased time and energy expenditure in travelling between scattered small plots. Figures for Kosona (NSSG 1971) indicate a household mean of approximately 3.3 ha divided into 20 plots, giving a mean plot size of under 0.2 ha. Such fragmentation is largely the result of the division of inheritances and dowries in a partible inheritance system over the generations (Forbes 1982:131-57). Moreover, because of the practice of exploiting trees and arable simultaneously in a polycropping system, trees and arable land on the same plot may be inherited and owned by different individuals, which further exacerbates the situation. Figure 7.2 illustrates the situation with regard to a particular Kosona household. Despite the obvious disadvantages of owning plots scattered so widely - disadvantages which have been stressed by both academics and the Greek government (Forbes 1982:324) the situation on Methana has much the same beneficial results as those found in Alpine environments (Rhoades and Thompson 1975:539). In the highly variable terrain of Methana, agricultural fragmentation ensures that each household exploits plots with a wide range of tolerances and susceptibilities to different climatic conditions. Hence, plots susceptible to unusually damp conditions will produce reasonable returns in drier-than-average years, and vice versa, ensuring adequate overall yields under most conditions. A quantified example of the effectiveness of this pattern is illustrated in Tables 7.1 and 7.2, which document in detail the wheat production of a single Kosona household over 12 years. Because of the two-year crop rotation system on Methana, plots are cultivated in alternate years - hence the need for two separate charts, each with a different set of plots. Significantly, the mean coefficient of variation for individual plots is substantially higher than the coefficient of variation for the total harvest in each of the two data-sets (28.95:16.30 for odd-year rotation; 27.14:16.11 for even-year rotation). Thus it is plain that for this household's par-

ticular mix of plots the variability in the total wheat harvest is reduced to nearly half of the variability of the average plot. This ensures a significant degree of security, although at the expense of notable loss of efficiency in terms of time spent travelling between plots. A comparable situation has been documented for households' holdings of vines (Forbes 1982:334-50). (c) Production strategies. Analysis of data on the amounts of wheat sown by Kosona households indicates that in 87 percent of reported cases wheat was sown with the conscious expectation of producing a significant surplus (over 10 percent above households' needs) under average weather conditions (Figure 7.3). The mean projected surplus above household requirements for all reported cases under such conditions was 63 percent (Forbes 1982:356-65). The substantial surpluses above households' needs would not be remarkable if cash sale were an important consideration. But both farmers' own statements and statistical testing of the available data strongly suggest that cash sale is not an important consideration in households' decisions concerning the amount of wheat to be sown. Only one household in Kosona professed to grow wheat for sale: it is represented by the two anomalous cases on the far right of the graph (sowings for the years 1971-2 and 1973-4) (Forbes 1982:365-70). Table 7.3 illustrates the long-term nature of this tendency to overproduce wheat by comparing some of the data in Tables 7.1 and 7.2 with an estimate of the household's actual consumption needs. The desire for cash sale does not, therefore, explain the significant tendency to overproduce wheat under normal climatic conditions any more than the coercion of surplus from the population does, since we have already noted that this latter phenomenon does not occur on Methana. Instead, the substantial net overproduction of wheat is directly related to the necessity of coping

Table 7.1. Household 30, Kosona: annual productivity of wheat, in sheaves per stremma (1 stremma =0.1 ha), for each plot cultivated. Even-year rotation. Plot

Plot size

no.

(stremmata)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

2.0 2.5 1.0 1.5 1.5 0.5 0.5 4.0 0.5 0.33 1.5 2.0 1.0 2.0 3.0

Coeff. of

Productivity in sheaves/stremma 1962

1964

1966

1968

1970

1972

1974

37 — 33 — 52 22 40 8.3 28 30.3 10.7 18.5 21 22 —

46 43.2 54 24 42.7 34 44 11 24 39.4 22.7 14 20 18 6.7

28 36 54 18 25.3 — 36 13.5 24 42.4 19.3 24 14 13.5 6.7

28 40.8 10 — 24 18 36 24.5 8 30.3 23.3 — 24 16 7.3

57 52 60 24.7 20 — 44 20.5 42 45.5 34.7 26 17 22.5 12

48 42 44 — — — 52 19 32 45.5 25.3 16.5 — — 13.3

42 52 54 17.3 18.7 — 36 10.5 36 27.3 17.3 17 10 7 —

variation 24.21 13.22 36.66 15.96 40.88 27.56 13.47 36.56 36.27 19.32 31.40 22.03 26.28 32.07 31.20

Mean coeff. of variation: 27.14 Total harvest (sheaves) Threshed wheat (kg)

511

600

501

462

713

1132

1861

1471

786

1908

489

493 ca.1500

16.11

93

Of grandfathers and grand theories

Table 7.2. Household 30, Kosona: annual productivity of wheat, in sheaves per stremma (1 stremma = 0.1 ha), for each plot cultivated. Odd-year rotation. Plot

Plot size

no.

(stremmata)

1 2 3 4 5 6 7 8 9 10 11 12

2.0 1.0 1.5 2.0 1.0 2.0 1.0 1.0 3.0 2.0 0.5 0.5

Coeff. of

Productivity in sheaves/stremma 1965

1967

1969

1971

1973

30.5 30 18.7 34 116 49.5 67 40 38 57 — —

31.5 52 13.3 30 110 45.5 60 32 36 35 — 16

24 66 35.3 33.5 118 55 46 34 30.7 34.5 144 8

24 66 14.7 48 66 46.5 71 48 23.3 22 196 30

5 38 14.7 — 78 40.5 25 16 21 24.5 160 12

variation 41.45 28.88 42.39 18.93 21.93 10.08 31.09 31.13 22.55 35.70 13.05 50.25

Mean coeff. of variation: 28.95 Total harvest (sheaves) Threshed wheat (kg)

737

674

779

737

468

2530

2190

1741

1780

1713

with uncertainties, particularly in the weather patterns. Only thus can a household expect adequate food supplies during inevitable poor years. Once again, however, security is 'paid for' by the extra work involved in producing such a surplus.

15 14 13 12 11 10 •

| |

87

5 • 4 3 2 •

-74- -49- -24- 0-24 25-4950-74 75-99 100- 125- 150" 175" 200- 225- 250- 275- 300-50 -25 0 124 149 174 199 224 249 274 299 324 Production over or under household requirements (in percent)

Fig. 7.3. Projected wheat production of households under normal weather conditions, compared to household requirements, Kosona 1971-4; all recorded cases. (Minus figures indicate a projected short-fall.)

16.30

The HRMs discussed here ensure that production is adequate for households' needs even in most poor years. But, as argued above, iffirst-defence mechanisms are a balancing act between the twin imperatives of efficiency and security, then it is inevitable that occasional climatic fluctuations occur when households can expect to lose their footing, as it were. The data in Table 7.3 show that over the 11 years documented, despite a policy of overproduction under average conditions, there were nevertheless two years in which production fell significantly below the household's needs. The comments of farmers on their harvests in the years between 1974 and 1985 suggest that two more exceptionally poor years have occurred during this period. Under such conditions, when the various first-defence mechanisms fail, the safety-net mechanism of storage strategies comes into play. Households aim to keep enough food staples (wheat and olive oil) in storage to ensure sufficiency if the following harvest should fail. In the case of wheat, this means a minimum of two years' supply (Forbes 1982:382-4). For olive oil it means four years' supply, because Methanites' olives, like those of many farmers in Greece, have a fruiting cycle which is both biennial and synchronised (Forbes 1982: 234-5). Storage strategies also have their own costs, however, since produce in storage deteriorates with time. Olive oil, in particular, suffers inevitable loss of quality over the years and thus this valuable asset (used by Methanites on occasions as a currency standard of a status equivalent to gold) loses cash value. Wheat is also liable to loss of quality during storage. Nevertheless, Kosonites prefer to retain food in storage and sell older produce at a lower price (if no emergency arises) as a means of ensuring adequate

Table 7.3. Household 30, Kosona: annual wheat production (in kg of threshed grain) compared with minimum household requirements, 1962-74. 1962

1963

1964

1965

1966

1967

1968

1969

1970

1971

1972

1973

1974

1132

?

1861

2530

1471

2190

786

1741

1908

1780

?

1713

c. 1500

Estimated minimum for household consumption: Estimated minimum seed-corn required: Estimated minimum threshing charge (in kind): Estimated milling charge (in kind)

1050 kg/yr 200 kg/yr 125 kg/yr 105 kg/yr

Estimated minimum total household requirement

1480 kg/yr

Of grandfathers and grand theories food supplies during the relatively rare occurrences of exceptionally poor years. Memories of the hardships associated with the German occupation of Greece during the Second World War provide examples of emergency mechanisms, which came into play in a situation where safety-net mechanisms failed. The Methana data show that at this level certain forms of behaviour contrary to the approved norms start to appear. The German occupation was a time when there was a genuine threat of death from starvation for some households on Methana. Responses included begging (which places people on a par with social pariahs like gypsies), eating animal fodder, and at least one instance of a poor family with many children marrying off a daughter without a dowry (social black mark number one), forcibly (black mark number two), to an ugly man much older than herself (black mark number three). This man, however, had adequate means to support the girl and thus freed the family of a mouth to feed. Such actions as these are not notably costly in energetic terms but they have high social costs in terms of the potential loss of family honour. (The maintenance of family honour is a concept of particular importance in Mediterranean societies, e.g. Peristiany 1965; Davis 1977:89-101). It might be tempting to accept Sahlins' (1974:127-30) view that such breaches of the social rules represent the appearance of anomie and social fragmentation caused by the loss of the social glue of surplus production. However, it is plain that the actions documented here, as well as comparable situations documented elsewhere (Sahlins 1974:127-30; Vayda and McCay 1975:301), can be recognised not simply as social breakdown, but as culturally patterned hard choices in the face of life-threatening hazard, for which lower-order HRMs are inadequate. Methanites are convinced that actions like these saved lives in certain circumstances. Methanites represent only an infinitesimal proportion of the world's population of peasant agriculturalists. That all such groups have comparable, though culturally specific, hierarchised repertoires of HRMs cannot be incontestably proved, but it seems reasonably safe to postulate it. There are both ethnographic case studies (e.g. Waddell 1975) and highly generalising and theoretical discussions of these phenomena (e.g. Rappaport 1979:145-72) which support this assumption. The Methana data, supported by those from other studies, suggest a generalising cross-cultural hypothesis concerning the characteristics of higher and lower levels of hazard response: Ij)wer-level HRMs will be characterised by: 1 2 3 4

continuous or frequent operation, which is energetically expensive. They are also well integrated into other aspects or subsystems of the culture or group so that they have low visibility qua HRMs, both to farmer and observer. Hence the 'grandfather response' (Alland 1975:70). Higher level HRMs will be characterised by:

1 2

infrequent operation and low energetic expense but

95 3 4

high social expense, since they tend to contravene normal social rules. The latter fact explains their high visibility both to' practitioners and to observers.

At first glance, this cross-cultural hypothesis might seem to stand the Halstead/O'Shea model on its head. This is not actually the case, however, for two reasons. (1) Different scales of analysis: whereas this chapter discusses responses only at the level of the individual household, Halstead and O'Shea are also concerned with larger social units up to and beyond the level of the polity. (2) Different terminological reference points: the terms used in this chapter ('efficiency', 'expense', 'costs') refer either to straightforward energy expenditure, or to social costs (loss of prestige or honour). Halstead and O'Shea, on the other hand, are also concerned with the frequency with which mechanisms are brought into play (i.e. cost/benefit), not simply with energy costs per se. The discussion so far has concentrated on the behaviour of individual households, with little reference to their articulation with the outside world. Consideration must now be given to the potential of larger social entities for creating and mitigating hazards in Methana. Thefirstthing to note is that the view Methanites have of the Greek state as it operates now, or has operated in the recent past, is of an entity anything but benevolent towards them. This is to a significant degree because the dealings that farmers in isolated villages have with the state apparatus are generally mediated through minor officials, the bases of whose decisions are either not fully understood, or else are considered to be arbitrary and capricious - and are in any case considered to be primarily to the personal advantage of those same officials. That such a problem in dealing with the state (and all other sources of power) is at least a Mediterranean-wide phenomenon can be realised from the system of patron-client relations which is endemic to the whole region and which attempts to control such sources of power to the benefit of the low-status individuals at the peripheries of the power network (e.g. Davis 1977:127-65). Secondly, direct experience over the years suggests to Methanites that the state is completely uninterested in their well-being in the event of a crisis. Even if on occasion help in dealing with crises is promised by the state, it often fails to materialise. For Methanites, the most memorable subsistence crisis this century was the period of the Second World War and the attendant German occupation. The state requisitioned large numbers of draught animals at the start of the war, leaving many households without a proper means of livelihood. This blow was further exacerbated the following year when crops failed owing to poor rainfall. Methanites are aware that large numbers of people throughout Greece, though especially in urban centres, died of starvation during the war years, because the state failed to intervene. At the same time they are aware that few if any farmers on Methana died of starvation: this entirely because of individual households' own successful responses to the crisis, not because of some coordinated action at the supra-community or even community level. There are two other indicators that the state has tradition-

H. Forbes

96

ally had little if any positive role to play in hazard response where Methanites are concerned. The first is the continued existence of the constantly operating lower level HRMs. Farmers' own selfinterest would ensure that such inefficient practices would not continue were they no longer needed.1 The second is the unwillingness of Methanites to be integrated more fully into the wider economy of the nation state, which reveals a lack of trust in the nation state as a reliable entity over the long term. A lack of commitment to a wider economic system, often seemingly unreasonable in the short term, is a recurring feature of farm households in most countries apart from the most highly industrialised nations of the western world. As Wolf has noted, despite the often unenviable life that many of these households lead, during wartime and depressions 'peasant holdings represent sanctuaries from the ravages which afflict people in cities and industrial centers . . . The peasant retains in his control of land and his capacity to raise crops on it - both his autonomy and his capacity to survive when others, more delicately dependent on the larger society, find such survival difficult' (Wolf 1966:17). The opposite side of this coin is that, both now and in recent centuries, the nation state has had difficulty in extracting much surplus production from Methanites if they do not wish it. In the past, villagers used to maintain store-rooms (ambaria) inconspicuously built into the hillside. (Though no longer used, some of them still exist.) Here the bulk of households' agricultural produce was kept: only supplies for immediate consumption were stored in dwellings. Thus, when the agents of the state (or other non-state power groups, like brigands) came to the community to collect surplus, they found little, if any. During the period of fieldwork from 1972-4, the modern nation state found itself unable to do much better. The government policy of forcing down the price of wheat resulted in a refusal by peasants in Methana and elsewhere to sell wheat at uneconomic prices. The drop in wheat supplies was aggravated by widespread rumours of impending major internal disruption within Greece, which caused many people, both farmers and nonfarmers, to hoard grain in the belief that, in the impending crisis, the state would fail to come to their aid. The combined effect of these actions resulted in quite serious bread shortages, which led to bread riots and more general political unrest in Athens. The threat of concerted action by a small but vocal section of the population close to the geographical centre of power forced the government to import wheat at above the home market price to alleviate the shortage (see also Jongman and Dekker, this volume). Conversely, the effect of the individual decisions of peasants and others in aggregate engendered incipient famine conditions in urban areas, despite the existence of adequate wheat supplies within the country as a whole. All the coercive powers of the military junta of the time were powerless to make households release their surplus grain supplies. For those attempting to identify state-level risk-reducing

mechanisms in the ancient world, the observations above may provide some suggestions for further consideration. The situation outlined here seems in substantial agreement with Garnsey and Morris (chapter 8). Polis governments in ancient Greece could do little to control the behaviour of their own food producers or distributors, and never developed a comprehensive institutional framework to protect the average individual from hunger and starvation. Adequate food supplies might be available within the polis territory, but the aggregate decisions of producers and distributors in withholding surplus stocks could result in food shortage. A second observation is that neither peasantry nor states make decisions. Decisions are made by individuals: hence the emphasis here on the farm household as the smallest individual decision-making unit, and the criticism made previously of attempts to identify peasant group dynamics.2 An important feature of the contribution by Garnsey and Morris is the discussion of the changing relationships between those at the base of the power pyramid and those at the top, as the ancient Greek polis develops over time. A similar element of change is built into the Halstead/O'Shea model (introduction to this volume). They argue that their higher-level mechanisms are characterised by a tendency towards embeddedness within the broader sociocultural system and so may provide a vehicle for important social changes. This observation is paralleled by Rappaport's (1979:145-72) view that responses at higher levels lead to permanent changes in the organisation of societies or communities. At issue here is the way certain HRMs become embedded into sociocultural systems, and their effects on those systems. When the scale of analysis is shifted from the state to the peasant household, a high degree of embeddedness is found to occur in the lower-level (hence, continuously operating) mechanisms, not higher-level ones. This means that, for the peasant household, those mechanisms which are highly embedded in other systems of society or culture are powerful factors for long-term stability and not for social change. Any change in the social systems within which these lower level HRMs are embedded will have an immediate impact on their effectiveness in buffering the population from the results of environmental variability. On the analogy of biological mutations, the chances of such effects being advantageous will be remarkably low. Hence the social and cultural institutions within which these HRMs are embedded can be expected to maintain considerable long-term stability. The particular characteristics of lower-level HRMs are thus a powerful explanation for the innate conservatism which has so impressed many of those who study peasant communities. (Studies which have explored the connection between risk and conservatism are reviewed by Barlett 1980a:556-7.) I would therefore argue that the HRMs that are well integrated within other institutions, which are identified here for peasant households, work differently from their

Of grandfathers and grand theories

97

counterparts at the state level, discussed by Halstead and O'Shea. Whereas at the state level they are seen as potential deviationamplifying devices, for the peasant household they are homeostasis-seeking types of behaviour, whose tendency is to maintain relatively unaltered the institutions in which they are embedded. Their success can be measured by their very longevity. The historical and archaeological record of Greece indicates a very long history for all the lower-level mechanisms identified from the Methana data (Foxhall 1986). The institutions of partible inheritance and dowry can be seen to extend back to before the Classical period, indicating a history of over two and a half thousand years for practices which encourage land fragmentation. Hesiod's poem Works and Days, written some time in the earlier first millennium BC, is an eloquent argument on behalf of increased energy expenditure as a way of surviving potential climatic hazards. More specifically, he seems to argue for just the overproduction and storage practices documented for Methana (Works and Days 30-2). Other ancient documentation amplifies this point (Foxhall 1986). Finally, present evidence suggests an Early Bronze Age date for the widespread establishment of olive cultivation - and, by implication, polycropping - in the Aegean area (Renfrew 1972:280-7). Thus, in this last case we are in the

happy position of being able to identify an HRM in the Aegean area with a pedigree of some 5000 years.

Notes The original Methana fieldwork was funded by a predoctoral fellowship from the University of Pennsylvania and a supplementary grant from the Department of Anthropology of that university. This paper has benefitted greatly from the comments of Lin Foxhall, Paul Halstead and John O'Shea. It has also benefitted materially from my being allowed to read a draft of the paper written for this volume by Peter Garnsey and Ian Morris. Guilt by association should in no way be attributed to any of these individuals for any shortcomings in this paper. 1 It might be argued, of course, that such practices as plot fragmentation continue as the result of inertia and the fossilisation of a now out-dated system. Such an argument would be to ignore modern anthropological research findings in the field of economic development generally, which emphasise that small farmers are neither irrational nor tradition-bound (Barlett 1980b:3). 2 That individuals make decisions does not contradict the emphasis on households. The farm household is the smallest decision-making unit, since its separate members tend to generate a common policy, followed and publicly supported by all (for Greece, see du Boulay 1974:19; Forbes 1982:22-3).

98

Chapter 8 Risk and the polis: the evolution of institutionalised responses to food supply problems in the ancient Greek state P. Garnsey and I. Morris

Hunger was never far away in ancient Greece. Conventional histories of the Greek cultural achievement rarely draw attention to the material bases of ancient society. In this chapter we aim to show how risk and uncertainty played a part in shaping the central institutions and practices of the Hellenic world, and how historians and archaeologists can use the Greek evidence to help them understand responses toriskin other complex societies. In particular, we will focus on the changing relationships between interannual climatic variability, population growth and the state from the eighth to the third century BC. During this half-millennium, two broad types of state can be distinguished within the Greek world, the polis and the ethnos. Most of our chapter will be concerned with cultural responses to variability and risk within the polis, a state formation based on the political relationship of citizenship.

Nothing is more shameless than an empty belly, which commands a man to remember it, even if he is sorely tired and pain is in his heart. (Homer, Odyssey 7:216-8)

I The polis and politics were invented in a land where agricultural producers laboured under certain natural disadvantages, related in particular to the seasonality and variability of production of staple crops. First, production in the Mediterranean environment is concentrated in large part in the winter months from October to March, which are mild and wet; summers are typically hot and dry and not conducive to the growth of the seed crops

which formed the basis of the diet of the ancients. Secondly, harvest fluctuations, associated above all with the variability of rainfall, are frequent and inevitable, if not precisely predictable. In Greece, rainfall is especially unreliable in the autumn and spring, the most critical periods in the growth cycle of winter cereals. The coefficient of variation at Kerkyra is below the high 60s percent only in March and November (62.5 percent, 45.7 percent respectively), and in Athens only in November (62.1 percent); in Thera it lies in the high 50s percent in February and March but otherwise is never less than 91.6 percent (in April). It is on the basis of such figures, combined with an estimated minimum rainfall requirement for crop growth, that one can produce a probability statistic for harvest failure for the places concerned. Thus in Attica (data from 1931-60), one of the driest areas of Greece, the probability of harvest failure in the case of barley (requiring a minimum rainfall of 200 mm during the season of plant growth) is 5.5 percent or about one year in 20, for wheat (requiring 300 mm) 28 percent or more than one year in four, and for seed legumes (requiring 400 mm) 71 percent or almost three years in four. Such calculations, based on long series of modern climatic data, may serve as a guide to past conditions - on the assumption, which has won general acceptance among scholars, that climatic conditions in the Mediterranean in classical antiquity and in the modern period are broadly similar, and in particular that interannual variability is a feature of both periods (Mariolopoulos 1962,1971; Gallant 1982:2-7; cf. Garnsey, Gallant and Rathbone 1984).

Risk and the polis To reach a fuller appreciation of the dimensions of the problem facing polis societies, we need to see these underlying physical environmental factors as interacting with the behaviour of men and of communities in specific historical contexts. Here demographic conditions come into play, notably the growth or decline of populations and therefore changes in the pattern of competition for largely finite natural resources. So do political relationships, internal and external: internal, the character of the relationship between social and economic unequals, the extent of exploitation and social unrest; external, the incidence of war and piracy. Against this background of a combination of natural and manmade constraints operating on the polis, we can construct a typology of social and political practices and institutions which might serve to head off or reduce the impact of subsistence crisis, and apply it to given historical contexts. The quality of the results to be hoped for from the application of this procedure will vary in accordance with the quality of the evidence. Thus, both the dimensions of the problem and the pattern of community response are more visible in the historical record for Greek society of the Classical and Hellenistic periods, that is to say, from about the beginning of the fifth century BC, than in an earlier period. The converse of this is that the challenge posed by the Archaic period in Greek history is the greater. It is not merely that the evidence is more exiguous. In addition the range of issues to be faced is wider. In particular there are the implications of the progression from pre-polis to polis society to be considered. What are the social, economic and political processes underlying this transition? How far did the emergence of the polis weaken the control exercised by noble houses over the distribution of the agricultural surplus, and how was this achieved? This relatively underexplored terrain may well prove as rewarding to survey as the more familiar landscapes of the later Greek world. But both need to be examined at least in outline if the investigation is to have any historical depth. II The government of a state society suffering from food crisis or seeking to avoid it had in principle a number of high-level options from which to choose, including the following: 1 Imperialism: to expand the territory under its control at the expense of other communities, nearby or at a further remove. 2 Extension of the cultivated area: to render productive more of its home territory. 3 Intensification: to seek higher returns from the same area of land by increasing labour input, employing more advanced technology or farming practices, introducing more productive crops. 4 Colonisation: to reduce the consumption needs of the community by exporting a proportion of its population. 5 Trade: to seek imports of staple food items through trade and other methods of exchange. 6 Regulation and redistribution: ad hoc measures, regular institutions, laws aimed at securing the situation of the more vulnerable citizen consumers through reallocation

99 of available foodstuffs. This might involve, at one end of the spectrum, 'social welfare' measures of various kinds undertaken by the political authorities or private individuals ('euergetism'), at the other end social revolution engineered from below. In the following section we consider the rise of the state in the eighth century BC and the use of these responses within its framework in the Archaic period (c. 700-500 BC). In section IV we discuss the Classical and Hellenistic periods (c. 500-150 BC), and in section V we draw out some of the implications of this study. Ill Any analysis of the early Greek state must begin with the collapse of the Mycenaean civilisation in the twelfth century BC. Its famous palatial centres were mostly consumed by fire around 1200 BC, and along with them a complex social system disappeared. The following period, approximately 1200-700 BC, is customarily termed the 'Dark Age'. The fall in numbers and sizes of known sites suggests that there may have been a 50-75 percent decline in population over the two centuries after the destructions. However, the prevalent tendency to see the Dark Age world as simple and egalitarian should be combated: the extent of the remains at Athens (covering 200 hectares), Knossos (100 hectares) and Argos (50 hectares) makes it clear that even at their nadir the population of the largest communities must have reached into four figures. Lower-level settlements with perhaps a few hundred inhabitants are also known (e.g. Asine, Karphi, Lefkandi), and larger numbers of tiny hamlets (e.g. Donoussa, Koukounaries, Nichoria). It is likely that the bulk of the population lived in small groups of no more than 30 to 50 people, but the continued existence of larger settlements, with all that that entails, must not be overlooked. Very little evidence for Dark Age subsistence strategies has been recovered. It has been suggested that the economy was primarily pastoral (e.g. Howe 1958; Snodgrass 1982:692-4), but the data available are woefully inadequate. Pollen analyses have so far yielded no decisive results. Seed evidence from Nichoria makes it clear that cereal agriculture did continue across the Dark Age (although the sample was very small: Shay and Shay 1978, Table 5.7), and this is supported by finds of eleventh-century bronze sickles from Karphi (Pendlebury 1937-8) and a probably contemporary pestle from Asine (Wells 1983:80). The faunal remains from Nichoria have been taken to indicate a shift in the use of cattle from providing milk to providing meat (Sloan and Duncan 1978), but it is important to note that this was based on a total of only 18 teeth found in the excavation. At Zagora it seems that cattle were being kept mainly for their secondary products (Cambitoglou 1981:81), while at Asine such evidence for meat-eating as was found came almost entirely from a single, probably ritual, context (Wells 1983:34). The scanty evidence does not allow us to establish what sort of balance there was between meat and cereals in the diet or how diet varied regionally, but the size of some of the larger settlements mentioned above certainly implies a heavily crop-based subsistence pattern in many parts of Greece. The evidence for trading contacts is ambiguous. There was

P. Garnsey and I. Morris

100

a decline after 1200 BC, and Snodgrass (1971:237-49) has argued that parts of Greece were largely cut off from outside contacts between c. 1025 and 950 BC, and that there was very little communication even within the Aegean in this period. Be that as it may, certainly by the late tenth century there is abundant evidence for travel within the Greek world, and it has even been suggested that the prosperous elite of Lefkandi were in direct contact with Egypt before 900 (Popham, Touloupa and Sackett 1982:247). By 800 there was a permanent Greek establishment at Al-Mina on the coast of Syria, and Greek pottery began to occur frequently in Italy and Sicily. The structure and organisation of Dark Age society are, if anything, still more controversial. The nineteenth-century view that early Greece was a 'tribal' society and that the state represents an evolution from blood to soil as the basis of the community has now been exploded (Bourriot 1976; Roussel 1976), and historians have argued back from eighth-century and later sources to suggest that society was dominated by aristocrats whose power rested on control of the land and their ability to attract dependents (e.g. Forrest 1966; Finley 1978). The archaeological evidence makes most sense within such a model. Throughout the Dark Age, burial practices emphasised a symbolic division of the community into an elite group, probably of nobles and richer peasants, and a nonelite of poorer peasants, whose dependence was represented through archaeologically invisible, informal disposal of the dead. Only the elite group had access to formal, bounded cemeteries. This situation only began to change in a few areas around 750 BC, with the admittance of the whole community to formal burial together in citizen cemeteries: in other areas, the integration may have come only in the sixth century; elsewhere, it never came at all (Morris 1987). Prehistoric archaeologists have often relied on Saxe's generalisation that restricted formal burial areas symbolise the monopoly of access to vital resources (usually the land) by unilineal descent groups (Saxe 1970:119-21; Goldstein 1981). This argument has been widely criticised (e.g. Hodder 1982:196-9, 1984:51-3), but the literary evidence from Classical Athens suggests that it has a certain relevance to ancient Greece. Having tombs in the ancestral cemetery was a necessary proof of descent, and, in order to inherit, an heir had to be able to show that his forefathers were buried in such a place (Isaeus 2.25,36-7; 4.19; 6.40; 8.21-37,39; Demosthenes 43.65; 44.32ff; Lysias 31.21). A fourth-century Athenian wishing to become a magistrate had to prove his citizen descent by pointing out his family tombs to an official (Aristotle, Constitution of Athens 55.3; cf. Xenophon, Memorabilia 2.2.13). The citizen body monopolised landholding, and it would be perverse to deny the relevance of Saxe's hypothesis. For these reasons, we would suggest that the dependency of the Dark Age peasantry was conceptualised as a form of serfdom, with the land somehow thought of as 'belonging' to the elite, who were partly or in some cases wholly supported by an upward flow of agricultural and industrial products and services. In keeping with the personalised nature of early Greek economic institutions, these flows were thought of as 'gifts' (Morris 1986:4-5).

There is little direct evidence for cultural mechanisms to override interannual variability. We can only assume that the elite's claim to the labour of their dependents was counterbalanced by responsibility for providing food in times of need, exploiting their capacity to store large surpluses and to tap surplusproducing areas abroad. Unfortunately, most excavation has focused on cemeteries, and large concentrations of storage facilities have not come to light. With the invention of the alphabetic script around 750 BC a whole new range of evidence becomes available. The most useful text for our present concerns is the Works and Days, a long didactic poem written by Hesiod of Ascra in Boeotia, probably in the early seventh century. Hesiod describes a wide range of low-level coping mechanisms within a society which still lacked the essential features of the polis (Ehrenberg 1937:155; Millett 1984:109 n.16, arguing against Luce 1978). Betraying a recognisably peasant mentality, he stresses direct storage within the independent household as the most honourable way to live, although reciprocity is also very prominent: 'It is good to take a measure from your neighbour, and good to pay him back the same, or better if you can, so that if you are needy afterwards you will again find him sure' (lines 349-51). There is no contradiction; Hesiod's maxims are typical of peasant wisdom from all over the world and all periods of history (see Millett 1984:93-103). He returns constantly to a few themes: work hard, fill your barns and storage jars, support yourself. After a string of Polonian maxims, Hesiod sums up his ideal: Thus the ears of your corn will bow down if Zeus himself finally gives you a good outcome, and you will sweep the cobwebs out of your jars; and I expect you will be happy when you take from the store you have gathered. You will have much until you come to bright springtime, and you will not look to others wistfully; but another man will need your help, (lines 473-8) Hesiod refers to seaborne trade as a source of gain (kerdos, line 632) and even wealth {chremata, line 686), but constantly stresses its dangers; he who goes sailing has a 'foolish heart' (646). Trade was surely held to be one way of mobilising a surplus, but in Hesiod it is presented largely in the guise of a high-level response to a shortage, which should be resorted to only in emergencies. He says his father used to voyage on the sea 'because he lacked the means of life', eventually migrating across the Aegean from Cyme to Ascra to escape 'wretched poverty, which Zeus gives to men' (634-8). Seafaring was an escape from 'debts and joyless famine' (647), whether for an individual household (this part of the poem is supposedly addressed to his brother Perses) or for an entire community (236-7). Any literary text gives only a partial view of what it describes. It is a model of and a model for society, in that its production implies action designed to influence the reproduction of social values. We cannot expect Hesiod to be an impartial, 'scientific' observer of his world, and Bravo (1977, 1983) and Mele (1979) have argued that he was in fact an aristocrat polemicising against nonnobles who involved themselves in trade. In this case, neither his-

Risk and the polis

101

torian's argument has a secure textual basis (Cartledge 1983; Millett 1984). But Hesiod's account of society does nevertheless seem very unbalanced, almost completely ignoring coalitions between unequals as contexts for high-level responses to scarcity. Yet without asymmetrical dyadic coalitions which function in this way no peasant society could survive (Wolf 1966:71-91). The only influence Hesiod's nobles are allowed over the community's subsistence is through their moral behaviour: if they are good, Zeus rewards them with agricultural abundance; if they are not, he sends dearth {Works and Days 225-47; cf. Homer, Odyssey 19.109-14). Otherwise, they appear only as 'gift-devouring lords' (38-9). Equally, we hear little about obligations to dependents, dmoes. Hesiod tells us what a man should give his dmoes during the 'normal' agricultural year (442, 559-60), but there is no reference to his responsibility in the shortages which must have occurred with unnerving frequency. Yet we must assume that the nobleman-wealthy peasant-serf chain functioned to provide higher-level flows of subsistence; it is impossible otherwise to explain the survival and reproduction ofpre-polis societies in the face of environmental constraints. The first important changes in the archaeological record come around 900 BC, with an extension of the settlement pattern into new areas. Demographic growth in the ninth to fifth centuries has been documented from Iran to Spain (Potter 1979; Adams 1981; Champion, Gamble, Shennan and Whittle 1984:251), and the first states began to appear in Mediterranean and temperate Europe from the eighth century onwards. The Greek developments can be understood only within this wider context, but the unique features of the polis are perhaps even more important than those features it shares with other Iron Age states. At the very moment of the appearance of archaeological indications of state institutions, the Dark Age systems of authority began to collapse in some parts of Greece. The processes at work are desperately obscure, but it seems likely that the Dark Age elite lost their ability to control the dependent 'serf population, and endemic rural unrest became social revolution (see Qviller 1981; Morris 1987, Ch. 10). The consequence was the emergence in some areas of a radically new form of state in the eighth century: the polis. In some parts of Greece the state took on the polis-structuve rather later; and in other areas the polis never appeared. The polis was a community in the strong sense, a group of men tied together by their political functions. This group of adult males, the citizen body, monopolised landholding and tended to endogamy. It embraced considerable inequalities, but the polis was its citizens: as far as they were concerned, there was an identity of state and society. Alcaeus, an aristocratic poet from Lesbos, summed up the idea around 600 BC: 'Neither finely roofed houses, nor the stones of well-built walls, nor even canals and dockyards make up the polis: but men do' (Lobel and Page 1955, Fragment Z 103). Serfdom was a direct contradiction of polis ideology and the overthrow of the Dark Age dependency was the defining feature of the polis as opposed to the ethnos as a state form. The ethnos was just as much a state as the polis: its leaders could raise armies, wage wars, make treaties, levy taxes, mint coins and pass laws just

like a polis, but the state was not centred on the citizen body. Serfdom was the norm in ethnos areas like Thessaly and in some intermediate forms of state like Sparta and many of the Cretan cities. The rejection of serfdom provided what Finley (1980, Ch. 2) has called the 'negative condition' for the emergence of the polis as the world's first true slave society, where chattel slavery became the normative form of surplus-producing labour. There were two sides to the central position of the polis, which roughly correspond to the classic anthropological distinction between social structure and social organisation. The polis provided a structure for social discourse, a set of guidelines to give meaning to experience; the/?o/w-relationship of citizenship allowed people to locate one another within an ideal structure. Citizenship meant that the polis was autonomous to an unparalleled degree, with the wellbeing of the citizen community replacing the wellbeing of any group within it as the ultimate good. Against this, we find that the state did not penetrate very far into society when it came to social organisation, actual day-to-day interaction. Many areas which the modern state controls were private in the polis. One such area was high-level cultural responses to risk and uncertainty. As Snodgrass has pointed out (1983a: 16-17), seafaring must have been very expensive in Archaic times, and Bravo's argument that the aristocracy controlled most long-distance trade, conducted by their agents, seems cogent to us. No doubt peasants could dispose of their own small surpluses over short distances and even out very localised or minor perturbations, transporting food by land or in small boats; but the larger ships which the wealthy could provide were both safer and more profitable, and many peasants may have allowed what surplus they could produce to be exchanged through elite networks. As Hesiod says, 'Praise a small ship, but put your produce in a large one' (Works and Days 643). The Archaic polis encompassed and redefined the pre-state exchange networks, substituting a new social structure rather than centralising distribution. The polis ideology was such that the elite should not use their control over distribution to create ties of serfdom through the poor's dependency, nor dispose of their surpluses overseas at a profit while citizens at home went short. There was therefore an essential tension in the polis, with the interests of the economically dominant group coming into conflict with the interests of the state. The balance between the two varied both spatially and temporally: at one extreme we have the Athenian democracy of the fifth and fourth centuries (and perhaps the Archaic tyrannies?), where the state was often positively antagonistic towards the elite, while other constitutions formed a spectrum gradually shading into the ethnos as the structural centrality of citizenship weakened in the face of aristocratic pressure. The reasons for this relationship between the polis and the material conditions of its reproduction seem to us to lie in the legacy of the pve-polis system of aristocratic society and in the ecological context of Archaic Greece. We would suggest that neither relative nor absolute population pressure ever constituted such a long-term threat to the survival of the community that the citizens of the Archaic polis were forced to create permanent, centralised institutions to provide cultural buffers; when the state did inter-

P. Garnsey and I. Morris vene, it almost always reacted to specific crises, the consequences of short-term fluctuations, and the state's involvement quickly passed away with the emergency itself. Only two hints of more permanent legislation are known from Archaic Greece, and it will be seen that both represent responses to human attacks on the polis structure, rather than to climatic or demographic pressures. Some historians have suggested that in fact population outran resources for long periods in Archaic Greece. The absolute overpopulation theory has been given a new lease of life by Snodgrass' argument that population growth reached 4 percent per annum in the eighth century (1980, Ch. 1; 1983b), while Camp (1979) has argued that a severe and very sustained drought in the eighth and seventh centuries created relative overpopulation. Both arguments rely on assuming a direct link between numbers of graves and ancient population levels, and are methodologically unsound (Morris 1987, Chs. 4-9). There is certainly evidence for greater use of most of the options described above, in section II, in the course of the eighth and seventh centuries - a great wave of overseas colonies, a sharp increase in long-distance exchange, several wars of annexation (Corinth/Megara, Sparta/Messenia, Chalcis/ Eretria, perhaps the destruction of Melie), the social revolution which overthrew dependency - but even the combination of all these does not give us a picture of a state dangerously out of balance with its natural environment so much as of a society involved in rapid expansion. Population pressure is not a 'prime mover'; it is itself part of the pattern rather than a monocausal explanation for it. Sahlins suggests that 'each political organisation harbors a coefficient of population density' (1974:131); the factors which encourage peasants to increase family size, and hence in the short term to lower the ratio of productive workers to non-productive consumers, are never simple (Caldwell 1981; Meillassoux 1983; Bulatao and Lee [eds.] 1983). Hesiod outlines the pros and cons: There should be just one son to feed the paternal house; for thus wealth will increase in the halls. But if you leave a second son you should die old. But on the other hand Zeus easily gives great wealth to more people. More people means more work and more increase' (Works and Days 376-80). This is not the place for a full discussion, and we will restrict our comments to colonisation, since this is generally taken as providing the most positive evidence for long-term overpopulation. In the fourth century, Plato (Laws 708B, 740E) suggested colonisation as a solution to population growth for his ideal state of precisely 5040 citizens. However, it seems that Archaic Greek colonisation was often a response to a short-term crisis, perhaps in much the same way as was the 'Ionian migration' into the Cyclades and Asia Minor during the Dark Age. The best-documented case is the foundation of Cyrene from Thera c. 630BC. Herodotus (4.151) says it followed a disastrous series of seven rainless years; the state intervened to compel one son from each household to sail to the new settlement. A fourthcentury BC version of the decree originally passed by the Therans further underlines the argument that such state activity was a response to a short-term crisis: anyone who refused to sail would be executed, but if the colony failed the settlers would be allowed to return to Thera after a period of five years, and to regain their cit-

102 izenship and property (Meiggs and Lewis 1969, no.5, lines 34-40; translated in Fornara 1983, no. 18). Less reliable later anecdotes about the foundation of Syracuse in 733 BC (Plutarch, Moralia 773A-B) and Zancle in the 720s BC (Strabo 6.1.6) also refer to drought and/or famine, which, given the climatic patterns of the Aegean, must again be interpreted as responses to short-term crises. In the less fully settled world of the Dark Age or in Archaic states like Attica and Boeotia where there were still in the seventh century opportunities to establish new settlements on reasonable soils, it was not necessary to react to shortages by sending people to the ends of the known earth (Herodotus [4.150] says the Therans were ignorant of the location of the site where the Delphic oracle had bidden them establish a colony - Libya). The eighth- to sixth-century colonisation movement, insofar as we can penetrate to its surely complex causes, suggests that greater organisation was possible within the polis and that the political, social and economic environment had changed, but not that there was a long-term population problem. The very scanty primary evidence for intervention by the Archaic state to resolve or head off subsistence crises serves to emphasise the extent to which high-level coping mechanisms generally remained outside the state's sphere of action. Apart from the foundations of colonies, where these can plausibly be linked with food crises, two regulations are worthy of attention, both tendentious, and both probably reflecting not a permanent imbalance between population and resources, but aristocratic manipulation of the distribution of available foodstuffs. First, an inscription from Teos in Asia Minor of c. 470 BC, that is from the period of transition between the Archaic and Classical Ages (Meiggs and Lewis 1969, no.30; translated Fornara 1983, no.63), lists curses to be repeated by the city's magistrates three times each year, among them the following: 'If into the land of Teos anyone prevents grain from being imported by any pretext or device, either by sea or by the mainland, or after it is imported repels (anotheoie) it, that man shall die, both himself and his family, (side A, lines 6-11). Bravo's interpretation in general outline is compelling: Clearly it is a case of rich and powerful aristocrats, who have sufficient physical means to prevent foreign cereals coming in which might be expected to reduce the scarcity of grain in Teos and the advantage they wanted to draw from it. Their behaviour is so deleterious to the interests of the community that it instructs its magistrates to pronounce a curse on them three times a year. (Bravo 1983:23) In brief, we are to envisage a bad year for crops on Teos (a common but not precisely predictable occurrence), aggravated by men with resources and power within the community. Unfortunately relatively little is known about the political background of late Archaic Teos. The Teians had abandoned their city in 545 BC to escape destruction by the Persian general Harpagus, and they set up new homes at Abdera and at Phanagorea in the Black Sea region; but it seems that at least some of them had returned to Asia Minor by 494 BC (Herodotus 1.168; 6.8; Strabo 14.1.30). These unusual circumstances could conceivably have contributed to the

Risk and the polis

103

ability of the aristocracy to resist the polis in the decades that followed. The involvement of men from Teos in the Greek trading post at Naukratis in Egypt (Herodotus 2.178) certainly indicates that an aristocracy with control over the distribution of foodstuffs was capable of disposing of their surplus abroad, and the language of a later inscription has been taken to suggest that the Teian nobility had controlled large estates worked by 'serfs' (Hunt 1947). This last point is not very well supported, but we should nevertheless be prepared to countenance the possibility that even in the early fifth century thepo/w/citizen state had not triumphed completely in Teos. The second case comes from Athens, the best-known seventh-century state. Drawing on the poems and laws of Solon, the sources (Aristotle, Constitution of Athens 2.2; Plutarch, Solon 13) report that serfdom rather than citizenship was the primary relationship within the state: the elite of seventh-century Athens had succeeded in their struggle against the ideology of the polis (Morris 1987). Solon accused the rich of bringing the state to the verge of ruin through their greed in exploiting the hektemoroi, dependents who worked the lands of the elite and handed over to them one-sixth or five-sixths of the produce. By the early sixth century, social revolution seemed imminent, and in 594 BC the aristocrats elected Solon as chief magistrate with powers to find a settlement. Plutarch (Solon 24) says that one of Solon's laws forbade the export of any Attic agricultural product except oil. Some have taken this as evidence that by 600 BC Attica was suffering from overpopulation and was dependent on regular grain imports from the Black Sea area. The notion that the Athenians of the time of Solon were unable to produce enough to feed themselves should be treated with the greatest scepticism. Modern historians have consistently underestimated the productive potential of ancient Attica (Garnsey 1988), and many have fallen into the normative trap of taking evidence for grain import as an indication of longterm subsistence problems rather than as a response to interannual variability. In fact, the direct evidence is very thin: it amounts to Xerxes' alleged sighting of Black Sea grain ships bound, according to Herodotus 7.147, for Aegina and the Peloponnese - not Athens. Similarly, Phrynon the Athenian's attempt to seize Sigeion from the Mytileneans in the late seventh century (Alcaeus, fragments H 28; X (7) 15-20; Z 105 (b) Lobel and Page 1955; Herodotus 5.94-5) is often interpreted (as indeed are other overseas adventures in which Athenians participated) as showing an early concern with the control of Black Sea grain trade routes. This is an improbable reconstruction. As Page pointed out (1955:158 n 2), Sigeion was poorly placed to influence such a trade route; no ancient source links the Athenian colony with interests in the Black Sea; and it is not even clear that Phrynon was acting on behalf of the state (Berve 1937:33-4; Frost 1984:288). In any case, there is no evidence that grain was exported from south Russia to Greece on a large scale before the fifth century (Noonan 1973).

terpreted in the same way as the Teian regulation, as an attempt by the state to limit the power of the large landowners to dispose of their food surplus overseas in circumstances in which the inevitable harvest fluctuations periodically exposed many domestic consumers to hunger, whether non-producing urban dwellers, or dependent hektemoroi already committed to running down their reserve fund for the profit of their landlords. The Solonian laws were a decisive step in the growth of the Classical Athenian polis.

Many of the so-called Solonian laws preserved in Plutarch are likely to be spurious, and we cannot be sure that the export regulation is any different. Supposing it to be genuine, it is best in-

IV Most of the relevant documentation for the Classical and Hellenistic periods concerns the final item in the typology outlined in section II above, state-directed laws, regulations and administrative practices designed to mitigate subsistence crisis - if we except imperialistic activity (often, though not invariably, with direct implications for food resources and supply), ranging from adventurism and aggrandisement by the more powerful Greek states, to petty territorial disputes between neighbours. No largescale colonisation occurred. Significant changes in land use within the territory of a city-state are hard to pin down in the epigraphical and archaeological record: the intensification which has been detected in fourth-century Attica (Jameson 1977-8) may be seen as a response to the challenge of maintaining high population levels (Hansen 1986) in the context of a much diminished empire and the increased insecurity of maritime transport. There is little documentation of trade in staples, in the limited sense of the import of essential food supplies, if we leave aside Athens, by far the best known but also the least typical polis. Few cities invariably raised enough grain for the needs of their populations, and it is obvious that each of them had developed over time regular trading contacts and networks to activate in lean years, even if little of this comes to light. When something of relevance does surface in the evidence, it does so in unpredictable ways: thus it was written into a treaty of 387 BC between Klazomenai, a modest town off the Asia Minor coast, and a resurgent Athens, that the former, in the event of food crisis (sitodeia), would still be able to seek help in certain specified cities, including neighbouring Smyrna (Tod 1948, no.l 14; translated Harding 1985, no.29). The role of the state in this matter of the import of essential foodstuffs requires further investigation. Fourth-century Klazomenai and Athens suggest in their different ways that public initiative and private enterprise complemented each other. Pseudo-Aristotle in the Oeconomica has this story about a food crisis in Klazomenai: The people of Klazomenai, suffering from food crisis and scarcity of funds, passed a resolution that any private citizens who had stores of oil should lend it to the state at interest; this being a produce which their land bears in abundance. The loan arranged, they hired vessels and sent them to the depots whence they obtained their grain, and bought a consignment on security of the value of the oil. (Oec. 1348bl7ff.) The matters of immediate interest here are the absence of a state

P. Garnsey and I. Morris merchant fleet and the presence of private traders and recognised potential sources of emergency grain. The existence of reciprocal exchange agreements, or any other form of official relationship between the city and its neighbours, must remain a matter of conjecture, in this case and in others. Fourth-century Athens too lacked a state merchant fleet (as did virtually all city-states in antiquity) and was dependent upon private traders who were furthermore for the most part not Athenian citizens. But the Athenians nevertheless took energetic steps to attract and bind traders to the service of Athens through a combination of diplomatic, imperialistic and legal measures. They negotiated favourable terms with the Bosporan kings, they supplied escort vessels when required, provided special courts for maritime disputes (dikai emporikai) and, on the other side, they imposed on traders using Athenian port facilities and financial backing the contractual obligation to unload their cargoes of grain nowhere else than at Athens on pain of death (Gauthier 1981; Garnsey 1988, Ch. 9). However, one thing the Classical Athenians never did was to establish a genuine grain market. This may seem to the modern mind a remarkable failing, but the political structure of the Greek city made such a step inconceivable (Finley 1970; Polanyi 1977:199-223). The economic institutions of the ancient world were a function of social practices (Finley 1985), and a simple decrease in transactions costs brought about by improved communications (North 1977) would not have been sufficient to 'disembed' the economy. This discussion has already drawn us into the area covered by item six in the typology, which concerns actions taken by governments to secure food for consumers. For these actions might include establishing or encouraging the establishment of a network of trade or exchange relationships to secure imports when needed, whether year by year or less regularly. There is one other detail to add before we move on to consider other aspects of official strategy, and that is the probability that there were recognised constitutional procedures for monitoring the food supply. The Aristotelian Constitution of Athens, composed in the second half of the fourth century BC, states that the assembly of democratic Athens had as standing items on the agenda of the principal meeting of the month (prytany) the following: 'to vote the confirmation of the magistrates in office if they are thought to govern well, and to deal with matters of food supply and the defence of the state' (43.4; cf. Xenophon, Memorabilia 3.6.13; Aristotle, Rhetoric 1.4.7,11). It is unclear whether this source is reporting a practice which antedates the mid-fourth century. In this period the food supply of Athens, a large importer of grain already for over a century, was less secure than in any previous epoch. It may be that a document from Samos of a little over a century later more faithfully reflects the situation in the 'standard' Greek polis: at Samos the food supply was reviewed every year in the month of Artemision, that is to say, in the spring, when presumably the size of the local harvest was roughly calculable (Sylloge 3.976; translated in Austin 1981, no. 116). If the prognostications for the harvest were poor, city governments had to hand by the second half of the fourth century or first half of the third a standard liturgy, the sitonia, which cir-

104 culated among the wealthier and more public-spirited members of the community. The sitones (or an equivalent official) had the job of searching out and acquiring grain in anticipation of or in the midst of shortages. Secondly, the public authorities were accustomed to evoke euergetism, or private benefaction. The sitonai typically worked in concert with private individuals, who put up either money for grain purchase or grain for sale at below-market prices if they had it - which they commonly did, as they were among the largest landowners. Euergetism, the public generosity of the rich, is the hallmark of i\iz polis in the Hellenistic age: with the universal disappearance of democracy by the end of the fourth century BC, it is the main safeguard of the common citizenry against hunger and starvation in a subsistence crisis (Veyne 1976; Gauthier 1985). There is another side to euergetism: the class which produced euergetists also produced speculators, and not infrequently they were one and the same. The inscription in honour of Polykritos of Erythrae points to a dual role of benefactor and hoarder: 'Later, when because of the grain shortage no-one was bringing grain in to the market, he promised the people to advance money for a reserve fund to those about to be appointed grain commissioners, and to bring into the market for feeding the people the wheat he himself held' {Inscr. Erythrae I Klazomenai 2%, lines 25 fl). Again, euergetism had definite limits: money was rarely given; rather, it was loaned, and often at interest (not, however, by Polykritos). Similarly, grain was sold, and profitably, even if at rates below the (elevated) market level. Priene's Moschion once actually gave grain (and twice offered it at reduced prices), a very rare occurrence in Hellenistic epigraphy {Inscr. Priene 108). Finally, the dependence of the cities on private benefaction only advertises the weakness of the public response to the inevitability of food shortage. In particular, contrary to the standard view (Tarn and Griffith 1952:107; Hands 1968:95), regular distributions of grain do not appear to have been a common practice in Greek states. The one such system which is well attested - according to which grain purchased with cash made available from a permanent fund set up by private donation was distributed to citizens - was established at Samos in c. 200 BC. Its duration is uncertain. More seriously, its effectiveness is problematic, since the amount of cash available appears to have been sufficient to purchase only a small proportion of the grain requirement of the citizen population. There was never any prospect that such an institution, whether fed by donation or indiction, would displace the essentially ad hoc and voluntary system of feeding the citizenry. Thus the elites ensured that politics would be dominated by competition among themselves for political and social power. As is implied in the discussion above, polis governments were not in a position to take a strongly independent stance against their own leading producers or for that matter distributors. There was no barrage of regulations protecting the consumer against speculation in foodstuffs. The most we can expect to come upon are ad hoc laws issued in a period of crisis regulating the movement of home-grown produce. Pseudo-Aristotle in the Oeconomica furnishes two instances involving the polis of Selym-

Risk and the polis

105

bria near Byzantium. The source first refers to a law passed by the people of Selymbria in time of food crisis forbidding the export of grain, and then continues:

cial revolution (this model is based on the general arguments of Theda Skocpol 1979). The polis was the typical, although not universal, Greek response to the need to harness the material and human resources of a larger and more complex society. The arrival of this particular form of community with its distinctive citizen ideology implied some loosening of the control exercised by the leading aristocratic families on the manpower and food resources of the community. To be sure, official responses to subsistence crisis, actual or threatened, were likely to reflect the aristocratic dominance of the state. Lower-class unrest was headed off or repressed, superfluous or unwanted members of the society were sloughed off in colonial ventures abroad, which incidentally gave leadership opportunities to individual aristocrats. On the other hand, central authorities were from time to time drawn into arbitration and regulation which restricted aristocratic abuse of power, if only to avoid the greater menace of tyranny. However, it appears that the Greek polis, neither in the first half millennium of its recorded history (the concern of this paper), nor subsequently, when it was absorbed into the Roman empire, developed a comprehensive framework of institutions and laws designed to protect the average citizen-consumer from hunger and starvation. It is true that we know nothing about possible institutional developments in democratic states outside Athens, which was active in this area, at least in the fourth century. The danger of generalising from the experience of Athens, a state with exceptional population levels, power and resources even when her influence was on the wane, is obvious. It is not unlikely that the standard Greek democracy of the Classical period was much more dependent than was Athens on the contributions of its wealthier members, whether in the form of one-off, ad hoc benefactions, or in the context of the liturgical system, that is, the regular services by which much of the religious, cultural and economic life of the city was sustained. In the oligarchic regimes which had displaced democracy everywhere in the Greek world by the last quarter of the fourth century, the crucial role was given to the elite, whether magistrates, liturgists or private benefactors. The citizens of the Hellenistic polis were beneficiaries (or victims) of the patronage (or profiteering) of the rich. In Rome, in contrast, private euergetism was held to be incompatible with the collective rule of the oligarchy, and subsequently with the one-man rule of an emperor. The Romans did develop a public grain distribution scheme for the benefit of the city of Rome itself, though for the first sixty-five years (123-58 BC) the grain was neither free nor available to all citizen-residents. But the distributions - and the taxes which fed them - post-dated and were a product of Rome's Mediterranean empire (Rickman 1980; Garnsey 1983a, 1983b).

On one occasion, however, they were in need of funds; and as they had large stores of grain, they passed a resolution that citizens should deliver up their grain to the state at the regular fixed price, each keeping for himself a year's supply. They then granted right of export to any who wanted it, fixing what they thought was a reasonable price. (1348b33fT.) A more complex structure of regulation is found only in democratic Athens, and it surfaces in the historical record not in the fifth but the fourth century, that is to say when the Athenian democracy was attempting to provide for a still densely populated city and territory against the background of a much reduced presence in the Aegean. In the typical oligarchy of the Hellenistic period, however, regulation was kept to a minimum, and to the extent that the gap was filled - and no governing class could have survived which was unprepared to feed the mass of the people in time of need - it was filled by euergetism. This is sufficient demonstration of the stranglehold exercised by Hellenistic elites over the political and social life of their states.

The emergence of the Greek polis is shrouded in mystery. Following the collapse of the centralised redistributive system of the Mycenaean palaces, power apparently devolved to petty lords, practising small-scale redistribution as part of a strategy of ironing out localised shortages. The word used by Homer and Hesiod for nobleman, basileus, probably derived from the Mycenaean word pa2-si-re-u, a relatively minor official in the Linear B tablets (Andreev 1979). The Dark Age basileis doubtless based much of their power on the need for coping mechanisms transcending the village level; a man starting out in the position of a chief would be in a far better position than any peasant to fit out a ship to travel for the purpose of finding or disposing of food, and would inherit guestfriends from his father (as Glaucus and Diomedes did in the Iliad, and Telemachus in the Odyssey), and thus ready-made exchange networks. Why this system broke down is not immediately obvious. One possibility is that, as the basileis extended their networks further afield, and as exchange opportunities with the Near East became easier following the expansion of the Assyrian empire under Ashurnasirpal II (884-859 BC), the possibilities for elite competition began to increase, together with the volume of 'gifts' demanded from the peasantry. Greater pressure on resources and frequent but unpredictable harvest failure, coupled with more intense warfare (see p. 102 above) could provide the context for a destabilisation of traditional power structures, culminating in so-

Note

We wish to thank Paul Cartledge for correcting some errors and tolerating our opinions.

106

Chapter 9 Monitoring interannual variability: an example from the period of early state development in southwestern Iran

Henry T. Wright, Richard W. Redding and Susan M. Pollock

This brief paper presents an example of a refuse-filled pit from a rural settlement of the Middle Uruk Period (mid-fourth millennium BC) in southwestern Iran, to illustrate the importance of, and potential for, recovering veryfine-grainedtemporal evidence from the archaeological record. The pit provides evidence of seasonal variation in production, consumption and administration. This information is used, in turn, to document the differing responses to successive good and bad years by the rural community.

Evaluating explanations of complex cultural processes often demands precise differentiation of temporal sequences of years, of seasons and - when we seek to utilise the remains of social, ritual or military action - even finer-grained event sequences. It is thus surprising that archaeologists interested in process still fail to analyse, or even to recover, such data. In fact, most archaeological records present opportunities to isolate samples deposited in short spans of time, and the techniques available for ascribing them to specific seasons or years are becoming more diverse all the time. The Susiana Plain region is an eastward extension of the steppe and alluvial desert environments of Mesopotamia proper. This plain contains more than 1400 km2 of well-drained cultivable soils; during the early periods discussed here, before the three major rivers of the plain entrenched their beds, much of this area could be irrigated with small canals. The plain has a winter rainy season when dry-farmed wheat or barley crops can be successfully grown two years out of three; the dry summer is extremely hot, and irrigation is necessary to sustain orchards, gardens and pas-

tures. To the northeast of the Susiana Plain, some 60 kilometres through a zone of barren gypsum and conglomerate foothills, are the ridges and higher valleys of the Zagros mountains, which provide rich summer pastures, as well as a range of woods, stones and other materials rare on the plain. French archaeologists have excavated for more than a century at the early urban centre of Susa and its immediate predecessors and dependencies on the west margin of the Susiana Plain. Since 1960, American and Iranian teams have concerned themselves with other centres and subsidiary settlements elsewhere on the plain and with various types of archaeological surveys. (See Adams 1961; Perrot [ed.] 1979; Carter and Stolper 1984; Dollfus and Wright [eds.] 1985; Hole [ed.] 1986 for overviews of these research efforts.) This work shows that, between 6000 and 4000 BC, there was an increasing density of small villages and hamlets, with the periodic florescence of small centres. Susa, the last of these centres to be founded, dominated the plain when this community pattern was in decline at the end of the fifth millennium BC. Its ten hectares or more of modest and elaborate domestic structures was dominated by a central mud-brick platform covering about one hectare, and rising to a height often metres. On this were storage structures, a large domestic structure and a possible shrine. There are indications that the elite of Susa controlled the storage and redistribution of goods there, but not that they had wide control on the plain or that Susa in general monopolised important crafts such as stone-working or the making of the elaborately painted

Monitoring interannual variability

107

Fig. 9.1. Uruk sites mentioned in the text

pottery which typifies this period (Wright 1984). After a poorly understood transition at the beginning of the fourth millennium, during the early part of the Uruk Period, Susa expanded to cover about 25 hectares, dominating a complex hierarchy of smaller centres and various kinds of rural settlements. Many craft activities were focused in the centres, as were newly developed institutions for storage and record keeping. At least by the Middle Uruk Period, most of the standardised plain pottery characteristic of the period was distributed from centres to smaller villages by several different means, and, at least by the Late Uruk Period, chipped stone tools were similarly distributed. The distribution at different kinds of settlements of seals and sealings, artefacts which allow control over access to goods and storage of information, indicate a control hierarchy with four levels of information processing. The Uruk transformation - and similar transformations around such centres as Nippur, Uruk and Nineveh elsewhere in Greater Mesopotamia - can be identified as the development of the first states (Johnson 1973,1978; Wright and Johnson 1975,1986). During the 1970s, much effort was devoted to documenting the economic and political organisation of the Susiana Plain during the Middle and Late Uruk Periods. One aspect of this effort was the excavation of Uruk remains at the rural site of Tappeh Sharafabad.1 Tappeh Sharafabad is an oval mound 130 m in length, and rising 11 m above the present surface of the plain, located on a

traditional route across the plain, roughly halfway between Susa and the smaller centre of Chogha Mish (Figure 9.1). We chose it for excavation because surface examination suggested that it had accessible Middle Uruk levels, and that there had been a building on the site decorated with clay cones, a marker of public buildings of various types. We hoped to find evidence of an administrative building. Unfortunately, most of the Middle Uruk deposit at Sharafabad had been disturbed by later occupants, farmers of the Sukkalmahhu Elamite period (Schacht 1975), and only a few Uruk rooms were intact; we therefore devoted our slim resources to the excavation of pits and middens, in the hope of at least recovering samples of animal bones, plant remains and domestic artefacts from a rural settlement. One pit, four by ten metres in extent on the present eroded north flank of the site, proved exceptionally rich in bone and unbaked clay sealings and other administrative artefacts. One day, Redding, archaeozoologist in charge of its excavation, reported that he had found an intact toad skeleton in a waterlain mud lens. 'So what', replied Wright, the frazzled director. T h e toad', replied Redding, 'probably entered the pit at the end of the rainy season, seeking a place to estivate (the opposite of'hibernate'), only to be sealed in forever by basketloads of broken brick and pottery thrown in during the summer.' Redding continued the initial one-metre wide test trench to a depth of four metres, measuring the volumes of each layer excavated so that the quantities of artefacts could be expressed as densities, and

Henry T. Wright, Richard W. Redding and Susan M. Pollock

108

Weathered silt

Mud-brick fragments

Pebbles

• .•..

Charcoal and ash

Greenish silt

* * ft

Silt with gypsum crystals

Bedded water-laid silt

// // //

Burned reddish silt

Burned construction fragments

* \

Sherds

One Metre Intact bevelled-rim bowls

Fig. 9.2. Section of Uruk pit (Feature 3) at Tepe Sharafabad dry-sieving all measured volumes. Wet-sieving would have been more thorough, but we feared it would destroy the unbaked clay administrative artefacts. The section revealed a succession of 26 layers (Figure 9.2), which could be aggregated into five broader groupings. 1 2 3

4 5

Layers (26) to (25) are composed of silt with lenses of sherds and charcoal-flecked ash. Layers (24) to (18) are composed of ashy silt with small mudbrick fragments and lenses of sherds. Layers (17) to (10) are deposits of ash and charcoal, alternating with layers of greenish silt, neither of which had mudbrick fragments; there was one lens of water-lain silt on the west edge of the pit. Layers (9) and (8) are composed of mud-brick fragments and silt with lenses of sherds. Layers (7) and above are weathered silts, some with charcoal and ash; there were two lenses of water-lain silt on the west side of the pit, under one of which were the remains of the unfortunate toad.

Two observations of the enduring realities of traditional rural life in Mesopotamia allowed us to interpret this section in terms of

seasonal succession. The first is that, during summer, there is much construction and repair in mud because mud brick and plaster cure well in the dry summer heat; it is difficult to make durable mud construction during the winters. The second is that during the damp, chilly winters there is a great need for fires; during the summer only the most essential fires, such as those for bread baking, are lit. Therefore, we can infer that layers with much brick debris, but little ash and charcoal, were deposited during the summer; and layers with much ash and charcoal, but little brick debris, were deposited during the winters. Given these criteria, Layers (26) through (18) would have been deposited during a summer and perhaps part of the preceding winter, Layers (17) to (10) would have been deposited during a winter, and Layers (9) and above would have been deposited during a summer and some part of the succeeding winter. Obviously, there was a need for independent testing of this construct, so we decided to use our dwindling resources to make a second and more precise 0.5 metre slice, using the section as a control, and taking many flotation and pollen samples for archaeobotanical study. Unfortunately, the sediments have not yielded pollen, and the seeds recovered do not permit seasonal assessments. Our best evidence of seasonality to date is that provided by the ages of the mandibles of very young sheep

Monitoring interannual variability

109

and goats, as indicated by tooth eruption. Assuming birth between November and January (a sensible strategy given local climatic stresses), the five mandibles from Layers (24) to (18) indicate slaughter in June to October, the three mandibles from Layers (17) to (12) indicate slaughter in November to January, and the two mandibles from layers (8) and (7) in September or October. The inferred slaughtering dates conform well to the seasonal ascriptions based on the sediment characteristics. The early stages of our analysis and a preliminary paper (Wright, Miller and Redding 1980) were focused on seasonal variation in domestic activities, agricultural production and articulation with the central administration. Even though we noted differences between the two years represented in our cut through the pit, first in the discard in certain pottery forms (Wright et al. 1980:274) and secondly intypes of administrative artefacts (Wright et al. 1980:278), questions about interannual variation were not asked. Even if one does not heed the Biblical story of Moses and the seven years of plenty and the seven years of famine, the historical and ethnographic records of traditional states amply illustrate the concern of administrators with the maintenance of their own infrastructures in spite of year-to-year variations, and modern environmental fluctuations in Mesopotamia underline the importance of such changes for the daily life of rural people. This further preliminary paper addresses the question of interannual variability using the evidence of the pit. We divide the year around midwinter, that is between Layers (13) and (14), primarily because this is the time when it first becomes evident to prudent farmers in Mesopotamia that the crops may not do well, and when they should begin to make compensatory decisions.

Table 9.1. Bone counts and densities by taxafor two years

Agricultural production

The Middle Uruk community at Sharafabad of around 3300 BC was involved in the cultivation of wheat and barley, and may also have raised lentils and linseed (Wright et al. 1980:275-6). Unfortunately, Naomi Miller's detailed study of the flotation samples shows that both the attenuated pathway by which carbonised plant remains entered the layers in the pit and the small size of the samples which we processed prevent useful statements about yearto-year variation in plant foods. One might think that the discard of sickle blades through time would indicate the amount of labour expended in harvest, but unfortunately the seasonal data show that those we found were discarded into the pit either in late winter just before harvest or in late summer after threshing. Most sickle replacement must have occurred in the fields during harvest in early summer. It is interesting to note that the maximum attested sickle discard into the pit is just before the harvest of the second year, perhaps indicating that sickles were heavily used during the previous harvest of the first year. Unfortunately, since we do not have clear evidence of the layers of late in the third winter, informing us about the harvest of the second year, we cannot safely use sickle discard to assess the relative amount of harvest labour during the two years. In short, we have no quantitative way to monitor cereal production directly. The Middle Uruk community at Sharafabad was also involved in the consumption of herded sheep, goats, cows and pigs, as

Bone counts

Sheep/goat Sheep Goat Cattle Pig Bird Fish

Bone densities

Year I

Year II

Year I

Year II

951 74 52 14 7 201 237

249 19 12 11 4 14 22

158.1 12.3 8.6 2.3 1.2 33.4 39.4

43.3 3.3 2.1 1.9 0.7 2.5 3.9

Note: These bone counts include samples from both slices through the pit, totalling 1.5 metres in width; the volumes are 6.02 cubic metres for Year I and 5.67 cubic metres for Year II. well as in fowling and fishing. Meat of most of these taxa, except for fish, was probably not conserved. We have no evidence of meat conservation in this area, neither archaeological, historical nor ethnographic. Hence, the bones would be regularly thrown into the pit in baskets of household refuse shortly after consumption. Though many bones probably never reached the pit, Redding's continuing studies indicate no differential bone selection, fragmentation or gnawing marks, suggesting that scavengers rarely entered the pit and removed what was deposited during the two years under consideration here. Table 9.1 presents the actual occurrences of identified bones and the bones/cubic metre for the two years under consideration. These figures indicate a drop in all forms of bone discard in the second year. Consumption of sheep and goat, the major meat sources in the diet, appears to have been about 26 percent of their level in the first year; cow and pig consumption were 80 percent and 60 percent respectively; fish and wildfowl consumption were less than 10 percent of their former levels. Table 9.2 and Figure 9.3 present the evidence of the ages at which sheep and goats were slaughtered. Keeping in mind that this is an aggregate of the somewhat different summer and winter slaughtering strategies (Wright et al. 1980:276-7), note that in the second year there is a surprising increase in the mortality rate of

% Surviving

6-8

13-16 18-24 Age (months)

30-36

Fig. 9.3. Survivorship curves for sheep and goat at Sharafabad

Henry T. Wright, Richard W. Redding and Susan M. Pollock

110

Table 9.2. Sheep and goat fusion data by element for two years Element

Age of fusion (mo)

Prox.radius Scapula Dist. humerus 1st phalanx 2nd phalanx Dist. metapod Dist. tibia Ulna Prox. femur Calcaneum Dist. radius Prox. humerus Dist. femur Prox. tibia

6-8 6-8 6-8 13-16 13-16 18-24 18-24 30-36 30-36 30-36 30-36 36-42 36-42 36-42

Year I

Year II

fused

unfused

index

fused

unfused

index

13 7 9 15 26 15 6 2 4 2 3 2 3 1

2 3 4 30 19 28 13 8 9 15 9 6 9 7

0.86 0.70 0.69 0.33 0.57 0.35 0.32 0.20 0.31 0.11 0.25 0.25 0.25 0.12

7 5 4 1 2 2 3 1 1 1 — 1 0 0

0 0 0 3 2 7 2 2 3 2 — 1 1 1

1.00 1.00 1.00 0.25 0.50 0.22 0.60 0.33 0.25 0.33 — 0.50 0.00 0.00

animals between 6-8 months and 13-16 months, but that few older adults were being killed. On the assumption that the animals would have been allowed to bear in early winter in this region, such mortality indicates a strategy of eliminating small or weak animals (even females, which one would normally conserve to ensure future breeding stocks), and of conserving proven breeding adults, in late summer. This strategy suggests that the herders realised that there would not be adequate fodder in the immediate future. In brief, the strategy of culling domestic stock and the decreased quantity of the culling in general indicates that the second year was one of agricultural crisis. The fact that the decrease in food supply from the domestic bovids was not matched by increased fishing, fowling or pig hunting - all of which could normally be locally pursued in wetter areas - suggests that these areas were desiccated, indicating a drought. This is an idea which might be tested in the future with studies of tree-ring growth, but, for the sake of argument, let us provisionally accept it and consider the impact of a drought year on village life and on the village's relations with the broader region. Domestic life Basic activities such as food preparation and serving, with some salient exceptions, do not appear to vary with changes in the agricultural economy (Table 9.3). The grinding stones used in the ceaseless daily preparation of flour show no change in density from year to year. The coarse-bodied round-rim jars, which often exhibit fire clouding indicating use in cooking, show little change in density. The various smaller jars, often spouted, and mediumsized bowls, associated with the serving of liquids and foods, similarly show little change. What do change are aspects of the domestic technology that related to the broader economy. First, the large band-rim jars, some of which we know arrived

from outside the community with sealed contents,2 are more than twice as common during the second year. Secondly, the discard of unbaked clay spindle whorls, which must find their way into a protected context quickly before they disintegrate, is relatively uncommon during the second year, suggesting that the spinning of thread, a late winter/early summer activity, was not much purTable 9.3. Domestic artefact counts and densities by year Artefact counts

Ceramic Vessels Round-rim jars Flat-rim jars Band-rim jars Round-rim bowls Flat-rim bowls Bevelled-rim bowls Conical cups Other artefacts Sickles Grinding stones Unbaked clay whorls

Artefact densities

Year I

Year II

Year I

Year II

29 31 5 10 6 3583 36

40 34 10 11 1 1391 19

15.4 16.4 2.6 5.3 3.2 1906 19.1

17.9 15.2 4.5 4.9 .4 624 8.5

3 9 9

10 7 3

2.1 4.8 4.8

4.5 3.1 1.3

Note: These counts include samples only from the second slice through the pit, 0.5 metres in width; the volumes are 1.88 cubic metres for Year I and 2.23 cubic metres for Year II. The counts are actual item counts except for those on the bevelled-rim bowls, which are rimsherd counts because it is difficult to match these sherds into vessels. One can make an adjustment by assuming that each bevelled-rim bowl breaks into an average of nine rimsherds.

Monitoring interannual variability

111

sued, perhaps because the surviving animals were in distant, lusher pastures. Thirdly, the quantity of small bowls, both conical bowls and the crude, handmade bevelled-rim bowls, which constitute 90 percent or more of all ceramic samples, drops to less than half its former level during the second year. The evidence from the Susiana and elsewhere supports the idea that these bowls are associated with mass labour projects, being used to issue rations or meals to workers (Nissen 1970; Johnson 1973:129-39; Le Brun 1980). If so, a lesser number of these bowls in refuse on a rural site is not informing us about local domestic life, but about either the decreased participation of villagers in mass labour elsewhere, or the lessened number in the Sharafabad area of labourers from elsewhere working on a mass project, or both, during the first year. Whatever the cause, the opportunities for recycling these mass-produced vessels in rural households were greatly reduced during the year suggested to be one of agricultural crisis.

Table 9.4. Chert artefact counts and densities by year

Craft production and exchange There is very little evidence for production other than agriculture. The limited occurrences of ceramic slag, spindle whorls, stone drills, etc. need indicate nothing more than limited production of ceramics, fibre products or wood products to meet local needs. However, the occurrence of quantities of items of unknown use, for example the rough limestone flakes rarely found in other Uruk sites, but common at Sharafabad, could indicate an unknown craft. In addition, there are crafts, such as mat weaving, which leave few durable traces. A craft which merits discussion is the production of chipped stone tools, particularly blade tools, the object of a special study by Pollock. Several kinds of usable stone occur in southwestern Iran (Wright 1981:266-7). Small pebbles of red and green jasper are ubiquitous, but tools of these stones suffer the disadvantages of being both brittle and quickly dulled. 'Medium Gray Chert' of coarse texture occurs in large cobbles in gravels on the northwest margin of the Susiana Plain and produces durable tools. The inhabitants of Sharafabad would have to walk about 25 km, fording the Dez River, to obtain cobbles of Medium Gray Chert. Imported raw materials, including fine cherts from as far away as the Syrian plains or the inner Zagros mountains and obsidian from Lake Van, could also be obtained. The pit contained both core working debris and discarded tools of both local and exotic materials (Table 9.4). Medium Gray Chert debitage was discarded during the winters. However, the quantities of decortication flakes and blades, discarded when cobbles are shaped into blade cores, are not present; instead, most of the debris is from the working of already prepared cores. The Medium Gray Chert cores found at Sharafabad must have been prepared at the quarry or at another settlement. The discard of tools of this local chert, in the form of blade segments, some of which were used as sickle blades, triples in the second year. There is also less use of the blade tools, as indicated by the number of edges utilised per piece before discard. At first it seemed to us curious that, in a difficult year, a material obtained

Artefact counts

Artefact densities

Year II

Year I

Year II

Medium Gray Chert Debitage 3 Plain blades 7 Utilised blades 10 1 Sickles — Backed blades

4 20 25 9 1

1.6 3.7 5.3 .5 —

2.2 9.0 11.2 4.0 .4

Exotic cherts Debitage Plain blades Utilised blades Sickles Denticulate/notch Burin Perforator

2 3 20 1 4 — 1

.5 .5 2.7 1.1 .5 .5 —

.9 1.3 9.0 .5 1.8 — .5

Year I

1 1 5 2 1 1 —

See note for Table 9.3

with some effort, albeit locally, would be used profligately, but this may be an ethnocentric interpretation. If the second year was indeed a poor agricultural year, the farmers of Sharafabad would have had plenty of time to engage in craft activities, either for their own immediate benefit or for exchange. However, we have not yet devised ways to evaluate these ideas. A few pieces of exotic chert debitage were discarded during the summers, in contrast to the local cherts. The discard of exotic chert tools, both as blade segments and such tool types as notched pieces, a perforator and a burin, also triples in the second year. Degree of utilisation, as indicated by the number of utilised edges per piece, does not change. Apparently, a somewhat greater quantity of these materials was obtained to facilitate an increase in certain cutting tasks. Future microwear studies may allow identification of these tasks. In summary, the inhabitants of Sharafabad seem to have obtained supplies of exotic materials for their domestic and productive tasks, in spite of the problems of the particular year. During the agriculturally difficult second year, they did engage in somewhat more petty craft activity, but to what extent this was for exchange rather than local replenishment is unclear. Administrative control It is a priori unlikely that a small agricultural community situated in a network of larger centres would be independent of these centres. Indeed, the evidence that domestic life was largely unaffected by agricultural difficulties, can be taken to indicate the existence of an authority which buffered the populace against dis-

Henry T. Wright, Richard W. Redding and Susan M. Pollock

112

aster. Though the remains of the building with clay cone decoration, perhaps the locus of an institution that directly represented the central authorities, were not found, we do have compelling direct evidence of central control over production and storage at Sharafabad. In the Middle Uruk pit, we found a range of artefacts which may be termed, sensu lato, 'administrative'. Among these were some of the tiny geometric counters that represented quantities of various goods (Amiet 1972; Schmandt-Bessarat 1977; LeBrun and Vallat 1978). We also found what may be fragments of unused clay envelopes, which would have been wrapped around sets of these counters and sealed to make a 'bulla', a closed record of a quantity of goods. However, we found no examples or fragments of actual sealed bullae, which are common at centres such as Susa. The implication is that transactions - whether the aggregation of goods for shipment, the local storage of goods or the local transfer of goods from one individual to another 3 - were recorded at Sharafabad, but that the records were sent elsewhere. Contrasting with such information-bearing sealings are the commodity sealings, which prevented unauthorised consumption of goods, a more common type of administrative artefact. Typically, these are small lumps of mud which have been placed over the knot on a cord holding the top on a bale, a basket or a jar full of goods or tying a storeroom-door peg to a wall peg. Some are lumps of mud which have been stuffed into the mouth of a small jar or bottle.

Table 9.5. Administrative artefact counts and densities by year

The seal of the official or individual who authorised closure had been impressed into the mud while it was still soft. Though only one very simple seal was actually found, 12 different seals are attested as impressions on sealings from the pit (Wright et al. 1980, Figure 6). Of these 12, six are not very useful, occurring as either very fragmentary or amorphous impressions. Of the remaining six, three are of different cylinder seals, each occurring only once: one on ajar, one on a bottle, and one on a door sealing. A fourth may have been a cylinder seal; its impressions occur on three bale sealings. These four seals were probably rarely, if ever, actually at the settlement. The other two were thematically similar stamp seals, showing an animal leaping over a kneeling human. Their impressions occur on both door sealings and bale sealings, and were probably seals held by individuals frequently involved with the settlement, though we cannot say that they were resident there, and it is possible that the seal-holders were not personally present when their seal was used. The consideration of year-to-year change in the seals used helps to clarify the position of the seal-holders. In particular, impressions of one of the two animal-human stamp seals occur on bales from the first year and a door sealing from the second year; it is oval in shape and shows much wear from protracted use. Impressions of the other animal-human stamp seal occur on bale and door sealings from the second year; it is rectangular in shape and is little used. The oval seal must have been replaced by the rectangular seal. The most parsimonious interpretation is that the motif was associated with an individual, who - finding his or her seal increasingly worn - had a new one cut with a similar motif. Another possibility is that the motifs were associated with institutions, and that one minor official replaced another. With additional cases,

Artefact counts

Artefact densities

Year I

Year I

Year II

Year II

Informational Counters Envelope fragments

3 8

3 3

.5 1.3

.9 .5

Storage Door sealings Bale sealings Jar sealings Bottle sealings

5 6 1 1

13 8. 3 1

.8 1.0 .2 .2

2.2 1.4 .5 .2

See note for Table 9.1.

the details of seal use, critical in evaluating overall administrative organisation, may be defined. Consideration of the overall proportions of different types of sealings allows us to evaluate year-to-year changes in the articulation between the village economy and the regional economy (Table 9.5). The opening of bales, which either came from elsewhere or from local storehouses, and of bottles, which probably came from elsewhere, increased slightly in absolute amount, during the second year. In contrast, the recording of quantities and sending out of information was relatively greater during the first year. Whether this was recording of goods locally stored, goods transferred from one individual to another, or goods shipped out with travelling invoices, we do not know; however, there were apparently more economic transactions to be recorded. In contrast, the opening of storerooms was relatively much more common during the second year, particularly during late winter and early summer, when domestic grain supplies would be running out. Apparently there was relatively more local need for stored goods, probably foodstuffs. In short, the evidence from the pit indicates that, during the year of difficulties in the agricultural economy, there was less recording of information about transactions, probably about the movement of agricultural products, and there was more opening of storerooms, probably to make food supplies available. This provides an indication of how domestic life could have continued with little change, even though domestic and wild animals were in short supply and there were other indications of agricultural crisis. While there is some indication of an increase in local petty craft activity, the basic sustenance of the village was ensured by the opening of centrally controlled storehouses. We do not yet know whether the institution or institutions whose officials sent goods in bales and jars to Sharafabad and authorised the sealing of storehouses were temples, palaces, estates or something yet unknown; we can say, however, that interannual variability was the concern of a hierarchical organisation with the authority to control both the storage of products and the information about such storage.

Monitoring interannual

variability

Concluding note

In summary, this presentation of the evidence from Sharafabad indicates that, in a good year, the inhabitants of this one rural community worked at their agricultural tasks and at mass-labour projects, that they stored the grain they did not consume both locally and elsewhere, that they prudently culled a large number of animals from their herds, and that they engaged only in local craft activities. In contrast, in a bad year - probably a drought - they did less agricultural work and mass labour, they utilised grain locally stored under the aegis of central administrators, the bulk of their flocks were elsewhere but they drastically culled the remaining animals, and they engaged in some unusual but minor craft activities. It is possible that it was the availability of stored grain which allowed the villagers to pursue a prudent, security-maximising strategy for their herds (Redding 1981), rather than covering the year's grain shortfalls by killing more domestic animals or by increased hunting and gathering. This example of the use of short-term depositional sequences to answer questions about the response of complex cultural systems to crises is based upon a single pit. We would like to confirm and expand upon these results with evidence from similar features in other rural sites, and to complement them with evidence from such features in the central towns. Would sampling from similar pits at Sharafabad, or even further slices sampled from the same pit, produce comparable evidence? How do other types of rural communities respond to such crises? How do rural communities respond to agricultural plenty and agricultural crisis when central control is weak? How do central institutions and elites act in such situations? If such questions about actual past processes are not asked, archaeologists will not be motivated either to excavate trash deposits carefully, as opposed to the buildings that remain their primary focus today, or to devise new methods of analysis for the evidence from such deposits. A few years ago, the director of the Sharafabad excavations

113 gave a lecture on this single pit filled with stratified garbage at a well-known archaeological institute. After most of the audience had dispersed, one of the students reported the latest news received from a friend participating in an excavation on an important Mesopotamian site: several similar pits, filled with carbonised debris, bones and sealings had been found. That's good news', said the lecturer. 'Yes', sadly replied the student, 'but the bad news is that each pit was excavated as a single locus.' This young archaeologist will not throw away such possibilities; I hope that others, as well, will heed this example. Notes 1 The work at Sharafabad was funded by National Science Foundation grant No. GS-3417. It was facilitated by the former Archaeological Service of the Government of Iran, then under the direction of Mr. A. A. Pourmand, represented by Mr. M. H. Khoshabi. The authors are grateful to our colleagues at the University of Michigan Museum of Anthropology and at many other institutions, who have provided assistance and critique of the ongoing analyses of the data from Sharafabad. This manuscript further profits from specific suggestions about anthropological economics from Paul Halstead. 2 An unpublished neutron activation analysis by M. James Blackman of the Conservation and Analytic Laboratory of the Smithsonian Institution shows that, while the door sealings and other items are of local clay, one of the sealings impressed against the rim of such a jar is of a non-local clay. 3 When these data were presented at the CNRS Colloquium of 13-15 June 1978 in Paris, we interpreted the evidence that bullae were prepared at Sharafabad, but sent elsewhere, as indicating that they were 'travelling invoices' sent to centres with shipments of goods. Alain Le Brun and Francois Vallat of the Susa Mission, have argued that the bullae they have excavated at Susa may be records of transactions or contracts between individuals (Le Brun and Vallat 1978:34-7). Until Sharafabad's bullae are found, either at Sharafabad or elsewhere, one cannot decide between these possibilities. For this reason, we leave the issue open in this summary.

114

Chapter 10 Public intervention in the food supply in pre-industrial Europe

Willem Jongman and Rudolf Dekker

The history of pre-industrial Europe shows that public authorities have almost invariably interfered with the urban food supply. In this paper, it is argued that there were sound reasons for such intervention. Fluctuations in supply would - in a free market - inevitably have led to enormous fluctuations in the price of the staple food. The mass of the urban population would not have been able to cope with these price fluctuations and would have demanded effective measures on the part of the authorities. Theoretically a number of different types of measures can be predicted, and they can all be documented from the historical record. Specific historical circumstances, however, determine which coping strategy will be prevalent. A comparative analysis of Classical antiquity (in particular Imperial Rome) and the cities of late medieval and early modern Europe demonstrates that costs and benefits of particular coping strategies may differ considerably. It also demonstrates that a purely economic analysis is not sufficient. The nature and extent of state power is at least as important, and so are changes in popular expectations.

In the Digression concerning the Corn Trade and Corn Laws in the Wealth of Nations, Adam Smith complains: The laws concerning corn may every where be compared to the laws concerning religion. The people feel themselves so much interested in what relates either to their subsistence in this life, or in their happiness in a life to come, that government must yield to their prejudices, and, in order to preserve the public tranquillity, establish that system which they ap-

prove of. It is upon this account, perhaps, that we so seldom find a reasonable system established with regard to either of those two capital subjects. (Smith 1976:1 V.v.b.40) We shall ignore his views about religion. But what was wrong with the prevailing system of corn supply? Once again, in his own words: T h e unlimited, unrestrained freedom of the corn trade, as it is the only effectual preventative of the miseries of a famine, so it is the best palliative of the inconveniences of a dearth' (Smith 1976:IV.v.b.7). But why had free trade not found such an eminent advocate before? Perhaps because it was not such a good idea after all (cf. Hont and Ignatieff 1983). The urban food supply has for most of pre-industrial Europe's history been an arena for public intervention. It is our contention that this was neither coincidence, nor a sign of traditional economic irrationality' stifling economic development. On the contrary, we shall argue that such intervention was a perfectly understandable response to the variability of urban food supply. The urbanisation of pre-industrial Europe is inconceivable without intervention in urban food markets. The argument will be in three steps. Firstly, we argue that the underlying features of the supply of and demand for food in pre-industrial cities are such that even relatively mild perturbations in supply will result in major hardship for the mass of the population. The second step consists of an analysis of conceivable mechanisms to cope with such problems.

Public intervention in the food supply in pre-industrial Europe

115

Not only is it possible to deduce a number of such strategies, but the European historical record provides ample illustration of their use. Thirdly, the prevailing form of coping mechanism relates intimately to the social organisation of the time. Public intervention was prevalent, but that still leaves a wide range of options. Both social norms and the nature and power of public authority severely restrict which theoretical options can be put into practice. In arguing our case, we employ two methodological principles. The first is the explicit use of modern economic analysis. This is not the place to reopen the formalist-substantivist debate (Leclair and Schneider 1968). We hope that the proof of the pudding is indeed in the eating, and equally we hope that we have not succumbed to the danger of formalist analysis: an anachronistic modernising of the past. We believe that our second methodological principle - the use of comparative historical analysis - has helped avoid the pitfalls of formalist anachronism. The comparative historical analysis clearly shows that coping mechanisms are not independent of social organisation, but are part of it. For practical reasons we have largely limited our analysis to two periods of European pre-industrial history. The first is that of classical antiquity, where we shall single out imperial Rome. The second is that of early modern Europe, roughly from the fourteenth to the eighteenth century (here, most of our examples come from the Netherlands). These are periods in which the coping mechanisms that we discuss are prominent.

Causes of food crises We may distinguish two groups of causes for pre-industrial food crises. The first is that of natural causes, such as drought or unusually high rainfall, epidemics of plant diseases, plagues of locusts or mice. Environmental variability, and climatic conditions in particular, have received considerable attention in recent years. Such variability creates an interesting situation, and also demands terminological precision. On the one hand, even modern meteorological science cannot predict weather conditions more than a few days in advance, let alone for a whole year. And yet the basic structure of such variability is quite well known. The peasant would know that he should expect a major drought, say, once every ten years, but he would not be able to predict in which year this would fall. The interesting point is that a given outcome may not be certain, but we can establish its numerical probability (Knight 1921, esp. Ch. VII, is classic). Economists call this situation 'risk', to distinguish it from uncertainty, where no such probability can be established. Only if an idea of the numerical probability is possible, 'may conduct in relation to the situation in question be ordered intelligently' (Knight 1921:213). Secondly we may distinguish 'cultural' causes of preindustrial food crises. For example, wars too could lead to the destruction of crops, or to the removal of the peasants who should have worked the land. Early modern observers often saw little difference between soldiers and the locusts of nature. But famines can also arise as the indirect effect of wars (even in non-belligerent countries). In the Second World War a large famine occurred in India, because the British had destroyed many private boats in Bengal as a security measure (the 'Boat Denial Scheme') (Greenough 1982:89f.). This made the normal transport of rice impossible. The closure of trade routes by belligerent parties could of course have similar effects. In the seventeenth and eighteenth centuries the Netherlands depended on shipments of grain from Poland. Some food crises in these centuries were the result of the dangers of shipping in the Baltic resulting from war in Scandinavia (van Dillen 1970:331-40). Piracy was important in the last century of the Roman Republic. Not only did it endanger the supply of grain to Rome, but it was clearly of such importance that at various times important Roman aristocratic leaders were given commands to deal with the pirates. Pompey's command from 67 BC shows both the economic necessity and political importance of putting an end to the problem (Rickman 1980:5If.). He was appointed by a law passed by a tribune of the people, against opposition from the Senate. As history has shown, the Senate's fears that success would lead to power were justified. And during the Civil War both Caesar and Pompey showed themselves aware of the importance of securing the grain supply of Rome. Speculation is always a prominent factor in the perception of contemporaries. In 189 BC the curule aediles (magistrates) Publius Claudius Pulcher and Servius Sulpicius Galba fined the grain dealers for hoarding (Livy 38.35.5). And the lex Mia de annona (corn supply law) was equally aimed against speculators who combined to withhold shipments of grain (Dig. 48,12,2). And yet, some doubts must be expressed. It is not surprising that the anger and frustration of the crowd and administration found a ready

Times of economic crisis have clearly inspired historians to look back into the past for apparently similar problems. This has been so in recent years, but it was no less true in the thirties when historians of classical antiquity such as Bolkestein and van Berchem wrote their famous works, and when early modern historians such as Labrousse laid the foundations for much subsequent work on crises and fluctuations (Bolkestein 1939; van Berchem 1939; Labrousse 1943). But we must avoid thinking of crises past and present as necessarily identical. As a gross simplification, when modern economists use the term 'economic crisis', they normally refer to a situation typified by insufficient demand for goods and services. Investments decrease, consumers become more cautious, the inflation rate drops, and nominal prices may even go down. Stagnating sales lead to a rise in unemployment. Historians working on pre-industrial crises deal with a different phenomenon, usually referred to as a 'subsistence crisis', or 'crise de type ancien' (Abel 1974). Through harvest failure or other causes the production of grain drops sharply. The shortage of food leads to a steep price increase. Consumers spend an everincreasing part of their budget on staple food (mostly grain), and this in a society where even in normal times the purchase of food represented some three quarters of the budget of the working family (Cipolla 1980:30, Table 1-7). As a secondary effect, no money is left to buy industrial goods, and sales of these dwindle, leading to unemployment in the industrial sector. But vagaries of the food supply are at the core of pre-industrial crises. Even though they have been called 'crises of the old type', the modern world affords parallels, such as the 1973 oil crisis. The oil crisis shares one characteristic of a pre-industrial crisis: the sudden reduction in the supply of an important commodity, with concomitant price rises. This was followed by an industrial recession.

Willem Jongman and Rudolf Dekker

116

target in the grain traders. But that does not necessarily make them the real culprits: would speculators have been able to create a famine in a year with a normal harvest? Many modern historians think this is unlikely. A dissenting voice has recently been raised by Steve Kaplan, the specialist in the history of eighteenth-century French grain trading and government regulation. He contends that speculation and collusion between dealers could indeed have been a real threat: the complot de famine need not have been solely the figment of frightened popular imagination (Kaplan 1976; and also Kaplan 1982). This is only likely, however, at a very local level: speculation can only raise prices effectively if it controls a very large part of the supply. The speculator operates by releasing stocks at the times and places of greatest scarcity. He holds on to his stocks because he believes things will be worse next month, and prices therefore higher, but this tactic would not be successful without real underlying shortage. Because his behaviour is guided by price levels, one might even argue that the speculator's actions have the beneficial effect of making grain available when and where it is most needed. Conversely, in the producing areas the appearance of grain dealers may drive prices up: much rural discontent with increasing freedom for grain traders was in areas that did not themselves suffer very bad harvests. The poor in such areas now saw their fate as determined by human factors (the fields had after all been full of grain), and the focus of their loyalties and moral expectations was local, not national (Thompson 1971; Tilly 1983). Such direct causes of famine (war, drought, etc.) can be identified quite easily, but the possibility of underlying structural causes is more difficult to evaluate. Were these crises also related to structural changes in economy and society? The opinions of historians differ; for some, among them Wilhelm Abel and Emmanuel Le Roy Ladurie, the answer is unequivocal: demographic growth is the culprit (Le Roy Ladurie 1966; Abel 1978). Crises would have originated because population growth outstripped the potential for increased production of food. A delicate balance emerges, where the slightest mishap is enough to cause famine. This model is often referred to as neo-Malthusian. We can see this Malthusian model as a model with two variables, one independent (production), and one dependent (population). Population growth is dependent upon prior growth of production. Periodically, population outstrips production, and the result is a pre-industrial crisis (Grigg 1980:49ff.). Not surprisingly this model has met with criticism, firstly from Marxists or historians inspired by Marx. Robert Brenner for example argues that ultimately the origin and development of preindustrial crises is determined by the agrarian class structure (Brenner 1976). He points out that it is the landowning class which decides how the land is cultivated. In addition, the landowning class keeps the cultivators of the soil in a dependent position and appropriates the largest possible share of their production ('surplus extraction'). But development economists have also added their criticism of the neo-Malthusian model. Ester Boserup has proposed a different model with the same two variables, but this time population is the independent and production the dependent variable (Boserup 1965). Population growth is viewed as a potential sti-

mulus to the growth of agricultural production. Crucial to her analysis is the view that it is usually possible to increase production per acre by increasing the frequency of cropping. But decreasing marginal returns to the other factors of production take their toll: increased production per acre usually involves a disproportionate increase in the amount of labour. So-called advanced agricultural systems such as annual cropping are therefore not employed by choice, but by necessity. One other trade-off has to be made. Intensification often goes hand in hand with taking bigger risks. Agriculture is a very complicated adaptation to the natural environment, frequently allowing a number of alternative production strategies. A fundamental choice is usually that between a regular but lower yield, and a higher average with larger variation around the mean. This invites scrutiny of the logic of risk-avoiding behaviour. The logic of risk avoidance If an economic analysis of the reasons for such behaviour is possible, it could perhaps help us to understand how economic subjects may go about achieving such security. Why is a stable income preferable to an income that is highly variable, but with an identical mean? To understand this, a distinction should be made between income and the utility of units of such income (Blaug 1985:332-4). The satisfaction from the first one hundred pounds of income (or kilograms of wheat for that matter) is greater than the satisfaction from the last one hundred pounds. Equally the hardship of losing the hundred pounds that are spent on theatregoing is far less than the hardship of losing the last one hundred pounds necessary for physical survival. In the jargon of the economist, income has a declining marginal utility. Therefore, the pain of losing one hundred pounds is larger than the satisfaction of gaining one hundred pounds. Thus variability causes utility loss, which is to be avoided. It pays, therefore, to take measures to diminish variability up to the point where the cost of such measures equals the utility gained. Obviously, in the pre-industrial world the marginal utility of income would decline steeply. The majority of the population lived very near subsistence level, and the difference between having just enough to survive and just starving is of no mean importance. Such a steep decline in the marginal utility of income suggests that major action would be undertaken to limit variability. Ignoring the logic of such expenditure can lead only to the false assumption of economically irrational behaviour. The subsistence peasant, who consumes his own produce, can directly tailor his production activity to his consumer interest in a stable food supply. When a market for food emerges, however, the interests of producers and consumers diverge. The market and its effects Of course, variability of supply from one year to the next (and within a year) is at the root of the problems that we are discussing. But its consequences are felt only where supply and demand meet, i.e. on the market for food. It is the structure of such a market which determines whether the vagaries of supply cause

117

Public intervention in the food supply in pre-industrial Europe Rye in Antwerp (groats]

60 580 oO Masons' laborer in Antwerp

. Saihcloth in Brabant (groats)

J 1520

1525

1530

1S3S

I

1

U

1540

1 5-4 5

1550

1555

Fig. 10.1. Price fluctuations in the Antwerp region in the sixteenth century (logarithmic scale, sic; nominal prices). Source: van der Wee 1963, vol. Ill: 50f.

substantial hardship for consumers or not. The market for subsistence food in pre-industrial societies is characterised by the fact that even small fluctuations in the quantity supplied result in very large fluctuations in price (in the jargon of the economist, the price elasticity of supply and demand is low). This is so for two reasons. The first is that, in the short run, rising food prices cannot easily generate increased supplies. Typically one would just have to wait until the next harvest. The second reason is that consumers cannot reduce their demand much when prices are soaring. Food is essential for survival. The seventeenth-century statistician Gregory King estimated that a 10 percent reduction in wheat output would result in price rises of about 30 percent, a 20 percent reduction would raise prices by 80 percent, and a reduction of 30 percent would create a staggering 450 percent price increase (Abel 1978:23). And we do indeed witness very considerable fluctuations in prices. Subsistence food seems to show the largest fluctuations, presumably because demand is least elastic there. The devastating fluctuation possible in the price of grain compared to other commodities is amply demonstrated by fourteenth- to sixteenthcentury data from the Antwerp region (Figure 10.1; see also Noordegraaf 1980). The resulting hardship for poor consumers is obvious. If in normal years they spent some three quarters of their income on subsistence food, it is hard to conceive how they survived the bad years (Figure 10.2). For farmers, however, a small harvest need not be a disaster. The importance of the emergence of an urban market for grain is that the farmer is often financially compensated for a disappointing harvest by a higher price. Any surplus grain can be sold at an attractive price (Abel 1978:23ff.). Thus, the more he produces for the market, the less incentive there is to avoid fluctuations in physical output: the urban consumers (and other buyers of food) are landed with the risk. An extreme example of this is provided by viticulture, especially in the production of quality wines in northerly areas such as Bourgogne. One hailstorm can wipe out any income for a whole year. On the other hand, where much of a vintage is destroyed, any producer whose vines are spared can make substantial profits. Ultimately, wine growers are gamblers, and are considered rather frivolous. Thus Theodore Zeldin writes of them: 'Some studies of individual budgets have

Fig. 10.2. The relationship between income and necessary expenditure on food in Antwerp in the fifteenth and sixteenth centuries. Source: van der Wee 1963, vol. Ill: 92f. The shaded area is what remains from annual income after expenditure on essential foodstuffs. For the poorer labourers, expenditure on food varied greatly and income often fell below the level required for bare subsistence.

shown how their lives were a series of hopes deceived and accumulated debts, but with the occasional marvellous harvest which cancelled it all out and allowed them to start again' (Zeldin 1973:168). The existence of a market has very different implications, therefore, for producers and consumers. Producers are to a large extent insulated from the consequences of fluctuations in supply. Consumers, on the other hand, are very vulnerable. Not only do they have to face the consequences of producers having less incentive to limit variation in supply, but they are also the direct victims of low price elasticity. Large fluctuations in output produce even larger fluctuations in prices. In pre-industrial Europe there was a substantial number of poor consumers who did not produce their own food. Their existence is a sign of a relatively complex economy (with substantial urbanism and commercial agriculture), and yet they were in an extremely vulnerable position, because they could not afford famine prices. Their vulnerability posed serious problems for the stability of society as a whole. Strategies for dealing with subsistence crises

Analytically the repertoire of possibilities for coping with these crises is not very complicated, though each one presupposes a certain level of organisation, and is thus dependent on social structure. The first possibility is to try to limit variability in physical production. It is possible to limit variation in output by growing grain types that are more resistant to unfavourable weather conditions. It is possible to grow a number of different crops, and on a number of different plots. Even if one crop fails, another is likely to succeed. With such a diversifying strategy, Robinson Crusoe could survive for many years on a desert island. These are typical strategies of the subsistence peasant. Apart from trying to stabilise production itself, one can also try to stabilise availability. One solution is storage, which copes with temporal variability by balancing good years against bad. The third possibility, again addressing the problem of availability, is to take advantage of spatial variability. Crop failure in one area may be balanced by a good harvest in a neighbouring area. Grain is exchanged between communities under what

Willem Jongman and Rudolf Dekker O'Shea has termed 'social storage' (O'Shea 1981). Underlying such exchanges is the establishment of a claim to reciprocal transactions when circumstances are reversed. Such claims are underwritten by a variety of norms and institutions. A transparent and highly developed institution of this sort is the modern grain market, in which money is the institutionalised form of the claim. Each of these three types of solution can be likened to an insurance policy, with its own costs and benefits. The choice of solution obviously depends both on the effectiveness of the benefits and the cost of the premium. These in turn depend on the given social and economic conditions and organisation. If storage is expensive and transport cheap, an emphasis on interregional trade may be expected. But if buying elsewhere only drives up prices, or is politically unreliable, storage may be cheaper or more effective. In reality these are of course not mutually exclusive strategies, so we may expect to find a range of combinations, with more or less emphasis one way or the other. For the subsistence peasant it is easiest and most effective to concentrate on limiting fluctuation of the physical output. We have noted earlier that the emergence of cities and the concomitant separation of the roles of producers and consumers of grain removes the incentive to keep physical output within a narrow band. Cities have to rely on other strategies to ensure availability of grain, and these strategies, such as large-scale storage and distribution or the maintenance of contacts with other communities, soon necessitate a higher authority - the state. Varieties of public intervention To understand the necessity of public intervention and its intimate relationship with social structure, we shall now compare the solutions adopted in early modern Europe and in classical antiquity (and particularly imperial Rome). The strategies that were adopted differed considerably. Differences in technical development may have played their part, but explanation relates largely to political systems. The concentration of imperial power in the hands of a small Roman elite provided the impetus for the emergence of an imperial capital with some one million inhabitants (Jongman 1988 esp. pp. 187-99 for inequality; Hopkins 1978:96ff. for the size of Rome). The result was an extraordinary demand, but large resources were available through imperialism. Elsewhere in classical antiquity and later in early modern Europe the problems were smaller, and so were the resources. Both in classical antiquity and in early modern Europe, peasants did not only produce for their own subsistence. The great historical separation between producers and consumers of grain had already taken place. In Roman Italy 30 percent of the population lived in cities (20 percent, even if we exclude the city of Rome itself). In Holland about half the population lived in cities in the sixteenth to eighteenth centuries, which is admittedly quite exceptional for early modern Europe. The consumers of grain, however, were not only the inhabitants of the towns, but also those farmers who produced cash crops (such as flax or madder in the early modern Low Countries). In a free market, bad grain harvests go hand in hand with substantial price increases. In the real experience of the poorest consumers this means that they are periodic-

118 ally driven off the market because they cannot afford to buy enough to feed themselves. In Keith Hopkins' terms, they have to adjust to market forces by fluctuations in their own body weight (Hopkins 1983:90). Only by public intervention in the market can this group survive. And governments did indeed intervene, not only from a sense of moral responsibility, or pity, but equally from fear of food riots. One only has to remember the recent food riots in Tunisia after the government tried to raise the maximum price of bread. And to quote Finley on Rome: The ancient sources are unanimous in their view of the dole as a form of poor relief won by the plebs after considerable struggle' (Finley 1985:201). After all, we are not speaking of small groups of marginal elements, but of large segments of the population, and often in a politically very sensitive location. Which methods did the authorities employ? They could tackle the problem on the demand side. This might entail rather drastic measures such as those of Augustus in AD 6 when, on the occasion of an acute dearth of grain, he evicted various groups from Rome, especially foreigners (Rickman 1980:63). An eighteenth-century memorandum from Leiden suggests that the masters of the city wards should keep lists of all strangers living in their wards in order that they 'may be sent back to where they came from, in case they would need any charitable support' (City Archive Leiden, secretarie na 1572, inv. nr. 5863). Another possibility is to exclude certain categories of buyers from the market, such as happened in 1630 to beer brewers or starch makers in Amsterdam, for whom grain was an important raw material (van Dillen 1915:14). Regulating the supply side was almost always more attractive to the authorities. The inventory of possible measures is long. In the first place the government could improve conditions for production. One of the earliest areas for legislation by the state and for the formation of a controlling bureaucracy was the production, trade and processing of grain. Canalisation of rivers, clearing of woodlands or the construction of a polder system for drainage were sometimes undertaken by the state itself, or stimulated by subsidies or tax exemptions. Between 1590 and 1635 the Estates of Holland regularly gave exclusive rights to groups of rich persons to drain lakes, with temporary exemption from taxes on land (van Dillen 1970:235). But all this is of course of only limited importance in the case of crop failure. The first measure taken was often that of a ban on exports. Well known from many classical Greek cities, there are parallel instances in the cities of early modern Europe (Francotte 1905:137ff.). Stocks in the stores of traders could not leave for other cities or other countries where famine was worse, and prices therefore higher. This measure had a very serious drawback, however, of which the sixteenth-century authorities of Amsterdam were well aware: a ban on exports made it less attractive for traders to import new supplies from elsewhere, because they preferred to bring their grain to where free market conditions led to higher prices. Therefore other measures were necessary. Similar problems of alienating the vital merchants occurred with the system of legally fixed prices for bread and other cereal

Public intervention in the food supply in pre-industrial Europe

119

products that was prevalent in early modern Europe. Prices (or rather the weight of the bread) were usually fixed weekly and were related to the price of grain on the free market. In periods of slight shortage, prices could be held down, at the expense of the bakers, who saw their profit margins decline, and who therefore sometimes went on strike, even in the thirteenth century. Manipulation of fixed prices was only possible within narrow margins. Fixing the prices too far below market levels reduces the quantity brought to the market, and stimulates demand. Supply could of course be kept up by subsidies: the traders get the free market price, and the consumers pay a normal low price. In AD 19 the emperor Tiberius paid two sestertii per modius subsidy to the corn merchants to compensate them for the imposition of a maximum price (Tacitus, Annales II87). If supply is inelastic this may involve very large sums of money, and in addition one still needs a rationing system to prevent the cheap grain from being bought up again by merchants, as Julian experienced in Antioch in AD 362 (Liebeschuetz 1972:127). Which other measures were taken? Official emergency stocks were often held, to be released at the prevailing high market price, at a reduced rate, or even free. And if no such official emergency stocks were available, members of the elite could act as benefactors and make their private stocks available (Garnsey and Morris, this volume). Numerous inscriptions from Hellenistic Asia Minor testify to popular gratitude for such euergetism by the more prominent members of society (Francotte 1905; Veyne 1976). No doubt refusal to act as benefactor would have had serious consequences for the social order. Free grain could be distributed directly or via the bakers. In the later Roman empire the bakers were an extension of the government apparatus, and the same has been noted of Amsterdam bakers under the Dutch Republic (Tengstrom 1974:73ff.; Pavis d'Escurac 1976:26ff.; Rickman 1980:205ff.; Kistemaker 1983). Building up stocks and distribution demanded a substantial infrastructure. Many European cities acquired municipal granaries, a fine example of which can still be visited in Metz. Yet these were dwarfed by the enormous structures that arose in Ostia, the harbour of imperial Rome (Rickman 1971). But then Rome had perhaps one million inhabitants. Under the Empire some 200,000 persons were each entitled to about 400 kg of wheat a year (for elaboration: van Berchem 1939; Rickman 1980; Garnsey 1983a) - a ration sufficient to feed about two persons, and given free. This was a system that did not come about overnight, but began with the arrival of tribute grain at the end of the third century BC. In 123 BC a law of Gaius Gracchus made rations available at a low price. In the subsequent century these grain distributions became a focus for political struggle, with reductions and abolition alternating with extension and free grain (Schneider 1974:361-91; Garnsey 1983b:62ff.). In making substantial rations available free, the system had obviously developed well beyond a mechanism for coping with variability, and had become a major factor in the growth of the city of Rome, demanding that clear boundaries be drawn as to eligibility.

market price, some system of rationing was necessary. Not surprisingly the military normally obtained preferential treatment. Apart from the military, rationing in early modern Europe usually aimed to help only the destitute. In Rome the picture is much less clear. There, distribution of grain was originally to all citizens in the city. In time the group that qualified was narrowed down to those who originated from the city, thereby leaving many of the poorer citizens out in the cold; while still allowing rations to other wealthier citizens. What was originally a right of all citizens in Rome developed into a privilege for a section within the citizen body of the capital. Beneficiaries and poor showed only a partial overlap. But this should not mislead us into thinking that the corn dole had nothing to do with poor relief, as has been suggested in recent years (Rea 1972:8; Turner 1975:21f.; criticism in Rowland 1976). The restriction of the number of recipients does not by itself imply that they are 'middle class' (Rea 1972:8). The reality of the case is that the corn dole was given to a section of the population. The majority of them were poor, and for them the distributions will have provided welcome relief. T o argue... that every senator also had the right to queue up for his ration, or to send a domestic to do so for him, strikes me as pedantic absurdity' (Finley 1985:201). And yet the system also reveals a fundamental distinction between antiquity and early modern Europe, in that not all poor qualified for relief. Although in practice ancient distributions were obviously of greatest benefit to the poor, they were linked to citizenship (or to privilege within the citizenry), and were not charity. Neither in Rome, nor elsewhere in antiquity, did the number of children that had to be maintained by a citizen beneficiary matter at all. The ration was the same. The link with citizenship is equally obvious in the early occasional distributions of grain in Rome, which came from war booty or a diplomatic gift to the community. Ancient citizenship had nothing to do with later ideals of universal human rights, and everything to do with privilege (cf. Patlagean 1977:18Iff.).

This brings us from the problem of storage to that of distribution. Because stocks were normally released on the market in times of dearth, and very often at prices below the prevailing

The scale of the distributions in Rome was made possible by imperialism. A large part of the grain was obtained directly from the conquered territories as a tax in kind. The first province to enjoy this doubtful privilege was Sicily, from about 200 BC (Rickman 1980:36ff.). In the following two centuries the population of the city of Rome was able to quadruple. Even more quickly, within a few decades, the Roman diet of porridge, made from varieties of grain unsuitable for breadmaking, had changed to one largely consisting of the luxury leavened white bread (Moritz 1958:67ff.). This gives an extra dimension to the expression 'bread and circuses'. Bread, like most games, was a luxury made possible by Roman military successes. It should be added that a system such as that in the Roman capital did not exist in other cities of the empire, with the exception of Constantinople, when it became capital, and equally of Alexandria and Antioch in the late Empire (van Berchem 1939:102ff.). The distributions in Antioch were probably introduced during the prolonged stay there of the emperor Valens (Liebeschuetz 1972:126ff.). Greek cities on the whole did not build up large stocks (Jameson 1983:10), but relied on a combination of measures to secure supply from elsewhere. The well-known figure of the

Willem Jongman and Rudolf Dekker

120

Table 10.1. The distribution of grain reserves in Pavia in 1555

ceive some grain. Such requests were met with pathetically small allowances. For cities in the Roman East it was a great privilege (and thus worthy of large inscriptions) to obtain imperial permission to receive some annona grain from Egypt (Worrle 1971). Municipal stocks were sold at prices lower than those prevailing on the market. In essence the early modern system is a system that depends mainly, and increasingly, on the possibility of obtaining supplies from elsewhere. That this increasing reliance on the market was viable is borne out by reduction over time of the differences in grain prices between various distant parts of Europe (Braudel and Spooner 1967:392 ff., 464ff.). The high prices which resulted in times of famine were reduced for the benefit of the poor. Another difference was that the Romans had incorporated their grain-producing areas into their empire by military force. The relationship between the Dutch Republic and an important source of its grain, Poland, was different. The Republic waged war in the Baltic only to ensure free shipping through the Sound, never to conquer Poland. The difference in market structure could be used to explain part of the differences in policies towards the grain supply. The city of Rome alone was so large that it absorbed the major share of the grain that was transported over long distances (Hopkins 1983:93). Therefore prices were no longer given. Additional grain could only be bought at steeply rising prices - or requisitioned with increasing difficulty. In such a position it is very attractive to have local stocks, and the greater the shortage the more attractive this is. Later pre-industrial cities were in a rather different position. They did not have the enormous political and military power of Rome, but in addition each of these cities individually had only a limited influence on the total market. This meant that in the event of shortage, additional supplies could be obtained at prices that would not rise sharply under the influence of the extra demand from one city. Rome, on the other hand, was largely equal to the 'world market'. This meant that the free market apart from the city's demand was of only limited size. For each additional quantity of grain that the emperor wanted to buy on this free market, he would have to pay considerably more, compared with later individual European cities. The following conclusions may be drawn. In the first place, public intervention in the food supply was the norm, rather than the exception. Secondly, the repertoire of measures that were taken can be understood within the framework of price theory. Rome was so large that it had to concentrate to a large extent on a storage strategy. This was far less the case with ancient Greek cities and the cities of early modern Europe.

Size of the reserves of grain per family

% families

% reserves

More than 20 bags More than 2, up to 20 bags Up to 2 bags None at all

2 18 20 60 100

45 45 10 100

Source: Cipolla 1980:11; after Zanetti 1964:71 euergetes, the benefactor, plays an important role in this respect. A private citizen or magistrate would personally pay to import a shipload of grain, or would make grain from his own estates available at a low price or free (Garnsey and Morris, this volume). The authorities could also requisition such private stocks from the wealthy. How far they could or would go depended on a complex struggle for power between elite, merchants and potentially riotous poor. In an inscription from the city of Ephesos (Dittenberger 1915-24, no. 354), a Rhodian merchant is honoured because he sold grain below the prevailing high market prices, 'persuaded by the magistrate in charge of the market and desiring to please the people'. The euergetes who gave from his own estates may have been a benefactor, but refusal to act in this way might create revolt. And requisitioning of grain from the attics of the early modern bourgeoisie almost always left them more than enough to get through such a period of famine (Table 10.1). Power and authority need to be legitimated by acts of benevolence in times of real crisis, and these also ensure that existing inequality is not undermined in any fundamental way. Roman and early modern systems compared Having run through the gamut of possible measures, we can make a modest effort to compare ancient and early modern systems. The Roman model was based on large-scale distribution in kind from the emperor's granaries, filled with grain that had mostly been brought in as taxation in kind from conquered areas, first from Sicily, later from North Africa and Egypt. In early modern Europe, with its split sovereignty and where measures were mostly taken for small geographical regions, the situation was quite different. Authorities, especially urban authorities, supported their citizens as parties on the market. The most important thing they stored was money. Of course, these cities usually had some municipal stocks, but these were proportionally far smaller than those in Rome: they were rarely sufficient to cover for more than one harvest failure, and often not even enough for that. As soon as wheat prices started to rise and high prices were expected, cities started to buy stocks on the market, which also brought them into competition with each other. One city's export ban meant hardship for others. In years of high food prices, the archive of letters sent to the city authorities of Amsterdam (the European staple market for the grain trade) contains large numbers of begging letters from other cities in the Republic asking that they might re-

The costs of the Roman imperial solution were enormous, and could only be borne by such an empire - an observation corroborated by a recent study of anti-famine policies in China in the seventeenth and eighteenth centuries. Here again we witness the ability of an administration to transport and store large amounts of grain (Will 1980). An important question, of course, concerns the effectiveness of these interventions. It is our belief that the system was on the whole effective. The very large fluctuations in free market prices indicate that such markets were unable to secure supplies in every year at prices that could be afforded by the poor. The system of

Public intervention in the food supply in pre-industrial Europe

121

intervention worked better than that of total non-intervention. Without it prices would have fluctuated so wildly that the mass of the urban population would have perished periodically. To put it more strongly, intervention made urbanisation possible. In Rome, the role of the free market was probably smallest and famines were apparently averted most successfully. Only through intervention in its grain supply could Rome reach a population size that was not matched in later European times until London reached one million inhabitants around 1800.

divided. The panis gradilis may have pleased the Roman plebs a lot; it was of course largely obtained at the expense of poor Egyptian and North African peasants. Just as Rome's position as a conquest state allowed, for a time, a measure of popular participation, it similarly allowed the Roman plebs to be fed (cf. Finley 1983). The Roman poor never questioned the virtues of imperialism. Their presence in large numbers in the capital ensured that their claims to food entitlement were ignored by the emperor at his peril. In eighteenth-century England many food riots were a protest against the consequences of liberalisation of the grain trade: food prices rose because of demand from elsewhere (Thompson 1971). There too the interests of urban and rural poor could diverge sharply. For the eighteenth-century liberalisation of the grain trade, new political ideologies were of no mean importance. More and more it was believed that the grain supply would be best served by leaving the market to its own devices. In that way food would be brought to where prices were highest, and where needs were presumably strongest. At first, this was an aim of only national policy; later the international market was also expected to follow these precepts. Yet, this decreasing intervention made the eighteenth century - all over Europe - the century par excellence of food riots and taxation populaire. The poor stuck to their traditional entitlement claims. Of course a score of provisions to care for the poor crept in through the back door, but regulated prices featured less and less amongst them. Even though the victory of nineteenth-century liberalism appears complete, closer inspection reveals that various countries retained a measure of intervention where food was concerned, in spite of outward allegiance to the liberal ideology. The transition to a more liberal food market was certainly eased by the choice of an alternative strategy by those most directly affected - the workers. Instead of concentrating on the expenditure side of the problem (and demanding price ceilings for food), they shifted their focus to the income side, and concentrated on demands for higher wages. Thus, taxation populaire was replaced by strikes as the preferred form of social protest (Dekker 1987). This transition would not have been feasible, however, without the increasing success of the grain market in keeping fluctuations within tolerable limits.

The context of public intervention As yet, little has been said about the context of intervention, in terms of social tension and political ideology. Coping with famines is not just a technical problem of economics, but is integrally bound up with the questions of who gains or loses from scarcity, and of who gains or loses as a result of coping strategies. A related question is what may be morally expected from political authorities. In the first place there is pressure from various social groups. Four groups are of special interest to us: authorities, merchants, bakers and the poor. The Dutch Republic may serve to illustrate the conflicts between these groups. From the sixteenth century onwards, merchants were demanding liberalisation of trade and diminution of the role of the state in anything other than the safe-guarding of free trade (Tracy 1983). A pamphlet of 1714 defends this liberal argument: imposition of an export ban will bring all import trade to a halt. Grain will move to regions where free trade offers better prices; intervention is therefore counterproductive (Dekker 1982:121-30, esp. 126). In 1630 and 1631 when a ban was imposed on exports from Holland, some Rotterdam merchants, caught smuggling, declared: The Estates General make silly laws and understand nothing of maritime trade' (Dekker 1982:127). A useful reminder that, even in Holland, the authorities did not always hold the same views as the merchants. Rather, they tried to balance the claims to entitlement from various groups, including the poor, and thus reduce social conflict. The poor did not have the same means to press their views as the merchants, but they could threaten with violence. Fear of riots was an important motivation behind the actions of authorities. Thus, in the eighteenth century, food riots often took the form of taxation populaire, i.e. the people took matters into their own hands when adequate public measures to stem the tide of rising prices failed to materialise. The people sequestrated grain stocks and organised their distribution at prices that they considered reasonable (Tilly 1971). In such a case it was the bakers, millers and the like who suffered. This group always found themselves caught between merchants (charging high prices), on the one hand, and the populace (often supported by the authorities), who fixed a maximum price, on the other. A Rotterdam baker and amateur historian living around 1700 put the blame for high food prices primarily on the merchants, but both he and others added that the authorities often fixed such low maximum prices that bakers were making a loss (Dekker 1982:128). Pressure from the poor is evidently the key to an understanding of public action. But often the interests of the poor were

Conclusions The relationship between public authority and the food supply is characterised by the fairly continuous presence of a number of constraints, and by variation in the practical solutions prevailing in the historical record. The value of our formalist economic analysis is that it has enabled us to avoid environmental determinism without succumbing to the irrelevance of an accumulation of historical examples, while a comparative historical framework has ensured that this analysis is embedded within the appropriate institutional context. Man's need for food and nature's biological laws pose severe constraints on human behaviour. This is reflected in substantial fluctuations of food prices. Even such high prices cannot raise production appreciably. The dismal logic of economics shows that it is not enough to do something about these high prices. That will not work, unless the supply of food itself (and

Willem Jongman and Rudolf Dekker

122

sometimes demand) is dealt with. And that is precisely what occurs. The coping mechanisms in Rome and in the cities of early modern Europe aim to supply more food than is available hie et nunc (they increase the price elasticity of supply). The same economic analysis shows the vast divergence that emerges between producers and consumers. Without adequate coping mechanisms for (especially urban) consumers neither Rome nor early modern Europe could have sustained such a high level of urbanisation and social complexity. But equally, only the emergence of effective state power could provide these coping mechanisms. Their cost was substantial. The size of the city of Rome made it difficult to rely on trade to deliver the goods in bad years; stocks built up in the past were a more practical source. Its political power made such accumulation feasible. This strategy was largely beyond the means of early modern authorities. However, they could more easily obtain supplies from elsewhere, in part because of smaller size. The increasing efficiency of interregional trade made this strategy more and more feasible, even to

the extent that authorities came to feel that their own role was becoming redundant. Environmental constraints and a technical economic analysis do not suffice to explain the variety of coping behaviour: state power and its origins and limits are essential ingredients. The final ingredient is that of popular power and expectations. Cities were dangerous places to live for the poor. Many lethal dangers may have been in the realm of fate or God, but famine was invariably felt to be a matter of public concern. Something could be done about it, and people felt they had a right to be protected, a right that was seen as a tradition of urban society. Meeting that obligation ensured allegiance to the social system, but failing the test could promote serious social protest. Note Earlier versions of this paper were read to audiences in Rotterdam and Belfast, and we gratefully acknowledge the many useful comments. In addition we should like to thank the late Sir Moses Finley, Peter Garnsey and Keith Hopkins.

123

Chapter 11 Conclusions: bad year economics

John O'Shea and Paul Halstead

At the beginning of this volume three questions were posed concerning the ways in which societies protect themselves against scarcity: (1) How do societies buffer themselves against periodic variation in food availability? (2) How do these coping activities influence other aspects of cultural organisation? (3) To what extent can coping strategies provide the impetus for social change? We can now return to these questions in the light of the varied cases that have been presented. (1) How do societies buffer themselves against periodic variation in food availability? In essence, buffering uses selected aspects of variability to dampen the effects of others -exploiting resource heterogeneity through diversification, spatial variability through mobility and exchange, and temporal variability through storage. The particular mix of responses employed, however, is influenced both by the nature and structure of variability and by a host of enabling and constraining cultural factors. So, for example, when considering storage as a buffering mechanism for high-latitude huntergatherers, Rowley-Conwy and Zvelebil emphasise the ecological and technological preconditions while Mine and Smith highlight the social and organisational requirements of such mass capture and storage systems. Collectively, the examples presented in this volume suggest a series of basic relationships that characterise risk-buffering systems. In particular, they emphasise regularities relating to scale

and to the way specific buffering activities are combined to form coherent coping strategies. The importance of scale, both spatial and temporal, is illustrated by the manner in which the various papers have subdivided and refined the four basic responses to scarcity: mobility, diversification, storage and exchange. This is most clearly shown in the case of mobility, where the logistic and residential movement regularly undertaken to cope with seasonal variation (Legge) may be contrasted with the irregular, opportunistic 'visiting' of distant kin or exchange partners in bad years (O'Shea) and with the wholesale migration of groups sometimes made necessary by longterm environmental change (Mine and Smith). In a similar manner, storage embraces not only the direct preservation of foods on an intermediate timescale, but also short-term storage of time through specialised technology (Rowley-Conwy and Zvelebil) and indirect storage in the form of livestock, valuable craft goods (O'Shea) and currencies (Jongman and Dekker) over progressively longer periods of time. Social distance can exert a similar scalar effect on risk buffering. The influence of social scale is clearly illustrated by the different varieties of exchange: sharing, trade, markets, theft and raiding constitute a continuum of increasing social scale which may or may not correspond with actual spatial distance (Halstead; Mine and Smith). The contributions also highlight interrelationships between the four categories of buffering mechanism. Some buffering mech-

John O 'Shea and Paul Halstead

124

anisms are naturally complementary while others are, in practice, incompatible. Mobility and diversification tend to be mutually reinforcing, while mobility may preclude effective use of storage. Storage and exchange are often employed in concert as part of a single strategy, with exchange operating as a functional equivalent of mobility - moving food instead of moving people. When differing buffering activities are combined in a strategy, compatible mechanisms tend to operate at the same levels of scale, while antagonistic mechanisms tend to be separated at differing hierarchical levels. So, for example, farmers typically maintain a range of storage, diversification and exchange options to cover usual seasonal and interannual variability, but may resort to mobility when shortages are particularly severe. This separation of incompatible buffering mechanisms is an important aspect of hierarchised risk-management strategies. Within hierarchised systems, low-level mechanisms to a great extent define the normal way of life of human communities and as such may not even be recognised as risk-buffering activities (Forbes). They also shape the means available for countering longer-term variability. Widespread day-to-day sharing, for example, may preclude the storage of food to counter seasonal shortages. Conversely, in settings where storage is regularly practised, overproduction and exchange can be effective responses to interannual variability in the food supply. Ideally, the effects of progressively higher-level mechanisms should be additive and complementary but this need not always be the case. In some instances, as in the paper by Jongman and Dekker, high-level mechanisms actually undermine or invalidate the lower-level responses. There also may be instances where a structural conflict exists between high- and low-level mechanisms, as when the selective forces acting on high-level decision makers are different from those on producers. This may occur, for example, when centres are largely insulated from shortages, or in the case of a parasitic elite. In such instances, the decisions and activities of upper-level mechanisms may only exacerbate shortage at the local level.

sification and mobility, both in normal subsistence and in emergency provisioning, and are of necessity dependent on storage and exchange. In addition, population concentration can reduce the efficacy of many low-level mechanisms and thus force greater reliance on high-level mechanisms, with all the attendant costs and risks that such dependence entails. Conversely, where stored foods are critical, large aggregations of people may be drawn together for cooperation in seasonally intensive harvesting or husbandry, or for mutual protection of vital stores. Similarly, the contributions underline the central importance of available technology in determining the character of the buffering strategies that can be employed by any specific society. One aspect of technology with particularly interesting implications for the organisation of buffering systems is transport technology, since it establishes the effective size and capacity of food exchange networks. The existence of a technology that allows the successful movement of bulk goods can greatly enhance the stability of a buffering system, since it allows spatial levelling of scarcity over a much broader area and permits the functional articulation of a greater number of productive systems. The key significance of such articulations over a very wide area was well illustrated by the examples of Jongman and Dekker. Of course, in such systems, the reliability of transport itself acquires a critical importance and, as the Roman emphasis on suppressing piracy suggests, social and political factors often determine the actual effectiveness of such technologies. The structure of social relations between individuals and between communities is a third major point of articulation for risk-buffering activities. Their important role in risk buffering is apparent in the universal use of food sharing and hospitality to define and confirm ties of kinship, partnership or patronage, and in the use of such relationships to establish rights to food resources. 'Getting one's feet under the table' is a sign of social acceptance. The unique social significance accorded to food provides both a vehicle for vital exchanges in times of need and a partial brake on the exploitation of such exchanges. As a low-level mechanism, kin-based networks of exchange and obligation represent an important (and virtually universal) response to scarcity. At higher levels of scale, real or fictive kin relations can be a powerful ideological device for initiating and maintaining exchange connections, providing a vehicle for interaction at scales beyond the normal scope of biological kinship. Other contributors emphasise the role of ideology in regulating and maintaining buffering systems. The dynamic relationship between the ideal of self-sufficiency, on the one hand, and the balancing act between household and supra-household buffering mechanisms, on the other (Forbes; Garnsey and Morris), is a good example. Emphasis on the virtues of self-sufficiency is not a proscription against mutual assistance above the level of the household, but rather is a reminder of the dangers of dependence on resources beyond the control of the prospective consumer. Similarly, Mine and Smith show how myths can be used to identify the key elements of resource variability, and thereby to encode information both about the nature and scale of subsistence risk in the region and about the ways in which spatial variability in re-

(2) How do coping activities influence other aspects of cultural organisation? All of the contributions provide examples of the articulation of buffering activities with other aspects of societal organisation, yet in each instance it is difficult to determine at what point a particular aspect of culture is simply accommodating a buffering mechanism as opposed to being a component of that mechanism. Furthermore, because of the opportunistic character of buffering activities, such points of articulation are found at all levels of society. Nonetheless, certain elements of cultural organisation repeatedly surface in the contributions as factors integral to the kind of coping strategies employed and to their effectiveness. Among these, demography, technology and the structure of social relations are most prominent. Few factors are more integral to the character of buffering systems than the size and distribution of population. Large aggregations of people may drastically reduce the potential of diver-

Conclusions: bad year economics

125

sources can be exploited to counter the effects of temporal variability.

may have been a vital ingredient in cementing inequalities of wealth and power in pre-classical Greece (Garnsey and Morris). Arguably the source of more radical long-term change, however, can be traced to the necessarily embedded character of high-level centralised buffering mechanisms. Both in the redistributive economies of the Bronze Age (Halstead; Wright, Redding and Pollock) and in the market economies of the historic period (Forbes; Garnsey and Morris; Jongman and Dekker), much of the 'normal surplus' occasionally deployed as a buffer against scarcity was usually committed to the production (and consumption) of non-subsistence commodities. This created an incentive for continued overproduction, even when a run of successful harvests or the apparent efficacy of centralised buffering mechanisms seemed to render this unnecessary. On the other hand, it obscures the riskbuffering role of the redistributive agency or market. Moreover, once these institutions are firmly established, they may assume a dynamism of their own. This last point is powerfully illustrated by the emergence, out of a market economy substantially concerned with the 'corn supply', of the enormously influential liberal ideal of'free trade' (Jongman and Dekker). Thus buffering mechanisms adopted as solutions to short-term or occasional instability have often instigated radical and sustained change.

(3) To what extent can coping strategies provide the impetus for social change? As the contributions to this volume aptly illustrate, uncertainty itself is often a recurrent stimulus to change because of its inherent tendency to pose unusual, and often unexpected, problems. Consequently risk-buffering behaviour is very prone to change, as old mechanisms fall into disuse during long periods of security, as new problems demand new responses and, on occasion, as new high-level mechanisms undermine the existing lowlevel responses. Moreover, since risk-buffering behaviour tends to have ramifications throughout the cultural system, changes in risk-buffering activities are likely to trigger a chain reaction affecting many other aspects of society. The resulting change may be catastrophic, cyclical or directional. Sooner or later, all populations encounter problems so severe or so unexpected that catastrophic change results, with greatly increased mortality and/or the collapse of the institutions, conventions and currencies which underpin the very existence of the society. The seemingly depopulated 'dark ages' which followed the collapse of many early states are a vivid example of this phenomenon. Change may also be cyclical in response to long-term climatic fluctuations, as with the alternation between terrestrial and marine resources among Alaskan Eskimo (Mine and Smith) or between farming and herding among the Wodaabe of Niger (Legge). In many cases discussed in this volume, however, change has been both cumulative and transformational. One such transformation is that from economies largely dependent on mobility and diversification to cope with the vicissitudes of the environment, to economies dependent on storage and exchange. Dependence on storage and exchange is often associated with, and reinforced by, increasing aggregation of population, so creating conditions favourable variously for the growth of warfare, intensified regional interaction and social stratification (Halstead; Mine and Smith; O'Shea; Rowley-Conwy and Zvelebil). In particular, the linking of durable valuables to buffering in 'social storage' provides a platform for the unequal accumulation of wealth, status and power. In complex societies, conflict arising from markedly unequal entitlement to food was often a source of political change. For example, in imperial Rome, early modern Holland and contemporary Greece, the presence of a large population of poor and vulnerable urban consumers, uncomfortably close to the seat of power, has at times subverted the normal inclinations or allegiances of the ruling authorities (Jongman and Dekker; Forbes). In the urban societies of ancient and early modern Europe, a number of constraints can be documented on the extent to which the rich could exploit food shortages afflicting less fortunate members of the population. But these constraints moderated rather than prevented exploitation; and the ability of local elites to mount hazardous expeditions overseas to dispose of or acquire surplus food

Retrospect and prospect: Finally, it is left to consider the implications and limitations of an approach emphasising uncertainty and strategies for buffering scarcity. A recurrent theme through the volume is the recognition that variability occurs on a series of more or less discrete scales and that buffering takes the form of various scale-specific transformations of four basic types of response - diversification, mobility, storage and exchange. Because of this scalar correspondence between variability and effective response, this framework provides a simple but useful tool for identifying and investigating buffering behaviour. Beyond an analysis of variability for its own sake, however, the constraints introduced by risk-buffering strategies provide a starting-point for the investigation of culture process and change. Indeed, because of their extensive ramifications, risk-buffering activities provide valuable insight into the operation of social systems even when they are not the primary focus of study. The emphasis on uncertainty and risk buffering which has guided the contributions to this volume has enriched our understanding in each of the specific cases presented. Although each study has been addressed in its own historical context, a number of common themes has emerged regarding both the problems faced and solutions adopted by different communities. The adoption of an explicitly comparative perspective has been fruitful, for example, in the contrast between imperial Rome and the cities of early modern Holland (Jongman and Dekker) or between the early crop-based economies of North America and their mixed farming counterparts in Europe (O'Shea). These studies only scratch the surface, however, of the potential of this approach for cross-cultural investigation. A particularly fruitful avenue is the study of trajectories of development within regions over long periods of time (Mine and Smith; Halstead).

John O 'Shea and Paul Halstead

126

Similarly, the importance of intermediate spatial scales has been emphasised in a number of the papers. With the recognition that the effects of environmental perturbation are not homogeneous over space, the papers have used the necessary scale of effective spatial response to look beyond the boundaries of particular societies to view the linking of often very different cultures (and economies) in supra-cultural buffering systems. In this sense, regional buffering networks represent the middle ground between individual societies and world systems. The character of these spatial networks, and the assessment of their capacities, remain important areas for future study. The contributions also illustrate a number of methodological principles which must be pursued if the potential of this approach is to be realised fully. Archaeologists must come to terms with the brief depositional episodes corresponding to the seasonal and interannual cycles of variability with which much of this volume has been concerned: this level of detail is essential not only for the analysis of buffering systems, but even for the basic functional identification of classes of artefacts, such as clay sealings (Wright, Redding and Pollock). Similarly, much can be gained from ethnographic studies focused on variability at the level of the individual household or field (Forbes), as well as on larger social units (Legge; O'Shea). Conversely, the brief observational frame of many ethnographic studies is a serious impediment to investigation of interannual and long-term variability. In response to this problem, Mine and Smith convincingly use palaeoclimatological and archaeological evidence to corroborate and extend ethnohistoric accounts from Alaska. Jongman and Dekker are blessed with contemporary archives from early modern Holland, but Garnsey and Morris make skilfully critical use of the more anecdotal information supplied for pre-classical Greece by the poet, Hesiod. Prehistorians, by definition, lack such written accounts, but modern meteorological records, crop-yield data and the like can be used as a starting-point for models of the basic structure of temporal and spatial variability in past environments (Halstead; O'Shea; Rowley-Conwy and Zvelebil). As an initial exploration of cultural responses to uncertainty, this volume has attempted to include a wide range of cultures, both simple and complex, contemporary and prehistoric. Yet, it is inevitable that a number of important issues or potential situations have not been addressed. The contributions in the volume have tended to focus on interannual variability and on societies in the temperate or highlatitude zones which are characterised by relatively marked seasonal variation. In one sense this emphasis is intentional, and represents a necessary corrective to the tendency to consider only annual rounds and to concentrate on remnant groups found mainly in equatorial settings. It is clear, however, that the nature of buffering systems among societies living in tropical environments with relatively low seasonal and longer-term variation will be rad-

ically different in form and focus. These warrant investigation, not least because small-scale societies from these relatively stable environments dominate the ethnographic literature and are widely used as models of universal relevance. Likewise, other scales of temporal variability can profitably be considered. To what extent are mechanisms beyond simple physiology and very localised sharing employed in response to very short-term variability? Over what time spans do cultures recognise recurrent or cyclical environmental phenomena and so are able to formulate a response? A related issue of considerable interest is how cultures actually come to embed buffering behaviour in forms such as mythology (Mine and Smith) or the 'grandfather response' (Forbes). A second issue not touched upon is the potential for uncertainty in the acquisition of resources other than food to elicit similar buffering responses. Food has been given priority in the papers because of its fundamental nature, and because of its analytical importance as a resource, the requirements and availability of which can be specified unambiguously. Yet, in some societies, particularly more complex ones, items other than food may be of sufficient necessity or 'strategic' importance to warrant similar kinds of responses. Access to copper and tin during the Bronze Age or to petroleum in modern times might qualify as such strategic items. Of course, true strategic materials must be distinguished from 'valuable' items which serve as maintaining elements in a regional exchange network. Such a determination may be rather more difficult than is the case with food. Although we believe that an approach emphasising variability and cultural responses to uncertainty offers great potential for understanding culture process and change, it clearly is not a solution to all problems, nor can risk-buffering activities be invoked to 'explain' all cultural phenomena. Yet, whether causal or not, environmental variability poses a set of problems that must be addressed by every society, regardless of time, place, social organisation or cognised view of the universe. It is the universality of the problems posed by environmental variation and the relatively limited ways in which human groups can respond to such unpredictability that gives this approach its analytical power. Finally, the diverse case studies presented in this volume also invite reflection of a more sombre nature. The contrast afforded between small-scale hunting-gathering and farming societies, on the one hand, and complex states and empires, on the other, highlights the vastly superior scale and power of the buffering mechanisms that can be mobilised by the latter. But the same large scale of these mechanisms is also their Achilles' heel. As recent events in Africa so tragically demonstrate, buffering mechanisms of even international scale may still prove grossly inadequate to counteract regional famine. And, ironically, such intervention may - as relief agencies have belatedly realised - actually prolong the effects of famine by disrupting existing low-level buffering arrangements and impeding their re-establishment.

127

References

Aaris-Serensen, K. 1980. Atlantic fish, reptile and bird remains from the mesolithic settlement at Vedbaek, North Zealand. Videnskabelige Meddelelser fra Dansk Naturhistorisk Forening 142: 139-49. Abel, W. 1974. Massenarmut und Hungerkrisen im Vorindustriellen Europa. Versuch einer Synopsis. Hamburg/Berlin, Paul Parey. 1978. Agrarkrisen und Agrarkonjunktur (3rd edn.). Hamburg/Berlin, Paul Parey. Adams, R. McC. 1962. Agriculture and urban life in early southwestern Iran. Science 136 (3511): 109-22. 1981. Heartland of Cities. Chicago, University of Chicago Press. Aigner, J.S. 1982. Cyclicity in caribou populations and prehistoric archaeology on the North Slope interior of Alaska (paper presented at the Symposium of North Slope Archaeology, Annual Meeting of the Society for American Archaeology, Minneapolis, April, 1982). Akritas, T. 1957. Methana: I Istoriki Khersonisos. O Ghiros tis Khersonisou. Athens, K. and I. Kaghaphas. Alexiou, S. 1961. Zitimata tou proistorikou viou: kritomikina'ikon emborion. Arkhaioloyiki Efimeris 1953-54(3): 135-45. Allan, W. 1965. The African Husbandman. Edinburgh, Oliver and Boyd. Alland, A. Jr 1975. Adaptation. Annual Review of Anthropology 4: 59-73. Allen, P.W. and Weedfall, R.O. 1966. Weather and climate, in N.J. Wilimovsky and J.N. Wolfe, eds., Environment of the Cape Thompson Region, Alaska (US Atomic Energy Commission). Washington, US Government Printing Office. Aller, W.F. 1954. Aboriginal food utilization of vegetation by the Indians of the Great Lakes region as recorded in the Jesuit Relations. The Wisconsin Archaeologist 35 (3): 59-73. Amberger, K-P. 1979. Neue Tierknochenfunde aus der Magula Pevkakia in Thessalien, 2: die Wiederkauer. Dissertation, University of Munich.

Amiet, P. 1972. La Glyptique Susienne. Memoires de la Delegation Archeologique Francaise en Iran XLII. Paris, Guenther. Amsden, C.W. 1977. A quantitative analysis of Nunamiut Eskimo settlement dynamics: 1898-1969. Ph.D. thesis, University of New Mexico. Andersen, S.H. 1985. Tybrind Vig. A preliminary report on a submerged Ertebelle settlement on the west coast of Fyn. Journal of Danish Archaeology 4: 52-69. Andersen S.H. and Malmros C. 1984. "Madskorpe" pa Erteb0llekar fra Tybrind Vig. Aarbogerfor Nordisk Oldkyndighed of Historie 1984: 78-95. Anderson, D.D. 1968. A stone-age campsite at the gateway to America. Scientific American 218(6): 24-33. 1972. An archaeological survey of Noatak drainage, Alaska. Arctic Anthropology 9 (1): 66-103. 1984. Prehistory of North America, in D. Damas, ed., Handbook of North American Indians: Vol. 5 (Arctic). Washington, Smithsonian Institution. Andreev, J. 1979. Konige und Konigsherrschaft bei den Epen Homers. Klio 61: 361-84. Andrews, J.T., Davis, P.T., Mode, W.N., Nichols, N. and Short, S.K. 1981. Relative departures in July temperatures in northern Canada for the past 6000 years. Nature 289: 164-7. Andrews, J.T. and Diaz, H.F. 1981. Eigenvector analysis of reconstructed Holocene July temperature departures over northern Canada. Quaternary Research 16: 373-89. Annual Reports of the Fisheries, Recreation and Amenity Officer, Welsh Water Authority, Dee and Clwyd Division. Arnold, C D . 1983. A summary of the prehistory of the Western Canadian Arctic. Musk-Ox 33: 10-20.

References

128

Arnon, I. 1972. Crop Production in Dry Regions, 2: Systematic Treatment of the Principal Crops. London, Leonard Hill. Arrhenius, B. 1984. Analysrapport gallande organiskt material pa keramikskarvor funna i Loddesborg, Loddekopinge sn, Skane, in K. Jennbert, Den Produktiva Gavan. Tradition och Innovation i Sydskandinavien for omkring 5300 ar sedan (Acta Archaeologica Lundensia, series in 4°, no. 16.). Lund, C.W.K. Gleerup. Asch, D.L. and Asch, N.B. 1985. Prehistoric plant cultivation in westcentral Illinois, in R.I. Ford, ed., Prehistoric Food Production in North America. University of Michigan Museum of Anthropology, Anthropological Papers 75: 149-203. Athens, J.S. 1977. Theory building and the study of evolutionary process in complex societies, in L.R. Binford, ed., For Theory Building in Archaeology. New York, Academic Press. Auer, V. 1925. Investigations of the Ancient Flora of Hame. Communicationes ex Instituto Quaestionum Forestalium Finlandiae Editae 9. Austin, M.M. 1981. The Hellenistic Worldfrom Alexander to the Roman Conquest. Cambridge University Press. Baier, S. 1980. An Economic History of Central Niger. Oxford, Clarendon Press. Bailey, H.P. 1979. Semi-arid climates: their definition and distribution, in A.E. Hall, G.H. Cannell and H.W. Lawton, eds., Agriculture in Semi-Arid Environments. Berlin, Springer-Verlag. Baker, R.R. 1978. The Evolutionary Ecology of Animal Migration. New York, Holmes and Meier Publishers, Inc. Balikci, A. 1968. The Netsilik Eskimos: adaptive processes, in R.B. Lee and I. De Vore, eds., Man the Hunter. Chicago, Aldine. 1970. The Netsilik Eskimo. New York, Natural History Press. Ball, N. 1978. Drought and dependence in the Sahel. International Journal of Health Services 8:271-98. Bamforth, D. 1986. Technological efficiency and tool curation. American Antiquity 51: 38-50. Barlett, P.F. 1980a. Adaptive strategies in peasant agricultural production. Annual Review of Anthropology 9: 545-73. 1980b. Introduction: development issues and economic anthropology, in P.F. Barlett, ed., Agricultural Decision Making. Anthropological Contributions to Rural Development. New York, Academic Press. Barnett, D.G. 1967. Long-range ice forecasting for Alaska's north coast. Sea Technology 17 (7): 24-7. Barry, R.G., Arundale, W.H., Andrews, J.T., Bradley, R.S. and Nichols, H. 1977. Environmental change and cultural change in the eastern Canadian arctic during the last 5000 years. Arctic and Alpine Research*. 193-210. Baumhoff, M.A. 1981. The carrying capacity of hunter-gatherers, in S. Koyama and D.H. Thomas, eds., Affluent Foragers. Semi Ethnological Studies 9: 77-87. Becker, C.J. 1939. En stenalderboplads paa Ordrup Naes i Nordvestsjaelland. Aarboger for Nordisk Oldkyndighedog Historie 1939: 199-280. 1941. Fund af ruser fra Danmarks stenalder. Aarboger for Nor disk Oldkyndighed og Historie 1941: 131 -49. Belanovskaya, T.D. 1975. K voprosu o rybolovstve v period neolita na nizhnem Donu. Kratkie Soobshcheniya po Arkheologii SSSR 140: 107-11. Bender, B. 1978. Gatherer-hunter to farmer: a social perspective. World Archaeology 10 (2): 204-22. Benecke, N. 1986. Some remarks on sturgeon fishing in the southern Baltic region in medieval times, in D.C. Brinkhuizen and A T . Clason, eds., Fish in Archaeology. Oxford, British Archaeological Reports (International Series 294). Benefice, E., Chevassus-Agnes, S. and Barral, H. 1983. Nutritional situation and seasonal variations for pastoral populations of the Sahel (Sengalese Ferlo). Ecology of Food and Nutrition 14: 229-47. Berchem, D. van 1939. Les Distributions de Ble et d'Argent a la Plebe Romaine sous VEmpire. Geneve, Georg & Cie. Berg, E. 1976. The Economic Impact of Drought and Inflation in the Sahel

(Discussion paper SI). Ann Arbor, Research for Economic Development. Bergerud, A.T. 1971. The Population Dynamics of Newfoundland Caribou. Wildlife Monographs 25. 1979. A review of the population dynamics of caribou and wild reindeer in north America, in E. Reimers, E. Gaare, and S. Skjenneberg, eds., Proceedings of the Second International Reindeer/Caribou Symposium, September 1979, Reros, Norway. Trondheim, Direktoratet for Vilt og Ferskvannskisk. Bernus, E. 1974. Possibility et limites de la politique d'hydraulique pastorale dans le Sahel Nigerien. Cahiers de VOffice de la Recherche Scientifique et Technique Outre Mer, Series Sciences Humaines 11: 119-26. Berve, H. 1937. Miltiades (Hermes Einzelschrift 2). Berlin, Weidmannsche Buchhandlung. Binford, L.R. 1978. Nunamiut Ethnoarchaeology. New York, Academic Press. 1980. Willow smoke and dogs' tails: hunter-gatherer settlement systems and archaeological site formation. American Antiquity 45: 420. 1983. In Pursuit of the Past. London, Thames and Hudson. Blakeslee, D J . 1975. The Plains interband trade system: an ethnohistoric and archaeological investigation. Ph.D. dissertation, University of Wisconsin. Blaug, M. 1985. Economic Theory in Retrospect (4th edn.). Cambridge University Press. Bleed, P. 1986. The optimal design of hunting weapons: maintainability or reliability. American Antiquity 51: 737-47. Bliss, L.C. 1971. Arctic and alpine plant life cycles. Annual Review of Ecology and Systematic^ 2: 405-38. Boaz, F. 1899. Property marks of Alaskan eskimo. American Anthropologist n.s. 1 (4): 601-14. Boessneck, J. 1962. Die Tierreste aus der Argissa-Magula vom prakeramischen Neolithikum bis zur mittleren Bronzezeit, in V. Milojcic, J. Boessneck and M. Hopf Argissa-Magula 1: das Prlikeramische Neolithikum sowie die Tier- und Pflanzenreste. Bonn, Rudolf Habelt. Bokonyi, S. 1970. Animal bones from Lepenski Vir. Science 167: 1702-4. 1972. Zoological evidence for seasonal or permanent occupation of prehistoric settlements, in P J . Ucko, R. Tringham and G.W. Dimbleby, eds., Man, Settlement and Vrbanism. London, Duckworth. Bolkestein, H. 1939. Wohltatigkeii und Armenpflege im Vorchristlichen Altertum. Utrecht, Oosthoek. Bolomey, A. 1973. An outline of the late epipalaeolithic economy at the 'Iron Gates': the evidence on bones. Dacia 17: 41-52. Boserup, E. 1965. The Conditions of Agricultural Growth. London, Allen and Unwin. Bottema, S. 1979. Pollen analytical investigations in Thessaly (Greece). Palaeohistoria

21:19-40.

1980. On the history of the walnut (Juglans regia L.) in southeastern Europe. Acta Botanica Neerlandica 29: 343-9. du Boulay, J. 1974. Portrait of a Greek Mountain Village. Oxford University Press. Bourriot, F. 1976. Recherches sur la Nature du Genos. Etude d'Histoire Sociale Athenienne - Periodes Archaique et Classique. Paris, Champion. Braham, H.W., Fraker, M.A. and Krogman, B.D. 1980. Spring migration of the Western Arctic population of bowhead whales. Marine Fisheries Review 42 (9-10): 36-46. Braudel, F. and Spooner, F. 1967. Prices in Europe from 1450 to 1750, in Cambridge Economic History of Europe (vol. 4). Cambridge University Press. Bravo, B. 1977. Remarques sur les assises sociales, les formes d'organisation et le terminologie du commerce grec a l'epoque archaique. Dialogues d'Histoire Ancienne 3: 1-59. 1983. Le commerce des cereales chez les grecques de l'epoque arch-

References

129

aique, in P.D.A. Garnsey and C.R. Whittaker, eds., Trade and Famine in Classical Antiquity (supp. vol. 8). Cambridge Philological Society. Breiwick J.M., Mitchell, E.D. and Chapman, D.G. 1981. Estimated initial population size of the Bering Sea stock of bowhead whale (Balaena mysticetus): an iterative method. Fishery Bulletin 78: 843-53. Brenner, R. 1976. Agrarian class structure and economic development in pre-industrial Europe. Past and Present 70: 30-75. Brinkhuizen, D. 1986. Features observed on the skeletons of some recent European Acipenseridae: their importance for the study of excavated remains of sturgeon, in D.C. Brinkhuizen and A.T. Clason, eds., Fish in Archaeology. Oxford, British Archaeological Reports (International Series 294). Broadbent, N. 1979. Coastal Resources and Settlement Stability (AUN 3). Uppsala, University Institute of North European Archaeology. Brogger, A.W. 1908. Vistefundet. En aeldre stenalders kjokkenmoddingfra Jaederen. Stavanger, Dreyer. Brokensha, D., Horowitz, M. and Scudder, T., eds. 1977. The Anthropology of Rural Development in the Sahel: Propositions for Research. New York Institute of Anthropology, Binghampton. Brown, J., Miller, P.C., Tiezen, L.L. and Bunnell, F.L. 1980. An Arctic Ecosystem: The Coastal Tundra at Barrow, Alaska. Stroudsberg, Dowden, Hutchinson and Ross Inc. Bryson, R.A. and Wendland, W.A. 1967. Tentative climatic patterns for some late glacial and post-glacial episodes in central North America, in WJ. Mayer-Oakes, ed., Life, Land and Water. Winnipeg, University of Manitoba Press. Bulatao, R.A. and Lee, R.D., eds. 1983. Determinants of Fertility in Developing Countries (2 vols.). New York, Academic Press. Burch, E.S. 1972. The caribou/wild reindeer as a human resource. American Antiquity 37: 339-68. 1975. Eskimo Kinsmen: Changing Family Relationships in Northwest Alaska. American Ethnological Society, Monograph 59. St. Paul, West Publishing. 1980. Traditional eskimo societies in northwest Alaska. Senri Ethnological Studies 4: 253-304. Osaka, National Museum of Ethnology1981. The Traditional Eskimo Hunters of Point Hope, Alaska: 18001875. North Slope Borough. Burch, E.S. and Correll, T.C. 1972. Alliance and conflict: inter-regional alliance in north Alaska, in L. Guemple, ed., Alliance in Eskimo Society (Proceedings of the American Ethnological Society, 1971, Supplement). Seattle, University of Washington Press. Caldwell, J.C. 1981. The mechanisms of demographic change in historical perspective. Population Studies 35: 5-27. Cambitoglou, A. 1981. Archaeological Museum of Andros. Athens, Goulandris. Camp, J.McK. 1979. A drought in the late eighth century. Hesperia 48: 397-411. Campbell, D.J. 1977. Strategies for coping with drought in the Sahel: a study of recent population movement in the Department of Maradi, Niger. Ph.D. thesis, Clark University. Campbell, J.M. 1962. Cultural succession at Anaktuvuk Pass, arctic Alaska, in J.M. Campbell, ed., Prehistoric Cultural Relations between the Arctic and Temperate Zones of North America. (Arctic Institute of North America Technical Paper No. 11). 1968a. Current research: Arctic. American Antiquity 33: 272-8. 1968b. The Kavik site of Anaktuvuk Pass, central Brooks Range, Alaska. Anthropological Papers of the University of Alaska 14: 32-42. 1968c. Territorially among ancient hunters: interpretations from ethnography and nature, in B. Meggars, ed., Anthropological Archaeology in the Americas. Anthropological Society of Washington. 1976. The nature of Nunamiut archaeology, in E.S. Hall Jr, ed., Contributions to anthropology: the interior peoples of northern Alaska. Museum of Man Mercury Series, Archaeological Survey of Canada, Paper 49: 1-51.

1978. Aboriginal human overkill of game populations: examples from interior north Alaska, in R.C. Dunnell and E.S. Hall Jr, eds., Archaeological Essays in Honor of Irving B. Rouse. The Hague, Mouton Publishers. Carroll, G.M. and Smithhisler, J.R. 1980. Observations of bowhead whales during spring migration. Marine Fisheries Review. 42 (9-10): 318-23. Carter, E. and Stolper, M. 1984. Elam: Surveys of Political History and Archaeology (Near Eastern Studies Vol. 25). Berkeley, University of California Press. Cartledge, P.A. 1983. 'Trade and Politics' revisited: Archaic Greece, in P.D.A. Garnsey, K. Hopkins and C.R. Whittaker, eds.. Trade in the Ancient Economy. London, Chatto and Windus. Cashdan, E.A. 1983. Territonality among human foragers: ecological models and an application to four Bushman groups. Current Anthropology 24:47-66. 1985. Coping with risk: reciprocity among the Basarwa of northern Botswana. Man 20:454-74. Casteel, R.W. 1976. Fish Remains in Archaeology and Palaeo-environmental Studies. London, Academic Press. Chagnon, N.A. 1968. Yanomamo: the Fierce People. New York, Holt, Rinehart and Winston. 1979. Mate competition, favoring close kin, and village fissioning among the Yanomamo Indians, in N.A. Chagnon and W. Irons, eds., Evolutionary Biology and Human Social Behavior: an Anthropological Perspective. North Scituate, Duxbury. Champion, T., Gamble, C , Shennan, S. and Whittle, A. 1984. Prehistoric Europe. London, Academic Press. Champlain, S.D. 1907. Voyages of Samuel De Champlain 1604-1618 (ed. W.L. Grant). New York, Barnes and Noble Inc. Chardon, P. 1932. Char don's Journal at Fort Clark 1834-1839. Pierre, State Department of History. Chayanov, A.V. 1966. On the Theory of Peasant Economy. Homewood, American Economic Association. Childe, V.G. 1954. What Happened in History. Harmondsworth, Penguin. Christodoulou, D. 1959. The Evolution of the Rural Land Use Pattern in Cyprus (World Land Use Survey Regional Monograph 2). Bude, Geographical Publications. Cipolla, CM. 1980. Before the Industrial Revolution. European Society and Economy, 1000-1700 (2nd edn.). London, Methuen. Clark, G.A. and Strauss, L.G. 1983. Late pleistocene hunter-gatherer adaptations in Cantabrian Spain, in G.N. Bailey, ed., Hunter-Gatherer Economy in Prehistory. Cambridge University Press. Clark, J.G.D. 1952. Prehistoric Europe. The Economic Basis. London, Methuen. 1975. The Earlier Stone Age Settlement of Scandinavia. Cambridge University Press. Clarke, D.L. 1976. Mesolithic Europe: the economic basis, in G. de G. Sieveking, I.H. Longworth and K.E. Wilson, eds., Problems in Economic and Social Archaeology. London, Duckworth. Clason, A.T. 1980. Padina and Starcevo: game, fish and cattle. Palaeohistoria 22: 141-73. Cleland, C.B. 1982. The inland shorefisheryof the northern Great Lakes: its development and importance in prehistory. American Antiquity 47 (4): 761-84. Climatological Data 1950. Climatological Data 36 (1-12). San Francisco, US Department of Commerce, Weather Bureau. Colson, E. 1979. In good years and bad: food strategies of self-reliant societies. Journal of Anthropological Research 35 (1): 18-29. Corbin, J.E. 1976. Early historic Nunamiut house types, in E.S. Hall Jr, ed., Contributions to Anthropology: the Interior Peoples of Northern Alaska. Museum of Man Mercury Series, Archaeological Survey of Canada, Paper 49: 135-76. Coues, E., ed. 1893. The History of the Lewis and Clark Expedition. New York, Dover Publications Inc. Cropper, J.P. 1982. Climate reconstructions (1801 to 1938) inferred from

References

130

tree-ring width chronologies of the North American arctic. Alpine and Arctic Research 14: 223-41. Cropper, J.P. and Fritts, H.C. 1981. Tree-ring chronologies from the North American Arctic. Arctic and Alpine Research 13: 245-60. Dahl, G. 1979. Ecology and equality: the Boran case, in Pastoral Production and Society. Cambridge University Press. Dahl, G. and Hjort, A. 1976. Having Herds: Pastoral Herd Growth and Household Economy (Stockholm Studies in Social Anthropology 2). University of Stockholm. Dauphine, T.W. 1976. Biology of the Kaminuriak population of barren ground caribou, Part 4: Growth reproduction and energy reserves. Canadian Wildlife Service Report Series 38. Davis, C.W., Linck, D.C., Schoenberg, K.M. and Shields, H.M. 1981. Slogging, Humping, and Mucking through the NPR-A: An Archaeological Interlude (5 vols.). Fairbanks, Anthropology and Historic Preservation Cooperative Park Studies Unit, University of Alaska. Davis, J. 1977. People of the Mediterranean. An Essay in Comparative Social Anthropology. London, Routledge and Kegan Paul. Davis, J.L., Valkenburg, P. and Reynolds, H.V. 1979. Population dynamics of Alaska's western arctic caribou herd, in E. Reimers, E. Gaare and S. Skjenneberg, eds., Proceedings of the Second International ReindeerI Caribou Symposium, September 1979, Roros, Norway. Trondheim, Direktoratet for Vilt og Ferskvannskisk. Dekin, A. A. Jr. 1981. Utkiavik Archaeology Project, 1981, management summary. Anchorage, National Park Service. Dekker, R.M. 1982. Holland in Beroering: Oproeren in de lie en 18e eeuw, Baarn, Ambo. 1987. Women in revolt. Popular protest and its social basis in Holland in the seventeenth and eighteenth century. Theory and Society 16: 337-62. Delille, G. 1977. Agricoltura e Demografia nel Regno di Napoli nei Secoli XVIII e XIX. Naples, Guida Editori. Dennell, R.W. 1974. The purity of prehistoric crops. Proceedings of the Prehistoric Society 40: 132-5. Diarra, S. 1975. Les problemes de contact entre les pasteurs peuls et les agriculteurs dans le Niger central, in T. Monod, ed., Pastoralism in Tropical Africa. London, Oxford University Press. Diaz, H.F. and Andrews, J.T. 1982. Analysis of the spatial pattern of July temperature departures (1943-1972) over Canada and estimates of the 700mb midsummer circulation during middle and late Holocene. Journal of Climatology 2: 251-65. Dillen, J.G. van 1915. Duurtemaatregelen te Amsterdam in de Zeventiende Eeuw, Amsterdam, Kruyt. 1970. Van Rijkdom en Regenten (5th edn.). The Hague, Martinus Nijhoff. Dittenberger, W. 1915-24. Sylloge Inscriptionum Graecarum (3rd edn., 4 vols.). Leipzig, apud S. Hirzelium. Dodd, C.F. 1984. Ontario Iroquois tradition longhouses. National Museum of Man Mercury Series. Archaeological Survey of Canada, Paper 124:181-437. Doerr, J.G. 1979. Modeling the population decline of two Alaskan caribou herds, in E. Reimers, E. Gaare and S. Skjenneberg, eds., Proceedings of the Second International ReindeerI Caribou Symposium, September 1979, Roros, Norway. Trondheim, Direktoratet for Vilt og Ferskvannskisk. Dollfus, G. and Wright, H.T. 1985. Actes du Seminaire CNRS/NSF de Bellevaux (24-29 Juin 1985). Paleorient 11 (2): 5-130. Dolukhanov, P. 1979. Ecology and Economy in Neolithic Eastern Europe. London, Duckworth. 1986. The late mesolithic and the transition to food production in eastern Europe, in M. Zvelebil, ed., Hunters in Transition. Mesolithic Societies of Temperate Eurasia and their Transition to Farming. Cambridge University Press. Dolukhanov, P., Romanova, Ye.N. and Semyontsov, A.A. 1970. Radiocarbon dates of the Institute of Archaeology II. Radiocarbon 12:

130-55. Donald, L. and Mitchell, D.H. 1975. Some correlates of local group rank among the Southern Kwakiutl. Ethnology 14: 325-46. Druker, P. 1951. The Northern and Central Nootkan Tribes. Bureau of American Ethnology, Bulletin 144. Duckham, A.N. and Masefield, G.B. 1970. Farming Systems of the World. London, Chatto and Windus. Dunbar, J.B. 1918. Letters from Rev. John Dunbar, September 6, 1831 to May 7, 1849. Collections of the Kansas State Historical Society 14: 570-689. Dunbar, J.B. Jr. 1880. The Pawnee Indians, their history and ethnology. Magazine of American History 4: 241 -81. Dunnell, R.C. 1980. Evolutionary theory and archaeology, in M.B. Schiffer, ed., Advances in Archaeological Method and Theory 3. New York, Academic Press. Dupire, M. 1962a. Peuls Nomades: Etude Descriptive des Wodaabe du Sahel Nigerien. Paris, Institut d'Ethnologie. 1962b. Trade and markets in the economy of the nomadic Fulani of Niger (Bororo), in P. Bohannan and G. Dalton, eds., Markets in Africa. Evanston, Northwestern University Press. Dyson-Hudson, N. 1984. Adaptive resource use by African pastoralists, in F. di Castri, F.W.G. Baker and M. Hadley, eds., Ecology in Practice, Part I, Ecosystems Management. Paris, UNESCO. Dyson-Hudson, R. and Smith, E.A. 1978. Human territoriality: an ecological assessment. American Anthropologist 80: 21-41. Ehrenberg, V. 1937. When did the polis rise? Journal of Hellenic Studies 57: 147-59. Eidlitz, K. 1969. Food and Emergency Food in the Circumpolar Zone (Studia Ethnographica Upsaliensia 32). Uppsala, Uppsala University Press. Ellen, R. 1982. Environment, Subsistence and System: the Ecology of SmallScale Social Formations. Cambridge University Press. Ellis, J.E., Jennings, C.H. and Swift, D.M. 1979. A comparison of energy flow among the grazing animals of different societies. Human Ecology 1 (2): 135-49. Ellis, J.M. and Calkin, P.E. 1984. Chronology of Holocene glaciation, central Brooks Range, Alaska. Geological Society of America Bulletin 95:897-912. Euler, R.C, Gumerman, G.J., Karlstrom, T.N.V., Dean, J.S. and Hevly, R.H. 1979. The Colorado Plateaus: cultural dynamics and paleoenvironment. Science 205: 1089-101. Fay, F.H. 1974. The role of ice in the ecology of marine mammals of the Bering Sea, in D.W. Hood and E.J. Kelley, eds., Oceanography of the Bering Sea. Fairbanks, University of Alaska. Finley, M.I. 1957. The mycenaean tablets and economic history. Economic History Review 10: 128-41. 1970. Aristotle and economic analysis. Past and Present 47: 3-25. 1978. The World of Odysseus (3rd revised edn.). London, Chatto and Windus. 1980. Ancient Slavery and Modern Ideology. London, Chatto and Windus. 1983. Politics in the Ancient World. Cambridge University Press. 1985. The Ancient Economy (2nd edn.). London, The Hogarth Press. Fitzhugh, W.W. and Kaplan, S.A. eds. 1982. lnua: Spirit World of the Bering Sea Eskimo. Washington, Smithsonian Institution Press. Flannery, K.V. 1969. Origins and ecological effects of early Near Eastern domestication, in P.J. Ucko and G.W. Dimbleby, eds., The Domestication and Exploitation of Plants and Animals. London, Duckworth. 1972. The origins of the village as a settlement type in Mesoamerica and the Near East: a comparative study, in P.J. Ucko, R. Tringham and G.W. Dimbleby, eds., Man, Settlement and Urbanism. London, Duckworth. 1974. The Olmec and the valley of Oaxaca: a model for inter-regional interaction in Formative times, in J.A. Sabloff and C.C. LambergKarlovsky, eds., The Rise and Fall of Civilizations. Menlo Park,

References

131

Benjamin/Cummings. ed. 1976. The Early Mesoamerican Village. New York, Academic Press. 1986. Guila Naquitz. Orlando, Academic Press. Fletcher, A. and La Flesche, F. 1911. The Omaha Tribe. Bureau of American Ethnology, Annual Report (1906) 27: 17-654. Fletcher, R. 1981. People and space: a case study on material behaviour, in I. Hodder, G. Isaac and N. Hammond, eds., Pattern of the Past, Studies in Honour of David Clarke. Cambridge University Press. Fleuret, A. 1986. Indigenous responses to drought in sub-Saharan Africa. Disasters 10 (3): 224-9. Foote, D.C. and Williamson, H.A. 1966. A human geographic study, in N.J. Wilimovsky and J.N. Wolfe, eds., Environment of the Cape Thompson Region, Alaska (US Atomic Energy Commission). Washington, US Government Printing Office. Forbes, H.A. 1982. Strategies and Soils: Technology, Production and Environment in the Peninsula ofMethana, Greece. Ann Arbor, University Microfilms. Ford, J.A. 1959. Eskimo Prehistory in the Vicinity of Point Barrow, Alaska. Anthropological Papers of the American Museum ofNatural History 47 (1). Ford, R.I. 1972. Barter, gift or violence: an analysis of Tewa intertribal exchange. University of Michigan Museum of Anthropology, Anthropological Papers 46: 21-46. 1977. Evolutionary ecology and the evolution of human ecosystems: a case study from the midwestern U.S.A., in J. Hill, ed., Explanation of Prehistoric Change. Albuquerque, University of New Mexico Press. Forde, C D . 1963. Habitat, Economy and Society. London, Methuen. Forge, A. 1972. Normative factors in the settlement size of neolithic cultivators (New Guinea), in P.J. Ucko, R. Tringham and G.W. Dimbleby, eds., Man, Settlement and Urbanism. London, Duckworth. Fornara, C.W. 1983. Archaic Times to the End of the Peloponnesian War (2nd edn.). Cambridge University Press. Forrest, W.G. 1966. The Emergence of Greek Democracy. London, Weidenfeld and Nicholson. Foxhall, L. 1986. Greece ancient and modern: subsistence and survival. History Today 36: 35-43. Francotte, H. 1905. Le pain a bon marche et le pain gratuit dans les cites Grecques, in Melanges Jules Nicole. Geneve. Franklin, U.M., Badone, E., Gotthardt, R. and Yorga, B. 1981. An Examination of Prehistoric Copper Technology and Copper Sources in Western Arctic and Subarctic North America. National Museum of Man Mercury Series, Archaeological Survey of Canada, Paper 101. Friedman, J. 1984. Tribes, states, and transformations, in M. Bloch, ed., Marxist Analyses and Social Anthropology. London, Tavistock Publications. Friedman, J. and Rowlands, M.J. 1977. Notes towards an epigenetic model of the evolution of'civilisation', in J. Friedman and M.J. Rowlands, eds., The Evolution of Social Systems. London, Duckworth. Fritts, H.C. 1976. Tree Rings and Climate. Academic Press, New York. Fritts, H . C , Biasing, T.J., Hayden, B.P. and Kutzbach, J.E. 1971. Multivariate techniques for specifying tree-growth and climate relationships and for reconstructing anomalies in paleoclimate. Journal of Applied Meteorology 10: 844-64. Frost, F.J. 1984. The Athenian military before Cleisthenes. Historia 33: 285-94. Gal, R. 1982a. Archaeological sites in the vicinity of the South Meade Test Well. Anthropological Papers of the University of Alaska 18 (1): 115. 1982b. Appendix I: an annotated and indexed roster of archaeological radiocarbon dates from Alaska, north of 68° latitude. Anthropological Papers of the University of Alaska 20 (1-2): 159-80. Gallant, T.W. 1982. An examination of two island polities in antiquity: the Lefkas-Pronnoi Survey. Ph.D. dissertation, University of Cambridge.

Gamble, C 1978. Resource exploitation and the spatial patterning of hunter-gatherers: a case study, in D. Green, C Haselgrove and M. Spriggs, eds., Social Organisation and Settlement. Oxford, British Archaeological Reports (International Series 47). 1981. Social control and the economy, in A. Sheridan and G.N. Bailey, eds., Economic Archaeology: towards an Integration of Ecological and Social Approaches. Oxford, British Archaeological Reports (International Series 96). 1982. Leadership and 'surplus' production, in C Renfrew and S. Shennan, eds., Ranking, Resource and Exchange. Cambridge University Press. 1986. The mesolithic sandwich: ecological approaches and the archaeological record of the early postglacial, in M. Zvelebil, ed., Hunters in Transition. Mesolithic Societies of Temperate Eurasia and their Transition to Farming. Cambridge University Press. Garfinkel, H.L. and Brubaker, L.H. 1980. Modern climate-tree growth relationships and climate reconstruction in sub-arctic Alaska. Nature 286:872-4. Garnsey, P.D.A. 1983a. Grain for Rome, in P.D.A. Garnsey, K. Hopkins and C.R. Whittaker, eds., Trade in the Ancient Economy. London, Chatto and Windus. 1983b. Famine in Rome, in P.D.A. Garnsey and C.R. Whittaker, eds., Trade and Famine in Classical Antiquity (supp. vol. 8). Cambridge Philological Society. 1988. Famine and Food Supply in the Graeco-Roman World: Responses to Risk and Crisis. Cambridge University Press. Garnsey, P., Gallant, T. and Rathbone, D. 1984. Thessaly and the grain supply of Rome during the second century BC. Journal of Roman Studies 74: 3 0 ^ 4 . Gauthier, Ph. 1981. De Lysias a Aristotle (Ath.pol.51.4): le commerce du grain a Athenes et les fonctions des sitophylaques. Revue Historique de Droit Francais et Etranger 59: 5-28. 1985. Les cites grecques et leurs bienfaiteurs, IVe-Ier siecles av. JC: contribution a Fhistoire des institutions. Bulletin de Correspondance Hellenique (suppl. 12). van Gennep, A. 1909. Les Rites de Passage. Paris, Emile Nourry. Gerlach, S.C 1982. A summary of archaeological research at Tukuto Lake, National Petroleum Reserve, Alaska, in E.S. Hall, Jr, ed., A Review of Cultural Resource Survey and Clearance Activities: National Petroleum Reserve in Alaska, 1977-1982. Anchorage, U.S. Geological Survey. Gerlach, S.C and Hall, E.S. Jr. n.d. The later prehistory of northern Alaska: the view from Tukuto Lake (unpublished ms.). Giddens, A. 1981. A Contemporary Critique of Historical Materialism, 1. Power, Property and the State. London, Macmillan. Giddings, J.L. 1941. Dendrochronology in northern Alaska. Laboratory of Tree-Ring Research Bulletin 1. 1942. Dated sites on the Kobuk River, Alaska. Tree-Ring Bulletin 9(1): 2-8. 1943. Some climatic aspects of tree growth in Alaska. Tree-Ring Bulletin 9 (4): 26-32. 1948. Chronology of the Kobuk-Kotzebue sites. Tree-Ring Bulletin 14 (4): 26-32. 1952. The Arctic Woodland culture of the Kobuk River. University of Pennsylvania, Museum Monographs 8. 1956. Forest eskimos: an ethnographic sketch of the Kobuk River people in the 1880's. University of Pennsylvania, Museum Bulletin 20(2): 1-55. 1962. Onion Portage and other flint sites of the Kobuk River. Arctic Anthropology 1 (1): 6-27. 1967. Ancient Men of the Arctic. New York, Alfred A. Knopf. Gilmore, M.R. 1977. Uses of Plants by the Indians of the Missouri River Region. Lincoln, University of Nebraska Press. Gimbutas, M. 1974. Akhillion: a neolithic mound in Thessaly; preliminary report on 1973 and 1974 excavations. Journal of Field Archaeology 1:277-302.

References

132

Goldstein, L. 1976. Spatial structure and social organization: regional manifestations of Mississippian society. Ph.D. thesis, Northwestern University, Chicago. 1981. One-dimensional archaeology and multi-dimensional people: spatial organisation and mortuary analysis, in R. Chapman, I. Kinnes and K. Randsborg, eds., The Archaeology of Death. Cambridge University Press. Goody, J. 1976. Production and Reproduction (Cambridge Studies in Social Anthropology 17). Cambridge University Press. Gossow, H. 1979. Preliminary studies in population ecology of the Svalbard reindeer in Nordenskiold Land, post-calving season 1972, in E. Reimers, E. Gaare, and S. Skjenneberg, eds., Proceedings of the Second International ReindeerI Caribou Symposium, September 1979, Reros, Norway. Trondheim, Direktoratet for Vilt og Ferskvannskisk. Graumlich, L.J. and Brubaker, L.B. 1986. Reconstruction of annual temperature (1590-1979) for Longmire, Washington, derived from tree rings. Quaternary Research 25: 223-34. Greenough, P.R. 1982. Prosperity and Misery in Modern Bengal. The Famine of 1943-1944. Oxford University Press. Griffin, J.B. 1960. Climate change: a contributory cause of the growth and decline of northern Hopewell culture. The Wisconsin Archaeologist 41 (2): 21-33. Grigg, D.B. 1974. The Agricultural Systems of the World: an Evolutionary Approach. Cambridge University Press. 1980. Population Growth and Agrarian Change. A Historical Perspective. Cambridge University Press. Gronnow, B. in press. Meiendorf and Stellmoor revisited. An analysis of late palaeolithic reindeer exploitation. Acta Archaeologica 56. Gronnow, B., Meldgaard, M. and Nielsen, J.B. 1983. Aasivisuit - the Great Summer Camp. Archaeological, Ethnographical and Zoo-Archaeological Studies of a Caribou-Hunting Site in West Greenland. Meddelelser om Gronland, Man and Society 5. Gubser, N.J. 1965. The Nunamiut Eskimo: Hunters of Caribou. New Haven, Yale University Press. Gurina, N.N. 1967. Iz istorii drevnykh remen Zapadnykh oblastei SSSR. Mia 144: 1-205. Haber, G.C. and Walters, C J . 1979. Dynamic of the Alaska-Yukon caribou herds and management implications, in E. Reimers, E. Gaare, and S. Skjenneberg, eds., Proceedings of the Second International ReindeerI Caribou Symposium, September 1979, Roros, Norway. Trondheim, Direktoratet for Vilt og Ferskvannskisk. Hadleigh-West, F. 1966. Archaeology of Ogotoruk Creek, in N.J. Wilimovsky and J.N. Wolfe, eds., Environment of the Cape Thompson Region, Alaska (US Atomic Energy Commission). Washington, US Government Printing Office. Hagberg, U.E. 1967. The Archaeology ofSkedemosse, I: The Excavations and the Finds of an Oland Fen, Sweden. Stockholm, Kungliga Vitterhets Historie och Antikvitets Akademien. Hall, A.E., Cannell, G.H. and Lawton, H.W., eds. 1979. Agriculture in Semi-Arid Environments. Berlin, Springer-Verlag. Hall, E.S., Jr. 1969. Speculations on the late prehistory of the Kutchin Athapaskans. Ethnohistory 16 (4): 317-34. 1970. The late prehistoric/early historic eskimo of interior northern Alaska: an ethnoarchaeological approach? Anthropological Papers of the University of Alaska 15(1): 1-11. 1971. Kangiguksuk: a cultural reconstruction of a sixteenth century eskimo site in northern Alaska. Arctic Anthropology 8(1): 1-101. 1975. An archaeological survey of interior northwest Alaska. Anthropological papers of the University of Alaska 17 (2): 13-30. 1976. A preliminary analysis of house types at Tukuto Lake, northern Alaska, in E.S. Hall, Jr, ed., Contributions to Anthropology: the Interior Peoples of Northern Alaska. National Museum of Man Mercury Series, Anthropological Survey of Canada Paper 49: 98134. 1978. Technological change in northern Alaska, in R.C. Dunnell and E.S. Hall, Jr, eds., Archaeological Essays in Honor of Irving B.

Rouse. The Hague, Mouton Publishers. 1981. Cultural resource site potential, in Cultural Resources in the midBeaufort Sea Region (Coastal Zone Management Plan). North Slope Borough. Halstead, P. 1977. Prehistoric Thessaly: the submergence of civilisation, in J.L. Bintliff, ed., Mycenaean Geography. Cambridge, British Association for Mycenaean Studies. 1980. Review of E. Hanschmann and V. Milojcic (1976) ArgissaMagula 3: die fruhe und beginnende mittlere Bronzezeit, and of V. Milojcic, A. von den Driesch, K. Enderle, J. Milojcic-v.Zumbusch and K. Kilian (1976) Magulen um Larisa. Journal of Hellenic Studies 100: 267-8. 1981a. Counting sheep in neolithic and brcnze age Greece, in I. Hodder, G. Isaac and N. Hammond, eds., Pattern of the Past: Studies in Honour of David Clarke. Cambridge University Press. 1981b. From determinism to uncertainty: social storage and the rise of the Minoan palace, in A. Sheridan and G.N. Bailey, eds., Economic Archaeology: towards an Integration of Ecological and Social Approaches. Oxford, British Archaeological Reports (International Series 96). 1984. Strategies for survival: an ecological approach to social and economic change in the early farming communities of Thessaly, N. Greece. Ph.D. dissertation, University of Cambridge. 1987a. Man and other animals in later Greek prehistory. Annual of the British School at Athens 82: 71-83. 1987b. Traditional and ancient rural economy in Mediterranean Europe: plus 9a change? Journal of Hellenic Studies 107: 77-87. 1988. On redistribution and the origin of minoan-mycenaean palatial economies, in E.B. French and K.A. Wardle, eds., Problems in Greek Prehistory. Bristol Classical Press. Halstead, P. and Jones, G. 1980. Early neolithic economy in Thessaly some evidence from excavations at Prodromos, Anthropoloyika 1: 93—117. Halstead, P. and O'Shea, J. 1982. A friend in need is a friend indeed: social storage and the origins of social ranking, in C. Renfrew and S. Shennan, eds., Ranking, Resource and Exchange. Cambridge University Press. Hands, A.R. 1968. Charities and Social Aid in Greece and Rome. London, Thames and Hudson. Hansen, M.H. 1986. Demography and Democracy. Denmark, Herning. Harding, P. 1985. From the End of the Peloponnesian War to the Battle of Ipsus. Cambridge University Press. Haresign, W., ed. 1983. Sheep Production. London, Butterworths. Harris, M. 1959. The economy has no surplus? American Anthropologist 61:185-99. Haugen, R.K. and Brown, J. 1980. Coastal-inland distributions of summer air temperature and precipitation in northern Alaska. Arctic and Alpine Research 12: 403-12. Hayden, B. 1977. Corporate groups and the late Ontario Iroquoian longhouse. Ontario Archaeology 28: 3-16. 1981. Research and development in the Stone Age: technological transitions among hunter-gatherers. Current Anthropology 22: 519-48. Hayden, B., Chisholm, B. and Schwarcz, H.P. 1987. Fishing and foraging: marine resources in the Upper Palaeolithic of France, in O. Soffer, ed., The Pleistocene Old World. New York, Plenum Press. Heidenreich, C.E. 1971. Huronia: A History and Geography of the Huron Indians, 1600-1650. Toronto, McClelland and Stewart. 1978. Huron, in B.G. Trigger, ed., The Handbook of North American Indians (vol. 15). Washington, Smithsonian Institution. Heller, J. 1964. Catch 22. London, Penguin. Hemming, J.E. 1971. The distributional movement patterns of caribou in Alaska. Wildlife Technical Bulletin 1 (Alaska Department of Fish and Game). 1975. Population growth and movement patterns of the Nelchina caribou herd: proceedings of the first international reindeer and caribou symposium. Biological Papers of the University of Alaska, Special Report 1: 162-9.

References

133

Hennigh, L. 1983. North Alaskan Eskimo alliance structure. Arctic Anthropology 20: 23-32. Henry, A. 1921. Alexander Henry's Travels and Adventures in the Years 1760-1776 (ed. M. Quaife). Chicago, R.R. Donnelley and Sons Co. Hickey, C.G. 1976. An economic view of adaptation, in E.S. Hall, Jr, ed. Contributions to Anthropology: the Interior Peoples of Northern Alaska. National Museum of Man Mercury Series, Archaeological Survey of Canada, Paper 49: 235-98. 1979. The historic Beringian trade network: its nature and origins, in A.P. McCartney, ed., Thule Eskimo Culture: An Anthropological Retrospective. National Museum of Man Mercury Series, Archaeological Survey of Canada, Paper 88: 411-34. Hill, P. 1972. Rural Hausa. Cambridge University Press. Hillman, G.C. 1973. Agricultural productivity and past population potential at As.van. Anatolian Studies 23: 225-40. 1981. Reconstructing crop husbandry practices from charred remains of crops, in R.J. Mercer, ed., Farming Practice in British Prehistory. Edinburgh University Press. Hinsdale, W.B. 1928. Indian Corn Culture in Michigan. Michigan Handbook Series 4. Hinz, G. 1979. Neue Tierknochenfunde aus der Magula Pevkakia in Thessalien, 1: die Nichtwiederkaiier. Dissertation, University of Munich. Hodder, I. 1982. Symbols in Action. Cambridge University Press. 1984. Burials, houses, women and men in the European Neolithic, in D. Miller and C. Tilley, eds., Ideology, Power and Prehistory. Cambridge, University Press. Hole, F., ed. 1986. Archaeological perspectives on Western Iran. Washington, Smithsonian Institution Press. Holen, S.R. 1983. Lower Loup lithic procurement strategy and the Gray Site. M.A. thesis, University of Nebraska-Lincoln. Hont, I. and Ignatieff, M. 1983. Needs and justice in the 'Wealth of Nations': an introductory essay, in I. Hont and M. Ignatieff, eds., Wealth and Virtue. The Shaping of Political Economy in the Scottish Enlightenment. Cambridge University Press. Hopkins, K. 1978. Conquerers and Slaves. Cambridge University Press. 1983. Models, ships and staples, in P.D.A. Garnsey and C.R. Whittaker, eds., Trade and Famine in Classical Antiquity (supp. vol. 8). Cambridge Philological Society. Horowitz, M.M. 1972. Ethnic boundary maintenance among pastoralists and farmers in the western Sudan (Niger), in W. Irons and N. Dyson-Hudson, eds., Perspectives on Nomadism. Leiden, E.J. Brill. Hourmouziadhis, Y. 1979. To Neolithiko Dhimini. Volos, Society for Thessalian Studies. 1980. Isagoyi sto neolithiko tropo paragoyis, 1. Anthropoloyika 1: 11829. Housley, R.A. 1981. The palaeoethnobotany of Servia: a neolithic and early bronze age Site in Western Macedonia, Greece. Dissertation, University of Southampton. Howe, T.P. 1958. Linear B and Hesiod's breadwinners. Transactions of the American Philological Association 89: 44-65. Hughes, R.C. 1974. Proposed Noatak National Arctic Range, Alaska: Final Environmental Statement. Alaska Planning Group, US Department of the Interior. Hughes-Parry, J. 1949. Fishing Fantasy. London, Eyre and Spottiswoode. Hunt, D.W.S. 1947. Feudal survivals in Ionia. Journal of Hellenic Studies 67: 68-76. Iakovidhis, S. 1977. The present state of research at the citadel of Mycenae. Bulletin of the Institute of Archaeology, London 14: 99-141. Indreko, R. 1948. Die Mittlere Steinzeit in Estland. Stockholm, Kungliga Vitterhets Historie och Antikvitets Akademien 66. Irving, W.N. 1962. 1961 field work in the western Brooks Range, Alaska: preliminary report. Arctic Anthropology 1 (1): 76-83. Jacobi, R.M. 1987. Misanthropic miscellany: musings on British early Flandrian archaeology and other flights of fancy, in P. RowleyConwy, M. Zvelebil and H.P. Blankholm, eds., Mesolithic Northwest Europe: Recent Trends. Sheffield, John Collis Publications.

Jacoby, G.C. and Cook, E.R. 1981. Past temperature variations inferred from a 400-year tree-ring chronology from Yukon Territory, Canada. Arctic and Alpine Research 13: 409-18. Jameson, M. 1977-78. Agriculture and slavery in classical Athens. Classical Journam: 122-45. 1983. Famine in the Greek world, in P.D.A. Gamsey and C.R. Whittaker, eds., Trade and Famine in Classical Antiquity (supp. vol. 8). Cambridge Philological Society. Jenkins, G.M. and Watts, D.G. 1968. Spectral Analysis and its Application. San Francisco, Holden-Day. Jenness, D. 1953. Stray notes on the eskimo of arctic Alaska. Anthropological Papers of the University of Alaska 1: 5-13. Jochim, M.A. 1976. Hunter-Gatherer Subsistence and Settlement: A Predictive Model. New York, Academic Press. 1981. Strategies for Survival: Cultural Behavior in an Ecological Context. New York, Academic Press. 1983. Palaeolithic cave art in ecological perspective, in G.N. Bailey, ed., Hunter-Gatherer Economy in Prehistory. Cambridge University Press. Johnson, G.A. 1973. Local exchange and early state development in southwestern Iran. University of Michigan Museum of Anthropology, Anthropological Papers 51. 1978. Information sources and the development of decision-making organizations, in C.L. Redman et al., eds., Social Archaeology: Beyond Subsistence and Dating. New York, Academic Press. Jones, G. 1987. Agricultural practice in Greek prehistory. Annual of the British School at Athens 82: 115-23. Jones, G., Wardle, K., Halstead, P., and Wardle, D. 1986. Crop storage at Assiros. Scientific American 254 (3): 96-103. Jongman, W. 1988. The Economy and Society of Pompeii. Amsterdam, Gieben. Jordan, B. 1975. Tierknochenfunde aus der Magula Pevkakia in Thessalien. Dissertation, University of Munich. Kaplan, S.L. 1976. Bread, Politics and Political Economy in the Reign of Louis XV (2 vols.). The Hague, Martinus Nijhoff. 1982. Le Complot de Famine: Histoire d'une Rumeur au XVIIIe Siecle (Cahiers des Annales No. 39). Paris, Colin. Kay, P.A. 1978. Dendroecology in Canada's forest-tundra transition zone. Arctic and Alpine Research 10(1): 133-8. Kayser, B. and Thompson, K. 1964. Economic and Social Atlas of Greece. Athens, National Statistical Service of Greece, Centre of Economic Research, Social Sciences Centre. Kelsall, J.P. 1968. The Migratory Barren-Ground Caribou of Canada (Canadian Wildlife Service). Ottawa, Queen's Printer. Kelsall, J.P. and Klein, D.R. 1979. The state of knowledge of the Porcupine Caribou Herd. Transactions of the North American Wildlife and Natural Resource Conference 44. Keramopoullos, A.D. 1930. Ai viomikhaniai kai to emborion tou Kadhmou. Arkhaioloyiki Efimeris 1930: 29-58. Khalikiopoulos, L. 1905. Wirtschaftsgeographische Skizze Thessaliens. Geographische Zeitschrift

11: 445-75.

Killen, J.T. 1984. The textile industries at Pylos and Knossos, in T.G. Palaima and C.W. Shelmerdine, eds., Pylos Comes Alive. New York, Lincoln Center, Fordham University. 1985. The Linear B tablets and the mycenaean economy, in A. Morpurgo Davies and Y. Duhoux, eds., Linear B: a 1984 Survey. University of Louvain Press. Kinietz, W.V. 1940. The Indians of the Western Great Lakes, 1615-1760. University of Michigan Museum of Anthropology, Occasional Contributions 10. 1965. The Indians of the Western Great Lakes, 1615-1760. Ann Arbor: University of Michigan Press. Kistemaker, R. 1983. Brood, bakkers en de stedelijke overheid. Ons Amsterdam 35:219-24. Klein, D.R. 1968. The introduction, increase and crash of reindeer on St. Matthew Island. Journal of Wildlife Management 32: 350-67. 1979. Range ecology and management - progress made and directions

References

134

for the future, in E. Reimers, E. Gaare, and S. Skjenneberg, eds., Proceedings of the Second International Reindeer/Caribou Symposium, September 1979, Roros, Norway. Trondheim, Direktoratet for Vilt og Ferskvannskisk. Klein, D.R. and White, R.G. eds. 1978. Parameters of caribou population ecology in Alaska: proceedings of a symposium and workshop. Biological Papers of the University of Alaska Special Report 3. Klein, R.G., Allwarden K. and Wolf C. 1983. The calculation and interpretation of ungulate age profiles from dental crown heights, in G.N. Bailey, ed., Hunter-Gatherer Economy in Prehistory. Cambridge University Press. Klindt-Jensen, O. 1957. Denmark. London, Thames and Hudson. Knight, F.H. 1921. Risk, Uncertainty and Profit. Boston/New York, Houghton Mifflin Company. Knight, J. and Starr, M. 1984. Water, in J.J. Swift, ed., Pastoral Development in Central Niger: Report of the Niger Range and Livestock Project. Niamey, MDR/USAID. Kotsakis, K. 1981. Tria oikimata tou oikismou tou Sesklou: anaskafiki erevna. Anthropoloyika 2: 87-108. 1983. Keramiki tekhnoloyia kai keramiki dhiaforopoiisi: provlimata tis graptis keramikis tis mesis neolithikis epokhis tou Sesklou. Ph.D. dissertation, University of Thessaloniki. 1986. The social and economic formation of neolithic Sesklo. (paper presented at the centenary conference of British School at Athens, Cambridge). Kroll, H.J. 1979. Kulturpflanzen aus Dimini. Archaeo-Physika 8: 173-89. 1983. Kastanas, Ausgrabungen in einen Siedlungshiigelder Bronze- und Eisenzeit Makedoniens 1975-1979: die Pflanzenfunde. Berlin, Volker Spiess. Kummerow, J., McMaster, G.S. and Krause, D.A. 1980. Temperature effect on growth and nutrient contents in Eriophorum vaginatum under controlled environmental conditions. Arctic and Alpine Research 12: 335-41. Labrousse, C.E. 1943. La Crise de I'Economie Francaise a la Fin de I'Ancien Regime et au Debut de la Revolution (vol. 1). Paris, Presses Universitaires de France. Lamb, H.H. 1965. The early medieval warm epoch and its sequel. Palaeogeography, Palaeoclimatology, and Palaeoecology 1: 13-37. 1966. The Changing Climate. London, Methuen. 1977. Climate: Present, Past and Future. 2: Climatic History and the Future. London, Methuen. Larsen, H. and Rainey, F.G. 1948. Ipiutak and the Arctic whale hunting culture. American Museum of Natural History, Anthropological Papers 42. Le Brun, A. 1980. Les 'ecuelles grossieres': etat de la question, in L'Archeologie de 1'Iraq: perspectives et limites de Finterpretation des documents. Colloques Internationaux du CNRS 580: 59-70. Le Brun, A. and Vallat, F. 1978. L'origine de l'ecriture a Suse. Cahiers de la Delegation Archeologique Frangaise en Iran 8: 11-59. Le Roy Ladurie, E. 1966. Les Paysans de Languedoc (2 vols.). Paris (Bibliotheque Generate de TEcole des Hautes Etudes. Vic section), SEVPEN. Leake, W.M. 1967. Travels in Northern Greece, 4. Amsterdam, Hakkert. Leclair, E. Jr and Schneider, H.K. 1968. Economic Anthropology. Readings in Theory and Analysis. New York, Holt, Rinehart and Winston. Lee, R.B. 1979. The IKung San: Men, Women and Work in a Foraging Society. Cambridge University Press. Lefebure, C. 1979. Introduction, in Pastoral Production and Society. Cambridge University Press. Legge, A.J. 1981. The agricultural economy, in R.J. Mercer, ed., Grimes Graves Excavations 1971-72. London, Her Majesty's Stationery Office. Legge, A.J. and Rowley-Conwy, P.A. 1988. Star Carr Revisited. A Reanalysis of the Large Mammals. London, Astragalus Press. Lehmer, D.J. 1954. The sedentary horizon of the northern Plains. Southwestern Journal of Anthropology 10(2): 139-59.

Lent, P.C. 1966. The caribou of the Cape Thompson region, Alaska, in N.J. Wilimovsky and J.N. Wolfe, eds., Environment of the Cape Thompson Region, Alaska (US Atomic Energy Commission). Washington, US Government Printing Office. Leonard, W.H. and Martin, J.H. 1963. Cereal Crops. New York, Macmillan. Lewthwaite, J.G. 1982. Cardial disorder: ethnographic and archaeological comparisons for problems in the early prehistory of the west Mediterranean, in Le Neolithique Ancien Mediterraneen (Archeologie en Languedoc, No. special). Sete, Federation Archeologique de l'Herault. Libbey, D. 1981. Cultural resource site identification, in Cultural Resources in the mid-Beaufort Sea Region (Coastal Management Plan). North Slope Borough. Liebeschuetz, J.H.W.G. 1972. Antioch. City and Imperial Administration in the Later Roman Empire. Oxford University Press. Lightfoot, K.G. 1979. Food redistribution among prehistoric Pueblo groups. The Kiva 44: 319-39. Lobel, E. and Page, D.L. 1955. Poetarum Lesbiorum Fragmenta. Oxford, Clarendon Press. Loze, LA. 1979. Pozdnii Neolit i Rannyaya Bronza Lubanskoi Ravniny. Riga, Zinatne. Luce, J.V. 1978. The polis in Homer and Hesiod. Proceedings of the Royal Irish Academy 78: 1-15. Luho, V. 1954. Der netzfund von Pori. Suomen Museo 61: 26-7. MacCormack, C.P. 1978. The cultural ecology of production: Sherbro coast and hinterland, Sierra Leone, in D. Green, C. Haselgrove and M. Spriggs, eds., Social Organisation and Settlement. Oxford, British Archaeological Reports (International Series 47). McCown, R.L., Haalan, G. and de Haan, C. 1979. The interaction between cultivation and livestock production in semi-arid Africa. Ecological Studies 34:297-332. McGhee, R. 1969-70. Speculations on climatic change and Thule culture development. Folk 11-12: 173-84. 1976. Paleoeskimo occupations of central and high Arctic Canada, in M.S. Maxwell, ed., Eastern Arctic Prehistory: Paleoeskimo Problems. Memoirs of the Society for American Archaeology 3 1 : 1 5 - 3 9 . McLaren, LA. 1958a. The biology of the ringed seal (Phoca hispida) in the eastern Canadian arctic. Fisheries Research Board of Canada Bulletin 118. 1958b. The economics of seals in the eastern Canadian arctic. Fisheries Research Board of Canada Circular 1. MacNeish, R.S. 1967. A summary of the subsistence, in D. Byers, ed., The Prehistory of the Tehuacan Valley (vol. 1). Austin, University of Texas Press. 1972. The evolution of community patterns in Tehuacan Valley of Mexico and speculations about the cultural processes, in P. Ucko, R. Tringham and G. Dimbleby, eds., Man, Settlement and Urbanism. London, Duckworth. Maher, W J . and Wilimovsky, N.J. 1963. Annual catch of bowhead whales by eskimos at Point Barrow, Alaska, 1928-1960. Journal of Mammalogy 44: 16-20. Maliki, A.B., White, C , Loutan, L. and Swift, J.J. 1984. The Wodaabe, in J.J. Swift, ed., Pastoral Development in Central Niger: Report of the Niger Range and Livestock Project. Niamey, MDR/USAID. Mariolopoulos, E.G. 1962. Fluctuation of rainfall in Attica during the years of the erection of the Parthenon. Geofisicapura e applicata 51: 243-62. 1971. Has the climate changed? Praktika Akademias Athenon 46: 3853. Markevitch, V.I. 1974. Bugo-Dnestrovskaya Kultura na Territorii Moldavii. Kishinev, Shtiintsa. Marquette, W.M. 1977. The 1976 catch of bowhead whales (Balaena mysticetus) by Alaska eskimos, with a review of the fishery, 1973-1976, and a biological summary of the species, in International Whaling Commission's Deletion of Native Exemption for Subsistence Harvest

References

135

ofBowhead Whales, Final Environmental Impact Statement. Wash39-113. ington, National Oceanic and Atmospheric Administration, US Minnis, P.E. 1985. Social Adaptation to Food Stress, A Prehistoric SouthDepartment of Commerce. western Example. University of Chicago Press. Mitchell, E. and Reeves, R.R. 1980. The Alaska bowhead problem: a comMarquette, W.M. and Bockstoce, J.R. 1980. Historical shore-based catch mentary. Arctic 33: 686-723. of bowhead whales in the Bering, Chukchi, and Beaufort seas. Marine Fisheries Review 42 (9-10): 5-19. Mithen, S. 1987. Prehistoric red deer hunting strategies, a cost-risk benefit analysis with reference to upper palaeolithic northern Spain and Marsh, G.H. 1954. A comparative survey of Eskimo-Aleut religion. Anthmesolithic Denmark, in P. Rowley-Conwy, M. Zvelebil and H.P. ropological Papers of the University of Alaska 3: 21-36. Blankholm, eds., Mesolithic Northwest Europe: Recent Trends. Mathiassen, T. 1935. Blubber lamps in the Erteb0lle culture? Acta ArchSheffield, John Collis Publications. aeologia6: 139-51. Mohl, U. 1970a. Oversigt over dyrknoglerne fra 01by Lyng. Aarbogerfor Matsuyama, T. 1981. Nut gathering and processing methods in traditional Japanese villages, in S. Koyama and D.H. Thomas, eds., Affluent Nordisk Oldkyndighed og Historie 1970:43-77. Foragers. Senri Ethnological Studies 9: 117-39. 1970b. Fangstdyrene ved de danske strande. Kuml 1970: 297-329. 1978. Aggersund-bopladsen zoologisk belyst. Svanejagt som arsag til Matz, F. 1933. Eine neue 'Pintadera' aus Thessalien. Archdologischer Anbosaettelse? Kuml 1978: 57-76. zeiger 48: 315-7. Moodie, D.W. and Kaye, B. 1969. The northern limit of Indian agriculture Mauss, M. 1970. The Gift. London, Routledge and Kegan Paul. in North America. The Geographical Review 59(4): 513-29. Meiggs, R. and Lewis, D.M. 1969. A Selection of Greek Historical InscripMoore, J.A. 1981. The effects of information networks in hunter-gatherer tions. Oxford, Clarendon Press. society, in B. Winterhalder and E.A. Smith, eds., Hunter-Gatherer Meillassoux, C. 1983. The economic bases of demographic reproduction: Foraging Strategies. University of Chicago Press. from the DMP to wage-earning. Journal of Peasant Studies 11: 501983. The trouble with know-it-alls: information as a social and ecolo61. gical resource, in J.A. Moore and A.S. Keene, eds., Archaeological Meinander, C.F. 1954. Die Kiukais-Kultur. Finska Fornminnesforeningens Hammers and Theories. New York, Academic Press. Tidsskrift 53: 1-191. Morais Arnaud, J. in press. The mesolithic communities of the Sado ValMeldgaard, M. 1983. Resource fluctuation and human subsistence: a zooley (Portugal) in their ecological setting, in C. Bonsall, ed., The archaeological and ethnographical investigation of a west GreenMesolithic in Europe: Proceedings of the Third International Symland caribou hunting camp, in J. Clutton-Brock and C. Grigson, posium, Edinburgh 1985. University of Edinburgh, Department of eds., Animals in archaeology: 1. Hunters and their Prey. Oxford, Archaeology. British Archaeological Reports (International Series 163). 1986. The Greenland Caribou - Zoogeography, Taxonomy, and PopulaMorgan, W.B. and Munton, R.J.C. 1971. Agricultural Geography. New tion Dynamics. Meddelelser om Grenland, Bioscience 20. York, St. Martin's Press. Moritz, L.A. 1958. Grain, Mills and Flour in Classical Antiquity. Oxford Mele, A. 1979. // Commercio Greco Arcaico: Prexis edEmporie (Cahiers du University Press. centre Jean Berard No. 4). Naples. Moritz, R.E. 1977. On a possible sea breeze circulation near Barrow, AlasMellars, P. 1978. Excavation and economic analysis of mesolithic shell ka. Arctic and Alpine Research 9:427-31. middens on the island of Oronsay (Inner Hebrides), in P. Mellars, Morlan, R.E. 1975. Kutchin prehistory, as seen from the middle Porcupine ed., The Early Postglacial Settlement of Northern Europe. London, drainage, northern Yukon Territory, in A. McFayden Clark, ed., Duckworth. Proceedings: Northern Athapaskan Conference, 1971. II: National 1985. The ecological basis of social complexity in the upper palaeolithic Museum of Man Mercury Series, Canadian Ethnology Service, of southwestern France, in T.D. Price and J. Brown, eds., PrehisPaper 27: 669—758. toric Hunter-Gatherers. The Emergence of Cultural Complexity. Morris, I.M. 1986. Gift and commodity in Archaic Greece. Man 21: 1-17. New York, Academic Press. 1987. Burial and Ancient Society: the Rise of the Greek City-State. Miliarakis, A. 1886. Gheographia Politiki Nea ke Arkhea tou Nomou ArCambridge University Press. gholidhos ke Korinthias. Athens, Vivliopolion Estias. Morrison, D.A. 1983. Thule culture in western Coronation Gulf, N.W.T. Miller, P.C., Stoner, W.A. and Tiezen, L.L. 1976. A model of stand National Museum of Man Mercury Series, Archaeological Survey of photosynthesis for the wet tundra meadow at Barrow, Alaska. EcCanada, Paper 116. ology 57:411-30. Morrison, F.B. 1954. Feeds and Feeding (abridged). Ithaca New York, The Millett, P.C.M. 1984. Hesiod and his world. Proceedings of the Cambridge Morrison Publishing Co. Philological Society 30: 84-115. Movius, H. 1937. A neolithic site on the River Bann. Proceedings of the Milojcic, V. 1955. Vorbericht iiber die Ausgrabungen auf der OtzakiRoyal Irish Academy C 23: 17-^0. Magula 1954. Archdologischer Anzeiger: 157-82. 1960. Hauptergebnisse der Deutchen Ausgrabungen in Thessalien, 1953- Murdoch, J. 1892. Ethnological results of the Point Barrow expedition. Annual Report of the Bureau of American Ethnology 9(1). 58. Bonn, Rudolf Habelt. Murie, O.J. 1935. Alaska-Yukon caribou. North American Fauna 54. 1971. Die Lage und der Erhaltungszustand der Otzaki-Magula, GraWashington, U.S. Department of Agriculture, Bureau of Biological bungsmethoden, Stratigraphie und Bauten der Flache III, in V. Survey. Milojcic Die Deutchen Ausgrabungen auf der Otzaki-Magula in Thessalien, 1: das Fruhe Neolithikum. Bonn, Rudolf Habelt. Myer, J.P. and Pitelka, F.A. 1979. Variations in summer temperature patterns near Barrow, Alaska: analysis and ecological interpretation. 1972. Neue deutsche Ausgrabungen in Demetrias/Thessalien, 1967-72. Arctic and Alpine Research 2:131 - 4 4 . Jahrbuch der Heidelberger Akademie der Wissenschaften: 61-74. Namias, J. 1970. Climatic anomaly over the United States during the 1976a. Die Grabung auf der Ayia Sofia-Magula, in V. Milojcic, A. von \96Ws. Science 170: 741-3. den Driesch, K. Enderle, J. Milojcic-v. Zumbusch and K. Kilian Die Deutschen Ausgrabungen auf Magulen um Larisa in Thessalien, Nasatir, A.P. (ed.). 1952. Before Lewis and Clark: Documents Illustrating the History of the Missouri 1785-1804. St. Louis Historical Docu1966. Bonn, Rudolf Habelt. ments Foundation. 1976b. Die Baubefunde, in E. Hanschmann and V. Milojcic ArgissaNelson, E.W. 1899. The Eskimo about Bering Strait. Bureau of American Magula 3: die Fruhe und Beginnende Mit tiere Bronzezeit. Bonn, Ethnology, Annual Report 18 (1). Rudolf Habelt. Nelson, R.K. 1969. Hunters of the Northern Ice. University of Chicago Mine, L.D. 1986. Scarcity and survival: the role of oral tradition in mediatPress. ing subsistence crises. Journal of Anthropological Archaeology 5:

References

136

Netting, R.McC, Wilk, R.R. and Arnould, E.J., eds. 1984. Households. Comparative and Historical Studies of the Domestic Group. Berkeley and Los Angeles, University of California Press. Newell, R.R. 1984. The archaeological, human biological, and comparative contexts of a catastrophically-terminated Kataligaaq house at Utqiagvik, Alaska (BAR-2). Arctic Anthropology 21 (1): 5-52. Nichols, H. 1974. Arctic North American palaeoecology: the recent history of vegetation and climate deduced from pollen analysis, in J.D. Ives and R.G. Barry, eds., Arctic and Alpine Environments. London, Methuen. Nissen, H.J. 1970. Grabung in den Quadraten K/L XII in Uruk-Warka. Baghdader Mitteilungen 5: 102-91. Noonan, T.S. 1973. The grain trade of the northern Black Sea in antiquity. American Journal of Philology 94: 231-42. Noordegraaf, L. 1980. Levensstandaard en levensmiddelenpolitiek in Alkmaar vanaf het eind van de 16de tot in het begin van de 19de eeuw. Alkmaarse Historische Reeks 4: 55-100. Norman, D.W. 1977. Economic rationality of traditional Hausa dryland farmers in the north of Nigeria, in R.D. Stevens, ed., Tradition and Dynamics in Small-Farm Agriculture. Ames, Iowa State University Press. Norman, M.J.T., Pearson, C.J. and Searle, P.G.E. 1984. The Ecology of Tropical Food Crops. Cambridge University Press. North, D.C. 1977. Markets and other allocation systems in history: the challenge of Karl Polanyi. Journal of European Economic History 6; 703-19. NSSG 1971. Agricultural census of 14 March 1971. National Statistical Service of Greece (unpublished). Oberg, K. 1973. The Social Economy of the Tlingit. Seattle, University of Washington Press. Ogilvie, A.E.J. 1984. The past climate and sea-ice record from Iceland, Part I: Data to A.D. 1780. Climatic Change 6: 131-52. Okakok, K.L., ed., 1981. Puiguitkaat: The 1978 Elders' Conference (Commission on History and Culture). North Slope Borough. O'Shea, J. 1981. Coping with scarcity: exchange and social storage, in A. Sheridan and G.N. Bailey, eds., Economic Archaeology. Oxford, British Archaeological Reports (International Series 96). 1984. Mortuary Variability: An Archaeological Investigation. New York, Academic Press. Oshibkina, C.V. 1978. Neolit Vostochnogo Prionezhia. Moscow, Nauka. 1983. Mezolit Basseina Sukhony i Vostochnogo Prionezhia. Moscow, Nauka. Paaver, K.C. 1965. Formirovaniye Teriofaunyi i Izmenchivost Mlekopytayushchikh v Pribaltiki v golotsene (German summary: Die Entstehung der Saugetierfauna und der Variabilitat der Saugetiere des Ostbaltikums in Holozan). Tartu, Akademiya Nauk Estonskoi SSR. Page, D.L. 1955. Sappho and Alcaeus. Oxford, Clarendon Press. Palsi, S. 1920. Ein Steinzeitlicher Moorfund bei Korpilahti im Kirchspiel Antrea, Lan, Viborg. Suomen Muinaismuistoydhdistyksen AikakauskirjaJFinskaFornminnesforeningens Tidskrift 28: 1-19. Pansiot, F.P. and Rebour, H. 1961. Improvement in Olive Cultivation. Rome, FAO. Parks, D.R. 1979. Bands and villages of the Arikara and Pawnee. Nebraska History 60(2): 197-213. Patlagean, E. 1977. Pauvrete Economique et Pauvrete Sociale a Byzance, 4e-7e Siecles. Paris/La Haye, Mouton. Pavis d'Escurac, H. 1976. La Prefecture de VAnnone. Service Administratif Imperial d'Auguste a Constantin. Ecole Franchise de Rome. Payne, S. 1973. Kill-off patterns in sheep and goats: the mandibles from A§van Kale. Anatolian Studies 23: 281-303. 1985. Zoo-archaeology in Greece: a reader's guide, in N.C. Wilkie and W.D.E. Coulson, eds., Contributions to Aegean Archaeology: Studies in Honor of William A. McDonald. Minneapolis, Center for Ancient Studies, University of Minnesota.

Pearson, H.W. 1957. The economy has no surplus: critique of a theory of development, in K. Polanyi, C M . Arensberg and H.W. Pearson, eds., Trade and Market in the Early Empires: Economies in History and Theory. New York, Free Press. Pendlebury, J.D.S. 1937-8. Excavations in the Plain of Lasithi III: Karphi. Annual of the British School at Athens 38: 57-145. Pequart, M. and Pequart, S.J. 1937. Teviec. Station-necropole mesolithique du Morbihan (Archives de l'lnstitut de Paleontologie Humaine 18). Peristiany, J.G.,ed. 1965. Honour and Shame: the Values of Mediterranean Society. London, Weidenfeld and Nicolson. Perlman, S. 1980. An optimum diet model, coastal variability and huntergatherer behaviour, in M. Schiffer, ed., Advances in Archaeological Method and Theory 3. New York, Academic Press. Perrot, J., ed. 1979. Actes de la rencontre international de Suse (Iran) du 23 au 28 Octobre 1977. Paleorient 4: 133-244. Polanyi, K. 1977. The Livelihood of Man (ed. H.W. Pearson, from unpublished papers written 1947-53). New York, Academic Press. Popham, M.R., Touloupa, E. and Sackett, L.H. 1982. Further excavations in the Toumba cemetery at Lefkandi, 1981. Annual of the British School at Athens IT 213-48. Porter, S.C. and Denton, G.H. 1967. Chronology of neoglaciation in the North American Cordillera. American Journal of Science 265: 117— 210. Potter, C.W. 1979. The Changing Landscape of South Etruria. London, Elek Press. Price, T.D. and Brown, J.A., eds. 1985. Prehistoric Hunter-Gatherers. The Emergence of Cultural Complexity. New York, Academic Press. Qviller, B. 1981. The dynamics of the Homeric society. Symbolae Osloen5^56:109-55. Rappaport, R.A. 1968. Pigs for the Ancestors. New Haven, Yale University Press. 1979. Ecology, Meaning and Religion. Richmond, North Atlantic Books. Ray, P.H. 1885. Report of the International Polar Expedition to Point Barrow, Alaska. Washington, US Government Printing Office. Raynaut, C. 1971. Soumarana - Notes sur le Terroir et I'Economie. Centre d'Etude et de Recherche Ethnologues, Universite de Bordeaux. Rea, J.R., ed. 1972. The Oxyrhynchus Papyri {vol. XL). Graeco-Roman Memoirs no. 56. London, published for the British Academy by the Egypt Exploration Society. Redding, R.W. 1981. Decision Making in Subsistence Herding of Sheep and Goats in the Middle East. Ann Arbor, University Microfilms. Redman, C.L. 1978. The Rise of Civilization. San Francisco, Freeman. Reimers, E. 1982. Winter mortality and population trends of reindeer on Svalbard, Norway. Arctic and Alpine Research 14: 295-300. Reisman, P. 1977. Freedom in Fulani Social Life. University of Chicago Press. Renfrew, C. 1972. The Emergence of Civilisation: the Cyclades and the Aegean in the Third Millennium B.C. London, Methuen. 1982. Polity and power: interaction, intensification and exploitation, in C. Renfrew and M. Wagstaff, eds., An Island Polity: the Archaeology of Exploitation in Melos. Cambridge University Press. Renfrew, J. 1966. A report on recent finds of carbonised cereal grains and seeds from prehistoric Thessaly. Thessalika 5: 21-36. Renouf, M.A.P. 1981. Prehistoric coastal economy in Varangerfjord, North Norway. Ph.D. dissertation, University of Cambridge. Reynolds, P.J. 1974. Experimental iron age storage pits. Proceedings of the Prehistoric Society 40: 118-31. Rhoades, R.E. and Thompson, S.I. 1975. Adaptive strategies in Alpine environments: beyond ecological particularism. American Ethnologist 2 (3): 535-52. Richards, A.I. 1939. Land, Labour, and Diet in Northern Rhodesia: An Economic Study of the Bemba. Oxford University Press. Ricklefs, R.E. 1980. Ecology (2nd edn.) Don Mills, Nelson. Rickman, G. 1971. Roman Granaries and Store Buildings. Cambridge University Press.

References

137

1980. The Corn Supply of Ancient Rome. Oxford University Press. Ridley, C. and Wardle, K.A. 1979. Rescue excavations at Servia 1971— 1973: a preliminary report. Annual of the British School at Athens 74:185-230. Rimantene, R. 1979. Sventoji. Vilnus, Mosklas. Rogers, J.C. 1978. Meteorological factors affecting interannual variability of summertime ice extent in the Beaufort Sea. Monthly Weather Review 106: 890-97. Rohan-Csermak, G. de 1963. Sturgeon Hooks of Eurasia. Chicago, Aldine. Rondiri, V. 1981. Ceramic style and distribution in neolithic Greece. M. Phil, dissertation, University of Cambridge. Root, D. 1983. Information exchange and the spatial configuration of egalitarian societies, in J.A. Moore and A.S. Keene, eds., Archaeological Hammers and Theories. New York, Academic Press. Rosenlund, K. 1976. Catalogue of Subfossil Danish Fishes. Copenhagen, Zoological Museum. Roussel, D. 1976. Tribu et Cite. (Centre de recherches d'histoire ancienne 23). Paris. Rowland, R.J. Jr. 1976. The 'very poor' and the grain dole at Rome and Oxyrhynchus. Zeitschrift fur Papyrologie und Epigraphik 21: 69-72. Rowley-Conwy, P. 1981. Slash and burn in the temperate European Neolithic, in R.J. Mercer, ed., Farming Practice in British Prehistory. Edinburgh University Press. 1983. Sedentary hunters: the Ertebolle example, in G.N. Bailey, ed., Hunter-Gatherer Economy in Prehistory. Cambridge University Press. 1986. Between cave painters and crop planters: aspects of the temperate European mesolithic, in M. Zvelebil, ed., Hunters in Transition. Mesolithic Societies of Temperate Eurasia and their Transition to Farming. Cambridge University Press, in preparation a. Report on the animal bones from Norsminde. in preparation b. Report on the animal bones from Ringkloster. Saario, D.J. and Kessel, B. 1966. Human ecological investigations at Kivalina, in N.J. Wilimovsky and J.N. Wolfe, eds., Environment of the Cape Thompson Region, Alaska (U.S. Atomic Energy Commission). Washington, US Government Printing Office. Sagard, G. 1939. The Long Journey to the Country of the Hurons (ed. G.M. Wrong). Toronto, The Champlain Society. Sahlins, M. 1974. Stone Age Economics. London, Tavistock Publications. Sauvage, H.E. 1875. On fishing in the reindeer period, in E. Lartet and H. Christy, Reliquiae Aquitanicae; being contributions to the archaeology of Perigord and the adjoining provinces of southern France. London, Williams and Norgate. Saxe, A. A. 1970. Social dimensions of mortuary practices. Ph.D. thesis, University of Michigan. Schacht, R. 1975. Preliminary report on the excavations at Tepe Sharafabad, 1971. Journal of Field Archaeology 2: 307-29. Schalk, R.F. 1977. The structure of an anadromous fish resource, in L.R. Binford, ed., For Theory Building in Archaeology. New York, Academic Press. Schmandt-Bessarat, D. 1977. An archaic recording system and the origin of writing. Syro-Mesopotamian Studies 1 (2). Schneider, H. 1974. Wirtschaft undPolitik (Erlanger Studien, Bd 3). Erlangen, Palm and Enke. Schwarcz, H., Melbye, J., Katzenberg, M. and Knyf, M. 1985. Stable isotopes in human skeletons of southern Ontario: reconstructing palaeodiet. Journal of Archaeological Science 12: 187-205. Schytte Blix, A. 1979. Temperature regulation and climatic influence on mortality in newborn arctic mammals, in E. Reimers, E. Gaare, and S. Skjenneberg, eds., Proceedings of the Second International Reindeer/Caribou Symposium, September 1979, Roros, Norway. Trondheim, Direktoratet for Vilt og Ferskvannskisk. Scott, M.F. and Gormley, B. 1980. The animal of friendship: an indigenous model of Sahelian pastoral development in Niger, in D. Brokensha, D.M. Warren and O. Werner, Indigenous Knowledge Systems and Development. University Press of America.

Shay, J.M. and Shay, C.T. 1978. Modern vegetation and fossil plant remains, in G. Rapp and S.E. Aschenbrenner, eds., Excavations at Nichoria in South West Greece I. Minneapolis, University of Minnesota Press. Sheehan, G.W. 1985. Whaling as an organizing focus in northwestern Alaskan Eskimo Society, in T.D. Price and J.A. Brown, eds., Prehistoric Hunter-Gatherers: The Emergence of Cultural Complexity. New York, Academic Press. Shelmerdine, C.W. 1985. The Perfume Industry of Mycenaean Pylos (Studies in Mediterranean Archaeology pocket-book 34). Goteborg, Paul Astrom. Sherratt, A.G. 1981. Plough and pastoralism: aspects of the secondary products revolution, in I. Hodder, G. Isaac and N. Hammond, eds., Pattern of the Past: Studies in Honour of David Clarke. Cambridge University Press. Shinkwin, A.D. 1979. Dakah De'nin's village and the Dixthada site: a contribution to northern Athabaskan prehistory. National Museum of Man Mercury Series, Archaeological Survey of Canada, Paper 91. Simpson, R.N. 1875. The western Eskimo, in A Selection of Papers on Arctic Geography and Ethnology (The Arctic Expedition of 1875, Royal Geographic Society). London, John Murray. Sivignon, M. 1975. La Thessalie: Analyse Geographique d'une Province Grecque (Memoires et Documents 17). Institut des Etudes Rhodaniennes des Universites de Lyon. Skaarup, J. 1973. Hesselo-Solager. Jagdstationen der sudskandinavischen Trichterbecherkultur (Arkaeologiske Studier 1). Copenhagen, Akademisk Forlag. Skocpol, T. 1979. States and Social Revolutions. Cambridge University Press. Skoog, R.O. 1968. Ecology of the caribou (Rangifer tarandus grand) in Alaska. Ph.D. thesis, University of California. Slaughter, D.C. 1982. The Point Barrow type house: Analysis of archaeological examples from Siraagruk and other sites in northern Alaska. Anthropological Papers of the University of Alaska 20 (1-2): 141-58. Sloan, R.E. and Duncan, M.A. 1978. The zooarchaeology of Nichoria, in G. Rapp and S.E. Aschenbrenner, eds., Excavations at Nichoria in South West Greece I. Minneapolis, University of Minnesota Press. Smith, A. 1976. An Inquiry into the Nature and Causes of the Wealth of Nations (The 'Glasgow* edn.). Oxford University Press. Smith, K.P. 1986. Ritual and resource variability: mechanisms for the transmission and storage of information regarding low-frequency resource cycles in hunter-gatherer societies (paper presented at the 19th Chacmool Archaeological Conference, University of Calgary, November, 1986). Snodgrass, A.M. 1971. The Dark Age of Greece. Edinburgh University Press. 1980. Archaic Greece: The Age of Experiment. London, Dent. 1982. Central Greece and Thessaly, in Cambridge Ancient History 111:1 (2nd edn.). Cambridge University Press. 1983a. Heavy freight in Archaic Greece, in P.D.A. Garnsey, K. Hopkins and C.R. Whittaker, eds. Trade in the Ancient Economy. London, Chatto and Windus. 1983b. Two demographic notes, in R. Hagg, ed., The Greek Renaissance of the Eighth Century BC. Stockholm, Skrifter Utgvina i Svenska Institutet i Athen. Soffer, O. 1985. Patterns of intensification as seen from the Upper Palaeolithic of the Central Russian Plain, in T.D. Price and J.A. Brown, eds., Prehistoric Hunter-Gatherers. The Emergence of Cultural Complexity. New York, Academic Press. Solecki, R.S., Salwen, B. and Jacobson, J. 1973. Archaeological Reconnaissances North of the Brooks Range in Northwestern Alaska. Department of Archaeology, Occasional Papers 1. University of Calgary. Spencer, R.F. 1959. The North Alaskan Eskimo: a study in ecology and society. Bureau of American Ethnology, Bulletin 171.

References

138

Speth, J.D. 1983. Bison Kills and Bone Counts: Decision Making by Ancient Hunters. University of Chicago Press. Speth, J.D. and Spielmann, K.A. 1983. Energy source, protein metabolism and hunter-gatherer subsistence strategies. Journal of Anthropological Archaeology 2:1-31. Spielmann, K.A. 1982. Inter-societal food acquisition among egalitarian societies: an ecological study of Plains-Pueblo interaction in the American Southwest. Ph.D. dissertation, University of Michigan. 1986. Interdependence among egalitarian societies. Journal of Anthropological Archaeology 5: 279-312. Spiess, A.E. 1979. Reindeer and Caribou Hunters. New York, Academic Press. Srejovic, D. 1967. Lepenski Vir. Beograd, Narodni Muzej. Srejovic, D. and Letica, Z. 1978. Vlasac. A Mesolithic Settlement in the Iron Gates, Vol. 1: Archaeology. Beograd, Narodni Muzej. Stanford, D.J. 1976. The Walakpa site: its place in the Birnirk and Thule cultures. Smithsonian Contributions to Anthropology 20. Stefansson, V. 1919. The Stefansson-Anderson Arctic expedition of the American Museum: preliminary ethnological report. Anthropological Papers of the American Museum of Natural History 14(1). Stenning, D.J. 1957. Transhumance, migratory drift, migration: patterns of pastoral Fulani nomadism. Journal of the Royal Anthropological Institute %1: 57-87. 1959. Savannah Nomads. London, Oxford University Press. Strong, W.D. 1935. An introduction to Nebraska archeology. Smithsonian Miscellaneous Collections 93 (10). Stuart, R. 1953. On the Oregon Trail. Robert Stuart's Journey of Discovery (ed. K.A. Spaulding). Norman, University of Oklahoma Press. Sturdy, D.A. 1975. Some reindeer economies in prehistoric Europe, in E.S. Higgs, ed., Palaeoeconomy. Cambridge University Press. Suttles, W. 1968. Coping with abundance: subsistence on the northwest coast, in R.B. Lee and I. DeVore, eds., Man the Hunter. Chicago, Aldine. Swift, J.J. 1973. Disaster and a Sahelian nomad economy, in R. Dalby and D. Harrison Church, eds., Drought in Africa I. 1977. Sahelian pastoralists, underdevelopment, desertification and famine, Annual Review of Anthropology 6:457-78. Swift, J.J., ed., 1984. Pastoral Development in Central Niger: Report of the Niger Range and Livestock Project. Niamey, MDR/USAID. Tapper, R.L. 1979. Individuated grazing rights and social organization among the Shahsevan nomads of Azerbaijan, in Pastoral Production and Society. Cambridge University Press. Tarn, W.W. and Griffith, G.T. 1952. Hellenistic Civilization (3rd edn.). London, Arnold. Tempany, H. and Grist, D.H. 1958. An Introduction to Tropical Agriculture. London, Longmans. Tengstrom, E. 1974. Breadfor the People. Studies of the Corn-Supply of Rome during the Late Empire. Stockholm, Paul Astrom. Testart, A. 1982a. Les Chasseurs-Cueilleurs, ou, Vorigine des inegalites. Paris, Societe d'Etnographie. 1982b. The significance of food storage among hunter-gatherers: residence patterns, population densities and social inequalities. Current Anthropology 23: 523-37. Theokharis, D.R. 1959. Pirasos. Thessalika 2: 29-68. 1961. Anaskafai Iolkou. Praktika tis Arkhaioloyikis Etairias: 45—54. 1971. Sesklon (oikismos). Ergon: 21-7. 1973. Neolithic Greece. Athens, National Bank of Greece. 1976. Anaskafi Sesklou. Praktika tis Arkhaioloyikis Etairias A: 153-62. 1980. To neolithiko spiti. Anthropoloyika 1: 12-14. Thing, H. and Clausen, B. 1979. Summer mortality among caribou calves in Greenland, in E. Reimers, E. Gaare, and S. Skjenneberg, eds., Proceedings of the Second International ReindeerI Caribou Symposium, September 1979, Reros, Norway. Trondheim, Direktoratet for Vilt og Ferskvannskisk. Thompson, E.P. 1971. The moral economy of the English crowd in the eighteenth century. Past and Present 50: 76-136.

Thwaites, R.G., ed. 1896-1901. The Jesuit Relations and Allied Documents: Travel and Explorations of the Jesuit Missionaries in New France, 1610-1791 (73 vols.). Cleveland, The Burrows Brothers Inc. Tiezen, L.L. 1973. Photosynthesis and respiration in arctic tundra grasses: field light intensity and temperature responses. Arctic and Alpine Research 5:239-51. Tilly, L.A. 1971. The food riot as a form of political conflict in France. Journal of Interdisciplinary History 2: 3-59. 1983. Food entitlement, famine, and conflict, in R.I. Rotberg and Th. K. Rabb Hunger and History. The Impact of Changing Food Production and Consumption Patterns on Society. Cambridge University Press. Tod, M.N. 1948. A Selection of Greek Historical Inscriptions (vol. 2). Oxford University Press. Tooker, E. 1964. An ethnography of the Huron Indians, 1615-1649. Bureau of American Ethnology, Bulletin 190. Topelius, G. 1912. Ett fiskstangsel fran stenalderen. Suomen Muinaismuistoyhdistyksen AikakauskirjajFinska Fornminnesforeningens Tidskrift 26: 227-32. Torrence, R. 1983. Time budgeting and hunter-gatherer technology, in G.N. Bailey, ed., Hunter-Gatherer Economy in Prehistory. Cambridge University Press. Tracy, J.D. 1983. Habsburg grain policy and Amsterdam politics: the career of sheriff Willem Dirkszoon Baerdes, 1542-1566. SixteenthCentury Journal 14: 293-319. Trigger, B.G. 1969. The Huron, Farmers of the North. New York, Holt, Rinehart and Winston. 1976. The Children of Aataentsic (2 vols.). Montreal, Mcgill-Queen's University Press. Tringham, R. 1971. Hunters, Fishers and Farmers of Eastern Europe 60003000 BC. London, Hutchinson. 1973. The mesolithic of south-east Europe, in S.K. Koslowski, ed., The Mesolithic in Europe. Warsaw University Press. Trolle-Lassen T. 1984. A preliminary report on the archaeological and zoological evidence offish exploitation from a submerged site in mesolithic Denmark, in N. Desse-Berset, ed., 2nd Fish Osteoarchaeology Meeting (CNRS Notes et Monographies Techniques 16). Paris. Tsountas, Kh. 1908. Ai Proistorikai Akropolis Dhiminiou kai Sesklou. Athens Archaeological Society. Turner, E.G. 1975. Oxyrhynchus and Rome. Harvard Studies in Classical Philology 79:1-24. Vakevainen, L. 1978. Mysterious stone settings at the swelling site of Pyheensilta in Mynamaki. Suomen Museo 85: 81-6. Vankina, L.V. 1970. Torfyanikovaya Stoyanka Sarnate. Riga, Zinatne. Van Stone, J.W. 1955. Archaeological excavations at Kotzebue, Alaska. Anthropological Papers of the University of Alaska 3 (2): 75-155. Vayda, A.P. 1979. Expansion and warfare among swidden agriculturalists, in A.P. Vayda, ed., Environment and Cultural Behavior. Austin, University of Texas Press. Vayda, A.P., Leeds, A. and Smith, D. 1961. The subsistence use of pigs in Melanesia, in V.E. Garfield, ed., Proceedings of the Annual Meeting of the American Ethnological Society. Seattle, University of Washington Press. Vayda, A.P. and McCay, B.J. 1975. New directions in ecology and ecological anthropology. Annual Review of Anthropology 4: 293-306. Veyne, P. 1976. Le Pain et le Cirque. Sociologie Historique d'un Pluralisme Politique. Paris, Editions du Seuil. Vickery, K.F. 1936. Food in Early Greece (Illinois Studies in the Social Sciences 20, 3). Urbana, University of Illinois. Vuorela, I. and Aalto, M. 1982. Palaeobotanical investigations at a neolithic dwelling site in southern Finland with special reference to Trapa natans. Annales Botanici Fennici 19: 81-92. Wace, A.J.B. and Thompson, M.S. 1912. Prehistoric Thessaly. Cambridge University Press. Waddell, E. 1975. How the Enga cope with frost: responses to climatic per-

References

139

turbation in the Central Highlands of New Guinea. Human Ecology 3: 249-73. Warrick, G.A. 1984. Reconstructing Ontario Iroquoian village organization. National Museum of Man Mercury Series, Archaeological Survey of Canada. Paper 124: 1-180. Watanabe, H. 1972. The Ainu ecosystem: environment and group structure. American Ethnological Society, Monograph 54. Watt, K.E.F. 1973. Principles of Environmental Science. New York, McGraw-Hill. Watts, MJ. 1983. Silent Violence: Food, Famine and Peasantry in Northern Nigeria. Berkeley, University of California Press. Wedel, W.R. 1961. Prehistoric Man on the Great Plains. Norman, University of Oklahoma Press. 1986. Central Plains Prehistory. Lincoln, University of Nebraska Press. Wee, H. van der 1963. The Growth of the Antwerp Market and the European Economy (3 volumes) (Recueil de travaux d'histoire et de philologie, ser. IV, fax. 28, 29, 30). Louvain, Publications universitaires de Louvain. Welinder, S. 1976. The economy of the Pitted Ware culture in eastern Sweden. Meddeldndenfran Lunds Universitetets Historiska Museum ns 1:20-30. Wells, B. 1983. Asine II.4.2-3. The Protogeometric Period. Stockholm, Skrifter Utgvina i Svenska Institutet i Athen. Weltfish, G. 1965. The Lost Universe. New York, Basic Books. Wheat, J.B. 1972. The Olsen-Chubbuck Site: A Paleo-Indian Bison Kill. Society for American Archaeology, Memoir 26. White, C. 1984. Herd reconstitution: the role of credit among Wodaabe herders in central Niger, in Pastoral Development Network (September 1984). London, Office of Development Information. White, K.D. 1970. Roman Farming. London, Thames and Hudson. White, R.G. and Trudell, J. 1980. Habitat preference and forage consumption by reindeer and caribou near Atkasook, Alaska. Arctic and Alpine Research 12: 511-29. Whitelaw, T.M. 1983. People and space in hunter-gatherer camps: a generalising approach in ethnoarchaeology. Archaeological Review from Cambridge 2 (2): 48-66. Wiens, J.A. 1976. Population responses to patchy environments. Annual Review of Ecology and Systematics 7: 81 -120. Wiessner, P. 1977. Hxaro: a regional system of reciprocity for reducing risk among the !Kung San. Ph.D. thesis, University of Michigan. 1982. Risk, reciprocity and social influences on !Kung San economics, in E.R. Leacock and R.B. Lee, eds., Politics and History in Band Societies. Cambridge University Press. Wijnen, M. 1982. The Early Neolithic I Settlement at Sesklo: an Early Farming Community in Thessaly, Greece. Leiden University Press. Wijngaarden-Bakker, L.H. van 1985. The faunal remains, in P.C. Woodman, Excavations at Mount Sandel 1973-77. Belfast, Her Majesty's Stationery Office. Wilhelm, P. 1973. Travels in North America, 1822-1824 (ed. S. Lottinville). Norman, University of Oklahoma Press. Will, G.F. and Hyde, G.E. 1917. Corn Among the Indians of the Upper Missouri. Lincoln, University of Nebraska Press. Will, P.E. 1980. Bureaucratie et Famine en Chine au 18e Siecle. Paris/ LaHaye, Mouton. Williams, L.D. and Wigley, T.M.L. 1983. A comparison of evidence for Late Holocene summer temperature variations in the northern hemisphere. Quaternary Research 20: 286-307. Wilson, G.L. 1917. Agriculture of the Hidatsa Indians, an Indian interpretation. The University of Minnesota Studies in the Social Sciences 9. Wilson, I.R. 1978. Archaeological Investigations at the Atigun Site, Central Brooks Range, Alaska. National Museum of Man Mercury Series, Archaeological Survey of Canada, Paper 78.

Winge, H. 1903. Om jordfundne fugle fra Danmark. Videnskabelig Meddelelserfra den Naturhistoriske Forening i Kjebenhavn 1903: 61-109. Wishart, D.J. 1979. The dispossession of the Pawnee. Annals of the Association of American Geographers 69 (3): 382-401. Wobst, M. 1974. Boundary conditions for paleolithic social systems: a simulation approach. American Antiquity 39: 147-78. 1976. Locational relationships in paleolithic society. Journal of Human Evolution 5: 49-58. 1977. Stylistic behavior and information exchange, in C.E. Cleland, ed., For the Director: Research Essays in Honor of James B. Griffin. University of Michigan Museum of Anthropology, Anthropological Papers 61: 317-42. Wolf, E.R. 1966. Peasants. Englewood Cliffs, Prentice Hall. Wood, W.R. 1974. Northern Plains village cultures: internal stability and external relationships. Journal of Anthropological Research 30 (1): 1-16. Woodburn, J. 1980. Hunters and gatherers today and reconstruction of the past, in E. Gellner, ed., Soviet and Western Anthropology. London, Duckworth. Woodman, P.C. 1978. The Mesolithic in Ireland: Hunter-Gatherers in an Insular Environment. Oxford, British Archaeological Reports (British Series 58). 1985a. Excavations at Mount Sandel 1973-77. Belfast, Her Majesty's Stationery Office. 1985b. Mobility in the early mesolithic of northwestern Europe: an alternative explanation, in T.D. Price and J.A. Brown, eds., Prehistoric Hunter-Gatherers. The Emergence of Cultural Complexity. New York, Academic Press. Worrle, M. 1971. Agyptisches Getreide fur Ephesos. Chiron 1: 325-^0. Wright, H.T., ed. 1981. An early town on the Deh Luran Plain. Memoirs of the Museum of Anthropology, University of Michigan. Ann Arbor, Museum of Anthropology. 1984. Prestate political formations, in T.K. Earle, ed., The Evolution of Complex Societies: The Harry Hojier Lectures for 1982. Malibu, Undena Press. Wright, H.T. and Johnson, G.A. 1975. Population, exchange and early state formation in southwestern Iran. American Anthropologist 77: 267-87. 1986. Regional perspectives on southwest Iranian State Development. PaleorientX 1:25-30. Wright, H.T., Miller, N. and Redding, R. 1980. Time and process in an Uruk rural center, in L'Archeologie de lTraq: perspectives et limites de Tinterpretation anthropologiques des documents. Colloques Internationaux du CNRS 580:265-84. Yarnell, R.A. 1964. Aboriginal relationship between culture and plant life in the Upper Great Lakes Region. University of Michigan Museum of Anthropology, Anthropological Papers 23. Yiannopoulos, N. 1913. Zwei praehistorische Siegel. Mitteilungen des Deutschen Archaologischen Instituts, Athenische Abteilung 38: 2930. Yorga, B. 1980. Washout: A Western Thule Site on Herschel Island, Yukon Territory. National Museum of Man Mercury Series, Archaeological Survey of Canada, Paper 88. Zanetti, D. 1964. Problemi Alimentari di una Economia Preindustriale. Cereali a Pavia dal 1398 al 1700. Turin (Universita degli Studi di Torino. Facolta di economia e commercio. Studi dell' Istituto di storia economica), P. Boringhieri. Zeldin, Th. 1973. France 1848-1945.1: Ambition, Love and Politics. Oxford University Press. Zvelebil, M. 1981. From Forager to Farmer in the Boreal Zone. Oxford, British Archaeological Reports (International Series 115). 1986. Postglacial foraging in the forests of Europe. Scientific American 254(5): 104-15.

140

INDEX

aggression, 33 agriculture see also farming agricultural buffering strategies, 57-8, 68-80, 87-97 agriculturally dependent societies, 59 agropastoralism, 81,83 animal husbandry and, 57, 58, 59, 66, 67, 73 cash cropping, 67, 85, 88, 89, 118 crop disease, 61, 63, 72 crop storage, 58, 61 -2, 65, 75-6 crop yields, 5, 61,63, 72, 73,90, 98, 116 cycles in, 57, 58, 59 diversification in, 58, 72, 83, 90, 117 failure in, 58, 59, 61,63,64, 65-6, 72, 73-4, 75, 78, 79, 80,93,95,98, 101, 102, 104, 111, 112, 115, 117, 118, 120 fishing and, 65 in the Great Lakes Region, 63-6 hunter-gatherers and, 60, 65 intensification of, 77,99, 116 labour and, 72, 75, 77, 78,90, 99,109,116 mixed, 71, 72, 81, 117 in modern Greece, 87-97 in the New World, 59,66-7 Plains, 60-4 polycropping, 90, 91, 97 reliability of, 59

root crops, 79 sedentism and, 58, 81, 84 seed crops, 98 specialised, 60 spread of, 66, 67 storage, 58, 61-2, 65, 75-6 subsistence on, 57, 59 surplus and, 45, 117 systems, 57, 58, 59-60, 62, 65, 66-7, 88,116 taxation on, 88 territoriality and, 58, 59, 66, 99 use of wild resources and, 58-60, 66, 67, 72, 75, 78, 113 viticulture, 117 agropastoralism, 81,83 Ainu, 45,48 Algonquin Indians, 65 Allan, W., 70 alliance Eskimo, 8, 20, 31, 34, 36 inter-regional, 20, 34, 36, 38 marriage, 74 storage and, 56 Archaic Period in Greece, 99, 101, 102 in North America, 67 Ankara Indians, 62 armour, 31,33, 34 Athabaskan Indians, 31, 33, 34, 36 begging, 90, 95 Bemba,74 ' Binford, L., 48,49 bronze, 99

Bronze Age social change and, 5, 6,96 access to copper in, 126 social organisation and, 115 centralised production in, 77 stability of, 96-7 Greece, 69, 71, 77, 97 state level, 96-7 mass capture technology in, t e c h n o ^ y and, 8, 124 55-6 buffering strategies mixed crops in, 72 combining, 90, 123 redistributive economies in, 125 community level, 10 surplus in, 125 effects of social distance on, 123 buffering mechanisms, 1 -7, 123-6 efficiency of, 5, 59 combining, 38, 87, 90, 124 flexibility of, 65 compatibility of, 124 hierarchy of, 124 costs of, 95, 118, 120 household level, 10 cross-cultural nature of, 90 implementation of, 10 culture and, 5, 38,95, 96, 124 implications of demography and, 66, 124 empirical, 6, 78 effectiveness of, 118, 120-1 heuristic, 6, 78 efficiency of, 6, 58, 66, 87, 89, theoretical, 6, 78 90, 93 population size and, 4, 5 emergency, 90,95 productivity and, 20-1 evolutionary significance of, 90 scale of, 10, 18, 90, 123, 125 hazard response mechanisms, socio-cultural context of, 10, 38 90,95,96,97 spatial, 58 hierarchy of, 10, 64, 90, 95, 96, stability of, 5-6, 66 124 technology and, 4, 5, 38, 124 high level, 4, 5-6, 90,93,95, 96, undermining of, 85 99, 100, 101, 124, 125 Bug-Dniester Culture, 53, 59 household level, 73, 76, 77, 78 burial practices, 100, 102 ideology and, 8, 124 Bushmen, 4 low level, 4, 5-6, 90,93,95, 96, 100 I 2 4 ' caches, 31, 32, 34, 36,48, 50 predictability of, 2, 5 caribou scale of response, 4, 5-6, 7, 9, c i i m a t e a n d ? 12-13, 14, 15, 21 87-8, 90, 93,95, 96,99, 100, concentrations of, 11-12, 26 101, 105, 124, 125 herd size, 16, 26, 30, 32, 35

141

Index

hunting, 8, 12,29,31,32,34 inua, 19,20 migration, 11-12, 33, 34, 35 subsistence on, 11, 24, 27, 28, 32, 33, 35, 37 variability in, 12-13, 14, 15,17, 21,26,28,29,30,32,34,35, 36, 37,42,45,46-7 carrying capacity, 42 Central Plains Tradition, 63 centralised administration administrative seals, 107, 112 of agricultural production, 109 archaeological evidence of, 107, 108,109,112 crisis mitigation through, 103 of food supply, 104-5,114-22 of surplus, 119 variability in, 106, 112 Childe, V.G., 68, 80 civilisations collapse of, 99 development of, 68 Mycenaean, 99,105 climate caribou and, 12-13, 14,15,21 Little Ice Age, 17 fluctuations in, 17, 26,28, 32, 35,38,89,97,125 micro-climate, 4, 5,61,63 palaeoclimate, 11, 14-17,26, 28,29,30,31,32,33 rainfall variability. 81-2, 89,98, 102, 115 reconstruction of, 13-17 sea mammals, 13, 15,26, 28, 30, 32,35 tree-ring chronologies and, 11, 13-17,38,39 colonisation, 78, 81, 84, 99, 102, 103, 105 copper Arctic, 30 Bronze Age access to, 126 use in Neolithic Thessaly, 77 trade for, 31 coracle, 41 craft goods craftsmen, 77 distribution of, 76 elite control of, 106-7 exchange of, 62, 77, 79, 111 increased production of, 78, 112 manufacture of, 62, 76 Sharafabad, 111,112,113 as storage, 50, 123 crops, 5, 58, 61-2, 63, 65, 67, 72, 73, 75-6,79,85,88,89,90,91, 97,98, 116, 118 cultural ecology, 5 cultural process, 106, 113 culture buffering mechanisms and, 5, 38,95,96,124 environment and, 2, 6, 78 deer Ainu subsistence on, 45,48 hunting, 62, 64 red deer, 54 reindeer, 55 diachronic analysis, 1, 2

diversification agriculture and, 58, 72, 83, 90, 117 archaeological evidence of, 21 as a buffering mechanism, 3,4, 9,10,21,27,28,29,30,31, 32, 34, 35, 37,45,46,48, 50, 58-60,62,64, 66, 67, 72, 73, 75, 77, 78, 83, 85, 86,90, 117,123,124 through domestic animals, 58, 73, 78, 83, 86 mobility and, 29, 124 regional level, 77 through wild resources, 58-60, 66, 67, 72, 75, 78, 83 division of labour, 56 dog sleds, 21,27, 31, 35 domestic animals agriculture and, 57, 58, 59,66, 67,73 camels, 81,83 cattle, 81, 83,85,99,109 culling, 113 disease, 73, 82, 83 diversification and, 58, 73, 78, 83,86 exchange of, 81, 84 as gifts, 79 goats, 81,83,109 herding, 72, 77, 85 horses, 62, 77, 83 labour and, 81 loaning, 79, 84, 85, 86 loss of, 82, 83, 85-6,95 marketing, 82-5 ownership of, 81 population aggregation and, 124

reliance on, 58-9, 70, 75 rinderpest, 85 secondary products of fertiliser, 74, 84 meat, 99 milk, 58, 70, 73, 83, 84, 85, 99 traction, 58, 59, 70-1,77, 81 wool, 58 sheep, 58, 77, 83, 108,109 as storage, 58-9, 70, 75 surplus and, 58-9, 79 taxation on, 85 wealth in, 81 drought, 58, 61, 64, 72, 74, 75, 76, 78,80,81-6,88,102,110, 115,116 . ecological systems, 90 economic crises, 115-17 Egypt, 100, 103, 120, 121 elites burial practices of, 100 competition among, 105 concentration of power in, 118 control of food by, 103 control of raw materials by, 77, 106, 107 control of storage by, 106 control of surplus by, 68-9, 77-8,80,99,103,106 control of trade by, 101, 102 dependence on, 77, 78, 100, 101, 104,105, 118, 119-20

development of, 75, 78 emergence of, 68-9, 75, 78, 80 hunting by, 77 shortages and, 124, 125 structure of, 116 support of, 69, 77-8 environment culture and, 2, 6, 78 high latitude, 40,41,42, 44-5, 49,51,56,79,126 marginal, 57, 59, 66 Mediterranean, 98 mid-latitude, 79 mobility and, 123 normative approach to, 87, 88 in Thessaly, 70 time-stressed, 49 tropical, 79, 126 environmental determinism, 121 Ertebolle, 54, 55 Eskimo Alaskan, 8-39,49,50 alliances, 8, 20, 31, 34, 36 cyclical change, 125 feasts, 20, 37 Greenlandic, 46-7 hunting rituals, 18-20 inua, 18-19,20 Ipiutak Phase, 39 karigl 18,20,21,31,33-5 kinship, 18 Netsilik, 48 Nuataqmiut, 45-6 Nunamiut,8, 11,12, 17,20, 34-6, 37 oral tradition, 17, 18,34 Punuk Phase, 26,28, 29 Tareumiut, 8, 11,12, 17,20 territoriality, 20 Thule, 21, 24, 26, 27,28, 30, 39 trading partners, 10,20, 30 umealiq, 18, 36 whale hunting, 8,13, 21-4, 29, 30,31,32,38

euergetism, 99, 104, 105, 119-20 exchange see also trade archaeological evidence of, 21 as a buffering mechanism, 3,4, 5,9,10,11,20,21,26,28, 30,31,34,35,36,37,38,45, 58, 59, 60, 62, 64, 65, 73-76, 77,79,81,83,84,85,86, 102, 111,117-18,124,125 commercial, 85 complementary, 65 of craft goods, 77, 111 of domestic animals, 81, 84, 85 of exotic materials, 21, 34, 36, 74, 76,111 offood,62,65,73,75,79,81, 84,85 gift giving, 37, 62, 79, 100 household, 74 of information, 83, 86 intensification of, 30, 32, 38, 102 inter-community, 26, 28, 62, 73, 76,117-18 inter-regional, 26, 28, 31, 32, 35, 37,65, 74 intra-regional, 26 kinship and, 124 of labour, 74,83,84,85,86

long-distance, 74,102 medium-distance, 76 mobility and, 20, 124,125 social networks and, 10, 34, 60 spatial variability and, 5 specialised production for, 30, 31,37,65 storage and, 45, 58, 124, 125 surplus and, 45, 73 technology and, 124 transportation and, 59,60 Evolutionary Theory, 90 famine see also hunger, starvation, 10,17,18,20,64,65,80,83, 102,114-16,118-22 farming see also agriculture agropastoralism, 81,83 altering cultivation practices, 3 cash cropping, 67, 85, 88, 89, 118

cereal, 79, 80, 88,98,99 crop disease, 61,63,72 crop storage, 58, 61-2, 65, 75-6 crop yields, 5,61, 63,72, 73,90, 98,116 dispersed field, 4,61, 72,117 family farms, 88-9,91-4,97 fixed plot, 79 Hausa,81 hierarchy of mechanisms in, 124 high-latitude, 79 knowledge of variability and, 58 mixed, 71, 72,81,117 mid-latitude, 79 polycropping,90,9I,97 pulse, 70, 71, 78, 79, 80 root crops, 79 seed crops, 98 storage and, 4, 71 in Thessaly, 68-80 tropical, 79 Uruk Period, 106-13 viticulture, 117 feasting, 20, 37, 50, 73 fish anadromous, 31, 36, 51-3, 57, 59,65 archaeological evidence of, 51-3,64,109 catadromous, 51-3 mesolithic use of, 52-3 neolithic use of, 52 reliance on, 28, 30, 35 riverine, 20 salmon, 41-6,48, 51, 52 sturgeon,52-3, 64 upper paleolithic use of, 52 fishing, 27, 29, 31, 33 agriculture and, 65 Huron Indian, 64-5 riverine, 28 technology, 41 folktales see also myth, 17, 18, 34 food available of, 3, 58, 79 crisis, 99, 103-5, 115-17,120 emergency, 4, 59, 64, 66, 69, 73, 104 exchange networks and, 67, 124

142

Index

exchanging, 65,73, 75, 79 preparation, 62, 83,110 preservation, 49, 54, 55, 61,64, 123,124 processing, 49, 55 riots, 119,121 sharing, 18,73,79 shortages, 1,17,18,20,67,72, 83,90,96,115,125 spoilage, 48,79,93 supply,104,105,114-22 surplus of, 4, 6,125 trade of, 20 transportation of, 59,60 Formalist Approach, 115,121 Fulani,81-6 Functionaiism, 1 gift-giving see also exchange, 37, 62,79,100 grandfather response, 87 Greece, 68-80,87-97,98-105 agriculture in, 88-97,^9 ancient, 98-105 archaic, 99, 101,102 classical, 97,99,102,403,105, 118,119-20 Dark Age, 99-100,102,105 ethnosin,98,101 grandfather response in, 87 Hellenic, 98,99, 103, 104, 105, 118 Methanites, 88-97 modern, 87-97 neolithic, 68-80 prt-polis system, 99,100-1,103 polis government in, 96,98-105 state, 95-7,98-105 Thessaly, 68-80 Hausa,81,83 hazard-reduction mechanisms (HRMs),90,95,96,97 Hesiod,97,100-1,105,126 Hidatsa Indians, 61,62-3,66 historical analysis, 115 hunger see also famine, starvation, 65,98,103,105,118 hunter-gatherers agriculturalists and, 60,65 Alaskan, 8-39 European, 40 high-latitude, 40,45,49, 56, 123 and long-term variability, 45 non-storing, 79 semi-nomadic, 48 semi-sedentary, 48 storing, 79 surplus and, 45, 79 hunting bison, 60,61,64 camps, 34, 53 caribou, 8,12,29,31,32,33,34, 35,45 communal, 31,32, 34,35, 38, 46,62 large mammal, 55,62 sea-mammal, 8,13,21-4,29, 30,31,32,33,35,36,38,51, 53-4 whale, 8,13,21-4,29,30,31, 32, 35, 38, 53, 54

Huron Indians agriculture, 63-6 economy, 57,63-6 fishing, 64-5 hunting, 64,66 storage, 64 Hxaro, 4 ideology, 124 imperialism, 99, 103,104,119, 121 inheritance, 91,100 Ipiutak Phase, 39 Iran state development in, 106-13 Susiana Plain, 106 iron, 26, 28,29, 30 Iron Age, 101 Iroquois Indians, 64,65,66 irrigation, 72,75,77 jade, 30, 31,34 kinship Eskimo, 18 exchange and, 124 fictive, 65 marriage alliances, 74 mediating disputes through, 31 networks, 124 scarcity and, 124 sharing and, 73 social storage and, 30 Kwakiutl Indians, 51 land dowries, 91 fragmentation, 90-1,97 inheritance, 91 landlords, 103 monopolies of, 100 ownership, 76, 88,91, 103, 104, 116 taxation, 118 wealth in, 81 Lepenski Vir Culture, 53 limiting factors, 2-3, 5,65 Little Ice Age, 17 loans, 79, 84, 85,86, 103, 104 logistical mobility, 27, 28, 29, 30, 32,48,53,65,123 Magdalenian, 54 maize agriculture, 60-7 exchange, 62,65 storage, 61-2, 64,65 yields, 5, 61, 63 market economies consumers and, 115,117, 118, 122 controlling variability in, 118 development of, 86, 117 fluctuations in, 117,120-1 modern, 115, 118 pre-industrial, 115-22 producers and, 115, 117,118, 122 social distance and, 123 specialisation, 115-16 structure of, 120 surplus and, 125 urban,117

markets dependence on, 85 food, 112, 114 free, 114, 118-19, 120 grain, 104 international, 121 location of, 81,89 Wodaabe uses of, 84 world, 120 Marxism, 116 mass capture mass kills, 54, 64 organisational requirements of, 123 population aggregation and, 124

storage and, 55 technology, 40,49, 56 enclosures, 54 nets, 41,44, 50, 51, 55 snares, 49 traps, 49, 55 weirs, 49, 51 Mesolithic,40,51-6 fish use in, 52-3 plant use in, 53 sea mammal use in, 53-4 sedentism in, 54 sites, 52-4 terrestrial mammal use in, 54-5 Mesopotamia, 68, 106-13 Methanites, 88-97 Middle Ages, 52 migration animal, 11-12,33,34,35,41, 43-4 human, 10,26, 37,83,86, 102, 123 mobility archaeological evidence of, 21, 37 as a buffering mechanism, 3-4, 9, 10,20,21,24,26,27,28, 29,30,31-2,33,34,35,36, 37,38,41,45,46,47-8,53, 56, 58,65, 74, 78, 79, 82-3, 86,102, 123, 124, 125 diversification and, 48, 124 exchange and, 20, 124, 125 frequency of moves, 10 increased, 26, 28, 29, 31, 36, 37 inter-annual variability and, 45 inter-community, 35, 65 inter-regional, 26, 32, 35 irregular, 123 logistical, 27, 28, 29, 30, 32, 48, 53,65,123 long-distance, 10 migration, 10, 26, 37, 83, 86, 102, 123

migratory drift, 82, 83, 86 political factors in, 83 reduced,28, 30, 31-2, 37 relocation, 20, 26, 37,45,46, 65, 74 residential, 28, 30, 32, 123 seasonal, 26, 28, 33, 34, 37, 38, 41,45,82-3 storage and, 36,47, 56, 124 territoriality and, 3, 34, 35, 58 transhumance, 82-3, 85 myth see also folktales creation, 18

identifying variability in, 18,20, 124 as a response to variability, 3 negative reciprocity appropriation, 4, 120 coercion, 77-8,96 exploitation, 99 piracy, 99, 115, 124 raiding, 4, 10,21,61,66,74, 123 theft, 4, 10, 123 warfare, 21, 34, 83, 89,95,99, 101,102,115,116,120,125 Neolithic, 68-80 economy, 59, 67, 80 fish use in, 52 plant use in, 55 sea mammal use in, 54 in Thessaly, 70-5 Netherlands, 114-22 Netsilik Eskimo, 48 non-materialist approaches, 1-3 normative approaches, 87, 88 Northwest Coast Indians Kwakiutl, 51 limiting factors for, 65 surplus and, 79 Nunamiut Eskimo alliance, 20 caribou hunting, 12, 37 mobility, 34-6 oral tradition, 17 resource base, 11 obsidian, 74, 111 olives cultivation of, 97 cyclical yields, 58 olive oil, 88, 89,93, 103 rainfall and, 89 Omaha Indians, 62 Onion Portage, 24, 27 overproduction, 45, 60, 70, 73, 79, 88,89-90,91,93,95,96,97, 124,125 pastoralism, 81-6 agropastoralism, 81, 83 in Dark Age Greece, 99 failure in, 82, 83, 85-6, 95 herd size, 83-4, 85 nomadic, 81-6 over-grazing, 82, 84 social networks and, 84 surplus generation and, 79 transhumance, 82-3, 85 Tuareg, 81 Wodaabe, 81-6 Pawnee Indians agriculture, 60-3 bison hunting, 60, 61, 64 Calumet ceremony, 62 economy, 57, 60-3, 66 storage, 61-2 Pearson, H.W., 68-70 peasants ancient Greek, 100 economic strategies of, 116, 117 Methanites, 88-97 modern Greek, 88, 96 wealthy, 101

143

Index pest infestation, 3,4,58,61,63,64, 65,71,72,83,85,115 piracy, 99,115,124 Pitted Ware Culture, 54 plant use archaeological evidence of, 109 barley, 72,75,88,98,106,109 cereals, 70,71,78,88,98,99, 115 domestic, 58 emmer, 71 einkorn,71,72,77 legumes, 98 maize, 5,60-7 in the Mesolithic, 55 in modern Greece, 88 in the Neolithic, 55 nuts, 55, 59,77 olives, 58,88,89,97 pulse crops, 70,71,78,79,80 seeds, 55 specialised, 59 inThessaly,71-2 in the Upper Paleolithic, 55 water chestnuts, 55 weeding and, 72,77 wheat, 72,75,88,89,92^, 96, 106,109 wild, 58,59 wild rice, 59 polls, 98-105 constraints on, 99 emergence of, 98,99,100,101, 105 euergetism, 99,104,105, 119-20 government, 96,98-105 pre-po/is systems, 99,100-1, 103 serfdom and, 101,103 territoriality, 96 Ponca Indians, 62 population aggregation, 34,38,75, 76,83 archaeological evidence of, 102 decreases in, 28,32,38,99 dispersal, 83 distribution of, 58,65, 75 diversification and, 124 effects of buffering strategies on, 66,124 exchange and, 124 growth, 28,32,34,72, 75,98, 99,101-2,116,119 mobility, 56,65,124 Neo-Malthusian Model of, 116 overpopulation, 102-3 pressure, 101-2 production increases and, 116 storage and, 124 pre-industrial Europe, 114-22 prestige development of, 50-1,56 items, 56 production beyond subsistence, 69,70 centralised, 77 control of, 112 development of social complexity and, 69 domestic, 69 elite groups and, 69-70, 77,88 household, 93

imbalances in, 79 intensification of, 69,72,116 over production, 45,60, 70,73, 79,88,89-90,91,93,95,96, 97,124, 125 population growth and, 116 Renfrew on, 69 Sahlins on, 69-70,73 social complexity and, 69 specialisation, 77 for subsistence, 69,70 underproduction, 69,103 variability in, 72,106,117 Punuk Phase, 26,28,29 r-selected species, 41 raiding, 4, 10,21,61,66, 74,123 reciprocity, 4,60,65, 74, 79, 85, 100,104,118 redistributive economies, 125 Renfrew, A.C., 69 residential mobility, 28, 30, 32,123 resource abundance, 11, 27,29, 30, 37, 41,45 alternative, 10,27,46,57 competition for, 99 complementary, 14,66 density, 41,59 fall-back, 27,37,38 monitoring, 45 monopoly of access to, 100 predictability, 9-10,21,27, 37, 41 primary, 50,63 productivity, 9-10,11,20-1,49, 92-4 reliability, 20 secondary, 28, 30,37, 50 specialisation, 40,54, 59 stability, 9 revolution, 103,105 risk see also variability a definition of, 115 logic of risk avoidance, 116 Roman Empire, 105,114-22 Sahlins, M. anomie, 95 domestic mode of production, 73 on overproduction, 73 on population pressure, 102 provisioning society, 70 on surplus, 69-70 salmon Ainu subsistence and, 45,48 catching, 41 in the Mesolithic, 52 seasonality in, 41 in the Upper Paleolithic, 52 variability in, 42-5, 50 sea mammals climatic conditions and, 13, 15, 26,28, 30, 32, 35 hunting, 8,13, 21-4,29, 30, 31, 32,35,38,51,53-4 use in the Mesolithic, 53-4 seals, 11,24,26,27,28,29,30, 42,45,53-4 subsistence on, 10, 13,28,29, 31,32,34,35,36,37,45, 53-4

variability in, 13,15,21,26,38, 42 walrus, 13,24,27,29 whales, 8,10,12,13,15,17, 19, 20,21-4,26,27,28,29,30, 31,32, 34,35, 37,38, 39, 53, 54 sealings administrative, 107, 112 changes in, 112 in Thessaly, 74 Uruk Period, 107,112 seasonality archaeological evidence of, 106, 107-9 in high latitudes, 41, 79 in mid latitudes, 79 in tropical latitudes, 79 Second World War, 95,115 sedentism increases in, 52,56 in the Mesolithic, 54 resource specialisation and, 51 technology and, 49-50 serial specialists, 48 settlement aggregation of, 29,77 centres of, 106,107 in Dark Age Greece, 99 density of, 74,75,106 dispersed, 34,67 growth of, 77 hierarchies of, 77 increases in, 27 location of, 27 permanence, 40,48,49, 52, 54, 70,78 nucleation of, 77 scale of, 21,34,99 seasonal, 30,32, 34 storage and, 49 in Thessaly, 70, 74, 76 Sharafabad, 106-13 sharing see also exchange decline in, 77,78 food, 18,73,79 inter-community, 10,76 kinship and, 73 social distance and, 123 storage and, 62, 79,124 surplus, 75, 79 universality of, 124 shortage catastrophic, 66 fodder, 73 food, 1,17,18,20,67,72,83, 90,96,115,125 kinship and, 124 localised, 17,20,50 long-term, 20 oral tradition and, 17, 18 scale of, 17 seasonal, 32,35, 36 short-term, 17,35 winter, 45,65 sickles,99,109,110, 111 Sioux Indians, 62-3,66 slavery, 101 Smith, A., 114 social change access to food and, 125 catastrophic, 125 cyclical, 125

in the Dark Ages, 125 directional, 11,21,38 high-level buffering and, 5,6,96 long-term, 78,125 risk buffering and, 5,6, 38,78, 96,124 storage and, 56 transformational, 78,125 Wodaabe,81,84-6 social complexity development of, 68-9,75-80 production and, 69 storage and, 40 surplus and, 68-9 in Thessaly, 80 social distance, 123 social networks agriculture and, 67 Eskimo, 8, 31 exchange and, 10,34,60 expansion of, 10,20 hostility and, 31 intra-regional, 21 kin-based, 124 in non-stratified societies, 10 pastoralism and, 84 pooling resources and, 10,21, 58 regional, 126 and risk buffering, 20,60,124 structure of, 124 social storage, 10,20,21,26,28,30, 32,40,50,56,58,74-5,76, 78,79,117,123,125 social stratification access to food and, 122,125 archaeological evidence of, 100 development of, 40, 56, 68, 78, 125 measuring, 21 social storage and, 40 storage and, 40 surplus and, 68-9 Star Carr, 54 starvation see also famine, hunger, 45,46,49, 58,69, 87,95,96, 105,116 state in Ancient Greece, 98-105 colonisation, 78,81,84,99,102, 103,105 formation, 98,99, 100, 101, 105,106-13, 122 emergent, 69 food crises and, 99, 103, 115-17 government, 98-105 imperialism, 99,103, 104, 119, 121 inter-annual variability and, 109 in Iran,106-13 Iron Age, 101 modern Greek, 95-7 polis state formation, 98,99, 100,101, 105 pre-state government, 99, 100-1,103 state-directed laws, 101,103 state-level buffering mechanisms, 96-7 storage, 40-56 alliances and, 56

144

Index archaeological evidence of, 21, 37,40, 51 -6 as a buffering mechanism, 3,4, 6,9,10,18, 20, 21, 26, 27, 28,29, 30,31, 32,34, 35, 36, 37,40-56, 58-9,61-2,64, 65, 73, 74-6, 78, 79, 83, 85, 90,93,97,100,101,103, 104, 105,106,107,112,117, 118,119-20,123,124,125 caches,31, 36,48, 50 ' control of, 106, 112 costs of, 93, 118 crop, 58, 61-2, 65, 75-6 defence of, 49, 56 dispersed, 31, 32, 34, 36, 37,48, 50,61 distribution of stores, 18, 20, 30, 101, 103,104,105, 119-20 exchange and, 45, 58, 124,125 facilities, 49, 51, 52, 75-6, 78, 100,106,112,119 indirect, 73, 79, 123 information, 107 institutionalised, 107 of labour, 49 livestock as, 58-9, 73, 79, 83, 85,123 long-term, 73,90,123 maize, 61-2,64, 65 mass capture and, 55 mobility and, 36,47, 56, 124 physical, 4, 21, 26, 28, 30, 32, 34, 35, 37, 58, 75 population size and, 124 predicting the need for, 27, 29, 45, 56 sharing and, 62,79, 124 short-term, 123 social change and, 56 social complexity and, 40 social storage, 10, 20, 21, 26, 28, 30, 32,40, 50, 56, 58, 74-5, 76,78,79,117,123,125 spoilage and, 49, 78,93 of surplus, 6,45,73, 75, 78, 79, 100,119,124 technology and, 10, 51, 52, 58, 59,65, 75-6, 78, 100, 106, 112, 1-19 temporal variability in, 45-50, 78,79,117 territorially and, 34 transportation and, 47,48, 120 worst case, 10,45, 50, 51, 56, 79 style boundary maintenance and, 21, 33,34 changes in, 21 differences in, 33, 34 distribution of, 2t, 74 ownership marks, 31,39 Punuk Phase, 26, 28, 29 territoriality and, 21, 33, 34 Thule Eskimo, 30 uniformity in, 27, 28, 29, 31, 37 zones, 28, 30 Substantivist Approach, 70, 115 Sukkalmahhu Elamite Period, 107 surplus, 68-80 absolute, 68-9 administration of, 119

Allan on, 70 agriculture and, 45,117 bronze age, 125 centrally controlled, 77 Childe on, 68,80 development of social complexity and, 68-9 distribution of, 18,20, 30,101, 103, 104, .105, 119-20, 121 domestic animals and, 58-9, 79 elite control of, 68-9, 77-8, 80, 99,103,106 emergence of elites and, 68-9, 75, 78, 80 exchange and, 45, 73 food, 4, 6, 125 hunter-gatherers and, 45, 75, 79 labour and, 93, 101 lack of, 26, 59 mobilisation of, 69, 79 pastoralism and, 79 Pearson on, 68-9 Renfrew on, 69 Sahlins on, 69,95 seasonal, 30 sharing, 75, 79 storage of, 6,45, 73, 75, 78, 79, 100,119,124 Substantivist Approach to, 70 taxation and, 77-8 technology and, 49 trade of, 100,101, 103 uses of, 10, 104-5, 112 whaling and, 21,37 Susa, 106, 107, 112 synchronic analysis, 1 Tareumiut Eskimos, 8 alliance, 20 Oral tradition, 17 resource base, 11 whale hunting 12 taxation agricultural, 88 animal, 85 fluctuation in, 89 land, 118 populaire, 121 s u r p i u s a n d,77-8 technology agricultural intensification and, 99 buffering strategies and, 4, 5, 3g^ ] 24 changes in, 11, 20, 31, 59, 110 exchange and, 124 mass capture, 40,49-50, 51, 54, 55-6 processing, 65 salmon fishing, 41 sedentism and, 49-50 storage and, 10,49, 51, 52, 58, 65, 75-6, 78, 100, 106, 112, i \9 surplus and, 49 territoriality and, 21 transportation, 21, 27, 31, 124 whaling, 24, 27, 31 Tehuacan Valley, 67 territoriality agriculture and, 58, 59, 66, 99 annexation of, 102, 119

defence of, 21, 33, 34, 37, 38 development of, 30 disputes over, 103 Eskimo, 20 lack of, 27 mobility and, 3, 34, 35, 58 polis, 96 storagea nd, 34 style and, 21, 33, 34 technology and, 21 Tewa Indians, 60 textiles, 77,111 texts Hesiod,97, 100-1,105, 126 information on buffering in, 97, 98, 100-5,118-19,120,121, 126 theft, 4, 10, 123 Thessaly archaeological sites in, 68-80 crop yields in, 72, 73 development of social complexity in, 75-80 diversification in, 72-3, 77 elite groups in, 68-9, 75, 77-8, 80 exchange in,73-4 farming in, 68-80 field ownership in, 76 natural environment of, 70 neolithic settlements in, 70, 75 sealings in, 74 Thule Eskimos, 21 archaeological sites, 27, 28 assemblage at Onion Portage, 27 economy, 24, 39 migration, 26 styles, 30 Tonga, 70 trade see also exchange in Dark Age Greece, 99-100 elite control of, 77, 101, 102 exports, 118, 121 fairs, 34 of grain, 103, 115-16, 118, 121 imports, 102, 103, 118 inter-regional, 20, 31, 118, 122 inter-tribal, 62 itinerant traders, 84, 85 laws regarding, 104,118-19 maritime, 100, 101, 103, 121 networks, 30, 34, 36, 103,104 partners, 10, 20, 30, 65,66, 74, 75,123 routes, 115 trading posts, 103 social distance and, 123 surplus andr 100, 101, 103 wealth and, 100 transhumance, 82-3, 85 transportation costs of, 118 dog sled, 21, 27, 31, 35 exchange and, 59,60 of food, 101, 115 long-distance, 120 problems with 48,60 reliability of, 124 storage and, 47,48, 120 technology, 21, 27, 31, 124 water, 47, 101,103,120 tree-ring chronologies, 16

palaeoclimate and, 13-17 predicting variability with, 17 using, 11,14-17,38,39 tribal society, 100 tribute, 119 Tuareg, 81, 84, 85 uncertainty see variability Upper Palaeolithic, 40, 51 fish use, 52 Magdalenian, 54 plant use in, 55 reindeer use in, 55 urbanism Childe on, 68 emergence of cities, 118 food supply and, 114-22 markets and, 85 in pre-industrial Europe, 114, 122 Uruk Period, 106-13 ceramics, 108,110-11 sealings in, 107 sickles, 109,110, 111 variability annual, 40,41,56,57,65 archaeological evidence of, 46, 126 buffering, 1-7,123-6 catastrophic, 65 cultural perceptions of, 11, 58 cyclical, 44, 58,126 heterogeneity in, 123 in high-latitude environments, 40,41,42,44-5,49,51,56, 79,126 homogeneity in, 3 intensity of, 1,3, 5, 10 inter-annuaMO, 14, 17,21,27, 28,29,34,35,40,41-8,50, 51,56,57,58,59,64,72,75, 78,79,81,82,98, 100,103, 106, 109,112,117 large-scale, 87 local, 43,44,45, 50, 78 long-term, 8, 17, 20,28, 32, 35, 40,44,45,51,56,74,75,82, 84,86,124,126 measuring, 9, 13 in mid-latitude environments, 79 modellings, 17,126 monitoring, 4 predictability of, 1, 3,4, 5, 9, 57-8,89 in rainfall, 81-2, 89, 98,102, 115 regional, 50 scale of, 18,58,82,86,87-8, 125,126 seasonal, 28,40,41,44,47, 57, 58,59,65,71,78,79,81,82, 124, 126 short-term, 8, 13,20,21,28,32, 78,86,102 social change and, 38 socio-political, 89,95 spatial, 1,3,5,10,40,41,42,45, 58,82,117, 123 storage and, 40, 47, 65 structure of, 1-6, 9, 17, 57, 58, 115

145

Index temporal, 1,3,5,10,18,58,82, 117,123 in tropical environments, 79 viticulture, 117 Wales, 41-5, 50 warfare, 21,34,83, 89,95,99, 101, 102,115,116,120,125 wealth accumulation of, 56, 76,79,81, 125 archaeological evidence of, 76 in cattle, 81

in land, 81 trade and, 100 whales beached, 54 climatic conditions and, 13,14, 15,26,27,28,30,35,38 hunting, 8,13,21-4,29,30,31, 32,38,54 inua, 19, 20 mesolithic use of, 53-4 migration, 12 settlement location and, 27 subsistence on, 10,13,28,29,

32,34, 35,37,39, 53-4 surplus and, 21,37 variability in, 13,15,17,21 whaling technology, 24,27,31 wheat, 58,72, 75, 88,106,109 production, 91-4 rainfall and, 89 sale of, 96 storage of, 93 surpluses of, 91-4 variability in, 91-4 Wodaabe agropastorialism, 81,83

cyclical change and, 125 diversification, 83, 85 economic change, 84-6 kinnidal, 83 mobility, 82-3, 85 nomadic pastoralism, 81-6 storage, 83-4,85 transhumance, 82-3,85 wuro, 81 Woodland Period ceremonialism, 67 on the Plains, 60

ALSO IN THIS SERIES Hunters in Transition Mesolithic societies of temperate Eurasia and their transition to farming

Edited by Marek Zvelebil Hunters in Transition analyses one of the crucial events in human cultural evolution; the emergence of postglacial hunter-gatherer communities and the development of farming. Traditionally, the advantages of settled agriculture have been assumed, and the transition to farming has been viewed in terms of the simple dispersal of early farming communities northwards across Europe. The contributors to this volume, however, adopt a fresh, more subtle approach. From a hunter-gatherer perspective, settled farming implies both advantages and disadvantages, and certainly involves organizational disruption during the period of transition and far reaching social consequences for the existing way of life. The hunter-gatherer economy and farming shared in fact a common objective: a guaranteed food supply in a changing natural and social environment. Drawing extensively on research in eastern Europe and temperate Asia, the book argues persuasively for the essential unity of all postglacial adaptations whether leading to the dispersal of farming or to the retention and elaboration of existing hunter-gatherer strategies. Contributors: Taker u Akazawa, Paul Dolukhanov; Clive Gamble; Stefan Karol Kozlowski and Janusz Kozlowski; James Lewthwaite; Gerald Matyushin; Peter Rowley-Conwy; Slavomil Vend; Marek Zvelebil.

The Archaeology of Prehistoric Coastlines Edited by Geoff Bailey and John Parkington The Archaeology of Prehistoric Coastlines offers a conspectus of recent work on coastal archaeology, examining the various ways in which hunter-gatherers and farmers across the world exploited marine resources such as fish, shellfish and waterfowl in prehistory. Changes in sea-levels and the balance of marine ecosystems have altered coastal environments significantly over the last ten thousand years, and the contributors assess the impact of these changes on the nature of human settlement and subsistence. An overview of coastal archaeology as a developing discipline is followed by ten case studies from a wide variety of places including Scandinavia, Japan, Tasmania and New Zealand, Peru, South Africa and the United States. Contributors'. Takeru Akazawa; Atholl J. Anderson; Geoff Bailey; Sandra Bowdler; Bill Buchanan; Margaret Deith; Jon M. Erlandson; Robert A. Feldman; Michael A. Glassow; Tony Manhire; Michael E. Moseley; John Parkington; Cedric Poggenpoel; Priscilla Renouf; Tim Robey; Judith Sealy; Judith C. Shackleton; Larry R. Wilcoxon; David R. Yesner.

Centre and periphery in the Ancient World Edited by Michael Rowlands, Mogens Larsen and Kristian Kristiansen The collaborative volume is concerned with long-term social change. Envisaging individual societies as interlinked and interdependent parts of a global social system, the aim of the contributors is to determine the extent to which ancient societies were shaped over time by their incorporation in - or resistance to - the larger system. Their particular concern is the dependent relationship between technically and socially more developed societies with a strong state ideology at the centre and the simpler societies that functioned principally as sources of raw materials and manpower on the periphery of the system. The papers in the first part of the book are all concerned with political developments in the Ancient Near East and the notion of a regional system as a framework for analysis. Part 2 examines the problems of conceptualising local societies as discrete centres of development in the context of both the Near East and prehistoric Europe during the second millennium BC. Part 3 then presents a comprehensive analytical study of the Roman Empire as a single system showing how its component parts often relate to each other in uneven, even contradictory, ways. Contributors: Colin Haselgrove, Lotte Hedeager, Philip Kohl, Kristian Kristiansen, Mogens Larsen, Mario Liverani, Leon Marfoe, Roger Moorey, Daphne Nash, Michael Rowlands, Carlo Zaccagnini.

Quantifying diversity in archaeology Edited by Robert D. Leonard and George T. Jones One of the enduring aims of archaeological research has been to explain why human material culture is so diverse, both across the world and through history. Recognising that diversity exists is not, however, to explain it nor to measure it effectively. The aim of the contributors to Quantifying Diversity in Archaeology is therefore to examine what we mean by diversity, to review the methods of measurement and formulae we can apply, and assess the pitfalls that exist. Richness and evenness, the two main components of the diversity measures developed in the biological sciences, are considered, as are the value of diversity measures in the study of style, ecology, cultural geography, and faunal, lithic and spatial analysis. Subsequent papers consider critically why the archaeological remains of particular cultures vary so markedly between sites, localities and regions. Contributors: George T. Jones, Robert D. Leonard, Peter T. Bobrowsky, Bruce F. Ball, David Rindos, Keith W. Kintigh, Michael B. Schiffer, Jan F. Simek, Charlotte Beck, Donald K. Grayson, David Hurst Thomas; Nan A. Rothschild, F.E. Smiley, Catherine M. Camerson, Prudence M. Rice; Margaret W. Conkey.