Title: Man the Hunter
Subtitle: The First Intensive Survey of a Single, Crucial Stage of Human Development—Man's Once Universal Hunting Way of Life
Date: 2009
Notes: This needs a lot of work error correcting and formatting, but I’m just uploading it at this early stage so that it can be word searched, plus so that specific essays can be linked to and skim read.
ISBN: 9780202330327
Publisher: Transaction Publishers
r-b-richard-b-lee-irven-devore-man-the-hunter-1.jpg

    [Front Matter]

      Hunters of the World

      [Title Page]

      [Copyright]

      [Dedication]

      Preface

      Contents

      Symposium on Man the Hunter

  Part I: Introduction

    1. Problems in the Study of Hunters and Gatherers

      Defining Hunters

      The Representativeness of Our Sample of Hunters

      Ethnographic Reconstruction

      The Subsistence Base

      Hunting vs. Gathering

      Social and Territorial Organization

      The Patrilocal Band

      The Problem of Corporation

      Flexibility and the Resolution of Conflict by Fission

      Order and Chaos in Hunter-Gatherer Society

      Marriage Arrangements in Australia

      Demography and Population Ecology

      “Nomadic Style”: A Trial Formulation

    2. The Current Status of the World’s Hunting and Gathering Peoples

      Mounted Hunters

      Sedentary Fishermen

      Incipient Tillers

      Africa

      East Asia

      Oceania

      North America

      South America

  Part II: Ecology and Economics

    3. The “‘Hunting” ’ Economies of the Tropical Forest Zone of South America: An Attempt at Historical Perspective

    4. What Hunters Do for a Living, or, How to Make Out on Scarce Resources

      Bushman Subsistence

      Abundance and VarieD’ of Resources

        Diet Selectivity

        Range Size and Population Density

      Demographic Factors

      Leisure and Work

      Caloric Returns

      The Security of Bushman Life

      What Hunters Do for a Living: A Comparative Study

        The Relation of Latitude to Subsistence

        The Importance of Hunting

      Conclusions

      Appendix

    5. An Introduction to Hadza Ecology

      Environment

      Subsistence

      Ease of Subsistence

    6. Coping with Abundance: Subsistence on the Northwest Coast

      Habitat

      Food-Getting Methods

        Food-Storing Methods

      Values and Social Organization

      Conclusions

    7. Subsistence and Ecology of Northern Food Gatherers with Special Reference to the Ainu

      Introduction

      The Problem of “Semi-Nomadism”

        Residential Shifts and the Food-quest

        Residential Stabili(y, Ecological Zones, and Habitat

        Variability of the Residential Stability at the Level of the Individual

      Hunting as an Occupation of Males

        Individualistic or Noncommunal Hunting of Larger Mammals

        Dife rentiation of the Activiry Field according to Sex

        Tendency toward Age Division of Labor

    8. The Netsilik Eskimos: Adaptive Processes

      Technological Adaptation

      Ecological Adaptation

        Order and Predictability in the Annual Round

        Alternatives and Substitutions in Subsistence Strategy

        Flexibility of Group Structure

        Sharing

      Demographic Adaptation

      Discussion

    9. Discussions, Part II

      9a. The Central Eskimo: A Marginal Case?

      9b. Notes on the Original Affluent Society

      9c. Does Hunting Bring Happiness?

      9d. Hunting vs. Gathering as Factors in Subsistence

      9e. Measuring Resources and Subsistence Strategy

  Part III: Social and Territorial Organization

    10. Ownership and Use of Land aIYlOng the Australian Aborigines

      The Classical Description

      The Recent Picture

        Composition of Communities

        Exploitation of Resources

        Territorial Boundaries

      Ethology and Primitive Hunters

    11. Stability and Flexibility in Hadza Residential Groupings

      The Geographical Divisions of the Country of the Eastern Hadza

      Camp Size, Composition and Nomadic Movement

      Regularities in Camp Composition

        Married Couples

        Parents and Children

      Conclusion

    12. The Diversity of Eskimo Societies

      Survey of Eskimo Societies

      Three Central Eskimo Societies

      Relationship to Other Studies of Hunting Societies

    13. The Nature of Dogrib Socioterritorial Groups

      Conclusions and Theoretical Considerations

    14. The Birhor of India and Some Comments on Band Organization

    15. The Importance of Flux in Two Hunting Societies

      Introduction

      Mbuti

      IK

      Comparisons

        Environment

        Land Use and Group Size

        Subsistence Techniques

      The Importance of Flux

      Political Authority

      Conclusions

    16. Southeastern Australia: Level of Social Organization

      Limitations

      Powerful Chieftainships

      Social Stratification

      Complexity of Social Organization

      Kamilaroi and Wiradjuri

        Powerful Headmen

        Hereditary Chieftains

        Number of Kinds of Sodalities

        Matrilineal Descent

        Conclusion

      Summary

    17. Discussions, Part III

      17a. On Ethnographic Reconstruction

      17b. The Problem of Lineage Organiza1 Ion

      17c. Analysis of Group Composition

      17d. Social Determinants of Group Size

      17e. Resolving Conflicts by Fission

      17f. Territorial Boundaries

      17g. Predation and Warfare

      17h. Hunter Social Organization: Some Problems of Method

        The Importance of Longer Field Studies

        Definition of Social Units

        The Application of Role Theory

        The Study of Ritual

      17i. Typology and Reconstruction: A Shoshoni Example

  Part IV: Marriage and Models in Australia

    18. Gidjingali Marriage Arrangements

      The Model

        Section 1: Kinship groups and terminology

        Section 2: Marriage rules

      The Facts

      Models, Facts, and Feeling Cheated

      Conclusion: Communication or Competition?

    19. “Marriage Classes” and Demography in Central Australia

      Appendix

    20. Demographic and Ecological Influences on Aboriginal Australian Marriage Sections

      Method

      Environment and Subsistence

      Tribes and Sections

      The Population Parameter

      Section Systems and the Population Parameter

      Australia. Upper: After Davidson; Lower: After Radcliffe-Brown.

      Discussion

      Conclusions

    21. Australian Marriage, Land-Owning Groups, and Initiations

      Introduction

      Marriage and the Land-Owning Group

        Sahlins’ Original Aifiuent Society,’ A Comment

      Marriage Gerontocracy in Australia

      Interpretation of Marriage Gerontocracy

      Initiation

      Conclusion

    22. Discussions, Part IV

      22a. Gerontocracy and Polygyny

      22b. Gidjingali Marriage Arrangements: Comments and Rejoinder

      22c. The Use and Misuse of Models

      22d. Statistics of Kin Marriage: A Non-Australian Example

  Part V: Demography and Population Ecology

    23. Epidemiological Factors: Health and Disease in Hunter-Gatherers

      Introduction

      Disease in the Population Equation

      Mortality and Disease in Hunter-Gatherers

      Conclusion

    24. Some Predictions for the Pleistocene Based on Equilibrium Systems among Recent Hunter-Gatherers

      Introduction

      The Density Equilibrium System

        The System

        The Evidence

        Predictions for the Pleistocene

        Areas of Needed Research

      Communication Eq.uilibrium System

        The System

        The Evidence

        Predictions for the Pleistocene

        Research Needed

      Local Group Eqiltillbriltm Systems

      The System

        The Nature of the Evidence

        Predictions for Pleistocene Hunters

        Required Research Jor the Future

      Equilibrium in Family Size

        The System

        The Evidence

        Predictions for Pleistocene Populations

        Research Needed

      Equilibrium in the Size of the Effective Breeding Population

        The System

      The Evidence

        Predictions for Pleistocene Populations

        Research Needed

      Conclusions

    25. Discussions, Part V

      25a. The Demography of Hunters: An Eskimo Example

      25b. Population Control Factors: Infanticide, Disease, Nutrition, and Food Supply

      25c. The Magic Numbers “25” and “500”: Determinants of Group Size in Modern and Pleistocene Hunters

      25d. Pleistocene Family Planning

  Part VI: Prehistoric Hunter-Gatherers

    26. Traces of Pleistocene Hunters; An East African Example

      The Occurrence of Archeological Evidence

      Location and Character of Occupation Sites

      Size of Occupation Sites

      Food Remains

      Species Represented

      Hunting Methods

    27. A Theoretical Framework for Interpreting Archeological Materials

      The Use of Analogy

      A Model Minimizing Analogy

      The Nature of Culture

      The Nature of Social Units

      A General Illustration of the Use of the model

      Conclusions

    28. Methodological Considerations of the Archeological Use of Ethnographic Data

    29. Ethnographic Data and Understanding the Pleistocene

    30. Studies of Hunter-Gatherers as an Aid to the Interpretation of Prehistoric Societies

      Prehistoric Settlement Sites and Survival Patterns

      Technology

      Historical Sources

    31. Discussions, Part VI

      31a. Hunters in Archeological Perspective

      The Subsistence Base

      The Social Unit

      Population Reconstruction

      31b. Archeological Visibility of Food-Gatherers

      31c. The Use of Ethnography in Reconstructing the Past

  Part VII: Hunting and Human Evolution

    32. The Evolution of Hunting

      The World View of the Hunter

      The Social Organization of Human Hunting

      Conclusion

    33. Hunting: An Integrating Biobehavior System and Its Evolutionary Importance

      Introduction

      Hunting as an Integrating

      Sequence Behavior Pattern

      Programming Children

        Scanning

        Stalking

        Immobilization, Killing, and Capture

        Retrieval

      The Physical Superiorities of Man

      Simplicity of Basic Technology

      Hunter’s Sophisticated Knowledge of Behavior and Anatomy

      Genetic Mechanisms in Hunting Societies

      Evolutionary Aspects

      Conclusions and Summary

    34. Causal Factors and Processes in the Evolution of Pre-farming Societies

      Methodological Trends

      Factor Determination

        Biological Factors

        Cultural Factors

        Environmental Factors

        Cultural-Historical Factors

      Evolutionary Factors and Processes

        Non-Cultural Adaptations

        Effects oj Cultural Factors

    35. Discussions, Part VII

      35a. Are the Hunter-Gatherers a Cultural Type?

      35b. Primate Behavior and the Evolution of Aggression

      35c. Future Agenda

  Part VIII: The Concept of Primitiveness

    36. The Concept of Primitiveness

  [Back Matter]

    References

    Index

    [Back Cover]

  Archivists Appendix: Digitizations of Tables

    Part II: Ecology and Economics

      Table 1. Numbers and Distribution of Resident Bushmen and Bantu by Waterhole

[Front Matter]

Hunters of the World

r-b-richard-b-lee-irven-devore-man-the-hunter-2.png
10,000 B.C.
World Population: 10 million
Per Cent Hunters: 100
r-b-richard-b-lee-irven-devore-man-the-hunter-3.png
1500 A.D.
World Population: 350 million
Per Cent Hunters: 1.0
r-b-richard-b-lee-irven-devore-man-the-hunter-4.png
1972 A.D,
World Population: 3 billion
Per Cent Hunters: 0.001

[Title Page]

MAN THE HUNTER

Edited by
Richard B. Lee
and
Irven DeVore
with the assistance of
Jill Nash-Mitchell


ALDINE DE GRUYTER / New York

[Copyright]

Copyright © 1968 by the Wenner-Gren Foundation for Anthropological Research, Inc.

All rights reserved. No part of this publication may he reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or ally information storage and retrieval system, without permission in writing from the publisher.

ALDINE DE GRUYTER
A Division of Walter de Gruyter, Inc.
200 Saw Mill River Road
Hawthorne, N.Y. 10532

Eleventh printing, 1987

ISBN 0-202-33032-X

Library of Congress Catalog Card Number 67–17603
Designed by Chestnut House
Printed in the United States of America

[Dedication]

To Claude Levi-Strauss

Preface

In November, 1965, Sol Tax asked us to organize a symposium on current research among the hunting and gathering peoples of the world. This meeting was conceived to follow logically from an earlier symposium on the Origin of Man held at the University of Chicago in April, 1965. There were a number of reasons why we felt a conference devoted exclusively to the hunting way of life would be appropriate at this time. Current ethnographic studies have contributed substantial amounts of new data on hunter-gatherers and are rapidly changing our concept of Man the Hunter. Social anthropologists generally have been reappraising the basic concepts of descent, filiation, residence, and group structure. In archeology the recent excavation of early living floors has led to a renewed interest in and reliance on hunter-gatherer data for reconstruction, and current theories of society and social evolution must inevitably take into account these new data on the hunter-gatherer groups. Finally, hunting and gathering, as a way of life, is rapidly disappearing and it was hoped that the symposium would serve to stimulate further research while there were still viable groups to study.

It was clear, however, that the first order of business was to present the new data on hunters and to clarify a series of conceptual issues among social anthropologists as a necessary background to broader discussions with archeologists, biologists, and students of human evolution. A preliminary canvass was made in January, 1966, of seventeen younger anthropologists who had recently completed field work among hunting peoples, and their enthusiastic response led us to contact an additional fifty scientists in social anthropology, human biology, archeology, demography, and ecology, particularly those who in their writings had shown an interest in hunter-gatherer data and theory. The positive response of this larger group insured the success of the symposium, and the Wenner-Gren Foundation for Anthropological Research generously agreed to sponsor the meeting and to cover traveling and living expenses for the participants.

The symposium on Man the Hunter was held at the Center for Continuing Education at the University of Chicago, April 6–9, 1966. In attendance were some 75 scholars from as far afield as India, Japan, Australia, Uganda, Kenya, and the German Democratic Republic, as well as from centers in France, England, Canada, Latin America, and the United States. In an intensive four-day meeting, those present heard 28 background papers and twelve formal discussions, and participated in over ten hours of open debate. Since most of the papers had been distributed in advance, it was possible to limit formal presentations to ten minutes each in order to leave time for discussion. Our impression is that most participants found this intensive format exhausting but rewarding.

With three exceptions, all of the original papers presented at the symposium are published here. The majority have been substantially revised in the light of our discussions in Chicago. In addition, two new papers, one by Yengoyan and the other by Washburn and Lancaster, are included in the book although they were not presented at the meeting.

The format of this book follows closely that of the symposium. The introductory section contains a general statement by the editors of the problems and issues to be discussed and an inventory by Professor G. P. Murdock of the current status of the world’s hunting and gathering peoples. Each of the remaining sections corresponds to one of the topical sessions of the symposium, and each section contains the relevant papers as well as the edited transcript of the symposium discussion. As editors we are aware that publishing verbatim material from symposia docs not always work out successfully; we hope that our attempt at editing has eliminated redundancy and inconsequential material while still conveying some of the excitement and the confrontation of different viewpoints that characterized the symposium itself.

On behalf of all the participants we want to thank the Wenner-Gren Foundation for Anthropological Research for providing the financial support that made possible the symposium and the preparation of this volume. Mrs. Lita Osmundsen and the excellent staff of the Foundation in New York solved the complex problems of travel, visas, and financing with their usual efficient dispatch, enabling all of us to converge miraculously on Chicago on the appointed day.

Sol Tax conceived the idea of a symposium on Man the Hunter and brought it to fruition. As chairman of the meetings he brought us together on time and skillfully kept speakers within their time limits; on behalf of a group of anthropologists who had a great deal to say, we offer Sol our affectionate resentment. The staff of Current Anthropology worked to distribute advance papers, to compile accurate transcripts and address lists, and generally to keep the symposium running smoothly.

Our thanks go as well to Lee Pravatiner and the staff of the Center for Continuing Education. With accommodations, food service, press, duplicating facilities, and meeting rooms all under one roof, the Center was an ideal place for a symposium of this size and complexity. Few of the participants realized the contribution that Mr. Pravatiner made to the success of the meeting; he attended to countless details of seating arrangements, reservations, messages, transcription, and mimeographing, and did so tactfully and unobtrusively. Personally we were struck by the contrast between the complex social and technological system that enabled us to meet and the much simpler systems of the hunting and gathering peoples we were discussing.

In addition we should mention the pleasant evening that many of us spent at the American Indian Center in north Chicago. Our thanks go to members of the Center and to its director Robert Rietz, for their hospitality.

The transcription and editing of the twenty hours of tapes from the sessions provcd to be a formidable job; we are particularly grateful to Paul L. DeVore of the University of Chicago for the 180 pages of transcription and to Jill Nash of Harvard University for her aid in editing the transcript and for many other editorial services. Miss Nash also r8ad through the entire manuscript and made many helpful suggestions which have been incorporated in the final version. Dr. David A. Horr of Harvard undertook the compilation of the references and handled many editorial details in the latter half of the book.

The drawings were executed by Judith Tempkin and Jane Britton; the former also assisted in the final preparation of the bibliography. lv1rs. Marcia Harding Anderson, Mrs. Elizabeth Burnham, and Mrs. Linda Moseley, all of the staff of the Department of Social Relations of Harvard University, typed portions of the manuscript; the final version was produced by Mrs. Nellie Miller.

Richard Lee offers a special vote of thanks to the students in Social Relations 251 and 252, graduate seminars at Harvard in the spring semesters of 1966 and 1967, respectively. The members of these seminars read most of the preliminary manuscripts and their lively discussions did much to clarify the core issues presented in the book.

Both of us wish to thank our colleagues in the Department of Social Relations at Harvard for providing a congenial intellectual atmosphere and for making a number of useful suggestions on specific portions of the manuscript; we appreciate particularly the help of John W. M. Whiting, A. Kimball Romney, Evon Z. Vogt and David H. P. Maybury-Lewis. We also want to thank our publisher, Alexander J. Morin, for his patience and forbearance while we labored to complete the manuscript before our departure for further field work among the Bushmen.

Finally we acknowledge our major debt to our contributors, many of whom are huntergatherers in spirit as well as in professional interest. No two anthropological field workers share exactly the same motivations and worldview, except perhaps a common interest in understanding the human condition. We cannot avoid the suspicion that many of us were led to live and work among the hunters because of a feeling that the human condition was likely to be more clearly drawn here than among other kinds of societies.

RICHARD B. LEE
IRVEN DEVORE

Contents

Part I: Introduction

  1. Problem in the Study of Hunters and Gatherers, Richard B. Lee and Irven DeVore

2. The Current Status of the World’s Hunting and Gathering Peoples, George Peter Murdock

Part II: Ecology and Economics

3. The “Hunting” Economies of the Tropical Forest Zone of South America: An Attempt at Historical Perspective, Donald W. Lathrap

4. What Hunters Do for a Living, or, How To Make Out on Scarce Resources, Richard B. Lee

5. An Introduction to Hadza Ecology, James Woodburn

6. Coping with Abundance: Subsistence on the Northwest Coast, Wayne Suttles

7. Subsistence and Ecology of Northern Food Gatherers with Special Reference to the Ainu, Hitoshi Watanabe

8. The Netsilik Eskimos: Adaptive Processes, Asen Balikei

9. Discussions, Part II

a. The Central Eskimo: A Marginal Case?

b. Notes on the Original Affluent Society

c. Does Hunting Bring Happiness?

d. Hunting vs. Gathering as Factors in Subsistence

e. Measuring Resources and Subsistence Strategy

Part III: Social and Territorial Organization

10. Ownership and Use of Land among the Australian Aborigines, L. R. Hiatt

11. Stability and Flexibility in Hadza Residential Groupings, James Woodburn

12. The Diversity of Eskimo Societies, David DaT(IQSs

13. The Nature of Dcgrib Socioterritorial Groups, June Helm

14. The Birhor of India and Some Comments on Band Organization, B. J. Williams

15. The Importance of Flux in Two Hunting Societies, Colin M. Turnbull

16. Southeastern Australia: Level of Social Organization, Arnold R. Pilling

17. Discussions, Part III

a. On Ethnographic Reconstruction

b. The Problem of Lineage Organization

c. Analysis of Group Composition

d. Social Determinants of Group Size

e. Resolving Conflicts by Fission

f. Territorial Boundaries

g. Predation and Warfare

h. Hunter Social Organization: Some Problems of Method

i. Typology and Reconstruction: A Shoshoni Example

Part IV: Marriage and Models in Australia

18. Gidjingali Marriage Arrangements, L. R. Hiatt

19. “Marriage Classes” and Demography in Central Australia, M. J. Meggitt

20. Demographic and Ecological Influences on Aboriginal Australian Marriage Sections, Aram A. rengoyan

21. Australian Marriage, Land-Owning Groups, and Initiations, Frederick C. C. Rose

22. Discussions, Part IV

a. Gerontocracy and Polygyny

b. Gidjingali Marriage Arrangements: Comments and Rejoinder

c. The Use and Misuse of Models

d. The Statistics of Kin Marriage: A Non-Australian Example

Part V: Demography and Population Ecology

23. Epidemiological Factors: Health and Disease in Hunter-Gatherers, Frederick L. Dunn

24. Some Predictions for the Pleistocene Based on Equilibrium Systems among Recent Hunter-Gatherers, Joseph B. Birdsell

25. Discussions, Part V

a. The Demography of Hunters: Eskimo Example

b. Population Control Factors: Infanticide, Disease, Nutrition, and Food Supply

c. The Magic Numbers “25” and “500”: Determinants of Group Size in Modern and Pleistocene Hunters

d. Pleistocene Family Planning

Part VI: Prehistoric Hunter-Gatherers

26. Traces of Pleistocene Hunters: An East African Example, Glynn L. Isaac

27. A Theoretical Framework for Interpreting Archeological Materials, L. G. Freeman, Jr.

28. Methodological Considerations of the Archeological Use of Ethnographic Data, Lewis R. Binford

29. Ethnographic Data and Understanding the Pleistocene, Sally R. Binford

30. Studies of Hunter-Gatherers as an Aid to the Interpretation of Prehistoric Societies, J. Desmond Clark

31. I. Discussions, Part VI

a. Hunters in Archeological Perspective

b. The Archeological Visibility of Food Gatherers

c. The Use of Ethnography in Reconstructing the Past

Part VII: Hunting and Human Evolution

32. The Evolution of Hunting, Sherwood L. Washburn and C. S. Lancaster

33. Hunting: An Integrating Biobehavior System and Its Evolutionary Importance, William S. Laughlin

34. Causal Factors and Processes in the Evolution of Pre-farming Societies, Julian H. Steward

35. Discussions, Part VII

a. Are the Hunter-Gatherers a Cultural Type?

b. Primate Behavior and the Evolution of Aggression

c. Future Agenda

Part VIII: The Concept of Primitiveness

36. The Concept of Primitiveness, Claude Levi-Strauss

References

Index

Symposium on Man the Hunter

WENNER-GREN FOUNDATION FOR ANTHROPOLOGICAL RESEARCH

PARTICIPANTS

Sol Tax, Chairman
University of Chicago

Alexander A. Alland
Columbia University

James N. Anderson
University of California, Berkeley

Asen Balikci
Universite de Montreal

Marco Bicchieri
Central Washington State College

Lewis R. Binford
University of New Mexico

Sally R. Binford
Los Angeles, California

Joseph B. Birdsell
University of California, Los Angeles

Jean Briggs
Memorial University of Newfoundland

Ralph Bulmer
University of Auckland

Napoleon A. Chagnon
Pennsylvania State University

J. Desmond Clark
University of California, Berkeley

Georges Condominas
Ecole Pratique des Hautes Etudes, Paris

William Crocker
Smithsonian Institution

Ralph Crocombe
Australian National University

David Damas
McMaster University

James Deetz
Brown University

Edward Deevey, Jr.
Dalhousie University

P. E. de Josselin de Jong
Leiden University

Irven DeVore
Harvard University

A. Jorge Dias
Lisbon

Frederick L. Dunn
Hooper Foundation
University of California, San Francisco

Fred Eggan
University of Chicago

Leslie G. Freeman, Jr.
University of Chicago

Peter M. Gardner
University of Texas

David M. Hamburg
Stanford University

June Helm
University of Iowa

L. R. Hiatt
University of Sydney

F. Clark Howell
University of California, Berkeley

Roger C. Owen
City University of New York

Angel Palerm
Pan American Union, Mexico

Arnold R. Pilling
Wayne State University

Merrick Posnansky
University of Ghana

Edward S. Rogers
Royal Ontario Museum, Toronto

Frederick G. G. Rose
Institut fur Volkerkunde, Berlin

James R. Sackett
University of California, Los Angeles

Marshall D. Sahlins
University of Michigan

Edward Schieffelin
University of California

Glynn L. Isaac
University of California, Berkeley

Irawati Karve
University of Poona

C. S. Lancaster
Rutgers University

Donald W. Lathrap
University of Illinois

William S. Laughlin
University of Connecticut

Duhyan Lee
Seoul National University

Richard B. Lee
University of Toronto

Claude Levi-Strauss
College de France

Toichi Mabuchi
Tokyo Metropolitan University

Lorna Marshall
Cambridge, Mas .

M. J. Meggitt
City University of New York

George Peter Murdock
University of Pittsburgh

Chie Nakane
University of Tokyo

David M. Schneider
University of Chicago

William Shack
University of California, Berkeley

Lauriston Sharp
Cornell University

Surajit Sinha
Indian Institute of Management, Calcutta

Richard Slobodin
McMaster University

Julian H. Steward
University of Illinois

Stuart Struever
Northwestern University

Wayne Suttles
Portland State College

Reina Torres de Arauz
Universidad de Panama

Colin M. Turnbull
Hofstra University

Sherwood L. Washburn
University of California, Berkeley

Hitoshi Watanabe
University of Tokyo

James Woodburn
London School of Economics

Aram A. Yengoyan
University of Michigan

John W. M. Whiting
Harvard University

B. J. Williams
University of California, Los Angeles

Part I: Introduction

1. Problems in the Study of Hunters and Gatherers

Richard B. Lee and Irven DeVore

Cultural Man has been on earth for some 2,000,000 years; for over 99 per cent of this period he has lived as a hunter-gatherer. Only in the last 10,000 years has man begun to domesticate plants and animals, to use metals, and to harness energy sources other than the human body. Homo sapiens assumed an essentially modern form at least 50,000 years before he managed to do anything about improving his means of production. Of the estimated ISO billion men who have ever lived on earth, over 60 per cent have lived as hunters and gatherers; about 35 per cent have lived by agriculture and the remaining few per cent have lived in industrial societies.

To date, the hunting way of life has been the most successful and persistent adaptation man has ever achieved. Nor does this evaluation exclude the present precarious existence under the threat of nuclear annihilation and the population explosion. It is still an open question whether man will be able to survive the exceedingly complex and unstable ecological conditions he has created for himself. If he fails in this task, interplanetary archeologists of the future will classify our planet as one in which a very long and stable period of small-scale hunting and gathering was followed by an apparently instantaneous efflorescence of technology and society leading rapidly to extinction. “Stratigraphically,” the origin of agriculture and thermonuclear destruction will appear as essentially simultaneous.

On the other hand, if we succeed in establishing a sane and workable world order, the long evolution of man as a hunter in the past and the (hopefully) much longer era of technical civilization in the future will bracket an incredibly brief transitional phase of human history—a phase which included the rise of agriculture, animal domestication, tribes, states, cities, empires, nations, and the industrial revolution. These transitional stages are what cultural and social anthropologists have chosen as their particular sphere of study. We devote almost all of our professional attention to organizational forms that have emerged within the last 10,000 years and that are rapidly disappearing in the face of modernization.

It is appropriate that anthropologists take stock of the much older way of life of the hunters. This is not simply a study of biological evolution, since zoologists have come to regard behavior as central to the adaptation and evolution of all species. The emergence of economic, social, and ideological forms are as much a part of human evolution as are the developments in human anatomy and physiology.

The time is rapidly approaching when there will be no hunters left to study. Our aim in convening the symposium on Man the Hunter was to bring together those who had recently done field work among the surviving hunters with other anthropologists, archeologists, and evolutionists who are interested in the results of these studies. But it was also clear that there were a series of issue8 among social anthropologists that required clarification before a dialogue with others could become meaningful. Therefore, the first half of this book is devoted to the presentation of new data on cOI1lemporary hunters, along with discussion and evaluation of current issues. The later chapters consider the relevance of these data to the reconstruction of life in the past.

A number of divergent viewpoints are represented in this volume and many of the issues to be raised remain unsolved. As both editors and partisans, it is our task to point out areas of general agreement while trying to avoid glossing over real differences where they occur. Considering the many points of view presented at the symposium, it would be impossible to touch on all of the interesting material. We make no apologies for our selection but offer it as a partial guide to the papers and discussions.

Defining Hunters

The symposium considered the definition of “hunters” but did not succeed in satisfying everyone. An evolutionary definition would have been ideal; this would confine hunters to those populations with strictly Pleistocene economies-no metal, firearms, dogs, or contact with nO!1-hunting cultures. Unfortunately such a definition would effectively eliminate most, if not all, of the peoples reported at the symposium since, as Marshall Sahlins pointed out, nowhere today do we find hunters living in a world of hunters.

Murdock (Chapter 2, this volume) and others took an organizational view which equated hunting and gathering with the “band” level of social organization. However, not all hunters live in bands. Suttles, for example (Chapter 6), documents the quite substantial non-agricultural tribal societies of the Northwest coast of North America. Judging from recent archeological evidence from the Paleolithic of France and of European Russia, a number of ancient hunting societies may have operated on a similar scale.

A further consideration was introduced by Lathrap (Chapter 3) and others who cited a number of hunting peoples who were “failed” agriculturalists. This readaptation to hunting, or “devolution” as it has been called, characterized such classic “hunters” as the Siriono of South America and the Veddas of Ceylon.

It was clear that there was much to be learned even from the ambiguous cases of tribal hunters, “devolved” hunters, and reformed hunters. To throw out impure cases was to lose the chance of gaining any significant understanding of modes of adaptation, group structure, social control, and settlement patterns. The symposium agreed to consider as hunters all cases presented, at least in the first instance. It was also generally agreed to use the’ term “hunters” as a convenient shorthand, despite the fact that the majority of peoples considered subsisted primarily on sources other than meat—mainly wild plants and fish.

The Representativeness of Our Sample of Hunters

There has been a burst of field research on hunting peoples in the last ten years, and this symposium was the first opportunity to bring together this diverse group of field workers.

The Australian contingent included L. R. Hiatt (Arnhem Land), M. J. Meggitt (Walbiri), Arnold Pilling (Tiwi), and Aram Yengoyan, all of whom have done field work in the 1950’s and 1 960’S, as well as Birdsell, Rose, and Sharp whose field work dates to the 1930’s and 1940’s. Workers from Africa included Bicchieri, Turnbull (Mbuti), Woodburn (Hadza), and Marshall, Lee, and DeVore on the!Kung Bushmen. Field work in Asia was reported by Dunn, Gardner, Sinha, Watanabe, and Williams. Lathrap and Crocker presented data on the part-time hunters of tropical South America. The largest group had done field work in North America: Balikci, Damas, and Laughlin on the Eskimos; Helm, Rogers, and Slobodin on subarctic Indians; and Owen and Suttles on Indians of the Pacific coast. Julian H. Steward, in many ways the founder of modern hunter-gatherer studies, submitted a paper, but was prevented by illness from attending the conference itself.

One of the first questions considered at the symposium was whether the sample currently available to ethnographers is representative of the range of habitats which hunters occupied in the past. Ever since the origin of agriculture, Neolithic peoples have been steadily expanding at the expense of the hunters. Today the latter are often found in unattractive environments, in lands which are of no use to their neighbors and which pose difficult and dramatic problems of survival. The more favorable habitats have long ago been appropriated by peoples with stronger, more aggressive social systems.

Taking hunters as they are found, anthropologists have naturally been led to the conclusion that their life (and by implication the life of our ancestors) was a constant struggle for survival. The three maps presented in the frontispiece show a radiGally different picture. In 10,000 B.C., on the eve of agriculture, hunters covered most of the habitable globe, and appeared to be generally most successful in those areas which later supported the densest populations of agricultural peoples. By A.D. 1500 the area left to hunters had shrunk drastically and their distribution fell largely at the peripheries of the continents and in the inaccessible interiors. However, even at this late date, hunting peoples occupied all of Australia, most of western and northern North America, and large portions of South America and Africa. This situation rapidly changed with the era of colonial expansion, and by IgOO, when serious ethnographic research got under way, much of this way of life had been destroyed. As a result, our notion of unacculturated huntergatherer life has been largely drawn from peoples no longer living in the optimum portion of their traditional range.

To mention a few examples, the Netsilik Eskimos, the Arunta, and the!Kung Bushmen are now classic cases in ethnography. However, the majority of the precontact Eskimos, Australian aborigines, and Bushmen lived in much better environments. Two-thirds of the Eskimos, according to Laughlin, lived south of the Arctic circle, and the populations in the Australian and Kalahari deserts were but a fraction of the populations living in the well-watered regions of southeast Australia and the Cape Province of Africa. Thus, within a given region the “classic cases” may, in fact, be precisely the opposite: namely, the most isolated peoples who managed to avoid contact until the arrival of the ethnographers. In order to understand hunters better it may be more profitable to consider the few hunters in rich environments, since it is likely that these peoples will be more representative of the ecological conditions under which man evolved than are the dramatic and unusual cases that illustrate extreme environmental pressure. Such a perspective may better help us to understand the extraordinary persistence and success of the human adaptation.

Ethnographic Reconstruction

Many of the best-known studies of huntergatherers have relied on ethnographic reconstructions of situations that were no longer intact. In the early years of field anthropology, Kroeber in California, Boas on the Northwest coast, and Radcliffe-Brown in the Andaman islands and Australia compiled the recollections of older informants to produce a picture of the culture and society. The “facts” of the case were thought to include the verbalized cultural tradition—language, myths, folktales, and kin terms—and concrete expressions— rituals, house types, tools, dress, and religious objects. As anthropology developed, however, a different view emerged about what constituted the ethnographic facts. More attention was paid to the study of individuals and groups in ongoing social systems. The interest in ideology was retained, but this was tempered by the task of comparing stated norms with observed behavior. When discrepancies between “real” and “ideal” came to light, the ethnographer could then make further inquiries and observations and attempt to explain “how the system really works.”

The shortcoming of the reconstructive method is that there is no means of testing and rechecking hypotheses. After the socioeconomic system ceases to function, the only check available is the test of internal consistency, and the early ethnographers were more or less successful in constructing models of social systems that were self-consistent.

The controversy on this question is very much alive in current social anthropology and the issue forms a central theme in this volume (Chapters 10, qa, 17c, 17h, 18, Ig, 22b, and 22(’). The relevance for methodology hinges on the question of how much of an ongoing system can be reconstructed 25–50 years after the event, solely on the basis of informants’ accounts. Birdsell and others take the pessimistic view that such important features as group structure and territorial relations vanish with contact (Chapter 17a). Williams, in the same vein, reports that he was unable to reconstruct the residence arrangements of a Birhor group whose campsite had been mapped in detail by another anthropologist only six years previously. Woodburn notes that the Badza present themselves as having a matrilineal descent system, when in fact the kinship and the group structure are bilateral; the fiction is used by the Hadza to emulate their agricultural, matrilineal neighbors. This fact is significant in itself, but more important, it would make impossible an accurate reconstruction of Badza social organization at a later date, if one had to rely solely on the memory of informants.

On the other hand, several participants were more sanguine about the persistenee of cultural traditions into the postcontact period. Levi-Strauss cited the example of the retention of marriage ideology by Australian aborigines in the face of acculturation and demographic reverses (Chapter 17b, 17c). In addition, careful reconstructions of precontact ecological situations were presented by Watanabe, for the Ainu of the 1880’S (Chapter 7), and by Balikci, for the Netsilik Eskimos in 1919–20 (Chapter 8).

Finally, June Helm pointed out that the group structure of such diverse peoples as the Dogrib Indians, the!Kung Bushmen, and the Nambikwara showed striking similarities in spite of widely differing social ideologies (Chapter 13). This suggests that small-scale societies may arrive independently at similar solutions to similar demographic and ecological problems. Without the opportunity of observing behavior, however, such an important point would be impossible to establish. This problem is discussed further by Anderson (Chapter 17c).

The Subsistence Base

Strictly speaking, hunting apd gathering refers to a mode of subsistence, and many of the conference papers discussed problems of ecology and economic organization. Several field workers pointed out that the subsistence base of hunters was much more substantial than had been previously supposed. It came as a surprise to some that even the “marginal” hunters studied by ethnographers actuallywork short hours and exploit abundant food sources. Several hunting peoples lived well on two to four hours of subsistence effort per day and were not observed to undergo the periodic crises that have been commonly attributed to hunters in general (Chapters 4, 5, and gb). Other peoples for whom detailed activity data were unavailable were nevertheless reported to show a “lack of concern” about the problem of finding food. This led the conference participants to speculate whether lack of “future orientation” brought happiness to the members of hunting societies, an idyllic attitude that faded when changing subsistence patterns forced men to amass food surpluses to bank against future shortages (Chapter gc).

Dissenting views were reported. Balikci spoke of a “constant ecological pressure” which caused real hardship and anxiety among the Netsilik Eskimos (Chapter 8). Williams (Chapter gc) found that the Birhor of India not only worked hard for their food but often went hungry. It was clear that Sahlins’ characterization of the hunters (Chapter gb) as the “original affluent society” would not apply in all cases. However Sahlins’ argument served to underline the point that anthropologists have tended to view the hunters from the vantage point of the economics of scarcity. Viewed on their own terms, the hunters appear to know the ibod’ire-sources of their habitats and are quite capable of taking the necessary steps to feed themselves. It is unlikely that hunters would have deliberately chosen the catch-as-catch-can existence that has been ascribed to them. Since a routine and reliable food base appears to be a common feature among modern huntergatherers, we suspect that the ancient hunters living in much better environments would have enjoyed an even more substantial food supply.

Hunting vs. Gathering

The hunting of mammals has been considered the characteristic feature of the subsistence of early man, and two chapters in this volume explore the implications of hunting for human evolution (Washburn and Lancaster, Chapter 32; and Laughlin, Chapter 33). Modern hunters, however, depend for most of their subsistence on sources other than meat, mainly vegetable foods, fish, and shell fish. Only in the arctic and subarctic areas where vegetable foods are unavailable do we find the textbook examples of mammal hunters. Over the rest of the world, hunting appears to provide only 20 to 40 per cent of the diet (Lee, Chapter 4).

Binford, Washburn and Lancaster, and others expressed the view that fishing, seedgrinding, and hunting with dogs are late adaptations, dating from the Mesolithic and therefore not characteristic of Pleistocene conditions. Thus, the eclectic diet of the modern hunters may tell us little about the food habits of early man. Our own view is that vegetable foods in the form of nuts, berries, and roots were always available to early man and were easily exploited by even the simplest of technologies. It is also likely that early woman would not have remained idle during the Pleistocene and that plant foods ^which are so important in the diet of inland huntergatherers today, would have played a similar role in the diet of early peoples. We agree, however, that hunting would become increasingly important as populations migrated out of the tropics into areas where plant foods are scarce. In addition, hunting is so universal and is so consistently a male activity that it must have been a basic part of the early cultural adaptation, even if it provided only a modest proportion of the food supplies.

Social and Territorial Organization

The analysis of the social structure of the hunting and gathering peoples proved to be a particularly difficult area of investigation because of the ambiguous conditions in which hunters are currently found. Fluidity of band composition appeared to be the most characteristic feature of the modern hunters, but participants disagreed on whether this flexibility was basic to the hunting way of life or whether it was a product of recent acculturation. We will briefly outline the history of band theory from Radcliffe-Brown, through Steward to Service and then go on to consider the evidence from recent field . research.

The Patrilocal Band

Radcliffe-Brown gave this concept its modern expression in “Social Organization of Australian Tribes,” where he described the patrilineal, patrilocal, territorial, exogamous horde as “the important local group throughout Australia” (1931, p. 35). The horde was a group of patri-laterally related males who lived and worked within their totemic estate and exchanged women with other such male-centered groups. Steward, in a general review of huntergatherers, admitted the horde or “patrilineal band” as one of three types of group structure, but he added the composite bands of the Northern Athapaskans and the family bands of Great Basin Shoshoneans (1936, 1955). This three-part division was later called into question by Elman Service (1962), who took the view that the “patrilocal band,” as he called it, was not only the characteristic form of local organization in Australia but was also the basic form for all hunter-gatherers in the past (1962, pp. 65–67, 107–109). The composite and family bands, in Service’s view, were artifacts of recent acculturation and breakdown.

Recent research has cast doubt on the model of the patrilocal band. Hiatt, reviewing the literature on Australian local organization (1962), failed to find a single indisputable example of Radcliffe-Brown’s horde. The existence of patrilineal totemic clans has been well documented, but these functioned in ritual contexts and not as residential or economic units. Hiatt concludes:

It is now clear that over a great deal of the continent the male members of totemic descent groups did not live together on separate pieces of land. They commonly lived in communities that contained male members of several totemic descent groups and regularly sought food over areas that included totemic sites other than their own. Investigators that failed to find the horde in particular tribes were not (as some of them thought) observing aberrant forms of local organization. They were probably looking for something that never existed in any tribe (1962, p. 286).

Other ethnographers who had worked among “patrilocal band” peoples also Jailed to find this form of organization operating now or at any discoverable period in the past. Turnbull among the Mbuti Pygmies (Chapter 15), Lee^and Marshall among the Bushmen (Chapters 4 and I7c) and Meggitt (Chapter 19) all repo.^t the occurrence of composite and flexible local groups. Others have studied the ethno-historic record of peoples that Service claimed had patrilocal organization in the past. Helm on the Athapaskans (Chapter 13), Damas on the Eskimo (Chapter 12), and Pilling on the southeastern Australians (Chapter 16) present evidence that the composite bands observed in the present were characteristic of the earliest contact periods as well.

The patrilocal band, however, is not an empty category; cases are presented in which a patrilocal organization was in evidence, including the Ona of South America (Bridges, 1949, cited by Lathrap in Chapter 9d), the Kaiadilt of Australia (Tindale, 1962b, cited by Birdsell in Chapter 17a) and the Birhor of India (Williams in Chapter 14).

On the basis of present evidence, it appears that the patrilocal band is certainly not the universal form of hunter group structure that Service thought it was. Anderson (Chapter 17c), Turnbull (Chapter 15), Eggan (Chapter 17i), and others pointed out that the fluid organization of recent hunters has certain adaptive advantages, including the adjustment of group size to resources, the leveling out of demographic variance, and the resolution of conflict by fission.These features are independent of the effects of acculturation and would have been no less adaptive in precontact situations. Given these advantages, few of the conferees would endorse Service’s view that the patrilocal band “seems almost an inevitable kind of organization” or that “the composite band ... is obviously a product of the near destruction of aboriginal bands after contact with civilization” (1962, p.108). In fact Service himself has recently revised his opinion that the patrilocal band is inevitable.

Service has correctly drawn attention to the effects of contact on hunter-gatherer social organization, and this remains a difficult problem for the ethnographers. Although fluid organization makes ecological sense for many hunters, and could be shown to be more adaptive than patrilocal organization, in itself this is not evidence for aboriginal conditions.Since all hunting societies have suffered to some extent from contact, we may never be able to prove conclusively that one form or another was typical of the past in any specific case.

The Problem of Corporation

Sharp (Chapter 17h) warns against the use of “prefabricated constructs” in the study of band organization.The prevalence of such constructs may derive from the application of concepts developed in tribal societies to the study of simpler and smaller-scale societies. The social units of tribal societies are based on the control and husbandry of real property usually in the form of agricultural land or livestock (Fortes, 1953). In addition, the formal political institutions such as courts, councils, and chieftainships lends a distinctive character to social relations of tribesmen that is lacking in the smaller-scale societies of hunters and gatherers. Radcliffe-Brown’s well-known views on the universal importance of lineal descent and corporation (1952) may have led him to impose a corporate unilineal model on Australian local groups.

However, the reports in this volume make it clear that the hunter-gatherer band is not a corporation of persons who are bound together by the necessity of maintaining property. A corporation requires two conditions: a group of people must have some resource to incorporate about and there must be some means of defining who is to have rights over this resource. Among most hunter-gatherers one or both of these conditions are lacking. In Australia and among the Birhor, the patrilineal descent groups are present but these groups do not maintain exclusive rights to territorial resources (Hiatt in Chapter 10, Williams in Chapter 14). Among the Mbuti Pygmies the territories are well-defined but the membership in the resource-using group is open and changes frequently (Turnbull in Chapter 15). Among the Dogrib (Helm in Chapter 13), the Bushmen (Lee in Chapter 4), the central Eskimo (Damas in Chapter 12 and Balikci in Chapter 8) and the Hadza (Woodburn in Chapter 1 I), both the composition of the group and the range they exploit may vary from season to season. The existence of this variety makes it difficult to accept a model of hunter local groups that encapsulates each group of males within a territory, with exchange of women as the primary mode of communication between groups.

Flexibility and the Resolution of Conflict by Fission

Turnbull (lg6Sb and Chapter 15) has defined an important means of social control among the Mbuti Pygmies. When disputes arise within the band, the principals simply part company rather than allow the argument to cross the threshold of violence. By this seemingly simple device, harmony is maintained within the cooperating group without recourse to fighting or to formal modes of litigation. The essential condition seems to be the lack of exclusive rights to resources; thus it is a relatively simple matter for individuals and groups to separate when harmony is threatened. The effect of this practice is to keep the population circulating between band territories. Such a form of conflict resolution would not be possible in situations where social units are strictly defined and firmly attached to parcels of real estate. Woodburn and Lee reported a similar mode of conflict resolution among the Hadza and!Kung Bushmen; judging from their generally flexible group structure, resolution of conflict by fission may well be a common property of nomadic hunting societies.

Order and Chaos in Hunter-Gatherer Society

One of the attractions of the patrilocal band model is the neatness of its formulation. In such a model the society is structured by the operation of a small number of fundamental jural rules: territorial ownership by males, band exogamy, and viripatrilocal postmarital residence. However, the model leaves scant room either for local and seasonal variations in food supply or for variance in sex ratios and family size within and between local groups. Nor does the model, in many cases, accord with the observed facts, and in this light it appears to be more a convenience for the analyst than a workable socioeconomic system.

Flexibility of living arrangements presents at first a confusing and disorderly picture. Brothers may be united or divided, marriage may take place within or outside the local group, and local groups may vary in numbers from one week to the next. Helm (lg6sa and Chapter 13) has made a valuable contribution to the methodology of band-composition analysis. By tabulating the primary relative linkages within Athapaskan groups, she has been able to show that married couples always join groups in which a sibling, parent, or child of one of the spouses is already resident. This produces a local group whose members are joined by a series of links through brothers, sisters, and spouses, often centered around a strong core of male siblings. However, the orientation is bilateral rather than lineal and serves to unite persons into effective work groups regardless of whether the kinship ties are patrilateral, matrilateral, or affinal. Application of this method to other hunters reveals a similar mode of affiliation among the Nambikwara, Bushmen, Hadza, and Arnhem Land aborigines. The bilateral nature of hunter local groups is not simply a matter of people failing to conform to their stated jural rules; several of the peoples just mentioned do not express an ideology of fixed membership groups. There is order in hunter-gatherer society, although not necessarily an order imposed by the operation of residence and descent rules.

Marriage Arrangements in Australia

The way Australian aborigines marry has held a central place in the anthropological study of kinship. The articulation between kin terms and cross-cousin marriage on the one hand, and moieties, sections and subsections on the other, have provided materials for a number of analysts, including Radcliffe-Brown, Leach, Murdock, Needham, and Levi-Strauss. Since the symposium was concerned with the hunting way of life and not with the study of marriage per se, the several papers on the subject attempted to place Australian marriage and section systems into the general context of hunter-gatherer ecology and demography. Hiatt, for example (Chapter 18) presents a model of the operation of an eight-section systern and then goes on to show that, given the small population of the Gidjingali of Arnhem Land, only a fraction of the marriages can be contracted between the cross-cousins specified by the model. The lack of women of the appropriate kin position makes it necessary for most men to seek a spouse elsewhere. Although over go per cent of the marriages are “morally proper, “ in other words contracted between persons of the correct subsections, the net result is that women do not function as a medium of exchange between descent groups. In his comment (Chapter 22b) Levi-Strauss acknowledges the importance of relating stated norms to demographic variables but he goes on to say that his work in Australian kinship has been “concerned with a different problem: to ascertain what was the meaning of the rules, whether they be applied or not.” He expresses the view that the evidence of a “collapsing Australian tribe” is not sufficient to make us discard the earlier formulations of RadcliffeBrown and others which demonstrated that marriage was indeed regarded by the natives as an exchange of women between descent groups.

Meggitt (Chapter 19) and Yengoyan (Chapter 20) both bring demographic data to bear on the study of section systems, but they arrive at different conclusions. Meggitt notes that a series of fifteen central Australian tribes all exhibit a division into eight subsections, despite the fact that the populations of several of the tribes are less than 200 persons, too small for marriages to be cOEtracted only between the members of appropriate subsections. This fact leads Meggitt to conclude that “among the Australian aborigines, subsections are not necessarily concerned in ...the regulation of marriage” (Chapter 19). Yengoyan argues that subsections do regulate marriage, but only very large tribes can successfully operate an eightsection system. A tribe of 1,100 persons, for example, would have no difficulty in arranging “proper “ marriages since each of the eight divisions would contain about 25 eligible mates. On the other hand, a tribe numbering 200 persons or less would not be able to arrange marriages without considerable deviation from the stated rules and, in fact, few of the small tribes in Yengoyan’s sample exhibited an eightsection organization.

Rose (1960b and Chapter 21) adds a further-demographic variable by demonstrating that among the Australians, men tended to marry women a generation younger than themselves. In combination with polygyny, this practice allows men in their forties and fifties to monopolize the majority of women of marriagable age. The function of this “gerontocracy” as Rose calls it, is to combine the women into effective food-producing units, centered around a male at the peak of his “political” career.

Recent demographic research has added a new dimension to the study of Australian systems. Aboriginal marriage now appears to be more explicable in terms of the economic conditions under which the people actually lived. Yet the basic question raised by Levi-Strauss remains: How are the stated rules of social life articulated into a meaningful whole? It is clear that more work is urgently required in Australia, although we suspect that it may already be too late to establish the relation of ideological forms to the behavior of persons on the ground.

Demography and Population Ecology

Many of the peoples discussed at the symposium live in small-scale societies in which the total population numbers a thousand persons or less. The social and ecological consequences of “smallness” were considered in several papers. Dunn (Chapter 23) made the interesting point that small groups of nomadic peoples living at low population densities would be much more resistant to epidemic diseases than farming peoples who live at higher densities in settled villages. Others speculated on the ideal sizes of hunter local groups and breeding populations, and two figures in particular came to be called the “magic numbers” of huntergatherer demography (Chapter 25C). Birdsell (Chapter 24) cites a figure of 500 as the modal size of tribes in aboriginal Australia; such a population was large enough to insure an adequate recruitment of mates and yet small enough for everyone to know everyone else. There was some evidence that tribes of larger size tended to subdivide into smaller and more manageable units. Twenty-five to fifty persons were the figures most frequently reported for the size of local groups or bands among modern hunter-gatherers, although we failed to arrive at any satisfactory explanation for this central tendency.

Perhaps the most important demographic issue was the question of what are the factors that keep the populations of hunters in check. Throughout the world hunter densities rarely exceed one person per square mile; most of the accurate figures reported at the conference ranged between one and 25 persons per hundred square miles. We feel that the one-per-square-mile figure is a useful estimate of Pleistocene carrying capacity. It was not until the development of agriculture that human populations were able to break this limit for the first time, while present agricultural densities in most parts of the world exceed the highest hunter densities by a factor of ten to one thousand.

Disease, malnutrition, and infanticide were each considered as possible control mechanisms for hunter-gatherer populations. Since food supply appears to be abundant in modern cases, the constant threat of famine has probably been overestimated as a population control. On the other hand, the management of fertility by means of long lactation, birth control, the killing of twins, and systematic infanticide may have been as frequent in early populations as they are among hunters today. Birdsell and Deevey (Chapters 22b, 22d) went so far as to suggest that a 15–50 per cent incidence of infanticide may have existed throughout the preagricultural history of man.

Helm and Washburn (Chapter 9a) following Bartholomew and Birdsell (1953) suggested that we consider what works in the long run. A population may thrive for several generations and give every appearance of being successfully adapted, Gnly to be cut back severely by an ecological reverse once in a hundred years. Probably the most successful long-term situation is for a population to become stabilized at 20 or 30 per cent of carrying capacity, but at the present it is not possible to determine how such an equilibrium may be achieved. Fortunately, it is still possible to gather the necessary d.emographic data among some of the modern hunter-gatherers, and we hope that these data, plus computer simulation of population models, will yield exceptionally useful results.

“Nomadic Style”: A Trial Formulation

Although the conference on Man the Hunter raised more questions than it answered, there seemed to be a widespread feeling among the participants that a useful beginning had been made in understanding the hunters better. A number of older theories were corrected in light of new data and, where issues were unresolved, we came away at least with a clearer appreciation of our differences, To attempt to draw a general pioture of the hunters at this time would certainly be premature; we would, however, like to offer a trial formulation of our views to serve as a starting point for future research and discussion.

We make two basic assumptions about hunters and gatherers: (i) they live in small groups and (2) they move around a lot. Each local group is associated with a geographical range but these groups do not function as closed social systems. Probably from the very beginning there was communication between groups, including reciprocal visiting and marriage alliances, so that the basic hunting society consisted of a series of local “bands” which were part of a larger breeding and linguistic community. The economic system is based on several core features including a home base or camp, a division of labor—with males hunting and females gathering—and, most important, a pattern of sharing out the collected food resources.

These few broadly defined features provide an organizational base line of the small-scale society from which subsequent developments can be derived. We visualize a social system with the following characteristics. First, if individuals and groups have to move around in order to get food there is an important implication: the amount of personal property has to be kept to a very low level. Constraints on the possession of property also serve to keep wealth differences between individuals to a minimum and we postulate a generally egalitarian system for the hunters.

Second, the nature of the food supply keeps the living groups small, usually under fifty persons. Large concentrations of population would rapidly exhaust the immediate resources, and members would be forced to disperse into smaller foraging units. It is likely, as Mauss observed (1906), that several bands would come together on a seasonal basis, resulting in a division of the year into “public” and “private” periods. Because of the small size of the living groups and the wide variance of family size, bands wax and wane in numbers. It is probably necessary to continually redistribute the population between bands in order to maintain food-gathering units at an effective level.

Third, the local groups as groups do not ordinarily maintain exclusive rights to resources. Variations in food supplyofrom region to region and from year to year create a fluid situation that can best be met by flexible organizations that allow people to move from one area to another. The visiting patterns create intergroup obligations, so that the hosts in one season become the guests in another. We think that reciprocal access to food resources would rank as equal in importance with exchange of spouses as a means of communication between groups. It is likely that food may antedate women as the original medium of exchange (cf. Levi-Strauss, 1949).

Fourth, food surpluses are not a prominent feature of the small-scale society. If inventories of food on hand are minimal, then a fairly constant work effort has to be kept up throughout the year. Since everyone knows where the food is, in effect the environment itself is the storehouse; and since everyone knows the movements of everyone else, there is a lack of concern that food resources will fail or be appropriated by others.

Fifth, frequent visiting between resource areas prevents any one group from becoming too strongly attached to any single area. Ritual sites are commonly associated with specific groups, but the livelihood of the people does not depend on such sites. Further, the lack of impediments in the form of personal and collective property allows a considerable degree of freedom of movement. Individuals and groups can change residence without relinquishing vital interests in land or goods, and when arguments break out it is a simple matter to part company in order to avoid serious conflict. This is not to say that violence is unknown; both homicide and sorcery are found among a number of current hunter-gatherers. The resolution of conflict by fission, however, may help to explain how order can be maintained in a society without superordinate means of social contro!’

At the symposium we presented some of these ideas under the general heading of the “Nomadic Style. “ Several others found this view plausible and suggested ways in which specific cases might have developed out of this baseline. L. Binford noted that northern adaptations required a more elaborate material basis, including fixed facilities such as fish J weirs and game fences, and more substantial dwellings, clothing, and tool kits. Steward pointed out that the development of traction! and water transport would allow northern peoples to own more and yet still retain mobility (Chapter 34). And Suttles (Chapter 6) showed how the fisherman of the Northwest coast may have originally started out in small nomadic bands, but were subsequently led into surplus accumulation, heavy facilities, and ceremonial exchange by the necessity to bank against vagaries in salmon supply.

It seems clear that when the means of production come to depend upon the exclusive control of resources and facilities, then the loose non-corporate nature of the small-scale society cannot be maintained. If this view is: correct, then a major trend in human affairs has been the transformations of social relations as advanced technologies and formal institutions have come to play a more and more dominant role in the human adaptation. The institutions of property, of clan organizado^ of government, and of the state did not spring full-blown in a divine creation. The study of the hunters may help to understand how these things came into being.

2. The Current Status of the World’s Hunting and Gathering Peoples

George Peter Murdock

Ten thousand years ago the entire population of the earth subsisted by hunting and gathering, as their ancestors had done since the dawn of culture. By the time of Christ, eight thousand years later, tillers and herders had ^ replaced them over at least half of the earth.

At the time of the discovery of the New World, only perhaps 15 per cent of the earth’s surface was still occupied by hunters and gatherers, and this area has continued to decline at a progressive rate until the present day, when only a few isolated pockets survive.

Many of the peoples who subsisted by hunting and gathering at the time of Columbus have disappeared entirely and been replaced by stronger peoples. Others have been reduced to dependency-on reservations or as servants or outcaste groups. Still others have made a transition, in some form or other, to modes of subsistence based on agriculture, animal husbandry, or industrial employment. Only a handful still live an independent life of hunting and gathering, nearly all of them under markedly altered conditions. It is time-in-deed, long past time-that we took stock of these dwindling remnants, for it is only among them that we can still study at first hand the modes of life and types of cultural adjustments that prevailed throughout most of human history.

In this paper I shall attempt to enumerate and briefly characterize the regions of the world where peoples with hunting and gathering economies survived long enough to be studied by modern ethnographers, with special reference to those groups amongst whom fundamental field research may still be possible.

The definition of hunting and gathering economies was the subject of considerable discussion at the symposium. For the sake of simplicity in presentation, I shall adopt a strict definition and merely allude here, by way of introduction, to three marginal subsistence categories which some would include with hunting and gathering and others would differentiate.

Mounted Hunters

Several groups of peoples who subsist primarily by hunting depend largely on domestic animals which they ride in pursuing or surrounding their game. Notable among them are the North American Indians of the Plains and the adjacent sections of the Plateau and Great Basin and the South American Indians of Patagonia and the adjacent Gran Chaco, both of which groups hunted with horses during the period of their ethnographic description. Comparable to them are several hunting peoples in Mauritania, including the Nemadi described by Gabus (1951–52), who employ camels in the chase. Mounted hunters approximate pastoral nomads in a number of respects, the chief difference, perhaps, being the fact that their domestic animals do not provide a majot part of their food supply.

r-b-richard-b-lee-irven-devore-man-the-hunter-5.jpg
Figure 1. Distribution of Recent Hunters and Gatherers.

Sedentary Fishermen

Various other peoples who lack both agriculture and large domestic animals depend not so much on hunting and gathering as on fishing, shellfishing, or the pursuit of aquatic animals. Where they lead an essentially nomadic mode of life, I classify them with hunters and gatherers. Sometimes, however, their marine food supply is so plentiful and stable that they have been able to adopt an exclusively or predominantly sedentary mode of life. Such conditions offer the possibility of attaining a considerable degree of cultural complexity otherwise achievable only with intensive agriculture. The Indians of the North Pacific coast, for example, seem to me to fall well beyond the range of cultural variation of any known hunting and gathering people. The Calusa, a sedentary fishing people of southern Florida, as Goggin and Sturtevant (1964) have shown, had even elaborated a political system with fairly sizable states. Along the Niger and Congo rivers in Africa, numerous societies which subsist primarily by fishing and trade have cultures in no way inferior in complexity to those of their agricultural neighbors. In northeastern Asia the Ainu, Maritime Chukchee, Gilyak, Kamchadal or Italmen, and Maritime Koryak possibly, though less certainly, fall into the same category.

Incipient Tillers

Not a few peoples, especially though not exclusively in lowland South America, practice agriculture but obtain more of their food supply from hunting, fishing, and gathering and show a clear preference for hunting over tillage. Several participants in the symposium, notably Crocker with reference to the Ge, characterized such incipient agriculturists as having a hunting-gathering ideology. While I freely grant their point, I have preferred to exclude such groups from my survey. After all, I know many American farmers who till the soil for a living but derive considerably more personal gratification from their subsidiary hunting and fishing activities.

In the listing of hunting and gathering peoples below, consequently, I omit all groups which I would categorize as mounted hunters, sedentary fishermen, or incipient tillers. I gratefully acknowledge the many suggestions made by participants in the symposium which have enabled me substantially to amplify and revise the original draft of this paper. Twenty-seven surviving groups of hunter-gatherers are enumerated in this survey. The geographical distribution of these peoples is plotted in Figure I.

Africa

  1. Bushmen

I shall begin my enumeration with the Bushmen of south and southwest Africa, of whom about 45,000 are reported to survive today (Tobias, 1956; Silberbauer, 1965). The vast majority of these, however, are attached to the Tswana as serfs or to Europeans as servants or laborers and have undergone intensive acculturation. Nevertheless, about 5,000 of them still pursue essentially their aboriginal nomadic mode of life. Lorna Marshall (1957, 1959, 1960), in particular, has demonstrated that field work of the highest quality is still possible among them. They are dealt with in greater detail by Richard B. Lee (Chapter 4, this volume).

  1. Koroca

On the desert coast and adjacent foothills of southwestern Angola live the Koroca and the related Kwepe and Kwise, who practice no agriculture. From very fragmentary reports, they appear to resemble the Bushmen in culture and the Bergdama in physique (Almeida, 1965). Remarks by Jorge Dias suggest that they merit more intensive study.

Desmond Clark believes that there may also be Bantu hunting tribes in the same general area.

  1. Pygmies

About 170,000 Pygmies are reported to survive in the tropical forest region of central Africa, nearly all in a dependent symbiotic relationship with the neighhoring Negro tribes. The majority, especially those in Rwanda, Burundi, and the central portion of the Republic of the Congo, have either adopted agriculture or otherwise become strongly acculturated. In the Ituri Forest region of the northeast, however, the Aka, Efe, and Mbuti groups still preserve large portions of their indigenous culture. For the Mbuti, in particular, we have a substantial monograph by Schebesta (1938–50) and the more intimate and revealing studies by Turnbull (1961, 1965b). There is certainly room for further field work here, and probably also among the inadequately described Pygmies of Cameroon and Gabon.

  1. East African Hunters

Until very recently, available information on the surviving hunting populations of east Africa has been exceedingly scanty. Now, however, we have a series of substantial accounts of the Dorobo from Huntingford (1951, 1954, 1955) and the still more recent field work among the Hadza or Kindiga by James Woodburn (Chapters 5, 3, this volume). The Teuso or Ik described by Colin Turnbull, though possibly related to the Dorobo, are not a hunting people in the strict sense adopted here. It is doubtless too late to expect new information on the Asa or Aramanic beyond the scraps contained in Merker(1904),but it still maybe possible to conduct rewarding field work among the Ariangulo, the Boni, or the Sanye.

  1. Ethiopian Hunters

William Shack reminded the symposium that there are half a dozen groups of hunters and gatherers in central Ethiopia, extending from the lakes of the Rift valley on the east to the edge of the plateau in the west. He called particular attention to the Fuga, whom he found still flourishing in the Gurage country. I omitted these peoples in my original survey because, from the scanty references available to me, they seemed more like despised outcaste groups than independent tribes. If they do in fact lead independent lives, they certainly deserve early and intensive study.

East Asia

  1. Siberian Hunters

The non-agricultural peoples of Siberia have long since been largely converted to a pastoral mode of life in the interior or, on the coast, have come to depend primarily upon fishing and have adopted a relatively sedentary settlement pattern. The only two Siberian peoples who seem to meet our strict definition of hunters and gatherers are the Ket or Yenisei Ostyak, on whom wehave a briefsketch byShimkin (1940), and the Yukaghir, on whom we have a full monograph by Jochelseon (1926). Both tribes have doubtless been so altered by Soviet acculturation policies that further field work among them is unlikely to yield rewarding results.

  1. Indian Hunters

There are in India a number of hunting societies, mostly dependent and often approximating the status of specialized castes. The fullest ethnographic account is probably that by Furer-Haimendorf (1943) on the Dravidianspeaking Chenchu, but Majumdar (1944) has contributed a brief account of the Korwa of Uttar Pradesh. B. J. Williams describes the Munda-speaking Birhor in this symposium (Chapter 14), and there is fragmentary information on the Irula and Kurumba of the Nilgiri Hills. Peter Gardner has provided me with a list of additional Indian hunting groups: the Kadar, Malapandaram, Paliyan, and Yanadi, all of whom have been studied, and the still undescribed Allar, Aranadan, Eravallar, Mala Vedan, and Vettuvan.

  1. Veddoid Hunters

It is almost certain that no further ethnographic information of consequence is obtainable from the Rock Yedda of Ceylon, who had already practically disappeared when the Seligmans (Igl I) made their classic study. I agree with Gardner that the Yedda are probably a case of “devolution. “ The Senoi or Sakai of Malaya, according to a personal communication from Robert Dentan on the basis of his recent field work among them, have everywhere adopted a substantial dependence upon agriculture. I have no recent information on the peoples of Indonesia who have been called Veddoid, very probably incorrectly, and who have been reported to practice very little agriculture, such as the Kubu of Sumatra (Hagen, Ig08; Schebesta, Ig28) and the Punan of Borneo (Furness, Ig02; Hose and McDougall, Igl I; Roth, 18g6).

9. Southeast Asian Hunters

Participants at the symposium have called my attention to several hunting and gathering peoples in Southeast Asia of whom I had previously been unaware. Thus Georges Condominias cites the Ruc of central Vietnam, Peter Gardner the Penang and Phi Tong Luang, and Lauriston Sharp the Mrabri or Yumbri of northern Thailand. The References at the end of this volume include the following supplied by Sharp on the Mrabri: Bodes (lg63), Flatz (lg63, 1964), Kraisri and Hartland-Swann (1962), Velder (1963, 1964).

10. Negritos

Most of the Negrito tribes of the Philippines practice at least a modicum of agriculture and thus fall outside the range of our discussion. We do, however, possess substantial accounts of two non-agricultural Negrito groups—the Andamanese by Man (1883) and RadcliffeBrown (1933) and the Semang of Malaya by Evans (1937) and Schebesta (1929). I suspect that further field work may still be possible among both of them.

Oceania

11. Australian Aborigines

The Tasmanians are long since extinct (although new ethnohistoric materials have appeared: Plomley, 1966), and the peoples of Melanesia, Micronesia, and Polynesia who lack large domestic animals and depend only minimally on agriculture are sedentary in consequence of their dependence on fishing, sago gathering, or trade, so that they fall outside our definition. This leaves the Australian aborigines as the sole surviving representatives of hunters and gatherers in Oceania. Most of them have either disappeared or have become strongly acculturated. Recent work among the Tiwi, in Arnhem Land, on Groote Eylandt, and elsewhere, however, demonstrates that the possibilities for productive field work have by no means been yet exhausted. Substantial additional information on the Australian indigenes is presented by Frederick Rose (Chapter 2 I), Aram Yengoyan (Chapter 20), L. R. Hiatt (Chapters io and 18), Mervyn Meggitt (Chapter ig), and Arnold Pilling (Chapter 16) in this volume.

North America

  1. Eskimo

Field research among the Eskimo has exhibited a remarkable spurt in recent years. Asen Balikci (Chapter 8) and David Damas (Chapter 12, this vol ume) present some of the results, which are perhaps especially illuminating in the field of social organization. Some Eskimo groups have long approximated a sedentary mode of life because of primary dependence upon the products of the sea, and others have recently been approaching this status because of acculturative influences. The majority, however, are still migratory for at least a part of the year, and a few inland groups are completely divorced from the sea. Among the latter are the Nunamiut of northern Alaska, on whom we have a recent substantial monograph by Gubser (1 g65).

  1. Northeastern Algonkians

The hunting patterns of the Algonkian tribes of eastern Canada were clearly modified considerably with the development of the fur trade. The available information is relatively good, though the input of new information appears to be dwindling. Whenever I begin to suspect that the law of diminishing returns is setting in, however, I am reassured by some significant new publication from the pen of Dunning (1959), Honigmann (I 962a, 1962 b), or Leacock (1954).

  1. Northwestern Athapaskans

A few years ago it appeared that the potential ethnographic crop for the Athapaskan tribes of northwestern Canada and interior Alaska had been almost completely harvested by Osgood, Honigmann, and McKennan. Appearances were deceptive, for recently we have had a succession of monographs and articles of the very highest quality from June Helm (1961, 1965a, Helm and Lurie, 1961). The prospects for further productive research in this area thus seem by no means closed.

  1. Plateau Indians

The Indians of the Plateau area resemble those of the adjacent Northwest coast in many respects, notably in the widespread dependence upon salmon fishing, but they devoted considerably more time and energy to seasonal migrations in search of roots, berries, and game. New information is still appearing in a fairly steady stream, though much of it is perhaps supplementary rather than fundamental. For the northern tribes of the area, notably the Lillooet and Shuswap, I suspect that basic field work of a quality comparable to that of Helm on the Northern Athapaskans may still be possible for a number of years to come.

  1. California Indians

Alfred Kroeber and his students, both through intensive field work and the Culture Element Survey, have probably exhausted most of the possibilities for fundamental ethnographic research among the tribes of the California area, so that only fragmentary information remains to be recovered, except perhaps across the border in Baja California.

  1. The Great Basin

The prospects seem somewhat better for the Shoshonean, Yuman, and other tribes of the Great Basin. My basis for this judgment is the surprising amount of new and valuable data which eight of my own graduate students have brought back from a summer each with the field training program conduc!ed for the past two years by the University of Nevada.

  1. Gulf Indians

On the Indian tribes who formerly inhabited coastal Texas and adjacent northeastern Mexico—the Karankawa, Tonkawa, Coahuilteco, etc.-we possess only fragmentary data, and it is much too late to obtain any new information. It is interesting to note, however, that fairly substantial archival material is beginning to come to light from sixteenth-century Spanish missionaries. The Drivers (1963) have recently summarized the information on the Chichimeca-Jonaz, and Kimball Romney informs me that he has unearthed remarkably full kinship data on the Coahuilteco.

  1. Apache

Thirty years ago Morris Opler was able to obtain full ethnographic data on most of the Apache tribes of the Southwest, by no means all of which has yet been published. Field work continues in this area, with results that are by no means always disappointing.

  1. Seri

The Seri Indians of the coast of Sonora on the GulfofCaliforniaare known from a famous but partly fanciful monograph by McGee (1898) and from a more sober sketch by Kroeber (193 1), both unfortunately based on extremely brief field trips. During the symposium Roger Owen characterized the Seri as “one of the black marks on American ethnography” and expressed the opinion that field work of good quality is still possible among them.

South America

  1. Warrau

As Wilbert (1958a) has demonstrated in a recent monograph, the Warrau Indians of the delta of the Orinoco River still adhere in large measure to an essentially aboriginal mode of life, and fundamental field work of high quality is still possible among them.

  1. Interior Marginal
    Tribes of Southern Venezuela

The so-called “interior marginal “ peoples of southern Venezuela and adjacent Colombia and Brazil have until recently remained practically unknown, but in the last few years preliminary reports have appeared on such tribes as the Guahibo, Piaroa, Sanema, and Shiriana, among at least some groups of which agriculture is absent or virtually so (Wilbert, 1958b, 1966; Migliazza, 1964). These reports indicate that fundamental field research is still possible, and, as a matter of fact, three of my own graduate students have been working recently among the Guahibo (Morey, 1967).

  1. Amazonian Hunters

As regards the “interior marginal” tribes of southwestern Amazonia, we already possess substantial accounts by Holmberg (1950) on the Siriono and by Levi-Strauss (1936, 1 940) on the Bororo and Nambicuara, and the Guato have been well described by Schmidt (1905, 1942). I understand, too, that new material of high quality has recently become available on the Guayaki (Levi-Strauss, this volume; cf. Bertoni, 1941; Vellard, 1939).

  1. East Brazilian Hunters

The non-agricultural peoples of eastern Brazil, of whom the Botocudo, Caingang, and Puri are perhaps the best known, are today either extinct or shattered. The last major ethnographic report from this area, that by Henry (19.41) on the Aweikoma, makes it clear that fundamental field work is no longer possible here.

  1. Gran Chaco

A fair amount of information, though not always of the highest quality, is available on the tribes of the Gran Chaco-the Chamacoco, Choroti, Mataco, Mocovi, Toba, etc. Though good field work is certainly still possible in this area, the aboriginal situation has altered substantially through the adoption of agriculture, domestic animals, and an increasingly ‘sedentary mode of life (Baldus, 1931; Boggiani, 1894; Karsten, 1932; Metraux, 1937; Norden- skii:ild, 1919).

  1. Ona

The Indians of the Argentinian steppe, like those of the Plains of North America, early adopted the horse and are thus as mounted hunters excluded from our consideration. However, their kinsmen in the extreme south, the Ona, never became dependent on the horse and thus qualify for inclusion. We are fortunate in having a fairly recent and extremely comprehensive account of the Ona by Gusinde (1931 -39) .

  1. Fuegians

Finally, we must list the Indians of the extreme southwestern tip of South America-the Alacaluf, Chono, and Yahgan-for they lived a migratory life despite their primary dependence on sheilfish, aquatic mammals, and other products ofthe sea. The Chono are ill described and extinct. There is fuller information on the Alacaluf and especially the Yahgan, but both tribes are now reduced to tiny remnants, and the prospects of obtaining further fundamental descriptive material are therefore extremely slim.

By way of concluding this rapid survey of the world’s hunting and gathering peoples who have survived long enough to be studied by modern anthropologists, I should like to stress the fact that their largest concentration is found in aboriginal North America, where they have been studied mainly by American and Canadian anthropologists. The publications of these scholars are thus especially crucial for any general or comparative study of the material culture and subsistence economy of hunting and gathering peoples.

Second, the great majority of these North American societies are characterized by social systems of a bilateral rather than unilineal type, and indeed constitute well over half of the known bilateral societies of the entire world, so that they are equally crucial for the comparative study of this rapidly expanding aspect of social organization. It is thus entirely understandable why people like Damas, Dunning, and Helm have recently been concentrating on this range of problems in their North American researches.

Finally, owing to the influence of Lewis Henry Morgan, North American ethnographers have paid special attention to kinship terminology and its social concomitants for several generations. As a consequence, we possess reasonably adequate kinship information on perhaps 85 per cent of all the North American tribes whom the white man has ever encountered, as contrasted with a comparable figure of not more than 15 per cent for the known tribes of Africa. Thus, future field research in this continent, despite the steadily decreasing range of its geographical possibilities, can be based on an exceptionally solid background of past accomplishment.

Part II: Ecology and Economics

3. The “‘Hunting” ’ Economies of the Tropical Forest Zone of South America: An Attempt at Historical Perspective

Donald W. Lathrap

The Amazon Basin contains the largest block of tropical rain forest in the world. The patterns of human utilization ofthis region and the history of how these patterns evolved through time is of considerable interest as a major chapter in the progressive expansion of human populations, and in the adaptation of such populations to diverse ecological settings.[1]

At the time of the first European contact, the Amazon Basin showed a considerable range of cultural pattern in terms of size and complexity of social units, complexity of material culture, and even in basic patterns of subsistence activity. The flood plain of the main stream of the Amazon and of its major tributaries sustained large, sedentary populations engaged in intensive root-crop farming combined with fishing and the hunting of aquatic mammals and reptiles. The exact size and complexity of sociopolitical units attained by these riverine societies, as of A.D. 1 500, continues to be a matter of controversy since all such societies were very early disrupted by the effects of slave raiding, missiOnization, and diseases introduced by the Europeans. While it is clear that none of these societies rivaled the kind of sociopolitical unit which had been typical of the central Andes for the last 500 to 1,000 years before Columbus, it is equally clear from the various reviews of the early sources (Metraux, 1948a; Nimuendaju, 1952a; Palmatary, 1960, 1965; Rowe, 1 952) that such riverine groups were far larger and more complex than the presently surviving remnant groups representing tropical forest culture. An even greater discrepancy exists between our knowledge of the contact period riverine societies of the Amazon Basin and the generalized notion of “tropical forest culture” or “tropical forest stage” which has been developed by Meggers and Evans (Meggers, 1954, 1957a, 1957b; Meggers and Evans, 1956).

The demographic and cultural situation in the slightly elevated regions between the major rivers was in sharp contrast to that of the flood plains. The interfluve areas were characterized by scant populations organized into small, widely dispersed and mobile social units. These groups typically showed simple and relatively unproductive agricultural systems and relied heavily on the hunting of terrestrial and arboreal game for their sustenance. Unlike the groups inhabiting the more accessible and economically more valuable floodplains, the simple, dispersed Indian societies of the interfluve have in many instances survived until recently without major modifications in culture (Carneiro, 1 964, p. 9).[2]

The presentation of a full, composite ethnographic picture of all such simple non-riverine groups is far beyond the scope of this paper, but certain key points in the above generalizations should be documented. In the area with which I am most familiar, the Peruvian Montana, groups such as the Cashibo, Ama-huaca, Remo, and Mayoruna are completely characteristic. A recent comparison of the agricultural practices of the Isconahua group of Remo, with those of the Shipibo, a riverine Panoan group, has shown the relative inefficiency of the Remo system and suggested that this inefficiency was an adjustment to the poorer agricultural potential of the interfluvial zone, the Remo’s accustomed habitat (Momsen, 1 964, pp. 76–77). Carneiro’s excellent discussion of Amahuaca agriculture indicates equally rudimentary practices. For instance, he says: “At best, the Amahuaca are reluctant weeders” (1964, p. 14). The Shipibo on the other hand appear to be rather systematic and finicky on these matters.

Points which are crucial to the present discussion are the great importance which the hunting of terrestrial and arboreal game had for these interfluvial groups, and the degree to which the dependence on such game affected all other aspects of their culture. Carneiro has repeatedly demonstrated (1960, 1961, 1964) that it is not agricultural potential which directly controls the size and mobility of tropical forest social groups. In the Amahuaca case he gives a dear demonstration that the agriculture system, rudimentary as it is, is still capable of producing far more vege”table foods than the people are able to use (1964, pp. 1718). It is the absence of significant aquatic resources (1964, p. IO) leading to a dependence on terrestrial hunting for the protein necessary to the diet which largely controls the nature of the social group. Carneiro is worth quoting at length on this point:

First of all, it should be kept in mind that the Amahuaca are still hunters almost as much as they are gardeners. Consequently, considerations having to do with the availability of game are very important to them. Even though three or four families settled in one locale for a year do not hunt out the game animals within the usual hunting radius of the settlement, they nevertheless probably make a noticeable inroad in their numbers. By the end of a year it has become necessary to walk farther to find game, and this is a decided inconvenience. If this inconvenience were counterbalanced by an equal or greater inconvenience in building a new settlement each year, the Amahuaca might not move so frequently. But Amahuaca houses can be built in three days and Amahuaca families are independent units perfectly free to pick up and move when and where they want to. Thus the Amahuaca, in deciding to move their settlements, do not have to overcome the inertia that would face a society which had a large village, substantial houses, and centralized political authority. In short, since the resistance to moving is small, the forces required to bring it about can likewise be small (1964, p. 16).

Holmberg (1950) paints a far bleaker picture of the results of the necessary dependence on terrestrial hunting by such interfluvial groups. While probably somewhat extreme in terms of the average situation for such groups throughout the Amazon Basin, his description of hunger and life. among the Siriono still stands as the most detailed coverage of the kind of economic patterns shared by such groups.

A final factor which characterized the primitive groups on the interriverine uplands is that they were typically denied access to the flood plain environment by the warlike activities of the riverine groups. It was the continual threat of slave raiding, head taking, and in some instances cannibalism which held the non-riverine groups in their less favorable environment. Nimuendaju’s discussion of the territory of the Tukuna illustrates the point admirably:

The Tukuna did not then inhabit the banks of the Amazon-Solimoes, for fear of the Omagua, who occupied the islands of that river throughout its course in Tukuna territory and even far beyond, while the banks of the Putumayo-I<;:i were held by Aruak tribes-the Mariate, Yumina, and Pase (1952b, p. 8).

The most feared enemies of the ancient Tukuna -one might say their only enemies-were the Omigua. [The Omigua] attacked by canoe, with great spears. They captured many: some of them were sacrificed to their idols and the rest served as slaves in their husbandary (1952b, p. 65).

Marcoy’s general discussion of the attitude of the riverine Shipibo toward the non-riverine Cashibb is equally telling, and his description of the crucifixion of a Cashibo unfortunate enough to fall into Shipibo hands is even more graphic (1873, vol. 2, pp. 162–63).

Concerning the relation between the riverine Arawakan and Tucanoan tribes of southeastern Colombia and the interfluvial Macu, Metraux says:

The Arawakan and Tucanoan tribes of the upper Rio Negro, Caiari-Uaupes, and Tiquie Rivers have since time immemorial waged merciless war against the Macu, whom they enslave or reduce to serfdom. Some small groups of Macu come to work for the sedentary Uanana and Desana and, after a few months, disappear again into the bush. The Tucano of the Tiquie River subjected a large group of Macu to their rule, but on the slightest suspicion of sorcery, they were prompt to attack them and sell their captives to the Whites (I948b, p. 866).

Carneiro also alludes to the constant hostility between the Amahuaca and their riverine neighbors as a factor tending to maintain their small, highly mobile, social units (1964, p. 16).

It is the culturally simple groups of the slightly elevated interfluvial regions of the Amazon Basin who are of possible interest to this symposium. It is dear from the foregoing discussion that they are typically dependent on the hunting of terrestrial and arboreal game for the essential protein complement of their diet, and arguments have been presented suggesting that the hunting practices of such people are a more important factor in conditioning their total way of life than are their relatively rudimentary and inefficient agricultural systems.

There has been a tendency to treat these people as unmodified representatives of a very early and primitive stage of tropical forest culture. Thus in a recent description of the Isconahua group of Remo, we find the authors, Whiton, Greene, and Momsen saying: “Their culture was that of incipient tropical forest slash-and-burn and close to neolithic man” (1964, p. 123). The suggestion has also been made that some, perhaps most, of these peoples represent groups of primitive hunters only slightly modified by trait-unit diffusion from their more advanced agricultural neighbors, and thus show’ a large degree of cultural continuity from a pre-Neolithic period or stage. Metraux is explicit about such an interpretation (I948b, p. 86 I), while such a view appears implicit in most of Steward’s discussion of the same problem (I948b, pp. 896–99). Some such view would also appear to be implicit in Needham’s emphasis on the anomaly of a system of asymmetric prescriptive alliances occurring in a “small-scale, hunting and collectingsociety” such as the Siriono (196 I, p. 252).

There is reason to question both of these interpretations and to suggest that many, perhaps all, of these simpler and less sedentary groups are the degraded descendants of peoples who at one time maintained an advanced form of tropical forest culture. In the remainder of this paper I hope to develop a simplified model of the historical background of such peoples.

I will assume that man entered South America with an economy oriented toward the hunting of big game, the large grazing animals typical of the grasslands and savannah. If the earliest inhabitants of South America indeed did function as predators on the grazing fauna, it is unlikely that they initially would have extended their range beyond that of the major element of their subsistence.

Such hunting groups appear to have entered South America before 15,000 B.C. (Rouse and Cruxent, I963a, p. 537) and to have extended their range to the Straits of Magellen by 9000 b.c. (Lanning and Hammel, 1961’I?’ 147). There is insufficient space to review the distribution of Late Pleistocene and Early Post-Pleistocene sites in South America, but the available evidence is compatible with the hypothesis that the earlier occupations . were confined to areas of relatively open vegetation: seasonal grasslands, high grasslands, open thorn forest, etc.

For groups with a technology and social system well adapted to cooperative hunting on the gras:;land, the tropical forest would not have been an attractive environment for permanent occupation or intensive utilization. The more common forms of game, both animals and birds, had as their typical habitat the upper levels of the forest canopy or were semi-aquatic to aquatic. In terms of mammalian fauna, the floor of the tropical forest away from the rivers supported a poverty of species and a low density of individuals. Gilmore comments on the generally unfavorable nature of the tropical forest as a hunting territory (1950, p. 354), and information on groups such as the Sirion6 suggests that even for a people with a technology specifically adapted to the problems of hunting within the tropical forest, the tropical forest is far from prime hunting territory.

One would expect that most of the areas of more open vegetation including the east Brazil highlands, the llanos of Venezuela and Colombia, and parts of the Guiana highlands would have been occupied before groups with a specifically hunting orientation would have attempted a direct penetration of the central block of tropical forest within the Amazon Basin. Indeed, there is no direct archeological evidence that such penetration ever took place. The earliest demonstrable penetration of the tropical forest of the Amazon Basin is oriented to the far richer riverine resources.

The more astute students of tropical forest culture, such as Sauer, Lowie, Steward, and Goldm::m, have all observed that tropical forest culture is less an adaptation to the forest as such, than to the riverine environment within the tropical forest. Since this point has been consistently ignored by those who tend to see the tropical forest as a uniform environment, it may be well to offer extensive documentation. Sauer’s emphasis on the importance of riverine resources suffuses his whole discussion of the origin of the tropical forest pattern of cultivation (1952, pp. 40–49). And Lowie comments:

At the core of the area the diagnostic features are: the cultivation of tropical root crops, especially bitter manioc; effective river craft; the use of hammocks as beds; and the manufacture of pottery.

The very wide distribution of certain traits in the area is correlated with navigation. Thanks to their mobility, the canoeing tribes were able to maintain themselves in the midst of boatless populations, to travel with ease over periodically inundated tracts, and to diffuse their arts and customs over enormous distances (1948, p. I).

Steward makes the same points even more strongly:

The distribution of culture elements and complexes reveals at least one broad pattern. The basic Tropical Forest cultures occur mainly in the areas accessible by water routes, both the coast and the great rivers The inference is clear that what is thought of as a typical Tropical Forest or selvan culture—a developed agriculture and a technology manifest in twilled and woven baskets, loom weaving, cotton hammocks, ceramics, and other material traits-flowed along the coast and up the main waterways stopping where streams were less navigable ... (I948b, p. 883).

Goldman’s statement ofthe Cubeos’ view of their ecological setting has an almost poetic force:

The river forms a wider community of related sibs. With some exceptions the sibs have aligned themselves along a river on the basis of degree of closeness and of rank . Sibs that have segmented from a parental sib ordinarily occupy an adjacent site. The river is the most important territory. It is a highway and a link between related sibs, the source of ancestral power, and the economic zone of the men, fish being the main source of animal protein. Even most land animals are hunted along the river banks.

The orientation of the Cubeo is toward the river and not toward the forest. Whereas the forest is undifferentiated terrain, the rivers are known to every turn and outcrop of rock or other feature. The river is the source of the ancestral powers, of benefits as well as of dangers. The forest is a source mainly of dangers (1963, p. 44).

The advantages of the riverine environment over the slightly elevated interriverine zone include not only the greater availability of animal protein, but also a much higher agricultural potential. Sauer discusses the difference between the two zones in terms of land form (1950, p. 324), while Sternberg stresses the very real difference in agricultural potential of the lands lying within the Amazon Basin:

Or take the prospect of an unlimited extent of rich crop land. The luxurious forest which mantles most of the region has commonly been interpreted as an unconditional promise of inexhaustible fertility. The very real productivity of bottomlands, periodically flooded and rejuvenated with silt-laden waters, has thus been ascribed to the enormous area corresponding to the generally acid and poor soils of the terras firmas, or uplands (1964, p. 13).

Thus it cannot be too strongly emphasized that the deep alluvial soils within the active flood plains have a far greater agricultural potential than the laterized and heavily eroded soils on the slightly elevated uplands back from the rivers. My Shipibo friends inform me that such deep, alluvial soils in certain parts of the central U cayali flood plain will support continuous root-crop agriculture for up to fifteen or twenty years.

Sauer has stressed that the tropical forest agricultural pattern is essentially one of carbohydrate production, with the fat and protein requirements of the diet supplied by a technologically sophisticated utilization of the prodigious fish, aquatic mammal, and aquatic reptile resources of the flood plain (I9YZ). Evolved tropical forest culture involves notjust developed root-crop agriculture, but a developed set of fishing practices, including fish poisoning and effective water craft. Archeological manifestations suggesting all of these accomplishments appear early along the major rivers within the Amazon Basin.

Along the flood plain of the central Ucayali River in the Montana of eastern Peru we find remains of such cultures as early as 2000 B.C. (Lathrap, 1958; 1962; I965a, PP. 797–98; I965b, p. 12). The elaborate and technologically sophisticated ceramic style designated Early Tutishcainyo cannot be derived from any early central Andean ceramic style. It does, however, show certain basic similarities to the earliest examples of tropical forest culture known from the flood plains of the Lower Orinoco in Venezuela, suggesting that both ceramic styles may have derived from some common source along the intermediate network of rivers (Cruxent and Rouse, 1959; Lathrap, 1963; Rouse and Cruxent, I 963b). The geographical setting in {he central Ucayali presents the grossest kinds of obstacles to a meaningful estimate of settlement pattern and density of occupation for the earliest complexes. Given the degree to which jungle litter tends to camouflage all land surfaces, given the fact that the activity of the more numerous colonial ants tends to bury all ancient land surfaces under three to six inches of culturally sterile deposit, and most importantly given the destructive power of the continually meandering river which erases all traces of early settlements not located at the precise edge of the flood plain, the available evidence suggests that Early Tutishcainyc already represents a dense occupation of the suitable ecological niche.

In the course of the long cultural sequence which follows Early Tutishcainyo on the central Ucayali there exists evidence for at least three instances of complete ethnic replacement; two of the intrusive ceramic styles, so strongly suggesting the entrance of new ethnic groups, are clearly derived from the central Amazon. The long archeological sequence on the central Ucayali fully substantiates the picture of cultural development which Steward sketched on the basis of ethnographic distribution alone:

Culturally, the Chuncha (the Indians of the Montana) belong with the Tropical Forest peoples. They appear to represent a series of migratory waves that had spent their force against the barrier of the Andes, where representatives of many widely distributed linguistic families ... subsided into comparative isolation (X948a, p. 507).

A series of waves of migration, basically in an upstream direction, is also suggested by the distribution of languages in two of the more fully studied South American linguistic stocks, Macro-Arawak and Tupf-Guarani. Noble (1965) has presented a detailed reconstruction of the proto-language of Macro-Arawakan including estimates of the degree of relationship among the various languages within Macro-Arawakan. Rodrigues(I958) has presented a brief statement as to the classification of the various Tupi-Guarani languages. A full treatment of the significance of these classifications for the understanding of South American culture history is beyond the scope of this paper, but I am at present engaged in preparing a manuscript which will attempt to evaluate the distribution of the languages of these two stocks in terms of past population movements. In attempting to summarize what is a complex picture, one might say that the more divergent, and presumably more anciently dispersed branches of Macro-Arawakan tend to be near the headwaters of the major western tributaries of the Amazon, while the more closely related Maipuran languages within Arawakan tend to dominate the mainstreams of the western segment of the Amazon, the Rio Negro, and the Orinoco. Likewise the more divergent branches of Tupi-Guarani are ranged near the headwaters of the various major southern tributaries of the Amazon, while the more closely related languages within the family dominate the broader flood plains of the lower reaches of these rivers, much of the mainstream of the Lower Amazon, the Atlantic coast of Brazil, and much of the flood plain of the Parana. The most economical explanation for these observations (especially when they are coupled with Noble’s demonstration of an ultimate relationship betweenProto-Arawakan and ProtoTupi-Guarani, relating to a period before the dispersal of these two stocks) would be that the proto-languages of the two stocks occupied adjacent stretches of the Central Amazon between 3000 and 2000 B.C. with Proto-Arawakan on the upstream side. From this hearth, colonization spread outward, mainly in an upstream direction, along all available waterways. Certain marked congruences between the distribution of ceramic styles and the distributions of particular linguistic families or subfamilies suggest the possibility of indicating the archeological concomitants of these major migrations demonstrable on linguistic grounds, and I hope fully to explore this possibility in the aforementioned forthcoming book.[3]

The pattern of outward migration suggested by all of these converging lines of evidence could best be explained by intense and continuing population pressures of the flood plain of the Central Amazon, the most favorable environment for the support of tropical forest culture. Such continuous expansion by groups moving out to colonize further areas of flood plain progressively pushed smaller or militarily weaker groups farther upstream or off the flood plains entirely. In presenting this view I am, of course, in complete accord with Vayda’s argument that warfare among tropical forest farmers has profound economic and demographic effects even though it may have a religious or recreational rationale (Vayda, 196rb).

This extreme and continuing competition for territory is understandable, if one realizes the limited amount of flood plain-the only ecological niche really suitable to tropical forest culture-available in the Amazon Basin. Sternberg summarizes the point nicely:

It has been seen how the uplands are generally poor, and it must be granted that they occupy the major part of the Amazon region, comprising more than a million square miles and leaving only a slight fraction of the area to the fertile vtirzeas, or floodplains. Nevertheless, one is dealing with lands of continental dimensions and this small fraction represents no insignificant area: the vtirzeas, rejuvenated every year by the silt brought down by the river, occupy some twenty-five thousand square miles ... (1964, p. 324).

Considering that these processes have been continuous from about 2500 B.C. up to the time ofthe contact, it is clear that a large number of ethnic groups were forced off the flood plains into less favorable environments.

Most of the primitive groups inhabiting the tropical forest uplands away from the major flood plains can be interpreted as the wreckage of evolved agricultural societies forced into an environment unsuitable to the basic economic pattern. Deprived of the riverine resources, such groups had to rely on the hunting offorest game to provide the protein and fat essential to the diet. A more intense orientation to hunting the relatively scarce game available led to more nomadism, a decline in agricultural productivity, and a still greater dependence on wild food.

I am convinced that this picture accounts for all the simpler groups within Tupf-Guarani, Arawakan, and Panoan linguistic stocks.

As we gain more knowledge the interpretation may be extendable to other groups, whose cultural and linguistic affiliations are less clear, such as the Tukuna, Mura, and Macu. The speed with which this kind of cultural devolution can take place is suggested by the relatively minor linguistic differences between the riverine Panoans such as the Shipibo and Conibo and the very “primitive” interfluvial Panoan groups such as the Cashibo and Remo.

While the “hunting” cultures of the tropical forest zone of South America offer highly explicit examples of the cultural and demographic effects of a dependence on hunting in an area where hunting is neither profitable or easy, they probably instruct us not at all about the nature of pre-Neolithic hunting cultures.

4. What Hunters Do for a Living, or, How to Make Out on Scarce Resources

Richard B. Lee

The current anthropological view of huntergatherer subsistence rests on two questionable assumptions. First is the notion that these peoples are primarily dependent on the hunting of game animals, and second is the assumption that their way of life is generally a precarious and arduous struggle for existence.

Recent data on living hunter-gatherers (Meggitt, 1964b; Service, 1966; and papers in this volume) show a radically different picture. We have learned that in many societies, plant and marine resources are far more important than are game animals in the diet. More important, it is becoming clear that, with a few conspicuous exceptions, the hunter-gatherer subsistence base is at least routine and reliable and at best surprisingly abundant. Anthropologists have consistently tended to underestimate the viability of even those “marginal isolates” of hunting peoples that have been available to ethnographers.

The purpose of this paper is to analyze the food getting activities of one such “marginal” people, the!Kung Bushmen of the Kalahari Desert. Three related questions are posed: How do the Bushmen make a living? How easy or difficult is it for them to do this? What kinds of evidence are necessary to measure and evaluate the precariousness or security of a way of life? And after the relevant data are presented, two further questions are asked: What makes this security of life possible? To what extent are the Bushmen typical of hunter-gatherers in general?

Bushman Subsistence

The!Kung Bushmen of Botswana are an apt case for analysis.[4] They inhabit the semi-arid northwest region of the Kalahari Desert. With drought every second or third year, this is, by any account, a marginal environment for human habitation. In fact, it is precisely the unattractiveness of their homeland that has kept the!Kung isolated from extensive contact with their agricultural and pastoral neighbors.

Field work was carried out in the Dobe area, a line of eight permanent waterholes near the South-West Africa border and 125 miles south of the Okavango River. The population of the Dobe area consists of 466 Bushmen, including 379 permanent residents living in independent camps or associated with Bantu cattle posts, as well as 87 seasonal visitors. The Bushmen share the area with some 340 Bantu pastoralists largely of the Herero and Tswana tribes. The ethnographic present refers to the period of field work: October, I 963-January, 1965.

Table 1. Numbers and Distribution of Resident Bushmen and Bantu by Waterhole{1}

r-b-richard-b-lee-irven-devore-man-the-hunter-26.jpg

The Bushmen living in independent camps lack firearms, livestock, and agriculture. Apart from, occasional visits to the fIerero for milk, these!Kung are entirely dependent upon hunting and gathering for their subsistence. Politically they are under the nominal authority of the Tswana headman, although they pay no taxes and receive very few government services. European presence amounts to one overnight government patrol every six to eight weeks. Although Dobe-area!Kung have had some contact with outsiders since the 1880’S, the majority of them continue to hunt and gather because there is no viable alternative locally available to them.[5]

Each of the fourteen independent camps is associated with one of the permanent waterholes. During the dry season (May-October) the entire population is clustered around these wells. Table 1 shows the numbers at each well at the end of the 1964 dry season. Two wells had no camp resident and one large well supported five camps. The number of camps at each well and the size of each camp changed frequently during the course of the year. The “camp” is an open aggregate of cooperating persons which changes in size and composition from day to day. Therefore, I have avoided the term “band” in describing the!Kung Bushman living groups.[6]



Each waterhole has a hinterland lying within a six-mile radius which is regularly exploited for vegetable and animal foods. These areas are not territories in the zoological sense, since they are not defended against outsiders. Rather they constitute the resources that lie within a convenient walking distance of a waterhole. The camp is a self-sufficient subsistence unit. The members move out each day to hunt and gather, and return in the evening to pool the collected foods in such a way that every person present receives an equitable share. Trade in foodstuffs between camps is minimal; personnel do move freely from camp to camp, however. The net effect is of a population constantly in motion. On the average, an individual spends a third of his time living only with close relatives, a third visiting other camps, and a third entertaining visitors from other camps.

Because of the strong emphasis on sharing, and the frequency of movement, surplus accumulation of storable plant foods and dried meat is kcpt to a minimum. There is rarely more than two or three days’ supply of food on hand in a camp at any time. The result of this lack of surplus is that a constant subsistence effort must be maintained throughout the year. Unlike agriculturalists who work hard during the planting and harvesting seasons and undergo “seasonal unemployment” for several months, the Bushmen hunter-gatherers collect food every third or fourth day throughout the year.

Table 2 The Bushman Annual Round

r-b-richard-b-lee-irven-devore-man-the-hunter-27.jpg

Vegetable foods comprise from 60–80 per cent of the total diet by weight, and collecting involves two or three days of work per woman per week. The men also collect plants and small animals but their major contribution to the diet is the hunting of medium and large game. The men are conscientious but not particularly successful hunters; although men’s and women’s work input is roughly equivalent in terms of man-day of effort, the women provide two to three times as much food by weight as the men.

Table 2 summarizes the seasonal activity cycle observed among the Dobe-area!Kung in 1964. For the greater part of the year, food is locally abundant and easily collected. It is only during the end of the dry season in September and October, when desirable foods have been eaten out in the immediate vicinity of the waterholes that the people have to plan longer hikes of 10–15 miles and carry their own water to those areas where the mongongo nut is still available. The important point is that food is a constant, but distance required to reach food is a variable; it is short in the summer, fall, and early winter, and reaches its maximum in the spring.

This analysis attempts to provide quantitative measures of subsistence status including data on the following topics: abundance and variety of resources, diet selectivity, range size and population density, the composition of the work force, the ratio of work to leisure time, and the caloric and protein levels in the diet. The value of quantitative data is that they can be used comparatively and also may be useful in archeological reconstruction. In addition, one can avoid the pitfalls of subjective and qualitative impressions; for example, statements about food “anxiety” have proven to be difficult to generalize across cultures (see Holmberg, 1950; and Needham’s critique, 1954).

Abundance and VarieD’ of Resources

It is impossible to define “abundance” of resources absolutely. However, one index of relative abundance is whether or not a population exhausts all the food available from a given area. By this criterion, the habitat of the Dobe-area Bushmen is abundant in naturally occurring foods. By far the most important food is the Mongongo (mangetti) nut (Ricinodendron rautanenii Schinz). Although tens of thousands of pounds of these nuts are harvested and eaten each year, thousands more rot on the ground each year for want of picking.

The mongongo nut, because of its abundance and reliability, alone accounts for 50 per cent of the vegetable diet by weight. In this respect it resembles a cultivated staple crop such as maize or rice. Nutritionally it is even more remarkable, for it contains five times the calories and ten times the proteins per cooked unit of the cereal crops. The average daily percapita consumption of 300 nuts yields about 1,260 calories and 56 grams of protein. This modest portion, weighing only about 7.5 ounces, contains the caloric equivalent of 2.5 pounds of cooked rice and the protein equivalent of 14 ounces of lean beef (Vlatt and Merrill, 1963).

Furthermore the mongongo nut is drought resistant and it will still be abundant in the dry years when cultivated crops may fail. The extremely hard outer shell protects the inner kernel from rot and allows the nuts to be harvested for up to twelve months after they have fallen to the ground. A diet based on mongongo nuts is in fact more reliable than one based on cultivated foods, and it is not surprising, therefore, that when a Bushman was asked why he hadn’t taken to agriculture he replied: “Why should we plant, when there are so many mongongo nuts in the world?”

Apart from the mongongo, the Bushmen have available 84 other species of edible food plants, including 29 species of fruits, berries, and melons and 30 species of roots and bulbs. The existence of this variety allows for a wide range of alternatives in subsistence strategy. During the summer months the Bushmen have no problem other than to choose among the tastiest and most easily collected foods. Many species, which are quite edible but less attractive, are bypassed, so that gathering never exhausts all the available plant foods of an area. During the dry season the diet becomes much more eclectic and the many species of roots, bulbs, and edible resins make an important contribution. It is this broad base that provides an essential margin of safety during the end of the dry season when the mongongo nut forests are difficult to reach. In addition, it is likely that these rarely utilized species provide important nutritional and mineral trace elements that may be lacking in the more popular foods.

Table 3. !Kung Bushman Plant Foods

r-b-richard-b-lee-irven-devore-man-the-hunter-28.jpg
Diet Selectivity

If the Bushmen were living close to the “starvation” level, then one would expect them to exploit every available source of nutrition. That their life is well above this level is indicated by the data in Table 3. Here all the edible plant species are arranged in classes according to the frequency with which they were observed to be eaten. It should be noted, that although there are some 85 species available, about 90 per cent of the vegetable diet by weight is drawn from only 23 species. In other words, 75 per cent of the listed species provide only 10 per cent of the food value.

In their meat-eating habits, the Bushmen show a similar selectivity. Of the 223 local species of animals known and named by the Bushmen, 54 species are classified as edible, and of these only 17 species were hunted on a regular basis.[7] Only a handful of the dozens of edible species of small mammals, birds, reptiles, and insects that occur locally are regarded as food. Such animals as rodents, snakes, lizards, termites, and grasshoppers, which in the literature are included in the Bushman dietary (Schapera, 1930), are despised by the Bushmen of the Dobe area.

Range Size and Population Density

The necessity to travel long distances, the high frequency of moves, and the maintenance of populations at low densities are also features commonly associated with the hunting and gathering way of life. Density estimates for hunters in western North America and Australia have ranged from 3 persons/square mile to as low as I personJ 100 square miles (Kroeber, “939; Radcliffe-Brown, 1930). In 1963–65, the resident and visiting Bushmen were observed to utilize an area of about 1,000 square mil.es during the course of the annual round for an effective population density of 41 persons/loo square miles. Within this area, however, the amount of ground covered by members of an individual camp was surprisingly small. A day’s round-trip of twelve miles serves to define a “core” area six miles in radius surrounding each water point. By fanning out in all directions from their well, the members of a camp can gain access to the food resources of well over 100 square miles of territory within a two-hour hike. Except for a few weeks each year, areas lying beyond this six-mile radius are rarely utilized, even though they are no less rich in plants and game than are the core areas.

Although the Bushmen move their camps frequently (five or six times a year) they do not move them very far. A rainy season camp in the nut forests is rarely more than ten or twelve miles from the home waterhole, and often new campsites are occupied only a few hundred yards away from the previous one. They do not undertake long marches of 30 to 100 miles to get food, since this task can be readily fulfilled within a day’s walk of home base. When such long marches do occur they are invariably for visiting, trading, and marriage arrangements, and should not be confused with the normal routine of subsistence.

Demographic Factors

Another indicator of the harshness of a way of life is the age at which people die. Ever since Hobbes characterized life in the state of nature as “nasty, brutish and short, “ the assumption has been that hunting and gathering is so rigorous that members of such societies are rapidly worn out and meet an early death. Silberbauer, for example, says of the Gwi Bushmen of the central Kalahari that “life expectancy ... is difficult to calculate, but I do not believe that many live beyond 45” (1965, p. 17). And Coon has said of the hunters in general:

The practice of abandoning the hopelessly ill and aged has been observed in many parts of the world. It is always done by people living in poor environments where it is necessary to move about frequently to obtain food, where food is scarce, and transportation difficult Among peoples who are forced to live in this way the oldest generation, the generation of individuals who have passed their physical peak is reduced in numbers and influence. There is no body of elders to hand on tradition and control the affairs of younger men and women, and no formal system of age grading (1948, p. 55).

The!Kung Bushmen of the Dobe area flatly contradict this view. In a total population of 466, no fewer than 46 individuals (17 men and 29 women) were determined to be over 60 years of age, a proportion that compares favorably to the percentage of elderly in industralized populations.

The aged hold a respected position in Bushman society and are the effective leaders of the camps. Senilicide is extremely rare. Long after their productive years have passed, the old people are fed and cared for by their children and grandchildren. The blind, the senile, and the crippled are respected for the special ritual and technical skills they possess. For instance, the four elders at!gose waterhole were totally or partially blind, but this handicap did not prevent their active participation in decisionmaking and ritual curing.

Another significant feature of the composition of the work force is the late assumption of adult responsibility by the adolescents. Young people are not expected to provide food regularly until they are married. Girls typically marry between the ages of 15 and 20, and boys about five years later, so that it is not unusual to find healthy, active teenagers visiting from camp to camp while their older relatives provide food for them.

As a result, the people in the age group 2060 support a surprisingly large percentage of non-productive young and old people. About 40 per cent of the population in camps contribute little to the food supplies. This allocation of work to young and middle-aged adults allows for a relatively carefree childhood and adolescence and a relatively unstrenuous old age.

Leisure and Work

Another important index of ease or difficulty of subsistence is the amount of time devoted to the food quest.[8] Hunting has usually been regarded by social scientists as a way of life in which merely keeping alive is so formidable a task that members of such societies lack the leisure time necessary to “build culture.”[9] The!Kung Bushmen would appear to conform to the rule, for as Lorna Marshall says:

It is vividly apparent that among the!Kung Bushmen, ethos, or “the spirit which actuates manners and customs,” is survival. Their time and energies are almost wholly given to this task, for life in their environment requires that they spend their days mainly in procuring food (1965, p. 247).

It is certainly true that getting food is the most important single activity in Bushman life. However this statement would apply equally well to small-scale agricultural and pastoral societies too. How much time is actually devoted to the food quest is fortunately an empirical question. And an analysis of the work effort of the Dobe Bushmen shows some unexpected results. From July 6 to August 2, 1964, I recorded all the daily activities of the Bushmen living at the Dobe waterhole. Because of the coming and going of visitors, the camp population fluctuated in size day by day, from a low of 23 to a high of 40, with a mean of 31.8 persons. Each day some of the adult members of the camp went out to hunt and/or gather while others stayed home or went visiting. The daily recording of all personnel on hand made it. possible to calculate the number of man-days of work as a percentage of total number of man-days of consumption.


Although the Bushmen do not organize their activities on the basis of a seven-day week, I have divided the data this way to make them more intelligible. The work-week was calculated to show how many days out of seven each adult spent in subsistence activities (Table 4, Column 7). Week II has been eliminated from the totals since the investigator contributed food. In week I, the people spent an average of 2.3 days in subsistence activities, in week III, 1.9 days, and in week IV, 3.2 days. In all, the adults of the Dobe camp worked about two and a half days a week. Since the average working day was about six hours long, the fact emerges that!Kung Bushmen of Dobe, despite their harsh environment, devote from twelve to nineteen hours a week to getting food. Even the hardest working individual in the camp, a man named ¥-oma who went out hunting on sixteen of the 28 days, spent a maximum of 32 hours a week in the food quest.

Because the Bushmen do not amass a surplus of foods, there are no seasons of exceptionally intensive activities such as planting and harvesting, and no seasons of unemployment. The level of work observed is an accurate reflection of the effort required to meet the immediate caloric needs of the group. This work diary covers the mid-winter dry season, a period when food is neither at its most plentiful nor abits scarcest levels, and the diary documents the transition from better to worse conditions (see Table 2). During the fourth week the gatherers were making overnight trips to camps in the mongongo nut forests seven to ten miles distant from the waterhole. These longer trips account for the rise in the level of work, from twelve or thirteen to nineteen hours per week.

Iffood getting occupies such a small proportion of a Bushman’s waking hours, then how do people allocate their time? A woman gathers on one day enough food to feed her family for three days, and spends the rest of her time resting in camp, doing embroidery, visiting other camps, or entertaining visitors from other camps. For each day at home, kitchen routines, such as cooking, nut cracking, collecting firewood, and fetching water, occupy one to three hours of her time. This rhythm of steady work and steady leisure is maintained throughout the year.

The hunters tend to work more frequently than the women, but their schedule is uneven. It is not unusual for a man to hunt avidly for a week and then do no hunting at all for two or three weeks. Since hunting is an unpredictable business and subject to magical control, hunters sometimes experience a run of bad luck and stop hunting for a month or longer. During these periods, visiting, entertaining, and especially dancing are the primary activities of men. (Unlike the Hadza, gambling is only a minor leisure activity.)

The trance-dance is the focus of Bushman ritual life; over 40 per cent of the men have trained as trance-performers and regularly enter trance during the course of the all-night dances. At some camps, trance-dances occur as frequently as two or three times a week and those who have entered trances the night before rarely go out hunting the following day. Accounts of Bushman trance performances have been published in Lorna Marshall (1962) and Lee (1968). In a camp with five or more hunters, there are usually two or three who are actively hunting and several others who are inactive. The net effect is to phase the hunting and non-hunting so that a fairly steady supply of meat is brought into a camp.

Caloric Returns

Is the modest work effort of the Bushmen sufficient to provide the calories necessary to maintain the health of the population? Or have the!Kung, in common with some agricultural peoples (see Richards, 1939), adjusted to a permanently substandard nutritional level?

During my field work I did not encounter any cases of kwashiorkor, the most common nutritional disease in the children of African agricultural societies. However, without medical examinations, it is impossible to exclude the possibility that subclinical signs of malnutrition existed.[10]

Table 4. Summary of Dobe Work Diary

r-b-richard-b-lee-irven-devore-man-the-hunter-29.jpg

Table 5 Caloric and Protein Levels in the!Kung Bushman Dietary, July-August, 1964

r-b-richard-b-lee-irven-devore-man-the-hunter-30.jpg

Another measure of nutritional adequacy is the average consumption of calories and proteins per person per day. The estimate for the Bushmen is based on observations of the weights of foods of known composition that were brought into Dobe camp on each day of the study period. The per-capita figure is obtained by dividing the total weight of foodstuffs by the total number of persons in the camp. These results are set out in detail elsewhere (Lee, 1967) and can only be summarized here. During the study period 410 pounds of meat were brought in by the hunters of the Dobe camp, for a daily share of nine ounces of meat per person. About 700 pounds of vegetable foods were gathered and consumed during the same period. Table 5 sets out the calories and proteins available per capita in the!Kung Bushman dietary from meat, mongongo nuts, and other vegetable sources.

This output of 2,140 calories and 93.1 grams of protein per person per day may be compared with the Recommended Daily Allowances (RDA) for persons of the small size and stature but vigorous activity regime of the!Kung Bushmen. The RDA for Bushmen can be estimated at 1,975 calories and 60 grams of protein per person per day. (Taylor and Pye, 1965, pp. 45–48, 463). Thus it is apparent that food output exceeds energy requirements by 165 calories and 33 grams of protein. One can tentatively conclude that even a modest subsistence effort of two or three days work per week is enough to provide an adequate diet for the!Kung Bushmen.

The Security of Bushman Life

I have attempted to evaluate the subsistence base of one contemporary hunter-gatherer society living in a marginal environment. The!Kung Bushmen have available to them some relatively abundant high-quality foods, and they do not have to walk very far or work very hard to get them. Furthermore this modest work effort provides sufficient calories to sup^ port not only the active adults, but also^ a large number of middle-aged and elderly people. The Bushmen do not have to press their youngsters into the service of the food quest, nor do they have to dispose of the oldsters after they have ceased to be productive.

The evidence presented assumes an added significance because this security of life was observed during the third year of one of the most severe droughts in South Africa’s history. Most of the 576,000 people of Botswana are pastoralists and agriculturalists. After the crops had failed three years in succession and over 250,000 head of cattle had died on the range for lack of water, the World Food Program of the United Nations instituted a famine relief program which has grown to include 180,000 people, over 30 per cent of the population (Government of Botswana, 1966). This program barely touched the Doll(’ area in the isolated northwest corner of the country and the Herero and Tswana women there were able to feed their families only by joining the Bushman women to forage for wild foods. Thus the natural plant resources of the Dobe area were carrying a higher proportion of population than would be the case in years when the Bantu harvested crops. Yet this added pressure on the land did not seem to adversely affect the Bushmen.

In one sense it was unfortunate that the period of my field work happened to coincide with the drought, since I was unable to witness a “typical” annual subsistence cycle. However, in another sense, the coincidence was a lucky one, for the drought put the Bushmen and their subsistence system to the acid test and, in terms of adaptation to scarce resources, they passed with fiying colors. One can postulate that their subsistence base would be even more substantial during years of higher rainfall.

What are the crucial factors that make this way of life possible? I suggest that the primary factor is the Bushmen’s strong emphasis on vegetable food sources. Although hunting involves a great deal of effort and prestige, plant foods provide from 60–80 per cent of the annual diet by weight. Meat has come to be regarded as a special treat; when available, it is welcomed as a break from the routine of vegetable foods, but it is never depended upon as a staple. No one ever goes hungry when hunting fails.

The reason for this emphasis is not hard to find. Vegetable foods are abundant, sedentary, and predictable. They grow in the same place year after year, and the gatherer is guaranteed a day’s return of food for a day’s expenditure of energy. Game animals, by contrast, are scarce, mobile, unpredictable, and difficult to catch. A hunter has no guarantee of success and may in fact go for days or weeks without killing a large mammal. During the study period, there were eleven men in the Dobe camp, of whom four did no hunting at all. The seven active men spent a total of 78 man-days hunting, and this work input yielded eighteen animals killed, or one kill for every four mandays of hunting. The probability of any one hunter making a kill on a given day was 0.23. By contrast, the probability of a woman finding plant food on a given day was 1.00. In other words, hunting and gathering are not equally felicitous subsistence alternatives.

Consider the productivity per man-hour of the two kinds of subsistence activities. One man-hour of hunting produces about 800 edible calories, and of gathering, 2,000 calories. Gathering is thus seen to be 2.4 times more productive than hunting. In short, hunting is a high-risk, low-return subsistence activity, while gathering is a low-risk, high-return subsistence activity.

It is not at all contradictory that the hunting complex holds a central place in the Bushman ethos and that meat is valued more highly than vegetable foods (Marshall, 1960). Analogously, steak is valued more highly than potatoes in the food preferences of our own society. In both situations the meat is more “costly” than the vegetable food. In the Bushman case, the cost of food can be measured in terms of time and energy expended. By this standard, 1,000 calories of meat “costs” 1.2 man-hours, while the “cost” of 1,000 calories of vegetable foods is only 0.5 man-hours. Further, it is to be expected that the less predictable, more expensive food source would have a greater accretion of myth and ritual built up around it than would the routine staples of life, which rardy if ever fail.

Eskimo-Bushman Comparisons

Were the Bushmen to be deprived of their vegetable food sources, their life would become much more arduous and precarious. This lack of plant foods, in fact, is precisely the situation among the Netsilik Eskimo, reported by Balikci (Chapter 8, this volume). The Ketsilik and other Central Arctic peoples are perhaps unique in the almost total absence of vegetable foods in their diet. This factor, in combination with the great cyclical variation in the numbers and distribution of Arctic fauna, makes Eskimo life the most precarious human adaptation on earth. In effect, the kinds of animals that are “luxury goods” to many hunters and gatherers are to the Eskimos the absolute necessities

’!f life. However, even this view should not be exaggerated, since most of the Eskimos in historic times have lived south ofthe Arctic Circle (Laughlin, Chapter 25a this volume) and many of the Eskimos at all latitudes have depended primarily on fishing, which is a much more reliable source of food than is the hunting of land and sea mammals.

What Hunters Do for a Living: A Comparative Study

I have discussed how the!Kung Bushmen are able to manage on the scarce resources 9f their inhospitable environment. The essence of their successful strategy seems to be that while they depend primarily on the more stable and abundant food sources (vegetables in their case), they are nevertheless willing to devote considerable energy to the less reliable and more highly valued food sources such as medium and large mammals. The steady but modest input of work by the women provides the former, and the more intensive labors of the men provide the latter. It would be theoretically possible for the Bushmen to survive entirely on vegetable foods, but life would be boring indeed without the excitement of meat feasts. The totality of their subsistence activities thus represents an outcome of two individual goals; the first is the desire to live well with adequate leisure time, and the second is the desire to enjoy the rewards, both social and nutritional, afforded by the killing of game. In short, the Bushmen of the Dobe area eat as much vegetable food as they need, and as much meat as they can.

It seems reasonable that a similar kind of subsistence strategy would be characteristic of hunters and gatherers in general. Wherever two or more kinds of natural foods are available, one would predict that the population exploiting them would emphasize the more reliable source. We would also expect, however, that the people would not neglect the alternative means of subsistence. The general view offered here is that gathering activities, for plants and shellfish, should be the most productive of food for hunting and gathering man, followed by fishing, where this source is available. The hunting of mammals ‘is the least reliable source of food and should be generally less important than either gathering or fishing.

In order to test this hypothesis, a sample of 58 societies was drawn from the Ethnographic Atlas (Murdock, 1967). The basis for inclusion in the sample was a 100 per cent dependence on hunting, gathering and fishing for subsistence as rated in Column 7–1 1 of the Atlas (Murdock, 1967, pp. 154–55). These 58 societies are plotted in Figures 1 and 2 and are listed in Tables 7 and 8 of the Appendix to this Article.[11][12]

The Ethnographic Atlas coding discusses “Subsistence Economy” as follows:

A set of five digits indicates the estimated relative dependence of the society on each of the five major types of subsistence activity. The first digit refers to the gathering of wild plants and’ small land fauna; the second, to hunting, including trapping and fowling; the third, to fishing, including shell fishing and the pursuit of large aquatic animals; the fourth, to animal husbandry; the fifth, to agriculture (Murdock, 1967, PP. 154—55).

Two changes have been made in the definitions of subsistence. First, the participants at the symposium on Man the Hunter agreed that the “pursuit of large aquatic animals” is more properly classified under hunting than under fishing. Similarly, it was recommended that shellfishing should be classified under gathering, not fishing. These suggestions have been followed and the definitions now read: Gathering-collecting of wild plants, small land fauna and shellfish; Hunting-pursuit of land and sea mammals; Fishing-obtaining of fish by any technique. In 25 cases, the subsistence scores have been changed in light of these definitions and after consulting ethnographic sources.[13]

In Tables 9 and io of the Appendix to this article, the percentage dependence on gathering, hunting, and fishing, and the most important single source of food for each society are presented. Such scores can be at best only rough approximations; however, the results are so striking that the use of these scores seems justified. In the Old World and South American sample of 24 societies, sixteen depend on gathering, five on fishing, while only three depend primarily on mammal hunting: the Yukaghir of northeast Asia, and the Dna and Shiriana of South America. In the North American sample, thirteen societies have primary dependence on gathering, thirteen on fishing, and eight on hunting. Thus for the world as a whole, half of the societies (29 cases) emphasize gathering, one-third (18 cases) fishing, and the remaining one-sixth (II cases) hunting.

On this evidence, the “hunting” way of life appears to be in the minority. The resuit serves to underline the point made earlier that mammal hunting is the least reliable of the subsistence sources, and one would expect few societies to place primary dependence on it. As will be shown, most of the societies that rely primarily on mammals do so because their particular habitats offer no viable alternative subsistence strategy.

The Relation of Latitude to Subsistence

The peoples we have classified as “hunters” apparently depend for most of their subsistence on sources other than meat, namely, wild plants, shellfish and fish. In fact the present sample over-emphasizes the incidence of hunting and fishing since some three-fifths of the cases (34/58) are drawn from North America (north of the Rio Grande) a region which lies entirely within the temperate and arctic zones. Since the abundance and species variety of edible plants decreases as one moves out of the tropical and temperate zones, and approaches zero in the arctic, it is essential that the incidence of hunting, gathering, and fishing be related to latitude.

Table 6 shows the relative importance of gathering, hunting, and fishing within each of seven latitude divisions. Hunting appears as the dominant mode of subsistence only in the highest latitudes (60 or more degrees from the equator). In the arctic, hunting is primary in six of the eight societies. In the cool to cold temperate latitudes, 40 to 59 degrees from the equator, fishing is the dominant mode, appearing as primary in 14 out of 22 cases. In the warm-temperate, subtropical, and tropical latitudes, zero to 39 degrees from the equator, gathering is by far the dominant mode of subsistence, appearing as primary in 25 of the 28 cases.

For modern hunters, at any rate, it seems legitimate to predict a hunting emphasis only in the arctic, a fishing emphasis in the mid-high latitudes, and a gathering emphasis in the rest of the world.[14]

The Importance of Hunting

Although hunting is rarely the primary source of food, it does make a remarkably stable contribution to the diet. Fishing appears to be dispensable in the tropics, and a number of northern peoples manage to do without gathered foods, but, with a single exception, all societies at all latitudes derive at least 20 per cent of their diet from the hunting of mammals. Latitude appears to make little difference in the amount of hunting that people do. Except for the highest latitudes, where hunting contributes over half of the diet in many cases, hunted foods almost everywhere else constitute 20 to 45 per cent of the diet. In fact, the mean, the median, and the mode for hunting all converge on a figure of 35 per cent for huntergatherers at all latitudes. This percentage of meat corresponds closely to the 37 per cent noted in the diet of the!Kung Bushmen of the Dobe area. It is evident that the!Kung, far from being an aberrant case, are entirely typical of the hunters in general in the amount of meat they consume.

Table 6 Primary Subsistence Source by Latitude

r-b-richard-b-lee-irven-devore-man-the-hunter-31.jpg

Conclusions

Three points ought to be stressed. First, life in the state of nature is not necessarily nasty, brutish, and short. The Dobe-area Bushmen live well today on wild plants and meat, in spite of the fact that they are confined to the least productive portion of the range in which Bushman peoples were formerly found. It is likely that an even more substantial subsistence base would have been characteristic of these hunters and gatherers in the past, when they had the pick of African habitats to choose from.

Second, the basis of Bushman diet is derived from sources other than meat. This emphasis makes good ecological sense to the!Kung Bushmen and appears to be a common feature among hunters and gatherers in general. Since a 30 to 40 per cent input of meat is such a consistent target for modern hunters in a variety of habitats, is it not reasonable to postulate a similar percentage for prehistoric hunters? Certainly the absence of plant remains on archeological sites is by itself not sufficient evidence for the absence of gathering. Recently-abandoned Bushman campsites show a similar absence of vegetable remains, although this paper has clearly shown that plant foods comprise over 60 per cent of the actual diet.

Finally, one gets the impression that hunting societies have been chosen by ethnologists to illustrate a dominant theme, such as the extreme importance of environment in the molding of certain cultures. Such a theme can be best exemplified by cases in which the technology is simple and/or the environment is harsh. This emphasis on the dramatic may have been pedagogically useful, but unfortunately it has led to the assumption that a precarious hunting subsistence base was characteristic of all cultures in the Pleistocene. This view of both modern and ancient hunters ought to be reconsidered. Specifically 1 am suggesting a shift in focus away from the dramatic and unusual cases, and toward a consideration of hunting and gathering as a persistent and well-adapted way of life.

Appendix

Table 7. The Sample of Hunter-Gatherers (excluding North America)

r-b-richard-b-lee-irven-devore-man-the-hunter-32.jpg

Table 8. The North American Sample

r-b-richard-b-lee-irven-devore-man-the-hunter-33.jpg
r-b-richard-b-lee-irven-devore-man-the-hunter-6.jpg
Figure 1. Old world and South American hunter-gatherers

Table 9. Subsistence Base qf Hunter-Gatherers (Old World and South America)

r-b-richard-b-lee-irven-devore-man-the-hunter-35.jpg
r-b-richard-b-lee-irven-devore-man-the-hunter-7.jpg
Figure 2. North American hunter-gatherers.

Table 10. Subsistence Base qf Hunter-Gatherers (North America)

r-b-richard-b-lee-irven-devore-man-the-hunter-36.jpg

5. An Introduction to Hadza Ecology

James Woodburn

The Hadza (also known as Hadzapi, Tindiga, Kindiga, and Kangeju) are a small group of nomadic hunters and gatherers living in the vicinity of Lake Eyasi, a salt, rift-valley lake not far south of the equator in Tanzania.[15]

Their language contains click consonants and several authorities have claimed that it may be related to Bushman languages in southern Africa (for example, Bleek, 1931a, 1931b, 1956; Greenberg, 1950; Westphal, 1956). My own work on the language[16] suggests that the material put forward as evidence for such a link is unsatisfactory and some research now in progress indicates that Hadza may possibly have distant affinities with languages of the Erythraic (Hamito-Semitic) group (A. N. Tucker, personal communication). Work carried out in 1966 and 1967 with N. A. Barnicot and F. J. Bennett on Hadza anthropometry and blood-grouping has not yet shown any evidence for a physical link with Bushmen, but only preliminary results are so far available. There is no satisfactory evidence for a cultural tie with Bushmen.

If we treat as Hadza all those who speak the Hadza language as their first language, there are three separate groupings of Hadza. In the tsetse bush to the east of Lake Eyasi, an area of, until recently, well over a thousand square miles, are some four hundred nomadic hunters and gatherers whom I shall refer to as the Eastern Hadza. South of these are more than a hundred settled Hadza who live, and have lived at least since the first Europeans visited the area in the 1890’s, by agriculture and by hunting, and who have intermarried intensively with the neighboring Bantu-speaking Isanzu tribe. The third group of Hadza, the great majority of whom live by hunting and gathering, are to be found in the neighborhood of Kimali to the west of Lake Eyasi and number about 250 people. Little contact occurs between the Eastern and the Western Hadza: only very occasionally have individuals moved temporarily or, more rarely, permanently from one of these areas to another. Almost all my field research,[17] carried out continuously from 1958 until i960, and again for short periods in 1961, 1965, 1966, and 1967, was done among the nomadic Eastern Hadza, and in this paper I shall refer almost exclusively to them.[18] Unlike most other East African hunters and gatherers, the Eastern Hadza are relatively independent of their agricultural and pastoral neighbors. Although they rely on trade and begging to obtain tobacco, cloth, beads, iron, and other goods, they have not entered into an elaborate dependence on, or interdependence with their neighbors. In this respect they stand in marked contrast to the Mbuti Pygmies of the Congo (Turnbull, 1965b and Chapter 15, this volume) and the Kenya Dorobo described by Huntingford (1954).

The small numbers of the Eastern Hadza should not be taken as evidence that they are a broken down remnant of some much larger group. Their numbers have been quite adequate to maintain their language, their culture, and their social organization; there is no evidence to indicate that their numbers were either much greater or much smaller in the past. No very serious epidemic, no large-scale war, no time of famine is remembered.

Environment

The country of the Eastern Hadza, dry, rocky savanna, dominated by thorn scrub and acacia trees and infested with tsetse flies, has been described by a traveler who walked through it in 1921 as “a barren land ...to all intents and purposes a desert” (Barns, 1923, pp. 71–72). Barren though it may be in places, the country is rich in wild foods. Animals are exceptionally numerous — and were certainly even commoner at the beginning of the century. Elephant, rhinoceros, buffalo, giraffe, eland, zebra, wilde — beeste, hartebeeste, waterbuck, impala, Thomson’s gazelle, warthog, baboon, lion, leopard, and hyena are all common, as are smaller animals such as anteater, porcupine, hare, hyrax, dik-dik, klipspringer, jackal, tortoise, and many others. All of these animals, apart from the elephant, are hunted and eaten by the Hadza. The amount of meat that could be regularly harvested without endangering the future of the game in the area is probably as great or greater than anywhere else in the world where hunters and gatherers live or have lived in the recent past.

Vegetable food-roots, berries, the fruit of the baobab tree, etc.-though not often obvious to the casual observer, is always abundant even at the height of the dry season in a year of drought. The type of vegetable food available is different in the six-month wet season from the dry season[19] but there is no period of shortage. The honey and grubs of seven species of wild bee are eaten; supplies of these vary widely from season to season and from year to year.

Sources of water are widely distributed over the country in the wet season but are very few in the dry season. The Hadza consider that about three or four miles is the maximum distance over which water can reasonably be carried and camps are normally sited within a mile of a water source.

Part of the country consists of open grass plains but the Hadza never build camps there. Camps are invariably sited among trees or rocks and, by preference, among both.

Subsistence

The Eastern Hadza assert no rights over land and its ungarnered resources.[20] Any individual may live wherever he likes and may hunt animals, collect roots, berries, and honey and draw water anywhere in Hadza country without any sort of restriction. Not only do the Hadza not parcel out their land and its resources among themselves, they do not even seek to restrict the use of the land they occupy to members of their own tribe. For at least the last fifty years outsiders have been steadily encroaching on land traditionally occupied by the Hadza, and by eliminating most of the game and destroying much of the vegetation in the areas they have occupied, they have denied the use of most of these areas to Hadza living by their traditional means of subsistence. However, in 1960 the Hadza still had plenty of land and abundant resources.[21]

In spite of the exceptional numbers of game animals in their area, the Hadza rely mainly on wild vegetable matter for their food. Probably as much as 80 per cent of their food by weight is vegetable, while meat and honey together account for the remaining 20 per cent.[22] In terms of calories, however, meat and honey represent far more than 20 per cent of the tota1.[23] Although a large number of species of plant food are eaten, only a few are systematically gathered and provide important quantities of food. The bulk of the vegetable food eaten by the Hadza is in fact obtained from only ten species of plant. The edible part of four of these plants is the root,[24] of five others the berry,[25] and of the remaining one its large fruit which contains edible pulp and seeds with edible kernels.[26]

Vegetable food is collected almost every day by the women of the camp who go out as a group or groups with their children. They make their way leisurely to the place where the vegetable food is to be found, which is usually within an hour’s walk of the camp. Whatever the type of vegetable food, a large proportion is eaten where it is gathered. Berries are quickly and easily collected and are eaten raw. Roots are obtained with rather more difficulty; they are dug up with a simple, sharpened wooden digging stick and are, in most cases, lightly roasted on an open fire. Only the food which -remains after the women and children have satisfied their hunger is brought back to camp and, even of the food brought back to camp, less than half is given to the men.

Men do not rely on the women to supply them with all the vegetable food that they need. They wander off into the bush individually for a while almost every day to satisfy their hunger. They gather vegetable food only for their own needs and normally bring none back to camp.

Hunting is done exclusively by men and boys and is an essentially individual pursuit. A man hunts only with bow and arrow; no guns, spears, traps, nets, or snares are now in use. Hunting procedure is simple and differs very little whether the target is a lion, a zebra, or a guinea fow!’ Once he has sighted his quarry the solitary hunter stalks it slowly and with care until he is, if possible, 25 yards or less away and then attempts a shot. For any animal larger than, say, a jackal, a poisoned arrow will be used. An animal hit with a poisoned arrow will usually be tracked down only after an interval of a couple of hours to allow the poison to kill the animal near the place where it is shot. To track it earlier might cause it to run for many miles before collapsing from the effect of the poison. Occasionally animals are shot ai night from hides over water and are tracked the following morning. The large animals which are most frequently killed are impala, zebra, eland, and giraffe.

The men, like the women, satisfy their hunger at the place where food is obtained. A man on his own will normally light a fire, cook, and eat on the spot any small animal he kills, and only after he is satisfied will he bring meat back to camp and, even there, a small animal is as likely to be eaten by the men as by the women and children. Men and women are thus unusually independent of each other in obtaining food. This arrangement has important implications for Hadza social structure which will be discussed in detail in my forthcoming monograph.

Although vegetable foods form the bulk of their diet, the Hadza attach very little value to them. They think of themselves and describe themselves as hunters. From informants’ assertions, one would gather that little but meat is eaten. In addition to being the preferred food, meat is also intimately connected with rituals to which Hadza men attach great importance.

Honey, although highly valued as a food, is especially important as a trading commodity. Neighboring tribesmen use honey for making beer and are eager to obtain it in trade.

The foods which Hadza eat, the size and the positioning of their camps, the activities within the camps differ markedly between the dry and wet seasons. These differences are not a simple product of ecological factors. In Hadza culture seasonal polarity is stressed. Activities which are characteristic of the wet season tend to be avoided during the dry season and vice versa. In the wet season the emphasis is on root gathering and hunting small game especially hyrax; in the dry season the emphasis is on gathering berries and hunting large animals. Camps are commonly small and widely dispersed in the wet season, large and concentrated near the few available sources of water in the dry season. In the small camps of the wet season men and women are not segregated to any great extent and live together relatively harmoniously, but in the large camps of the dry season sexually segregated activities (such as gambling) are stressed and the opposition, even hostility, between men as a group and women as a group is reiterated.

Ease of Subsistence

Hunting and gathering tribes are often described as living on the verge of starvation. It is easy to gain such an impression after living for a short while with the Hadza; often by nightfall every scrap of food in the camp has been eaten unless a large animal happens to have been killed recently. Moreover the Hadza place such emphasis on meat as proper food and treat vegetable foods as so thoroughly unsatisfactory in comparison that they are apt to describe themselves as suffering from hunger when they have less meat than they would like. In fact, there is never any general shortage of food even in time of drought. The range of foods in the bush is so great, if one knows what these are and how to obtain them, that if weather conditions should cause the failure of some type of root or berry, or the migration of some of the game, some other type of food is always available. For a Hadza to die of hunger, or even to fail to satisfy his hunger for more than a day or two, is almost inconceivable.

I have already mentioned the exceptional abundance of game animals in this area. Although Hadza, in common probably with all other human societies, do not eat all the types of animal available to them—they reject civet, monitor lizard, snake, terrapin among others- they do eat an unusually wide range of animals including predators such as lion, leopard, serval, wild cat and scavengers such as l1yena, jackal, vulture.[27] With their very powerful bows and their poisoned arrows they are able to kill without any great difficulty all the animals in the area with the sale exception of elephant, which are too large to be killed by the type of poison they use. In spite of the large number of species which they are both able to hunt and regard as edible, the Hadza do not kill very many animals and it is probable that even in the radically reduced area they occupied in 1960 more animals could have been killed of every species without endangering the survival of any species in the area.

The low opinion which Hadza have of the vegetable food which makes up the bulk of their diet is not surprising when its unpalatability is taken into account. Roots are, in general, tough, fibrous, and have little taste; many of the berries are hard and dry and contain large stones which are swallowed whole; undushi berries leave a dry, sticky residuum in the mouth; k’alahai berries split the lips and tongue if eaten in quantity. Very little of the vegetable food is eaten with much enthusiasm. But the advantage of vegetable food over meat (or honey), and the basic reason why it constitutes the bulk of the diet of the Hadza, is that

It can be obtained quickly and, above all, predictably. When they go out to look for vegetable food, the women can be sure that they will find some type without undue effort. Hunting, even by a highly skilled hunter, on the other hand, is always an unpredictable pursuit and therefore one which is less suitable as a basis for day-to-day subsistence.

With food of some sort always available, the Hadza give little attention to the conservation of their food resources. When women dig up roots, they do not attempt to replace any portion of the plant to grow again. When they gather berries, heavily laden branches are often torn from the trees and carried back to camp. When a woman is building the framework of her grass hut, she is as likely to use branches from berry trees as from any other type of tree that happens to be near at hand. The Hadza may inadvertently assist in propagating berry trees in their locality by their practice of swallowing the berry stones which pass intact through the digestive tract, but they make no deliberate effort to conserve the trees and other food-producing plants.

When a nest of wild bees is found and raided for its honey, no portion of the comb is left to encourage the bees to stay on. Moreover, little effort is made to leave the nest suitable for reoccupation. If the nest is in a hollow tree and the entrance has been cut open with an axe to extract the comb, the hollow will not again be used by bees unless the hole through which the comb has been extracted is partially blocked. When there happens to be a stone of suitable size near at hand, this may be wedged into the hole in the hope that the bees will later return and reoccupy the hollow. As often as not, though, the hole will be left unblocked; people do not feel obliged to attend to this task and do not bother about it unless they can do so very easily.

In hunting no attempt is made at systematic cropping. A man out hunting will shoot any animal he comes across. There are no inhibitions about shooting females (even pregnant females) or immature animals. Adult males are preferred but only because they are generally larger. Hadza do not often kill more animals than they need but this is only because hunting requires effort and they see no virtue in hunting unless they are hungry for meat. Meat may sometimes be wasted: if a large animal is killed by a man of a small camp, some of the less palatable portions may be left behind for the vultures and hyenas. Iftwo animals are killed on the same day, the more distant one may be abandoned. When tracking an animal wounded by a poisoned arrow, it will commonly be abandoned if it is not found in one day’s tracking.

In camp meat is widely distributed and rapidly consumed.[28] The Hadza are familiar with techniques for drying meat in the sun and could, if they wished, preserve stocks for months, at least in the dry season. But in practice meat is rarely preserved: it is eaten in large quantities until it is finished. Meat should be shared with those who ask for it. In particular pregnant women have the right to eat meat belonging to anybody.[29] To eat meat slowly, to preserve it and store it would be largely wasted effort; other people would simply demand meat when their own was finished and it would be wrong to refuse them.

Hunting is not a coordinated activity. Men hunt individually and decide for themselves where and when they will go hunting. When a man goes off into the bush with his bow and arrows, his main interest is usually to satisfy his hunger. Once he has satisfied his hunger by eating berries or by shooting and catching some small animal, he is unlikely to make much effort to shoot a large animal. Of course, if he sees an animal close by which can easily be hunted, he will almost always take the opportunity. Men most often return from the bush empty-handed but with their hunger satisfied.

During the dry season Hadza men spend much more time gambling than hunting. The gambling game, played with bark disks, is a game of chance in which skill plays hardly any part. The most usual and most acceptable stake is a metal-headed arrow, preferably of the poisoned type. In the course of a day arrows (and other possessions which are staked) will change hands hundreds of times. In practice,

Many men are, in any case, quite unprepared or unable to hunt big game even when they possess the necessary arrows. Large animals are killed by a small minority of the adult men. Perhaps as many as half of the adult men fail to kill even one large animal a year. The active hunters are usually men in their late teens, or in their twenties or thirties; few men continue to hunt large game actively after the age of 45. There are some men who have killed scarcely a single large animal during their entire adult lives. Whether a man hunts is his own affair. Other men will not put pressure on him. He may, though, find it more difficult to marry a wife, or, once married, to keep a wife, if he is unsuccessful in hunting big game.

In spite of the fact that the Hadza make scarcely any attempr to conserve the food resources of their area, that they rapidly eat all the food which comes into camp without preserving it, that they do not cooperate very much or coordinate their food gathering activities with each other, that they make hunting difficult for themselves by using their arrows for gambling, that a high proportion of men are failures at hunting, they nonetheless obtain sufficient food without undue effort. Over the year as a whole probably an average of less than two hours a day is spent obtaining food. [30] There are seasonal variations and variations from year to year but not very considerable ones.

We have good evidence that the food they eat is adequate nutritionally. In 1960 I was visited in the field by D. B. Jelliffe and F. J. Bennett of Makerere College Medical School, and they examined 62 Hadza children. According to their report, “the clinical nutritional status of all the children was good by tropical standards; in particular, the syndromes of kwashiorkor and nutritional marasmus, rickets, infantile scurvy, and vitamin B deficiency syndromes were not seen.” In 1966 and 1967 Dr. Bennett, this time in cooperation with N. A. Barnicot and myself, did further work on Had7.a nutrition. More than 450 Hadza were examined in detail; some of these were resident in settlements, others were still living in the bush by their tcaditional means of subsistence. The nutritional status of those who were resident in the bush was again found to be good by tropical standards.

My impression is that, over the year as a whole, the Hadza spend less energy (and probably less time) obtaining their subsistence than do neighboring agricultural tribes, but until detailed comparative research is done the matter must remain in doubt. From a nutritional point of view the Hadza again appear to be better off than their agricultural neighbors although to establish this, too, more research is needed. I t is clear that agriculturalists are liable to suffer from recurrent famine in this area while hunters and gatherers are not. In the early years of this century, before the colonial government provided effective famine relief measures, numerous Isanzu took to the bush from time to time and lived like the Hadza on bush produce until the famine eased and they were able to return to cultivation. Some stayed on and intermarried with the Hadza. In general the bush plants on which the Hadza mainly rely yield with great regularity; failure is far rarer than it is with vulnerable, introduced cultivated plants such as maize and other cereals grown by neighboring peoples. In comparison with their agricultural neighbors the Hadza are well protected against the dangers of famine. They are doubly protected by the diversity of the food supplies available and by the lesser vulnerability of wild plants to such natural hazards as drought, insects, and birds. Hadza prefer maize and millet meal porridge to their own unappetizing berries and roots but, until the recent settlement schemes were initiated in 1964, they did not in general think these foods were worth all the effort and uncertainty of cultivation.

Perhaps largely because of the temporal priority of hunting and gathering, there has been a widespread tendency to see it as a hard and demanding way of life in which the necessities of the food quest dominate people’s lives. With the Hadza this is clearly not the case and judging from some of the other papers in this volume, they may not be exceptional. I have sought to show that the Hadza meet their nutritional needs easily without much effort, much forethought, much equipment, or much organization. The social implications are important and will be discussed partly in a second paper in this volume and partly in my forthcoming monograph.

6. Coping with Abundance: Subsistence on the Northwest Coast

Wayne Suttles

Although the aboriginal peoples of the Northwest coast of North America were not hunters so much as fishermen, they seem especially worth including in a survey of rv1an the Hunter for two reasons: First, their rich, maritime, temperate-zone habitat is a type in which few food-gathering peoples survived until historic times, partly because this very type of habitat elsewhere saw the growth of more advanced forms of subsistence.[31] Second, the Northwest coast peoples seem to have attained the highest known levels of cultural complexity achieved on a food-gathering base and among the highest known levels of population density. The Northwest coast refutes many seemingly easy generalizations about people without horticulture or herds. Here were people with permanent houses in villages of more than a thousand; social stratification, including a hereditary caste of slaves and ranked nobilit y; specialization in several kinds of hunting and fishing, crafts, a.nd curing; social units larger than villages; elaborate ceremonies; and one of the world’s great art styles. The area appears to have been matched in population density, among food-gathering areas, by only two or three areas adjacent to it-California and parts of the Arctic and Plateau culture areas. (Kroeber, 1939; but it would be good to ha\“C comparable data for the Ainu and the Lower Amur peoples.) These feature of Northwest coast culture and demography are generally thought to have been made possible, or even inevitably produced, by the richness of the habitat of the area and the efficiency of the subsistence techniques of its peoples. Perhaps, then, the study of Northwest coast subsistence can offer some guidance in estimating the possibilities of cultural development under comparable conditions in prehistoric times. (See Chapter 1, this volume.)

In a short paper I cannot hope to do justice to the variety and complexity of Northwest coast culture and its historic relations (for a general survey, see Drucker, 1955, 1965),[32] nor to go into the problem of aboriginal population sizes and densities except to comment that Mooney’s and Kroeber’s figures (Kroeber, 1939, pp. 131 ff.) have generally been revised upwards (Baumhoff, 1963, pp. 157—61; Duff, 1964, pp. 38–46; H. Taylor, 1963). I shall simply take the cultural complexity and population density as proven, and discuss our knowledge of subsistence and its relationship to them. But as you will see, I do not consider the relationship a simple one and so I must range through several sorts of phenomena. In general, my thesis is that while the habitat was undeniably rich, abundance did not exist the year round but only here and there and now and then,and that such temporary abundances -though they may well be a necessary condition for population density and cultural development of the sort seen on the Northwest coast -are not sufficient to create them. Equally necessary conditions were the presence of good though limited food-getting techniques, foodstoring techniques, a social system providing the organization for subsistence activities and permitting exchanges, and a value system that provided the motivation for getting food, storing food, and participating fully in the social system. I shall deal with each of these conditions in succession.

r-b-richard-b-lee-irven-devore-man-the-hunter-8.jpg
Figure 1. Indians of the Northwest Coast.

Habitat

The Northwest coast was an area where one could find, on a single occasion, <tuite literally tons of food. Salmon ran into the smaller streams by the thousands and into larger streams by the tens and hundreds of thousands. Waterfowl came to the marshes by the tens and hundreds of thousands. A single sturgeon can weigh nearly a ton, a bull sea lion more than a ton, a whale up to thirty tons. But this aspect of the Northwest coast habitat is only too ‘well known. It has been so emphasized that the implications of the phrase used above, “on a single occasion,” have been ignored and the habitat has been presented as a constant source of plenty. But as I have said elsewhere (Suttles, 1960, 1962), it was not constant. It did not provide an ever-reliable abundance of natural resources simply there for the taking. Abundance there consisted only of certain things at certain places at certain times and always with some possibility of failure. Describing the central coast Salish[33] habitat, I wrote:

The environmental setting of native culture was characterized by four significant features: I) variery of rypes of food, including sprouts, roots, berries, shellfish, fishes, waterfowl, land and sea mammals; 2) local variation in the occurrence of these types, due to irregular shorelines, broken topography, differences between fresh and salt water, local differences in temperature and precipitation; 3) seasonal variation, especially in vegetable foods and in anadromous fishes; 4) fluctuation from year to year, in part due to the regular cycles of the different populations of fish, in part to less predictable changes, as in weather (1960, p. 302; see also Suttles, 1962, pp. 527–29).

In their subsistence activities, the central coast Salish had to cope with these variable features of the habitat. It seems likely that farther north on the coast there were fewer types of resources but greater concentrations and thus possibly greater dangers in failure, through human error or natural calamity, of resources to appear at the right place at the right time (1962, pp. 530–33) . Farther south I would expect the opposite to hold. But these subareal differences have yet to be worked out.

One cause of unexpected hardship was purely climatic. Because of the Japan Current and prevailingly westerly winds from the ocean, the weather is usually mild and damp. Summers are cool and during many winters the temperature only now and again drops below the freezing point. In the central coast Salish area, winters without frost are not unknown. But occasionally masses of extremely cold continental air break out through the coast mountains to bring periods of as much as ten days of near-zero (Fahrenheit) weather to the outer coast. In British Columbia such severe outbreaks have been recorded at twelve- to eighteen-year intervals (Young, 1943, pp. xli-xlii).

From a number of places along the coast we have indications that there were times when food was scarce. For groups dependent upon the open ocean, the cause was usually a stretch of bad weather; for groups on rivers, it was tardiness or failure in a fish run. In both situations it was of course also a matter of human failure to have accumulated enough food for the emergency.

For the Haida of the Queen Charlotte islands we have a statement by an early observer, Poole, quoted by Niblack, who writes:

Some of these berries are collected and dried for winter’s use, forming, with dried fish, the principal winter’s supply. Poole, (1863) says of the Haida, that they often, through feasting or improvidence, eat up all the dried berries before spring, and “were it not for a few bulbs which they dig out of the soil in the early spring-time, while awaiting the halibut season, numbers of Indians really would starve to death” (Niblack, 1890, pp. 276–77).

For the Tsimshian of the Nass and Skeena rivers we have Boas’s statement, derived from his analysis of Tsimshian mythology, that “sometimes when the olachen were late in coming, there would be a famine on Nass River” (Boas, 1916, p. 399). In his comparison of Tsimshian data with Kwakiutl data in Kwakiutl Culture as Reflected in Mythology, Boas writes:

The difficulties of obtaining an adequate food supply must have been much more serious among the Tsimshian than among the KWakiuti, for starvation and the rescne of the tribe by the deeds of a great hunter or by supernatural help are an ever-recurring theme which, among the Kwakiutl, is rather rare. One story of this type is clearly a Tsimshian story retold Starvation

stories of the Kwakiutl occur particularly among the tribes living at the heads of the inlets of the mainland, not among those who dwell near the open sea, where seals, sealions, salmon and halibut are plentiful (Boas, 1935, p. 1 7 1; see other references in Piddocke, 1 965, p. 247) .

For the Nootka, who occupy the west coast of Vancouver Island in less sheltered waters than the Kwakiutl, we have Drucker’s statement that a poor dog salmon or herring run followed by weather bad enough to prevent people from going out for cod or halibut “quickly brought privation” when people sought foods they ordinarily disdained. He cites Father Brabant’s report of “two successive springs at Hesquiat when pickings were lean and children cried out with hunger, until the weather abated enough for the fishermen to go out.” He also cites family traditions of the hardships of winters on the outer coast before alliances had been made with groups with salmon streams (Drucker, 1951, pp. 36–37).

Swan says of the Makah at Cape Flattery:

The ease with which these Indians can obtain their subsistence from the ocean makes them improvident in laying in supplies for winter use, except of halibut; for, on any day in the year when the weather will permit, they can procure, in a few hours, provisions enough to last them for several days (1870, p. 30).

Yet he later (p. 76) describes a period in April of 1864 when the weather did not permit going out fishing or whaling and how the Ozette were concerned enough about this to accuse an old man of sorcery and threaten to kill him if he didn’t stop his incantations and make fair weather. The Quileute, Swan was told, had killed an alleged sorcerer for bringing bad weather during the halibut season only a few summers earlier.

Discussing relations between the upper and lower classes among the Chinook of the mouth of the Columbia River, Ray (1938, p. 56) says that his principal informant “declared that the upper class could infringe as much as it pleased upon the lower classes and added that famine was unknown to the former since the food of the latter was appropriated in such a circumstance.” Ray adds that the informant’s specific examples were the acts of chiefs. In the context of Ray’s other statements, it appears that the Chinook village chief received, or seized, tribute and redistributed food to his people, favoring the upper class. My point in citing this is simply to indicate that shortages were conceivable.

Finally, we have Gould’s (1966) recent reanalysis of the wealth quest among the Tolowa, an Athapaskan-speaking people on the Oregon-California border. This is the group described by DuBois (1936) as having distinctly separate “subsistence” and “prestige” economies in the paper that first introduced these terms into the study of primitive economies. According to DuBois, subsistence was not a problem; scarcity existed only in “treasures” (dentalia shells, obsidian blades, and woodpecker scalps); food circulated freely in a “subsistence economy, “ while men bent their efforts in gaining prestige through amassing wealth in the “prestige economy. “ But Gould has found no separate “subsistence economy” in that there is no evidence for specialization in subsistence or the exchange of foods. There do appear, however, to have been differences in the productivity of households due in part to differences in the number of women available to process foods in season. It is true that “treasure” items were not constantly used to buy food, nor ideally nor publicly so used, but they were commonly used to buy food during periods of shortage. In fact, storing food to sell to others was recognized as an important method of acquiring wealth; hence the interest the Tolowa show in acquiring women, the processors offood. Wealth is converted into prestige by its use in ceremonial displays and in payment of bride price, which establishes the social status of children to the marriage (Gould, 1966, pp. 70, 77, 86). Thus recognition of the existence of periods of shortage and of problems in subsistence has made possible a reinterpretation of the data that give them a greater coherence than in the earlier analyses.[34] There is really only one Tolowa economy; in it food is converted into wealth and wealth is converted into prestige. Occasional shortages of food provided the occasions for converting wealth into food. Hungry people wanted food; greedy people wanted wealth. Why did the greedy want wealth? Because wealth brought prestige. Gould does not ask why the Tolowa sought prestige. But it seems likely that it was because indirectly prestige-seeking enabled hungry people to obtain food. Or, if this seems to imply conscious purposiveness, we may say it is because populations that have unconsciously stumbled on ways of feeding hungry members survive better than those who let them starve.

I have dwelt at length on the existence of times of scarcity on the Northwest coast precisely because recognizing their existence may be essential to our understanding of the complexities of Northwest coast culture. Some years ago Bartholomew and Birdsell pointed out that it is a firmly established ecological generalization, known as “Liebig’s law of the minimum,” that the size of a population is determined not by the mean condition but by the extremes. “A semi-arid area may have many fruitful years in succession, but a single drought year occurring once in a human generation may restrict the population to an otherwise inexplicably low density” (Bartholomew and Birdsell, 1953, pp. 487–88; see also comment by Sherwood L. Washburn and June Helm, Chapter 9a, this volume). Perhaps, also, a single, once-a-generation failure of a major fish run or prolonged period of severe weather may explain an otherwise inexplicable practice such as the Northwest coast search for prestige.

What we need to know for the whole coast is not merely what resources were present and when and where they were found under ordinary conditions, but also the minimal occurrences in space, time, and quantity. This means we should pay special attention to what foods are, or were, available in winter, even though winter was culturally defined by most peoples of the area as the ceremonial season during which people lived on stored food. As we have seen from the quotations given, stored food was not always available. It also appears that there were differences between inland river environments, sheltered saltwater bays, and the open ocean shores. There were probably also differences from south to north. None of these has been worked out in detail, but I am confident, that, with the growing body of biological and climatological data on the area, we shall one day be able to state fairly accurately what the resource base of each people’s territory was.

Of course it is necessary to distinguish between what foods could have been hunted, fished, and gathered, and those foods actually obtained in practice. The difference between potential and actual is in part a function of the efficiency of food-getting methods.

Food-Getting Methods

The state of our knowledge in this respect, too, is certainly not as good as we would want it. We do not have and probably shall never have any figures on the quantities of food actually collected and, consumed per capita, on the ratio of meat, .fish, and vegetable food in the diet,[35] on the number of man-hours spent in the food quest, and on percent of the population supported in leisure or other non-subsistence activities. Moreover, for much of the area we do not even have the facts in the form of good ethnographic accounts of just how food was obtained. This is not to say there are no good works at all. Boas’s material on Kwakiutl subsistence activities, mainly in the form of native texts with translations (Boas, 1909, pp. 461516; 1921, part I), gives the native view of a number of food-getting and food-storing techniques and a great many recipes. In these texts we see what was important to the informant, but do not always find the answers to questions the ethnographer would ask. From some, however, we can make inferences about quantities of food taken and consumed. Among the ethnographies of the area, Drucker’s (1951) on the Nootka (see also Drucker, 1965, chap. 7, for a dramatic account of Nootka whaling) and Elmendorf’s (1960) on the Twana (a coast Salish group) are outstanding. The only specialized monograph is Kroeber and Barrett (1960) on fishing in northwestern California. But for much of the area between the Klamath and the Columbia and north of the Kwakiutl, we have only very sketchy accounts.

The reason for the deficiency in figures on subsistence is not hard to find. The purely aboriginal way of life had been greatly altered in most parts of the Northwest coast by the end of the last century and quantitative data are, as Kroeber pointed out (1939, p. 3), not ordinarily recoverable by the method of ethnographic reconstruction we must use in western North America. The reasons for the scant attention often paid to subsistence techniques are also easy to find. Some of the techniques had disappeared by the beginning of this century or had survived in forms altered by the introduction of European goods. But, also, the very complexity of social forms and richness of art and ceremony that draw attention to the area are likely to draw attention away from mere subsistence. Thus when McIlwraith (1948) had the opportunity to play a part himself in the Bella Coola winter ceremonies, he did so and the results form a good part of his two-volume work on that people. I find this quite understandable, but still wish we had more on Bella Coola salmon fishing than the pluses and minuses in Drucker’s (1950) element list. Another reason for the neglect of subsistence probably lies in the assumption that the habitat was so rich that subsistence simply was not a problem. Finally, an understanding of just how any given subsistence technique works requires moving outside of culture and seeing the technique as part of an ecosystem; this is not easily done if one is unfamiliar with the natural history of the area and has only a few summer months to spend in it.

But the situation is by no means hopeless. In spite of the fact that year by year the purely aboriginal way of life recedes into the past; I fully expect that a few years hence we shall know much more about Northwest coast subsistence than we know today. I have several reasons for this optimism. First, contrary to gloomy predictions by colleagues that I’d find all was lost, when I began working with the coast Salish I discovered that it was still possible to do ethnography and to get much new data on subsistence. I am sure that my experience could still be duplicated in other areas, if for no other reason than that some activities survive yet in modified form. Second, data from the related fields of linguistics, archeology, and ethnohistory and from the ethnography of other areas can be increasingly used in interpreting the ethnography of one’s own area. For example, some of my informants and the informants of others have believed that the gill net is not aboriginal in the Salish area. However, a comparison of native terms for the gill net and the record of an earlier observer suggest’ that it probably was, in fact, aboriginal. I expect that further probing will settle the matter. Third, the growing body of biological and climatological knowledge can be used to interpret the ethnographic data. For example, the testimony of informants that their ancestors fished by trolling in winter as well as spring is supported by the work of researchers in fisheries (for example, Rounsefell and Kelez, 1938, pp. 749–50). Their work indicates not only that spring salmon are present in winter but that Indians did troll for them. Also, other biological work shows that herring spawn in this area in winter and early spring (Clemens and Wilby, 1946, pp. 79–80) and would therefore be available during times of the month when tides are low. Finally, meteorological data could undoubtedly be obtained indicating average number of days during winter months when weather would permit trolling in each of several areas.

When I express this optimism about the possibility of increasing our knowledge of Northwest coast subsistence, I do not mean to imply that we shall surely be able to reconstruct the quantitative data. But what we surely can do is define more narrowly the requisites and limits of each technique so that we can make some estimates of the relative effectiveness of the different techniques under similar conditions and the same technique under different conditions.

Food-Storing Methods

The techniques for preserving food are certainly as important as those for getting food. Thousands of salmon swimming upstream in September would not make winter a time of ceremonial activity if people lacked the means of preserving them, nor would several tons of blubber on the beach. No doubt some people would not have survived some winters without storage methods.

A few foods do not keep well at all. Mussels and salmonberries, for example, seem to be too soft and watery. But most fishes and meats, some shellfishes, and some berries can be preserved by drying, either with or without smoke. Clams and some meats were steamed and salal berries were sometimes cooked before drying.

Whether fish or meat can be dried outside in the sun or must be smoked indoors depends of course upon the season of the catch and the climate of the area. In the central coast Salish area, where summers are relatively dry and sunny, summer catches of sockeye and humpback salmon, as well as halibut, lingcod, and other fish, were generally dried on outdoor racks at summer fishing camps, where the fishermen could live in mat shelters. Fall runs of springs, cohoes, and dog salmon, however, usually had to be smoked indoors because of the danger of rain. Occasionally separate structures may have been built for smoking fish, but evidently the usual “smokehouse” was the ordinary winter plank house and in local English usage this is what it is still called. The importance of this difference in preserving methods for social relations is easily seen. In the drier season a nuclear family, on its own, could store up a considerable amount of food with a minimum of shelter, but in the wetter season it would be bound to the plank house of its extended family. Because of its use as a smokehouse, the plank house is thus an important instrument of production, and the ownership or control of a house at the site of a fish weir used in the fall may have had as important social and economic consequences as ownership or control of the weir itse1f. In the Salish area, weirs were in fact usually public property but hOllses were not.

In many other parts of the coast, climatic conditions probably did not permit as much preserving of food olltside as in the Salish area. During the three months of summer, Victoria (in the Salish area) gets an average of 14 days of rain (out of an annual total of 133) while Port Hardy (in Kwakiutl country) gets 38 (out of a total of 204), Masset (Haida) 41 (out of 210), and Prince Rupert (Tsimshian) 46 (out of 2 I 5). During an average year Victoria gets a total of 2,092 hours of sunshine while Prince Rupert gets 1,019, or roughly half (figures from Kendrew and Kerr, 1955). Precipitation is also greater on the outer coast to the south than in the central coast Salish area. These differences certainly have implications for food preserving and possibly for social relations.

It should be noted that not all species of salmon, perhaps not even all populations of the same species, keep equally well. My Salish informants say that fatter fish last longer and thus sockeye and dog salmon are their favorites. Other species may not last through the winter months.

Another very important method of storage is the rendering of fat into oil, which can then be kept in such containers as seal bladders (in the Salish area), kelp bulbs (Kwakiutl), or wooden boxes (Chinook). Throughout most of the coast, dried fish or meat was eaten after being dipped in oil; sea-mammal oil was used from the Salish and Nootka southward, and eulachon oil was used from the Kwakiutl northward. On the Fraser, where seal oil was less plentiful, salmon oil was also used (Duff, 1952, p. 66). This constant use of oil seemed excessive to Europeans but it may have compensated for the scarcity of carbohydrates (Drucker, 1951, p.62; Rivera 1949, p. 34; see also note 5 of the present paper).

In the north, oil was also used for preserving some kinds of berries. From the Chinook southward, meat was preserved by pulverizing it and mixing it with grease. Salmon eggs were first allowed to get “high” and to form a kind of cheese-like substance, and eulachon were also allowed to decay before rendering; however there was evidently no general practice of allowing fish to decay. Storage pits are reported for a few areas and so are raised caches, but probably throughout most of the area preserved food was stored in boxes, baskets, and bags placed on racks inside the dwelling house.

A number of foods, such as dog salmon, must have required far more effort in the storing than in the taking. Thus limits in the exploitation of times of abundance may have been set less by people’s capacity to get food than by their capacity to store it. How much of a very heavy fall run of dog salmon could be stored must have depended on the number of hands available for the work of cutting and skewering, the drying-rack capacity of the houses at the site, fuel for smoke, and finally the number of containers available. As Gould has stressed, preserving food was largely the work of women and. for this reason rights over women and bride prices were so important in the economy of the Tolowa. The role of women in subsistence no doubt deserves reappraisal elsewhere on the Northwest coast. But it must be remembered that women need not be wives; they may also be slaves.

Values and Social Organization

Of course the possibilities offered by the environment and the techniques of food-getting and food-storing could be realized only through the work of people, organized by their social systems and motivated by their value system. To survive that occasional period of scarcity, people had to have not only the earlier periods of abundance; the weirs, nets, etc., to take the food with; and the drying racks, houses, etc., to preserve and store it with; but they also had to have a reason for doing it and the ability to mobilize the labor for it.

Probably the knowledge that there is an occasional period of scarcity is not reason enough to make most people store up food for every winter. But the Northwest coast peoples had “better” reasons. From the Yurok and Tolowa northward there was the ultimate goal of prestige, into which food was eventually converted. From the Chinook northward there was also the cultural definition of winter as a ceremonial season, when people should not have to seek food. (This last reason for storing food was also found through much of the Plateau as well.) Swan’s account of sorcery among the Makah (1870) suggests that some people did not expect random variation in their environment anyway; they attributed bad weather to human malice and coped with it by threatening the supposed sorcerer. If people stored food for “cultural” reasons rather than rational recognition of possible failure in the environment, this may explain why they sometimes did improvidently eat up all their stores in feasting, as Poole reported for the Haida. Nevertheless, the “cultural” reasons may have enabled larger populations to survive in this habitat than rational planning alone would have done.

In social organization, there seems to have been a rough sort of south-to-north gradient of increasing tightness of structure and size of social unit. Kroeber found the Yurok so individualistic that he declared:

Property and righ ts pertain to the realm of the individual, and the Yurok recognizes no public claim and the existence of no community. His world is wholly an aggregation of individuals. There being no society as such, there is no social organization. Clans, exogamic groups, chiefs or governors, political units are unrepresented even by traces in northwestern California (1952, p. 3).

Nevertheless, Kroeber’s data suggest possibly three kinds of social groups engaging in subsistence activities: the members of the household, the joint owners of a fishing place, and the large group from several villages that built the Kepel weir. This last group is an example of what Anastasio (1955) has called a “task grouping,” in which members of social units defined by other criteria come together for a particular purpose at a particular time and place. These task groupings are generally under the leadership of a person who has the requisite-technical and/or ritual knowledge. Treide (1965) has recently analyzed the function of such “salmon chiefs” as the ritual director of the Kepel weir for much of western North America.

Northward, on the Oregon coast the village seems to have consisted of an unnamed patrilineal kin group formed by patrilocal residence and to have functioned as a unit under some circumstances (Barnett, 1937, elements 1345, 1400). Among the Chinook at the mouth of the Columbia, thtTe was a village chief (who may have been simply the head of a patrilocal kin group) who evidently exacted tribute from his villagers and redistributed it (Ray, 1938, P. 5 6).

The central coast Salish social organization was seemingly looser than that of the Chinook. Village exogamy was preferred but residence was ambilocal so neither the household nor the village formed any kind of definable kin group. There were, however, cognatic kin groups perhaps best understood as “stem kindreds,” that is, the personal kindreds of successive generations of “owners” of certain ceremonial rights and one or more of the more productive natural resources. One of these men might tend to dominate a village but there was no village chief as such. The village was recognized as a unit when it came to certain types of sharing and certain ceremonial activities, but in general the household was autonomous. Task groups were directed by the “owners” of resources, such as fishing sites and clam beds; by owners of special gear, as the net for a deer drive; or simply by skilled specialists in the activity. Such subsistence activities and also ceremonial activities often brought together people from several villages over areas which crossed dialect and even language boundaries, but there were no structural principles that allowed for the definition of discrete social units.

Wakashan (Nootka and Kwakiutl) social structure seems to have been similar to that of the Salish in that both lacked rules of unilocal residence and unilinear descent. The Nootka and Kwakiutl, however, were much more structured through the principle of ranking of titles in series. Within the cognatic kin group, individuals, at least the more important ones, held positions ranked ill relation to olle another. Within the yillage (or “tribe”), kin groups were so ranked. In a few areas “tribes” formed confederacies through the same principle. Among the Wakashans, as among the Chinook, “chiefs” (in fact, kin group heads) evidently received tribute and acted as redistributers. (For the Nootka, see Drucker, 1951, especially pp. 220–21, 251–57.)

The highest development offormal organization with permanent discrete social units was that found among the northern peoples. The Tsimshian, Tlingit, IIaida, together with the Haisla (the northernmost Kwakiutl), had a system of matrilineages, sibs, and phratries or moieties. The largest of these kin groups were the phratries of the Haisla and Tsimshian and the moieties of the Tlingit and Haida. These were simply exogamous units and served to classify every person in the whole northern area for marriageability. Except for the Tsim-shian, the largest unit with economic and political functions was the lineage. But the Tsimshian of the lower Skeena and Nass rivers went farther and acknowledged the chief of the leading lineage of a village as the village chief Garfield believes that this probably occurred early in the eighteenth century. Then, early in the nineteenth century, villages up the Skeena established colonies near its mouth and these each rernained under the authority of the chief of the mother village, who in some cases also moved to the colony. “At this stage,” writes Garfield, “tribal chieftainship emerged and the tribal chief was regarded as the active leader of his tribesmen regardless of where they lived.” The village chief and the tribal chief, like the earlier lineage chief, received tribute and made decisions regarding moves to fishing sites and other subsistence activities (Garfield, 1951, PP. 34–35).

The greater development offormal organization in the north was probably accompanied by more effective control of larger labor forces. This difference in turn may account for the remarkably different views we find on the economic importance of slavery on the Northwest coast.

Barnett, writing primarily of the Kwakiutl and coast Salish, suggests that slaves were “in bondage ... as much a liability as an asset and ... useful primarily as overt demonstration of the ability to possess them” (1938, p. 352). Drucker writing about the area in general says:

It is difficult to estimate the slave population of the area, but if was certainly never very large, for slave mortality was high. Slaves’ economic utility was negligible. They gathered firewood, dug clams, and fished, but so did their masters (1965, p. 52).

Kroeber, listing exchange values among the Yurok, expressed mainly in strings of dentalia, writes:

A slave was rated at only one or two strings. Evidently the Yurok did not know how to exact full value from the labor of their bondsmen, not because the latter could not be held to work, but because industry was too little organized (1925, p. 27)-

The Yurok did not take slaves in war nor buy them from other peoples; all Yurok slaves were fellow Yurok enslaved by debt. The greatest number of slaves held by one man did not exceed three (p. 32).

On the other hand, Garfield (1945, p. 628; 1951, p. 30) has strongly opposed the view that slaves were kept merely-for prestige value and al’gued that their contribution must have increased the productivity of their owners’ households. A potlatching Tsimshian chief was not expected to mention what his slaves contributed in labor, but if he gave them away he boasted of what he had paid for them. During the nineteenth century a slave was worth from two hundred to a thousand dollars. They were mainly war captives and their descendants. The numbers held were much higher than among the southern tribes.

Ten to twenty slaves are reported as belonging to each of the nine tribal chiefs of Port Simpson in the middle nineteenth century. Each of approximately fifty Port Simpson lin:.:age heads is also reported to have owned from two to as many as ten slaves (Garfield, 1951, p. 30).

It does seem that the mere fact that slaves performed the same task as their masters does not ruie out their being of economic importance. If adding a wife or daughter-in-law can increase the productivtiy of a Tolowa household and give additional economic advantage, as Gould reports, why not adding a slave to a Tsimshian household? Kroeber may have been correct regarding the organization of labor among the Yurok, but can’t we expect more from the Tsimshian? Driver (1961, pp. 24546, 387–38) has pointed out that at the southern end of the Northwest coast a master could not kill a slave nor did he have sexual rights over a female slave, while in the north he held both these rights. This too suggests the possibility of greater control over their labor in the north.

Slavery may also be seen as one of several possible ways of making the human population fit the resources. The most drastic of these is infanticide in times of scarcity. As far as we know this was not practiced on the Northwest coast. The least drastic is fluidity of social groupings, allowing “surplus” people at a place with poor resources to move in with those at more favored places. This fluidity was a prominent feature of the coast Salish social system. With local exogamy, a man’s four grandparents were likely to have been from four different villages; at eacb of these he had residence rights and he also had the option of living with his wife’s people. For men of property it was certainly preferable to stay in one’s father’s village, but no rule required it. The Wakashans also seem to have been Hexible in the residence of lower-ranking people. For the northerners, the sib system may have facilitated the mobility of sib-identified groups. But for northwestern California and the Oregon coast we get consistent reports of patrilocal residence arid also of impoverished parents selling their children into slavery. Perhaps if residence rules are rigid, selling “surplus” children is the only peaceful alternative to infanticide.

Finally, in looking at social organization in relation to subsistence we should note that the local group was not simply a task force for the exploitation of resources. In one season it could function as a producer and distributor of surplus and in another season as an absorber of surplus produced by other groups. Elsewhere (Suttles, 1960, 1962) in discussing the central coast Salish socioeconomic system, I have pointed out how the preference for local exogamy and the exchanges that occured between affines in different villages may have been adaptive under conditions of variable occurence of natural resources. A man with a temporary abundance of any food had three choices: (i) he could share it with his fellow villagers, if they could consume it (which they could not ifthey too had the same abundance); (2) he could preserve it, if it was preservable and he had the labor force and time before the next harvest of fish, berries, etc., was due; or (3) he could take it to his in-laws in another village (where this particular food might be scarce) and receive in return a gift of wealth, which he might give later to in-laws bringing food to him. If he got more wealth than he gave, he could always potlatch and convert the wealth into glory, which of course might attract more prosperous in-laws. Thus, exchange between affines made it possible for a household not to store food and still take advantage of times of abundance.[36]

Vayda (I96Ia) and Piddocke (1965) have suggested that the Kwakiutl potlatch system may also be seen as an adaptation to a fluctuating environment rather than the “absurdly wasteful” epiphenomenon it had sometimes been labeled. From Piddocke’s review of the data, it appears that the Kwakiutl converted food into wealth by selling it to those in need (somewhat as Gould reports the Tolowa having done) and then (unlike the Tolowa) converted the wealth into prestige through competitive potlatching. Piddocke argues that the entry of new wealth into the system and the reduction of the population through disease increased the size and frequency of potlatches to the “fantastic” level of the classic descriptions.

Weinberg (1965) has criticized and expanded on Vayda’s suggestion and worked out a model of Kwakiutl culture as a self-regulating adaptive system in which the stability of culture is dependent on fluctuation in the environment within certain limits. When, in the nineteenth century, the limits were exceeded in the direction of surplus, the spectacular growth of the potlatch that followed was essentially an effort to maintain the stability of the culture (cf. “Romer’s Rule” of Hockett and Ascher, 1964).

It has been implicit in the previous discussion of the coast Salish that items of “wealth” (blankets, canoes, hide shirts, etc.) did not constitute all-purpose money. As in several other parts of the world, there were restrictions on the occasions when wealth items could properly change hands. These restrictions seem to have been Jess severe among the Tolowa, but still present in that only exchanges that related to marriage and litigation were publicized, while those that were purely commercial were not, hence the appearance of a separate “prestige economy.” Among the Kwakiutl, too, giving was certainly more honorable than selling., The areas where commercial transactions were most open and honorable seem to have been the north and, most especially, the lower Columbia. But even here commerce was not wholly free for it is reported that chiefs (lineage heads?) held monopolies over trade in their territories. It seems possible that such monopolies may have had their origins in exchanges between affines in areas of different resources. In both cases the most important trade was between the coast and interior—the Tlingit with the Athapaskan hinterland and the Chinookans with the Plateau Sahaptins. Perhaps the most important question, at present unanswerable, is: Why didn’t a market economy develop and spread out from these areas? Or were these commercial developments so recent that there was no time left for further growth? This may be, yet archeological evidence suggests that some form of trade goes back to the early occupation of the area.

Conclusions

I am afraid I have raised more questions about the Northwest coast than I have been able to answer. If! can offer any conclusion it can only be this: The Northwest coast material suggests that where people are faced with great seasonal and local variations in the amount of food offered by their habitat, their success in exploiting the abundance depends on more than technology alone. They must also have: (i) the organization of labor for getting and storing food (marital rights? kin obligations? property rights? monopoly of technical and/or ritual knowledge?); (2) some means of redistributing population on the habitat (wide kin ties and fluidity of residence? slavery?) and/or of redistributing the bounty ofthe habitat among the population (barter in food? exchange of food for durable goods? markets?); and (3) some motivating value (prestige?). It seems to me that these factors together with an adequate technology, are necessary conditions for coping with abundance, regardless of whether it appears in the form of fish, sea mammals, land game, or even vegetable food.

From the data presented on the “simpler” hunters by other members ofthe symposium, it appears that those hunters often are capable of organizing drives and other activities that would take quantities of animals if available. They also seem to have, and necessarily so, a fluidity of organization that permits redistribution of population and at least rudimentary trade. Perhaps the greatest point of contrast is that the simpler hunters lacked the characteristic Northwest coast feature of motivation to achieve and maintain superior status through production of surplus. For this reason it seems to me that the “hunting ideology” or nomadic style (Lee and DeVore, Chapter I, this volume) may turn out to be well worth the attention it seems to have attracted. Finally, I would ask, when we find archeological evidence of unexpected cultural complexity and population density, is it altogetiwr hopeless to seek ethnographic parallels from the Northwest coast? This is not to say that I believe the Upper Paleolithic Europeans held slaves and gave potlatches. But I expect that some day we will be in a position to say whether they possessed the functional equivalents.

7. Subsistence and Ecology of Northern Food Gatherers with Special Reference to the Ainu

Hitoshi Watanabe

Introduction

The symposium on Man the Hunter devoted a great deal of attention to problems of analyzing group structure and subsistence activities. The relations between these two areas and their spatial correlates, however, did not receive the attention they deserve. The purpose of this paper is to consider these two questions in more detail. First, I discuss the arrangement of settlements with reference to resources and attempt to clarify the concepts of nomadism and sedentarism. Second, I discuss the allocation of subsistence activities by sex and age, in order to reconsider the issue of division of labor and how it evolved. For documentation I draw upon field research among the Ainu (Watanabe, Ig64a, Ig64b, Ig64c) and other ethnographic data on Arctic and subarctic food gatherers.

The Problem of “Semi-Nomadism”

Residential Shifts and the Food-quest

Gjessing (1944) introduced the concept of a “semi-nomadism” in contemporary Arctic and subarctic food gatherers in his study of house and settlement types of the Circumpolar Stone Age. The so-called semi-nomadism of those peoples, however, shows such remarkable variation that it is difficult to apply the concept. A simple typology of three terms (nomadic, semi-nomadic, and sedentary) is not sufficient to describe the wide variety observed in patterns of residential shift. To make a systematic investigation of semi-nomadism, I have attempted a classification of the residential shift patterns of northern food gatherers (Table I).

This classification shows that the residential shift patterns of those peoples termed semi-nomadic actually vary across the spectrum. A feature common to all such people is settled life in winter. Thus, the variation in residential shift pattern is, in a large measure, brought about by the differences in the residence arrangements in summer. Such peoples as the Ainu (Watanabe, Ig64a, 1 964b, Ig64C), the Koryak (Jochelson, 1905–08), and the Bering Strait Eskimo (Nelson, 1899), attained the highest stability of residence in the sense that they occupied one and the same permanent dwelling all the year round. Of these peoples only the Ainu subsisted largely on land mammals, the rest chiefly subsisting on sea mammals and/or fish. Other northern gatherers living largely on land mammals are all nomadic wanderers in the summer. The period of the longest inhabitation at the same place, however, varies greatly in different groups. The Nunatarmiut (Larsen and Rainey, 1948), for instance, stay in winter villages for as long as nine months, although younger group members wander about and live in temporary camps even in winter. Oil the other hand, the Copper Eskimo (Jenness, 1922) lives in the winter settlements for only about three months.

Table 1. A Tentative Classification of Residential Shift Patterns

r-b-richard-b-lee-irven-devore-man-the-hunter-37.jpg

The classification presented here may also be applied to southern food gatherers, provided the division of the seasons is replaced by dry and rainy seasons. It is almost certain that

r-b-richard-b-lee-irven-devore-man-the-hunter-9.jpg
Figure 1. Representation of Ecological Zones of a River Valley as the Territory of an Ainu Group.

among food gatherers outside the Arctic and subarctic regions there also exists a wide variation of the residential-shift pattern. One extreme of the variation is represented by the Semang who move about throughout the year (Forde, 1934, 1954) and the other by the Owens Valley Paiute who inhabited permanent villages more or less year round (Steward, 1938). Significant here is the fact that both extremes represent groups subsisting on plant collecting. Other southern hunter-gatherers show considerable variation in the residential pattern. The case of the Vedda (Seligmann and Seligmann, 191 i) demonstrates that the residential shifting can be restricted to only two sites occupied alternately. On the other hand, the Forest Andamanese (RadcliffeBrown, 1933) are a good example of southern hunter-gatherers with a permanent settlement as a base for seasonal hunting and collecting.

In southern regions, however, it is difficult to find any hunter-gatherers who maintain as permanent a year-round residence as do the Ainu. Is the high residential stability of the Ainu a result of their subsistence on salmon? The economic importance of this fish for the Ainu is certainly great. But the deer also had an important place in their diet, and there is evidence to suggest that in some areas more than half of their animal food supply was derived from deer alone. In a settlement on the upper Tokapchi, for instance, the annual catch of the deer per family was not less than 300 while the annual store of fish per family consisted of 500–600 dog salmon and 600–800 cherry salmon. The figures are not likely to be exaggerated. The deer was hunted not only for the meat but also for the skin which was an important item of trade. The meat was consumed both by the Ainu themselves and by their hunting dogs. They also habitually prepared extra stores of food as insurance against the year when resources failed. The Tushipet valley in Tokapchi had no runs of dog salmon and the Ainu there lived chiefly on deer. In spite of their dependence on deer, even the Tushipet Ainu maintained perennially inhabited settlements.

Residential Stabili(y, Ecological Zones, and Habitat

One of the fundamental factors relevant to the high stability of residence among the Ainu may have been the distribution of the ecological zones within narrow river valleys they inhabited (Fig. I). Ecological zones refer to the zones of exploitative activities classified in terms of physiography and biota. The subsistence activities of the Ainu were conducted in the following ecological zones, each of which yielded specific resources in specific seasons: (1) The river: cherry salmon fishing (summer, in the main stream and some tributaries); dog salmon fishing (autumn, in the main stream); (2) The river banks: collecting of wild plants (spring to autumn); (3) The river terraces: deer hunting (autumn); plant collecting (spring to autumn); human habitation (all year); (4) The hillsides along the river course: dcer hunting (early winter, at or near the animals’ winter quarters); (5) The mountain region around the source of the river: bear hunting, specialized (spring and autumn); collection of elm bark for clothing (usually spring) .

Zones 1–3 were exploited from a single center, namely the permanent settlement which is usually situated on the edge of the river terrace. The outermost zones 4 and 5 were hunting areas, each exploited from a different hunting hut. Some Ainu, especially those living downstream, did not engage in bear hunts and consequently had no hunting hut in zone 5. Such a settlement group as Tonnika of the Azuma valley exploited the zone 4 directly from their permanent settlement since the Azuma valley is so narrow that zone 4 was within easy walking distance of the base camp.

Seasonal migrations to the hunting huts were generally undertaken by the male members of each household in the settlement. When men moved, their wives and children were left behind; the remaining family members had various tasks, of which the collection of edible plants in zones 2 and 3 was the most important. The periodic migration occurred each spring and autumn; men moved to the same hut to stay for about one month during each hunting season. Thus men shifted residence usually twice every year; however, the total period of their stay away from their homes was only about two or three months per year. The permanent houses in the settlement were occupied by the entire family for nine to ten months of the year and by housewives and their children nearly all the year.

From 1884 onward, the Japanese government encouraged the Ainu to take up agriculture. Each Ainu household was granted a plot of land, was given agricultural implements and seeds, and received technical instructions from official Japanese specialists. Consequently, they were suddenly transformed from a gathering life to a farming life. Even before this change, however, the coastal Ainu and part of the interior Ainu, probably under Japanese influence, had farmed by shifting cultivation, although on a very small scale. The old techniques were simpler and the farm work was usually performt;d by women. There are concrete data (Watanabe, 1964a) showing that those groups who practiced farming gathered wild roots with less frequency and in smaller quantity than those who did not. But there is no evidence that the introduction of the primitive (pre-1 884) agriculture significantly increased the stability of residence among them.

A similar distribution pattern of ecological zones and a high stability of residence are found among the Northern Paiute in Owens Valley (Steward, 1938). In both Hokkaido and Owens Valley, it is the presence of narrow valleys that permitted the maintenance of year-round residence.

The residential shift pattern depends upon ecology, that is, the relationship between habitat and subsistence technology. The ecological study of the Basin-Plateau tribes by Julian Steward (1938) clearly indicates that even within the same cultural group the residential shift pattern may vary a considerable degree according to the variation of habitat. Among the Upper Tanana of the Yukon, the pattern of residential moves varied even from band to band (McKennan, 1959).

If the variation of habitat can affect the residential shift pattern, there may be a possibility that some form of the residential shift pattern can be defined for populations of early man. Karl Butzer (1964) pointed out the existence in southwestern France of Mousterian and late Paleolithic caves which he interpreted as perennially inhabited dwellings, and suggested that the Neanderthalers of the less severe forest-tundra environments were able to maintain settlements based on permanently available animal resources. This is an indication that under certain favorable conditions even early man may have been able to inhabit certain sites with little shifting of residence. In this connection it is noteworthy that the mountain valleys of inland Alaska described by William Irving (1953) provide a set of biotic zones compactly arranged within small territories which suggest suitable conditions for Arctic hunters to maintain stable residence. Modern Alaskan groups have no year-round residence, but at least some of the Nunatarmiut maintained their quasi-permanent villages for the greater part of the year, perhaps nine months (Larsen and Rainey, 1948).

Variability of the Residential Stability at the Level of the Individual

An aspect of the complexity of the northern semi-nomadism is the fact that a residential shift does not always mean movement of the entire social group, since settlements are not always completely deserted. In other words, there is the variability in residential stability at the level of the individual as well as of the family and band.

An analysis of ethnographic reports would reveal that among many of the northern hunters there are some non-moving, sedentary individuals or families who occupy the same dwelling all year, while there are others who shift their residence. Such variation of residential stability is correlated with differentiation in subsistence activity at the level of the individual or the family, which will be mentioned below. Peoples who maintain some sedentary members at the base camp are those who make, instead of nomadic wanderings, seasonal expeditions from the permanent dwelling to other sites. Such definite evidence is found among the Ainu (Watanabe, 1964a), the Northwest coast I ndians (Drucker, 1951; Gunther, 1927) the Ingalik (Osgood, 1958), the Tareumiut (R. Spencer, 1959), and the Bering Strait Eskimo (Nelson, 1899). Indirect evidence suggests that perhaps other peoples, such as the Koryak (Jochelson, 1905–08), belong to this category. Groups without such sedentary members are those who make nomadic wanderings whether they have a permanent winter dwelling or not. The examples are the Copper Eskimo (Jenness, 1922) and the Upper Kolyma Yukaghir (Jochelson, 1926) .

As far as evidence is available, the sedentary persons appear to be the less able-bodied, the aged, women, and children. These persons among the Ainu and lngalik (Osgood, 1 940) were engaged in collecting and/or fishing in areas near their settlement while others were away from the settlements hunting and/or fishing at distant localities. But among the Ainu and Tareumiut (R. Spencer, 1959, 1965), at least, even some able-bodied men would remain perennially in their permanent settlements owing to the circumstances of gathering. Among the Ainu, some male members of a nuclear family seasonally moved to a mountain hut for bear hunting, while others remained in their settlement, engaged on fishing and deer hunting. Among the North Alaskan coastal Eskimo (R. Spencer, 1959, 1965), some nuclear families moved to the summer camps for fishing and hunting land mammals, while others remained in the winter villages hunting sea mammals. Even among nomadic wanderers such as the Nunamiut (R. Spencer, 1959) and the Chipewyan (Birkett-Smith, 1930), there is a considerable degree of occupational differentiation at the level of the individual or the family. R. F. Spencer (1965) pointed out the Tareumiut’s freedom to choose the summer

work best suited to their needs or their skills. To what extent such differentiation and the freedom of choice in subsistence activity exists among other food gatherers is a problem worthy of further exploration.

Among some northern food gatherers, as previously mentioned, there is a tendency for some individuals or families to remain in the same permanent dwelling all the year although the proportion of the moving population to the sedentary one is yet to be investigated. The non-moving members are less able-bodied persons or the aged, women, and children. Among nomadic food gatherers abandonment of the aged as well as the sick is a common practice (Simmons, 1945; Watanabe, 1966). Possibilities for the aged, the sick, and the infirm to survive or live longer may be greater among sedentary food gatherers than among nomadic ones. This implies that opportunities for such persons to remain perennially settled may have had adaptive significance (Watanabe, 1966). One of the possible effects is greater life span, which, in turn, would influence the age composition of the population. Another possible effect is in selective pressure. Sedentary life may give greater possibilities of survival to the sick and infirm or pregnant. These effects may vary and change according to the variation and change in residential stability as well as other ecological conditions inside and outside the dwelling. S. L. Washburn and I. DeVore (1961a) pointed out the importance of the horne base for populations of early man, and its evolutionary implication concerning hominisation. During the Pleistocene, variation in the residential shift pattern must have had some effects on the stability of the home base in both its earlier and later stages and consequently must have influenced human evolution (Watanabe, 1966).

Hunting as an Occupation of Males

The development of meat-eating and the establishment of a hunting way of life are regarded as important ecological factors in hominid evolution (Washburn and Avis, 1958; Washburn and Lancaster, Chapter 32, this volume; Zuckermann, 1933). A related problem is the consistent allocation of hunting to males and collecting to females. Sex differences in psychology and physiology alone would not be sufficient to explain this division of labor. The problem will be further pursued by making reference to data from modern huntergatherers.

Individualistic or Noncommunal Hunting of Larger Mammals

Hunting of small animals by women is not a rare phenomenon. Among such peoples as the Shoshoni (Steward, 1938) and the aboriginal Australians (Berndt and Berndt, 1964; Spencer and Gillin, 1927), it is, in fact, a woman’s occupation.

In case of relatively large mammals there is information from various peoples that women take part in communal hunts (Turnbull, 196sa, 1 965b). Occasional cases are not unknown of women hunting large mammals alone. Some Copper Eskimo women (jenness, 1922) hunted seal and the caribou occasionally; Ainu women and children (Watanabe, 1 964a) sometimes hunted deer with sticks, i ropes, and/or dogs when they had opportunities. But there is no society in which individualistic or noncommunal hunting of larger mammals is the socially recognized regular occupation of women. It is this individualistic hunting of larger mammals that is invariably the task of males. Communal hunts, however, do not always exclude females.

Among modern hunter-gatherers, exclusion of females from the individualistic hunting of larger mammals seems to be closely related to the making and using of hunting weapons, and associated economic and/or religious ideas. Women have no weapons of their own which are specially made to hunt animals. If they want to hunt they must do so without weapons or otherwise with some provisional weapons such as sticks. Rarely do they use specially made hunting weapons such as harpoons or spears, although these might be borrowed temporarily from males. Under these restrictions women’s hunting activities are confined to small animal hunts, communal hunts in which they take part in driving, and, very rarely, individual hunts of larger mammals.

Ethnographic data suggest that perhaps the development of hunting weapons and ideas associated with them is one of the factors relevant to the tendency towards the exclusion of females from hunting.

r-b-richard-b-lee-irven-devore-man-the-hunter-34.jpg
Figure 2. Seasonal changes in the activity fields of male and female Ainu.
Dife rentiation of the Activiry Field according to Sex

Another possible factor pertinent to the allocation of hunting to males is the differentiation of the activity field according to sex. The activity field means the area of land covered by each individual in his subsistence activities.

Among the Ainu (vVatanabe, Ig64a, 1964b), the male’s activity field is clearly and markedly differentiated from the female’s (Fig. 2). The male’s field extends far beyond that of the female and the two fields have little overlap. The extension of the male’s field is the result of hunting, while most of the female’s activity field is the area near the base camp covered in her collccting activities.

The degree of the differentiation of the two fields seems to vary from people to people. Among the Semang (Forde, 1934, 1954), who are mostly vegetarian and whose hunting is sporadic and confined to smaller game, the differentiation seems to be much less marked than among the Ainu. But among the Kuchin (Osgood, I 936a), men are active full-time hunters of large game, and the activity field of males markedly deviates from that of females even in the way in which camps are moved. A Kuchin woman trudges slowly on with her burden, following the direct line from one camp to another, while her husband carrying only a gun circles widely through the hills and covers many more miles in transit.

It seems to be certain that the differentiation of the activity field according to sex is a universal phenomenon among modern food gatherers. It may be one of the ecological characteristics of man. Owing to this differentiation the structure of man’s home range differs from that of nonhuman primates.

Table 2 Allocation of Hunting Tasks by Age

r-b-richard-b-lee-irven-devore-man-the-hunter-38.jpg

Plants stand still, but animals move; the distribution of specific plants does not always coincide with that of specific animal species. Owing to the ecology of biotic species it is usually difficult to get edible plants and large mammals in the same place at the same time. Consequently hunting ground for the animals and collecting fields for the plant food are not identical. Synchronous exploitations of both resources by one and the same group necessitates the splitting of the unit.

Data from the Ainu and other food gatherers imply the possibility of an interpretation of the process of the allocation of hunting to males in the following way: (1) long and active pursuit of larger mammals brought about the differentiation of the activity field into the collecting and the hunting fields; (2) the necessity of synchronous exploitation of these spatially separated fields forced -the group to split; (3) this splitting and the selection for hunting ability on the basis of physical and psychological differences resulted in the allocation of hunting to males and collecting to females.

It may be assumed that the selection for hunting skills began when early man took up long and active pursuits of larger mammals. The Yukaghir (Jochelson, 1 926) and Ingalik (Osgood, 1958), for instance, clearly exhibit the importance for larger mammal hunters of physical and psychological capacities such as the ability of continuous running and endurance. The sexual division of labor might be reversible in work other than individualistic hunting of larger mammals, but in this kind of hunting no reverse case is found at all.

Tendency toward Age Division of Labor

One of the most neglected phases of the economic life of food gatherers is age division of labor. Careful examination of ethnographic data reveals that among food gatherers there is a tendency for the more strenuous work to be allocated to younger members of the population study, although this division of labor is not so striking as the sexual division of labor. However, it is known that in hunting, at least, some northern food gatherers practice division of labor between the younger and the older. It is sometimes division of tasks within an occupation and sometimes a division of whole occupations. Although it is uncertain to what extent the division is rigid and fixed, it is apparent that the division of labor crosscuts the sexual division of labor to form an important part of the organization of food getting. In Table 2, the allocation of hunting tasks by age is cited for seven northern groups.

The age division of labor shows a tendency towards the differentiation of the activity field in which the younger men exploit more distant fields and the older men fields nearer to base camp.

Age division of labor must have some relationship to the age composition of the society concerned. In view of the shorter life span of early man (Vallois, 1962), here is an interesting problem in the history of that division of labor.

8. The Netsilik Eskimos: Adaptive Processes

Asen Balikci

The Netsilik Eskimos inhabit a vast area over 300 miles in length and over 100 miles in depth north and west of Hudson Bay.{2} The region is barren, covered with innumerable rocky hills, lakes, sea inlets, and vast stretches of flat tundra. The climate is arctic, characterized by extremely cold winters and short cool summers. The local tundra bears no trees and its vegetation consists of creeping shrubs, tufted grasslike plants, lichens, and mosses. When traveling through this cold desert in 1923, Knud Rasmussen counted 259 Eskimos of whom 150 were males and only 109 females. They lived a nearly traditional life, having had only minimal contact with Europeans up to that time and having obtained guns only a few years before. The Netsilingmiut hunted the caribou, the musk-ox, and the seal, they fished the salmon trout and, with the exception of some berry picking in August, did no gathering. The question asked in this paper is: What were the adaptive processes and survival strategies utilized by the Netsilik Eskimos in this extremely harsh environment? It is assumed here that adaptation can be analyzed at three distinct levels: (1) the technological, (2) the socioecological, and (3) the demographic.

Technological Adaptation

Netsilik technology consists of four major complexes, on the basis of the raw materials utilized: the snow-ice complex (snowhouses, ice-houses, ice-caches, etc.), the skin complex (clothing, kayaks, sledges, tents, etc.), the bone complex (tools and weapons), and the stone complex (cooking utensils). Netsilik technology is thus adaptive in two ways: first, through the ingenious manufacture of a large number of specialized artifacts from a small number of locally available raw materials; and second, in reference to the adaptive features of the finished artifacts related to specific environmental conditions. It is primarily these forms of easily visible technological adaptation that drew the admiration of ethnographers.

Ecological Adaptation

Order and Predictability in the Annual Round

Among the four socioecological processes of adaptation to be considered is the predictability and order implicit in the annual migration circuit (see map, Fig. I) . In winter, after the caribou had left the country, the Netsiling -muit had to rely almost exclusively on the seal for survival. Since the seal keeps open a large number of breathing holes, it was advantageous to have numerous hunters, with harpoons in hand, to control an equal number of holes. Large hunting parties had maximal chances for a speedy catch. Hence the rationale for extended igloo communities consisting of over sixty people and comprising several distantly related extended families. vVinter was the season of intense social life, spectacular shamanizing and other ceremonial activity (Fig. I, Camps 1 and 2).

r-b-richard-b-lee-irven-devore-man-the-hunter-39.jpg
Figure 1. Reconstruction of the Annual Migration Cycle of the Arviligjuarmiut, a Netsilik Subgroup, for the Year 1919.

Early in June when the seals came lying on the ice, the large winter camp broke up into smaller social units having usually an extended family as a core. Tents were set up on the shore and group sealing continued at the large breathing holes (Camp 3). In July, the restricted groups moved inland, fishing with the fishing harpoon and occasionally hunting caribou with the bow and arrow. After the first week of August, the stone weirs were built across rivers and the salmon trout running upstream were caught in large amounts with leisters (Camp 4). At the end of August began the vitally important caribou hunts from kayaks at the caribou crossing place (Camp 5). This was a collaborative activity involving a division of tasks between beaters and spearers. Early in October the restricted groups moved camp to some larger river and fished salmon trout with leisters through the thin river ice (Camp 6). As winter came on the larger groups re-formed for sealing through the breathing holes (Camps 1 and 2).

During spring, summer, and autumn the restricted social units were isolated in the tundra and social life was markedly less intense. The Netsilik migration route thus reveals a basic winter-marine and summer-inland dichotomy involving a dual pattern in social morphology in harmony with the classic analysis of Marcel Mauss on the seasonal variations of Eskimo society (Mauss, 1906). There is thus a clear correlation between the movements of society and the known distribution of game, permitting the exercise of prediction (Balikci, 1 964) . It clearly indicates an adaptive pattern, the migrating groups taking advantage of particular game concentrations through the different seasons.

Alternatives and Substitutions in Subsistence Strategy

The endless search for alternatives, the practices of substitution and ad hoc inventiveness constitute another class of adaptive strategi(‘s. Each seasonal hunt knows of some alternative. If the fall caribou hunts with spears from kayaks failed, the bow and arrow was used. In winter, musk-ox hunting could be substituted for sealing. In summer, iffishing with the lcister was unrewarding, the fishing harpoon or the sleeping hooks were employed. There always seems to have been a way out of a difIicult situation. The alert hunter had to take advantage of any changes in game availability and specific local conditions such as wind direction and topography and adapt his hunting strategy with ingenuity. In hunting, no conditions are static and quick ad hoc decisions wcre nccessary.

Flexibility of Group Structure

Constantly changing opportunities and pressures favored flexibility in residence and determined temporary fission of co-residential groups. At all seasons the co-residcntial groups functioned as production and distribution units, yet in various stress situations the community would split, separating extended families. Or a nuclear family would go on extensive fishing or trading trips. Or two brothers would hunt temporarily in different areas and thus maximize game returns, the possible failure of the first being compensated for by the success of the second. Later the original group would reassemble. Clearly such fissive practices are adaptive in character and appear to be common to many of the hunting peoples discussed at the symposium.

Sharing

Delayed consumption involving elaborate caching techniques and various sharing patterns equalized food distribution both through time and within the community. Sharing patterns ranged from informal gift giving and occasional or regular commensality to rigidly structured, community-wide division of seal meat. Such rigorous sharing was based upon a system of named dyadic partnerships involving all hunters in camp and cutting across kinship alignments. Thus an unlucky hunter was always certain by right to obtain a portion of the daily catch secured by any one of his partners (Van de Velde, 1956). Clearly such a system is adaptive. Camp fellows share together or starve together.

The adaptive processes described here can be best understood in relation to the almost continuous ecological pressure to which the Netsilik were subjected. Traveling and moving camps was a very arduous task. Lack of dog feed severely limited the keeping of dogs to only one or two per family. The heavy sledges had to be pulled or pushed by both men and women. Only very small children were allowed to sit on the sledge. Old people had to drag themselves behind, and were often left behind to sleep out on the ice if they had not caught up with the others. Seal hunting involved a motionless watch on the flat ice under intense cold maintained for many hours. In order to build the stone weir, the stones had to be carried in the ice-cold water for several days. Beating the fast-running caribou over great distances in the tundra was an exhausting task. Stalking the caribou with the bow and arrow also involved endless pursuits across the tundra, the hunters lying on the wet grass and trying to approach the game while hiding behind tufts of moss. The nomadism of the hunting life in this extremely rigorous climate imposed a constant strain on the humans. Hunting was a never-ceasing pursuit, the game had to be brought to camp at all cost, and the hunter had to stay out until a successful kill: “The man that is wise never lolls about idle when the weather is good; he can never know when bad days may eat up his meat caches and drive him and his family into starvation” (Rasmussen, 1931, p. 134). At a camp of Kuggup-pamiut, Rasmussen observed that for the whole of the winter twelve hunters caught about 150 seals: “this may well be said to have been dearly bought food.”

Demographic Adaptation

Such continuous pressures and hardships necessitated internal, sociodemographic controls. The most rigorous was female infanticide (Balikci, MS.). On King William Island, Rasmussen counted 38 girls killed out of 96 births, in a genealogical sample of eighteen marriages. He concluded: “Despite the high birth rate the tribe is moving towards extinction if girl children are to be constantly suppressed” (1931, p. 141). The Netsilingmuit considered life as harsh and short, and girls were thought to be less productive than boys who were the future hunters. An informant said to Rasmussen, “Parents often consider that they cannot afford to waste several years nursing a girl. We get old so quickly and so we must be quick and get a son” (1931, p. 140). In the same context, a Pelly Bay Eskimo pointed to his dog team and added, “Bitches simply don’t pull as hard as the dogs do.” This high rate of female infanticide generated the sex imbalance noted by Rasmussen and resulted in great difficulties in finding a wife. So anxious were the parents to secure future wives for their little boys that they betrothed them to any one of their female cousins (Balikci, 1963). Cousins, both cross and parallel, were preferred because the parents of the promised couple were closely related, trusted each other, and kept their word; their children thus had maximal chances to become husband and wife. Furthermore, suicide in crisis situations, invalidicide and senilicide were additional responses to harsh pressures (Balikci, 1960). We may consider these forms of demographic controls as adaptive, in the sense that unproductive members of society were eliminated, the size of the family adjusted to the capacity of the provider, and the survival chances of future hunters were maximized.

Discussion

It is assumed here that the combination of adaptive processes and controls functioned successfully under conditions of optimal ecological pressure. Of course we admit with Rasmussen that: “Everything is naturally focussed entirely and absolutely upon the requirements that the increasing search for the daily meat necessitates” (1931, p. 142). Yet with peak ecological pressure, disaster occurred. For the two years preceeding his journey of 1923, Rasmussen counted 25 deaths due to starvation, equivalent to 10 per cent of the tribal population. Order and predictability, together with flexibility, certainly existed, yet they did not leave the Netsilik with much leisure and evidently failed on some occasions to assure his survival.

A brief comment may be necessary concerning the notions of harsh environmen.t and ecological adaptation and pressure. Harsh environment refers more closely to objective geographic realities. In the case of the Netsilik area, it includes primarily a long winter characterized by intense cold, lack of timber, and shifting and occasionally unpredictable game concentrations. The term ecological adaptation is the effect of the most striking features of environment on human life, as evidenced in subsistence techniques, economic organization, and settlement patterns. It is obviously a dynamic process and relative to the technical equipment of a society. Ecological pressure is but one phase of the funeral adaptive process and implies diminution of food availability, and increased difficulty in securing a livelihood. Such periods of hardship were consciously recognized by the Netsilik who reacted accordingly by trying alternative strategies. The work/leisure time ratio may be considered as one objective criterion for measuring the degree of ecological pressure. Food availability over prolonged periods expressed in quantitative terms may be another such measure. Depending on a multiplicity of factors, pressure phases may be of greatly varying irregular occurrence, or they may be cyclical and seasonal. They may attain a critical level which results in starvation for productive members of society.

Now within the vast Eskimo area there are great regional differences in the processes of ecological adaptation. For instance, both northern Alaska and the Netsilik country may be considered to have harsh environments, yet the abundance of larger sea mammals in Alaska has allowed a much denser population to develop relatively sedentary patterns. Greater food availability seems to have reduced prolonged ecological pressure among the Alaskan Eskimos and made possible the growth of complex ceremonialism. In the Netsilik area, on the other hand, prolonged pressure phases appear to have been much more frequent and severe. The relatively low seal availability and the erratic character of caribou migrations seem to have been mainly responsible for this state of affairs. The Caribou Eskimos dwelling inland illustrate another case of precarious ecological adaptation with frequent pressure phases.

The Bushmen (Lee, Chapter 4, this volume) seem to enjoy a regular availability of resources allowing for considerable leisure time. The Bushmen are primarily collectors of plant foods and only secondarily hunters. In their region, pressure phases seem infrequent. Clearly the Netsilik Eskimos stand in sharp constrast to the Bushmen. The Netsilik case can be considered as an extreme example of human adaptation to a harsh environment. In our opinion, both Bushmen and Netsilik constitute extreme cases, and neither can be considered as typical as far as ecological adaptation of hunters-gatherers is concerned. It is rather obvious that in each geographic zone inhabited by hunters-gatherers, different adaptive forms are going to be found with different qualities of pressures. A regional ty. pology should have a greater heuristic value than blunt assumptions about global lack or presence of ecological pressure in hunting and gathering societies.

9. Discussions, Part II

9a. The Central Eskimo: A Marginal Case?

Speakers: Balikci, Damas, Eggan, Helm, Washburn.

DAMAS: I would like to question one or two points in Dr. Balikci’s paper. I think that he is right in indicating that winter communities comprise several extended families. But from the group charts that I have been able to reconstruct for the early 1 920’S, the families were connected by primary ties. And in an earlier period, the Netsilik winter groups were considerably larger, probably averaging 100–150, implying a more substantial subsistence base. Between 1900–20, about 150 people moved down to trade at Repulse Bay, and the Chesterfield area, which accounts for the smaller band size of the later period.

Dr. Balikci’s treatment of alternative means of food procural seems to imply that they may all be equally fortunate choices. I suggest instead that in the winter, hunting through the breathing hole was the only practical way for the group to survive. At times, the Netsilik made long trips for polar bear, and in extreme cases they jigged for tomcod. But these were not reliable means of survival.

I think that the quotation from Rasmussen in Balikci’s paper regarding the harried existence of the Kuggup is not altogether typical of the Central Eskimo. There seems to be quite a variation within the Netsilik area in the game potential. The Thorn Bay, Bellot Strait, and Queen Maud Gulf areas were probably better endowed than the area in which Rasmussen found the Kuggup.

Finally, apropos of the deaths from starvation mentioned in Dr. Balikci’s paper, it might be interesting to note that starvation was usually related to bad weather. In the particular case cited, there was a late freezing of the sea accompanied by storms, which made the ice rough and unsuitable for normal winter hunting.

Balikci: Concerning the first point in reference to the primary ties linking extended families, I am in agreement with Dr. Damas. This is to be expected for small, bilaterally organized communities. My aim was to emphasize that on many occasions extended families among the eastern Netsilik behaved as discrete social units, independent of the close ties between particular members of different extended families. As for the restricted size of winter settlements in the early 1920’S, emigration may probably account for this, at least partially. Yet a figure as high as 150 individuals seems difficult to accept in view of the limited subsistence base. More probably, large groups occupying an area were divided into two or three camps located close by to allow visiting. It seems reasonable to assume that there was considerable flexibility in the size and duration of winter settlements. As for my data, I took the four or five years before Rasmussen’s journey of 1923 as baseline.

I did not imply that alternative means of food procural were as good as the main subsistence technique for a particular season and site. My point is that substitutes were available as a temporary device to avoid disaster. The effectiveness of the alternatives depended on particular local conditions. The very use of the terms of alternatives and substitutes implies their quality as “second-best” choices.

In reference to regional variation in game availability within the Netsilik area, I agree with Dr. Damas’ comment. Of course there was considerable variation: Pelly Bay was good for fish, Thom Bay was good for seals, the area behind Tksiaktorvik was good for caribou, and the western Netsilik could obtain a good deal of tomcod. But I wished to produce a global scale for the Netsilik and did not have place for regional variations.

HELM: I refer to a paper published some years ago in which Bartholomew and Birdsell (1953) said, in effect, that it does not matter how well you eat during eleven months of the year if you have no food during the twelfth month. I think that this holds true in terms of generations. If a group maintains or increases its population for, say, three generations, but there is then a severe population loss, there is apt to be a cutback in incipient development and even the continuity of cultural forms. Has this happened among the Central Eskimo?

BALIKCI: I t is difficult to answer this for very long time depths. However, with the exception of occasional cases of people getting lost in traveling or getting caught in bad weather as in the example from Rasmussen-widespread starvation of a magnitude required to induce culture loss, is a very rare phenomenon.

WASHBURN: Dr. Balikci has shown that there can be a real struggle for existence in some extremely marginal areas, while, on the other hand, others have spoken about hunters that were not living marginally and who did not harvest all the food they might. I would like to suggest that these are not necessarily contradictory notions. If we look at what happened to the idea of the struggle for existence in genetics, we find that it started in the conception of an all-out tooth-and-claw struggle and then changed to a consideration of what works in the long run.

I think this is the kind of shift that we need to make when we look at human economies. It is not so important to worry whether people are going to die tomorrow or whether they are eating to their maximum capacity. What we want to know and understand is the kind of system that works and keeps on working. For a quantitative example, Bourliere (1963) estimated the amount of meat on the hoofboth in the Congo forest and in the savanna on the edge of the forest. The amount in the forest proved to be only about 5 per cent as great as that in the savanna. So objectively speaking, there is a great deal more to be harvested in the savanna.

I would like to shift for a moment to Paul K. Anderson’s study (a seminar given to the Zoology Department, Berkeley, in 1965) of mice in Canada living near great grain depots. Now, from the point of view of mice, these depots represent an infinite supply of food, so there is no reason why the mouse population should not soar to infinity. But this is not what happens at all. The mice build up a group of about twenty, then subadult and young adult males migrate away from the food supply. Here is a built-in genetic system devised, so to speak, to keep these animals moving out. They have not been evolved to eat tons and tons of grain and are, therefore, completely unable to adjust to the presence of it. I think that the lesson of this is that an economy may be completely adaptive when those involved are eating only 20–30 per cent of the available food. It is not adaptive to eat the complete food supply, since-aside from the troubles brought on even by a minor variation in supply-anything like a major seasonal variation will break down the system. So if through warfare the Pygmy group occupies twice the area that it needs for its normal food supply, this may actually be an adaptive situation in the long run. We have, then, to abandon the notion of an immediate struggle-even though it shows up among groups like the Eskimo-in favor of a far more subtle approach, in which we try to determine just what percentages of the food supply would be used in a good adaptation by a particular group.

EGGAN: I want to follow up one of Washburn’s points on adaptation versus flexibility. There is a series of norms or sanctions with regard to the sharing of food. I t seems to me that they often differ with the kind of foodmeat, fish, or vegetable-—with the size of the animals involved, and with the latitude of the location. I remember looking at the Siberian central Asian data, and in that area foodsharing increases to the north and decreases to the south. I wonder whether these patterns of sharing, ifwe worked them out, might not give us something relevant to the problem of adaptation, even an index to it.

9b. Notes on the Original Affluent Society

Speaker: Sahlins

Sahlins: I should like to pick a point that is embedded in Colin Turnbull’s discussion, implied in Suttles’, elaborated in Lee’s, and given ultimate explanation I think in Washburn’s comments, particularly his suggestion that a 20–30 per cent usc of productive capacity may prove quite adaptive over the long run. What I want to talk about is, as it were, the inner meaning of running below capacity, the consequences for the quality of hunting-gathering economical life.

If economics is the dismal sciencf’, the study of hunting-gathering economics must be its most advanced branch. Almost totally committed to the argument that life was hard in the Paleolithic, our textbooks compete to convey a sense of impending doom, leaving the student to wonder not only how hunters managed to make a living, but whether, after all, this was living? The specter of starvation stalks the stalker in these pages. His technical incompetence is said to enjoin continuous work just to survive, leaving him without respite from the food quest and without the leisure to “build culture.” Even so, for his efforts he pulls the lowest grades in thermo-dynamics-Iess energy harnessed per capita per year than any other mode of production. And in treatises on economic development, he is condemned to play the role of bad example, the so-called “subsistence economy. ”

It will be extremely difficult to correct this traditional wisdom. Perhaps then we should phrase the necessary revisions in the most shocking terms possible: that this was, when you come to think of it, the original affluent society. By common understanding an affluent society is one in which all the people’s wants are easily satisfied; and though we are pleased to consider this happy condition the unique achievement of industrial civilization, a better case can be made for hunters and gatherers, even many of the marginal ones spared to ethnography. For wants are “easily satisfied, “ either by producing much or desiring little, and there are, accordingly, two possible roads to affluence. The Galbraithean course makes assumptions peculiarly appropriate to market economies, that man’s wants are great, not to say infinite, whereas his means are limited, although improvable. Thus the gap between means and ends can eventually be narrowed by industrial productivity, at least to the extent that “urgent” goods became abundant. But there is also a Zen solution to scarcity and affluence, beginning from premises opposite from our own, that human material ends are few and finite and technical means unchanging but on the whole adequate. Adopting the Zen strategy, a people can enjoy an unparalleled material plenty, though perhaps only a low standard of living. That I think describes the hunters. I

The traditional dismal view of the hunter’s fix is pre-anthropological. I t goes back to the time Adam Smith was writing, and maybe to a time before anyone was writing. But anthropology, especially evolutionary enthropology, found it congenial, even necessary theoretically, to adopt the same tone of reproach. Archeologists and ethnologists had become Neolithic revolutionaries, and in their enthusiasm for the revolution found serious shortcomings in the Old (Stone Age) Regime. Scholars extolled a Neolithic Great Leap Forward. Some spoke of a changeover from human effort to domesticated energy sources, as if people had bet’n liberated by a new laborsaving device, although in fact the basic power resources remained exactly the same, plants and animals, the development occurring rather in techniques of appropriation (i.e., domestication. Moreover, archeological research was beginning to suggest that the decisive gains came in stability of settlement and gross economic product, rather than productivity of labor.)

I. I realize that the Netsilik Eskimo as described by Balikci constitute an exception in point. I shall not speak to this case here, but second Richard Lee’s explanation (or disposition) of it (Chapter 4, this volume).

But evolutionary theory is not entirely to blame. The larger economic context in which it operates, “as if by an invisible hand,” promotes the same dim conclusions about the hunting life. Scarcity is the peculiar obsession of a business economy, the calculable condition of all who participate in it. The market makes freely available a dazzling array of products all these “good things” within a man’s reach-but never his grasp, for one never has enough to buy everything. To exist in a market economy is to live out a double tragedy, beginning in inadequacy and ending in deprivation. All economic activity starts from a position of shortage: whether as producer, consumer, or seller of labor, one’s resources are insufficient to the possible uses and satisfactions. So one comes to a conclusion-“you pays your money and you takes your choice.” But then, every acquisition is simultaneously a deJ’!rivation, for every purchase of something is a denial of something else that could have been had instead. (The point is that if you buy one kind of automobile, say a Plymouth fastback, you cannot also have a Ford Mustang-and I judge from the TV commercials that the deprivation involved is more than material.) Inadequacy is the judgment decreeed by our economy, and thus the axiom of our economics: the application of scarce means against alternate ends. We stand sentenced to life at hard labor. It is from this anxious vantage that we look back on the hunter. But if modern man, with all his technical advantages, still hasn’t got the wherewithal, what chance has this naked savage with his puny bow and arrow? Having equipped the hunter with bourgeois impulses and Paleolithic tools, we judge his situation hopeless in advance.

Scarcity is not an intrinsic property of technical means. It is a relation between means and ends. We might entertain the empirical possibility that hunters are in business for their health, a finite objective, and bow and arrow are adequate to that end. A fair case can be made that hunters often work much less than we do, and rather than a grind the food quest is intermittent, leisure is abundant, and there is more sleep in the daytime per capita than in any other conditions of society. (Perhaps certain traditional formulae are better inverted: the amount of work per capita increases with the evolution of culture and the amount of leisure per capita decreases.) Moreover, hunters seem neither harassed nor anxious. A certain confidence, at least in many cases. attends their economic attitudes and decisions. The way they dispose of food on hand, for example-as if they had it made.

This is the case even among many present marginal hunters-who hardly constitute a fair test of Paleolithic economy but something of a supreme test. Considering the poverty in which hunter and gatherers live in theory, it comes as a surprise that Bushmen who live in the Kalahari enjoy “a kind of material plenty” (Marshall, 1961, p. 243). Marshall is speaking of non-subsistence production; in this context her expiication seems applicable beyond the Bushmen. She draws attention to the technical simplicity of the non-subsistence sector: the simple and readily available raw materials, skills, and tools. But most important, wants are restricted: a few people are happy to consider few things their good fortune. The restraint is imposed by nomadism. Of the hupter, it is truly said that this wealth is a burden (at least for his wife). Goods and mobility are therefore soon brought into contradiction, and to take liberties with a line of Lattimore’s, the pure nomad remains a poor nomad. It is only consistent with their mobility, as many accounts directly say, that among hunters needs are limited, avarice inhibited, and—Warner (1937 [1958], p. 137) makes this very clear for the Murngin—portability is a main value in the economic scheme of things.

A similar case of affluence without abundance can be made for the subsistence sector. McCarthy and McArthur’s time-motion study in Arnhem Land (1960) indicates the food quest is episodic and discontinuous, and per capita commitment to it averages less than four hours a day (see accompanying Figs. 1 and 2). The amount of daytime sleep and rest is unconscionable: clearly, the aborigines fail to “build culture” not from lack of time but from idle hands. McCarthy and McArthur also suggest that the people are working under capa-city-they might have easily procured more food; that they are able to support unproductive adults-who may, however, do some craft work; and that getting food was not strenuous or exhausting. The Arnhem Land study, made under artificial conditions and based only on short-run observations, is plainly inconclusive in itself. Nevertheless, the Arnhem Land data are echoed in reports of other Australians and other hunters. Two famous explorers of the earlier nineteenth century made estimates of the same magnitude for the aborigines’ subsistence activities: two to four hours a day (Eyre, 1845, 2, PP. 252, 255; Grey, 1841, 2, pp. 261–63). Slash-and-burn agriculture, incidentally, may be more labor-intensive: Conklin, for example, figures that 1,200 man hours per adult per year are given among the Hanunoo simply to agriculture (Conklin, 195 7 p. 151: this figure excludes other food-connected activities, whereas the Australian data include time spent in the preparation of food as well as its acquisition). The Arnhem Landers’ punctuation of steady work with sustained idleness is also widely attested in Australia and beyond. In Lee’s paper he reported that productive members of !Kung Bushman camps spend two to three days per week in subsistence. We have heard similar comments in other papers at the symposium. Hadza women were said to work two hours per day on the average in gathering food, and one concludes from James Woodburn’s excellent film that Hadza men are much more preoccupied with games of chance than chances of game (Woodburn and Hudson, 1966).

r-b-richard-b-lee-irven-devore-man-the-hunter-11.jpg
Figure 1. Hours per day in food-connected activities: Hemple Bay group (constructed from McCarthy and McArthur, 1960, p. 190).
r-b-richard-b-lee-irven-devore-man-the-hunter-12.jpg
Figure 2. Hours per day in food-connected activities: Hemple Bay group (constructed from McCarthy and McArthur, 1960, p. 191).

In addition, evidence on hunter-gatherers’ economic attitudes and decisions should be brought to bear. Harrassment is not implied in the descriptions of their nonchalant movements from camp to camp, nor indeed is the familiar condemnations of their laziness. A certain issue is posed by exasperated comments on the prodigality of hunters, their inclination to make a feast of everything on hand; as if, one Jesuit said of the Montagnais, “the game they were to hunt was shut up in a stable” (Le Jeune’s Relation of 1634, in Kenton, 1927, I, p. 182). “Not the slightest thought of, or care for, what the morrow may bring forth,” wrote Spencer and Gillen (1899, p. 53). Two interpretations of this supposed lack of foresight are possible: either they are fools, or they are not worried-that is, as far as they are concerned, the morrow will bring more of the same. Rather than anxiety, it would seem the hunters have a confidence born of affluence, of a condition in which all the people’s wants (such as they are) are generally easily satisfied. This confidence does not desert them during hardship. It can carry them laughing through periods that would try even a Jesuit’s soul, and worry him so that-as the Indians warn-he could become sick:

I saw them [the Montagnais] in their hardships and their labors, suffer with cheerfulness I found myself, with them, threatened with great suffering; they said to me, “We shall be sometimes two days, sometimes three, without eating, for lack of food; take courage, Chihine, let thy soul be strong to endure suffering and hardship; keep thyself from being sad, otherwise thou will be sick; see how we do not cease to laugh, although we have little to eat” (Le Jeune’s Relation of 1634, in Kenton, 1927, I, p. 129).

Again on another occasion Le Jeune’s host said to him: “Do not let thyself be cast down, take courage; when the snow comes, we shall eat” (Lejeune’s Relation of 1634, in Kenton, 1927, I, p. 171). Which is something like the philosophy of the Penan of Borneo: “If there is no food today there will be tomorrow”—expressing, according to Needham, “a confidence in the capacity of the environment to support them, and in their own ability to extract their livelihood from it” (1954, p. 230).

9c. Does Hunting Bring Happiness?

Speakers: Alland, L. Binford, Crocker, Helm, Hiatt, Marshall, Owen, Suttles, Turnbull, Williams, Woodburn.

Hiatt: I think one may follow even further the point of Marshall Sahlins’ paper and raise a question about the mentality of huntergatherers. Lee and DeVore have said that one of the points that we might consider would be “nomadic style. “ In light of what Sahlins said, I wonder whether we ought also to talk about the “parasitic mentality.” What he said certainly fits the aborigines; they are, so to speak, a happy-go-lucky people. I wonder whether this description applies to the other hunters and gatherers that we have heard about during this conference.

Williams: I think I may have an aberrant group in many senses; they seem not to be so well off as some of the groups described here; they hunt and are hungry most of the time; and they really do not laugh a lot. However, I will give Marshall Sahlins the point that they do not worry much about it or try to change the situation.

HELM: I do not think the thesis of confidence in the yield of tomorrow holds for the boreal forest peoples. Rather, my reading of northern Indian equanimity is that it is fatalistic rather than optimistic: Ifeither the day or the morrow does not provide for itself, there is little one can do about it.

Owen: No one has ever accused the Yuman-speaking nomads of being especially jolly. The group that I worked with in Baja California are now deculturated and are peasant farmers in a certain sense. Perhaps their concern with farming has some relevance to whatJune Helm has just suggested, because it is precisely tomorrow that they are unconcerned about, since they have no real control over it. They muster enough energy to plow a field and sow it, but then their uncertainty about the future creeps in and they kt it go. The cattle cat the crops, and that’s God’s will, by God. Now, this would suggest to me that these populations— who have never known anv great abundance— are unconcerned with tomorrow because they cannot control it, rather than because they expect that it will take care of itself.

Suttles: In response, again, to Hiatt’s question, the Northwest coast Indians have never been called happy-go-lucky, so far as I know. Of course, they do plan for the future and may have been trapped into this, in Sahlins’ terms. It is this, I think, that makes them different from others that we have discussed.

ALLAND: As I understand it, the Naskapi also worry and care about the morrow, and they handle this in a magical way by throwing a scapula into the fire and thus determining the direction to be taken in future hunts. And Omar Moore (1957) has suggested that this incidentally helps them to maximize hunting, since it randomizes the hunting territories they exploit so that the animals do not become accustomed to being hunted in a particular place. At any rate, whether or not this model works, there is evidence again that a hunting group may be anxious about the future and will try to do something about the morrow.

BINFORD: It might be interesting to speculate about the “happiness” of the people responsible for the archeological remains recently uncovered by R. S. MacNeish (1964) in the Tehuacan Valley. These people were coping with their environment during the period of initial use of domesticated plants. Judging from the archeology, they lived in an arid area which did not compare favorably with the Bushman habitat, particularly in regard to the availability of native tuberous roots and other dependable wild plant sources. The impetus to invest in the development of storage technology and in the horticultural perfection of certain plants must have occurred in the context of “some concern” for the future. I was pleased to hear Dr. Suttles say that the hunters of the Northwest coast also showed concern, since they depended heavily on resources which were exploited for limited periods and stored for future consumption. The TehuaGi-nos’ similarly increased the time utility of their resources through storage and preservation techniques. J suspect that the people of the Tehliacan Valley were not very “happy” in the sense suggested earlier.

The happiness factor may not, after all, be so much a direct function of hunting; happiness should vary directly with the degree to which subsistence input levels are maintained in a steady state despite seasonal fluctuations in yield from the standing crop of available resources. Increased sedentism together with storage and preservation techniques, as in the cases of the TehuaGlnos and the Northwest coast Indians, was, at least initially, a response to the failure to maintain such a steady state. It is expectable that such adaptations would be coupled with “concern for tomorrow. ”

In contrast, one might expect to find a more happy-go-lucky attitude among groups which have achieved a homeostatic relationship with regard to food resources. Such a group might move from one resource area to another, consuming immediately available products; or a group might gain time utility from resources through storage and preservation techniques and stabilize population below the point at which exploitation of the standing crop would threaten productivity. The point here is that “happiness” in the sense of little cate for tomorrow would be more a function of homeostasis than of hunting per se.

lf we can judge from the tremendous conservatism and stability of many Pleistocene hunting adaptations and from the stability and tenacity of some groups of modern hunters, there are many hunting groups which have achieved a balanced equilibrium with their environment. I am suggesting that it is this tendency of hunting adaptations to reach homeostasis which makes for the “happy hunter.”

CROCKER: I thought we had buried the Gespeaking peoples today with our emphasis on pure hunting subsistence, but I find that the hunter’s mentality now being described fits the Canella. Yet they have a great deal of agriculture and may, in earlier times, have had even more. This raises a problem. If we are going to talk about the hunter’s mentality, maybe we have to talk about peoples who have other modes of subsistence—even, perhaps, fully agricultural peoples. It is a question of the yardstick that we use and the purposes we are pursuing.

WOODBURN: It may be worth mentioning a lack of commitment different from the lack of commitment to the morrow and yet, Iperhaps, related to it. The Hadza are strikingly uncommitted to each other; what happens to other individual Hadza, even close relatives, does not really matter very much. People are often very affectionate to each other, but the affection is generally not accompanied by much sense of responsibility. If someone becomes ill, he is likely to be tended only so long as this is convenient.

Now, in the literature on hunters and gatherers, abandonment of the sick is quite frequently mentioned and is almost invariably interpreted as deriving directly from ecological pressures. The notion is that an individual is abandoned only through necessity, when the survival of other people is at stake. But this explanation does not make sense of the case material I collected among the Hadza and I strongly suspect that it also does not adequately explain the incidence of abandonment in most other societies from which the custom is recorded. In one of the cases of which I have records, a boy became paralyzed below the waist, probably from poliomyelitis. He was carried for quite a number of camp moves. Then, on one occasion, the people of the camp moved to a site where they had expected to find water and found none. They were perhaps only about four or five miles from water, but they had reached the point at which it had become rather tedious to carry him for the unexpected additional move when they were rather tired and rather thirsty. So they left him behind with food and water and with his bow and arrows-not that these were of much use to him—and he did not survive. Those who abandoned him included his own mother and certain other close relatives. A similar lack of commitment tc other people is apparent in many other contexts.

For an explanation, I think we should look beyond the obvious ecological factors. We should give far more attention to the consistency of this behavior with the degree to which social relationships bind individuals to each other. With a few exceptions, a Hadza does not depend on specific individuals standing in particular relationships to him for access to property or to adult status, or for assistance in cooperative activities. In this respect the Hadza and similar hunters and gatherers differ very fundamentally from most other tribal societies.

I would suggest that much of the sort of behavior which Marshall Sahlins has described— the listlessness, the apathy,, the merry squandering of windfalls, the lack of concern for the morrow-is intimately bound up with this relative lack of responsibility to specific other individuals, the relative absence of binding obligations to them. I am not suggesting for one moment that this applies to all hunters and gatherers but I certainly do think that it may apply to many of them. It may also have applied to many ancient hunters and we ought to bear this in mind when considering why their populations apparently rerr.ained stable over so vast a period.

Turnbull: I would like to support Woodburn’s observations with data on the Ik. There is an almost exact parallel, and I have seen ‘abandonment without any kind of emotion whatsoever. One person was shot in the stomach in a raid at about one o’clock in the morning and was still conscious when I was first told about it at six o’clock in the morning, and people had done nothing about him except to fold him into the death position. They only began to take an interest in him when I began to wonder what I should do about it.

I think this also bears on what Hiatt said about the mentality of the hunter. Both the Pygmies and the Ik laugh a lot; in fact, they get quite hysterical. But this does not mean that they do not also cry. At any rate, they are violently emotional. They fight continuously, although the fighting is without malice. In some cases, a Pygmy camp may be characterized by innumerable disputes, many of which are almost deliberately created. They are created with the specific objective of concealing a real dispute.

One last point, not previously mentioned, is that both these hunting communities are characterized by an almost complete lack of magic, witchcraft, or sorcery. I would like to hear more about this from students of other societies.

MARSHALL: In response to Colin Turnbull’s last remark, I would like to say that the!Kung we worked with were free from a belief that there were witches among them. They had an apparently vague idea of a sorcery-like power, but spoke of other people far away having such power and using it for evil.!Kung informants insisted that their “medicine men” only tried to cure and protect with their “medicine. ”

9d. Hunting vs. Gathering as Factors in Subsistence

Speakers: Alland, Bicchieri, L. Binford, Lathrap, Laughlin, R. Lee, Marshall, Struever, Washburn

BICCHIERI: I would like to point out that in my doctoral dissertation on the ecology of food-gathering peoples (1965) I utilized a sample of thirty societies very similar, in the choice of units, to the one used by Lee. My analysis also emphasized the need for determining the actual dependence of a specified group on hunting versus gathering. It appears to me that it is essential, for meaningful cross-cultural comparison, that the relative amount of a group’s dependence on hunting as contrasted to gathering to be specified. The term “hunters”-as used by J. H. Steward-tends to obscure the actual exploitative condition, and, as a consequence, inhibits inferences regarding socioeconomic covariations. One basic implication, for example, lies in the fact that the chances of productive failure, everything else being equal, are halved by a mixed hunting-gathering exploitation. Practicing such mixed exploitation, a group can, more often than not, fall back on the alternative technique when necessary. Thus the Kalahari Bushmen, with a hunting and gathering subsistence base are likely to resist environmental failures more readily than, for instance, the Copper Eskimo who have to rely solely on hunting.

STRUEVER: This distinction between hunting and gathering interests me. It may be useful for solving some problems, but irrelevant to others. For example, if our object is to understand the considerable complexity of a particular collecting group, this is a problem of productive potential of their environment visa-vis plants and/or animals and of the efficiency of their exploitative. and maintenance technology. A hunting-gathering distinction has no value for solving this particular prQblem.

Also, we have not really defined what we mean by the hunting-gathering distinction. Are we talking about differences in the food substances themselves—plant materials vs. flesh, or is the distinction a technological one, i.e., hunting and gathering technologies can be distinguished from each other? If the distinction is technological, would you call the collecting of migratory waterfowl by means of traps or cast nets gathering or hunting? Would you call herding fish onto mud flats or draining lakes and then collecting the fish population, gathering or fishing?

But the problem of distinguishing hunting from gathering seems peripheral to our major interests which center more on the productivity of an environment vis-a-vis nondomesticant resources and how human groups structure themselves socially and demographically to use a technology of measurable efficiency to exploit this environment.

LAUGHLIN: I t is interesting to notice the kinds of information that are discarded if the catching of sea mammals is classified as fishing, rather than hunting. One of these losses is in our understanding of the development of some kinds of technology. Now, most of the sea mammals are hunted from kayaks, which have enormous advantages over open boats like umiaks (which, in turn, have great advantages over intertidal zone collecting). When it comes to input-output analysis, we must count the extensive training of a child to become.a kayak hunter. A second major time investment is in the scanning operation; though this is still quicker and more efficient from a kayak than, say, scanning on foot along a coastline.

Binford: While I am in complete sympathy with obtaining quantitative data on subsistence, I think that the emphasis on the percentage contribution of gathering versus hunting to total subsistence may divert us from a major point of ecological studies—the dynamics of human adaptations in various environments. For example, if we are concerned with trying to understand human radiations

from a relatively limited range in the basal Pleistocene, it becomes critical to understand variations in the structuring of subsistence resources through the phasing of an annual cycle among the major environmental zones which man successively penetrated. We might reasonably expect that hunting played very different roles in particular habitats. Man’s radiation was, in general, from a relatively dry semitropical park savanna through zones of increasing year-‘round dampness and annual cold. In environments of the latter type permanent resident fauna include a relatively high proportion of predators; herbivores are either seasonal residents, or are very rare in species numbers, or they are represented by species which hibernate during the months when plants are dormant or blanketed with snow. In penetrating the more northerly environmental zones, man must have put greater reliance on hunting (predation) and on the development of storage and preservation techniques. A small amount of available meat during periods of plant food scarcity might have been more crucial to survival than the vast quantities of plant foods available during times of plenty.

To point out that non-agriculturalists living in relatively arid and sub-tropical environments are heavily dependent on plant foods tells us little about the importance of hunting in the evolutionary history of man and culture, nor does it provide a basis for generalizing about the role of hunting in adaptations to different environments.

ALLAND: I agree with Lee’s input-output model, and I think it makes sense in terms of nutritional requirements. People require a rather large number of calories in order to subsist, but they require only a rather small amount of protein-45-50 grams per day of animal protein and perhaps a bit more in vegetable protein. These things should also be measured. It seems to me that the investment of calories in getting animals may be greater than the output in calories; yet if the caloric output of gathering vegetables shows a surplus, you find a surviving society.

LEE: To carry on the points made by Binford and Alland, I might say that when we think of man the hunter, we generally think of the higher latitudes: the temperate and arctic zones; while the concept of man the eclectic subsister calls to mind the middle and low latitudes, i.e., the tropical zones. This eclecticism and emphasis on vegetable foods appears even in the game-rich environments, as the Hadza case shows.

To take up a physiological point: eating a lot of meat just does not make physiological sense in the Bushmen case, because it takes five times as much energy (with heat generated in proportion) to process by digestion a unit of meat as it takes to process a unit of vegetable food (Bourliere, 1964, p. 336). So in a cold area, it makes good sense to eat a lot of meat; but in a hot area, the reverse is true.

Washburn: Another nutritional point is that man the animal is adapted to certain kinds of food. He has to have certain things in his food supply, and this makes measurements by quantity tricky. We are dealing with an animal that has relied on meat as part of his diet for at least a couple of million years, so 20 per cent, or even a smaller quantity of meat, may be a very important constituent of the total diet.

LATHRAP: I would like to inject a word of caution here. Most of the hunters whom we have been discussing so far during this conference are peoples who live in environments not particularly favorable to hunting, or who share their territory with more complex groups. The Ona bands of northern Tierra del Fuego survived into the first decade pf the present century in an unacculturated condition. Here we find people with good hunting territories which they j’ealously defended -from all outsiders. The larger part of their caloric intake was from game (the guanaco being the prime source), and only minimal use was made of vegetable foods.

Interestingly enough, all of the classic stigmata of “hunting peoples,” which have been brought into question at this symposium, were strongly developed among the several Ona bands, which were in fact patrilineal. Patrilineages were clearly defined and strictly exogamous. Each held a sharply delineated territory with absolute rights over the included economic resources. The facts of the case are reasonably clear since we have a good account of the culture written by Bridges (1949) who actually lived with Ona bands while the culture was still flourishing. There are also more formal

and detailed ethnographies culled from the memories of Ona informants who had participated in the aboriginal situation (Gusinde, 1931–39; Lothrop, 1928).

With regard to the question of the “hunter’s mentality” the Ona are particularly instructive. Ona culture stressed emotional control, and Bridges’ account makes it clear that this particular ethical imperative was observed far more often than it was breached. There were a number of ingenious institutions which were effective in minimizing intraband hostilities and in channeling agression toward outsiders or toward wives, who in strict sense were outsiders. The Ona both preached and practiced the support of the aged.

When we consider projecting cultural traits and modal personality structures backward to a time when most hunters presumably found themselves in favorable circumstances, we might find the example of the Ona more instructive than that of the handful of still extant hunting groups who are scarcely making a go of it as hunters.

MARSHALL: I hope that there will be more discussion about the attitudes of huntinggathering peoples toward gathering. Assuming that there is;nore food to be gathered, why do they not gather more at a time? The Hadza gather for only two or three hours a day, the!Kung, perhaps more hours, but only on an average every other day, or sometimes one in three days. It has been suggested that because they do not have to work every day they can be said to have an “affluent society. “ This is a bon mot, but does not add to the understanding of the reasons. I have pointed out that the!Kung we worked with are all thin and that they constantly expressed concern and anxiety about food. There must be reasons why they do not gather and eat more. I think that energy for digging runs out and the daylight hours come to an end, for one thing. It has been suggested that they cannot eat more roots, berries, and seeds than they do, because the roughage is too much. Also, I believe we might look more into a possible social reason. If a woman gathered very much more than her family needed at a given time, would it turn out that she was working for others? Would she draw envious attention upon herself and be blamed for not being more generous if she had an excess of food and kept it for herself? (!Kung custom requires that meat be shared, but not veldkos.)

9e. Measuring Resources and Subsistence Strategy

Speakers: Alland, L. Binford, Birdsell, Deevey, Eggan, Gardner, Sahlins.

BIRDSELL: There has been a great deal of discussion about carrying capacity and productive potential and about how much of a given food source could be reaped or killed. I suggest that we really do not know this in any specific case. I do not think that ecologists know what the combined human/predator load would be for a given anima! in a given situation. But, against this, there is considerable evidence that as one food source is overstrained, people turn to another in compensation. This is particularly true of shellfish in certain cases where one species is over-reaped and there is a change to a less favored species. In other words, the empirical proof is testable, but the theoretical description of what constitutes overdraft would be difficult. We would have to know the total ecology of the total biota to come to a specific statement, a task which probably would not be worth the trouble.

DEEVEY: By wayofpoking a little fun, let me note as an ecologist that it is refreshing to find that anthropologists have now discovered for themselves that the Arctic tundra, the temperate deciduous forest, the tropical savanna, and the tropical rainforest are not all the same kinds of environment, even from the point of view of subsistence. That they are different seems self evident to an ecologist, although certainly we have had trouble quantifying these differences. It is difficult to measure what we call productivity, and even when we measure with modest success the caloric content of the vegetation, the apportionment of the caloric content among the various species is not easy to measure. So we have no clear idea ecologically as to what we mean by the change in diversity between arctic, temperate, and tropical conditions. It is refreshing, then, to see anthropologists now measuring quantities that bear on this.

Let me frame a question for ecologically minded anthropologists to answer. A few years ago I naively thought it might be good to total up the measured anirnal productivity in a temperate deciduous forest, just by way of seeing in theory how much human protoplasm a given area of the forest could support on a hunting and gathering economy. Not being professionally a terrestrial ecologist, I was surprised to find that the animal productivity that goes into deer and other such “catchable” animals is a negligible fraction of what goes into songbirds, which, in turn, is a negligible fraction of what goes into insects. And then I looked at mice and found that—although they too are less productive as an ongoing portion of the ecosystem than are insects-mice are considerably more productive than either deer or songbirds. And I calculated that if men were capable of utilizing the annual yield of mice, there ought to be population densities of the order of 100 times as great as those that there actually were in such forests.

It is known that in a closely regulated predator-prey situation, the more the predator takes, the more he can get (up to a certain point, naturally), This is because the predator presses down the demographic structure of the prey population, so that it is in its maximum productive stage. The predator can then take the annual yield, which otherwise would be lost to natural mortality. So I ask the ethnographic experts whether there is any evidence that people anywhere have li\Td the year round solely on mice,

Eggan: Some well-meaning group of whites tried to move a group of Shoshone out of a desolate valley, and the Shoshone answered, ill effect: ““Vhat? Leave all these wonderful mice around here?”

Gardner: I have wondered how the South Indian Paliyans managed to maintain a population density that must be nearly two persons per square mile. DeeH’Y gives us an awswcr— because they catch mice, rats, bats, ... and let the big game go!

Binford: In light of Dr. Deevey’s statement, we may be able to restore the Tehuacan people to the list of happy natives. A Canadian by the name of Callen (Callen and Cameron, 1 960) has studied the human feces recovered from the Tehuacan excavations and was able to get quantitative information on the diet of the people-a rare thing indeed. One of the more common inclusions in the coprolites was undigested bone and hair of mice and other small rodents.

Alland: A simple example from psychology might be applicable here. If rats are presented with a learning situation in which two stimuli are rewarded randomly--one 30 per cent and the other 70 per cent of the time—they soon learn that it is a better strategy to press the 70 per cent bar all of the time than to try to figure out the game. Unfortunately, humans do not do thi:;; they try to beat the game. So, when there are alternate resources of deer and mice, I suspect that they will always take the deer and leave the mice for emergencies.

SAHLINS: With regard to Professor Birdsell’s remarks made at the outset about carrying capacity, we do have for slash-and-burrt agriculturists a whole series of estimates of their relation to critical capacities, as measured by their agricuitural needs-land in usc at a particular time, lane! that they need to have available, etc. These techniques, of course, were developed in Africa and applied to native reserves, where it was almost always discovered that the people were over capacity. But they have since been applied to peoples not on reserves, for instance in the South American tropical forests and the results nearly everywhere indicate that population densities are about 20–50 per cent of capacity—in other words, in the range mentioned earlier by Dr. \\!ashburn.


{1} In some of the cases marked, the subsistence percentages have been changed from those published in the Ethnographic Atlas. The categories have been redefined so that shell fishing is included under “Gathering, “ and pursuit of sea-mammals under “Hunting.” In the Atlas, both are included under “Fishing.”

{1} In some of the cases marked, the subsistence percentages have been changed from those published in the Ethnographic Atlas. The categories have been redefined so that shell fishing is included under “Gathering, “ and pursuit of sea-mammals under “Hunting.” In the Atlas, both are included under “Fishing.”

Part III: Social and Territorial Organization

10. Ownership and Use of Land aIYlOng the Australian Aborigines

L. R. Hiatt

Aboriginal territorial organization is currently a subject of controversy (Hiatt, 1962, 1966; Stanner, 1965). I shall first present the classical description, then indicate the different picture at present emerging, and finally raise some speculative issues suggested by moderrl ethological research.

The Classical Description

Radcliffe-Brown (1931) distinguished three kinds of grouping—tribe, clan, and horde. A tribe consisted of people who spoke the same language and occupied a continuous tract of land. It was divided into patriclans; each owning a defined territory that included a number of totemic sites. A horde consisted of the male members of a patriclan, their wives, and unmarried female members. It was independent and autonomous, managing its own affairs and acting as a unit in its relations with other hordes.

The Recent Picture

Elkin (1950, pp. 17–18) stated that “the local organization in many tribes was not the clearcut patrilineal patrilocal exogamous group occupying a definite territory which some textbooks imply. “ Meggitt (1962) was the first tc assert specifically and in detail that some of Radcliffe-Brown’s generalizations did not hold for at least one tribe (the W albiri). My own field investigations between 1958 and 1960 (Hiatt, 1965) among the Gidjingali of Arnhem Land also failed to confirm Radcliffe-Brawn’s view; and subsequently I drew attention to similar material in reports of earlier workers who merely thought they had discovered isolated aberrant forms (Hiatt, 1962).

There is no quarrel with Radcliffe-Brawn’s statements ahout the tribe as a linguistic division or the patriclan as a land-owning unit. The argument is about the horde—more precisely, about residential associations, exploitation of resources, and group movements. What follows is a summary of my earlier paper (1962).

Composition of Communities

Radcliffe-Brown asserted that the local unit was based on a single patriclan. Yet at least seven observers since 1930 have described more or less stable communities that included members from two to twelve patriclans. Several writers mention dispersal and reunion according to seasonal conditions, visiting, and occasional


joint participation of several communities in the same ceremony. The aborigines themselves saw these events as temporary separations from the main group or temporary amalgamations with other groups.

Exploitation of Resources

Radcliffe-Brown stated that each horde had to obtain most of its foods and water from its own territory. He also wrote: “Acts of trespass against this exclusive right of a horde to its territory seem to have been very rare in the social life of the aborigines but it appears to have been generally held that anyone committing such a trespass could justifiably be killed” (1952, pp. 33–34). Yet eight observers since 1930 have reported that unrestricted movement of foodseekers over broad regions that included the totemic sites of many patriclans. This does not mean that all comers had free access to every locality. A Walbiri community, for example, included members of from six to twelve patriclans who habitually ranged over a defined area that included the totemic estates of the constituent clans. But they entered the territories of othftr communities on invitation (Meggitt, 1962, chap. 5).

Territorial Boundaries

Radcliffe-Brown said that known boundaries defined each patriclan’s territory, on which were situated its totemic sites. Some later observers have reported such boundaries, others have described them as uncertain or ill defined, and others again have denied their existence. The subject is important mainly in relation to Radcliffe-Brown’s assertion that a patriclan exercised an exclusive right to the natural resources of its territory. In tribes like the Walbiri and Gidjingali, patriclans did not claim such rights and accordingly did not establish demarcations around their totemic estates for the purpose of controlling access.

I should make it clear that I do not regard these empirical refutations of Radcliffe-Brown’s generalizations as somehow adding up to a new generalization. In view of the wide range of ecological variation in Australia, it seems unlikely that aboriginal local organization conformed to a single type, though now it is probably too late to establish with confidence the main varieties.

To anticipate the objection that RadcliffeBrown was in a better position than later investigators to observe the real thing, I would remark that the main studies of local organization in Australia have been carried out on the frontier, not beyond it, and that Radcliffe-Brown’s research was no exception. We have all faced the problem of reconstructing a nomadic past from a mission-, station-, or townbased present. Elsewhere I have suggested that Radcliffe-Brown’s “cell-like” account of local organization was based on his unverified assumption that an acknowledged ritual relationship between a patriclan and its estate implied an exclusive economic and residential connection as well (1962). Here, in support of the kind of account given by Meggitt and myself, I mention Elkin’s view (as yet unpublished) that the congregatron of aborigines around European settlements was, in terms of local organization, a change not of kind (as commonly supposed) but of degree. At least in the early phase of contact the natives regarded these establishments as “super-water-holes.”

Ethology and Primitive Hunters

Carpenter wrote that: “It would seem that possession and defense of territory which is found so widely among the vertebrates, including both the human and subhuman primates, may be a fundamental biologic need” (1941, p. 158). It is now evident that there are important exceptions. Schaller (1965) reports that groups of mountain gorillas restricted their activities to home ranges, but these areas were not exclusively occupied by one group. He concludes: “The almost complete overlap of some ranges and observations on peaceful interactions between groups indicate that gorillas have no territory in the sense of an exclusive area defended against others of the same species” (p. 342). Reynolds and Reynolds (1965) state that chimpanzee groups in the Budongo Forest were constantly changing membership, splitting apart, meeting others and joining them, congregating, or dispersing. And Goodall (1965) says that because chimpanzee groups in the Gombe Stream Reserve freely united from time to time, they could not be divided into separate communities. Even baboon territorial organization seems less rigid than previously supposed. Whereas Ardrey ( 1 96 I, p. 77) spoke of the troop as a society which “demonstrates all those hostile traits normal to the individual territorial proprietor,” DeVore and Hall (1965) in a recent report describe cases of close daily contact between different baboon groups without displays of intergroup aggression.

Modern ethology thus raises no difficulty for an account of aborigines which implies that they lacked territorial instincts. Nevertheless, patriclans did have inalienable rights to pieces of land. Given the evidence that they did not also exercise exclusive rights of use over these properties, it seems legitimate to ask why tribal areas embraced a number of separate clan estates. I shall approach the problem by selecting a Gidjingali patriclan as an example and discussing two questions: ( I) What did a clan estate look like to its owners?; and (2) What did it feel like?

  1. The Mararagidj patriclan (which I shall abbreviate to “Mara”) in 1960 comprised 25 members. It owned seven named sites distributed over roughly one square mile of coastal fiat. Five sites were totemic (several shell middens and parts of a tidal creek, all connected with a mythological Dingo), two were non-totemic (a mud flat formed by tidal inundation, and a place called “Where-the-plum-tree-died”). Only one site was a camping place.

The Mara belonged to a community of about 135 people called the Anbara, whose home range comprised six contiguous estates owned by the constituent clans. They usually camped together at one or other of two, possibly three, places suitable for a group of this size. In other words, the local community consisted of members of six different patriclans, and at any given camp the majority of members were living outside their estates.

The men fished with spears along the coast and tidal creeks and also hunted land animals, while women gathered plant foods and shell fish. Men often built fish traps across creeks running through their own estates. But there were no restnctlOns on movement, and the camp seems to have been a base from which people might range anywhere over the entire Anbara region (about twenty square miles), returning at dusk with the day’s catch. The Mara estate itself did not have a site suitable for such a large encampment, but from time to time clan members temporarily left the Anbara community to spend a period on their homeland. I should add that folk always said they had been born on their estates, though further questioning often indicated that they were not certain.

2. Each clan owned a number of designs representing its totemic sites and the mythological creators supposed to be living beneath or near them. The designs were used in initiation and mortuary rites, which were not necessarily held on the clan’s estate.

It is thus true to say that there was a ritual tie between ‘l clan and its estate. But there was also an emotional bond with the land itself that seems to some extent independent of mythological associations. Totems were regarded with awe rather than affection, and when an aborigine spoke lovingly of his clan estate, he described its natural features, not its supernatural underworld.

We have now reached the core of the problem. Given that the aboriginal owners of a piece of land were not solely or even largely dependent on it for their survival, and that they often spent more time away from it than on it, how do we account for the strength of their emotional attachment? I do not pretend to know the answer. The only relevant facts I can bring forward are that members of a Gidjingali clan thought of their estate as their birthplace and the birthplace of their forefathers (though they did not believe, as some other tribes did, that their totemic sites were repositories of spirit-children); and as a home to which they periodically wished to return. The evidence is clearly against the existence in aboriginal clansmen of an instinct to occupy and defend their territory. But it points to a strong impulse to establish and maintain territorial ownership.

An easier question to answer is why Gidjingali clans did not confine themselves to their estates. There are clearly adaptive advantages to having access to a broader subsistence range. Centripetal forces in communities such as the Anbara included desires for variety in diet and human company, the safety of larger numbers, and the economic security against poor seasons afforded by a bigger hunting expanse. The operation of these forces was facilitated by an ethic of genero5ity and hospitality which defined the good man as the generous man. But the persistence of small clan estates in the face of communities of clans and communal usc of land indicates the presence of a counterforce. Ritual responsibility was undoubtedly a faetor, but another component was an emotional bond with the land itself. The source of this emotion is not clear.

11. Stability and Flexibility in Hadza Residential Groupings

James Woodburn

In another paper in this volume (Chapter 5) I describe how the Eastern Hadza obtain their food relatively easily by hunting and gathering. Their nutritional needs are met quickly and without much effort, coordination, or cooperation.

In this paper the characteristics of Hadza residential groupings are briefly described. This account is only an introduction; a full description with the necessary supporting numerical data will be given in my forthcoming monograph on the Hadza.

Hadza residential groupings are open, flexible, and highly variable in composition. They have no institutionalized le«dership and, indeed, no corporate identity. They do not own territory and clear.-cut jurally defined modes of affiliation of individuals to residential groupings do not exist. The use of the term “band,” with its connotations of territorial ownership, leadership, corporateness, and fixed membership is inappropriate for Hadza residential entities and I prefer to use the term “camp,” meaning simply the set of persons who happen to be living together at one place at one time (Woodburn, 1964).

I consider first the area in which the Hadza live and the geographical regions into which they divide it. The nature of the camps, their composition and their flexibility are then examined. Synchronic and diachronic census material is presented to show the way in which groupings change through time and to demonstrate that in spite of their flexibility and variability, important underlying regularities persist.

The principal purpose of the paper is simply descriptive, but my intention is also to stress that we cannot hope to understand the residential arrangements of hunters and gatherers, especially those with flexible residential arrangements, unless we make use of numerical data, in particular diachronic numerical data.

The Geographical Divisions of the Country of the Eastern Hadza

The area the Eastern Hadza occupy consisted in the recent past of well over 1,000 square miles of land. Until the 19308 the Hadza had this country to themselves, but then two villages of outsiders were founded at fertile places right inside Hadza country and within a few years their population far outnumbered that of the Hadza. At the same time there has been constant loss of land on the edge of their country to neighboring tribes and more especially to the Iraqw. But the serious incursions are all relatively recent and had not had any effect on the Hadza system of land tenure until the Hadza were settled in 1964 and 1965.

The Hadza commonly use certain terms to refer to regions of their country and for the people associated with these regions. For example, Mangola is used for an area around the Mangola River. This area cannot be precisely defined, for the use of the term depends partly on context. If one asks a Hadza living far from Mangola about the whereabouts of some person and is told that he lives at Mangola, this means that he is living somewhere within an area of some hundreds of square miles around the Mangola River. On the other hand, if one is told by a man living only five miles away from the Mangola River that some person is living at Mangola, this normally means that the person is living within a mile or two of the lower reaches of the river.

The word Sipunga, meaning the region around Sipunga Mountain, has similar variations in meaning according to context. A third term Tli’ika (literally “the west”) is commonly used to refer to a stretch of high ground running northeast from Isanzu between the Yaida Valley and Lake Eyasi to about halfway above the take where the fourth and tinal region, Han!abi (literally “the rocks”), begins. Han!abi, which is, as its name suggests, a mountainous, rocky area, extends along the same ridge between Lake Eyasi and the Yaida plains up to the mountains overlooking the Mangola River which are generally considered to be part of the Mangola region. Unlike Mangola and Sipunga, neither Tli’ika nor Han!abi have a focal point in a single geographical feature, but all four regions resemble each other in having no clearcut boundary. To draw boundaries would be quite artificial; the regions grade into each other.[37]

On the other hand, no part of the country in which the Hadza live to the east of Lake Eyasi lies outside the areas to which these terms may be applied when used with their widest connotations. We cannot separate them off from each other, but between them they cover the whole country. Only the extensive open Yaida plains are outside the classification, but this is not country in which people live. Hadza never live in the open plain, but only in the bordering thickets and hills which fall within the named areas.

The division of Hadza country into four named, roughly defined localities bears some relation to Hadza social arrangements. At any one time each of these areas is likely to contain between 50 and 150 Hadza. One commonly talks of the Sipunganebe (people of Sipunga), Mangolanebe (people of Mangola), Tli’ikanebe (people of the west) and Han!abi- cehe (people of the rocks) . These names do not refer to a fixed body of people in each area but simply to the aggregate of individuals who at any particular time or in any particular context are associated with one locality rather than another.

Each of these areas contains sufficient sources of food and water to maintain its inhabitants throughout the year and many people, especially the elderly, restrict their nomadic movements for years at a stretch largely to a single one of these four areas. At the peak of the dry season, when water is scarce and when the berries of Cordia gharaf Ehrenb. and Salvadora persica L. ripen in huge quantities near to these few sources of water, the majority of the inhabitants in each of these areas may be found concentrated in a few large camps within perhaps two or three miles or less of each other and close to the berries and water supplies of their area.[38] I have never heard of any occasion on which all the inhabitants of one of these areas joined togetqer to form a single camp, although their mode of subsistence would not of itself make this impossible.

In one of the four areas, Sipunga, there is a marked cleavage into two divisions, one lying close to the Sipunga mountains and the other centering on Yaida River. For much of the year the camps of the Yaida division are far from those of the Sipunga division, but at the peak ofthe dry season most of the people living near Yaida usually come to livc in camps in close association with the camps of the Sipunga division among the berry trees under Sipunga Mountain. The division of the country into four areas is conspicuously linked with the fact that, wherever they may be in the wet season, camps are reasonably clearly grouped into four widely separated clusters at the peak of the dry season.

People, singly and in groups, move freely from region to region. Any individual Hadza may live, hunt, and gather anywhere he or she likes without any sort of restriction and without asking permission from anyone. Neither individuals nor groups hold exclusive rights over natural resources, over land and its ungarnered produce, and there is never any question of people who have long been associated with a particular area having exclusive or even prior rights there. Even the expansion of neighboring agricultural and pastoral tribes into Hadza country is not opposed, and each year much land is lost.

The inhabitants at any particular time of each of these regions do not share a joint estate; nor do they unite to perform any activity; nor do they acknowledge any bond uniting them in opposition to other areas. They cannot be described as constituting a group and to use the term “band” or “horde” to describe them would be completely inappropriate.

Camp Size, Composition and Nomadic Movement

At any time, the 400 or so Hadza who lead a nomadic hunting and gathering life to the east of Lake Eyasi are living in camps containing very varied numbers of people—from a single person to almost a hundred people. The average camp contains about eighteen adults. But the members of a camp do not constitute a stable unit, either in place, in numbers, or in composition. People do not camp continuously at a particular site tor more than a few weeks and they usually move much more often. At the time members of a camp all move, they may go together to a new site; they may split up and form camps at two or more new sites; they may go as a body to join some existing camp, or they may divide, some joining an existing camp and others building a camp at a new site. Even while people are living together in a camp at a particular site, the composition of the camp changes: some people move in and some move out. In spite of this flexibility, there are interesting consistencies in the composition of camps which will be described later.

The individual members of a camp at a p ar t ic u l ar time are not present on account of their common attachment to a body of resources; nor does a camp have an acknowledged leader to whom common allegiance is owed. A camp at any particular time is often known by the name of a well-known man living in it at that time (for example, (ets’a rna Durugida-Durugida’s camp). But this indicates only that the man is well enough known for his name to be a useful label, and not that he acts either as a leader or as a representative of the camp. At any particular time a camp is an agglomeration of individual members tied to each other by a variety of kinship and affinal ties.

In my doctoral thesis I use data derived from two censuses, which I describe as my synchronic and diachronic censuses, to establish which relationships of kinship and affinity are important in residential arrangements. For the synchronic census, I took down particulars of the membership of ten camps in the Sipunga and Mangola regions in June and July, 1960. These camps contained approximately 60 per cent of the Eastern Hadza. The diachronic census consists of a record of the changes and consistencies in the composition of a series of camps. Taking one particular old widow named Bunga, the mother of two married sons and two married daughters, I recorded, on 25 separate occasions spread over more than three years, the people who were living in the same camp as she was.

Bunga was, in the estimation of the Hadza, a respectable and conventional women who lived rcgularly, by Hadza standards, with her close kin and affines and who moved camp less frequently than many other peoplc, especially younger people. On all 25 occasions, the camps were in the Sipunga region. The particulars of the members of these camps tend therefore to stress the continuities rather than the discontinuities in Hadza residential groups. Yet even with this emphasis on continuity, there were radical differences in the size and composition of these camps. The minimum number of adults present was 3 and the maximum was 37. There were less than 10 adults present on four occasions, between 10 and 19 adults present on twelve occasions, between 20 and 29 adults present on seven occasions and between 30 and 39 adults present on the remaining two occasions. A total of 67 adults, more than a quarter of the entire Eastern Badza adult population, lived with Bunga in at least one of these camps. The Hadza apply the word huyeti (visitor) freely to anyone living in an area ih which he does not normally live or living with people with whom he does not normally associate. But the use of the term differs according to the speaker and to the context, and it is quite impossible to divide up the members of a camp into those that are residents and those that are huyeti. However, an examination of the diachronic census does permit the isolation of a set of people who were regularly associated through time, and the nature of the bonds linking these persons will be mentioned in a later section of this paper.

With the rich food resources of their country and their knowledge of how to exploit these resources, it may seem hard to understand why the Hadza live in small, unstable nomadic groups which move at frequent intervals and constantly change in size and composition. In my monograph on the Hadza I shall discuss the way in which individuals and groups move from place to place far more frequently than is strictly necessary if movement is seen simply as a means of providing the best possible access to supplies of food a.nd water. We ourselves are so tied down by, among other things, the sheer quantity of our posses3ioiis, that we with justification regard the movement of a household from one place to another as difficult and not to be attempted without substantial reason. But we must be careful not to allow our ethnocentrism to creep into our ethnography. The Hadza, like many other nomadic people, value movement highly and individuals and groups move to satisfy the slightest whim. Their possessions are so few and are so easily carried that movement is no problem. Indeed people often find it easier to move to the place where a game animal has been killed than to carry the meat back to their camp. The Hadza may move camp to get away from a site where illness has broken out, to obtain raw materials—stone for smoking pipes, wood for arrow shafts, poison for arrows, herbal medicines and so on, to trade, to gamble (see Chapter 5, this volume), to allow the realignment of the huts in a camp after changes in camp composition, to segregate themselves from those with whom they are in conflict, and for many other reasons. People do often move primarily because food and water are less readily available than they would like; and even where some other motive is present, they will of course at the same time try to improve their access to food and water. However, movement normally takes place long before it is essential, long before shortages have become in any way serious.

The variations in the size and in the composition of camps again cannot be interpreted in simple ecological terms. There are no grounds for suggesting that the constant and considerable changes in camp size and composition neatly parallel similarly frequent and substantial fluctuations in the available sources of food or in the arrangements used to procure food. At any season a wide variety of sizes of camp is ecologically viable. During the wet season, camps are, on average, smaller than in the dry season, but not, I think, because of any important difference in the rigor of the environmental pressures in the two seasons. Perhaps the most important single factor creating the larger camps of the dry season is that at this time of the year more large game is being killed. The meat of a large animal is by custom widely distributed through the camp and people tend to congregate in camps where there are skilled and successful hunters. The optimum size, considering this factor alone, will be the largest number of persons who can share an average-sized game animal and feel that they have had an acceptable quantity of meat. In the wet season, when few large animals are killed, there is less incentive for people to congregatc in large camps and the divisive effects of quarrels tend to keep the size of the camps small. The general emphasis on seasonal polarity, mentioned briefly in Chapter 5, this volume, is also relevant. Many other factors, some ecological, some not, affect camp size, but cannot be discussed here. The limits on camp size set by the availability of food and water and the techniques used to obtain them are very broad and permit wide fluctuations; wide fluctuations in fact occur.

We must beware of any tendency to treat fixed, permanent ties linking together aggregates of people as normal, and loose, impermanent bonds as abnormal and requiring special explanation. Because the Hadza do not join in groups to assert exclusive rights over portions of land, or other property, because they do not unite to defend either resources or their own persons, because they cooperate very little with each other in their subsistence tasks, there is little to bind individuals to specific other individuals. People do depend on obtaining meat from other people, but they are entitled to a share of meat simply by being in a particular camp at a time when a large animal is killed there and do not have to rely on specific categories of kin or other specific individuals to supply them. With a few important exceptions, which will be described below, people depend simply on living with any other Hadza and not on living with a particular set of people standing in specific relationships towards them. They range widely, associating themselves now with one collection of people, now with another, according to their desires and whims of the moment.

Regularities in Camp Composition

I cannot here give much of the detail which is set out at length in my doctoral thesis and which will be published in my forthcoming monograph about the regularities that occur in camp composition. But the more important regularities are as follows:

Married Couples

The Hadza refer to any man and woman who cohabit and who publicly acknowledge their cohabitation as being husband and wife, whether or not any ceremony of marriage had taken place, and whether or not the man and the woman accept the obligations which husbands and wives normally accept towards each other and each other’s kin. This is not to say that these obligations are of little consequence: they are, in fact, exceedingly important and a marriage is most unlikely to last unless they are accepted.

Ceasing to live together, even for a matter of a few weeks, puts a marriage in jeopardy. A husband goes away on a visit to some other part of the country without his wife at his peril. He may well find that when he returns she has either married someone else or that she has repudiated him by putting on the dress of an unmarried girl to indicate her availability for marriage. People who are stated by informants to be married are therefore almost always coresident with their spouses. In the synchronic census I encountered 115 spouses of monogamous marriages and of these 112 were living with their spouses. The remaining three were married according to Hadza informants; that is, informants believed that the separations were temporary and that the couples would resume cohabitation. All three of the marriages were, however, demonstrably unstable. Separation, even for short periods, may be both an indication that the marriage is not entirely successful and a contributory factor to its eventual breakdown. At the same time, not all separations are necessarily either symptomatic or damaging; spouses of successful marriages do occasionally live apart for a while, but they risk their marriages by doing so.

Compared with many other societies, marriage is relatively unstable. The calculated divorce rate[39] is 49 per 1,000 years of marriage, though this is only a very rough approximation. Few figures for other East African societies are available in this form. The Amba have a rate of 31.5 (Winter, 1956, p. 39) but no figures are available for tribes near to the Hadza. The rate for England and Wales, 1950–52 was 2.8 and for the United States, 1949–51, it was 10.4 (Demaographic rearbook, 1968).

In comparison with other Hadza relationships, the noteworthy aspect of the marital relationship is not its instability, but on the contrary its stability and strength. A. marriage is broken by divorce on average only about once ill twellty years of married life. In general, most Hadza men settle down and live for many vears with a particular wife; to sustain the marriage they do not leave hn for long on her own and they fulfill onerous obligations, described below, to her and also to her mother.

Once they marry for the first time, usually by their early twenties, very few men live for long unmarried. After the death of a wife or permanent separation from her, the husband will soon remarry. In the ten camps of the synchronic cenSllS, there was only one man who had previously been married who then lacked a wife. On the other hand there were thirteen previously married women who at the time of the synchronic census were unmarried. Of these, eight were past the menopause.

A few men have more than one wife. This complicates their residential arrangements and their position cannot be discussed here.

Parents and Children

My census data shows that children whose parents arc separated almost invariably live with their mother. In adult life both men and women often live with their parents, but the emphasis is on residence with the mother. If the parents are separated orthe mother is dead, it is unusual to find a man or a woman living in the same camp as the father.

Both husband and wife value co-residence with their mothers. However, residence with the wife’s mother is considerably more frequent. Considering only monogamous marriages in which the spouses were living together, there were in the synchronic census 28 husbands and 34 wives with mothers alive. Twelve (that is, 43 per cent) of these husbands and 23 (68 per cent) of the wives were living with their mothers. But the emphasis on residence with the wife’s mother is greater than might appear from these figures.

Let us examine the residence of those monogamous individuals who are co-resident with their spouses and who have both a living mother and a living mother-in-law. In the synchronic census there were 42 such people (that is, 2’ couples):

‘5 couples were living with the wife’s mother

7 couples were living with the husband’s mother

4 couples were not living with either mother

But five of these couples were living in camps containing both mother and mother-in-law. From these five cases, no indication can be gained of any difference betwet’n the attraction of living with the wife’s mother and the attraction of living with the husband’s mother. If these cases are omitted, the figures are as follows:

10 couples were living with the wife’s mother alone

2 couples were living with the husband’s mother alone

4 couples were not living with either mother

In other words the monogamous couples of the synchronic census who were faced with a choice between living with the husband’s mother and living with the wife’s mother, had exercised their choice in favor of the wife’s mother five times as often as in favor of the husband’s mother.[40]

A high valuation is placed on residence with the mothers of husband and wife, yet four couples were found living in camps in which neither mother was present. As it happens, in each of these four cases, there are unusual circumstances[41] and in quite a high proportion of cases observed at other times in which a couple is living away from both mothers, there is an obvious reason to account for their behavior.

The diachronic census data support the findings from the synchronic census and allow the matter to be taken an important step further. 1t enables us to sec whether decisions about which mother to live with are lasting. In my diachronic census, only one couple among those with both mothers alive who lived for a time with the husband’s mother did not at some other period live with the wife’s mother. In this one case the wife’s mother was living not among the Hadza but with a neighboring tribe.

Marriage involves an obligation, and one which is observed, to live with the wife’s mother and, except in special circumstances, the only time a couple with both mothers alive will live for long away from the wife’s mother is when they are living with the husband’s mother. For a marriage to survive, a husband must live regularly not merely with his wife but also in the same camp as her mother. If his own mother is alive he and his wife will usually (but not always) visit her from time to time. There is no question of any obligation to live with her.

A man has important property obligations toward his wife and his mother-in-law. Long strings of bridewealth beads should be, and usually are, given by the bridegroom to his parents-in-law. They are taken and worn by the mother-in-law around her waist. Thereafter, throughout his life of the marriage, the husband should keep his wife and mother-inlaw supplied with meat and with trade goods. Almost all trade is carried on by men; women rely mainly for access to these goods not on their own kin but on their husbands and sons-in-law. Beads, tobacco, and cloth are especially keenly desired and demanded. Husbands cannot afford to fail too often if they wish to preserve their marriages.

The pattern of dyadic ties that I have described links persons into a grouping which has the same genealogical composition as a matrilineal or matrilocal extended family, that is a woman and her married daughter (or daughters), together with their husbands, and their unmarried offspring of both sexes. In a few cases in which women have survived to a great age, the grouping will contain three generations of married women: the apical woman will have both married daughter(s) and married granddaughter(s) in her grouping. These groupings are not joint or extended families since they are not corporate groups and do not have a recognized head.[42] I prefer to use a neutral term and to refer to them as “simple residential units. ”

By stating that a simple residential unit has no recognized head, I mean that there is no institutionalized position of head of the unit, no person who exercises authority over other members of the unit whether by virtue of his or her genealogical position or by any other qualification. In some units one or more individuals may stand out as influential persons, but in other units this is not the case. However, the unit does have a genealogical focus in the person of the apical women. We might, therefore, describe the units as matrifocal, but even this is, in some cases and contexts, misleading. Sometimes, especially if the wife’s mother is old or of weak character, has no husband, and has only one married daughter, it can be seen that the mother is aligning herself residentially with her married daughter and her son-in-law, rather than they with her. But usually it is the mother who, more than any other person, serves as the genealogical focus around which a simple residential unit clusters. At any particular time a camp will usually contain one or more simple residential units together with a very few people who are not members of these units. When two or more units are present in a single camp, they may be linked by kinship and affinal ties or they may not. If no tie links them, they are most unlikely to stay together long. Units which are regularly residentially associated are usually bound together by a number of close kinship or affinal ties. Not even the more important ties linking together units-such as the tie between a woman and her married son or the tie between a pair of siblings-are very highly valued or involve significant property obligations. The kinship and affinal ties of the apical woman are temporally prior to those of the junior members of the unit and they tend to be more often used in establishing links between the component units of the camp.

Conclusion

The flexibility and variability of Hadza residential arrangements are fundamental; the regularities within the flux that is Hadza society emerge clearly only from an examination of residential arrangements through time. Such regularities as do exist are linked with the continuing property obligations of certain pairs of persons to each other, above all husband and wife, and mother-in-law and son-inlaw. In Hadza society, individuals are bound to each other by dyadic property ties between relatively few categories of persons, and the type of unit that is a simple product of these ties must be distinguished from the corporate domestic and other groups, whose members cooperate, possess important joint interests, and accept authority within the group, which are characteristic of most small-scale societies.

The highly flexible social arrangements of the Hadza are by no means unique among hunters and gatherers. At this symposium, data has been presented on a number of peoples whose residential arrangements show some striking similarities to the Hadza. For instance, the!Kung Bushmen, the Mbuti, the Ik, the Dogrib, the Netsilik Eskimos, and even the Gidjingali of northern Australia all live in groupings that change frequently in size and composition.

The widespread occurrence of this flexibility brings us back to the problem of field method. The analysis of group structure is a crucial area of research, and, if comparative studies are to be of any value, it is essential that the field workers concerned develop a common body of techniques. One of the most important results to come out of the symposium on Man the Hunter may be the development of measures of general use for describing and analyzing the open and fluid social groups of hunting and gathering peoples.

12. The Diversity of Eskimo Societies

David Damas

Survey of Eskimo Societies

It is not necessary for the student of hunting societies to extend his survey further than the Eskimo area to gain an appreciation of their diversity.[43] The large permanent villages of the North Alaskan and Bering Sea whale hunters and the fragmented mobile groups of the Caribou Eskimo represent extremes of settlement and community types occurring within an area of considerable cultural and linguistic uniformity (Fig. I) .

Two important characteristics of social groups are the size and the duration of aggregates. Aggregations exceeding 50 and often 1OO persons formed at one time or other during the year in almost all Eskimo regions. The maximal aggregates of the Caribou Eskimo which occurred at caribou drives appear to have been smaller units (Birket-Smith, 1929, p. 74). Spencer (1959, p. 132) feels that caribou drives of the similarly adapted Nunamiut attracted as many as 200 or 300 individuals, but Gubser (1965, p. 167) thinks that 50 to 150 was the usual range. These short-lived aggregates generally lasted only a few weeks. The snowhouse encampments of most Central Eskimo groups appear to have often included more than 100 persons and at times as many as 200 (Boas, 1888, p. 425) for a usual duration of two to four months.

Considerable variation can be noted in the duration of occupancy and the size of villages composed of permanent structures (if. Watanabe, this volume). The North Alaskan villages of up to 300 inhabitants (R. Spencer, 1959; Van Stone, 1962) were occupied for about ten months of the year, with some individuals remaining at the sites during the entire year. Villages of the stone and sod house dwelling Polar Eskimo (Ekblaw, 1948), Coast of Labrador Eskimo (G. J. Taylor, 1965), and East Greenlanders (Holm, 1914; Thalbitzer, 1914) generally averaged fewer than fifty inhabitants and were occupied for four to seven months. In these cases the Eskimo merged into larger groups at some time during the year’s cycle. Economic, social, and ceremonial motives were probably all involved in this tendency to aggregate.

Kinship plays an important role in Eskimo social organization, just as it does in the organization of other hunting groups. Centrai Eskimo groupings reveal extensive linking of overlapping kindreds (Damas, 1969). In the absence of adequate data from other regions, continuity of kinship ties within most local groups might be expected to exist in consistency with the apparent general absence of rules of local exogamy. Matri-patrilocal residence has been reported for the Nunivak Eskimo (Lantis, 1946) and 8t. Lawrence Eskimo (Hughes, 1960) while matrilocality is attributed to th(; Nunamiut (Pospisil, 1964) and neolocality to the Copper Eskimo (jenness, 1922). Patrilineages occur in the Bering Sea region (Hughes, 1960; Lantis, 1946) and the patrilineal principle is projected to the clan level on St. Lawrence Island and among the Siberian Eskimo (Menovshchikov, 1964; Hughes, 1965). Elsewhere the chief kinship based segments within the community were the nuclear family, which seldom operated as an independent unit, the extended family, and units of the whaling crew and the karigi or associated men’s house segmented the communities. The karigi existed without association with boat crews on Nunivak (Lantis, 1946), but, on the other hand, the boat crew was organized according to lineages on St. Lawrence Island (Hughes, 1960) and among the Siberian Eskimo (Menovshchikov, 1964). Whaling crews were recruited from the members of adjacent villages in the case of the Chugach Eskimo (Birket-Smith, 1953).

Other structural features were peripheral to or independent from kinship. In North Alaska (R. Spencer, 1959) and in the MacKenzie delta (Stefansson, 1914) the voluntary in the case of the Eskimo of the Coast of Labrador, the polygamous joint family (G. J. Taylor, 1965). Throughout the Eskimo area spouse exchange and adoption provided a growing edge to the network of kinship.

r-b-richard-b-lee-irven-devore-man-the-hunter-13.jpg
Figure 1. Distribution of Eskimo Societies.

Partnerships of various sorts were important social features in much of the Eskimo area. Trading partnerships crossed the ecological boundary between coastal and inland groups in North Alaska (R. Spencer, 1959). Seal sharing ‘practices provided essential elements to Netsilik (Balikci, 1964, pp. 36–37; Van de Velde, 1956), and Copper Eskimo (Damas, in press) social structures. Partnerships in dancing were also prominent in Central Eskimo society (Damas, in press).

Whatever the chief segments or sorts of relationships that may have existed within Eskimo local groups, they were always identified by name with a territory. The postbase -miut or -meit together with a base which indicated a region, was a designation applied to major and minor local groupings among the more nomadic Eskimo as well as to permanent villages. This identific;ation with territory seldom, ifever, implied a defense of hunting area.

Three Central Eskimo Societies

Table 1 has been included in order to make possible more detailed comparison with other hunting societies. It covers the general social and economic features of three major Eskimo groups studied by the writer.[44] Each of the groups represented genealogical bodies with extensive linkage within and sharp separation from each of the others. These factors together with differences in dialect and culture justify separation of the groups for purposes of analysis. The term “tribe” has been usefully applied (Helm, 1965a, p. 384) to such entities without implying political aspects of the term.

Aboriginally the 450 to 500 Netsilik (Boas, 1888, p. 454; Damas, 1969), 500 to 550 Iglulik (Damas, 1963, p. 23; Manning, 1943, p. 103) and 800 Copper Eskimo (jenness, 1922, p. 42; Rasmussen, 1923, p. 70) were divided into groups averaging 100 members who lived part of each winter in villages established on the sea ice. These were the largest yearly aggregates and appear to have been strongly associated with the ecology of the ringed seal. However, ceremonial and social motives for assemblage probably were also important.

Fluidity in membership occured among the groups of each tribal region but the winter aggregations of two successive years were comprised of a stable core of 60 to 70 per cent of the members. I have indicated (Damas, 1969) that the hunting groups that merged in such winter camps formed the “band” and that the aggregation itself was the “band assemblage.” These designations seem consistent with accepted ethnological usage. It would probably weaken the typological value of the band concept to extend its application to large permanent settlements like those’ found in West Greenland or the Alaskan and Bering Sea areas.

The seasonal economic cycles of the Copper and Netsilik Eskimo were identical in main features (see Table I). The Iglulik Eskimo cycle differed in that there was greater emphasis on sea mammal hunting and less on fishing.

Band composition appears to have been uniform in the three regions. Table I refers to composition as “bilateral with prominence of male relevant ties.” This generalization has been abstracted from charts of primary kin ties based on censuses of the Fifth Thule Expedition of 1921–24 (Mathiassen, 1928; Rasmussen, 1931, 1932). These charts reveal the predominance of male-male primary kin ties over female-male or female-female ties. Enough of of the latter two types of links occur to inject considerable bilaterality into the genealogical shape of the band (Damas, 1969).

Criteria for family type used in the table are those of residential arrangement and minimal sharing units. Among the Netsilik and Iglulik tribes the families are characterized as “viri- oriented extended.” These groups occupied multi-family dwelling or compounds of single dwellings. They were indivisible sharing units. Usually one or more of the nuclear family heads in these units was attached uxorilocally. Factors of population structure, especially isolation through death of agnates, and bride service appear to account for such cases (Damas, 1963)’ The Copper Eskimo are one of the few Eskimo groups whose family structure exemplifies the nuclearity usually represented for the Eskimo in interpretive writing (Linton, 1936, p. 163; Murdock, 1949, p. 227; Service, 1962, p. 99).

Table 1. Central Eskimo Social and Territorial Organization

r-b-richard-b-lee-irven-devore-man-the-hunter-40.jpg

The structure of the family i s reflected in the makeup of the hunting groups which cO.m- prised the chief local groupings during a large part of the year’s cycle. In the spring the Iglulik and Netsilik winter band aggregates split into extended family clusters averaging perhaps fifteen to twenty persons each. The Iglulik split further during part of the summer when the younger men roamed inland after caribou while pairs of older men hunted from kayaks in the sea. The nucleation ofthe Copper Eskimo family was expressed in the irregular composition of summer hunting groups. Lacking the high degree of cohesiveness of extended family structure found in the other two regions, the Copper Eskimo hunting groups were made up either of a core of siblings, or of spouse exchange partners, or of actually unrelated or distantly related persons. In all three areas aggregations of an intermediate size in relation to the winter sealing village and summer hunting groups gathered for short periods at the sites of caribou drives or fish runs.

The diversity in the organization of hunting groups is apparent in the other features shown in the table: leadership, sharing practices, and commensal units.

Leadership was weakly developed among the Copper Eskimo and personal qualities alone appear to have been responsible for whatever authority certain individuals enjoyed. The Netsilik extended family was headed by the oldest male member, the ihumataq, but there appears to have been no overall band leader. This family leadership pattern was extended to cover the entire band in the Iglulik Eskimo region. The band isumataq was usually the head of the largest extended family unit in that area (Damas, 1963).

Sharing practices and commensal practices together formed complementary systems of distributing the catch in each area. Among the Copper Eskimo, individual nuclear family heads formed partnerships which determined the sharing of specific parts of the seal. The network of distribution was extended in the area of consumption by the practice of communal eating on a village-wide basis. In the Netsilik region, seal-sharing partnerships also existed between individuals, but the meat thereby acquired became the property of the extended family. The two commensal units, the men and the women, ate separately in adjoining domes of the family’s snowhouse compound. The catch of the Iglulik Eskimo was distributed by the band isumataq among the lesser isumatat. Their commensal units were variable and based on expediency. Sometimes the returning hunter ate alone; other meals were restricted to the nuclear family, and some were shared by the entire extended family.

The character and magnitude of variation within these three regions may not seem significant when viewed in worldwide perspective, but certainly they are important enough to require explanation.

There is an apparent integration or coordination of social features that can support the contention that the three societies were balanced, internally consistent units. The nucleation of Copper Eskimo families is compensated for in economic matters by highly structured sharing and communal eating. The need for communal eating appears less urgent among the Netsilik since the extended family acts as a pool for resources. Partnerships in sharing appear to be largely unnecessary among the Iglulik since the distribution of food among extended families throughout the band was provided for by the band isumataq.

Looking beyond the internal arrangement of these social and economic features to external explanations, an examination of the table lends little support to facile correlations. If one assumes that the chiefsource of regional variation of societies will be ultimately related to variations in ecology, two hypotheses could be posited for the Central Eskimo regions discussed here: ( I) Uniformity of resources and exploitative patterns in the three regions are so similar that one would expect great similarity in social features. I have argued against such social uniformity. (2) Similarity in the economic life of the Copper Eskimo and Netsilik in contrast to the somewhat divergent economy of the Iglulik Eskimo should be reflected in social features. This correlation cannot be strongly supported by the table.

While a doser examination of ecological factors may reveal correlations with social structure, in the opinion of the writer that approach will fall far short of total explanation of regional variations in the Central Arctic. Historical factors of common heritage, diffusion, and cultural drift will have to be considered in this context.

This summary of some of the important social features in three Central Eskimo regions indicates that, in addition to the general diversity found among the major Eskimo areas, variability in patterns of community organization occurs within an area exhibiting highly similar or even identical patterns in the dispersal of people.

Variability provides the raw material for studies of “concomitant variation” (Nadel, 1951) or “controlled comparison” (Eggan, 1954), approaches which will be rewarding if applied in the ‘Eskimo area. This discussion of three Eskimo groups is only a beginning to more detailed comparative studies. To extend useful comparative studies throughout the entire Eskimo area requires more detailed information about a number of groups than is available at present. However, even such a casual review of social diversity as I have given suggests further channels for enlarging the understanding of hunting societies in general.

Relationship to Other Studies of Hunting Societies

As indicated above, the concept of the band does not seem to be applicable in all Eskimo regions. Certainly within the realm of hunting societies there are peoples similar in their sedentary existence to the North Alaskan or Bering Sea Eskimo who fall outside the scope of a useful application of “band.” Even if one restricts comparison to nomadic groups, difficulties arise in the use of the term “band.” During a recent conference on band organization[45] it was found that the term “band” had been applied to such widely variant groups as the Birhor of India with average size groups of 25 and the Peel River Kutchin group of 300 to 400 members. In trying to establish common criteria for usage, it was generally agreed that two major types of units could be subsumed under “band,” those that aggregated at some point in the year’s cyete and those that comprised only a conceptual unity.

A number of other factors which complicate attempts to isolate the band in nomadic hunting societies were also revealed during that conference. A careful examination of a larger sample of these societies is needed before such problems of band organization can be resolved.

Whatever term is assigned to groups that comprise a society, certain bases for comparison can be posited. Helm ( 1969 ) states the following regarding such comparisons:

By anthropological convention, band societies are set apart from non-band societies of all kinds. But I do not believe the features conventionally used to identify a band society—nomadism, lack of central tribal authority, and hunting-gathering pursuits ... make the problems of methodology in the analysis and classification of community composition inherently different from those encountered by perceptive ethnologists working in non-band societies.

It is possible to compare the kinship makeup and the family organization of the permanent winter villages of North Alaska with the band assemblages and family organization in the Central Eskimo regions. On a larger scale, units showing useful bases for comparison will certainly be found throughout much of the hunting and gathering world.

Reference to voluntary relationships such as Central Eskimo seal sharing partnerships or North Alaskan whaling crews or karigi organization suggest that classifications of the social structures of hunting groups have perhaps- placed too much emphasis on kinship factors. Classifications or typologies of hunting societies should be enlarged to include features that lie outside kinship considerations. Residence and marriage practices have generally been the chief criteria for classifying bands but perhaps these features are not always the chief distinguishing attributes of band societies.

Classification of bands in the Central Eskimo area based only on kinship composition, which is uniform, tends to obscure important differences in the family organization, the patterns of leadership and food distribution practices. The selection of criteria for band types from such diverse features requires considerable care.

In addition, classifications can be meaningful only if the ethnography of individual groups is sound and approaches completeness. Just as there is no great uniformity in the quality or completeness of information about all Eskimo societies, there is no great uniformity in the data on hunting peoples throughout the world. Considerable additional research is needed before statistical comparisons of hunting groups can be highly reliable. The rapid disappearance of such groups places the problems of salvage foremost. In many cases a closer examination of ethnohistoric sources will be the chief means available for reconstructing clear pictures of such societies.

Problems of function a]“o appear to he compounded by the variability that has been noted here. The integration of such features as karigi organization or seal sharing practices into societies differing in a number of other features suggests that possibly in the world at large alternate features will be integrated into similar systems and similar features integrated into otherwise contrasting systems.

Finally, problems of the adaptation of society to environment seem to become more complex when the diversity noted here is taken into account.[46] This diversity did not always conform to a concomitant variation in subsistence or ecology. Although strong arguments can be advanced for a close harmony existing between the pattern of exploitation and certain features of the settlement pattern in the Eskimo area and probably in the world of hunters as a whole, the relationships that exist between these settlement patterns and variable community patterns appear to be less rigid and perhaps will allow alternatives in every conceivable case.[47]

While the diversity of the social forms found within hunting and gathering societies may comprise only a narrow segment of the total spectrum of social types, a proper appreciation of this diversity will be an important step toward their proper classification and eventually toward an understanding of the problems of process or evolution.

13. The Nature of Dogrib Socioterritorial Groups

June Helm

This is an attempt to characterize and categorize socioterritorial groups among the Athapaskan Dogrib Indians of the northwestern Canadian subarctic.[48] By socioterritorial, I specify those kinds of groups which conjoin settlement pattern with community pattern (Helm, 1969). That is, I am concerned both with “any form of human occupation ... over a particular locale” or range (Chang, 1962), and with the nature of the social group constituted by the occupants of the locale or range.

The three kinds of socioterritorial groups which I identify among the Dogrib[49] I have termed the regional band (comparable to the “macrocoslPic unit of social organization” identified by Honigmann [1946, p. 64] for the inland Northern Athapaskans), the local band (equivalent to Honigmann’s “microcosmic unit”), and the task group.

Membership in these three kinds of socioterritorial entities is in no wise mutually exclusive. An individual may at the same time have social identity as a member of a regional band and of a local band, and, by the simple fact of his presence, also be a member of a task group. Segments of the local band and regional band operate as task groups (as defined below). These three kinds of groups, however, may be segregated analytically by time-and- space dimensions-namely, temporal duration and spatial (settlement) cohesion—and by three foci around which membership’ revolves.

The foci of membership identification and recruitment are (I) territorial range, (2) specific resources and resource locales within the range, and (3) kinship. I see the ecological- subsistence complex of range and resources conjoined with an ego-based kinship network as the basic forces in the creation and structuring of the three sorts of socioterritorial groups. Range, resources, and kinship not only affect, but efect, socioterritorial organization. Furthermore, each in turn serves as the prime focus of identity and affiliation for regional band, for the task group, and for the local band, respectively. For the regional band, the prime focus of identity is the territorial range; a task group forms in response to a specific and localized[50] resource; the strength of primary kin bonds holds the local band together in space and through time.

r-b-richard-b-lee-irven-devore-man-the-hunter-14.jpg
Figure 1.

The varying weight of these three factors affects the temporal and spatial dimensions of the three kinds of socioterritorial groups.

The triangle diagram (Fig. I) attempts a visual construction of the interplay of the specifying and discriminating features of regional bands, local bands, and task groups. The size of these kinds of socioterritorial groups is a function of the identity-affiliation factors and the time-space dimensions. Table I presents a chart of the main features differentiating the three kinds of groups.

In the persons exploiting the various and variable resources within a recognized range or territory, the regional band has its social identity as a “people.” The total territory yields sufficient materials for the necessities of life (famine periods excepted) so that within its domain the regional band can endure as an identity for generations. (Table 2 presents documentation identifying and placing Dogrib regional bands.) Characteristically, the total constituency of the regional band is not physically together. The families that make up the regional band are most apt to all come together when operating as a task group exploiting a resource which allows large assembly, as at a fall fishing camp. Most of tht’ time, physical co-presence and social communication is shifting, intermittent, and piecemeal. Within the regional band there is kin-connectedness in the form of a network of primary affinal and consanguineal ties between members.[51] But neither kinship nor a single specific resource is the focus of total group identity per se.

Table 1. Diagnostic Features of Dogrib Socio-Territorial Groups

r-b-richard-b-lee-irven-devore-man-the-hunter-41.jpg

The local band also exploits a range, often as a subunit within a regional band, and, like the regional band, is “concerned with the total problem of living within its territory” (Slobodin, 1962, p. 75). But it is distinctive as a spatial grouping-together of kinsmen, characteristically structured around a core sibling set, including females as well as males, with their spouses. To nucleated-settlement pattern and core-kin focus must be added some degree of temporal duration to set the local band apart from a kind of task group.

The task group is prt’-l”‘lIlinently a grouping- together of persons concentrating upon the exploi tation of a specific seasonal resource, ordinarily at some locale or along some route within the total range of a regional band. Wht’ther activities arc coordinated and/or produce shared depend upon thl”’ particular purpose of the task group (Helm, 1965£1). By definition, the task group lacks temporal duration beyond a few weeks. The social composition of the task group may either be based on the basic social building block, the family, composed of a conjugal pair with dependents, or it may be all male. It is apparent that kin linkages serve as a mechanism of affiliation in forming a task group, but there is also evidence that friendships and special aptitudes sometimes play a role.

Task groups vary in size and in sex-age composition, depending largely upon the nature of the resource which is the focus of task group creation. Today, two or three nuclear families (usually linked by sibling and/or parent-child ties) may join together for periods of two weeks to two months of trapping, combined with general subsistence activities. The distinction between such a seasonal, linked- family task group and what I identify as a local band can be made only on the sliding scale of duration. I have suggested elsewhere that the aboriginal equivalent oftoday’s local band may have been a relatively short-lived small grouping of this sort (Helm, 1965a, p.:379).

At the appropriate season, a task group may also be formed of many nuclear families around a fishing locale or along the route of a caribou migration. But task groups need not always be constituted of family units. Today, most hunting expeditions, for example, are all male, with the women and children left behind in camp or village.

Conclusions and Theoretical Considerations

In conclusion, I must make some effort to clarify the nature of the kinship principle which I sec operative in Athapaskan band affiliation —specifically, in post-nuptial band residence. I believe the same principle also to be, if not critical, significant in task-group assemblages.

From the Dogribs, as with other Mackenzie Drainage Alhapaskans, it is impossible to elicit “rules” regarding preferential marriage or local or regional band endogamy or exogamy. So also, with post-nuptial residence, except for the traditional understanding that the young husband should “work for” (and thereby reside with or by) his father-in-law until the birth of the first child. Add to all this the fact that the people move around the land “visiting” and that occasionally some do not return from whence they came.

The results of a search for residential pattern in this seeming flux have been presented in an earlier paper (Helm, 1965a). From inspection of the kinship connections between conjugal pairs in a sample of Athapaskan bands, I adduce the covert directive of primary relative relationship serving as a determinate limit on local and regional band residence-affiliation of the conjugal pair after marriage. Almost without exception, no married pair is residing with a local band or within a regional band domain unless at least one spouse of that pair, male or female, had at time of entry a primary consanguine, of either sex, already resident within the band.

r-b-richard-b-lee-irven-devore-man-the-hunter-15.jpg
Figure 2. Ranges (approximate) of Dogrib Regional Bands (after Petitot, 1891; Osgood, 1932, 1936; and Helm, 1962 field notes).

Some local band-hamlets are indicated by numerals (for identification, see Table 2). The eastern reaches of Band C, Edge-of-the-Woods People, are barren grounds, exploited only for brief seasonal caribou hunts (and, in the past, musk-ox hunts).

Table 2 Dogrib Regional Bands: Documentation over a 70-year period

(Regional Band designations are italicized. Local group designations are not complete, but are included when noted in the literature. Designations of local bands are not italicized. The capital letters in parentheses preceding references to local groups refer to the site of their hamlets or main camps located on the map in Figure 2.)

r-b-richard-b-lee-irven-devore-man-the-hunter-43.jpg
r-b-richard-b-lee-irven-devore-man-the-hunter-44.jpg

In practice, this bilateral primary linkage principle allows multiple residence alternatives for the conjugal pair. From the position of male Ego, for example, the man may marry within the local[52] and/or regional band where he resides with his parents. Or he may, after bride-service, bring an “alien” wife back to his band. Or he and his wife may, early or later in conjugal life, elect to live with the band where his wife’s parents or one or more siblings are established. Or he and his wife may link into a band where his brother or sister, married to a member of that band, resides. In later years, the man and his wife may “join” the band where a married child resides. And so on.

I have recently begun to look at other hunting societies for which ethnologists have provided data on the kin-composition of actual local groupings, short-lived or long-lived.[53] I find that Professor Levi-Strauss’ (1940) fluctuating Nambikwara bands (a I and b I) display primary bilateral linkage by conjugal unit around a sibling set with a brother-leader, comparable to the Northern Athapaskan local band. The pattern of kin-composition of a sample of !Kung Bushman bands (Marshall 1960, pp. 323–33) is indistinguishable from that of Athapaskan local bands, as is Marshall’s description of the chain-like residential linkage through bilateral primary ties (pp. 344–45). From Austraiia, the information provided by McCarthy and McArthur (1960) on the kincomposition of two Arnhem Land task groups reveals bilateral linkages between constituent conjugal pairs (with the result that men of different “country,” totem, and moiety are in co-residence). Hiatt has recently specified a number of Australian aborigine societies in which the “usual common residence groups,” as constrasted with the patri-line land-owning units, were composed of men of several patrilineal descent groups and in which the “small food-seeking units [task groups] commonly contained nonagnatic kinsmen”(i962,p. 285). In terms of who is actually “in community” with whom-for example, brother-in-law with brother-in-law rather than brother with brother—the picture of de facto residence association and economic interaction is quite different from that to be deduced from attributions of de jure patrilocality.

From these few cases alone, I submit that, for full comprehension of the economic and social dynamics of hunting societies and for cross-societal comparisons, the empirical patterns of post-nuptial co-residence, stable or shifting, demand as much attention as do formal “rules” of kin alliance and residence assignment.

14. The Birhor of India and Some Comments on Band Organization

B. J. Williams

This paper falls into roughly three sections. The first is a preface explaining a particular approach to the study of hunting-gathering groups and why this is used by a person whose primary interests are in physical anthropology. The second part constitutes data on a particular group, the Birhor of India, who became a subject of investigation in the pursuit of these interests. The third part explores some implications of such a view of hunting-gathering society.

My interests are those of a physical anthropologist, using, primarily, the approaches of population genetics in the study of human evolution. Some areas in which these interests are closely allied to interests of ethnology are (I) social organization as it effects mating patterns and (2) the extent to which a single model of hunting-gathering society might apply to Paleolithic groups. The problem is not whether it is possible to have more or less complex societies based on a hunting economy, but what is the modal form and how extensive might such a form have been at a given time in the past. The question is one of whether or not the findings of the ethnologist can potentially be useful in making statements about past social structures and population demography which have a bearing on genetic parameters such as total mating population size, degree of isolation, rate of gene flow, and rate of inbreeding.

Not being an ethnologist, I have proceeded in what may seem to the ethnologist a slightly cavalier fashion-to set up a model of hunting society based on certain assumptions, then to seek evidence on certain points of the model. Field investigations among the Birhor of India were a part of this endeavor. Some of the assumptions are sketched in the following paragraphs.

The broader assumptions are as follows:

1 (a). Most animals will exhibit either sexual or economic territoriality-or both.

1 (b). Animals whose variation in population numbers is strongly influenced by variation in food supply or other economic variables will exhibit economic territoriality.

1 (c). Such territoriality implies spatial aggregation.

2. Counter to the above is the ecological-economic desideratum of distributing the population homogeneously with respect to the distribution of food supply or other vital resources.

The optimal resolution of these two forces is a local organization whose size is the minimum necessary to exploit the collected foods, maintain territorial integrity and absorb or accommodate all but the rarest kinds of chance fluctuation in resources, sex ratio, and so forth.

To state a conclusion in advance of the argument, I feel that what was called -the patrilocal horde by Radcliffe-Brown (1931), the patrilineal band by Steward (I 938b, 1955) or the patrilocal band by Service (1962) represents such an optimal solution for hunting-gathering economies over a very broad range of environmental diversity. To what extent this kind of local organization was realized is what was debated at the symposium and in the ethnographic literature today.

This is not to agree entirely with the conceptualization of the hunting-gathering band as presented by Radcliffe-Brown, Steward, and Service, nor is it to say that the patrilocal band is the only possibility. As one interested in population genetics I am interested in the existence of a predominant form-not the full range of possibilities that might be realized, as perhaps would be the interest of the ethnologist. And I believe that what I call the lineage-band, a local group whose central members are of a single lineage, represents such a structure for hunting-gathering economics. The argument leading to this conclusion, based on the preceding assumptions, runs as follows: To go back to the concept of territoriality for a moment, we find that only among the Canidae and the Primates can multiple, sexually mature adult males be accommodated within the same, territorially autonomous, social group. This phenomenon reaches its highest development in the Primates. This has occurred in both cases through the evolution of a dominance hierarchy which can be learned by individual males and which is strong enough to override, with a certain effectiveness, sexual impulses.

In the genus Homo alone this system is extended to permit greater cooperation among males by the assignation of defined sexual rights which are relatively exclusive between a male and one or more females. The stable emotional ties engendered. permit cooperative behavior on a scale not found in other Primate societies. These affective ties are further extended outward in all contemporary human societies by means of what we term a kinship system, that is, symbols are attached to certain affective emotions early in the life of the child which can then be extended by extending the symbols to a group larger than the immediately co-resident household.

Furthermore, in contemporary societies in which kinship systems provide an important part of the jural structure of society, marriage bonds are channeled in such a way as to extend these affective ties between suprafamilial social units permitting cooperative or collaborative interactions which could not otherwise be maintained and which we recognize as forms of affiliation.

I suggest that the most efficient form of local organization for^ hunting-gathering societies, in terms of the criteria set forth previously, is one in which marriage rules lead to exogamic marriage practices between territorially autonomous groups. This makes possible a smaller territorial group, that is, a population whose distribution is more closely correlated with the distribution of resources than if all matings had to be contracted within the group and if marriage alliances were not maintained with all surrounding such groups.

These marriage alliances serve a number of functions leading to demographic stability: (I) they reduce conflict between territorially autonomous groups (in a sense they reduce autonomy); (2) they can provide access to “windfalls” for all the affiliated bands at certain times of the year, for example, beached whales, fish run sites, passes used by migratory herds, etc.; (3) they provide a form of insurance in that matrilateral connections can be utilized (assuming patrilocal residence) in seeking haven in times of hardship, that is, the population can temporarily redistribute itself.

I suggest, therefore, that the territorial social unit most often to be expected in societies with hunting economies is an exogamous band of the minimum size necessary to maintain continuing marriage relations with all surrounding such bands. This criterion is met by a band size of 35–75 individuals. The criterion of exogamy is met in unilineal bands by unilateral cross-cousin marriage. This form of marriage was not the case among the Birhor, as will be shown below. Other practices detailed in the ethnographic literature seem to serve also to extend affiliation to surrounding groups. Among the Andamanese an unusual elaboration of adoption practices seems to have fulfilled this function. This may also have been the case to a certain extent among Eskimo groups.

The bonds of alliance created by marriage exchange are reinforced by the annual aggregation of some or all of the bands so affiliated, at a season when such gatherings are economically feasible, and under circumstances hedged by ceremony. Assuming that the autonomous local groups are patrilocal lineage-bands, the major” ideological thrust of such gatherings would be to reinforce, symbolically objectify, and ceremonialize the matrilateral bonds affiliating the lineage-Gands whose de facto existence needs little symbolic mediation.

I t is for this reason that when the impact of Western civilization is such as to disrupt the norlllal dispersal of hunting populations- collecting them around trading posts, mission stations, cattle stat ion s or schools—the local band organization disappears almost completely and very quickly. The matrilateral Gonnections, enshrined in ideology, remain. They remain ill sOille cases to organize the part-society then existing; they remain for the salvage ethnographer to retrieve. As RadcliffeBrown said with respect to Australian groups “the local organization is the first part of the social system to be destroyed by the advent of the Europeans” (1931, p, 35).

What has been presented above is a model of hunting society, perhaps sketched so briefly as to be cry pt ic, which is not quite the inflexible band which seemed implicit in Steward’s formulation, The territorial group, hcnce the group with primary autonomy, is the patrilocal lincage-band. It is capable of splitting into even smaller, non-territorial family bands in response to seasonal scarcity. It is, on the other hand, allied with a larger group of bands which permits and even requires larger aggregates for at least short periods of time. This larger grouping, because of the circulation of females and the i siting patterns involved in marriage affiliations, tends to form a cultural unity, linguistically and otherwise.

Implicit in such a model is the idea that the size of the basic social unit (or units) is not a function of environmental richness, but instead a function of the system of filiation. Corollary to this, geographic variation in environmental richness is met by variation in the size of the territory occupied by the lineageband.

Investigations were carried on among the Birhor of Hazaribagh District, South Bihar, India, during 1961–62. The Birhor of this region arc still engaged in hunting-gathering activities and provide some information on the hypotheses proposed. In some important ways the Birhor do not meet the conditions assumed in the model of h unting - g ather i ng society. They are neither politically autonomous nor are they economically autonoIllOllS,

They live in an area that has been inhabited by tribal agriculturalists for a very long period of time. During the past loo-plus years, the area has also seen a large influx and growth of population of Hindu and Muslim agriculturalists who now far outnumber the tribal population.

The Birhor trade hunted and collected items to the villagers in exchange for rice. This exchange does not usually take place in markets. The principal gaIIlc animal nctted and traded by the Birhor is the hare. These, if they are to be traded, han’ to be traded on the sanll’ day they are caught, before they die. This necessity rules out the normal village market as an outlet. The Birhor also spend some time making rope from the inner-bark fiber of certain vines. These they also trade for rice.:VIany bands take such ropes lO village markets; others still prefer to avoid markets,

Not only do the Birhor live in a form of economic parabiosis with the agriculturalists, but they are in some ways a politically subjugated min o rit y. Their primary political dcknse, vis-à-vis villagers, is a carefully cultivated anonymity. This is a modus vivendi which fits quite well the migratory hunting pattern of the Birhor.

These conditions which arc a result of intensive interaction ‘with dominant groups makes the Birhor less than ideal as a basis for inferences about possible forms of social organization in hunting groups living only among other hunters, On the other hand, they have the great advantage of being hunters now. Therefore, the data on the Birhor is not a reconstruction from the memory uf informants. The difficulties inherent in working with “memory cultures” have, I believe, been particularly important in discussions of local organization among former hunting groups and this has not been sufficiently emphasized in the literature (cf. Chapter 17a).

The Birhor band is composed of relatives, and an individual can normally visit other bands only in which he has relatives. The ability to visit another band also implies the potential of being able to join that band on a more permanent basis. lVIost of the males of a given band will be members of a single patrilineal descent group. These descent groups are differentiated in having slightly different pantheons, slightly different ceremonies concerning these deities, and different sacred mountains.

In a field census of 25 bands it was found that thirteen bands had only one lineage represented; eight had two lineages; three had three lin.eages; and one band had four lineages represented. The latter band was atypically large and scmi-settled. In terms of number of individuals, even multi-lineage bands are predominantly of a single lineage. Also, since this is census data, it includes families who are involvt’d in extended visits or who otherwise do not feel themselves permanently attached to the group in which they were censused.

The residence arrangements found in these bands often do not fit simple descriptive categories such as patrilocal, matrilocal, or the more esoteric compounds. A brief description of the composition of four bands with which the author worked most intensively is given .

below.

Band one: At its usual maximum, this band contained nine households. Three household heads were brothers. One was the widow of a fourth brother. One was (he husband of a sister of these brothers. The seventh household head was a classificatory “brother” of the core sibling group, but his genealogy and theirs could not be traced reliably to a common ancestor. The eighth was a “son” of the classifi- catory brother. The ninth, who joined this band in 196 I, was the brother-in-law of the eighth.

Rand two: This band consisted of four households. The two elder household heads were brothers. One household head was a son of these brothers. The jc)Urth head was a son-inlaw of these brothers.

Band three: This was an unusual case in that the head of this group had been orphaned when very young and raised in a band not of his father’s lineage. It is difficult, under such circumstances, for a man to secure a bride and even more unusual that he can form a band around himself. This band consisted of six households. The household heads included the old man, two of his married sons, two sons of his first wife (by an earlier husband of hers), and the son of one of the latter men.

Bandfour: Structurally, this was the simplest band. If consisted of the household of an old man and the six households of his married sons, the offspring of his two co-wives.

Although the genealogical connections between families in some of the bands varies widely, and in some cases could not be traced, the situation is not viewed as being this complex by the Birhor. For them, the bands were m.ade up almost entirely of brothers and fathers and sons with their families. Of the 25 families involved, 22 fit this pattern, given the Birhor kinship system. The only three household heads who could not be fitted into a brotherbrother or father-son relation wen’ kinds of brothers-in-law. And these three, it should be noted, were not living with their wife’s mother’s natal band but with their wife’s father’s natal band. In sum, I believe that the category “patrilocal” is, at least, less misleading as a description of the residence system than any other rubric we might apply.

In a previous paper (Williams, in press), I presented evidence that extended familiarity with a local hunting area does not appear to be a factor favoring patrilocality by virtue of increasing hunting efficiency. Among the Birhor, a person who is a better-than-average hunter tends to be so whether he is in familiar or unfamiliar territory.

The Birhor do try to remain in a familiar area and, as this is by band, this takes on aspects of what could be called a “home range.” These ranges overlap extensively and if territoriality is defined in terms of boundary defense, there is no territoriality. Because of their subordinate position the Birhor are in no position to exclude anyone from any area. One suspects that if they were not subordinate in this way, the situation might be different. There is some resentment towards a hand which moves outside its traditional area and encroaches closely on another band. Also, there is occasionally considerably hostility toward the agricultural Santhal who sweep through on their annual “big hunt” which can remove quite a bit of game in one day.

In my earlier paper (in press) major portions of which were circulated prior to this symposium, two other points were made of interest in the present context. The first was that the descent unit characterizing the band is best termed a lineage. The second point was that the band, which has as its core a patrilineage, is extremely flexible in size and therefore not disrupted by seasonal variability in resources. By this, I mean that each lineage-bound splits into smaller band units during a part of the year and reaggregates, as the hunting improves, to again attain maximal size in May.

Marriage practices among the Birhor are fairly simple. There is only one clearly stated rule: a person does not marry a relative. Of course there is more to the situation in practice. A person also does not marry a nonrelative if raised in the same band. Brother’s wife’s “sister” is marriageable, and in fact is almost inevitably chosen when available. Although a number of forms of marriage are named and can be described by the Birhor, only one kind commonly occurs; this is sister exchange. Often a lineage “sister” will be used.

The prevalence of this practice seems to be connected with the difficulty of arranging marriage. Most bands in the near vicinity are ruled out as they are related by marriages in the previous one or two generations. The primary interest of the elder males is to provide a wife for a son in order to keep him in the band. Therefore, daughters become important in an exchange for daughters-in-law. Since the bands between whom the marriage is being arranged are at least semi-strangers, and thus not too well trusted, the equivalence of the exchange is assured by immediate exchange of “sisters” between two men who are seeking spouses. Once this has been successful, other marriages are easily arranged by similar exchanges between these bands for the remainder of the generation. This practice is also encouraged by the fact that women are less likely to run away from a band if they have “sisters” who are also residing there.

Marriage arrangement is also rendered difficult by the fact that market items such as cloth and perhaps cheap jewelry are required as wedding gifts. In these exchanges, there is a reasonable equivalence, but the fact that they must be made seems to insure that a man without support from several families cannot arrange a marriage. For this reason, orphans often marry older widows, where such exchanges are not required. Such men will also accept residence in the wife’s father’s band in return for a wife, although such instances are not common.

Such marriage practices deviate from those set out in the model given in the introductory paragraphs of this material. But they deviate in predictable ways. A strong tendency toward unilocal residence is present and exogamy is present. The model, however, requires unilocal residence and exogamy by means of crosscousin marriage to maintain ties of affiliation where stable territorial relations are present.

Among the Birhor the possibility of territoriality is completely ruled out as mentioned earlier. Therefore, for various reasons, bands do tend to “drift” over long periods of time, although they prefer to remain in a familiar area. Given these circumstances, the rule that one does not marry a relative maximizes the probability of developing kin ties with any new band encountered in the course of migration. This means that the marriage rule in practice among the Birhor effects the maintenance of marriage affiliation with surrounding groups in a long-standing situation of some territorial fluidity, much as does the rule of cross-cousin marriage in a situation of territorial stability.

Similar marriage rules whose effect is to produce obligate exogamy between territorially autonomous groups have been recorded for many hunting societies. It must be emphasized that this practice has to be viewed as different in some respects from exogamic practices between clans or other social units which exist within the context of a supraordinate political organization.

Were it not for its widespread ethnographic documentation, such a system of territorial exogamy would seem intuitively not very likely, precisely because the units involved in the exchange are territorially autonomous. One would expect marriages between such units to be more difficult to arrange than marriages within such units. In effect, the participants are sending daughters to, and are expecting daughters-in-law from, groups over which they have no control except that via sentiments which might exist as a result of prior exchanges of this kind. Further, such obligate exogamy, to my knowledge, is not found operating between territorially exogamous societies except in hunting bands. In the smallest tribal societies such as those of highland New Guinea, we find a certain, perhaps sizable, percentage of marriage between territorial groups; in royal lineages we find a great deal of marriage between territorial groups. This may be functionally equivalent in many ways to exogamy in hunting bands; that is, it serves to filiate otherwise autonomous groups.

If such is the function of marriage at the band level of social organization, we can draw further inferences concerning band organization. In human society it is not always possible to locate a strictly endogamous group. But, on the other hand, an exogamous unit can always be located. I t may be exogamic in both prescription and practice, or it may be so only in practice. A further implication of marriage as filiation is that the exogamous unit must be able to maintain cohesion indefinitely without the necessity of reinforcing this cohesion by marriages within the unit. From this point, two questions immediately suggest themselves: First, what is the largest group in a kinship-based society having a hunting economy in which social cohesion can be . maintained indefinitely without internally contracted marriages? Second, what is the minimum size of the exogamous group which can maintain such a system indefinitely?

The answer to the latter question can be given by considering a simple demographic requirement. That is, the exog-amous group must, on the average, provide marriage connections with the average number of territorially contiguous groups-as a minimum. This must also take into account vagaries of the sex ratio and other stochastic processes which will raise the minimum size above that required on the basis of a strictly mechanical model of man. Since the numblT of nearest neighbors, considering a large area, will be between 5–6 groups, this sets a fairly uniform minimum size, a size I suggested earlier as being 35–75 individuals.

With regard to the question of maximum exogamic band size, I would guess from reading ethnographic accounts and from my own work that this average maximum size will not be far from the average minimum size. The suggestion is that this band grouping which is exogamic readily tends to fission when it is of such a size as to be able to operate as two separate bands, given the minimum size criterion.

If this is the case, it is of great interest to the archeologist and the physical anthropologist as well as to the ethnologist. Since the number of nearest neighbors is a geometrical “given,” it is the same for 25,000 years ago as today— the same for a rich environment as for a poor environment. If there is not a great difference between the maximum and minimum size of such groups, then the archeologist, with an estimate of population density, should be able to say what minimum distance between sites would be expected to show culturally differentiated elements. If such predictions can be confirmed archeologically, the next question will be to ask how far back they can be confirmed.

For the physical anthropologist, the implications are clear. If such a system existed in Paleolithic populations, the rate of inbreeding and of gene flow are given directly from such a model. Inbreeding and gene flow are important factors in determining the rate at which populations can become physically differentiated and the ability of populations to maintain genetic polymorphisms.

This rather simplified hypothesis of factors determining band size in such groups can be, if it is a part of the story at all, only a part. The Birhor, as I have indicated, are patrilocal and exogamous but not territorial, and they do not practice cousin marriage. Yet the Birhor clearly strive to maintain their groups above a certain size. The critical minimum seems to be somewhere just below 25 individuals. At such a level, factors other than the demography of marriage exchange apparently come into play. These may be ecological, they may even be psychological factors. In any event they seem to need fuller exploration in the studies of ethnologists.

15. The Importance of Flux in Two Hunting Societies

Colin M. Turnbull

Introduction

This paper examines the flexibility of social organization characteristic of two East African small-scale societies, the Mbuti Pygmies of the tropical rain forest of the northeast Congo and the Ik of the dry, open mountains ncar the borders of Cganda, Kenya, and Sudan. Both the Ik and Mbuti, as hunters and gatherers, work within their respective environmcnts rather than attempting to alter them. Because neither is under the rigid controi a truly marginal economy might impose, each is able to maintaip a fluid band composition, a loose form of social structure, and to utilize flux as a highly effective social mechanism.

By flux I mean the constant changeover of personnel between local groups and the frequent shifts of campsites through the seasons. This apparent instability is, in fact, the very mechanism that gives these societies their cohesion. Both the Mbuti and the Ik arc composed of many constantly shifting clements forming established patterns. Flux is expressed as recurrent fission and fusion which affects the composition of local bands. A similar state of flux is found in a number of hunting and gathering societies, such as the Hadza and Bushmen, and may be characteristic of the majority of peoples described at the symposium. Before undertaking a comparison of flux in Mbuti and Ik society, a brief ethnographic sketch of each will be given.

Mbuti

The Mbuti Pygmies have heen previously described by Schebesta (1938–50), Putnam (1948), and Turnbull (1 gG r, 1965a, 1965b,!9^5C)- They live by hunting and gathering in the interior of the Ituri forest. There arc two economic divisions: the net-hunters who live ill large camps of seven to thirty families, bas(^d on communal or cooperative hunting, and the archers, who live in much smaller groups, and who hunt individually with bow and arrow. The residential units of both net-hunters and archers constantly change in composition as individuals and families circulate between the territorially based bands.

The total population of Mbuti appears to stand at about 40,000 and all of these are more or less closely associated with their Sudanic and Bantu neighbors, the Lese, Mangbetu, and Bira. This association has existed for at least several hundred years and as a consequence all pygmies speak only the languages of the village cultivators adjacent to them. What makes the continuation of their hunting life possible is the existence of the forest itself, which their neighbors fear to penetrate and to which the pygmies have become closely adapted. Their adaptation to the forest is conspicuously expressed in their technology and subsistence but is also deeply rooted in their ideology.

The Mbuti of the Epulu area, with whom I did field work in 1954 and 1957–58, practiced no agriculture, although they did engage in trade with the villagers, and spent brief periods living with these neighbors and eating their foods.

The villagers regard the Pygmies as their vassals, but, as I have argued elsewhere, the Pygmies manipulate this relationship to their advantage and maintain their primary orientation to the forest.

IK

Very little is known about the Ik (Gullivers, 1953; ‘Wayland, 1931), and to all outward appearances, they are predominantly cultivators. Yet, because one year out of every four or five brings a drought which destroys all crops, the Ik periodically abandon cultivation and revert to hunting. Thus, although they are cultivators for 75 per cent of the time, for the remaining 25 per cent they are 100 per cent hunters and gatherers.

The Ik, who number somewhat over 1,500, have previously been known as Teuso or Teuth, a name they reject as being applied to them only by the Dodos. They distinguish themselves from all pastoral tribes, such as Dodos, Turkana, and Karimojong, and they associate themselves with nearby hunting tribes, such as the Niangea and Napore, and even the more distant Tepes. They do so, despite a mutual unintelligibility of languages, by describing themselves and their fellow hunters as kwarikik, meaning “mountain people”; but they claim no closer connection with the other “kwarikik” than the common mountain environment. There is no thought of common origin. Neverthless, though separated by hundreds of miles, the environmental factor seems important enough to all kwarikik for them to recognize and feel a unity, against all others who are not mountain people.

The Ik live during the agricultural cycle in small, isolated, heavily stockaded villages that are almost as discrete as their hunting bands. Even within each village there are numerous divisions that prevent any solidification into effective corporate units much larger than the nuclear family. The divisions and subdivisions do not necessarily follow any lines of kinship, nor are they permanent. The groups may not even last the maximum duration of any one village site, namely, two years. Clan membership is only significant as a convenience for a rough reckoning of whether or not a proposed marriage is allowable. Marriage within the clan, however, does occasionally take place, and then the clan is said to have divided. One side of the dividing line is called “big,” and the other “small." These features are all directly comparable to the Mbuti band, and when the Ik village converts itself into a hunting and gathering band, the same principles of organization operate with equal effectiveness, and equal fluidity.

Comparisons

Environment

The environments of the Mbuti and Ik contrast sharply. The forest of the Mbuti, for the most part, is flat with occasional hilly regions. Its climate varies scarcely at all throughout the year. The maximum annual temperature range is 20G, and the normal daily variation is usually far less, perhaps 5” (between 75° and 80°F.). Rain falls evenly over the’ entire area, and is evenly spread throughout th;» year. Game and veg,etable supplies are similarly uniform in distribution, and are abundant throughout the area. There is nothing that makes one part of the forest more or less desirable than any other part at any time of the year.

On the other hand, the mountain territory of the Ik is for the most part arid, though studded with richly wooded valleys and gorges. Some areas are much more wooded while others may verge on desert. Far from the absence of seasons characteristic of the Ituri forest, the Ik habitat is one of violent climatic extremes. During the summer, the sun scorches the ground dry with its fierce intensity, while in the winter, the clouds obscure the sun, making the air cold and damp. Violent gales blow up and down the valleys, sweeping the torrential rains with them. Dry gorges become swollen rivers in minutes and, as quickly, become dry again. Small, tightly localized areas can become devastated by wind and flood, while all around these areas, it may be calm and dry. A region that fares well one year may fare badly the next. The game population is seasonally nomadic and edible vegetable supplies are tightly localized rather than evenly distributed.

In light of this contrast between the Mbuti and the Ik environments, a curious feature emerges. The Mbuti treat their stable e’nviron- ment as though it were unstable, cre’ating imaginary seasons of plenty and scarcity, while the Ik treat their highly unstable environment as though it were stable, and consequently bring upon themselves alternating periods of real plenty and scarcity.

However, it should be stressed that the usual notion that hunters live a marginal existence does not apply in either case. The economics, at first glance, may give the impression of being precarious, but in fact this is not the case. Famine or anything approaching it is utterly unknown to the Mbuti who have an axiom that “the only hungry Mbuti is a lazy Mbuti.”

The same axiom would have applied to the Ik, up to the drastic encroachments of both the cen’tral administration and the neighboring pastoral tribes in recent years which have hemmed them in, confined them, and hampered their traditional economy in every possible way. Yet even now it would be true to say that the only hungry Ik is one who is unable or unwilling to run the gauntlet of the various armed forces that forbid them access to major portions of their former subsistence range.

Land Use and Group Size

Naturally, the methods of exploitation of the resources vary according to the environment, as does the hunting organization and the gathering pattern. Both Mbuti and Ik hunting bands fluctuate in size from seven to thirty nuclear families, but whereas an Mbuti band is territorially circumscribed, and is so defined, it is much less so with the Ik. An Mbuti band can comfortably confine its movements from one year to the next within 100 square miles, due to the uniform distribution of the game in the forest. An Ik band is forced by climatic fluctuations to undertake seasonal movements over large areas, up to about 1 0,000 square miles.

Differences due to migration patterns of game and to seasonal differences in the availability of vegetable foods have social consequences as well. The Mbuti band, as a territorial unit, retains independence for most of the year and rarely needs to share its territory with another band, while the Ik bands often come together in larger concentrations in order to exploit the localized resources.

Subsistence Techniques

What is more remarkable than the differences between these two areas is the range of subsistence organization within each area. Among both the Mbuti and the Ik, two styles of hunting organization are in use. The Mbuti are divided into net-hunters and archers, and the Ik into net-hunters and spe’ar-hunters .

Regarding the Mbuti, there is no environmental reason why half of them should be net-hunters and the other half archers, although there may be some historical explanation. Nonetheless, the forest is so divided and each division regards the other as somewhat quaint and wonders how the other can survive. Yet on investigation it became clear that each half is well versed in the other’s hunting technique; visitors from one division to the other have no difficulty at all in adapting to the different hunting methods. And it is strange indeed that in the same environment and with equally adequate technologies, the net-hunters regard the brief honey season as a time of plenty, while the archers see it as a time of scarcity. Each group takes appropriate measures to meet the perceived situation, the net- hunters splitting into small units, and the archers congregating into larger ones!

Among the Ik, it is equally difficult to discern why some should hold to the net-hunt (of the Acholi type) while others say that hunting with spears is more effective. Again, there appears to be no discernible differences in habitat which would explain the different styles of hunting.

Part of the explanation may lie in the fact that in both cases, the environment is generous enough to allow alternative hunting techniques. In common with other tropical huntergatherers discussed at the symposium, gathering of vegetable foods is fully as important a factor in diet as is hunting. Although hunting is a more prestigious activity, gathering strategies play a large part in determining the band’s daily and seasonal movements. Even in the direction of their all-male spear hunt, the Ik take the distribution of vegetable foods into consideration and women are called on for advice. Thus a common basis of abundant vegetable food sources may make possible alternate and equally felicitious styles of hunting.

The Importance of Flux

Among the Mbuti, the constant process of fission and fusion that characterizes every band, is more predictable than among the Ik. The flux appears as a systematic pattern which I have analyzed elsewhere (Ig65a, I g65b, 1965c). The focal point of the process is the honey season, during which the net-hunters spread out into fragmented subbands, sometimes uniting siblings, sometimes dividing them, but always separating antagonistic elements. At the end of the honey season, the band begins to re-form, carefully avoiding any lines of fracture that remain unhealed. The net- hunters normally form bands larger than those of the archers, for a net-hunt demands cooperation between a minimum of six or seven nuclear families, and allows a maximum of thirty. The honey season, they say, is a time of such plenty that the game can easily be caught by . hand, and so there is no need for the large cooperative net-hunt. During the previous ten months, the necessity of constant cooperation, and the proximity and intimacy in which all band members must live with one another, invariably gives rise to numerous latent antagonisms and even a few open hostilities. If hostilities were to go unchecked, it would destroy the essential unity of the band and consequently ruin the success of the hunt. Thus the honey season is an important safetyvalve, allowing for the radical reconstitution of face-to-face groups.

The archers, in exactly the same environment, do precisely the reverse. They hunt in maximal bands only during the honey season, and for the rest ofthe year split up into minor and ultimately minimal segments. Their stated rationale is that the honey season is a time of poor hunting, and thus, maximum cooperation is demanded. The reality is that whereas the net-hunting band is of necessity united throughout the bulk of the year, the archer band is splintered into tiny segments, sometimes only two or three families, each independent of others. The ideal number of archers for either tracking or ambushing game is three; five would already be felt as unwieldy. The composition of these small segments does not remain constant throughout the year, anymore than does the composition of the net-hunting band, but the net-hunters must endure close association with much greater numbers of people than do the archers.

When hostilities come into the open, the solution is for one or the other disputant to pack up and leave. But the very size of the minimal archer segments means that the healing of internal disputes is of much less concern to them during the honey season than it is to the net-hunters. The problem for the archers is how to maintain any semblance of band unity under such conditions, and band unity must be expressed, for territory is defined by bands, just as bands are defined by territory. It is essential that territorial boundaries should be known so as to avoid any conflicts that might arise due to allegations of trespass or poaching. The need, then, is for each band to draw together all its scattered segments and to act as a band, within its territory, for at least some part of the annual cycle. On the excuse that hunting is poor at the honey season (I could find no justification for the excuse, though I prefer to leave it an open issue), the archers of each band gather together and participate in the begbe, or communal beat, which in technique is similar to the net-hunt.

The Ik band is similarly in a constant state of flux, and again, its composition is determined primarily by non-economic factors.

The process, though less predictable than that of the Mbuti, clearly does not follow any unilinear descent pattern in fission. Age and territoriality arc prime factors in the regulation and definition of the band, and the conflictresolving aspccts of its shifting composition are predominant. The Ik also have a honey season, which comes at a time of general abundance. Communal activity breaks down, even during the agricultural cycle, and major changes ill band composition are likely to take place in adjustment to politiral factors by separating antagonistic segmcnts. The honey season is followed by a termite season which breaks the band into its minimal segmen ts and takes these far afield—during the hunting year. At the end of this season, the bands gradually re-fOl’m into agricultural units, if rain promises well, and cooperate in the building of new villages and the hoeing of new fields. However, each village retains the internal divisions of the hunting bands, though, as with the Mbuti, the I k seldom revert to previous groupings.

Political Authority

There is no form of centralized authority in either society, apart from the nominal authority vested by the European administrations in their appointed headmen. This lack is perhaps at first surprising, considering that both Mbuti and Ik are divided into virtually autonomous bands with ho formal system of interrelationship, and that each band is constituted in such a way as to create the minimum of cohesion within itself. Even nuclear families are not the solid social entities we might expect them to be. It is difficult to see what are the threads that hold each society together as such, for despite the apparent lack of cohesion, both the Ik and the Mbuti are strongly united, as peoples, in opposition to their neighbors.

On the contrary, in each case there is a strong hostility felt towards any individual who aspires to a position of authori t y or leadership. Hostility is even shown towards those who, without any such aspirations, are plainly better fitted to lead than others by virtue of sheer ability.

Part of the answer lies again in the environment, for while it determines certain aspects of the mode of economic exploitation, within that general framework it leaves a large margin of latitude and always allows the certainty of, at the very least, a sufficiency of food, if not a surplus. This assurance of food is clearly seen in the unpredictability of the daily hunt. Even when game is known to be close, a band (in either society) may well at the last moment decide not to go hunting at all. It may find an excuse, such as a trivial dispute, or it may not bother to find any. The result may even be hunger on that day. Yet when already sated with successive days of good hunting and feasting, a band rnay decide to set off in the most unlikely direction and return, at the end of a long and arduous day, with nothing to show for it. Sometimes rain will make hunting impractical, and again hunger might ensue since neither society (the Ik considered in their hunting cycle, here) will store or keep anything for the morrow. After two or three consecutive rainy days, which is about the maximum that can be expected, a fine day may dawn and find the band still determined to stay in camp and attend to unimportant chores.

A band may hunt as a single unit, or it can split up and hunt as several independent units. There is no predicting which it will do among the Ik, where seasonal variations are strongly marked, yet among the Mbuti where there are virtually no seasons (except for the two months of the years during which honey is available in addition to all the usual food supplies) the process of fission and fusion is clearly predictable.

It all seems to be a rather topsy-turvy world for both Mbuti and Ik, where the things that happen are those that could least reasonably be expected. I am perhaps somewhat overemphasizing the lack of organization, but here that seems important, for all too often hunters and gatherers are expected to set a shining example of social organization at its simplest and its clearest, and to illustrate with crystalline lucidity the fundamental principles upon which all respectable social systems are based.

The fact is that both the Mbuti and the Ik work within their permissive environments rather than attempting to control them. They are unencumbered by the rigid imperatives that would be imposed by a truly harsh environment. Thus they are able to maintain a fluid band composition and a loose social structure; and are able to utilize this flux as a highly effective social mechanism, providing scope for action in all aspects of social life.

Conclusions

In summing up, I want to stress three points. First it seems clear that it is environmental permissiveness and not environmental rigor that makes possible the system of flux. Second, the major function of flux is not ecological adaptation but what could be called political adaptation. And third, the fission and fusion of individuals and groups does not follow lines of kinship.

The relative abundance of food that obtains in both traditional situations permits the system of flux. It allows for the fluidity of composition that affords a simple means of averting disruption without imposing a legal mechanism that would in itself be even more disruptive to the total society. The predictability offood supplies allows for neighboring bands to plan their movements so as to avoid conflict.

It is plain that in each case the process of fission and fusion followS’ lines of dissent rather than those of descent, and that the major function is conflict-resolving. At no time during the process can any trace ofthe concept of unilinear descent or affiliation be found. The band can only be defined as that group of individuals living and hunting within rec9g- nized territorial boundaries at a given time. A cross-cutting loyalty is formed by attachment and cooperation between age-mates, age being the vital bond that takes clear precedence over kinship, among both the Mbuti and the Ik.

Another function of flux might be said to be religious, for, by deemphasizing stability in interpersonal relations, the process throws the people into closer recognition of the one constant in their lives, the environment and its life-giving qualities. Under such conditions of flux where band and even family relations are often brittle and fragmentary, the environment, in general, and one’s own hunting territory in particular, become for each individual the one reliable and rewarding focus of his attention, his loyalty, and his devotion.

16. Southeastern Australia: Level of Social Organization

Arnold R. Pilling

In 188g, when only those aboriginal tribes in the eastern half of Australia were well known, Alfred William Howitt, one of the great pioneers of Australian ethnology, proposed that the well-watered coastal areas of Australia had an “advanced social development.” Howitt’s nineteenth-century evolutionary reasoning is best expressed in his own words:

With the exception of that part of NorthEastern Queensland where the Kamilaroi type touches the coast, the whole of the coast tracts, speaking blOadly, between the Great Dividing Range and the sea, both in Queensland and New South Wales, and between the Murray River and the sea in Victoria and South Australia, were occupied by communities having abnormal types of class system which in most cases count descent through the male line. These coast tracts, taken as a whole, are the best watered and the most fertile parts of Australia, and moreover, the richest in animals and plant food for an aboriginal population.

This coincidence of advanced social development with fertility of country is not without some significance. The most backward-standing types of social organization, having descent through the mother and an archaic communal marriage, exist in the dry and desert country, the more developed Kamilaroi type, having descent through the mother, but a general absence of the Pirauru marriage practice is found in the better watered tracts which are the sources of all the great rivers of East Australia; while the most developed types having individual marriage and in which, in almost all cases, descent is counted through the father, are found along the coasts where there is the most permanent supply of water and most food (Howitt, 1889a, pp. 33–34).

For those unfamiliar with the ethnographic descriptions available in 188g, it should be stated that the “Pirauru marriage practice” is a variety of ceremonial sexual license reported for the Dieri of the arid Lake Eyre area of northeastern South Australia, which Howitt and others in that era interpreted as the only occurrence, surviving into modern times, of the hypothetical original form of human mating, so-called “group marriage,” in which a group of males was alleged to be married to a group of females.

Today most anthropologists reject the Dieri pirauru as evidence of group marriage; in fact, during the twentieth century, group marriage has been rejected by the majority of social anthropologists as an ancestoral form of marriage. Likewise dismissed as invalid is any assertion that either matrilineal descent or patrilineal descent, by itself, may be taken as evidence of an advanced social stage.


Yet there still lingers on the suggestion that southeast Australia, the area specified by Howitt,.in some way represents a unique level of social development.

Some of the writers on aborigines claim as features of the aboriginal social structure of the Southeast such traits as powerful chieftairtships, social stratification as exemplified by hereditary chieftainships, and “a more advanced type of political organization” and, by reporting such occurrences, appear to validate for the Southeast a level of social organization not known elsewhere in Australia. To be specific, powerful “chiefs” or “kings” were noted by several. early chroniclers; for instance, Howitt (1904, p. 337) cites “‘illiam Thomas, a Guardian of .-\borigines in the Colony of Victoria in the 1850’s (W. Thomas, 1878, p. 118), as the author of such statements. Howitt, himself, alleged an emergent variety of social stratification when he suggested that the chieftainships of southeastern Australia tended to be hereditary. In 1940, Korman B. Tindale (p. 150) specified the Kamilaroi and \\,iradjuri tribes of southeastern Australia as having “the beginnings of a more advanced type of political organization”; while, in 1953, Tindale’s assertion was accepted by Joseph B. Birdsell, who attributed this advance to “matrilineal descent” (pp. 175, 193).

Southeastern Australia includes a relatively well-watered coastal area. Therefore, a proposition made by Elman Service becomes significant when the social and political level of this region is discussed. Service writes:

[An] interesting characteristic of Australia is that a regular progression from the rich, rainy, coastal areas with theif large and relatively sedentary local groups to the desert interior with its widely scattered, small, wandering population is -equally a progression from least formality and complication in the former to greatest in the latter area (1962, p. 74).

This assertion of less formality and comrli- cation for rich, rainy, coastal areas is in apparent contradiction to any claim of a more advanced social level for coastal southeastern Australia. However, both are claims of a special social and/or political level for this area. In contrast to such suggestions of a unique variety of social organization for all or a part of the southeast, in 1950, Birdsell (p. 279) stated that “‘the low level of social organization attained in Australia was constant and nowhere permitted the successful governing of large groups of people.”

Clearly, there is need for a review of the data on social aAd/or political level in southeastern Australia. The question which the present paper will try to answer is: Are the social and political patterns known for any of the aboriginal tribes of southeastern Australia sufficiently distinct to warrant a placement of these hunters at a different social level than other Australian aboriginals?

In answering this question it is relevant to consider (i) the data on powerful chieftainships, (2) the material concerning social stratification and hereditary chieftainships, (3) prohlems relating to the complexity of social organization in southeastern Australia, and (4) the evidence for a more advanced political organization among the Kamilaroi and Wiradjuri tribes.

Limitations

Before turning to the major topics, two serious limitations of the present paper should be made explicit.

  1. I have not been in a position to carry out actual on-the-spot ethnographic investigation of problems prior to preparation of this paper. In fact, probably no investigator working in the last fifty years could have collected much meaningful data in southeastern Australia on the subjects under consideration, for our interest is in the pre-European patterns as they existed in the Australian state of Victoria primarily, prior to 1835. Any field worker undertaking the present study in 1966 would suffer from the same lack of firsthand experience with the pre-European societies being considered, for as early as 1880 Howitt (p. 224) stated that the information he could collect on Kurnai local groups was “imperfect,” but was “all that ...[could then be] rescued from oblivion, and must suffice.”

  2. The number of primary sources available to me is to a degree inadequate’ for the study at hand; some of the rarer published sources have not been accessible. Only by months of touring local libraries, archives, and newspaper offices can one collect the data for detailed ethnohistorical analysis such as should properly be the foundation of the present paper. But even with the references available, tentative conclusions about the social level of the Southeast can be reached.

Powerful Chieftainships

Some early writers on southeastern Australia stressed the power of native leaders by using labels such as “king” (Howitt, 1904, p. 307, citing ‘William Thomas). For instance, Thomas Honery writing about the Wailwun tribes of western New South Wales in 1878, (p. 250) refers to their headmen as “kings”; while R. H. Mathews (1896c, P. 296), reporting events of the same tribe about fifteen years later, notes that the local whites use the “king” designation for a native local leader. Mathews (1894, pp. I 10–1 I) also mentions the use of the “king” label by a Kamilaroi for a headman during a ceremony in 1894; while a northwestern New South Wales kalf-caste informant of Beckett (1958, p. 104) still retained the usage in the 1950’s.

Howitt discusses the use of the terms “king” and “chief” by early writers and states:

I have chosen the term Headman as being less likely to be misunderstood than that of Chief, which has associations not applicable to the Australian savage. But if we must use the word Chief to imply a person having power to direct the people of his class or tribe, and that his directions or orders are obeyed by them I say ... [there] are Chiefs. Although when compared with those of well-known tribes in other parts of the world, their power is limited, yet it is an actual power command, coupled with a certain measure of ability to compel obedience (1904, p. 320).

Similarly, Le souer states, concerning the tribes of Victoria:

A good deal has been written and said about chieftainship, but nothing of the kind exists; there are certainly a few men-generally the boldest and strongest and very often the most mischievous-who acquire some ascendency in their tribe, but they have not recognized authority as the American Indian chiefs have (1878, p. 295).

It would appear that any assertions of powerful chieftainship for Victoria are overstatements of the type of leadership present; leadership was essentially of the same type in southeastern Australia as elsewhere on the continent.

Social Stratification

The assertion by Howitt (1904, p. 319) that in some tribes the local group leadership passed from father to son might be taken as a suggestion of a kind of minimal occurrence of social stratification. Howitt, indeed, states:

Most of the Headmen were related to each other by marriage, and thus in a family ... where there was a tendency for authority to become hereditary, there was a germ of a practice which, under favourable circumstances, might have established a privileged family, sueh as some of my correspondents have spoken of (1904, pp. 310–1 I).

However, I would suggest two possible circumstances under which transmission of local group leadership from father to son should not be taken as the emergence of social stratification as an occurrence in the aboriginal Southeast. The first ofthese explanations would acknowledge the existence of such transmission, but would question its presence in the preEuropean era. Beckett (1958, p. 104) notes that in some parts of Australia local cattle and/or sheep station owners formerly appointed a special local tribal elder as “king.” It would seem possible that the station-owning family would also impose English primogeniture as it applies to titles when the office of “king” was again vacant; indeed, the aboriginal group may have borrowed the British pattern of inheritance for kingships in imitation of English royal pattern without any overt statement by the station owners as to how the kingship should pass on to the next generation. As Howitt’s data were collected at least thirty years after the initial settling of the Southeast by Europeans, it seems that such a European pattern may have, by then, become accepted pattern in some tribal groups, without it necessarily reflecting pre-European patterns.

A second possible explanation rests in part upon the general presence in Australia of “gerontocracyjpolygyny,” as discussed by F. C. C. Rose (1960b) for Arnhem Land and other Australian groups. Le souer reports gerontocracyjpolygyny, apparently for all Victorian Aboriginal groups, when he writes:

Polygamy is universal; but it is generally the old men of the tribe who have the greatest number of wives. The reason of this is that they exchange their young daughters for young wives for themselves (1878, p. 291).

Rose (1960b, p. 119) notes the same pattern for the Kurnai.

A survey of the pre-European data for Tiwi (Pilling, 196 I, Table I) showed that five “big men” (or headmen) allegedly sired 76 children; while 14 “ordinary married men” allegedly sired 42 children, meaning, in this instance, that about two-thirds of the children born were sired by headmen. That is, a man with several wives and many children was more likely to become a leader, a leader had the prestige which brought him more wives, and more wives produced more children. On an average, such a headman, at his death, had more adult sons; there would be more votes cast by these sons than by all other men in the group; and this feature alone would favor one of these sons, usually the eldest, bein’g chosen as the next leader. That is, in southeastern Australia, as well as elsewhere on the continent, a son may inherit his father’s headmanship, without such an occurrence being a suggestion of social stratification, when the majority of men are the sons of headmen, as was probably often the case in aboriginal local groups.

Complexity of Social Organization

Elman Service’s statement (1962, P. 74) that there is “a regular progression from the rich, rainy, coastal areas with their large and relatively sedentary local groups to the desert interior with its widely scattered, small, wandering population is equally a progression from least formality and complication in the former to greatest in the latter area” needs detailed evaluation. Yet such a task is hampered by an uncertainty as to what is meant by “formality” and “complication” in his statement. The way in which one is to measure the complexity of social organization in hunting and collecting groups needs explicit statement. Does Service mean that there is a larger number of status terms, as he implies at one point (1962, p. 73)? Does he mean there are more non-kinship status terms?

What Service means by “formality” and “complication” becomes obvious by reference to his page 74, where he states that his criterion is “more kinds of sodalities,” which he seems to define as non-residential special-purpose groups (1962, p. 21).

To test Service’s generalization, let us compare the number of “kinds of” “non-residential special-purpose groups” (1962, pp. 74, 2 I) for several coastal areas with those for the desert tribes. Service lists for desert tribes three kinds of non-residential special-purpose groups: (i) initiation-centered sodalities, (2) totemic “clans,” and (3) blood-brotherhoods (Service, 1962, p. 74). Reference to the Kurnai for southeastern Australia shows at least two, and probably three, varieties: (I) initiation-centered sodalities (Howitt, 1904, p. 616–38), (2) totemic sex moieties (Howitt, 1 904, pp. 148–50), and (3) totemic “clans” (Howitt, 1904, p. 135). The Gournditchmara tribe of western Victoria also has two, and possibly three, types of sodalities: (I) totemic matrimoleties, (2) totemic matriclans (Howitt, 1904, pp. 124–25), and (3) initiation-centered sodalities (Howitt, 1904 p. 690). Clearly, one cannot say that the number of “kinds of ... non-residential special-purpose groups” is dramatically higher for the arid parts of Australia than for the two southeast Australian groups noted.

However, the invalidity of Service’s generalization becomes even more conspicuous when one lists the number of kinds of sodalities for the Tiwi, a north coastal group: (I) “one- grandfather” or aminiyati groups, (2) patrilineal dance groups, (3) totemic matriclans, (4) totemic supermatriclans, (5) totemic matri- phratries, (6) unnamed matrimoieties, (7) named subdivisions of matriclans, dependent upon recognized alternating genealogical generations, and (8) initiation-centered sodalities (Pilling, 1958, pp. 43- 73).

Let the preceding suffice to illustrate that there is no justification for claiming that Australian desert groups have a significantly more complex social structure than every Australian coastal group, including those of the Southeast. In fact, I would stress that we are very far from knowing how to measure complexity of social organization among Australian, as well as other, hunting groups.

Kamilaroi and Wiradjuri

The assertions of Tindale (1940, p. 150) and Birdsell (1953, pp. 175, 193) that the Kami- laroi and Wiradjuri tribes have “a more advanced type of political organization” are worthy of evaluation. I t should be noted initially that these recent assertions of an advanced type of political organization relate not to coastal tribes in southeastern Australia, but to the tribes of the headwaters of the great Murray-Darling river system. Tindale states that a general positive correlation between amount of rainfall and density of population does not seem to hold

among the Kamilaroi and Wiradjuri, where especially widespread communities seem to have developed in relatively fertile country (1940, P. 150).

Birdsell writes:

Consolidation among tribal groups seems to have occurred in the central interior of New South Wales, among such tribes as the Kamilaroi and Wiradjuri. This trend appears to reflect the development of a more advanced type of political organization based upon matrilineal descent (1953, P. 175)-

The statement of Tindale above does not indicate why he believes the Kamilaroi and Wiradjuri may be said to have “the beginnings of a more advanced type of political organization.” We must, therefore, attempt to test the validity of his claim without any hint as to the basis he used to determine the political level of these tribes. The relative power of the headmen, the existence of hereditary headmanships, and the number of kinds of sodalities (as defined by Service) will be our measures.

Powerful Headmen

In regard to the power of headmen, it is noteworthy that the term “king” was used by a native in 1894 for an image of a Kamilaroi headman, but in a context which seemed to disclaim that such a “king” was then viewed as awesome by young aboriginal males (Mathews, 1894, pp. 110-u). Mathews’ (1897b, p. 151) description of a northern Kamilaroi Bora, or initiation, ceremony in the I8go’s states that at one point during the ceremony “one or more of the head men of each tribe stand out in their turn.” Mathews (1894, p. 106) also reports the Kunopia local group of the northern Kamilaroi as having two leaders ill 1 894, while the Gundabloui group of the same tribe had but one headman. Elsewhere, Howitt writes of the northern Kamilaroi:

Mr. Cyril E. Doyle, writing to me of the Kamilaroi blacks of the Gwydir River, says that the old men of the tribe formed a council which managed all matters of importance. It was they who sent messengers to distant places (1889b, p. 322).

Concerning the southern Kamilaroi, Howitt states:

As to the southern tribes of the Kamilaroi, situated to the northward of Maitland, I have evidence dating back to about 1830. There might be two or three Headmen in each division of the tribe. Their position was one of influence and authority, and depended on the valour of the individual (Howitt, 1904, p. 302).

In regard to the statements above, it should be noted that both Mathews and Doyle seem to use the term “tribe” for local group; while Howitt employs “division of the tribe” to designate such a unit. If these interpretations of terms are correct, close analysis reveals new information. The early reports on the Kami- laroi, that of Doyle and that for the area north of Maitland, do not mention supreme local group leaders, but only a kind of council of several elders; the occasional occurrence of a single local headman is not mentioned until the 1890’S, more than fifty years after the area was first settled by Europeans.

Apparently somewhat in contrast to the above opinion, Howitt writes:

The oldest Headman would be the chief or principal man in the council of elders. He could carry a measure by his own voice, as the Kamilaroi have great respect for age. The Headman had a great amount of authority, and all the disputes among the members of his division of the tribe would be settled by him (1904, pp. 302–303).

The context of this statement becomes clearer when one refers to other of Howitt’s words:

The greatest Bora of all the Kamilaroi tribes was always held at Terryhaihai. All the Headmen were there, and the oldest was the principal, or president of them, and he could carry some decisions by his own voice (1904, p. 594).

Howitt’s reference to the power of the oldest or principal headman may easily be no more than a statement in relation to his functioning at the time of Bora ceremonies, and probably relates to the era after the provision of government rations during Boras allowed greater attendance at Bora ceremonies.

The concentration of power in the hands of the oldest headman at a Bora suggests that in this context a single male became the Kamilaroi leader. However, in contrast to Howitt’s statement suggesting that all Kami- laroi assembled for Boras, Mathews’ accounts (1894, 1895, 1 8g6c, 18g7b) of Boras in the 1890’s make it clear that such ceremonies did not bring together males from all Kamilaroi local groups, so that such a ceremonial leader did not become even temporarily the Kamilaroi leader.

In summary, it appears that Kamilaroi local groups may never have been led by a single senior male in the pre-European era, but were occasionally considered to be so controlled in the era after the Europeans appointed “kings. “ At Kamilaroi Bora ceremonies, the oldest headman present assumed the leadership of the whole assembled group, at least by the late nineteenth century. Such a pattern of leadership, is not especially complex or elaborate for aboriginal Australia, nor does it suggest an especially powerful occurrence of leadership for the Kamilaroi. For instance, Howitt (1904, pp. 297–98) notes that among the Dieri of the arid interior in the 1880’s that “some one [of the headmen] is superior to the others” of the tribe, and by then had become essentially the tribal leader; while Howitt (1904, pp. 314, 526) suggests for the Yuin of the southern coast of New South Wales that each local group had only one leader, one of whom assumed the headship of the Yuin tribal Kuringal ceremony.

What little is known of local group leadership for the Wiradjuri suggests that this tribal group had no more powerful local leaders than the Kamilaroi. Howitt says:

In the southern Wiradjuri a Headman is called Bidjabidja, and, as I have heard him described, is one who “gives orders to people, “ there being a Headman of this sort in each local division....

If, for instance, he were the oldest Yibai [one of the four intertribal sections-A.R.P.] ... he would be the head of the Yibai sub-class; but, assuming that he also became the Headman of the local group, then all the people of that division, not only of this “Rudjan” (sub-class), would obey him. Each totem also had its Headman. I have heard the Bidja-bidja spoken of by a term equivalent to “master, “ this being the analogue of the Biamban of the Yuin (1904, p. 303).

Another passage by Howitt causes one to realize that there were usually several headmen in a Wiradjuri local group:

In each locality the oldest man of any totem [named division of the four intertribal sections— A.R.P.] is the head of the totem there, and the oldest man of any totem is the head of that as a whole. Important messages, such as those relating to the initiation ceremony ... are sent by a headman. The messenger must be of the same totem as the sender, and the messenger is sent to the oldest man of the same totem in the locality to which the message is sent. This oldest man is the head of the totem there (188gb, p. 32 I).

Mathews (1897d, p. 118) in his discussion of the Burbung initiation ceremonies of the Wiradjuri indicates that the local headman of the host locality acts as the “initiator, “ that is the person who initiates arrangements for the Burbung ceremony. Beyond this minor office, no single individual seems to have held a special role during Wiradjuri initiations.

In general, we may say that the role of the headman among neither the Kamilaroi or the Wiradjuri seems outside of the range of political power found elsewhere in southeastern Australia and seems somewhat less powerful than that held by the Dieri tribal headman.

Hereditary Chieftains

The existence of families which have produced many chiefs or a strong tendency for hereditary chieftainships might have caused Tindale and Birdsell to assert the existence of a more advanced political organization for the Kami- laroi and Wiradjuri. However, the evidence does not support an unusual development of hereditary leadership for either group. Howitt says of the Kamilaroi office of headman:

It was not hereditary, but a man who distinguished himself as a warrior or orator would become a leader by mere force of character, and his son, if valiant, would be very highly thought of (1904, p. 302).

Of the Wiradjuri, Howitt writes:

The office of Headman was in a sense hereditary, for a son would inherit the position of his father, if he possessed any oratorical or other eminent ability. But if not, then the son of the brother of the deceased Headman would probably hold the position, and failing him some qualified relative of the same sub-class. But this was with the consent of the community, for the office went in fact by election in each division (‘904, p. 303).

Quite obviously, neither the Kamilaroi or Wiradjuri had sufficiently hereditary chieftainships to mark them as distinct from other Australian groups.

Number of Kinds of Sodalities

Service’s suggestion that social complexity should be judged in terms of the number of kinds of non-residential special-purpose groups causes one to turn to this topic in order to measure the sociopolitical level of the Kami- laroi and the Wiradjuri. This area is well docu., mented for the two tribal groups. The Kamilaroi have five varieties: (1) matrimoieties; (2) four sections, with determination of membership based on the maternal section (Radcliffe-Brown, 193 I, p. 230; Mathews, i897e, p. 160; 1904, p. 214); (3) totemic matriclans (Mathews, i897e, p. 159; RadcliffeBrown, 193 I, p. 230); (4) a second matri- moiety system, which was probably reflected in the residential pattern, and consisted of the “blood” and “shade” divisions (Mathews, 1904, p. 209; Radcliffe-Brown, 1931, p. 232); and (5) Bora-centered sodalities (Mathews, 1894, 1895, I 896c, 1897b). The Wiradjuri of the upper portion of the Murrumbidgee River (Mathews, i897e, p. 175), likewise, had four, or possibly five, kinds of sodalities: (1) matrimoieties; (2) four sections, wi th determination as among the Kamilaroi (Mathews, 1896a; 1897a; i897e, pp. 173–74; Radcliffe-Brown, 1931, p. 230), (3) named, totemic subdivisions of these sections (Mathews, 1897e, pp..17374), (4) Burbung-centered sodalities (Howitt, 1884, pp^ 438, 443, 445, 447, 452, 455, 456, 458; Mathews, 1896b, pp. 38, 39; 1896d; 1897c; 1897d), and possibly (5) “blood” and “shade” matrimoieties (Radcliffe-Brown, 1931, p. 232). The remainder of the Wiradjuri tribal area is alleged to have the Kamilaroi pattern of organization (Mathews, 1897e, P. 175).

If one rates societies in terms of the number of kinds of non-residential special-purpose groups, the interior Kamilaroi and Wiradjuri tribes show an incidence of a couple of more types of units than either the desert Aranda or the coastal Kurnai; they have fewer types of sodalities than the north coastal Tiwi. The Kamilaroi and Wiradjuri are not significantly outside of the range of other Australian groups, ifone measures political advancement in terms of the number of kinds of sodalities.

Matrilineal Descent

Birdsell asserts as mentioned before that the alleged Kamilaroi and Wiradjuri political advance is “based upon matrilineal descent” (1953, p. 175)- His phrasing suggests two possible interpretations: ( I) the occurrem:e of matrilineal descent among these tribes caused Birdsell to classify them as more advanced, or (2) the allegedly “more advanced type of political organization” was the product of some aspect of the matriliny of the social patterns of these tribes. Few modern anthropologists would accept as meaningful the classification of a tribe as more advanced simply because it had matrilineal descent, so we may safely exclude the first possibility, and limit our discussion to the second.

Can the existence of matrilineal descent have been the sole cause of whatever feature of Kamilaroi and Wiradjuri society is taken by Tindale and Birdsell as indication of “the beginnings of a more advanced political system” among these two tribes? A negative response would seem to be the only one which would be given by a person familiar with the social patterns of the tribes of eastern Australia. At least half of the present states of South Australia, New South Wales, and Queensland were occupied by aboriginal groups having matrilineal descent (Howitt, 1889a, opp. p. 31; 1904, opp. p. 90; Radcliffe-Brown, 1931, opp. P. 42). Within this area, many of the tribes adjacent to the Kamilaroi and most of the tribes of Queensland have the “Kamilaroi Type” of matrilineal descent (Howitt, 1 889a, opp. p. 31). Yet neither Tindale nor Birdsell suggest that the other tribes in this area have a social structure with an advanced political organization. Clearly the Kamilaroi type matrilineal social organization was not viewed by Tindale and Birdsell as sufficient grounds for classifying all these tribes as more advanced.

Conclusion

Only one conclusion seems possible about any assertion that singles out the Kamilaroi and Wiradjuri as especially advanced compared to other Australian tribes, especially the Queensland tribes to their north. A review of the major ethnographic accounts of these two tribes does not suggest that “the development of a more advanced type of political organization” was sufficiently under way in these groups to have been noted by the ethnographers and other chroniclers who considered these tribes during the last century when many informants were alive whose parents, if not themselves, had known the pre-European social patterns. Whatever political development has occurred to cause these tribes to be classified as “more advanced” must, indeed, be very recent.

Summary

A review of sources on southeastern Australia, an area where the environment was clearly richer than in many other parts of Australia, strongly suggests that the aboriginal tribes of this area did not have any especially advanced, or simple, type of social organization. Schemes proposing that, throughout, this area had either a significantly elaborate or especially simple social and/or political structure seem not to be based upon a detailed and careful analyses of the best available sources.

The most nearly accurate previously-published statement on the actual social level of southeast Australia is to be found in a 1950 paper by Joseph B. Birdsell (p. 279); “The low level of social organization attained in Australia was constant.”

17. Discussions, Part III

17a. On Ethnographic Reconstruction

Speakers: Birdsell, Gardner, Hiatt, R. Lee, Jovleg Itt, Rose, Williams, Woodburn

LEE: A central theme in the papers we have been hearing today concerns the reconstruction of ethnographic- situations that are no longer intact. We can take off from Hiatt’s paper (1962 and Chapter 10, this volume) in which he PQints out that Radcliffe-Brown among the Kariera was working with a detribalized situation. This field situation undoubtedly contributed to Radcliffe-Brown’s excessive emphasis on formal rules. The fact is that formal rules are a residuum in a memory culture, where the local organization has broken down. Naturally, the kind of material they feed the anthropologist will be their ideology, not the description of day-to-day movements. However, we have heard of a number of cases at this conference in which ethnographers have observed the local organization still functioning, and these observations-as distinguished from interviewing informants in a memory situation-show that groups in Australia, Africa, North America and India, are flexible and composite, and one sees people exercising options according to a number of variablesnot only kinship connection, but also personal associations, and ecology. This is quite a different picture from the simple model of the patrilocal, territorial, exogamous, band that has come to us via Service (1962), ou t of Radcliff’e- Brown (1931). What we are actually observing is the work group, and it is important for us to distinguish, as Leach suggests, between the conscious models of ideal group structure and statistical models derived from the real choices of real people. No one would deny that the conscious model is also a part of reality; it is. But it is equally dangerous to assume, as Radcliffe-Brown did,. that the conscious model alone was an accurate or even approximate reflection of group structure.

BIRDSELL: I think Dr. Hiatt has raised a very important point, which I would like to discuss both indirectly and directly. The question is not so much his quarrel with’ Radcliffe-Brown -which Stanner answered last year (Stanner, 1965) although he left it hanging on this note: “All the large communities that have been quoted in the 1962 paper by Dr. Hiatt, are indeed, modern ones and post-acculturational and, in fact, tied—to one degree or another— to European economy.”

Now, before approaching that, let me say that my own belief is that precontact Australians can have lived in as large a group as that for which we can define a subsistence base. Reconstruction is- a tricky business, and I would like to present and discuss three recent examples of this approach.

  1. One is the very elegant reconstruction of Ainu social organization and local grouping by Dr. Watanabe (1964 and Chapter 7, this volume). It is a very careful reconstruction going back almost a century in time. In it he has used ecological data and genealogies to provide the demographic variables.

  2. Ted Strehlow in a recent publication (1965), rather strikingly reinvents Radcliffe- Brown’s geometrical pattern of patrilineal local groups. I should note that Strehlow was born in Aranda territory, grew up speaking as much Aranda as he did English and German, and he has devoted his life to the study of the Western Aranda. In short, I think his data are as impeccable as data collected this late could be. He does find rigorous territoriality, defiaed in terms of people and totem associations and having definite boundaries.

  3. Recently Norman Tindale (1962b) published a reconstruction of the Kaiadilt people and their country. This is a group of (originally) 120 people living on Bentinck Island in the Gulf of Carpentaria. He found them about four years after they were removed from the island; previously, they had been totally isolated and, therefore, were still within twelve years of precontact structure. Interestingly enough, as far as you can define marine densities by terrestrial areas (which is not a good measure), these are as densely situated a people as any in Australia; they come to about three persons per square land mile. Among them, he could reconstruct eight bands, each territorially resident in a fixed area-even to the detail of discovering one disputed interband territory. Band size, ranging from something like 8 to 35, is interesting, since they lived by generalized food resources. These, then, were small bands in enormous density. Though, perhaps, the most advantaged of tropical island peoples, they show the same rigorous territorial pattern projected by earlier modelists.

Now, I think no one will quarrel with the Sydney workers in terms of the reconstruction of social organization; they have done it superbly. But, as Stanner also points out (1965), there are almost no workers in Australia who have fitted real men into real country. So I suggest that what we need is the insertion of significant ecological analyses, based upon significant demographic data. I would say, then, that one cannot go to a modern mission station, government station, or mining town and-without inserting demography and ecology-reconstruct previously functioning local group structure, or, for that matter, territoriality. These things vanish with contact.

HIATT: I would like to deal with several of the points raised by Dr. Birdsell. It is at least questionable that Stanner (1965) has “answered” my 1962 paper; it was more an exegetical attempt to show that RadcliffeBrown had not said the things I attributed to him. In one part of his paper, he does say-as Birdsell noted-that the modern studies have dealt with postcontact communities. This is a curious assertion, since, as far as I know, all the studies of local organization in Australia have involved communities that have been seriously affected by European contact. RadcliffeBrown was in no better position to study territoriality than the rest of us.

I do not think that my rebuttals of Radcliffe- Brown’s ideas add up to a new gen::ralization. Given the ecological diversity in Australia, it is unlikely that there was a single type of local organization. And I suspect that it is too late now to establish the main varieties. However, if Tindale has shown that there is a horde in the Gulf of Carpentaria, I am quite happy about it. I did not say categorically that there was not one, but only that so far no one has ever seen one.

Concerning Strehlow’s work-the Hermannsburg Mission was established a long time ago, about 1875 (Strehlow, 1965, p. 142), and his father’s notes refer to a period twenty years after contact. As far as I have been able to make out from Strehlow’s written material, he provides no empirical support for Radcliffe- Brown’s model. In fact, I quote him in my 1962 paper as giving evidence that is inconsistent with the model.

No one, I think, is satisfied with our methods for reconstructing the past. The people that I worked with in Arnhem Land had been living a nomadic life until a year before r arrived. I do not suggest that they were completely unaffected by European contact; some of them had been wandering to neighboring missions or even walking 200 miles into Darwin. In 1957, the government established a settlement in their home country and, once the word got around, they all moved in very quickly, so that it was not possible for me actually to see an indigenous, nomadic, ongoing economy. But I did live for a month away from the settlement with a group of about fifty people, plotted totemic sites carefully, asked about boundaries (there were none), asked about use of resources around totemic sites, and, in general, observed about fifty people from several different patricians living harmoniously together- specifically, not arguing about the fact that some members of the group were living on totemic estates that did not belong to them. Similarly, I recorded many disputes that had occurred, roughly, between ‘940 and 1957—- when the people were still living in the area largely unaffected by European settlement. These disputes were narrated in great detail, and I found that one of their favorite mnemonic devices for recalling disputes was the memory of the places where the disputes happened and the people who were there. So in the course of describing a dispute, an informant would name a dozen or so characters; and this makes it possible to say of a particular waterhole belonging to a particular clan at a particular time, there were many people present who did not belong to that particular clan. All in all, I don’t think my opportunities and methods were vastly inferior to Tindale’s on Bentinck Island.

MEGGITT: I should point Ollt in passing that Stanner’s 1965 paper fails to rebut Hiatt’s paper (1962). It might be better construed as an exercise in ancestor worship.

ROSE: 1 also wish to address myself to Birdsell’s remarks. I think one should realize that Strehlow was dealing with the Western Aranda in the early part of the century, and they had in fact been in contact with a white settlement since about 1875 (an area to the west of them was settled about that time). He also used the genealogical method, which I, at least, try to avoid.

I must confess that I have not yet read Tindale’s paper on Bentinck Island, but I would like to comment on Tindale’s reports on the Groote Eylandt people, who were in a pristine state when he inv(‘stigated them in 1922 (Tindale, 1925). Tindale took photographs of groups of aboriginals whom he describes as belonging to a particular patrilineal totemic group, and I had these photographs identified in 1948. The identifications entirely confirmed Hiatt’s and Meggitt’s conclusions, for it turned out that each of the photographed groups belonged to several totemic groups.

Williams: I want to return to Birdsell’s ideas about reconstructing local residence arrangements, about which I take a pessimistic view. B. K. Sen published a paper in 1955 (Sen and Sen, 1955) in which he mapped the site of a group of Birhor and specified the membership of each household. Working with the same people, only six years later I was unable to reconstruct the situation Sen had described. Furthermore, after keeping track of every move over a fourteen-month period, I tested again by walking around old sites with the men and asking them where the houses had beeIl and who had occupied them, and could not get consistent answers from man to man. The lack of ideological reification of the patrilocal group is such that I suspect that if the Birhor were put all in oiie place, fifty years from now, no one would know that he had occupied a patrilocal group. Therefore, I have no confidence in our ability to reconstruct these things, once the de facto arrangement has ceased to exist.

WOODBURN: I abo want to talk about formal rules and problems of reconstruction. I think that with the Hadza it would be quite impossible to get an accurate idea of their system, even in the present, if one were to rely merely on informants. Their ideas about what they do differ very substantially from what they actually do. One (and only one) of the factors that influences this is the fact that the Hadza are very eager to gain the esteem of members of other tribes. A great deal of what they say is affected by this desire, especially when they are talking through interpreters. When asked about aspects of their own amorphous and elusive social organization, which they with some justification fear would be ridiculed by most outsiders, they commonly take refuge in describing the formal rules of neighhoring tribes and particularly of the Isanzu, \’. “ nil they know best. The Isanzu have a matrilineal system which is only imperfectly understood by most of the Hadza and the sort of description that one is likely to get as a result of questioning rather than observation is of a decidedly odd and rather incoherent form of matrilineal organization, whereas in fact they have a nonunilineal system.

Gardner: Several papers have raised the question of whether devolution, detribaliza- tion, or cultural simplification has led to retention of formal structures (clan names, formal rules, and so on) for which there are no ongoing practices that can be observed. Other discussants have noted the very frequently occurring situation in which discrepancy between rule and practice can be ascribed to additions rather than losses. The Paliyans of South India are another case in point. They told me that they were patrilocal, when by Helm’s procedure ( 1969) for determining de facto bilaterality they are 53 per cent virilocal and 47 per cent uxorilocal. They say that they universally marry cross-cousins and cross-nieces, which 21 per cent of them actually do, but 19 per cent marry parallel-cousins and parallel-nieces, which according to the “rules” would be classificatory incest. They claim to llse Iroquois kin terminology, which in fact they do use in front of outsiders— including anthropologists—but by hearing “mistakes” and conversations where my presence was not noted I found that they use Hawaiian terms among themselves. In order to appear respectable in culture contact situations, they attempt to be orthoprax, in terms of the culture of their neighbors; but this is all very superficial and recent, the “rules” are not a residue of old, abandoned institutions. This same situation apparently obtains among the Mbuti Pygmies, and among the Yanadi and Chenchu of India.

17b. The Problem of Lineage Organiza1 Ion

Speakers: Damas, Deetz, Sinha, Williams

DAMAS: In deciding the issue of the presence or absence of lineal organization in hunting people, we ought to consider further the question of acculturative influences. For instance, among the Birhor there seems to be some evidence that the lineages themselves may be borrowed from surrounding agricultural peoples—who have the same name for the lineage. This raises a question in regard to Williams’s model as to how essential the lineage is. Specifically, one wonders whether the Birhor would have worked out some alternative solution to this adaptive problem if they had not had an opportunity to borrow a solution from their neighbors.

WILLIAMS: I think we agreed at the 1965 conference on Band Organization’ that patrilocal and patrilineal bands should be separated, and specifically over the matter of lineage. Perhaps, we can leave aside the terminological questions and discuss specifically what has been borrowed. The Munda and the Santal have named clans, and these are the names that the Birhor have borrowed. I can assign no function to the names themselves; they do not regulate marriage. That is, they are not exogamic units, because one finds marriages within named clans. These are supposedly patrilineal, but I also found men who were agnatically related but gave themselves different names. These, then, are borrowed and do not fit the model.

On the other hand, this is not what I call a lineage. What I do call a lineage—rightly or perhaps wrongly, as I discover from talks with Woodburn—is a group of males who inherit property. There is very little symbolic objectification of a lineage, so to distinguish a lineage required a lot of looking. It is true, however, that each group of males inherits a pantheon, and each learns a different way of treating it. Also, members of a lineage band behave differently towards one another than they do in their interactions with people outside the lineage. For example, if they are patri- lineally related males and another group of patrilineally related males come to visit, they move in directly; while if the visiting group is related in any other way-e.g., in-laws through the wife—they first establish a camp several hundred yards away and then feel out the social situation before moving” in. The distance of their camp from the resident group is a function of the distance of their kinship relation.

SINHA: I want to draw upon the earlier reports by Roy on the Birhor of the Ranchi

I. Held at the National Museum of Canada, Ottawa, under the chairmanship of David Damas. district, the district neighboring on Hazari- bagh. It appears in Roy’s report (1925) that the totemic clans, rather than the lineages, seem to be quite important as exogamous units. Of course, we cannot be very sure how specific his information was. But if we go back to our previous idea of devolution, it could possibly be the case that the Birhor at some time in the past, like many of the other Munda-speaking group of tribes, had totemic clans with a greater depth of lineages when they were practicing shifting cultivation. Then we would suppose that only the symbolic names persist; while the operational units have become lower-depth lineages with bilateral compromises. Does this sound feasible?

WILLIAMS: It does sound feasible, at least. Dr. Sinha has already cited evidence on some east-central hunting groups which, thirty years ago, were not hunting as much as they do today; they were shifting agriculturalists. This raises a suspicion that this condition has been prevalent throughout India. Also, there is a general feeling that one simply should not find hunters in an area in which there has been a high civilization for thousands of years. In any case, if they were agricultutalists in the past, I was not able to find any evidence for it beyond the names already mentioned.

Now, as to Roy’s data: I did not work in his area, nor do I know what techniques he used. I can say, however, that I did ask the Birhor for these clan names and when I sat down with an informant and asked him into which of these groups he would marry, he would give me a definite answer, an answer which, however, bore no relation to actual marriage choices.

DEETZ: I would like to comment on the distinction between patrilocal and patrilineal bands. It occurs to me that a territorially based single patrilocal extended family is a de facto patrilineage or, at any rate, is so perceived by us. If, however, we compare this to a Nuer patrilineage, we should look at the type of corporate ownership that we are dealing with. In the case of the Nuer, the ownership is of cattle, or of something else tangible. But in most cases in which we speak of corporate ownership on the part of hunting-type patrilocal bands, the thing that is owned and shared is something infinitely expandable. For example, if I have eight boys and -sixteen cattle,

I can give each boy two cows; while if I have sixteen gods, I can give all sixteen to each of the eight boys. Therefore, we should distinguish carefully betwecn what we might call corporate property as opposed to corporate ideology. The real need for sharing tangible material that cannot be split down and is passed on unilineally is of a different order from the treatment of something like totems, pantheons, legends, or far-distant geographical places, which can be diffused in all directions and do not demand the same sort of corporate action.

17c. Analysis of Group Composition

Speakers: Anderson, Balikci, Helm, Laughlin, R. Lee, Marshall, Slobodin, Woodburn

Slobodin: I have a question for Dr. Helm. I wonder if kinship is talked about in two senses in her paper—kinship in posse and kinship in esse\ i.e., the range of possible kinship, as distinguished from kinship actualized in activities. When a couple or family moves into a group (and it turns out that there is almost always a sibling of one of the spouses in the group), they are actualizing one of many possibilities. Among the Kutchin-for whom I present data (Slobodin, 1962) similar to that used by Dr. Helm—first cousins are equated with siblings, and it is in fact very difficult for a couple to move into a group where neither spouse has a sibling. So it has been enlightening to me for her to emphasize the importance of this sibling relationship. Pehrson (1957) did this for a northern pastoral group. But in both cases, the relationship is specific to a purpose actualized there. It is not a princi pIe of kinshi p in general; there is kinship involved in other kinds of groups, too.

HELM: I do not see the distinction made by Slobodin between the range of possible kinship as distinguished from kinship actualized in activities as being a point of confusion in my paper. I have used the generic term “kinship” in the same way that I have used the generic terms “resource” and “range.” These are merely the most convenient and simple designations that I could muster to project the broadest perspective. I have taken pains in this paper and others (e.g., Helm ig65a, ig65b) to specify the particular kinds of kin relationships that are “actualized in activities,” the activity in question being co-residence or contiguous residence as a socioterritorial identity. The “actualized” kind of kinship connection is the primary bond of sibling-to-sibling or child-toparent as it is “activated” bilaterally, from either spouse of the conjugal pair, to allow residence entree ofthat pair into one or another of the groups where primary consanguines reside.

One non-sociological variable that I see to be quite as important in the creation and duration of Dene local bands as Pilling sees it in his Australian material is progenitiveness and the physical survival of the resultant sibling set. Three or four or more siblings who survive into adulthood along with a forceful father or brother tend to stay together around that figure, and they are apt to pull in male affines as co-resident group members, as well. Thereby, the group tends to endure in time and space, focused around the bilateral primary linkages between the conjugal pairs that result from marriages by the sibling set, including the females as well as males of the set among the Dene. Herein lies the bilaterality of the Dene system.

As to the fact that cousins are called “siblings” among the Kutchin, this is equally true of the kinship terminology system of the Dogribs (and three out of every four Slave Indians give this nomenclature as well). But, like the Kutchin, every Dogrib knows his cousins from his siblings. Dr. Slobodin makes that same point I do when he says that “it is, in fact, very difficult for a [Kutchin] couple to move into a group where neither spouse has a sibling.” In my paper, I deal with real siblings, not terminological “siblings.” It is true that sometimes the “effective sibling set” (as discussed in Helm, Ig65a) may include more than uterine siblings. I have cases of an adopted sibling and of a parallel cousin becoming the leader of a uterine sibling set. I think these cases simply demonstrate that the basic two- sex sibling unit may be socially extended for the uses and purposes of a dominant male. The fact that the kinship terminology allows cousins to be termed as siblings may indeed aid in the extension of the fraternal sentiment, when it is socially or economically advantageous to do so.

Woodburn: I have been asked to describe the place of sibling ties in Hadza residential arrangements and also to specify their system of cousin terminology.

In my second paper in this volume (Chapter I i) I have described how the most important bonds in the composition of Hadza camps are those between husband and wife and those between mother-in-law and son-in-law. These two types of dyadic bond in combination give rise to a non-corporate unit which has the genealogical form of a matrilineal extended family and which I describe as a simple residential unit.

As a consequence of the two types of tie that make up the simple residential unit, young married sisters commonly live together while their mother is still alive, although they do not have important property obligations to each other. After the death of their mother they may well have married daughters of their own, and residentially their focus of interest will be on these daughters rather than on their sisters. Two sisters each of whom is at the apex of a simple residential unit will often live together from time to time but are unlikely to live together regularly unless their units are additionally linked by other ties. If the son of a woman at the apex of a unit marries, the unit he marries into is unlikely to live regularly with the unit of his mother and sisters unless, again, there are multiple ties between the units. Quite often two or more brothers marry into the same unit or a brother and a sister marry a brother and a sister. In one instance two sisters and two brothers had married two sisters, a brother, and a halfbrother. In cases of this sort, where multiple bonds link a pair of units, they are to be found together much more frequently than would otherwise be the case.

Since a man normally lives with his wife’s mother, he will not find it easy to maintain two marriages simultaneously unless he marries two sisters, which is possible but disapproved. When a man does have two wives, it is very often a transitory stage between two monogamous marriages. This is apt to bappen when the first marriage is not going well: the man marries another wife and eventually chooses to maintain one or the other of the marriages.

Rather surprisingly, the cousin terms of the Hadza are of Crow type, although in other respects the system is not Crow. The Hadza say that a man should not marry a first cousin of any sort, but one occasionally finds cases of marriage with either the parallel or the cross-cousin. The preferred marriage is a cross-generational one with the classificatory sister’s daughter, though, as with any cross- generational marriage, it is rare in practice.

LEE; I shall be speaking about the Dobe- area!Kung Bushmen of Botswana (Lee, 1 965) and their neighbors, the Nyae Nyae!Kung of South-West Africa, studied by Lorna Marshall (1957, 1959, 1960). Unlike the Hadza, among whom marriage seems to be a transitory union, marriage among!Kung Bushmen is very stable, and the nuclear family is the basic cohesive unit that stays togethe,’ throughout the year. This does not mean the!Kung have a family level of organization. They live in multifamily groups of 1 0–50 individuals with modes at 15 and 25 persons.

One must realize that censuses of a camp will show different totals from day to day; the 1 5–25 figure, however, is a consistent central tendency. Thus the two basic units of !Kung society are the nuclear family, and the multifamily living group which I prefer to call the camp, although others may want to call it a band.

When one tabulates the links within the camp using the method developed by June Helm ( 1969), one finds male-male, malefemale, and female-female sibling links, as well as affinal links of both sexes. Thus in terms of genealogical composition, the Dobe!Kung camp looks very much like the northern Athapaskan microband.

Helm also pointed out that among the Dogrib, a strong group of siblings will tend to attract male affines to the task group. This is a prominent feature of !Kung groupings and such a mode of affiliation may be characteristic of other hunter-gatherer peoples as well. Among the!Kung, demographic factors such as family size and survivorship account for much of the variance observed in social groupings. Large families tend to attract others, and small families attach themselves where they can; with people moving as often as they do, it is impossible to discern the operation of jural rules of residence. The!Kung camp does not constitute in any way a patrilocal band. Following the suggestion of Knight on the Naskapi (1965), I prefer to call their residence arrangements “pragmaticolocal.”

We should not be surprised to find that where jural rules do occur, they have built-in options. Consider the Bushman name relation, first described by Lorna Marshall (1957). The ascription of personal name is ambilateral, with a patrilateral emphasis. The Bushmen say that the first child of each sex should be named after the paternal grandparents and subsequent children after the maternal grandparents; in practice, half of the people have patTi-names and half matri-names. Since each sibling group is split by the name relationship, no unilineal corporate entity can emerge. What is more important is the fact that the name relationship restructurcs kinship terminology. If you share a name with another individual, then the kinship terms that you use radiate from the fact of the common name, and these name partnerships leapfrog genealogical connections. This makes it difficult to talk about kin term typologies since Ego’s cousin terms will be Eskimo, or Crow, or Omaha, depending on his name relationship!

In addition to the!Kung, name sharing is found among the Eskimo (Guempie, 1965) and among the Kaingang of South America (Henry, 1941 [1964]). It might be interesting to determine whether name sharing is common to other hunter-gatherer groups.

MARSHALL; The Nyae Nyae!Kung and Dobe-Area!Kung have much in common. I found that among the people we worked with, the ideal naming pattern stated by Richard Lee was followed consistently. After using the names of all the grandparents in, proper order, a further child was usually named after one of the father’s siblings or a sibling’s spouse. I might note, regarding the effect of common naming upon kinship terminology, that the perspective that a person adopts is primarily that of the specific individual after whom he is named. Since there are relatively few names to go around, the chance sharing of a name with several people does not involve quite the same number of rules and patterns in the use of kinship terms. The people we worked with claimed that they did not marry lineal kin or collaterals up through first cousins, and this seemed to be quite consistently maintained. However, they did frequently break another stated marriage rule to the effect that one should not take a spouse having the same name as one’s own parent.

BALIKeI: With reference to the Netsilik Eskimos of the Arctic coast, I have asked myself the question: What keeps these people together in a band? And I have found several reasons . for it. There are, of course, such things involved as kinship ties, the vital necessity of cooperative acquisitive activities, and some ceremonial activities. But there is another class of social facts that tie people together; namely, patterned dyadic relationships, which seem to be pretty important in this area of the Arctic coast. One can enumerate a number of these dyadic partnerships—wife-exchange, . meatsharing (which are extremely rigorous and involve named dyads), joking, avoidance, trading, drumming-singing (which sometimes go together with wife-exchange partnerships), wrestling, old-friend, and name-sharing. All these partnerships are reciprocal, and most of the time they are contractual; although meat partners are given to children soon after birth. These are peace-making devices in the camp that both cut across kinship lines and tie extended families together. Sometimes such dyadic relationships are ambivalent, especially in the case of wife-exchange partnership. After exchanging wives over a period of months, husbands are likely to begin disliking one another and wind up as enemies instead of friends. This is manifest in the songs sung at the drum dances. I wonder whether dyadic relationships play as important a role in other hunter-gatherer bilateral societies.

LAUGHLIN: I think it is interesting that dyadic and name relationships have survived in rather robust form in the Aleutian islands, where the population has been severely limited and where there have been some serious changes. One of the most common practices is that of naming after a common event. There is also a reciprocal use of single kinship terms, so that two persons may be calling each other uncle. Another important relationship that shows up occasionally among the Eskimo is called anakiisak. When a person is born, an older persons is named his anakiisak or protector. The older person is supposed to protect the infant through all the stages of life and in embarrassing circumstances (hence, if a child wets on the floor, an old woman may do the same thing, in order to partake in the shame). When the young person gets older and his anakiiisak gets very old, the younger one helps the older one out of this world. This seems to function to a considerable degree today, more, perhaps, in the Aleutians than in Kodiak or in western Alaska. There are other such relationships. The two men that went out in the two- hatched kayak were “partners”; they had to get along. Some magic was possible there, and some deaths resulted.

ANDERSON: Concerning the analysis of group composition, the issues which I wish to emphasize represent the other side of the coin of arguments presented by Sahlins (1965) on the relationship of group ideology and group composition. Of course, this is not an either/or matter. Deriving structure from ideology or composition involves different activities but both have value. These analytical activities are complementary. Acknowledging this, I will emphasize the group composition side in order to relate a number of points which have emerged in the papers and discussions.

One of the quite general findings concerns what has been termed fluidity in group composition and group size over time. Though shifting group affiliations do not surprise us, the high frequency of their occurrence among hunters and collectors does warrant further investigation. From the examples we have been presented, they appear to occur (though probably with differing frequency) among relatively settled as well as nomadic peoples, where orientation is lineal as well as bilateral, where territorial attachment of groups is tight or loose, and where the habitat is relatively richly or poorly endowed. Indeed, if fluidity of group affiliation is as general as it seems, then hunting and gathering peoples would appear to be not only what Sahlins terms the “original affluent society” but also the “original mobile society,” at least in terms of residence arrangements. And if in the original affluent society wants are restricted, then in the original mobile society shifting alignments permit individuals and families to maximize fulfillment of their wants (material and non-material).

By “fluidity” in the composition of social units and “residential mobility, “ I am summarizing an apparently widespread but not universal feature of hunting and gathering societies which is to be distinguished from strictly seasonal fragmentation or shifts. The Ainu have seasonal residential shifts as do all hunters and gatherers, but Dr. Watanabe’s reports do not indicate fluidity in camp composition. Neither Hiatt nor Williams report it either, though Balikci, Damas, Helm, Lee, Marshall, Slobodin, Suttles, Turnbull, and Woodburn have reported a situation of fairly continuous realignments of individuals or families. In fact, a number of the latter indicate that fluidity of alignments is so great that the entities formed do not meet the usual definition ot band or group. Balikci, Laughlin, Sharp, Turnbull, and Woodburn have emphasized the relative importance of various patterned reciprocal dyadic relationships over groups integrated by normative or functional ties. This is more than mnely semantics. The use of the term “group” or “band” for local or task units among some of the cases may even lead to a faulty conception of observed social interactions. The terms “aggregate, “ “unit, “ or “network” would seem to be more appropriate in describing certain local or activity-oriented units. On reexamination, some units identified as “bands” could turn out to be what Dr. Sharp has characterized as overlapping “egocentered sets of societies. “ And even where groups are established, reciprocal dyadic ties which frequently reinforce or provide additional bases of alliance are worthy of fuller investigation (cf. Chapter 17h).

Three other corrective findings are provided from the papers and discussions. First, exclusive territoriality with territorial defense by a fixed group of people is rare at best. However, association between a named land area and a fairly stable social unit is common, whether the centripetal factor is a totemic spot, a resource place, a water hole, or a “home base,” i.e., a relatively limited range within a wider area to which the activities of aged, young, sick and reproducing persons are restricted. Second, in a number of hunting and gathering peoples it is difficult to elicit norms of residence, descent, or marriage, the commonly significant rules in group structure. Third, there appears to be more specialization for certain tasks among hunters than has been stressed up to now.

With these findings in mind, I wish to return to the question of why fluidity in group composition may be such a common feature for hunters and gatherers. It has already been mentioned that chance demographic factors may impose difficulties on those peoples having well-defined norms of group structure to follow. Among hunters whose rules are less precisely stated, certain patterned but less formal practices confer a degree of constancy in group composition. With the Dogrib, the!Kung, and the Tiwi, sibling and affinal bonds are of primary importance in affiliating families. With the Hadza, sibling ties are of less importance than those uniting a mother and her married offspring (especially mother-daughter) and those between spouses. Age ties and sibling ties are factors among the Mbuti Pygmies. Among other hunting peoples, spouse exchange, adoption, ritual obligations, and friendship provide some continuity. I find especially intriguing Dr. Woodburn’s and Dr. Turnbull’s remarks on the one hand about the lack of commitment or loyalty of hunter to one another, and on the other hand that interpersonal bonds are purely expedient. If such a lack of commitment to others and a’ view of relationships as expedient has wider currency among hunters, then the importance of dyadic ties and of fluidity of group composition are more easily understood. I think that we can reject Dr. Woodburn’s lead that lack of commitment is related to lack of dependence on others for access to things. The fact that the majority of Hadza men are “hangers on” as regards kills of meat belies a lack of independence, especially since size of meat portion distributed is a factor in budding off. Maintenance of a broad series of reciprocally validated exchange bonds with kinsmen, affines, friends, partners, etc., certainly provides fullest security in circumstances which are apt to be rather constantly changing. Where alternatives for access to wants change not only with fluctuations in resources but also with fluctuations in demography of the social unit, it is difficult to plan ahead. But what better security is there for the future than to keep a wide range of strategic social ties “warm”?

The emphasis on informal and dyadic alignments and the fluidity of many hunter and gatherer peoples can thus be seen to have some advantageous consequences. The presence of a range of options by which individuals and families pursue their best interests at any time may be viewed as a workable, if imperfect, cybernetic system. Shifting arrangements based upon best options within the guidelines of social rules or preferences (as far as is possible) may most effectively relate a hunting population to short-term fluctuations in resources and conditions of life, as Lee, Alland, and others suggeM. Social flexibility allows for optimal exploitation of the preferred effective resources by the total population, including ill and aged as well as the able. And, as a number of conferees have suggested, fission may be the principal means of regulating interpersonal conflict. On the other hand, as Dr. Balikci suggests, reciprocal dyadic relationships—which crosscut social bonds established on kinship or family ties-are peacemaking devices. Clearly more data on these relationships are needed.

Finally, that the above is possible does not militate against the suggestion by Sahlins and others (e.g., Rappaport, 1 966) that ideology and rules may have adaptive advantages over longer term fluctuations.

17d. Social Determinants of Group Size

Speakers: L. Binford, Helm, Williams, Woodburn

BINFORD: I would like to ask Bob Williams to elaborate on his comment that there may be size limitations to various organizational features of society. For example, what are the maximum sizes of groups that may cooperate without having affinal ties or other articulations which arise not out of face-to-face interaction but ‘out of some symbolic partitioning of society? I t seems that many of the people who have presented material on composite bands have mentioned the size range of 1 5–25 people. What factors would operate to limit local group size? What factors seem to lead to an optimum size of 15–2.5?

Williams: In my own group, it is difficult to tell what is cause and what is effect. For instance, the marriage requirements are such that a man cannot arrange to bring a bride for his son without help. He cannot get the needed help unless he is co-resident with the people who are to help him. And I have seen people move back in with others of their patriline in order to get this kind of help. This would suggest that such a type of cooperation is necessary, but one can see the causality in two possible ways. This might be a device that ensures a certain group size, or—taking a functionalist point of view—one might say that the bride’s family is ensuring that there is a kinship group of adequate size to take care of their daughters. This, like many other things, regulates the size of the band, but it is always hard to distinguish cause and effect. But the very presence of such non-economic factors- as mentioned also by Turnbull—assures us that group size is not determined merely by ecology and economy.

Woodburn: I argue that the environment and technology of the Hadza set only very broad limits to the possible sizes of camps. In practice, one finds camps containing from one to a hundred people; the average camp contains about eighteen adults. One factor that seems to set a rough upper limit on the number of people in a camp is the size, and not the number, of animals killed. By custom the meat of an animal is divided widely among the members of a camp. Hadza have a fairly definite notion of what constitutes a reasonable individual portion of meat and if they do not get this amount, people may begin to move out. The good hunters are especially likely to leave.

Helm: Is there any way in which a good hunter can control a large group of strangers following him?

Woodburn: None at all. It is striking that among the Hadza most of the hunting is done by a small proportion of the men. As I have mentioned earlier, perhaps as many as half of the adult Hadza men kill fewer than one large animal per head per year. Many men eat more meat than they contribute and the good hunters just put up with this situation.

17e. Resolving Conflicts by Fission

Speakers: Hamburg, Hiatt, Turnbull, Woodburn

HAMBURG: Dr. Turnbull has put considerable emphasis on the utilization of fluid band composition for the resolution or avoidance of conflict. Can you say something more about the sources of such conflict?

TURNBULL: Among the Mbuti, conflicts usually involve anxiety over marital problems. Perhaps due to the composite nature of the band, there is a good deal of sister exchange, and this leads to obvious conflicts of loyalty between brothers or between a man and his wife. Occasionally, there is conflict over food but very little over property—some- times a minor theft, perhaps. The sources of contlict seem rather trivial and sometimes are even invented. I think they represent simply the great internal tensions that build up during a period of hunting under very close conditions.

HIATT: I wonder whether Dr. Turnbull could identify the cohesive factors—both of ideology and self-interest.

TURNBULL: Among the Mbuti-where practically all the tendencies seem to be towards fission—the one really strong cohesive factor is simply the forest itself. This is where the religious aspect comes in. The forest is the only constant in their lives; there are no allegiances among themselves as individuals or as families to which they can look for a feeling of constancy. As I said before, a band can only be defined as that group of individuals who are living and hunting in a certain territory at a certain time. The composition of the band is always changing, but the band as an ideal continues to exist within a given territory. So there is a feeling of loyalty towards the territory. In times of fission, the Mbuti are thrown upon themselves and upon this relationship to the forest. Their activities become less and less communal until finally, at the height of the honey season, individual men, women, and even children go out alone and live off the forest. There is a clear idea of the enormous generosity of the forest, and the structural importance of this recognition is seen in the fact that the bounty and benevolence of the forest form the major topics of discussion in times of crisis, just as they form the baSIS of the great religious songs, which more than anything bind Mbuti society together.

Although I am not yet sure, I strongly suspect that the Ik have a similar feeling about their environment. In their own lives there is no room for affection as we understand it; couples and families break up, and women abandon children to decrepit grandparents— apparently without remorse. They are a rather cold and calculating people, and relationships are seen in terms of expediency. But one does sense a ‘ feeling of affection and of a unifying bond when one sees them sitting at the band’s sitting place—a rocky outcrop, chosen for its view. In times of anxiety, they will sit there and gaze across the mountains and valleys and seem to feel affection for them. Like the Mbuti, the Ik also recognize that the environment provides them the necessities of life.

Woodburn: The Hadza make use of environmental explanations to justify nomadic movements which ease contlict. When people do move or leave a camp because of conflict, they very often give an ecological reason for their movements. They suddenly discover that the berries somewhere else are better. By giving such a neutral ecological explanation, they solve disputes simply by refusing to acknowledge them.

17f. Territorial Boundaries

Speakers: Hiatt, R. Lee, Pilling

Pilling: Regarding territoriality, to say that boundaries are totally absent among hunters is an oversimplification. Let me draw examples from my knowledge of the Tiwi of North Australia. Certain sectors of the boundaries of territories belonging to some of their local groups are fairly exact; however, this is true primarily where the boundaries coincide with wide bodies of water, and even in these instances—I’ should note—I never heard of boundaries having a fixed location on the surface (or bottom) of such a body of water.

To understand the location of boundaries on land, it should be noted that among the Tiwi the countryside is covered by a large number of small named localities with only vaguely

defined limits, somewhat in the way that a farmer in America or England has names for localities on the land he owns or works. In most instances, these Tiwi-named localities were within the territory of one local group. However, a few were neutral, or buffer, localities between the territories of different local groups; in such areas members of either local group could camp, even when members of the other local group were also camped there.

On one occasion, I was traveling with some men near the center of l\1elville Island, and one said: “This is where (so-and-so named locality) ends, and this is where (so-and-so other named locality) begins. “ I already knew that one of these localities was a netural area, and the other was considered to be in the territory of a specific local group. Therefore, I was hearing a statement that in this one instance the land boundary was a sharp boundary. When I asked my Tiwi traveling companions about the sharpness of this boundary, they showed their impatience at my ignorance; for the name of the neutral area was ecologically specific, meaning “the area where (so-and-so tree) grows, “ and that grove happened to have a sharp edge. Therefore, the application of the specific local geographical term meant that there was a sharp boundary. Thus, there are few instances among the Tiwi where there is a very precise edge to the territory of a local group.

But it should be stressed that in Tiwi thinking these boundaries seem to be a by-product of application of an ecologically related word, and not the product of the presence of a concept of precise, fence or line-like boundaries.

HIATT: I found something similar to this. The typical situation, I think, is one in which there is a cluster of named totemic or non- totemic sites that form what Stanner (1965) has called a “heartland. “ Between one heartland and the next, there is usually no sharp boundary; there is an area between the two about which no one cares much. But I did find one or two cases of the sort of thing that Dr. Pilling was talking about. One totemic site was a sand ridge which was said to have been formed by a totemic sea monster. The people in talking of this site could say quite explicitly, “This is the end of Yiridja and the beginning of Dua.” But this situation is very uncommon.

LEE: Culturally defined boundaries do not necessarily imply sanctions against trespass. You can achieve the same ecological and social effect by having rules for accommodating people across boundaries as when there are no boundaries at all. All of the hunting peoples we have been discussing have the institutionalized means for moving from group to group. So if we find boundaries in a given case, we should not commit the frequent error of assuming that they enclose a defended and exclusive territory.

17g. Predation and Warfare

Speakers: Chagnon, Hiatt, R. Lee, Owen, Pilling, Woodburn

OWEN: One of the factors in the reduction of the western North American hunters and gatherers has been the attrition on them from surrounding agriculturalists. No one has mentioned yet that hunters and gatherers sometimes hunt domestic animals, which often results in the domestic people hunting the wild ones, so to speak. I wonder whether anyone else has observations to make about population control exerted by neighbors.

LEE: The 425!Kung Bushmen of the Dobe area share their range amicably with 340 Herero and Tswana pastoralists. This situation has not developed into the “patron-client” relationship of the Mbuti, but neither is it one of constant warfare. They have reached a modus vivendi over a period of forty years, and cattle theft by Bushmen occurs now on an average of only once in several years. The last case of a Bushman murdering a Mutswana occurred in 1946, and this dispute involved women, not livestock.

WOODBURN: The Hadza do not often come into conflict with their neighbors. There have been a few violent encounters with the pastoral Tatoga, who gain the esteem of their fellow tribesmen by killing members of other tribes. As many Tatoga as Hadza seem to have died in these encounters, but they are not an important cause of mortality for either tribe.

The Hadza talk of a time before the arrival of Europeans when they had to conceal their camps more than they do today and when they built their fires in holes in the ground to avoid detection at night. They say they had many enemies then. But it is now impossible to judge whether many people were killed by enemies. In recent years there have been one or two cases of Hadza hunting domesticated animals and in one of these a Hadza was murdered.

CHAGNON: So far, we have concerned ourselves with settlement patterns and distribution of peoples largely in terms of economics. Now it seems that we have drifted into the context of conflict and warfare of hunters in a world of non-hunters, as Sahlins puts it. I would like to hear from those who have studied the Australian aborigines about the effects of warfare between bands on the distribution of the groups. By “warfare” I mean any hostile relationship between autonomous groups that may alter their seasonal pattern. For example, while the distribution of food might make it advisable for them to split up during the year, perhaps fear of night attack or of witchcraft from a neighboring group might preclude their breaking up into indefensible collecting groups.

PILLING: On the question of warlitre and feud, I would like to draw on material from the very vivid memory culture of Tiwi informants who had lived this way of life in their early adulthood. The night raids were effectively terminated, about 1912, when Sir Baldwin Spencer was inadvertently injured by a Tiwi during a spear-throwing demonstration. It was apparently made very clear at that time that the continuation of spear-throwing and the use of guns (which had come into use for night raidin.g) would no longer be tolerated by the Europeans. This Spencer incident was correlated with the end of night raiding and sneak attacks and it appeared to have stopped pitch battles producing death. But, in fact, as late as 1948 death-causing battles with clubs were still occurring. My chief informant, the major leader of one faction, survived this 1948 battle only by having nearly all the doctors in the Darwin Hospital in attendance upon him for many hours, performing brain surgery. (It appears that the administrators were fearful that the Tiwi, armed with guns, might cause a major stir, if this leader died. National and international news releases of such a story might reflect unfavorably upon the administrators.) Club battles lasted, although without death-causing injuries, until at least 1953, even in the face of the threat of strong sanctions against them hom Europeans.

Under the old pattern, sneak attack was sufficiently common that informants spoke of special ecological adjustments to it. When local group A managed in a raid to kill a man in opposing local group B, retaliation was to be expected. Normally, a female intermediaryliving in local group B, but born in the territory of local group A-came to A to warn them of future planned retaliation by B. At this, the threatened group A was likely to move to the mangroves, a very specialized and unpleasant ecological niche with, among other things, crocodiles and a sloshy mud floor. Some of my older informants, who had lived under these conditions, considered it a blessing when the pattern of sneak attack was terminated in 1912. Apparently, camps in the mangroves were never larger than that of a married couple with children, and they might have to live in this manner for as long as several months. The hiding in mangroves ended when an adult male of the group hiding was apprehended and killed by the enemy, or when an exchange of messages indicated another means than sneak attack would be used to even the score. These alternative means of evening the score included pitched spear-battles or a planned context for counter-injury, but in either of these cases deaths would have been considered improper.

It is important to note the incidence of fatalities associated with the old pattern of sneak attacks and the way of life with which it was correlated. In one decade (1893–1903), at least sixteen males in the 25-to-45 age group were killed in feuding; either during sneak attacks or in arranged pitch battles. Those killed represented over 10 per cent of all males in that age category, which was the age group of the young fathers.

17h. Hunter Social Organization: Some Problems of Method

Speaker: Sharp

Sharp: Let me continue to try to sum up some of the issues that have been raised here.

One thing that strikes and humbles me is the smallness of our sample, both in the number of peoples studied by those present here and, even more generally, as shown by Profe^;sor Mur- docFs paper. And for many of the tribes that we do know about, we still have only very fragmentary data, which points up the urgency of ethnographic research to increase the sample upon which we can draw in considering such a problem as the nature of the social unit.

A second major defect in our data is the lack of situational material, material in context. This lack raised a number of problems earlier, for example, the problems of ethnographic reconstruction. Should we try to limit outselves to pure forms, and reject material that comes to us only in complex and accultur- ative situations? These questions arise in any ethnographic research and have to be dealt with as they arise. For the future, I am fairly sanguine about the help that we can get from our archeological, ethnohistorical, and historical colleagues ifwe leave the channels open for .collaboration and communication—which, of course, is one of the prime reasons for this symposium.

The Importance of Longer Field Studies

To overcome the evils of the old hit-and-run ethnography of which many of us have been guilty, we need longer-term field studies and follow-up studies to get descriptions of process. This is necessary also to give us longer runs for statistical handling. For example, I have never discovered whether there is any validity to the fascinating statistic that I drew from two censuses, three years apart, of the Yir Yoront on Cape York, which showed a birth ratio of 125 males to every 100 females (a ratio that was corrected in later age as males died from snakes, sharks, fighting, etc.). Do the anxieties of hunting life somehow affect the chemistry of conception, or is it simply the accidental result of a two-year census, as opposed to a longer one? The dangers of extrapolating from our few data are great.

Longer-term studies will tell us something more about the matter of devolution. Societies that seem to be going downhill have all the greater claim to be studied. Dr. Gardner mentioned the Phi Tong Luang in northern Thailand (Bernatzik, 1938 [1958]), who are undoubtedly hunters. But, apparently, they are the survivors of a domestic slave group freed by a prince of Nan and sent into the bush. They had no technical knowledge beyond sweeping floors and arranging flowers, and, presumably, fell back to a hunting subsistence level. They are being studied, but only inadequately; yet we know that they were reported very early, long before Bernatzik first published on them. Aside from the matter of their presumed origin, we shjuld be interested by the perseverance of this group in the midst of the rice culture and court civilization of Siam.

Definition of Social Units

The question of units still needs a good bit of work, both in the study and in the field. Family and kin structures appear to be manageable; we can build band units, hunting groups, etc., either in terms of an aggregation of family units or as segments of family groups. But we still have trouble with the suprafamily or the reconstituted subcommunity units. Again, this problem is not limited in ethnographic studies to hunters alone; Leach (1943, 196 I), Lehman (1963), and others have worried about it with regard to tribal studies. It is even a problem— at least, in Southeast Asia—in regard to states; we Europeans come in with a conception of states with fixed boundaries and find that in precontact times there was a much more flexible and overlapping system in effect. The whole matter is still biased by the imposition of prefabricated constructs—the patriclan, the horde, and the band—which we go looking for.

The construct of the society has given me trouble, and I do not know whether we can talk about this in all cases. In studying the aboriginal population on Cape York Peninsula, I simply could not find a society. I would have to describe it in terms of an ego-centered set of societies; no one individual was the center of a system of networks that overlapped iso- typically with anyone else’s. Not only were there networks and chains, but these were also institutionally divided. A man might participate in a trade network or chain that went southward; whereas his ritual networks might go northward, and with no overlap between the members of the two groups. The same would be true for symbolic and even for some technological behaviors. All of these, incidentally, were phrased in terms of kinship, which gave the basis for moving outward from the series of individuals with which one started.

The Application of Role Theory

These problems all point to deficiencies in our descriptive methods and the resulting content of our descriptions; some participants have called for development of sociometric techniques. Given my bias toward roles as useful descriptive units, and since we have to look at the interaction of something, I suggest that a further elaboration of role constructs would allow us to go into any society—whether highly structured, like the Australian societies, or less structured, like Eskimo groups-and, by an inductive process, build up the networks of roles. We should not only include masculine roles, feminine roles, children’s roles, totemic leaders’ roles, etc.; but we should also see roles in terms of behavioral activity, including ideational and emotional aspects. This might help us get really comparative information.

The roles would have to be seen in spatial and temporal, as well as in social contexts. There are diurnal and seasonal patterns. When a woman is out with her children and with other women digging yams during the daytime, is she a wife, or is the role of wife latent then? What are the possibilities for the manifestation and latency of roles, especially of children engaged in anticipatory learning? I think that the role approach forces us to look at contexts-the setting, duration, and frequency of behaviors. We would find that a certain woman is the head of a matrifocal family eight or twelve hours a day and head of a nuclear family during the rest of the time. Most of the ethnographic accounts do not give us that kind of information. Ifwe break up the 24-hour day and 365-day year in terms of roles and networks of roles, we will come up with more realistic and exact descriptions for comparative use. This would allow us to tackle a number of the problems that have merely been touched upon at the symposium—problems of definition, correlations, criteria of nomadism (“nomadic style”), territoriality, descent, and devolution.

The Study of Ritual

It seems to me that the main gap in the material presented so far-and I suppose that this is true of a great deal of the material being collected today-occurs in the area of symbolic or ritual behavior. I could reemphasize the point made by Dr. Helm that we must deal with socio-territorial groups. This is partly a matter of having enough personnel in the field to be in about eight places at once and enough time to remain through several seasonal cycles. We have to get at the socio- symbolic or ritual aspects of behavior, the idea categories that are imposed on the land and the people, thus linking specific people with specific bits of land-as in the totemic complexes that Dr. Hiatt mentioned. So far as I know, the Eskimos do not have land ideologies and rituals of this type, but one wonders if this may be because those who have described them had too narrow a conception of symbolic and ritual behavior. The character of the idea system is, of course, of vital importance in the survival of these groups. Given a universal rigid kind of system, as in Australia, a change in one part of the system is likely to bring about change elsewhere, notwithstanding Dr. Drucker’s recent statement about the Haida (1965). On the other hand, one gets the impression that the Alaskan groups, and band groups in general, have a much more elastic ideological system. And, as Dr. Deetz pointed out, the ideological elements are manageable and provide a major means of adjustment if they are of importance to the social behavior of the group. Myth can be adjusted to reality, and to a certain extent reality can be adjusted to myth.

These considerations relate to a whole series of more specific problems. I do not see, for example, how one can claim to have found corporate structures without the names or symbols of corporateness that provide for stability and maintenance. (Only “he who thinks that he’ll be the same tomorrow does not need a name.”) Names furnish ideas about land ownership and boundaries. Apropos of this last, I do not quite understand how Dr. Hiatt’s groups can have named plots of land and yet not have boundaries, however wide or vague. It is difficult to say when corporate ness is needed by a group. What conditions would reqUire a group to take on corporate characteristics, and what are the ideological correlates of such corporate structures?

17i. Typology and Reconstruction: A Shoshoni Example

Speakers: Eggan, Hiatt

EGGAN: I have several comments, beginning with some brief remarks, on typology. Julian Steward was the first to bring some typological order out of the chaos of band organization (1936, 1955). You will recall that he set up three types, partly empirical and partly ideal: the patrilineal band, the composite band, and the matrilineal band. And he attempted to relate them to one another and to cultural ecology. Elman Service more recently remodeled Steward’s band typology (1962); he keeps Steward’s two main types but sees the composite as an acculturative product of the patrilocal or patrilineal band, and he adds an “anomalous band” to take care of the Shoshoni and Eskimo.

I think it is clear that our sessions have demonstrated that Service is going in the wrong direction. Bands show a much greater variety than either Steward or Service allow. If social organization is adaptive, then this variety is exactly what we would expect. We have heard about a great variety of bands and will probably hear of more in the future. However, I think there will prove to be a limited number of viable combinations; although this is still to be determined.

In contradiction to Service, I think that patrilineal bands are a specialization on the more flexible, amorphous base than most of us have been describing. That is, I do not think that it is some sort of primary unit that has degenerated in late times.

Another aspect of this problem has been raised a number of times for North America and other areas. There is seasonal alternation between regional or macrobands and local or microbands, and we need to sort these out systematically. This offers us an opportunity to do better comparative work than we normally have done, since, after all, one has the same people under different ecological or seasonal differences rearranging themselves in different groupings. We should have some closer control on such situations, so that we can specify more clearly the factors operative in them.

I would like to make some remarks on the Shoshoni of the Great Basin, since they have been mentioned by several people and might cast a little more light on what has happened in Australia. I might say that the Shoshoni, in relation to their neighbors to the north and south, are in a peaceful situation like that described by Dr. Turnbull. This is true both within linguistic groups and, for the most part, across the linguistic boundaries, as well. The Shoshoni and their linguistic neighbors, the Ute and the Northern and Southern Paiute, throw some light on the development of band organization under extreme ecological conditions. This is an area of long occupation by what archeologists call the Desert Culture, a combination of hunting of small game on the part of men and gathering by women. This has gone on since at least 8,000 b.c., perhaps even earlier. There has been a nearly continuous cultural record; even though there is evidence of movements into and out of the area.

The Great Basin area itself is quite different from most other culture areas. It is dish-shaped with a culture of bare essentials in the center and a series of specializations at the margins, including incipient agriculture or use of the horse, according to the opportunities offered by the local habitat. Basic social structure is essentially a loose network of ties based on kinship and marriage. It is completely bilateral and runs across linguistic boundaries, and it functions both to supply spouses and to spread information about the sporadic appearance of new food supplies.

The winter group was made up of two to five families or ten to twenty people, who were almost invariably kin. There might be a different kin composition from one year to the next, but the size of the group was fairly stable. The basic ties were sibling bonds, parent-child bonds, and marriage ties. Residence throughout the area was generally matrilocal, at least initially. Brother-sister exchange was universal and was accompanied by polygyny (usually sororal) and fraternal polyandry (sometimes temporary, sometimes institutionalized). Brother-sister exchange gave double bonds or “knots” at certain places in the network.

In a few central regions along the Humboldt River, one found a more stable food supply, which made it possible for groups of two to five families to remain in the same place year after year. In summer, they made a circular trip of 25 to 150 miles, each group stopping at a series of places known to them. In these localities it was possible for parents and children to remain together for more than one generation, and in the second generation the normal process of brother-sister exchange made for the institutionalization of cross-cousin marriage. This trend was strong enough to affect the kinship system, resulting in a Dakota pattern for the terminology and a joking and permissive relation between cross-cousins. This marriage was thought to be of value, since children reared there learned the area thoroughly.

In areas to the north and east, there were areas of rather more stable food supplies. Salmon came up the Snake River, and the introduction of the horse made possible larger groupings. Interestingly enough, cross-cousin marriage here gave way to what Steward calls “pseudo-cross-cousin” marriage, where one marries a step-cross-cousin, a sociological cross-cousin, rather than a real one. This makes possible integration on a broader scale and allows for, say, the assimiliation of a mother with a daughter coming into a band. This dates back before 1 7°O, as shown by the fact that the Comanche had pseudo-cross-cousin marriage too.

Now, around the margins of this area there is another set of specializations. In the Owens Valley, where seed resources and other resources were more abundant, the density was increased from one person per twenty square miles to one person per two square miles. And due to the matrilocal residence of the women concerned with seed-gathering, there was a development of matrilocal bands which—as Steward says somewhere-are almost, but not quite matrilineal clans, as he defined them.

Toward the south, the Southern Paiute, with incipient agriculture, also had relatively large bands. And I predict that when we get the archeological sequences worked out for protoHopi and the Tusayan Area, we will find that in Basketmaker I there is a transition from the Basin-type Desert culture to incipient matrilocal bands and, probably, matrilineal clans in association with agriculture.

I happen to have grown up, anthropologically, on Australia. And I am somewhat surprised to find that many of the same straw men are being demolished, and in some cases by slightly dubious statistics. I am also surprised at the cavalier dismissal of the work of earlier scholars. Given the state of knowledge of Australian social organization in 1930, it seems to me that Radcliffe-Brown’s contribution was a very considerable one. It was too narrow and did not take into account the systems of Arnhem Land, nor did it treat technology and religion. But it seems to me a base upon which to build and elaborate, rather than something to be destroyed.

Hiatt: To reply to Dr. Eggan’s comment about our knocking down Radcliffe-Brown’s straw men, which is very nearly what Stanner says, I want to say for the record that the horde and the subsections are, at the very least, very sturdy straw men. I make no quarrel with the fact that Radcliffe-Brown was a worthy pioneer in aboriginal studies, but this does not mean that one should not take issue with him on particular points.

EGGAN: Just for the record, I was referring to the last paragraph in Dr. Meggitt’s paper.

Part IV: Marriage and Models in Australia

18. Gidjingali Marriage Arrangements

L. R. Hiatt

Levi-Strauss, whose distinguished contribution to anthropology we honor at this conference, has paid a good deal of attention to the Australian aborigines. In this paper I shall discuss some of his views by relating them to the Gidjingali tribe of northern Arnhem Land.[54]

“Gidjingali” is the term for the language spoken by people who, until recently, lived south of Cape Stewart and around the mouth of the Blyth River. When I carried out my field work between I 958 and I 960, most of them were living about thirty miles from their homelands at Maningrida, a government settlement established in 1957 at the mouth of the Liverpool River. Elsewhere in this volume I have described some aspects of their traditional local organization.[55]

People of northeastern Arnhem Land referred to the Gidjingali as the “Burera,” whom Warner (1937, pp. is, 36) included as one of the eight tribes he designated collectively as the “Murngin.” Although he admitted that they had a very different language from the rest and were affiliated linguistically with tribes further to the west (p. 37), he apparently. believed that they shared with other of his “Murngin” tribes a kinship system based on matrilateral cross-cousin marriage. This was not so. The Gidjingali had an Aranda-type kinship system which was based on marriage between certain kinds of second cousins (e.g., MFZDS-MMBDD).[56] A total of four categories of male relatives were distinguished in Ego’s grandparental generation.[57]

Levi-Strauss (1949, 1963 [1958]) attaches fundamental importance to mechanisms of reciprocity. He believes that marriage regulations and kinship systems can best be understood by regarding them as “a kind of language, a set of processes permitting the establishment, between individuals and groups, of a certain type of communication” (1963, p. 6 I). The mediating factors in this case are not words but women, who are circulated between clans, lineages, or families.

Gidjingali kinship terminology and marriage rules can be adequately described within the limits of a hypothetical system of exchangemarriage among four patriclans. The model bears little relation to marriage arrangements in contemporary practice, and so I shall use the method of conjectural history in presenting it. I shall leave till later the question whether it has any explanatory, as distinct from expository, value and, in particular, whether there is any evidence for supposing it to represent a communication system that uses women as units.

The Model

Suppose that on some past occasion on aboriginal law-maker named Plato[58] proclaimed the following constitution:

Section 1: Kinship groups and terminology
  1. Descendants in the male line of four men in Plato’s grandparental generation-his FF, MF, FMB, and MMB-henceforth belong to four separate clans, named p, g, r, and s (Fig. I).

  2. A person should apply a different kinship term to his FF, MF, FMB, and MMB; and to their respective sisters, sons, and daughters (Fig. 2).

  3. Everyone should abide by the principle of the equivalence of siblings of the same sex. For example, a man should apply the same term to both his F and FB; another term to both his M and MZ.

  4. Everyone should abide by the principle of the combination of alternate generations. For example, within his own clan a man should apply the terms “FF” or “FFZ” to all members of his FF’s genealogical generation and generations alternating with it; and the terms “F” or “FZ” to all members of other generations. The same principle should be applied to the other three clans, giving a total of sixteen terms as indicated in Figure 2.

Section 2: Marriage rules
  1. Men should not marry within their own clans.

  2. The right to dispose of a girl in marriage is vested equally in her M and MB. She may be bestowed an infant but should not live with her husband till puberty.

  3. Men should abide by the principle of niece-exchange. That is, when a man receives a woman in marriage, he should subsequently bestow his own ZD on his WMB.

  4. Sons should follow in the same tradition of exchange as their fathers. Thus, as Plato’s MMB gave a niece to Plato’s F (Fig. I), so Plato’s MMBS should give a niece to Plato; and Plato’s F and Plato should in return give nieces to their respective WMBs. Figure 3 illustrates the results of such exchanges. Arthur and Jack have exchanged their nieces, Jill and Ada. Jack has married his MMBDD, Arthur his FZDDD. Arthur’s DD, Laura, is related to Jack’s son, Simon, as both MMBDD and FZDD. By extending the diagram further it can be shown that all niece-exchanges between the patrilineal descendants of Arthur and Jack ad infinitum result in marriages to women related to their husbands as both MMBDD and FZDDD.

  5. The rule that sons should follow in the same tradition of niece-exchange as their fathers should be interpreted broadly in terms of clan affiliation, not solely in terms of actual F-S connections. Thus, clans p and s should exchange nieces; and, likewise, clans q and r should exchange nieces.

  6. It follows from the two previous rules that a man has rights to the daughters not only of his true MMBD but also of all the other women in her patrician classified as her sister; conversely, he has rights to the daughters not only of his true FZDD but also of all the other women whose MMs belong to his own clan.

  7. The oldest of a group of men sharing rights to women as defined in the previous clause has priority, the second oldest second priority, and so on. Polygyny is permitted, but an older man should not take additional wives by insisting on his prior right if to do so means subjecting a younger unmarried man to unreasonable hardship.

h) No one should marry a person who is not a member of p, g, r, or s. If all the principles set out in this and the previous section of the constitution are obeyed, a dosed system of niece-exchange will be set in motion along the lines indicated in Figure 4.

r-b-richard-b-lee-irven-devore-man-the-hunter-17.jpg
Figure 1.
r-b-richard-b-lee-irven-devore-man-the-hunter-16.jpg
Figure 2.
r-b-richard-b-lee-irven-devore-man-the-hunter-18.jpg
Figure 3.

The Facts

If Gidjingali marriages once conformed to the principles of this system, they were no longer doing so when I arrived in 1 958. The basic kinship terms and the definition of marriage rights were as described. But there was no rule stating that marriages should occur within cIosed systems of four patricIans. In fact they did not. More important, only a small percentage of men managed to marry women to whom they had rights. In this section I shall try to explain why.

In 1960 the Gidjingali numbered about 300. They were divided into four communites and 28 patricIans (mean 10.6, range 2 to 29). Marriages occurred both within and between communities, and, indeed, between Gidjingali and members of neighboring tribes. A person applied one of the four sets of kinship terms (Fig. 2) to members of every patricIan he encountered.

I shall limit my analysis to the Anbara community, which contained a total of 48 adult males and 47 adult females (by “adult” I mean over the age of about 14). I shall use the expressions MMBDD and FZnDD to mean the women to whom Ego had marriage rights, not just his true MMBDD and FZDDD (see Fig. 5)-

At the time of observation 25 men were without wives (56 per cent were under 25 years of age). The remaining 23 men (practically all of them aged over 25) were married to a total of 35 wives, only six of whom were related to their husbands as MMBDD or FZDDD. Two questions arise: (I) why were there so few unions consistent with the exercise of marriage rights (only 17 per cent)? and (b) how did the other unions come about? The answers are relevant to a general question raised by Firth (1954, p. 12) when he asked what happened in Australian systems when individuals lacked relatives of the kind they were expected to marry.

r-b-richard-b-lee-irven-devore-man-the-hunter-19.jpg
Figure 4.
r-b-richard-b-lee-irven-devore-man-the-hunter-20.jpg
Figure 5.

Examination of genealogical and age data disclosed that, when 18 of the 35 women married, adult males with rights to them were nonexistent. That is, in order to marry upon reaching puberty, these 18 women had to wed men without rights to them. A similar analysis showed that 19 of the 48 Anbara men did not have a MMBDD or fZdDD of marriageable age. The remaining 29 had a total of 27 mature relatives of these kinds, but the distribution of relationships was such that only 2 I male-female pairs could be formed. In short, at the time of observation demographic factors alone limited the number of men married to a MMBDD or FZDDD to a possible 44 per cent.

Now in cases where there was no one with a right to a particular female child, it was accepted that her M and MB could without delay choose a man with no right to her. The only formal requirement was that he should belong to the same terminological category as would a man with a right (that is, the girl should be his classiScatory MMBDD); or, alternatively, he should be an actual or classificatory cross-cousin of the girl. The advantage to bestowers in making an early promise to a mature male was that they could from that moment expect him regularly to send them presents, support them in quarrels, and eventually give a niece to his WMB.

Sometimes men with no rights to any women acquired as wives females to whom no other men had rights. Sometimes men with rights waived them in a display of generosity in favor of men without rights. But sometimes interests clashed and fights broke out. It is impossible in this short paper to analyze these conflicts (which occurred more frequently than any other type), or even to outline the opportunism with which bestowers and seekers of wives tried to advance their own interests. I must simply assert that, in a context in which there was a chronic disproportion between the demand for wives and the supply of female relatives specified by the rule defining marriage rights, the rule was often ignored. These breaches, in combination with demographic factors already described, helped to reduce the percentage of men married to a MMBDD or FZDDD even further. I did not observe anything remotely approaching a closed system of exchange-marriage among four patriclans of the type postulated in Figure 4.[59]

Models, Facts, and Feeling Cheated

It is not clear from Levi-Strauss’ writings whether he would regard these facts as of any significance for his theory on kinship and marriage. He stated in his inaugural lecture at the College de France in 1960: “In admitting the symbolic nature of its object, social anthropology does not thus intend to cut itself off from realia.... One cannot study ... social rules independently of the things which correspond to them” (1966, p. I 15). Although the word “correspond” might be interpreted narrowly to mean only behavior that conforms with the rules, the context suggests that he was referring to the relationship between rules and what actually happens. In the same year, however, he described Maybury-Lewis as a “prisoner of the naturalistic misconceptions which have so long pervaded the British school,” because “when he is presented a structural model which departs from empirical reality, he feels cheated in some devious way” (1960, P. 52). The two statements are not contradictory, but they leave us wondering how much trust Levi-Strauss is seeking. After suggesting that the different aspects of social life may constitute phenomena whose inmost nature is the same as that of language, he raised the question: “How can this hypothesis be verified?” (1963, p. 62). Before being led into the non-empirical wilderness, we might also like to know how it can be falsified. Or if the theory is merely a heuristic model, in which case normal criteria of falsifiability might properly be regarded as inappropriate, we might still reasonably ask how we are to tell whether or not we are being misguided.

It is at least of passing interest to note that Levi-Strauss’ theoretical approach to social phenomena bears some striking resemblances to that of Freud.[60] First, both writers treat social institutions as manifestations in some degree of certain unconscious properties of the human mind. Levi-Strauss has written: “The road will then be open for a comparative structural analysis of customs, institutions, and accepted patterns of behavior ... the so-called ‘collective unconscious’ would, in the final analysis, be no more than the expression, on the level of individual thought and behaviour, of certain time and space modalities of the universal laws which make up the unconscious activity of the mind” (1963, p. 65). The difference is that the unconscious processes postulated by Freud are conative and affective, whereas those postulated by Levi-Strauss are cognitive.

Second, in dealing with empirical material both Levi-Strauss and Freud distinguish between manifest and latent content. Levi- Strauss has argued that “a theoretical QYpo- thesis which deviates from the manifest content of the ethnographical data is substantially upheld, if we are able to discover, in the latent content supplied by the myths, the religious representations, etc., some data showing a remarkable parallelism between the native categories, and those arrived at by the way of theoretical reconstruction” (i960, p. 50). To grant such conditions is (as well- known from the psychoanalytic debate) at least to give verification the inside-running; and when special hypothesis-saving devices are introduced as well (such as those used by Ernest Jones with the Trobriand material),[61] falsification is left hobbled at the starting-gate. Levi-Strauss has not, as far as I know, resorted to psychoanalytic devices when informants have failed to give the required answers. Instead, he has postulated an interference with their cognitive processes. In his reply to Maybury-Lewis, he explained a discrepancy between the structural model and ethnographic reality by asserting that the natives had “let themselves be mystified by their own system” (1960, p. 51).

I shall try to anticipate how Levi-Strauss might deal with the assertion that, among contemporary Gidjingali, women did not circulate between groups in systems of marriage-exchanges. One possibility would be to argue that this fact in itself does not disprove the hypothesis that the kinship terminology and marriage rules constituted an apparatus originally ‘set up lor the purposes of communication and used as such. It merely indicates that the apparatus had fallen into disuse. In 1958 60 the lines of communication were still there as an integral part of the kinship system, but, for demographic and other reasons, messages were no longer bcing sent.

The hypothesis would thus be saved by making it refer to an unobserved state of affairs in the past. In that case my presentation of the model in Figure 4 as ail imaginary historical event might have been apt, though whether Levi-Strauss would accept sllch a concrete account is another matter (but see note 5). That he is prepared in principle to usc the method of conjectural history is indicated ckarly by his statement: “One may well ask oneself, then, if Radcliffe-Brown’s mistrust of historical reconstruction did not correspond to a stage of scientific dcvelopmen t which will soon have passed” (1966, p. 115)- This allows us to add yet another similarity between Levi-Strauss and Freud.

Another deknse of the theory might 1)(’ to say that the postulated communication apparatus is a latent or unrealized aspect of the kinship system, created and perpetuated by unconscious mental structures present in all Gidjingali, but prevented from functioning by demographic and other factors. In this view there would be no critical difference between past states and present, and hence no need to postulate a conception once perfectly realized hut now barely acknowledged. To put this another way, the first version of the theory would represent contemporary Gidjingali kinship terminology and marriage rules as “survivals” from a past period when the “true” meaning or purpose of the system was realized in actual behavior (see “The Concept of Archaism in Anthropology, “ Levi-Strauss, 1963, p. I 17). The second version would postulate the “true” meaning or purpose of the system as a chronically unrealized unconscious ideal.

Explanations of the observed in terms of the unobserved are notoriously susceptible to charges of circularity (e.g., “There is a present in my stocking because Santa Claus came last night; Santa Claus exists because there is a present in my stocking”). We could not place much faith in an explanation of Australian kinship systems in which the effect is the only evidence for the occurrence of the alleged cause. The absence of independent evidence has always been the difficulty for conjectural history, and (as far as the evolution of kinship systems is concerned) I fail to see how such evidence might be provided by “stratigraphic excavations, the introduction of statistics into archaeology, the analysis of pollens, and the use or carhon-1 4, and above all the closer and closer collaboration between ethnologists and sociologists, on the one hand and archaeologists and prehistorians, on the other” (Lcvi- Strauss, i <)ti6, p. i 15).

Levi-Strauss has provided more substantial instructions for discovering independent evidence of unconscious causal factors by directing us to myth, ritual, and religious symbolism (i960, pp. 50, 53). I did not detect such material during my investigations, but perhaps I was not looking for it. The only data that appear to support Levi-Strauss’ theory were undisguised.

First, the people had in relation to marriage arrangements a definite notion of reciprocity between individuals: namely, that if X gives a niece to Y, then Y should subsequently give a niece to X (Fig. 3). Second, marriage rights were defined partly in terms of clan membership (Fig. 5). Third, occasionally men spoke of traditional exchange relationships between clans. Do these three facts indicate that at least some Gidjingali consciously harbored an ideal image of marriage arrangements of the sort represented in Figure 4?

A full discussion of the question would require an analysis of the corporate functions of Gidjingali patriclans. Elsewhere I have argued that they were neither local nor political units and that their main functions were ritual (Hiatt, 19(5). But here I shall confine myself to marriage.

The Gidjingali were polygynous, and the ratio of men to women was about equal. This meant that the demand for wives always exceeded the supply. Bachelors worried about their chances of acquiring spouses, married men of losing them. Men and women with girls to dispose in marriage used them to further their own interests. In this context, the system of kinship and marriage to some extent regulated the distribution of scarce resources. Marriage rights stated in terms of specific genealogical links (Fig. 5) located for a man a few women to whose daughters he could make strong claims. Only 17 per cent 6f unions in the sample I discussed under the heading “The Facts” were consistent with the exercise of such rights. But, as well as narrow rules defining marriage rights, there were also broad rules stated in terms of kin categories. These located for a man a much larger number of women (all those classified with his MMBDD and with his FZD) with whom marriage would be morally proper (or “straight,” as it has often been expressed). His difficulty here was that other men had genealogically defined rights to many of these women, and he was accordingly at a jural disadvantage.

Because some confusion has arisen from a failure to distinguish between genealogically defined marriage rights and moral attitudes to marriages stated in terms of kin categories, 1 shall restate the position in a different way. If a man acquired as a wife (e.g., by elopement) a women belonging to one of the following six categories-Z, FZ, M, MM, MMBD, FMBD-the marriage was regarded as immoral, whether or not the husband had in the process infringed the genealogically defined right of another man. If he married a woman belonging to one of the remaining two categories-MMBDD, FZD-the marriage was regarded as morally proper, again whether or not he had infringed the marriage right of another man. About go per cent of Gidjingali marriages were “straight” in this sense. Occasionally, as a result of previous irregularities, a man had a genealogically defined right to a woman who belonged to a category of female relatives with whom marriage was immoral. If he married her, people said that, strictly speaking, the union was improper, but they admitted that the situation was a mIX-Up.

Given the struggle for wives, it is not surprising that bestowers tended to be hardheaded. Sometimes they bestowed nieces in response to a legitimate claim, sometimes in defiance of it; but usually they aimed at placing the recipient under an obligation. In these circumstances niece-exchange appeared as a contractual arrangement between individuals rather -han a symbolic transaction between groups.

I said that marriage rights were defined partly in terms of clan membership. This meant that a man might regard all women of a specified category within a specified patriclan as his potential mothers-in-law and that he might legitimately make a claim on any one of them. I t did not mean that the clan as a whole had a joint responsibility to provide him with a wife, and I never saw clansmen speak or act in these terms. I also said that. occasionally men spoke of traditional exchange relationships between two patricians (though never four, as in the system represented in Fig. 4). Analysis of genealogies failed to bear out such assertions, and the people did not take them seriously. Possibly we have here, emerging from the unconscious, a half-formed ideal of how the system should work. Another possibility (which might put the previous one in a different light) is that such assertions were instances of special pleading voiced during disputes in order to bolster dubious marriage claims. Of this I have evidence.

The discussion would be incomplete without some mention of moieties and subsections (indeed, Levi-Strauss might consider that without them it has not started). The Gidjin- gali were divided into two exogamous patri- moieties, but the people had no notion of marriage-exchange between them. In view of the facts that a man’s ZD belonged to the opposite patrimoiety, and that the system ruled out both daughter-exchange and sisterexchange (to which I shall return later), this is not surprising.

The eight Gidjingali subsections were named divisions corresponding to the eight “brothersister” terminological categories indicated in Figures 2 and 4. They apparently diffused into the area within the last century (Elkin, 1 950); and they have significance mainly in ceremonies of recent adoption. They are neither more nor less significant for marriage regulations than the terminological categories to which they gave objectivity (i.e., in principle it means the same to say “I and everyone I call ‘B’ may marry anyone we call ‘MMBDD’ or ‘FZD’ “ as to say “I am in WAMUD subsection, and WAMUD men may marry either BANGADI or GODjOG women”). The men of a subsection belonged to many different patricians and tribes over a large area of Arnhem Land.

I must point out that informants were able to indicate the give-and-take of women in subsection terms. In order to relate the following example to Figure 4, I shall put the appropriate kin category in’ brackets after the subsection: BALANG (MMBS) gives Bangadidjan (MMBDD) to WAMUD (B), and WAMUD (B) gives Ngaridjan (ZD) to BALANG (MMBS). People demonstrated with blocks of wood representing individual men and women. I should point out that folk regularly used subsection terms to address or refer to particular individuals. They knew that men of one subsection did not have corporate rights of disposal over all the women of another, and they did not think of the subsections as solidary groups exchanging nieces, sisters, or any other female kin in marriage (Meggitt, 1962, Ch. 10; also Meggitt, Chapter 19, this volume).

Conclusion: Communication or Competition?

I have demonstrated the possibility of constructing from Gidjingali kinship terminology and marriage rules a hypothetical model of marriage-exchange that verifies Levi-Strauss’ theory that systems of kinship and marriage are really communication systems with women as the mediators. I argued that, among contemporary Gidjingali, women did not circulate in marriage along lines set out in the model, that the people were not using their kinship system as a communication apparatus of the kind postulated, and that they gave little indication of recognizing I his as a possibility or of holding it as an ideal. The discrepancy between the model and the facts is wide. Here I concur with Firth: “I am in sympathy with the view that the nature of the evidence, the ethnographic fact, is often less important for the development of our subject than the inference, the generalization, the theoretical argument presented. This can be treated as a ‘model,’ an heuristic instrument not meant to depict or mirror social reality. But at a certain point, if the argument is not meant as a purely intellectual exercise, it must apply to some set of demonstrated facts” (1966, p. 2).

I said that, in some degree, the Gidjingali system of kinship and marriage regulated the distribution of scarce resources in women. The question that interests Levi-Strauss is why the distribution should have taken its characteristic Australian form. His faith in reciprocity as the key to the problem may well be justified. But substantial modifications to his theory would have to be made before it would apply even roughly to the Gidjingali. Here two main poi,nts would need to be taken into account. First, the right to give a girl in marriage was vested in her M and MB.[62] Patricians, subsections, and patrimoieties as corporate groups had nothing to do with the decision, and there was no rule or ideal saying that they ought to. A model of marriage-exchange using these groups as units departs seriously from reality. Second, the Gidjingali had a clear notion of reciprocity in marriage. I have referred to this as “niece-exchange,” by which I mean that a bestowal placed the receiver under an obligation to return in kind. Gidjingali men did not have rights to give their sisters or their daughters in marriage.

The second fact is incompatible with Levi- Strauss’ generalization that “in human society a man must obtain a woman from another man who gives him a daughter or a sister” (1962, p. 46). It also raises a question of some significance for all Aranda-type systems (the most widespread kind in Australia). They cannot be accounted for in terms of daughterexchange because a man’s DD belongs to the category of his cross-cousins. An explanation in terms of sister-exchange, on the other hand, is likely to be at variance with the facts of bestowal. Typically, females are promised in marriage as infants by relatives in a senior generation, not by their male siblings. Hypotheses about sister-exchange drawn as inferences from anthropological kinship charts or subsection diagrams are unlikely to have any counterpart in attitudes of the aborigines themselves.

The recent work of Rose on age relationships has an important bearing on this subject.[63] Starting with Levi-Strauss’ view that reciprocity is the keystone of the kinship structure, it. is possible by following Rose’s method to demonstrate a number of interestmg consequences relevant to Australian systems. Aboriginal girls usually married at about 15 years of age, whereas men were lucky to have a wife by the age of 25. The discrepancy was largely a product of polygyny in communities where the sex ratio was about equal. Let us examine the significance of this state of affairs for three possible types of marriage-exchange, regardless of the formal features of the kinship system —(i) sister, (2) daughter, (3) niece.

  1. A reasonable working assumption is that few women would be more than ten years younger than their oldest brother and that few men would be more than ten years older than their youngest sister. It would be uncommon, therefore, for a man married at 25 to have a sister aged 15 and his I 5-year-old wife to have a brother aged at least 25. These would be the conditions necessary for a sister-exchange to take place in accordance with the minimal age difference between a girl married at puberty and her husband. In most cases the sisters of a man married at 25 to a pubertal girl would be either older than the bride’s brothers or, at any rate, not much younger.

  2. If a man married at 25, he could not expect to receive a DD in return for his D until he was at least 57 (25 plus 16 plus 16).

  3. Suppose a man has a twin sister who marries at 15. By the time he is 31, he might have a I 5-year-old ZD. If she marries a man aged 25, also with a twin sister who has married at 15, her husband might have a ZD aged g. If he fulfills his obligation, her father will acquire a second wife by the age of about 37. (I have used the unlikely example of twin sisters to strike a rough mean between older and younger sisters.)

If it is fair to say that Australian marriage systems were characterized both by a demand for plural wives by dominant older men and by individual reciprocity for mutual advantage (and this I believe to be true), then it might be argued that niece-exchange would have satisfied the requirements better than other possible exchanges.[64] This suggestion, based more on man’s animal nature than his spirit, might in the long run provide a better starting-point for an exercise in conjectural history than the view that kinship and marriage are best understood as systems of communication.

19. “Marriage Classes” and Demography in Central Australia

M. J. Meggitt

I n comparative studies of primitive peoples the Australian aborigines have long held a special place and, perhaps for this very reason, a number of misconceptions concerning their society and culture have entered the anthropological literature. This paper, which is essentially negative in tone, aims at attacking one of these propositions.

Anthropologists and others, wntmg in general terms about the aborigines, commonly emphasize one or other of two points. Either they single out aborigines as the prime example of simple hunters and gatherers whose way of life is effeGtively and completely adapted to their environment, or they remark on the complexity of aboriginal social organization.

Thus, from the nineteenth century onward there have been many statements such as the following:

The Australian ... who never thinks of the day to come and does not possess other sources of food than those offered him by the bush in the shape of game and of scanty fruit, develops an eminent skill in obtaining his prey at any moment without the help of others and without any complex contrivance .... This mode of living makes him the most independent of men, but it also stands in the way of all progress, forcing him to roam restlessly over wide areas (Semon, 1899, p. 154).

What is disconcerting is the way in which assertions about the efficiency of the aboriginal hunter early achieved the status of stereotypes; the same few literary references have been cited many times, with no real understanding of the need to take even the broadest ecological factors into account when generalizing about aboriginal custom. In particular, failure to specify environmental, demographic, economic, and technological differences has led to the acceptance of a picture of the Australian aboriginal everywhere as a highly nomadic, desert-dwelling kangaroo-hunter, with all that such exotic features imply in characterizing a culture. For instance Basedow says:

The aboriginal’s ideal of life is attained when he finds himself in hot pursuit of the game which shares with him the wilds of his ancient haunts.... The love of the sport, the keenness of the senses, and the astounding powers of endurance are natural attributes, which the aboriginal alone knows how to use to their fullest. These are the hereditary gifts of man which characterize the primitive hunter (1925, p. 2).

And it is of interest that, in an otherwise excellent description of aboriginal culture, Basedow devotes ‘to an account of men’s hunting more than five times the space that he gives to women’s food-gathering, whereas recent studies indicate that women’s contribution to the daily diet is usually far greater than that of the men (see, for instance, references cited by Meggitt, 1964a, 1 964b) .

Ashley-Montagu goes even further in fixing this stereotype of the Australian as the desert hunter. In Chapter 2 of Coming into Being among the Australian Aborigines (1937), entitled “The Arunta, the type pattern of Australian culture,” he says that, because of the common origin of Australian cultures, “any particular culture may in a general way be taken to represent the pattern of Australian culture, the type of which is represented by the culture of the Arunta.”

Only in the last two decades have anthropologists demonstrated any great sophistication in recognizing and analyzing regional variations in demographic and environmental factors in Australia and their effects on technological, economic, and social practices. It is unfortunate that, because of the effects of European contacts, this insight has come too late.

In the same way, there has developed through the years a series of stereotypes as early speculations, later hypotheses, and subsequent models concerning aboriginal social organization have been hypostatized into elements of a society that never was.

Perhaps the most obvious example of this process has been the way in which the notion of the horde became accepted as the statement of a great truth (Radcliffe-Brown, 1930, 1951). Ignoring the caveats entered long ago by such investigators as Elkin (1938, p. 40; 1953) and Piddington and Piddington (1932), anthropologists, especially those unacquainted with aboriginal society at first hand, simply assumed that the patrilineal, patrilocal horde was a real entity to be found everywhere in Australia. It is only in the light of recent statements to the contrary by field workers such as Berndt (1959), Hiatt (1962), and Meggitt (1962) that this neat but misleading picture is being replaced—albeit reluctantly[65]—by an acceptance of the great variability that actually existed in aboriginal local organization.

A similar sequence of conceptual hardening, followed by slow recognition of the need for empirical studies, has occurred in the investigation of systems of kinship and marriage among the aborigines. Although flcld workers such as Elkin (1932, 1938, 1938–40) and Kaberry (1939) urged that clear statements be made of the actual degree of variation or choice incorporated in these systems, once again anthropologists abroad simply reified the schematic statements of kin terms and of marriage prescriptions into concrete elements of social behavior and assumed that this was somehow the essential nature of aboriginal societies, that is, that they were in constant equilibrium.

In particular there has been a failure to appreciate the significance of demographic differences within and among these smallscale societies and their importance in conditioning marriage arrangements. Only in the last few years have ficldworkers published quantified data that bear directly on these problems (Rose, 1960a, 1960b; Goodale, 1962; Meggitt, 1962, 1965b; Hiatt, 1965).[66]

But, while the results of empirical research in these several areas have slowly broken through the accretion of reiflcations that has grown out of former misconceptions, there remains a set of social phenomena about which many anthropologists have been firmly in error. The phenomena are the modes of classifying the world in social and ritual terms which take the form of moiety, section, and subsection systems; the mistake is to regard these schematic categories of cosmic classification as substantive or causally effective “marriage classes, “ to see them as the primary determinants of marriage arrangements.

The confusion clearly stems from an earlier anthropological preoccupation with the derivation of monogamy from putative group marriage. For exponents of such hypothetical schemes the discovery in the nineteenth century that certain aboriginal tribes were divided in halves, quarters, or eighths, and that these divisions were in some sense exogamous, was of great moment. Here, at a stroke, they could add the missing rungs to their evolutionary

ladders. Thus, from an initial stage of group marriage in which all males of the group (= “tribe”?) might mate indiscriminately with allY of the females, there developed (for unknown reasons) a situation in which half the women (= that half including mothers and sisters?) were excluded as mates. The presence of moieties among some aborigines was taken to be a survival of that period. Similarly, the existence of sections among other tribes attested to an earlier extension of the prohibitions on mating to three-quarters of the women of the tribe, just as subsections reflected an extension to seven-eighths of the women.

After this, various specifications of mates in terms of classificatory kinship reduced mating to limited polygyny and finally, although not in aboriginal societies, one and only one woman at a time could be the mate of a given man.

r-b-richard-b-lee-irven-devore-man-the-hunter-21.jpg
Figure 1. Locations of Aboriginal tribes in Central Australia.

Obviously such a simple explanatory scheme was unlikely to survive long but, even though it succumbed to reason, it had significant effects in that subsequently most anthropologists, whatever their theoretical orientation, continued to assume that moieties, sections, and subsections were “marriage classes” in substantially the sense intended by writers such as Frazer, Fison, and Morgan. Thus this assumption not only importantly conditioned the early attempts of men such as N. W. Thomas (1906) to treat Australian data from the armchair; it also significantly biased the observations made in the field by investigators and travelers as varied as Howitt ( I889a, 1904), Spencer and Gillen (1899, 1904, 1927), Basedow (1925), and Chewings (1936).

But even as the speculation that moieties, etc., were “marriage classes” was congealing into received dogma, several anthropologists engaged in intensive field research among aborigines were discovering data that led them to express serious doubts about the proposition. Elkin, guided by his usual meticulous concern for the facts of field work, was one of the first to suggest the need for a reassessment. Thus in 1933 he said:

too much emphasis has been placed on (sections and subsections) as methods of regulating marriage and of classifying kin at the expense of understanding their totemic significance and function ...though their number and incidence is in ideal intimately related to social behavior (marriage and kinship), they are just as much concerned with the ordering of man’s relationship with nature. The latter function, indeed, may be primary (1933, p. 66; see also his comments in The Australian Aborigines, 1938).

Considerable support for Elkin also came from the observations of Kaberry (1937, 1939), Stanner (1936), and Warner (1937). All three offered firsthand evidence to show that there is no necessary connection between the presence of sections, etc., and the regulation of marriage in aboriginal society. Especially relevant was the demonstration that a tribe such as the Kariera, the Murngin, or the Murinbata could readily adopt a section or subsection system that was incompatible with its prescriptive rules of marriage, and that nevertheless both usages could be maintained. Subsequent field work by Falkenberg (1962), Meggitt (1962), and Hiatt (1965) confirmed these kinds of conclusions.

Despite presentation of such evidence in the 1930’s, however, the view persisted in some quarters (and apparently is still held today) that among aborigines moieties, sections, and subsections are groupings, membership in which crucially determines marriage arrangements or choices.[67] Thus Lawrence (1937) argued this at some length and with much ingenuity, although it is difficult to see what he proposed as evidence for his assertions. Later Murdock (1949) took Lawrence’s position as being in some way conclusive, but again it is hard to discern the empirical basis for this.

In the light of the history I have so briefly sketched, I want to devote the rest of this paper to presenting further evidence (albeit of a negative sort) to support the view that subsection systems are not primarily concerned to regulate marriage.[68]

The argument is simple. As I have already remarked, data published by Rose, Goodale, Hiatt, and myself indicate that demographic factors, or variations in them, may in particular cases exert considerable and fairly direct influence on patterns of marriage in aboriginal society. But, as I shall show, demographic differences have little or no effect on the presence or even on the functioning of subsection systems. The reason seems clear-marriages, as well as marriage prescription, concern finite numbers of real people in interaction; if, as can easily happen in a situation of change, there are not enough individuals to fill or to exemplify the pertinent statuses or categories, the system simply fails and soon ceases to be. On the other hand subsections, like moieties and sections, are fundamentally conceptual schemata for the classification, not merely of spouses or even of people at large, but of the entire world that the aboriginal knows. That is to say, as long as that universe exists, or substantial parts of it persist, all categories can be occupied (although not necessarily all by people) and the system .can continue. It follows then that demographic changes and changes in local and economic organization (even though considerable) of the kind that have commonly been associated with European contacts need not entail a breakdown in the subsection system of a particular tribe.

Table I Central Australian Tribes-Statistical Summary

r-b-richard-b-lee-irven-devore-man-the-hunter-42.jpg

The sample of tribes from which I draw my data is fairly small-fifteen tribes in all. They are all located in the Northern Territory, and all are hinterland groups that normally reside in arid to semi-arid country (see Fig. I). Thus these tribes are reasonably comparable in terms of important environmental variables and in terms of population densities. I deliberately avoided using groups from the more richly endowed coastal regions. Perhaps the easiest way to locate the sample is to note that these people are all immediate or close neighbors of the Walbiri, the tribe with which I have worked intensively and which I know best.[69]

Another point, relevant to my argument, is that members of this sample have experienced different degrees and kinds of contact with Europeans, ranging from mildly to markedly disruptive. But whatever the effects that these changes have had on local organization, economic conditions, or marriage arrangements, none of the tribes in the sample has abandoned its subsection system.

The population figures I use here derive from the 1957 Register of Wards compiled by the Welfare Branch of the Northern Territory Administration, which indicates-where appropriate—section, subsection, or clan affiliation as well as tribal membership. Comparison of the Walbiri figures given there with my own suggests (to me at any rate) that the census as a whole is tolerably accurate. If anything, it tends to overestimate some tribal numbers because the mobility of the aborigines may lead to double counting.

I summarize the relevant figures in Table I . The symbols I use have the following meanings:

N = the number of members of the tribe.

Xe = N/16 = the number of males or the number of females expected in each subsection if they were evenly distributed; that is, N18 = the number of people ideally in each subsection.

IDI =:Re — Xo = the difference (ignoring arithmetical sign) between the expected and the observed number of males or of females in each subsection of the tribe.

I./D/ = the sum of the deviations, ignoring arithmetical sign.

Id = I./D/“5<:. e = the sum of the deviations divided by the expected mean.

= the Index of Deviation for the tribe.

C% = the ratio of children under the apparent age of 14 yearstothetotal size ofthe tribe.

F% = the ratio of females of all ages to the total size of the tribe.

The first point to note is the range of tribal size in the sample-from a mere 54 members to 1,570 members. But despite these differences in size (which in some cases at least are directly connected with different kinds of encounters with Europeans), every one of these tribes still has a subsection system. Indeed, the names of the categories are cognate from tribe to tribe, and tribesmen are able to employ this resemblance when visiting each other.

The second important point is that in every tribe the observed distribution of males and females (both adults and children) through the subsection system differs appreciably from the expected mean. Although I am not convinced of the propriety of utilizing the Chi- squared test of significance in such an inquiry,[70] application of the test indicates that probably only in one case (that of the Kaitish) is the difference between the observed and expected distributions not statistically significant. I can given no reason for the one exception.

Third, when Indices of Deviation are computed for all fifteen tribes and are correlated with the actual size (X) of the tribes in terms of the Rank Difference coefficient, the result Rho1d, N = — .37 is not statistically significant. That is to say, there appears to be no significant relationship, positive or negative, between simple size of the tribe and the distribution of tribal members through the subsections.

Fourth, when the Indices of Deviation are correlated with the ratios of children (C%) in the fifteen tribes, the result Rho1d,C% = + .20 is also not statistically significant.

Fifth, when the Indices of Deviation are correlated with the ratios of females (F%) in the fifteen tribes, the result Rho1d. F% = +.I I is again not statistically significant.

These seem to me to be the only calculations that may safely be made on the basis of the specifications of the available sample of tribes, but I am persuaded that they enable us to say confidently that, for these hinterland tribes at any rate, intertribal demographic differences have no discernible effects on the presence or absence of subsection systems. A fortiori, then, these results may be taken to support empirically the proposition that, among the Australian aborigines, subsections are not necessarily concerned in a substantive manner in the regulation of marriage.[71]

Appendix

Examples of actual distributions of individuals through subsections and moieties.

r-b-richard-b-lee-irven-devore-man-the-hunter-45.jpg
r-b-richard-b-lee-irven-devore-man-the-hunter-46.jpg
r-b-richard-b-lee-irven-devore-man-the-hunter-47.jpg

20. Demographic and Ecological Influences on Aboriginal Australian Marriage Sections

Aram A. Yengoyan

Aboriginal Australian social structure has been traditionally regarded as a complex cultural elaboration which integrates whole societies or “tribes.” Most sociological interpretations have stressed the formal aspects of structure, although Kroeber (1938, p. 217) has attempted to demonstrate that formal aspects of Australian social organization (such as clans, moieties, descent groups, and totemism) are secondary structural patterns which rely on primary patterns such as subsistence and locality. In summary, Kroeber (1938, p. 218) concluded that the elaborate system of Australian marriages takes the form of unconscious experiment in the manner of fads and fashions. Though there is no disagreement with Kroeber’s distinction between primary and secondary structures, one can examine the functions and interrelationship between section systems and environmental and demographic factors.

The specific purpose of this paper is to examine the ecological and demographic factors which regulate different section systems.[72] The following points will be investigated:

  1. A correlation between different section systems and certain environmental conditions is present and empirically observable.

  2. Population size is an important feature in the operation of section systems and in part determines their ideal functioning.

  3. Section systems may also function in regulating interaction over a delimited area, thus servimg as a means of “insuring” utilizable resources.

The analysis of such problems is always limited due to the scarcity of reliable quantified data on population structure, areas of exploitation, etc. Population models, which are constructed from a generalized knowledge of hunting and collecting groups, may be utilized as an operational device to illuminate certain problems and areas of investigation. The population pyramid in this paper represents such a device, since accounts of precontact aboriginal population structure were unobtainable.

Method

The population parameter (population size, area, and density) represents the primary unit in relating environmental factors to section 186 Marriage and Models in Australia systems, and also the pivotal point in the ideal operation of section systems as regards their function in marriage regulation. Birdsell (1953, pp. 205–206) has demonstrated a .8 correlation between mean annual rainfall and population density for 123 ecologically homogeneous tribes. Since water is the most important environmental variable, it has a direct influence on the type of local biota and cultural variables which regulate a group’s extractive efficiency and population structure.

Population size, area, and density are then correlated with different section systems. The population pyramid is utilized to demonstrate how tribal size may set specific limitations on the functioning of any marriage system. Popu- -lation area figures are not always available nor reliable. However, one readily notices from Tindale’s (1940) map that the largest tribal areas are located in extreme arid conditions and are commonly characterized by the occurrence of subsection systems.

Environment and Subsistence

Rainfall represents the determining variable for the existence of different vegetation belts in Australia. Mean annual rainfall varies from 2–3 inches to 150 inches per year. Highest aridity occurs in the center of the continent; as one moves outward toward the coastal regions rainfall increases reaching a maximum in the Cairns rain forest of Queensland (G. Taylor, 1943, P. 83).

Mean annual rainfall is not always an adequate criterion for available surface waters. For example, unearned or secondary waters may fall in one tribal area but are utilized in another. This phenomenon is present where one finds large rivers wi1 h substantial seasonal drainages. An example of how secondary waters affect assumed population densities occurs in the region of the Murray and Darling river drainages. The mean rainfall in this area is 5–10 inches per year; however, population densities are exceedingly high on the riverine frontages. Both rivers have their sources in the high dividing ranges to the east which are characterized by a 30–50 inches of rainfall. The runoff from these drainage systems is sufficient enough to produce a flourishing biota and a relatively high population density downstream (Commonwealth of Australia, 1953).

Even in areas with an adequate rainfall, deserts may be common. This phenomenon occurs in regions lacking zinc and copper as trace elements, thus making any type of vegetation quite sparse (Anderson and Underwood, 1959, p. 97h

The mapping of surface water supplies represents a difficult task. Johnston (1940) marked available water supplies along aboriginal routes for the western portion of South Australia. Although this account is quite limited, it does provide an adequate description of routes and names dealing with water supplies. Routes in the Ooldea region are from one water source to the next, thus the utilization of a route varies with seasonal differences in water supplies. Water is derived from rock holes, soaks, wells, claypans, dams, roots of the desert oak, and the mallee tree. The more important water supplies have some totemic significance and play an important part in — aboriginal ceremonial and social life (Johnston 1940, pp. 34 35)-

Since water supply distribution are unavailable for the remainder of the continent and mean annual rainfall data are not precise enough, effective annual rainfall may be utilized. Effective annual rainfall is the amount of rain utilized after evaporation is excluded. Hence, in the Australian case, rain in excess of one-third of the evaporation is termed effective rainfall. Evaporation in most areas is high and in the central regions it may exceed 100 inches per annum. Approximately 76 per cent of Australia has a growingperiod offour months or less (Commonwealth Scientific and Industrial Research Organization, 1960, p. 30). Australian vegetation patterns are directly related with effective annual rainfall. Rainforests exist in coastal Queensland, eastern New South Wales, southern Victoria, and the southwestern portions of West Australia, while spinifex desert areas are located in the central and western parts of the continent.

The diet of the Australian aborigine is basically vegetarian (Meggitt, 1964b), though nearly all edible animals and plant foods are consumed. Diets vary from desert areas to coastal regions, yet in both cases the samples substantiate the claim that the aborigines exploit practically all sources of available energy and cover the full range offoods which may be economically utilized. The Walbiri of central Australia represent a good example of how a desert group has extensively utilized the spini- fex environment (Sweeney, 1947, P.289). Native plant foods are divided into three groups: bush fruits, root foods, and seeds.Bush fruits consist of wild oranges, wild figs, desert plums, berries, desert tomatoes, desert raisins, desert yams, and onion grass, while seeds consist of pod seeds, grasses, and wild herbage. The dominant animal foods are kangaroos, euros, desert marsupials, rock wallabies, lizards, carpet snakes, and rabbits. Honey, eggs, gall insects, and grubs are other miscellaneous foods (Cleland and Johnston, 1933, p.114).Annual precipitation is about eight inches, thus the growing period is quite short though it is longer than in the Lake Eyre region to the south.

Even in high precipitation areas such as southeast Australia, no single grain is obtainable in vast quantitites. All forms of life compete for available resources, yet one notes a regulation of activities among animals and man which to a certain extent are mutually interdependent relationships. As one moves from the center to the well-watered fringe areas, resources are more abundant and seasonal availability is longer. Utilized resources increase in quantity and quality on the coastal regions where a new biota is exploited.Coastal tribes have access to marine resources such as fish, whales, dugong, turtles, and shellfish, as well as different plant foods such as coastal yams, lilies, pandanus, cycad nuts, and numerous varieties of roots and grasses.

For a relatively dry continent such as Australia, ranging from tropical to desert habitats, the variation in flora will be dependent to a large degree on rainfall. The biomass of the flora will be expected to vary directly with the intensity of annual rainfall. Since man stands at the peak of the food chain, exploiting all lower trophic levels (animals and plants) population densities must also respond to rainfall variation.

All technological implements and skills revolve around the primary subsistence patterns of hunting and food collecting. Wood, bone, and stone are the most commonly utilized materials, with the proportion of utilization varying with the habitat. The two most important extractive devices are the stone axe or chopper and the digging stick. Sharp (1952, p.69) stated that among the Yir Yiront, the stone axe was not only an important male implement, but also possessed a high degree of ritual value since it symbolized masculinity and age. Tindale and Noone (1941, p.122) observed similar conditions in central and western Australia.Stone is used for flakes, scrapers, bulbs, and adzes; however, in most cases each of these tools may be remanufactured from one type to another (Tindale, 1957, P.40). The adze is a crucial tool since it is used to make boomerangs, spearheads, sacred objects, water bowls, winnowing troughs, etc.Women’s tools are digging sticks, clubs, and occasionally stone implements.

Both sexes are fully involved in extracting daily food supplies. Hunting activities are predominantly men’s work, while women collect plant foods, seeds, eggs, and occasionally hunt rabbits. Women are often better “trackers” than men, although if large game is seen or water is located, the women will tell the men who, in turn, will decide what course of action to follow.Females will collect foods up to four or five miles from a campsite, while men will go nearly seven or eight miles from a camp in order to hunt large game.When foods are exploited within a 6- to 8-mile radius from a camp, the local group will shift to another site. Collecting is generally foreign to men’s work, although hunters will gather plant foods to fill immediate needs while away from the camp.

All food is distributed among the local group and promptly consumed. Ideally the local group claims everything in their own territory; however, if certain foods are abundant, any individual may partake in it.

Tribes and Sections

The population unit employed in this paper is the tribe. Though there is much controversy about definitions and formal aspects of aboriginal local organization (Birdsell, 1958; Hiatt, 1962; Stanner, 1965) and about the concept of the tribe (Berndt, 1959; Davidson, 1938a), the use of the term tribe is primarily arbitrary.Since the majority of trained anthropological field workers as well as other observers have presented demographic and cultural data under this or that “named” tribe, the problem of defining a tribe to fit all particular cases is not necessary. However, a tribe may be considered as a reference group by which individuals identify with each other either through common language, ritual and legend, kinship and marriage, or locality.

The Australian tribe possesses no effective political and economic control. Although tribal functions arc minimal as regards sociocultural factors, it does represent an effective breeding unit. Intratribal marriages constitute from So-go per cent of all unions (Tindale, 1953), and it may be suggested that rates of intcrtribal marriages may vary with demographic factors. Whereas the local group possesses mechanisms for socioeconomic activity and social control, the tribe as an “ideally” endogamous unit, though not always, would designate a genetic and demographic unit.

Sections and subsections should not be fully equated with marriage classes, nor do sections only function in the regulation of marriage choices (l\leggitt, Chapter 19, this volume). The same author ( I 964a) has noted that sections and subsections “were essentially ritual categories,” while Elkin has stated that “what the [sections] do is to summarize relations into four groups, in one of which a possible spouse is normally found; but in the same group various other types of relations are also included.” The determination of marriage mates is based on kinship relationships.

Thus, the section system is a shorthand index for combining kin relations into categories for multiple purposes, be it possible marriage mates, ritual, or economic activity. Finer distinctions within each section category are left to genealogical connections and locality. Meg- gitt (1962, p. 169) also recognized sections as simplified combinations of individuals when he stated that “basically, [the Walbiri] regard subsections as a summary expression of social relationships that may sometimes be practically useful but just as often is dangerously ambiguous.” Hiatt’s (1965, p. 50) interpretation of Gidjingali subsections followed Meggitt’s analysis.

While marriages arc based on kinship relations and Jocality, the analysis of marriages is commonly summarized in terms of sections or subsections. Thus, preferred, alternative, and wrong marriages stem from the kinship system but are presented within a section-subsection framework. In this paper, sections and subsections are utilized as a summary index of marriages, and not as the determining factor in regulating marriage choices.

The Population Parameter

For operational purposes, a parameter is defined as all key quantifiable variables that structure a population. Furthermore, a parameter not only designates the population of a particular unit but also circumscribes one unit from other comparable ones. Thus, to a geneticist, the parameter delimits effective breeding units which may be compared to one another for the demarcation of clines. In utilizing the tribe as a unit of comparison, the governing variables which characterize the demographic structure are population size, area, and density. Size and area co-vary with environmental factors, while density is the function of both variables.

Since Birdsell (1953) has demonstrated a correlation between population density and mean annual rainfall for ecologically homogeneous tribes, a relationship may also exist between population size and area, and rainfall. Tindale (1940, p. 150) noted that the size of tribal areas was inversely proportional to rainfall. In regions where tribal areas were large, the annual rainfall was remarkably low. There was also a high degree of correlation between tribal limits and ecological-geographical boundaries. Lines of communication and migration tended to follow natural lines of least resistance across the continent, often clinging to open plains, creeks, rivers, and lines of water along ranges, and avoiding dense forests and rugged mountains. Tindale’s (1940) tribal map reveals that the largest tribal areas occur in the central and western deserts, while the small, more fragmented areas are along the coast and in other favorable habitats.

The average Australian tribal size is given by Krzwicki (1934, p. 305) as 500, with a range from 150 to 2,500. Elkin’s (1964, p. 25) figure for the tribal average is 500 to 600, with a range from 100 to 1,500. Tindale (1940) also recognized an average tribal size of 500, though he noted that it may IJe high. Lumholtz regarded tribal size as about 200 members in regions adjoining rain forests (Wheeler, 1910, p. 33). In general, one would expect to find tribes whose sizes are lower than the continental average on the coastal and interior coastal areas. Although the coastal tribes have less area and fewer individuals, they still possess the highest population densities.

In order to observe variations in population size and density, different environmental areas will be briefly discussed. In West Australia, from the Kimberleys, southwest to Eighty Mile Beach, a number of tribal groups on the coast and interior coastal area utilize both marine and terrestrial resources. The coastal Kariera are one of the best reported groups (RadcliffeBrown, 193 I). The Kariera occupied an area of 3,750 square miles with an estimated population of 750. The tribe was divided into 20 to 25 local groups, each occupying a defined area. Band size varied from 30 to 40 individuals. Population density in this relatively favorable environment was one person per five square miles. Elkin estimated that the well,watered portion of the Kimberley district may have had an overall population of 10,000 living on 65,000 square miles, or one person per six or seven square miles (Meggitt, 1 964, p. 166). Reports by Hernandez (1941), Stanner (1932), and Webb (1933) for other coastal and interior coastal groups indicated similar population conditions.

Further southwest, near the Swan River, Stirling estimated one native per two square miles during the early years of the settlement, while Radcliffe-Brown (1930, p. 689) calculated one person per four square miles. Population densities become progressively lower in the eastern portion of West Australia (that is, Great Sandy Desert, Gibson Desert). The Rawlinson Ranges area, from Lake Christopher east to the western edge of the Petermann Ranges, may have had a density of one person per 30 to 40 square miles. This area of approximately 5,400 square miles contained a population of 172 persons.[73] For the whole of West Australia, Radcliffe-Brown (1930, p. 695) estimated a former aboriginal population of 52,000 and an overall population density of one person per 18.8 squat.’: miles.

The northern section of the state of South Australia (that is, north of Port Augusta) represents one of the lowest rainfall areas in Australia. This area is approximately 300,000 square miles and has a rainfall average of less than ten inches per year. Tribes with higher population densities occupied the area of the Murray drainage. Radcliffe-Brown (1930, p. 690) estimated an aboriginal population of 6,000 for the southern portion while the northern part of the state may have had population densities as low as one person per eighty square miles. The Musgrave Ranges in the northeast had a population density of one person per 30–35 square miles, though yariations between the ranges and plains alter this estimate. The Ooldea groups in western South Australia were tribally fragmented due to conflicts over trade routes and sacred waters. From Berndt’s (1942) account, one deduces that the Ooldea area may have supported a relatively large population though the overall density was exceedingly low. Howitt (189 I, p. 33) estimated the Dieri of the eastern lakes region at I, I 00 members with a density of one person per twenty square miles; however, this density figure may be too high due to the exceedingly poor environmental conditions. A total population of approximately 3,000 for the whole of the lakes region may have been reasonable enough (Elkin, 1931, pp. 51–52).

Southeast Australia (Victoria and eastern New South Wales) represents the most extensive cultural elaboration and highest population densities in Australia. Curr (1886–87) estimated the total Gippsland population at 1,500, and Howitt’s (1904, p. 23) figures for whole Gippsland area yielded a population density of one person per 9–10 square miles. Radcliffe-Brown (1930, p. 691) assumed an irreducible minimum figure of one person per fifteen square miles for the Gippsland area. For the Mt. Burr and Mt. Gambier areas, Campbell (1939, p. 35) concluded that 2,000 aborigines lived in this region in 1895 though they were being exterminated at a rate of 50 per cent every five years. On the southern frontage of the Murray and Darling rivers, Howitt (1904, p. 53) designated an areaof2,500square miles occupied by ten tribes totaling 1,200 individuals. Population density in this area was approximately one person per two square miles. For the same area, Curr estimated one person per 2.3 to 3 square miles (Radcliffe-Brown, 1 930, p. 69 I) and Wallace in the 1870’s reported that local groups along the Murray drainage consisted of 60 to 100 people (Campbell, 1934, p. 28).

Coastal New South Wales had an estimated population of 25,000 with a denser population on the north coast (Radcliffe-Brown, 1930, p. 695). Meggitt (1964, p. 165) noted that early observations of the Sydney district from Botany Bay to Broken Bay suggested a population range of 1,500 to 3,000 occupying a narrow coastal strip of 400 to 500 square miles.

The Queensland coastal area was generally similar in tribal densities to the southern coastal regions. From Broad Sound to the Clarence River, the population density was one person per three to four square miles (RadcliffeBrown, 1930, p. 694). On the Cape York Peninsula, the Yir Yiront had an average density of one person per 4.8 square miles, but then density was twice as high (one person per 2. 4 square miles) if only utilized land is considered (Sharp, 1940, p. 486).

In general, Arnhem Land population densities were unusually high and are comparable to the coastal tribes of Victoria, New South Wales, and Queensland. The Groote Eylandt inhabitants in 1939 numbered 300 to 350, with a density of one person per three square miles (Rose, 1960b, p. 12). Hiatt (1965, p. 17) reported Gidjingali population density at two persons per square mile on an occupied area of 150 square miles. The Carpentarian Gulf tribes also had an overall density oftwo persons per square mile according to Thomson (I 934a). The Kaiadilt of Bentinck Island showed comparable and even higher population densities. Tindale (1962b, p. 304) stated that

The density of population for the whole of Bentinck Island area in 1940 (the last year when all were present) was 1.7 persons per square mile of total land and reef surface, or over 6.8 persons for each square mile of reef. Since part of the total area is inaccessible, and only used at some risk (as indicated by two tragic episodes accompanying efforts to reach Allen Island in 1940 and 1947) only about 14 square miles of reef were in constant use, i.e., over 8 persons obtained their food on each square mile of reef (1962b, p. 304).

Murngin population density was about one person per eight or nine square miles (Warner, 1937, p. 16). On the Daly River, 100 to 150 Murinbata occupied 500–700 square miles (Stanner, 1936, p. 187).

As one moves into the more rigorous environments of central Australia, densities become lower since desert populations need more land to support themselves. The Aranda, who occupied the more productive portions of central Australia, numbered less than 2,000 and occupied over 25,000 square miles (Spencer and Gillen, 1899). For the Walbiri, Meggitt (1962, p. 32) concluded that the precontact population ranged from 1,000 to 1,200 in an area of 35,000 square miles. The average population density was one person per 35 square miles. These low densities are comparable to other desert areas such as the Rawlinson Ranges and the Pitjandjara of the Musgrave-Mann-Tom- kinson Ranges.

The total aboriginal populatiqn of Australia had an overall average density of one person per twelve square miles. Radcliffe-Brown’s (1930) enumeration of the aboriginal population by states is as follows:

r-b-richard-b-lee-irven-devore-man-the-hunter-48.jpg

The remainder of the population data, which has been extracted from Curr (1886–87), Krzwicki (1934) and other sources, is listed in Tables I, 2, and 3.

From the variation in population size and densities, one notices how ecological factors interplay in limiting and preserving environments. “Primitive” man’s territoriality is not only limited by cultural factors, but also by environmental and demographic components which regulate population size and tribal interaction.

Section Systems and the Population Parameter

The correlation between section systems and each component of the population parameter is listed in Tables I, 2, and 3. In each table, the absence of sections, which is labeled “0,” the moiety organization (“2”), sections (“4”), and subsections “(8”) are horizontally demarcated. Table 1 lists tribal areas by square miles, while Tables 2 and 3 deal with population sizc and density respectively. Since the samples are relatively small, the median rather than the average or mean has been utilized in each case. From Table I, note that with an increase in tribal area, there is a complementary change in the number of sections, except from “0” to “2.”

Tribal sizes also increase with the number of sections. With a sample of 76 cases, this correlation may represent a higher degree of assurancc. However, the subsection systems are represented by only seven cases, thus limiting the quantitative base of our statements. further investigation on the relation of tribal size to number of sections follows In the Discussion.

In the plotting of median points for tribal area and size (Tables 1 and 2), the same pattern emerges in each case. With a corresponding increase in tribal area and size, one notes an increase in tribes with sections and subsections. However, this relationship may not hold true for those tribal groups wbere the most important marriage restriction is local group exogamy. Since the “0” tribes occupy the environmentally favorable coastal areas, tbe hypothesis as stated earlier may not fully explain this type. As is expected, the correlation between section systems and population density also patterns in the same direction as the two previous cases.

Table I Sections and Tribal Areas

r-b-richard-b-lee-irven-devore-man-the-hunter-49.jpg

Figure 1 presents the spatial distribution of section systems in aboriginal Australia. Both Davidson’s (1928a) and Radcliffe-Brown’s (1930) maps are basically similar, yet in a few particulars there are notable differences. Davidson’s information on the Kimberleys and Arnhem Land was practically nil, while RadcliffeBrown had access to the most recent field work of Elkin, Hart, Stanner, and Warner. Furthermore, Davidson did not distinguish between marriage and descent functions of the section system, a distinction which Radcliffe-Brown fully utilized. However, Radcliffe-Brown designated the Murngin as having a subsection system, yet Warner (1937) noted that although Murngin descent is based on an eight-subsection system, their marriage restrictions are on a four-section basis. Both authorities made systematic errors in certain localities such as the Cairns rain forest of coastal Queensland. For this area, Davidson and Radcliffe-Brown interpreted the tribal social organization as

Table 3 Sections and Tribal Density

r-b-richard-b-lee-irven-devore-man-the-hunter-50.jpg
r-b-richard-b-lee-irven-devore-man-the-hunter-22.jpg
Figure 1. Geographic distribution of section systems in Aboriginal

Australia. Upper: After Davidson; Lower: After Radcliffe-Brown.

possessing section marriage restrictions; however, Tindale and Birdsell (1941, p.6) noted that the Cairns people had only moiety restrictions on the coast and section systems in the interior.A more recent map and discussion on the distribution of section systems is available in Berndt and Berndt’s ( 1964, pp.53–60) new volume on the Australian aborigines.

From the data in Tables I, 2, and 3, the following conclusions may be tentatively stated:

  1. Those tribes characterized by subsections possess the largest tribal areas and sizes, and the lowest population densities.

  2. Following the first point, as the number of sections decrease from eight to four to two, tribal areas and sizes also decrease and population density goes up.

  3. Since environmental factors strongly influence the population parameter, and since the latter is related to the number of sections into which a tribe is divided, it may be tentatively concluded that an indirect relation exists between sections, population, and environment. The remainder of this paper is devoted to the investigation of the processes by which section systems are regulated by population size.

Discussion

In this section, population size is explored in terms of its functions and limits on the “ideal” operation of marriage sections. A population pyramid (Fig. 2) is utilized in analyzing the interrelationships between population size and marital unions. The pyramid is based on generalized hunting and gathering groups, and on a few assumptions to be spelled out.

Figure 2 shows the age and sex breakdown of a tribe numbering 100 individuals. First, a balanced sex ratio for each five-year interval is assumed, although specific ethnographic cases may not verify this assumption for all age levels. Second, population equilibrium is assumed. In a population where equilibrium with the carrying capacity is stabilized, each adult must reproduce himself, consequently a married couple must replace themselves with two offspring who reach reproductive age.

Populations control mechanisms include natural infant and child mortality, abortion, infanticide, birth spacing through late weaning, and so forth. For example, infanticide rates among aboriginal Australians may have been as high as 30 per cent (Birdsell, 1958).

The age of first marriage amohg Australian males varies from the early twenties to the mid-thirties, and the average is between the ages of 25 to 30. Marriage age for females varies from eigh’t to ten years among the Walbiri and some Arnhem Land groups to the late teens among the Pitjandjara.Although age of females at marriage in many groups is relatively early, menarche and conception usually occur later. A young female may not effectively reproduce until 15 or 16, and only after the birth of her first offspring will her economic role become important. Thus, females may become “effective” marital partners—for economic and reproductive purposes — at about 15 or 16 years of age. For convenience, an overall age bracket in which practically all females would be married is 15 — 19, though one must recognize earlier marriage in some groups.

If all able males over 25 are of marriageable age, then a total of thirteen males in a population of 100 may take a spouse at any one time. Any women over 15 may acquire a spouse at any particular time, thus resulting in 24 eligible females. Older widows are often given to younger men as first spouses and elder males often have primary access to younger females (Rose, Chapter 2 I, this volume).

To balance the model with ethnographic fact, three alternative stipulations concerning marriage are set forth. The first stipulation is that a male marries only females younger than himself though all males over 40 have access to all eligible females. This results in an average of seventeen eligible females at any one time.The second stipulation, that males marry females younger and up to five years older, results in 19.2 (= 19) eligible females. The third stipulation is males marry females younger and up to ten years older. This yields 20.8 (= 2 I) eligible females at any one time in a population of 100. The computations for assessing eligible females are in Table 4.

Of the three stipulations, the third one is closest to ethnographic reality. While older males control the younger women, the younger males are often provided widows as their first spouse. In fact, an estimate such as the third stipulation is still conservative (see Goodale, 1962; Meggitt, 1965; Rose, 1960b). The age difference for the first marriage of either sex to their first spouse is quite marked.

r-b-richard-b-lee-irven-devore-man-the-hunter-23.jpg
Figure 2. Population Pyramid for a Hypothetical Tribe Numbering 100.

The interrelationship between the population pyramid and different marriage restrictions is found in Table 4. Assuming a hypothetical number of ten local groups and an ideal situation where each group has equal representation as regards females, a local group would then have 2.1 eligible females per male. With only local group exogamy (the “0” condition), a male has a total 18.g possible mates. When moiety restrictions are imposed the number of potential mates, and possibly potential competitors, is reduced by one-half. Subsequently, with section marriage restrictions, the number of eligible females is reduced to onefourth of the “0” condition, and with subsection marriage to one-eighth. Thus, with a tribal size of 100 and the existence of subsections, a male would ideally have only 2.36 possible marriage mates.

Table 4 Number of Possible Mates under Four Sets of Marriage Restrictions

r-b-richard-b-lee-irven-devore-man-the-hunter-51.jpg

The correlation between number of eligible mates and population size is presented in Figure 3. Extrapolating from the tribal model of 100 to larger, more realistic population sizes, allows us to ascertain the number of eligible mates under varying tribal sizes and section restrictions. A tribe of 500 people with subsection marriage rules yields an ideal figure of nearly twelve eligible females per male at any one time.

When the results of the model are compared with the empirical data in Table 2, a few interesting observations are noted. First, is the occurrence of a tribe of 50–100 individuals with a subsection system (listed in Table 2 as “?”). The reality of this case is doubted since the tribe was occupying approximately IO,500 square miles thus resulting in an extremely low population density of one person per ioo-2OO square miles. A small population with low densities operating on a set of highly selective marriage restrictions would be most impractical.

Also, the presence of ideally two eligible mates would be improbable since local groups occupy large areas with wide intervening distances.

From Table 2, the medians for tribes with moiety, section, and subsection restrictions are 250–300, 550–600, and 1,100–1,150. When the median figures are plotted on Figure 3, the result is that in each case a figure of 25–27 eligible mates per male is derived. With the existence of 25 eligible mates and only moiety restrictions, an ideal tribal size figure is approximately 262 which falls between the median range of 250 to 300. Twenty-five eligible mates in a section system yields a size of 530 while the empirically derived median is 550 to 600. With subsections and 25 eligible mates, an ideal tribal size of 1,070 is postulated. Once again, the median of 1,100–1,150 is remarkably close to the predicted figure from the model.

If the existence of 25 eligible mates represents an operational number, then it is suggested that those populations with sections and subsections whose tribal sizes are significantly lower than the predicted range could not statistically confirm ideal marriage rules. Three cases may be briefly discussed to illustrate this point. Walbiri population size in 1954 was approximately 1,400 (Meggitt, 1962, p. 31), a figure well above the predicted range of 1,070 for tribes with subsections. As might be expected, preferred unions with matrilateral second cross-cousins occurred in 91.6 per cent (566/617) of all cases (Meggitt, 1962, p. 86) and thus confirm the stated marriage rule. At Borroloola near the Gulf of Carpentaria, Reay ( 1 962) discussed subsections and semimoieties among eight small tribal groups of which the Garawa and the Anyula are the largest. The total population of these groups in 1 958–59 was 268, a figure well below the predicted range of 1,070. Consequently, out of 88 unions only 51 (57.95 per cent) marriages were regular. Furthermore, the natives were fully conscious that appropriate subsection partners were not always available and that highly frequent irregular unions did not follow the ideal marriage pattern (Reay, 1962, pp. 109–110).

r-b-richard-b-lee-irven-devore-man-the-hunter-24.jpg
Figure 3. Relationship between population size and the number of eligible mates.

Among the Lunga, Djaru, Kunian, Punaba, and Wolmeri of the Kimberleys, Kaberry (1939, p. I IS) noted a range of 47.2 per cent to 73. I per cent for all regular marriages. However, the tribal population of these groups is not listed, though detribalization was in process.

It may be concluded that population size sets definite limitations on the operation of section sy;tems. Thus when tribal size significantly decreases from the predicted range, certain compensating factors such as increasing intertribal and theft unions with consequent destructive conflicts, become prominent. However, it must recognized that subsection systems in all probability have never operated entirely according to the ideal pattern..

Intertribal exchanges, theft unions, and wrong and alternative marriages may occur as adaptive factors when ideal marriage partners are infrequent through decreasing tribal size and other factors. However, all intertribal marriages can not be attributed to the unavailability of designated mates. Tribes such as the Pitjandjara and Ngadadjara regard intertribal marriages as usual events which fit into existing marriage patterns (Tindale, 1953, P. 173)-

Marriages with alternative and prohibited mates vary in frequency and may be utilized as an index of population decline. The attitude towards marriages of this type differs from group to group, and in some cases, a wrong marriage is “corrected” after the union through partial realignment of kinship relations and terminology.

Another form of adjustment to population size is the fusion of two or more groups whose tribal identities are diminishing. In the coastal areas of the Kimberleys, one finds small tribes who socially recognize the acquiring of marriage mates from adjacent populations.

Conclusions

Population size is a critical factor in the ideal operation of section systems. When tribal numbers significantly decrease and the number of eligible mates is few and spatially dispersed, certain adaptive changes may arise to provide marriage mates. Since population size is regulated by ecological factors, the environmental influences on marriage patterns are important variables, though indirect. Furthermore, tribal units with exceedingly small and fragmented populations and subsections have been recorded, but statistically it is highly improbable that these groups could conform to stated marriage rules.

Many tribes with subsections occupy vast areas in the central arid regions of the continent. As Elkin has repeatedly noted, subsection groupings are simplified combinations of kin relations, and class terms are employed primarily among individuals and groups who are not in close association and whose interaction with one another is sporadic. Among the Pitjandjara and the Jankuntjatara for example, section terms are used in referring to individuals whose kin relationships are either ambiguous or non-existent. Among immediate and genealogically “close” kinsmen, kin terms are always employed. Strehlow also- noted similar use of class terms among the Aranda when he stated:

Since the usc of these kinship terms accurately indicated the social obligations that existed between persons who use them toward each other, the class-names or anbanerinlja were of prime importance when first contacts were being established between members of a local group and new arrivals who had come as guests, also between men belonging to a number of different local groups who had assembled on one ground for a ceremonial gathering (1965, p. 135).

Service’s analysis of egocentric and sociocen- tric relationship terms in Australia concludes with the following statement:

Where meetings are less frequent and the people are relative strangers to each other, the need for sociocentric terminology is greater—a stimulus is present. This would mean that in inland Australia social conditions exist which make the naming of the sections more likely there than in more productive areas (1960, p. 431).

One of the characteristics of section systems is the extension of section terms to distant kinsmen and non-kinsmen, who may occupy adjacent and distant areas of exploitation. The expansion of section terms not only combines spatially distant groups into meaningful units for marriage and ceremonial functions, but also permits reciprocal movements of related groups into larger territorial domains. For the Aranda, Fry (1934, p. 20) stated that “marriage with a distant kin not only minimizes the danger of infringing the tabu on parents-in- law, but also gives a certain right of entry to other districts in addition to that of the mother.”

Since sections and subsections commonly occur in tribal units with large areas, the degree of spatial distribution of one’s relations is greater than in small tribal areas. This network of related individuals and groups permits local groups to move into different areas of exploitation, especially during severe droughts and periods of economic hardship. The extension of kin relations via section groupings thus “insures” for each local group the ability of movement into adjacent areas, as well as reducing the possibility of non-tribal members crossing traditional boundaries.[74]

It is thus suggested that the ability of local populations to survive under rigorous environmental conditions requires a certain flexibility in local organization which allows for maximum mobility of exploiting groups, and a means of linking vast numbers of individuals and groups through sections and subsections into a network which permits small mobile units to spatially expand and contract under varying environmental conditions. Consequently, one of the functions of sections and subsections may have been an economic factor in allowing populations to “insure” a vast territorial domain from non-tribal groups, as well as “insuring” each local group’s future against economic hardship by permitting access into more favorable environments. On the coastal regions of Australia, subsections were not necessary, but in the central arid areas, sections and subsections linked spatially and socially distant groups together for mutual interaction and mobility. A somewhat analogous situation is reported by Sahlins (1958, p. 236) for the Polynesian atolls. The main problem on the atolls is the survival of a dense population in an area of sporadic or limited surplus production. Since economic specialization can not be sustained, every individual is interconnected by some alignment principledescent, age grades, residence-to every other type of socioeconomic group. Thus, everyone shares in the distribution and consumption of what is produced. Sahlins (1958, p. 245) postulates that the development of a number of interlocking social groups, each dedicated to exploitation of different resources, may be expected on atolls as adaptations to selective pressures limiting surplus production. In aboriginal Australia, one may regard the interlocking of kinsmen through section and subsection groupings as allowing for greater interaction and mobility over exploitable lands and thus “insuring” utilizable resources.

21. Australian Marriage, Land-Owning Groups, and Initiations

Frederick G. G. Rose

Introduction

At the symposium on Man the Hunter we discussed the Australian land-owning groups and marriage arrangements. Except indirectly, initiation found no place in the discussion, let alone its relevance to the relations between social groups, marriage, and resource use. The purpose of the present contribution is to show that marriage, initiation, and the relations between the land-owning groups are mutually interdependent and that an appreciation of the reciprocal relations between them takes one a considerable way to understanding the organization of aboriginal society as a whole.

Marriage and the Land-Owning Group

Because our views are conditioned by our own experience, we tend to think of marriage arrangements among, for instance, the Australian aborigines as essentially a means of regulating biological relations between men and women necessary for the propagation of the species. This notion has little relevance to the Australian aborigines because they do not or did not know the connection between sexual relations and the birth of children. It is not intended to argue here for or against the evidence on the aborigines’ ignorance of the facts of life reported by various field workers. It is merely intended to cite a piece of indirect evidence in support of this alleged ignorance of the aborigines.

An ever-recurring theme of cult life and of art throughout Australia, north, central, and south, is the fertility of nature and of man. In the north of Australia, however, probably through the contact with the Indonesians, and possibly other Asians before them, the aborigines learned of the connection between the birth of children and sexual intercourse, although they retained in the background as a kind of theology the usual Australian concept of spirit children. Anyone who compares the fertility cult life and concepts of the north with those of the center and the south will immediately be struck with the erotic form the cults and beliefs of the north have taken. A presupposition to the development of the erotic in the fertility beliefs is clearly a knowledge of the significance of the sexual act for the birth of children. We are convinced that this knowledge was lacking in the south and center and until a few centuries ago was probably also lacking in the north.

For the aborigines, therefore, a woman was a being who throughout the period of her life from puberty to the menopause regularly produced children, and in the eyes of the aborigines the man with whom she was living had no relevance to the biological facts. A woman who did not produce children was an anomaly in the aborigines’ scheme of things.

The relation between the man and the woman in marriage, from the viewpoint of aboriginal society, was essentially a reciprocal economic one, the man primarily providing the meat and the woman the vegetable foodstuffs for themselves and her children.

If this economic element is the essential one In aboriginal marriage, to understand its dynamics it is necessary to consider two other aspects of their life which reciprocally influence one another and, in their different ways, have a bearing on marriage. These two aspects or institutions are, first, the relationships between the land-owning (i.e., the political) groups or clans and, second, the initiation of the males: both are the obverse of the coin to the marriage arrangements or, perhaps better expressed, the other two sides of the triangle.

A generation ago it was considered almost as axiomatic that marriage arrangements were determined by kinship relationships, perhaps reinforced-in a somewhat obscure way-by the so-called “marriage classes,” where these existed. Thus we were told that the Aranda- type of kinship system was the most prevalent in Australia, in which a man married his mother’s mother’s brother’s daughter’s daughter (who at the same time was the other three second cousins), and if he married anyone else, then it was considered an “exceptional” marriage. This all made for excellent “modeling” and kinship algebra but got us little nearer to understanding why aboriginal society was integrated the way it was.

However, the main determinant of aboriginal marriage is not the kinship system or the “marriage classes.” Rather, each individual marriage is determined ultimately by the reciprocal rights and obligations existing between the actual political land-owning groups and to a less extent the reciprocal rights and obligations between members of these different landowning groups. These reciprocal rights and obligations aggregate into an extraordinarily complicated network linking groups and individuals throughout the society.

The kinship terminology-at least as far as the Australian aborigines are concerned and whatever might be true of other societies—is primarily an expression of these rights and obligations between individuals as members of groups, and does not express blood relationship (real or fictive) which we social anthropologists in our superior wisdom and in our categories of thought have arbitrarily imposed on it.

What is more, the network of which we have spoken is continually changing in the temporal dimension. The lasting service of Hart and Pilling (1960) is that they were able to show just how complicated this network was, and what factors ultimately determine a marriage.

It is, of course, much simpler to determine the kinship structure of an aboriginal society than to determine what are the economic, diplomatic, political (or call them what one will) relations between groups and individuals. As far as we know, Hart and Pilling’s work, is, of its kind, unique, while there are scores of accounts of Australian kinship systems. As a consequence it has become almost a dogma that kinship was the ultimate determinant of marriage.

If it is ultimately the economic relations between land-owning groups that determine what marriages take place, how are these relations manifested? Here again we have to clear away another dogma. The land-owning group in the drier, near-desert areas (perhaps it would be better to speak of the waterhole-owning group) is an abstraction, for it never comes together except, almost incidentally, when very large numbers of aborigines meet. Although an abstraction in the sense mentioned, it is nonetheless real. In Australia it is the fundamental political unit, deriving its power from the common ownership of the land[75] (or more exactly the game that lives on it, and the vegetable foods which grow in or on it) which is the most important means of production. It was early recognized that the land-owning group was exogamous and that the men of the group received their wives from other groups and, in their turn, the women born into a land-owning group went to other groups at marriage. However, it was assumed that the men of a landowning group worked together as a foraging group, horde, or band on their own land, except on rare occasions when they visited other land-owning groups. In other words, we had a virtual identification of the foraging group with the land-owning group.[76] This is an incorrect conception and one that can easily be shown to be incorrect when the members of foraging groups are classified according to membership in land-owing groups (Hiatt, 1962, Chapter 10 and 18, this volume). The foraging group is an extremely variable entity not only in respect to size but also in respect to land-owning group membership.

It is variable in respect to size mainly because the availability of food in any area is extremely changeable, not only between seasons but also from year to year. At a particular time of year, when food resources are scattered, the foraging group breaks down even to the individual family; at other times, when food is abundant in a small area, the groups come together to form aggregations of two hundred or more aborigines at a single camp. These large groups persist sometimes for weeks or even months at a time (Meggitt, 1962; Strehlow, 1967). Abundant food in a limited area was clearly a prerequisite for the carrying out of large-scale initiation and increase rituals.

The foraging group is variable in respect to male membership of land-owning groups because, given low population density, an estate belonging to a particular land-owning group simply cannot provide continuously for twelve months of the year, year in and year out, sufficient foodstuffs for its members. The land of one group may provide an abundance of food during the spring, for example, while another may provide an abundance of food during the winter. This leads to reciprocity between the land-owning groups. Inevitably, there are seasonal movements of population from one area to another. The reason for this is ecological and is determined essentially by distribution of resources.

A man can hunt on his own land. He also has certain rights to hunt on the group-land of his mother and of his wife, but in order to hunt on the land of other land-owning groups, he has to be invited or to obtain permission.[77] It is the sum of the obligations a man incurs by being allowed to hunt on another group’s territory, and the rights that he possesses to allow men of other groups to hunt on his territory, that constitutes the network of obligations and rights mentioned earlier as being the essential factors in determining marriage arrangements. The land-owning group does not, in fact cannot, live by itself; it usually has some kind of economic, political, or social relationship with some ten to fifteen other landowning groups. This would appear to be the mean radius, under Australian hunting-collecting nomadism, that a man can extend his contacts. Because of the social contacts existing within this radius, a certain language and cultural homogeneity develops, and the sum total of land-owning groups existing within this radius is often spoken of as a “tribe,” although there is in Australia no political organization at a “tribal” level. The “tribe” numbers some 400 or so aborigines on the basis of a mean membership of thirty for the land-owning group.[78]

One often reads in the literature of exchange of wives between individuals or between groups. While this occurs, of course, the actual exchange of wives is only one manifestation of a much wider reciprocity between land-owning groups. However, because it is a phenomenon that can easily be observed-- r logically, although not necessarily correctly, deduced from observation (cf. Rose, 1960b, p. I 19)—it is reported as if it were independent of the wider reciprocity.

Sahlins’ Original Aifiuent Society,’ A Comment

At this point a digression will be made to follow a train of thought stimulated by Marshall Sahlins’ “Notes on the Original Affluent Society” (Chapter gb, this volume). If Sahlins is right and those like Tokarev (1g61, p. 683) who conceive of the hunter-collectors as living continuously on the borderline of starvation are wrong, why are the Australian aborigines so preoccupied with the fertility of man and nature? There seems to be a contradiction here. If they need to spend only three or four hours a day hunting or collecting, there is obviously plenty of foodstuffs available and increase rituals for kangaroos (or other foodstuffs) just do not make sense. Why was there also a preoccupation with human fertility? Were the women peculiarly sterile?

Quantitative records such as we have of remnants of tribes on the fringes of civilization, where the aborigines were riddled with disease and were living under bad sanitational conditions and with inadequate diet, show that the women were indeed peculiarly sterile.But such detribalized aborigines were in the process of dying out. What was the situation under tribal conditions unaffected by the deleterious influences of civilization? Were the women peculiarly sterile here, so that it was necessary to invoke magical or religious means to increase the population and so prevent the society dying out? Records of aboriginal (or other hunter-gatherers for that matter) birth rates under traditional conditions are almost nonexistent. Although the birth rate, as such, is not record6d, the writer’s data (1g60b) on a tribe that was still living under traditional conditions indicate how high the birth rate must have been. Far from being peculiarly sterile, the women were quite the reverse. If a woman lived the full span from puberty to the menopause, she had at least eighteen children. Surely this stands in logical contradiction to the aborigines’ apparent obsession with human fertility? One might expect a greater interest in birth control! But on Groote Eylandt, where the writer worked, there was no birth control and little infanticide was practiced.

Sahlins is, nevertheless, right for Groote Eylandt subsistence: for the greater part of their time they only needed to work three or four hours a day. There was an abundance of meat and an abundance of vegetable foodstuffs, and the population certainly did not starve in spite of the women’s high fertility.

Let us examine more closely the second aspect—the maintenance of the population. It was stable and an extraordinarily high birth rate implies in these circumstances a very high death rate among infants and small children so that only a fraction reached the reproductive age. This was, indeed, the situation among the Groote Eylandt people. The writer estimated that at least 60 per cent of infants died before they reached their first birthday and there was a correspondingly high death rate among young children. In these circumstances the aborigines’ concern with the fertility of women was well founded.[79]

But why the concern with the fertility of nature if there was plenty of game and vegetable foods? The aborigines know from their own experience, if they are older men, or from what they have been told if they are not, that in the past there have been times when the land was gripped by drought so that the kangaroos became scarce and thin and the yams were not to be found. The people then were truly faced by starvation. The aborigines’ prime interest in the fertility of nature is to maintain the status quo where there is an abundance of kangaroos and yams, and above all to avoid a recurrence of the situation when the land was gripped by drought. They also know that in spite of all the magical and religious means they employ in their increase rituals and no matter how beneficently nature may be smiling on them today, perhaps next year, the year after, or in five or ten years the land will again be visited by drought and once again they will face starvation.

At the low technical level of hunting and gathering, man is a parasite on nature. Apart from his meager equipment, the attribute he possesses which prevents his own population numbers being subjected to the violent fluctuations of the other fauna in his environment is his social organization and the reciprocity and co-operation it implies. Most of the time the need for cooperation and reciprocity, specifically of the political land-owning groups, is not immediately apparent; but when drought or other natural catastrophe occurs, cooperation and reciprocity in the division of the few available food resources on the tracts of land owned by the various groups is a matter of life and death for the community as a whole. It is only at these times of food shortage that infanticide has the sanction of the society. It is better to destroy an infant or young child whose chances of survival are small anyway than to hinder the mother unnecessarily in her task of food collecting. When times are better, she will breed again. The belief in spirit-children gives her the solace that her dead child’s spirit will return to the store of spirit-children;lwaiting to be reborn.

We have digressed somewhat from the main theme of this contribution to resolve the apparent contradictions implied by Sahlins’ paper. By doing so we have underlined how cooperation and reciprocity is a matter of life and death for aboriginal society.

Before dealing with the question of male initiation-the third side of the triangle-we shall discuss what we have called “marriage gerontocracy” in Australia. This was the limited subject of the writer’s paper read at the Chicago symposium and the following with suitable additions and comments arising from the discussion largely follows what was presented to the symposium.

Marriage Gerontocracy in Australia

What do we understand by Australian marriage gerontocracy?[80] Eylmann in a few words gave a good description.[81] He wrote that “marriage relations of the Australian Aborigines are remarkable in that an eighteen to twenty-five year old man often has no wife, or if he does she is old enough to be his grandmother, while the oldest, most influential men in the community, on the average have most wives, amongst them not infrequently very old women and girls in their early teens” (1902, p. 13 I).

Although probably the best brief statement of the matter, we^ would not entirely agree with Eylmann’s description. For instance, we would not say that in traditional conditions it was the oldest men who had most wives. Rather, it was the older men in their forties or fifties. Eylmann’s statement that it was the “most influential” men who had most wives would support this viewpoint.

From what he wrote, one could draw the conclusion that girls were married at puberty. It is our observation that under traditional or near traditional conditions a girl went to live with her first husband when she was considerably younger-about 9 years old, with intercourse delayed until she had reached puberty. Going to live with her first husband should not be confused with the fact that a girl was promised to a man when she was much younger or even before she was born.

With these minor strictures the situation that Eylmann describes has some important consequences. We see that the young girls go as wives to the older men while the old women can be wives of men of any age. One can say that the older men have a monopoly of the younger women, while also having some of the old or older women. The normal population pyramid of an aboriginal society-as, of course, in our own society-shows that there are considerably more females in their teens and twenties than men in their forties or fifties. As an obvious consequence these older men are married polygynously. This frequently reported fact deserves mention here since polygyny is a corollary of gerontocracy and it would perhaps be better to speak of “polygyny-gerontocracy” and not merely of “gerontocracy. ”

Another logical consequence ofthe situation described by Eylmann, but one which is more rarely reported, is that a woman in the course of her lifetime has several husbands. Hart wrote for the Tiwi of Melville and Bathurst Islands that “while a woman goes to live with her husband at about thirteen, a man is lucky to get his first wife under the age of 35. I t follows that in the great majority of cases the father of children dies long before the mother. Therefore ...a woman passes through the possession of quite a number of different men during her lifetime” ( 1930, p. 282). On Groote Eylandt a woman had at least four husbands in the course of a full lifetime. This number is quite apart from clandestine sex relations. By the same token a man had at least four wives—for the most part polygynously-in the course of his lifetime. Marriage in Australia is thus seen to be an extraordinarily fluid institution. Although at any one time polygynous unions are in the minority, most men who lived to old age would spend part of their lives married poly- gynously. Similarly a woman lived at least part of her lifetime inside a collective of co-wives. In this sense we can say that polygyny is the rule and monogamy the exception.

Before discussing the attempts made to interpret this fluidity of Australian marriage, something can be said on the distribution of marriage gerontocracy inside Australia. The evidence is at times ‘ contradictory; nevertheless it would appear that gerontocracy was a feature of marriage throughout Australia and Tasmania (Rose, 1960b, pp. 91–99, 234–36, 482). It is a remarkable fact that some observers with wide experience of aborigines under traditional conditions did not report gerontocracy in areas where it is known definitely to have existed. For instance, in his monograph Economic Structure and the Ceremonial Exchange Cycle in Arnhem Land (1949), Thomson does not once mention gerontocracy although it clearly would have had considerable bearing on the subject with which he was dealing. Similarly Sir Baldwin Spencer (1914) does not mention it among the Tiwi. One can explain these omissions by the assumption that these particular observers had become so used to gerontocracy as part of the fabric of aboriginal society that they simply ignored it. In the writer’s own case it was not until he examined the actual ages of man and wife in the Groote Eylandt family units that he realized how pronounced it was. Without the actual age data he would also probably have ignored it, although, of course, in general terms he was quite aware that in Australian societies a man tended to marry a woman considerably younger than himself.

The evidence for polygyny-gerontocracy is further complicated by the fact that it can disappear very rapidly under contact situations. A report of its absence, or the fact that the aborigines married monogamously, by no means disproves the existence of polygynygerontocracy under traditional conditions.

Elkin (1954, p. 127 note) has noted its disappearance under contact situations, and the writer has shown that it is particularly the change in the economic activity of the women under such conditions that leads to its disappearance (Rose, 1965b, pp. 75–83 passim). As a consequence, in evaluating the older writers it is necessary to determine under what contact situation the observations were made. Taking the reports at their face value is unsatisfactory; a certain historical research is also necessary.

In spite of the popular opinion that Australia is virtually a land of near deserts and steppes, the ecological conditions are, in fact, extremely varied. The universality of polygyny-gerontocracy is surprising and would indicate an extremely low degree of specialization in the sbcial organization to meet the particular ecological situations.

Interpretation of Marriage Gerontocracy

Let us now return to the question mentioned earlier, the fluidity of the marriage relation as the result of polygyny-gerontoeracy. A problem which has occupied tqe writer since he first described in statistical terms the polygynygerontocracy of the Groote Eylandt people and its universality in Australia has been finding an interpretation of the custom.

A perfectly rational explanation is, of course, that it is “natural” for the older, most influential men to monopolize the younger and more desirable women to the exclusion of their younger clansmen (Fison and Howitt, 1880, p. 355). But a more objective explanation was sought; that is to say, a reason that was independent of the will or wishes of the individual. The explanation mentioned appeared to stand the issue on its head. To explain the power of the older men to monopolize the women in terms of its being “natural” for them to do so, was, in fact, no explanation. It was unsatisfactory, first, because it did not explain how the older men obtained this power and secondly, because, as had frequently been observed, polygyny-gerontocracy disappears extremely quickly once the economic basis of the society changes, apparently without any effective resistance from the older men. This latter consideration would seem to point to a material, possibly ultimately a relatively simple economic explanation of polygyny-gerontocracy.

On the basis of his own data, the writer propounded an hypothesis which ran generally as follows. The women tended to aggregate themselves in collectives of co-wives around the men at the peak of their productive capacity and this tendency reached its maximum when the women had their greatest child-bearing and child-rearing burdens.[82] It will be immediately noted that in this explanation it was the women and not the men who took the active role to establish polygynous units, and that this type of family provided the optimum conditions for the rearing of the younger generation inside the collectives of co-wives of the polygynous families.

The writer went further and suggested that the men were at the peak of their productive capacity when they were in their early forties and the women had their greatest burdens to bear when they were about 24 (Rose, 1960a; Ig63)^

Unfortunately for this neat theory, Ashley- Montagu (1957) has provided certain biological facts that showed that the women’s maximum burden in caring for children must have occurred when they were considerably older than 24. Then Meggitt (lg65b) produced comparative figures on polygyny-gerontocracy for the Walbiri. On this particular point his results did not confirm the original hypothesis, and Meggitt sought a “multifactoral” explanation of maximum polygyny; Nevertheless, it would seem that the difference between the writer’s and Meggitt’S conclusions on this point are more apparent than real. While the explanation of maximum polygyny may, as Meggitt expressed it, be multifactoral, the writer is not convinced that the essential explanation of polygyny-gerontocracy in Australia cannot be given in the general terms originally formulated.[83]

Initiation

We can now return to the question of male initiation and its relation to marriage and to the cooperation between the land-owning groups.

The discussion above on marriage gerontocracy has shown that the Australian aboriginal society is on the horns of a dilemma. On the one side there is the tendency for the women to aggregate themselves around men at the peak of their productive activity with this tendency reaching a maximum when the women have their greatest maternal burdens to bear. This situation leads to the older men- past their physical and sexual peak-monopolizing the women in polygynous units. On the other hand, the younger men-at their physical and sexual peak-are virtually without women, or at most with a woman old enough to be their grandmother. Economic necessity, expressed as a need to provide collectives of co-wives in which to raise the infants and young children, comes into conflict with the sexual drive of the younger men who are denied women. This contradiction in Australian society has been resolved by the older men controlling the younger through the elaborate system of male initiation found in Australia.

It would take us too far to consider the various forms initiation takes, but even though the forms vary considerably, they all have certain features in common. The first is that the full series of initiation covers a span of a young man’s lifetime from before puberty until his middle or late twenties.

The second feature is that the initiate cannot form an economic alliance for the foundation of a family with a woman (i.e., marry her), until he has completed his initiation.

Third, the initiate is instructed by didactic methods both in the ecology of the country in which he lives and in the art of obtaining the wherewithal of life.He ahsorbs this knowledge by learning the tribal myths and by witnessing the cult ceremonies centered on the wanderings and creative feats of the heroes of the Dreamtime.

Fourth, the initiation of a youth is not simply a domestic matter of his own land-owning group, but is an issue of intergroup importance. Relatives who can and frequently do play an important role in a youth’s initiation are his mother’s brother, his male patrilateral crosscousin (who is sometimes his sister’s husband), his sister’s husband when he is not the initiate’s patrilateral cross-cousin, and a representative of the land-owning group from which he will later obtain a wife.All of these relatives belong to land-owning groups other than the initiate’s own, and each of these groups in its own way has a vested interest in the youth’s initiation.

Fifth, during part of his initiation extending over some years, a youth is attached to, and is under the strict control and tutelate of, a guardian, a considerably older man. This guardian is frequently the youth’s sister’s hus- band.This relationship has definite advantages for the guardian.He not only has the responsibility of continuing by practical example the instruction of the initiate in the hunting of game and the other manly arts, but he also has rights to dispose of what foodstuffs the initiate brings back to camp. The guardian’s interest in controlling an initiate is not greatly dissimilar to his interest in controlling the initiate’s sister as wife. It is indicative of the relationship existing between initiate and guardian that on Groote Eylandt the missionaries frequently referred to the initiate as a “boy slave” or alternatively as a “boy wife.” Stealing of young wives from their husbands was a feature of traditional life on Groote Eylandt, and it is significant that initiates were stolen in the same way and as frequently from their guardians (Rose, 1960b, p. 73).

Sixth, the initiate is secluded for long periods in the bush, during which time he has to fend for himself. Moreover, he is obliged to travel over and live on the country of several landowning groups other than his own.

Conclusion

We have seen how the preoccupation of the aborigines with maintaining human fertility has a real material basis. In order that adequate population replacements survive the dangerous first decade of life, women with infants and young children to care for, tend to live in collectives of co-wives (i.e., in polygynous units) centered around experienced men at the peak of their economic production. This economic necessity is in conflict with the fact that the younger men at the peak of their physical strength and sexual activity have virtually no women available to them. The contradiction is resolved socially by the control exerted through initiation of the young men. To this extent marriage and initiation are complementary.

We have also seen how the preoccupation of the aborigines with maintaining the fertility of nature has a real material basis. The very survival of the society depends on the reciprocity and cooperation between the political units-the land-owning groups-in the sphere of food resource utilization. Similar1.y, the survival of the society depends on the maintenance of the type of marriage and initiation described, and the reciprocity and cooperation of the land-owning groups are also manifested in this sphere.

These two spheres of reciprocity and cooperation are not two separate categories for the aborigines as they may appear to be to the ethnologist; they form an integrated unity. This is achieved by the marked sexual division of labor found generally throughout Australia -the woman normally collecting the vegetable foodstuffs while the man is responsible for providing meat. The giving and receiving of females in marriage is not merely the reallocation of a person who will produce babies that will eventually help to maintain the group’s numbers, but it is also the reallocation of a gatherer of vegetable foodstuffs who plays an important part in food resource utilization.

The giving of a girl in marriage implies that her husband and her children have a virtual right to utilize the resources of her natal group’s land. It is therefore natural that the girl’s fellow clansmen have a great interest in her future husband and thus assume certain responsibilities in his initiation.

It is on the above basis, with all the contradictions it involves, that the whole superstructure of aboriginal cult and religious life has been logically erected by the aborigines. A discussion of this superstructure and its connection with the basis was, however, not the purpose of this contribution.

The method used here’ at first glance may appear to be a functional, nonhistorical interpretation of various institutions of aboriginal society in the classical sense that they are described as a functioning whole. What, in fact, has been described is the interconnection between certain relations of production, in the historical materialists’ sense of the expression. The historical materialists’ view is that the relations of production—and essentially, as we have dealt with them here, marriage, initiation and the land-owning group relations as such—are determined by the forces of production. The development of these forces of production is the motor of human history. Although it was not the purpose of this contribution to examine the development of these forces of production, it can be remarked that, in so far as the Australian aborigines are the least developed of any known people, they occupy a key position in our understanding of earliest human social development. To elaborate on this theme would take us far afield. However, it has been dealt with cursorily elsewhere (Rose, ig6oa).

22. Discussions, Part IV

22a. Gerontocracy and Polygyny

Speakers: Bicchieri, Bulmer, Gardner, Isaac, Pilling, Rose, Slobodin, Whiting, Woodburn

ROSE: I would like to make one point about the possible significance of gerontocracy. It is interesting that gerontocracy is found throughout Australia and in Tasmania, and I think the question must be asked whether gerontocracy/ polygyny was not also possibly a characteristic of the Paleolithic.

WOODBURN: I think it is essential that we look for a term to replace “gerontocracy.” This has been with us for some time, but I do not think it is too late to make a change. The term literally means “government by old men.” Now, in the first place, we are dealing with men aged 30 and up. The second objection is to the connotation of political power. In fact, we are dealing with differential age at marriage and not necessarily with the implication that political power is involved.

SLOBODlN: To respond to Woodburngerontocracy may or may not be applicable to an ethnographic situation on empirical grounds, but the word does not necessarily, or even usually, mean “rule by the aged.” The morpheme -cracy is from Greek kratein, which means “to rule,” but also it more generally means “to exert power,” from kratos, “strength.” Surely those Australian 42-year- olds exert power!

WHITING: In a cross-cultural study of polygyny and its correlates, I found that wherever polygyny occurs the ages at marriage are about 25for men and 15for women; while the averages in monogamous societies are 19 for men and 17 for women. Now, these statistics hold fairly constant irrespective of the presence or absence of gerontocracy; polygyny alone is the variable that accounts for some of the difference in ages.

Although polygyny is not directly related to hunting and gathering, it does seem to be related chiefly to nomadism, where the task of carrying children makes child spacing necessary. And it is most frequent when postpartum sex taboo is the means of child spacing. If the postpartum sex taboo is two or three years, the man takes a second wife to help carry the burden of the child and for purposes of sexual access.This leads to enforced bachelorhood for the younger men, thereby glVIng rise to the idea that polygyny and gerontocracy are correlated.

GARDNER: However one can have geronto- gamy without gerontocracy. The Paliyans had gerontogamy of both sorts with both older males and older females. Very frequently this evolved without ceremony out of adoptive parent-child relationships. The child, when attaining freedom from the authority of anyone else (at about 9 years of age), quickly slips into the spouse’s role. The point is that there is no authority difference between spouses nor any use of marriage as a device for obtaining authority, and the suffix “-ocracy” would be most inappropriate. In reply to Whiting’s point I might add that the combined incidence of polygyny and polyandry is under 5 per cent, while 10 per cent of the Paliyan marriages (and a different 10 per cent) are gerontogamous. It would be impossible in this case, and perhaps in others, to call gerontogamy a simple result of polygamy.

ISAAC: I would like to ask whether it is normal for a man’s food-productive capacity to reach a maximum as late as 42 years of age among hunting people.

ROSE: I did not have any direct evidence for the specific age of 42, but I found that there was tendency for a woman to marry a man of 42 (or 41, or 39) in Australia, quite independently of her own age. (There is a correlation coefficient approaching zero between the ages of husband and wife in Australia, while, of course, in our own society it approaches one.) And I assumed that the reason for this was that the man was at his peak of production. I did not try to measure production, however.

PILLING: I think it would be fruitful to review Dr. Rose’s material with this question in mind: Are the figures the same when a woman is marrying a man who already has a wife as when she is marrying a single man? I suspect that the motivation for polygyny may be less the sexual gratification of the male than the carrying of co-wife’s burden during intensive nursing or late pregnancy, when a man’s wife is less productive as a gatherer.

RosE: My data on that areavailable and can be examined from that viewpoint (Rose, 1960b). Strehlow gives some data on this particular issue and makes the point that a man often takes another wife when his first wife is in an advanced stage of pregnancy.

BULMER: I am a little puzzled by our new mystical number for peak production in a man’s early forties. I can easily understand that he might be at the peak of his political capacity at that age, but it would be hard in any society to demonstrate that his economic capacity was greatest at such an age. If, in fact, political or social influence is at stake, this would seem to hinge back upon marriage arrangements and create a rather circular system.

BICCHIERI: I would like to pursue Dr. Bulmer’s reaction to Dr. Rose’s idea that man is at peak production in his early forties. This is a case in which it is not desirable to separate economic from political and social factors. Social organization can be regarded as a technological mechanism, so a man of 45- while not in his physical prime-may be able to manipulate other people and add a share of their production to his own, gained by physical manipulation of the environment. So 45 nlay be an ideal age in the sense of peak ability to coordinate productive activities.

22b. Gidjingali Marriage Arrangements: Comments and Rejoinder

[Note: This exchange of correspondence between Levi-Strauss and Hiatt was received by the editors after the symposium on Man the Hunter.]

LEVI-STRAUSS: It is quite natural that an Australian scholar should ignore that in the French anthropological usage, the word realia (Levi-Strauss 1966, p. i 15) faked after the Latin, does not mean “what really happens” as he believes, but material objects of the kind unearthed by archeologists such. as buildings, statues, coins, tools, and weapons. Besides, when I suggested that new techniques such as pollen analysis and carbon- I 4 may improve what Radcliffe-Brown once called “conjectural history” I was certainly not referring to kinship systems. Thus in two instances Hiatt’s poisonous darts fall widely off the mark and I am afraid that the same can be said, for the rest of his paper.

I have not made any claim to interpret the kinship system and marriage rules of the Gidjingali of which I had never heard before and I do not know either how they themselves explain their marriage rules. I know, however, in which terms informants from other tribes have explained their system to men such as Spencer and Gillen, Strehlow, RadcliffeBrown, Sharp, Stanner, Elkin, Warner, and many others and I have tried to articulate these scattered testimonies in a meaningful whole. If Hiatt’s recent observation of what is left of a collapsing Australian tribe should carry more weight than the bulk of the older literature, then let us burn the books. It would probably be more sensible to seek in demographic and other factors wh.y the Gidjingali cannot adhere to their own norms. I agree that it is important to find out what really happens in any society, notwithstanding its proclaimed rules. However, my past work on kinship has been mostly concerned with a different problem: to ascertain what was the meaning of the rules, whether they be applied or not. For rules have their own life if only in the mind of the people who are able to state them and to translate them back and forth at will in the language of preferred degrees or in that of sections and subsections. This should suffice to convince us that these two languages are not as wholly disconnected as it has become fashionable to pretend. Finally, it takes more than mere naivete to believe that since the theoretical model of an Aranda system requires only four patrilines, the system itself could not ensure marriage exchange between 28 clans or many more. This empirical complexity has become commonplace at least since the time of Strehlow, and the Gidjingali case teaches nothing new in this respect. For the network can be woven on a shifting basis where the four lines do not need to remain the same ones. Unless the Gidjingali marry their daughters or sisters, which apparently is not the case, my statement that “in human society a man must obtain a woman from another man who gives him a daughter or a sister” (the French wording was more elaborate: “qui la lui cMe sous forme flUe ou de soeur”) holds true for them too. I never claimed that this was happening without intermediaries and I even took pains to show that a delayed exchange remains an exchange none the less.

HIATT: If Levi-Strauss is putting forward a general theory of Australian kinship (as I have always supposed), then the fact that he has never heard of the Gidjingali until now is of as little account as the fact that Newton was never hit on the head by a falling paw-paw. His insinuation that the Gidjingali do not count because they are the remnants of a collapsing Australian tribe must also be rejected. I have already dealt with a similar objection made by Birdsell to my criticisms of Radcliffe-Brown’s generalizations about aboriginal local organization (Chapter 17a, this volume). Levi-Strauss supposes that some observers were in a better position to describe traditional aboriginal social life simply because they were earlier on the scene. This ignores the fact that European settlement has advanced slowly over the continent and that anthropologists have regularly worked on the frontiers, never beyond them. The Gidjingali system of kinship and marriage was, at the time of my observations, in no worse state of dislocation than those studied by the investigators Levi-Strauss mentions.

Levi-Strauss suggests that it would be more sensible to seek in demographic and other factors an answer to the question why the Gidjingali cannot adhere to their own norms. This is a distressing statement. It indicates the great difficulty I shall have in conveying the distinction that exists among the Gidjingali between marriage rules stated in terms of broad kin categories and rules that specify in narrow genealogical terms particular women within these categories to whom a man has marriage rights. Anthropologists in the past have concentrated upon rules of the former type, and it is upon their material that Levi- Strauss has based his analysis. I must insist that in this respect the Gidjingali are not aberrant, and I repeat that roughly go per cent of their marriages are in accordance with the rules. The nonconformity referred to by Levi-Strauss occurs in relation to rules of the second type, and I have already spent a good deal of time trying to explain it in precisely the terms Levi-Strauss recommends. I regret that my effort to do so was unsuccessful.

Levi-Strauss says that it has become fashionable to deny that rules stated in the language of preferred degrees and in the language of sections or subsections can be translated back and forth by the aborigines at will. If by “preferred degrees” he means categories of kin with whom a marriage would be morally proper, then it is clear from my paper that I do not subscribe to the fashion. What I deny is that a marriage rule stated in terms of kin categories or subsections is equivalent to a rule defining a man’s right to only certain women within these categories or subsections.

I apologize for failing to grasp the meaning of Levi-Strauss’s fake Latin and for assuming that he had actually said that one cannot study social rules independently of what really happens. The fact remains, as his final remarks indicate, that he does believe that his analytical models have empirical counterparts; in particular, that when he describes kinship systems as systems of communication in which women are mediators, he means that actual women are being exchanged by actual social groups. I have argued that there is little evidence of either a statistical or ideological kind to support such a view in relation to the Gidjingali. In my paper at this conference I have compared the facts with a hypothetical model of four patriclans; elsewhere I have tried to demonstrate that there is even less evidence for an extended network of exchange (Hiatt, 1965, p. 133).

I am baffled by Levi-Strauss’s final point. I said that Gidjingali men exchange their sister’s daughters but do not have rights to bestow their own daughters or sisters. Levi-Strauss insists that they do exchange daughters or sisters, not directly hut through intermediaries. Suppose for the sake of argument that we were able to discover cycles of the postulated kind. We might then say: “Gidjingali men exchange their sister’s daughters. They also exchange their daughters or sisters, but they are quite unaware of it.” The question is whether this would be a meaningful use of the word “‘exchange.’”

22c. The Use and Misuse of Models

Speakers: Alland, Helm, Lathrap, Laughlin, Uvi- Strauss, Meggitt, Pilling, Schieifdin, Sharp, Whiting.

ALLAND: I would like to address myself to marriage and residence rules. We find rules that do not work or are misapplied, rules that do not exist, and variations on rules that might work in some way or other. In short, a model is not a model. It is possible, I suppose, that mechanical models could be completely false a joint creation of the anthropologist and a zealous informant. It is more likely, however, that the models which some anthropologists have extracted from informants do exist, but the researcher’s application of the models may be based on a misunderstanding of what they stand for.

What of the existing variation, for which we now have a good deal of empirical evidence? Is it capricious, or does it reflect scientific laws? I prefer the latter interpretation. First, we must consider those cases in which the demographic facts of life make it impossible for a people to follow their rules, and so they do the best they can. Lounsbury (1962) in his review of Needham’s Structure and Sentiment, suggests that “the best they can do” may, indeed, be good enough to maintain an ongoing system. Computer simulation study—particularly that published by Kunstadter on cross-cousin marriage (Kun- stadter et al., 1963) indicates that “the best they can do” certainly is not a replication of the model.

It seems to me that it is, at least, possible that some of the variations in residence and marriage reflect adaptations to changing conditions and are, therefore, only short-term fluctuations. Meggitt has, I think, suggested this in another context. These variations may reflect the operation of supporting variables acting to maintain a state of some system. The statistical model, then, may reflect a certain degree of resource management, including the allocation of men and women. Thus, a diachronic analysis of variation in relation to ecological variables becomes a vital area for research.

In my own analysis of residence patterns among the Abron of the Ivory Coast, I stated that a mechanical model of residence told me more about social structure than did a statistical analysis of residence frequency, in which irregular forms of residence (i.e., those that did not fit the model) would be counted along with predictable types. How might this view be reconciled with my suggestion above that statistical rules might reveal ecological relationships? I think there are two answers to this, either of which might apply in the proper circumstances. The two models reflect different aspects of the system. One is more ernie, and the other more etic (if one likes those terms).

  1. Traditional categories are imprecise and show no regularity, because they do not reflect an actual, ordered situation. But there may be some undiscovered order in the data. Thus, in the Abron material, I was forced to consider— as Fortes had done for the Ashanti (1949)- residence rules against domestic cycles. Martial residence alone was inadequate. Unlike Fortes, I then extracted rules that I considered to be significant for the social structure. Deviations from these rules were seen as accidental or temporary aberrations due to such minor factors as housing shortages.

  2. A statistical model may be linked into self-regulating systems, in which statistical variations can be predicted on the basis of changing values of some other variables, such as population size, age- and sex-pyramids, allocation of resources, etc. The mechanical model may, as other conferees have noted, support an ideology that may or may not be related to marriage and residence-even if informants say that it is--or it may support partial adherence to marriage rules that are enough to sustain a social system.

I would suggest offhand, as a hypothesis, that as technology grows in complexity (with the concomitant complexities in social systems), strategies of maximization related to direct confrontation between man and environment are complicated by strategies related to property relationships and other social aspects that flow from these. Stated rules then, may be violated less frequently among peoples with higher technologies than among huntergatherers, whose pragmatism is closely linked with the vagaries of the base environment.

LEVI-STRAUSS: There are a few remarks that I would like to make on the papers by Messrs. Hiatt and Meggitt. I do not feel that I have written anything original about the Aranda system; I utilized the existing literature and reasserted the interpretation given years ago by Durkheim, on the one hand, and van Gennep on the other. This is, namely, that the entire set-up of the system can be explained by a series of commutative operations between residence and clan names, so that a satisfactory interlocking of all units concerned will be assured, generation after generation.

I view with great interest Mr. Meggitt’s statistical correlations, but there is something that I could not find in his accounts; i.e., what the natives themselves say about their subsection system and its relation to marriage. Trying to explain such things merely by making statistical runs and trying to build up correlations seems to me to amount to very much the same situation as that of a man who, wishing to understand the rules of the game of chess, eschewed a treatise on chess in favor of statistical runs on hundreds of thousands of games played by clumsy players. Of course, the clumsy players would have made mistakes, and the analyst would have no way of distinguishing these from moves made in accordance with the rules.

My acquaintance with Australian studies comes by way of the literature, and I have been very much impressed by the fact that whenever there are verbatim accounts of what native informants say, it is perfectly clear that they can express their marriage rules in terms of sections and subsections. I do not mean that sections and subsections are nothing other than marriage classes. But the natives are able, if need be, to translate the marriage rules in terms of these categories. And it seems to me that the great originality of the Australian native thinkers is shown by their ability to devise a coherent system in which there is the possibility of translation at every level between the social organization, the marriage rules, and the terminology.

Now, I am also aware that Elkin discovered that there is not a close correspondence between the subsection system and the marriage rules and, further, that the subsection system expresses a great number of things other than marriage rules. Mr. Meggitt has said that the system expresses primarily the order of the universe and only incidentally expresses the marriage rules. I question this. Do the natives say this? Is it necessary that we establish a hierarchy of importance and exclude the possibility that in this case cosmology and marriage rules may be on a par with one another? Even if the section and subsection system expresses in some way the order of the universe, as conceptualized by the natives, and even if the marriage rules normally can be used to maintain and perpetuate the social order, it nevertheless seems quite obvious that if-for demographic or other reasons—the marriage rules do not work, it will seem more important to the people concerned to maintain the set-up than to follow the rules. This, I take it, is what Elkin has pointed out. Yet there is nothing ex- traordinaryabout this. Every society, including our own, would do exactly the same. What is important for us, as for the Australian, is the moral order and the representation that we have of the universe. And if the rules do not succeed in maintaining that order, then they may be ignored. This does not mean, however—speaking of our own society—that law, institutions, and constitutions are entirely negligible because we violate them every time it is useful to do so.

MEGGITT: For better or worse, the Australian aborigines are in many matters far more sophisticated than anthropologists. By their own reports, they have a hierarchical view of the universe and give primacy to the rules (such as subsection rules) that structure the universe. This is one model. This model is unchanging, as is the universe. ‘Where they differ from us is in the fact that they distinguish a second model for marriage arrangements, while we keep trying to impose the universal model on the marriage realm. Moreover, they point out that their marriage model is in general accord with what actually happens; and they take care to see to this, insofar as demographic and other factors allow them to. For them, mating is material (hence, a part of the universal model), but marriage is not. So they are able to maintain their marriage rules; even though they may be mating against these rules.

LEVI-STRAUSS: I would like to make a distinction here, borrowing an illustration from the natural sciences. There are shapeless mollusks living in the sea that secrete marvelous shells, which have mathematical properties. Now, it is a problem to discover how a certain type of mollusk secretes a certain type of shell. That is, if you like, a problem of what happens in the field, and it is a very important kind of problem. But it is perfectly legitimate never even to look at the mollusk, and to try to learn the mathematical transformations involved in passage from one type of shell to another. Either type of investigation is worthwhile, and they are entirely distinct from one another. So we may make use of both, but it is a mistake to mix the two operations or to suppose that we are doing one type when, in fact, we are doing the -other.

HELM: I cannot see, in the case either of a mollusk or of a human community, the legitimacy of separating the pattern or map and the actual behavior. I should think that it is not only permissible but altogether necessary that we keep our attention on the relations of the two. The manipulation of models in the abstract or the refusal to test them against events “on the ground” will never tell us to what degree, if any, societies are systems.

Alland: Ashby (1960) has said that “a system is anything that the investigator cares to define as one,” but I would add that the investigator has the responsibility of proving that the system is at least interesting. Some events which the anthropologist rather arbitrarily fits into a class of behavior, like residence, may be more significant than others, at least for some system that he has in mind. This is what I tried to do with the Abron data, and I think that in my treatment of it I went beyond Fortes’ treatment of his data on the Ashanti (1949). He wound up with a range of residence rules that showed that the Ashanti were not just matrilocal or dual-lo.cal, but that there was matrilocality, dual-locality, etc. I tried to relate this to a system of social relationships having to do with lineal stresses in the society.

Now, there are two things that interest me in rule systems when the rules systems are related to a single set of phenomena: ( I ) the question whether the natives try to approximate the rule system as we think that they might approximate it; and (2) the variations, which may be very significant—and not just capricious. So both rules and variations may be related to culture.

Whiting: It seems to me that the structure in the native’s mind that corresponds to what, in the anthropologists’ mind, we might call a model, is really better described as a cognitive view of the universe that serves him as a mnemonic device and an aid to communication with other members of his own society. Now, it is no more an empirical matter to find out who marries whom than it is to try to discover how much consensus there is in a society in the cognitive systems that they espouse. However the latter question is one which we have failed to investigate.

Alland: We have had a blush of success from componential analysis, but it has tended to go back to the single informant approach of earlier anthropology. This is unfortunate in many cases, although it seems to work pretty well with kinship systems, since most people know their systems pretty well and agree on them. But I think Whiting’s point is a good one. I remember Sapir’s story about the time when he had completely worked out a system and then presented it to one of the Indians, Two Crows, who told him that the system simply was not like that. There is this sort of variation in system, even ideal ones, within a single society; and it is significant to know how many “Two Crows” there are in any particular society.

LAUGHLIN: I recall an early paper by Steve Hart, published in the American Anthropologist (Hart, 1943) in the 1940’s, in which he analyzed the marriage rules for the Natchez Indians. As I recall, they had an exotic system that differentiated Suns, Honors, Nobles, and Stinkards. And by purely arithmetical means he demonstrated that the system could not work. Some years later, the mathematician Robert Busch demonstrated again, this time by means of matrix algebra, that the system could not work. This might make us wonder how well other systems would stand a test of internal consistency—quite apart from the question as to the conformity of the people to the systems.

LATHRAP: I think that the demonstrations just mentioned were made on the assumption that the system was intended to function as a closed one, while the Natchez and other groups with somewhat similar systems, such as the Mbaya of the Gran Chaco, were continually adding to their lower class through conquest. The details of the setting and the opnation of such a system must be taken into account.

ALLAND: It seems significant that this should be a model that cannot work unless something unusual happens, which seems especially important if we are studying, say, gene flow.

SHARP: There may be different models and different kinds of models that should have been considered here. The Australian aboriginal, however abstract his thinking may become, is not really concerned to distinguish between Dr. Birdsell’s linguistic community and Dr. Alland’s breeding one, which might include two to three linguistic communities. It is up to us not to be taken in by stereotyped con- structs-either the natives’ or our own-or to apply them to the wrong group or the wrong activities.

SCHIEFFELIN: While there seems to be agreement on the fact that the scientist’s model is not identical to that in the native’s mind, it has not yet been seriously asked whether the native’s construct is even the same sort of thing as the anthropologist’s model. We appear to assume that the native does live in accordance with some sort of system of explicit and implicit beliefs to be brought out by careful questioning; while it may well be that, in fact, the native synthesizes and systematizes his beliefs for the very first time in the act of explaining them to the anthropologist. It might be useful to look at the “model” in the native’s mind in the same way that Thomas Kuhn (1962) looks at paradigms in scientific theory. Instead of being strict models, according to which scientific theories are elaborated, they are bodies of assumptions that are more or less consistent with one another and include exemplary examples, so to speak.

Pilling: In view of the monographs published on Australia, I might stress the importance of graphic representations of “models,” which the aborigines often make in the sand (in circumstances which suggest that such graphic representations are not a new,, accul- turative phenomenon). In studying some of these, I found that they were of a very different nature from the anthropologist’s presentation of the same material. As ethnographers, we ought to be certain that we do not overlook the aid to be had from these graphic representations as supplements to the verbal explanations.

LEVI-STRAUSS: I would like to cite an example of the clarity of the model that sometimes is to be found in the mind of the native. One of my students has just returned from Africa, where she studied a Voltaic people with a kinship system of the Crow-Omaha type (but with bilateral descent and, accordingly, numerous clan interdictions). Now, the use of the terms is extremely complicated, and in order to explain why in describing a certain degree of kinship they use one term rather than another, the natives expounded a set of rules resembling very much those developed by Lounsbury in his analysis of the Crow-Omaha system (Lounsbury, 1964). Is it not striking that Lounsbury, by sophisticated techniques of analysis, should have actually rediscovered rules that do exist in the mind of a native in a quite different part of the world?

22d. Statistics of Kin Marriage: A Non-Australian Example

Speaker: Helm

HELM: The bulk of our discussion today has concerned marriage among Australian aborigines. It hardly needs saying that the Australians are not the only hunter-gatherers who marry, nor are they the only ones who exhibit cross-cousin marriage. To broaden the perspective it may be useful to consider marriage data from other areas.

I present part of the results of an effort to identify kin-pattern in marriages among the Hare Indians. The Athapaskan Hare live near the Arctic circle in western Canada and have numbered around 300 for the last several generations. In my few weeks with the Hare in 1957, I found Iroquois cousin terms, terminological equivalence of father’s sister and mother’s brother with mother-in-law and father-in-law, and the apparent fact that terms for cross-cousins of opposite sex derive from husband and wife terms. These forms suggested that, at least in the past, cross-cousin marriage might be practiced (MacNeish, Ig60a). In a subsequent study of the Hare, Hiroko Sue (Ig64, p. 424) indicates that first cross-cousin marriage was formerly the preferred type and that “more distant cross-cousins do marry each other. ”

In considering the Hare material, the marriage patterns and terminology systems of Northern Algonkians (Cree, Northern Ojib- wa, Naskapi) came to mind as immediate models (see Eggan, 1955, pp. 519 ff.). Hallowell and Dunning, for example, havc givcn us quantified data on kin-connected marriages in two Northern Ojibwa groups (with populations comparable in size to the Hare). Although at pains to point out that persons falling in the “marriageable” class may be distant relatives of Ego, Hallowell and Dunning demonstrate that consanguineal first and second crosscousin marriages are high. Such marriages comprised 40.8 per cent of three generations of marriages among the Island Lake Salteaux (Hallowell, 1937, Pp. 98, 101) and 50.0 per cent of the marriages taking place among the Pekangekum Ojibwa in a five-year interval in the 1950’s (Dunning, 1959, p. 154). The kinship terminology system of the Hare differs from the cited Northern Ojibwa groups only in that sibling-in-law categories are identified by distinct terms. In his comparative analysis of Algonkian terminology systems, Eggan (1955, p. 529) argues that “separation of siblings-in-law from cross-cousins is an important marker in the breakdown of cross-cousin marriage” and for several cases suggests “white acculturation” as a critical factor. If the distinct sibling-in-Iaw terms in the Hare system represent an acculturative shift, then a reasonable expectation, by analogy from the Algon- kian material, is that marriages in earlier generations at least would show a fair incidence of consanguine cross-cousin marriage of first and second degree.

Data on Hare genealogies and all marriages from the beginning of the Mission Period (that is, 1860) were available to me through the records of the Oblate missionaries. The marriages of almost 1,000 individuals were recorded. The genealogical and marital data were subjected to computer processing. Given sufficient data, the computer yields a print-out of all chains of primary consanguineal and/or affinal links to the tenth remove that may obtain between spouses.[84] For 457 Hare marriages falling within the interval between 18go and 1 956, there was sufficient genealogical background to demonstrate all linkages between the spouses.

The computer print-out revealed that almost every Hare was linked, within ten or less removes, to his partner in several or many ways (maximum: 148) if affinal connections are included. But the surprising result was that only 35 or 7.6 per cent of the marriages were between consanguines. Out of these there were twelve cases—2.4 per cent of all marriages—of consanguineal cousin marriage, all between second cousins. Eleven of the twelve were terminologically “cross-cousins.” Compare this with the 40 to 50 per cent of first and second cross-cousin marriages in the two Ojibwa populations cited above.

The rates of consanguineal marriages through six and a half decades (1890 to 1956) remain remarkably constant for the Hare. However, half of the second-cousin marriages (six) occurred in the first two decades. In sum, it appears that the Hare suppress marriage between consanguines, and did so at a period when it is very unlikely that white accultura- tive press would have had much effect. Specifically, if it were simply a matter of choosing among a number of kin-types, including con- sanguineal cross-cousins, that fall within the class of “marriageable relative,” distant or near, the Hare should come considerably closer to approaching the proportions of first and second cross-cousin marriages found among the Ojibwa groups reported by Hallowell and Dunning.

Part V: Demography and Population Ecology

23. Epidemiological Factors: Health and Disease in Hunter-Gatherers

Frederick L. Dunn

Introduction

A series of generalizations concerning health, disease, and mortality in hunter-gatherers are reviewed and discussed in this paper. Some are long-established and widely accepted; others are proposed in the light of certain considerations of ecological diversity and complexity. I shall refer particularly to the published record for health and disease in the Bushman, the Australian aborigine, the Eskimo, the African Pygmy, and the Semang (Malayan Negrito). In the course of the discussion some epidemiological or “medical ecological” issues will be raised which pertain to the study of diseases: (I) as agents of natural selection, and (2) as dependent or independent variables influencing size and stability of human populations. I shall emphasize the role of disease in population regulation rather than in selection, but much that will be said relates to both areas of inquiry. Disease-particularly parasitic and infectious disease-has attracted considerable attention from human geneticists and anthropologists in recent years (Livingstone, 1960; Motulsky, 1960; Neel et al., 1964; World Health Organization, 1968) but the role of disease in population regulation has been somewhat neglected.

Whether one’s concern may be the destiny of the modern hunter-gatherer or the reconstruction of prehistoric hunting and gathering conditions, hunter-gatherer studies are really urgent today. This urgency is emphasized in a succinct World Health Organization report, “Research on human population genetics” (1968), which also outlines research needs, procedures, and the kinds of demographic, ethnographic, genetic, and biomedical data whose collection is so essential. These procedural matters need not be discussed here, but a brief review of the means available for gathering data and developing a picture of disease patterns in prehistoric hunter-gatherers may be useful. Only four approaches come to mind:

  1. We may study health and disease in modern hunter-gatherers and project these findings into the past. This approach assumes that disease patterns in populations still removed from substantial contact with outsiders are to some extent similar to ancient disease patterns. The extensive pilot study of the Xavante Indians of the Brazilian Mato Grosso by Neel and his colleagues (1964) is a good example. Their report provides some tentative hypotheses of human genetic and evolutionary interest; many indications exist that additional studies along the same lines may also provide data of value in the interpretation of prehistoric populations and demographic conditions.

  2. We may review first-contact records- historical and ethnographic within this century —for information about specific diseases and conditions of health. Much of our knowledge of aboriginal health conditions in the Americas has been built up in this way (Cook, 1955, for example).

  3. We may make use of increasing refinement of archeological technique and extensive multidisciplinary collaboration in archeological investigations, which are beginning to provide extraordinary dividends in knowledge of the prehistoric environment. Paleoecological studies now make it possible to assess prehistoric environmental potentials for transmission of infectious and parastic agents of disease. Paleo- pathological investigations are becoming increasingly sophisticated, and coprolite studies, which also provide valuable dietary and ecological information, have opened the way to epidemiological assessment of intestinal helminthiases in prehistoric populations (Callen and Cameron, 1960; Samuels, 1965). Recently, in collaboration with Dr. Robert Heizer of the U niversity of California I have had the opportunity to study a series of more than fifty coprolites from Lovelock Cave in Nevada. We found no eggs or larvae of parasitic helminths, but because of the extraordinary state of preservation of certain pseudoparasites-mites and nematodes-we were able to conclude, with a fair degree of certainty, that the ancient people represented by these specimens were in fact free of a whole series of intestinal helminths, including flukes, tapeworms, and important nematodes such as hookworm and Ascaris. This conclusion was not altogether surprising in the light of our knowledge of helminthiasis in modern Bushmen and other hunter-gatherers of arid regions.

  4. We may finally approach the study of prehistoric patterns of disease in man through the study of disease patterns in non-human primates today in a variety of ecological settings, some of which are shared with modern hunter-gatherers (Dunn MS-I, 1966). In this relatively unexplored field we can only expect to discern broad patterns of similarity for man and other primates. I shall return briefly to this approach when I discuss ecological complexity and diversity.

Disease in the Population Equation

To examine the role of disease in a human population, we may refer to the “equation of recruitment and loss” of Wynne-Edwards (1962), defining the variables contributing to the condition of population stability:

Recruit- + Immi- = Uncon- + Emi- + Social ment gration trollable gration Mortality

Arising Los es

from Reproduction

In this equation, “uncontrollable losses” are identified as the independent variable and include losses due to predation, parasitism (infectious and parasitic diseases), accidents, starvation and old age (chronic diseases, diseases associated with aging). Social mortality comprises that mortality arising from “social stress”: strife, stress diseases, cannibalism, infanticide, war, homicide, suicide, etc. Wynne- Edwards, with an acknowledgement to Carr- Saunders. (1922), discusses population stability in hunter-gatherer and subsistence agriculturalist societies in terms of this equation. Births and immigration are balanced by emigration and two kinds of mortality, the socially controllable (social mortality) and the socially uncontrollable (losses due to independent variables). Within the realm of controllable (sometimes even “acceptable”) mortality is population regulation by abortion and infanticide, by reduction in numbers (sacrifice, cannibalism, head-hunting and other forms of warfare, geronticide, and even homicide and suicide), and by territoriality (rules regulating the utilization of space and resources, land tenure, marriage regulations, etc.). In a brief discussion of mortality due to predation and disease, Wynne-Edwards points out that neither of these variables can be considered wholly density-dependent or density-independent. Predation not only contributes to uncontrollable loss but is also density-dependent insofar as the prey may “cooperate” in making its surplus members available to predators. Similarly, infectious and parastic diseases may be either uncontrollable or density-dependent, as socially stressed individuals are selected out by disease because of their “depressed” physiological state. Thus, as variables contributing to mortality, predation and disease due to parasitism are to some extent controllable and density-dependent, to some extent uncontrollable and independent. The disease variable is obviously important in the regulation of human population but its impact is constantly modified by sociocultural and ecological factors. Other variables in the demographic equation-notably accidents, predation, and certain forms of social mortality-may affect population size more profoundly than disease in certain settings.

Mortality and Disease in Hunter-Gatherers

Some of the characteristics of the hunting and gathering way of life (as seen today and in historic times) that may influence the prevalence and distribution of various diseases and other causes of mortality are these: the size of the social group is small, and contact between groups is limited, in part because of the relatively large area needed to support each small population unit; the group utilizes environmental resources intensively but with minimal permanent disturbance of the environment; the individuals are well adapted to the conditions of the ecosystem in whieh they belong; the individual lives in intimate contact with his fellows and the environment; the group is characteristically mobile, within certain territorial limits; dwellings are often rudimentary or temporary; the dietary range is relatively wide, at least potentially-less desirable foods are available as a reserve for times of hardship; in general the diet may be said to be well balanced in the sense that minimal nutritional requirements are apparently met; in the tropics and subtropics the vegetable component of the diet normally exceeds the animal component in quantity ifnot in quality; in the arctic and in certain maritime and riverine settings the animal component may bulk larger than the vegetable; occupational specialization except along lines of age and sex is relatively slight.

With these characteristics in mind we may consider some generalizations about health, disease, and mortality in hunting and gathering peoples (prehistoric, precontact, or isolated modern population).

  1. Patent (and perhaps even borderline) malnutrition is rare. We should expect this in stable, well-adapted hunter-gatherer populations, modern or prehistoric. Dietary resources, even in arid environments, are diverse: typical sampling of these resources by modern huntergatherers in ecosystems relatively undisturbed by outsiders seems to provide at least the minimal protein, carbohydrate, fat, mineral, and vitamin requirements. Many workers have commented on the relatively good nutritional status of hunter-gatherers in comparison with neighboring agriculturalists or urban dwellers (for example: Ackerknecht, 1948; Maingard, 1937; Turnbull, 1965b; Woodburn, 1959). Neel (1962) postulated a dramatic gorging- and-fasting way of life for hunter-gatherers through “99 percent of hominid existence” in developing his hypotheses of the “thrifty” diabetic genotype. In the light of recent studies of the Bushman (Lee, 1965), this portrayal of major fluctuations in the quantity and quality of the hunter-gatherer diet is probably somewhat exaggerated.

  2. Starvation occurs infrequently. While huntergatherers are rarely exposed to relative dietary deficiencies leading to malnutrition, they may, exceptionally, be faced with gross deficits. In tropical and temperate regions of moderate to heavy rainfall, starvation has undoubtedly been the exception, occurring only in individuals incapacitated for other reasons (Turnbull, 1965b). In the arid tropics only an unusually prolonged drought may be expected to imperil the food supply, but failure of the water supply may select out the aged and sick before gross food shortages can have their effect. In the arctic and subarctic winter, on the other hand, starvation has probably always been a relatively important cause of death (Ackerknecht, 1948). In general it seems that agriculturalists, more or less dependent on one or several staple crops, are more liable to food supply failure and famine than are hunters and gatherers. Famine risk may even be increased as the effects of dry seasons are exaggerated through agricultural modification ofthe natural vegetational cover.

  3. Chronic diseases, especially those associated with old age, are relatively infrequent. Birth rates arc high, but the population pyramid for hunter-gatherers is broad only at the base because life-expectancy at birth is low for males and even lower for females (Ackerknecht, 1948; Krogman, 1939; Krzwicki, 1934; Neel et at., 1964; Polunin, 1953; Roney, 1959)’ Shorter female life expectancies presumably reflect maternal losses in childbirth, stresses associated with multiple pregnancies and deliveries, and in certain cultures male-female dietary disparities. Although life expectancies of hunter-gatherers are low by modern European or American standards, they compare favourably with expectancies for displaced hunter-gatherers, many subsistence agriculturalists, and impoverished urbanized people of the tropics today (Ackerknecht, 1948; Billington, 1960; Duguid, 1963; Dunn, MS-1; Maingard, 1937; Polunin, 1953). Few huntergatherers survive long enough to develop clinical cardiovascular disease or cancer, major causes of mortality in America and Europe today. These diseases do occur infrequently, however, and preclinical manifestations of cardiovascular disorder may be detected in such populations (Mann et ai., 1962). Occasional writers have claimed that certain primitive populations are “cancer-free” or “heart disease-free,” but sound evidence to support such contentions is lacking, and evidence to the contrary has been steadily accumulating. Chronic diseases of early onset, including genetic disorders, do of course occur regularly in all human populations; upon survey the more severe of these appear less frequent in huntergatherers, probably because prolonged survival of incapacitated individuals is less likely in mobile than in settled situations. Such persons have sometimes been abandoned or killed, particularly in the Arctic (Ackerknecht, 1948).

  4. Accidental and traumatic death rates vary greatly among hunter-gatherer populations. It is sometimes said that accidents constitute a major cause of death in hunting and gathering societies. Although this is indeed the case for certain peoples, it is not universally true for all hunter-gatherers, past and present. Deaths due to drowning, burns, suffocation, exposure, and hunting accidents have probably always weighed heavily in the population equation for Eskimos and other peoples of polar and subpolar regions (Ackerknecht, 1948; Hughes, 1965; World Health Organization, 1963). The Australian aborigines and Bushmen, on the other hand, cannot often drown or suffocate or fall from trees, and are physiologically tolerant of an environment that is climatically “constant” compared to that of the Eskimo. Turnbull (1965b) has recently commented on the low incidence of accidental injuries and deaths among the Mbuti Pygmies. He recorded falls from trees, falls into campfires, attacks by animals, falling on a spear, bee stings, and snakebites as Congo forest hazards. Similarly, Billington (1960) noted the infrequent occurrence of trauma among “settled” aborigines in Arnhem Land. Paleopathologists have nevertheless demonstrated abundant traumatic pathology in skeletal series for certain hunters and agriculturalists; most of these series have been recovered from archeological sites outside the tropics.

  5. Predation, excluding snakebite, is a minor cause of death in modern hunter-gatherers p predation may have been relatively more important in the past. Haldane (1957) and Motulsky (1960) suggest that through much of hominid evolution predation and other selective agencies were far more important than infectious and parasitic diseases, which became major agents of natural selection only after the introduction of agriculture with its concomitant increases in population density, community size, and community contact. Livingstone (1958) has provided an interpretation of sickle cell gene distribution in West Africa that supports such an hypothesis, but little other evidence is available at this time to assist us in evaluating Haldane’s suggestion. We cannot rule out the possibility (a strong probability, in my opinion) that some infectious and parasitic diseases were important selective agencies in prehistoric hunting and gathering populations, as they undoubtedly are and have been in modern hunter-gatherer societies. Other parasites and microorganisms have undoubtedly become important in selection only in recent millenia in response to culturally induced environmental change. We do not know the relative importance of predators at various points in human evolution, although common sense suggests that predators constituted a greater hazard for early man than for modern hunter-gatherers. Accurate predation data for the great apes would be of interest in this connection, but it would be rash to utilize these data in generalizing about predation in early hominid life. In reviewing some of the literature on health in huntergatherers, I have come upon occasional references to snakebite fatalities, which were apparently particularly frequent in Australia (Cleland, 1928), but references to animal attacks have been few, even for the ‘Arctic (Ackerknecht, 1948; Turnbull, 1965b).

  6. Mental disorders if huntlng and gathering peoples have been so little investigated that no generalizations about incidence can be justified. Descriptions I have encountered of presumed mental disorder have been presented in terms of the psychological norms of the observerrecorder (for example, Cleland, 1928; Basedow, 1932). Ackerknecht (1948) and others have suggested that the rates for phobias and hysteria were higher for the precontact Eskimo than for other societies; discussions of mental disorder in other hunter-gatherers usually emphasize that mental health was good, at least in the aboriginal state. Facts to support such statements are essentially non-existent. Cawte and his colleagues in Australia are pioneering in their ethnopsychiatric investigations of ac- culturated and unacculturated aborigines. A few limited psychological investigations of Bushmen and Pygmies have been cited by Doob (1965).

  7. Ample evidence is available that “social mortality” has been and is significant in the population equation for any hunting andgathering society. Cannibalism, infanticide, sacrifice, geronticide, head-hunting, and other forms of warfare have largely disappeared today, but the early contact records provide abundant evidence of the importance of these practices in many societies (for example, Ackerknecht, 1948; Krzwicki, 1934; Neel et al., 1964; RadcliffeBrown, 1933). Other forms of social mortality, such as homicide, suicide, and stress diseases, were apparently less frequent prior to contact; today they are replacing the old forms and increasing in frequency in displaced and accul- turated hunter-gatherers as in so many other human societies (Doob, 1965; Duguid, 1963; Hughes, 1965; Polunin, 1953; World Health Organization, 1963).

  8. Parasitic and infectious disease rates if prevalence and incidence are related to ecosystem diversity and complexity. Although many if these diseases contribute substantially to mortality, no simple, single generalization is possible Jor all hunter-gatherers.

  9. Introductory comments: Injection. Infections by microorganisms and parasites[85] may be classified for epidemiological purposes in two primary subdivisions: asexual injections (organisms reproducing asexually) and sexual infections ( organisms with some form of sexual reproduction; Macdonald, 1965). Each subdivision may be further divided into direct and indirect categories. An indirect infection requires some development outside the definitive host in normal transmission. This type of infection is therefore partially dependent upon conditions of environmental temperature and humidity; many but not all are specifically tropical.

Asexual infections may be introduced into a population by a single dose in a single individual. Multiplication in the community occurs most readily if the agent is rare in that community, and particularly if that agent has not circulated in the community previously. Malaria and yellow fever, whose agents are arthropod vector-transmitted, are good examples of indirect asexual infections. Direct asexual infections are numerous; they include many of the common viral and bacterial infections of mankind (measles, smallpox, diphtheria, etc.). Asexual infections often produce partial and sometimes complete acquired immunity to repeated infection.

Sexual infections cannot normally be introduced into new populations by single doses. The agents multiply and maintain themselves in the population more readily if they occur in large numbers in a high proportion of available hosts-that is, if the prevalence rate is high in the community and if the intensity of infection in the individual is great. Indirect sexual infections include many of the helminthiases: schistosomiasis, filariasis, hookworm infection, etc. Direct sexual infections are not numerous, but include such well-known parasitic helminths as the pinworm. Sexual infections stimulate poor to partial acquired immunity at best.

B. Introductory comments: Ecological diversity and complexity. The concepts of ecological diversity and complexity are well known to ecologists (Holdridge, 1965; Odum, 1959) but little attention has been paid to their epidemiological implications until recently. The starting point for discussion is the climax tropical rainforest, which is characterized by many species of plants and animals per unit area and by few individuals per species in the same unit area. A one-hectare plot in the Brazilian tropical rainforest, for example, COIl- tained 564 trees (greater than 10 cm. diameter at chest height) belonging to 60 species. Even the commonest species was represented by only 49 individuals, and 33 species were represented by single individuals (Dobzhansky, 1950). I need not stress the contrast with the tree species/individual relationships in boreal forest. As Dobzhansky (1950) has said, where diversity of inhabitants is great, the environment is rich in adaptive opportunities. This diversity in climax rain forest applies not only to trees, and to other plants, but of course also to birds, snakes, insects, mammals, and many other forms of life, not omitting potential or actual vectors, intermediate hosts, and alternative hosts for infective organisms. We are now beginning to see that this diversity also extends to parasites and microorganisms.

The ratio between the number of species and the number of individuals per species has been called the diversity index (Odum, 1959). In the tropical rain forest the diversity index is high; wherever man disturbs this forest the index is lower; wherever physical factors are severe and limiting, as in the Arctic, the desert, or at high altitude, the diversity index is very low—few species but large numbers of individuals per species. At an “edge,” that is, the line of contact between a forest and a field, the diversity index may be somewhat higher than on either side of the line. Wherever man has substantially altered the environment, and the diversity is accordingly low, some of the species represented by large numbers of individuals are commonly called “weeds.” Actually all plants and animals (including man) are weeds in such circumstances.

The complexity index recently developed by Holdridge (1965) provides a measure of vegetation complexity (and secondarily of other biological complexity), but should not be confused with the diversity index. The complexity index is based on height of forest stand, basal area of stand, number of trees, and number of tree species, but it does not take into account the number of trees per species. Holdridge has recently published a useful classification of world plant formations; it provides complexity indices for many of these formations (Holdridge 1965).

Table 1

r-b-richard-b-lee-irven-devore-man-the-hunter-52.jpg

Table 2 Parasitic Helminths and Protozoa in Four Hunting and Gathering Peoples

r-b-richard-b-lee-irven-devore-man-the-hunter-53.jpg

C. Injection and ecological diversity and complexity: hypothetical conditions. We may now contrast two “ideal” and undisturbed ecosystems including human populations of similar group sizes, similarly dispersed. One is complex-with high diversity and complexity indices; the other is simple-with low indices.

D. Generalizations: infectious and parasitic disease in hunter-gatherers. The ideal conditions outlined for a complex ecosystem are believed applicable in any consideration of infection in unacculturated, undisplaced Congo forest Pygmies or Malayan Negritos (Semang). The conditions for a simple ecosystem are presumably applicable to the Eskimo, the Bushman, or the Australian aborigine. These conditions should also be relevant in any epidemiological study of infection in non-human primates: in the more or less arboreal rain-forest species on the one hand, and in the more or less terrestrial open-country species on the other. I have recently reviewed records for parasitism in African primates, setting the findings against the conditions outlined above. The results are consistent with these conditions (Dunn, 1966). The epidemiological record for infection in hunter-gatherers is of course considerably “contaminated” by contact with outsiders, introductions of infections, environmental disturbances, and displacements. Careful study of historical and ethnographic records will probably provide additional data for examination against the hypothesized conditions. At the moment, however, I can offer only limited supporting data on helminthic and protozoan parasite species numbers in four hunting and gathering peoples (Table‘2). Records for the Bushmen, Australians, and Pygmies have been extracted from the literature (primarily: Billington, 1960; Black, 1959; Bronte-Stewart et al., 1960; Casley-Smith, 1959; Cleland, 1928; Crotty and Webb, i960; Heinz, 1961; Maingard, 1937; Mann et al., 1962; Price et al., 1963). Records for intestinal protozoa are probably incomplete; in other respects the records appear adequate. All available records for the Australian continent, representing a variety of habitats (mainly arid), are shown in one column; findings for one nomadic and isolated population in Central Australia are also shown separately. Semang data are based primarily upon my own surveys in 1962–64 (Dunn and Bolton, 1963), and on the work of Wharton et al. (1963) and Polunin (1953). A few species of helminths and intestinal protozoa in modern acculturated, semiurbanized Eskimos have been recorded, but I have not so far encountered any records of parasitological survey for truly isolated Eskimo populations.

E. Summary. The complexity and diversity of the ecosystem must influence-perhaps profoundly-the patterning of infection and disease in hunter-gatherer populations. Prevalence, incidence, intensity, distribution, transmission efficiency, vector abundance, intermediate or alternative host variety and abundance are all affected by these ecological factors. For some hunter-gatherers many kinds of parasitic and infectious diseases may be responsible for moderate rates of morbidity and mortality; for others only a few kinds of disease may be responsible for much morbidity, and even mortality; some diseases appear at long intervals, causing high mortality; others are ever present, causing more or less morbidity and mortality depending on a complex array of ecological and cultural factors. If diseases caused by parasites and microorganisms serve as agents of natural selection and population regulation, then any patterning of these diseases will mirror, or be mirrored by, similar patterning of disease-related selection pressures and population-regulating mechanisms.

Conclusion

The hunter-gatherer is an element in an ecosystem and cannot isolate himself from his environment. Little or no buffering stands between him and the other components of the system. His relationship to the land, to its flora and fauna, and to his fellow man is intimate. Although he is never perfectly adapted to the conditions in his environment, there is a degree of adaptive stability-of ecological conserva- tism-which does not exist in a modern urban setting. When forces of change appear, however, rapid and profound destabilization may follow. The medical sphere today provides many examples of this destabilization among the surviving hunting peoples as they encounter outsiders, lose their lands to agricultural and pastoral encroachment, and even suffer permanent displacement from these lands. “New” diseases and disorders appear, endemic diseases either disappear or become more prevalent and more damaging as social and ecological change enhances their transmissibility, and former causes of mortality are displaced by new causes.

Hunters today do not live in wholly aboriginal or “prehistoric” states of health, and historic or ethnographic records offer little data upon which to base speculations about prehistoric conditions of health. We can, however, take note of the patterning of present-day health problems among the surviving hunters. This patterning is inevitably affected by the ecological setting of the hunting group, for, as I have previously noted, the hunter lives in a truly intimate relationship with his surroundings. If we can determine and understand the factors that contribute to patterning of diseases and causes of death in modern peoples who live in an unbuffered relationship to their environment, we can apply this understanding to the interpretation of the prehistoric and evolutionary record, even in the absence of knowledge about which (specific) diseases may or may not have been present in an ancient population.

A discussion of epidemiological patterning for the infectious and parastic diseases has been presented; this patterning is linked to ecological factors, including diversity and complexity. Elsewhere in this review I have mentioned striking differences in incidence for predation, accidents, a,nd other causes of death for huntergatherers in various geographical settings. A few ecologically linked patterns of morbidity and mortality have emerged; others will no doubt be added as old information is screened and new data become available. The findings already emphasize the fallacy in generalizing about “hunter-gatherers” as though they were some kind of homogeneous cultural-genetic- ecological unity. They are diverse, their hunting territories are diverse, and so are their diseases and ways of death.

24. Some Predictions for the Pleistocene Based on Equilibrium Systems among Recent Hunter-Gatherers

Joseph B. Birdsell

Introduction

The conceptual use of equilibrium systems is ancient and honorable in the social sciences. But in practice such systems have usually remained at the level of diagramatic models. A stable equilibrium system provides a suitable basis for prediction from a number of points of view. It ordinarily remains in a steady state, or if displaced returns to it. An analysis of those forces which contribute to its balanced state allows for the quantification of the forces and encourages the exploration of the interrelations between them. Especially suitable for predictive projections are equilibrium systems whose mean values show a high degree of correlation with one or more external variables. In the brief outline below, a series of equilibrium systems are discussed which will assist directly, or with further elaboration, in reconstructing the dynamics and attributes of prehistoric populations, both socially and biologically, and in defining the ecological niche occupied by hunters in the Pleistocene period.

The Density Equilibrium System

The System

Density is a ratio expressing the relationship between people and space. Among hunters it is usually expressed in terms of the number of square miles or square kilometers needed to support one individual. It contains the implicit assumption that the population is representative in numbers, in sex ratio, and in age distributions. To be meaningful, it further presumes an essential constancy in population numbers; that is, neither rapid expansion nor decline is involved. For our purposes it always refers to precontact conditions, and some systematic form of reconstruction of numbers and territorial usage will necessarily be involved.

A density equilibrium system is based upon the broad biological assumption of a balanced condition in nature for all life, including man at a hunting and collecting level of economy. Short-term oscillations are not precluded, but are a necessary part of the system, and reflect various forces applying corrective measures after the equilibrium system is disturbed. The system is usually visualized as representing a population approaching the carrying capacity of a given environment at their given level of extractive efficiency and pattern of social behavior. Among even the simplest of peoples, social factors seem to operate to stabilize population numbers below the absolute level of saturation. Where food resources are scattered and generalized, and methods offood preservation limited or absent, the equilibrium system can be expressed in rather simple and direct terms. Where concentrations of food allow specialization, together with its preservation, cultural deviations in behavior are to be expected.

The primary force toward an expansion of population numbers is no more complicated than the inherent excessive fertility which characterizes all successful species. For simple hunters, the intrinsic rate of increase has been calculated as a doubling of numbers per generation (Birdsell, 1957) but this should be raised to a tripling in numbers per generation, with the mean generational interval extended to about thirty years. The depressant forces of the system which balance the fertility force are highly varied and include all of the climatic, biotic, and social factors which directly or indirectly limit human numbers. They may exist both as short term seasonal forms, involving stochastic variations, as well as those of a long trend nature. While the real depressant system represents a complex multifactorial series of forces, reasonable approximations in some regions may be simple in nature.

The Evidence

The general nature of the system of density equilibrium has long been known to zoologists, geographers, and anthropologists. But as expressed in purely qualitative terms, this knowledge resulted in few useful predictions. The first detailed quantification was based upon Australian aboriginal tribal distributions (Tin- dale, 1940) as analyzed by Birdsell (1953). A series of 123 ecologically uniform tribal territories yielded a logarithmic relationship with mean annual rainfall. The coefficient of curvilinear correlation reached the very high value of 0.8. The remarkable functional relationship between rainfall and tribal area (I) demonstrated the existence of an equilibrium density system, (2) validated the previously necessary assumption that these hunters existed at the approximate carrying capacity of the environment, and (3) with further analysis, showed that the size of the dialectical tribe statistically tended toward a constant value which was estimated as approaching 500 persons. The nature of this density system allowed statistically accurate predictions for aboriginal Australia. But it was pointed out explicitly that the constants for this equation would change markedly in terms of differences in biota, environmental variables, and methods of extractive efficiency for hunters in other regions.

Two of my students have made similar analyses. M. Vorkapich (unpublished) found among the Great Basin Shoshoni tribes a density equilibrium system comparable to the Australian one, with an equivalent coefficient of correlation for the restricted rainfall values involved. J. Hainline (1965) investigated data in Micronesia and found that population numbers were a direct and highly correlated function of the land area in that region. It is to be hoped that someone will undertake a similar analysis based on the scanty remaining African data.

Some evidence exists to show how the system operates when densities have been reduced well below carrying capacity. Four populations which serve as models under these circumstances have been quoted by Birdsell (1957) and are the basis for calculating the intrinsic rate of increase in economically simple populations. These empirical data reveal that where unoccupied ecological space exists, the expanding population may begin to bud off when values from 60 to 70 per cent of carrying capacity are realized. Since man is a long-lived mammal, the extrapolation from other sources suggests that his population numbers may overshoot carrying capacity before being depressed back to the long-term equilibrium value.

Predictions for the Pleistocene
  1. Pleistocene men everywhere belonged in a long-term density equilibrium systems for which the constants would change with biotic, climactic, and technological variables.

  2. Oscillations in such an equilibrium system would quickly be restored upward to normal values by human fertility factors on the one hand, or depressed when necessary by the complex of forces existing in the environment.

  3. The intrinsic rate of increase in man is sufficiently high to assure that depleted numbers would be completely restored within a few decades. In Pleistocene time perspectives, the recovery was instantaneous.

  4. The thrust of the fertility force is sufficient so that human populations would tend to overshoot carrying capacity as they restore their numbers from a depressed condition.

  5. While ecologically available empty space was present, whether due to colonizing of new areas, or due to the loss of previous occupants, it is predicted that budding-off of colonizing groups would occur well below the density equilibrium value.

Areas of Needed Research

Starting with a basic territorial unit, in which both food and water resources are locally earned, densities have a constant meaning. But as one proceeds to ecologically more variable regions, added factors begin to complicate the basic concept. Densities at the mouth of the Murray River were forty times greater in fact than would be predicted from the rainfall regression curve. This is a dramatic manifestation of the influence of locally unearned water. Marine resources, ranging all the way from sedentary shellfish to free-ranging fish and pelagic mammals, represent a locally unearned food resource. Since they directly affect human density, where they are technologically available, a system must be devised to translate these marine energy sources into terms that are consonant with purely basic terrestrial densities. Among the Kaiadilt of Bentinck Island, in the Gulf of Carpentaria (Tindale, Ig62b), high population densities prevail, when measured in terrestrial area, but almost all of their food resources were derived from tidal waters around the island. Neither the statement that one square mile of tidal reef will support eight Kaiadilt, nor the figures that are based upon one square mile of island surface can be meaningfully translated into the density system based upon inland tribes living upon locally earned resources. Where broad regional deviations from the basic ecological conditions occur, as with the seasonal availability of salmon, whether it be among the Ainu, or with the Indians of the British Columbia Coast, ecological density systems no doubt prevail, but they need redefining to be equated with terrestrial ratios.

Archeologically, sites based upon local concentrations offood resources will systematically be better represented in the recovered data than will scattered occupancies based upon generalized sources of food. The Pleistocene record will be biased heavily in favor of the former. A priori, it would be difficult to predict the number of elephant bog-trap sites on the one hand, or the migratory pathways that channeled big game herds, such as that at Solutre, on the other hand. Their existence may be determined empirically, but there remains the difficult problem of translating their ecological impact into broad regional density values.

Numerous studies of subhuman mammals have shown that food resources are not the sole determinant of population densities. Behavioral factors influence the equilibrium densities, and under some conditions become the critical factor (Calhoun, 1962). Even in aboriginal Australia the densities attained reflect some behavioral influences. Research is urgently needed among high-density hunters, especially those living upon locally unearned food resources, to determine the impact of social factors upon equilibrium population densities.

Communication Eq.uilibrium System

The System

Density is an expression for the distribution of people in space. It contains no implications as to the nature of their social or cultural aggregations in that space. A priori, spatial distribution among hunters might be expected to fall into two polar types. Distance alone might act as an isolating mechanism, so that variations in behavior occur as some fairly constant function in space. On the other hand, behavioral patterns might be nucleated so that cultural and social space take on a cellular characteristic. Most of the living hunting and gathering peoples for whom the reasonable precontact reconstructions can be made, show a cellular arrangement in terms of language variations. The Australians, the Andamanese, the Kalahari Bushmen, and even the territorially diffuse Shoshoni reveal a honeycomb pattern, where dialectical homogeneity provides the best form of group identity. Both the natives and the anthropologists have recognized the dialectical tribe as a reality.

In its clearest form, the dialectical tribe lacks all aspects of political authority, and simply consists of an aggregation of local groups in spatial proximity. The Australian data show a useful constancy in size for the dialectical tribe, which statistically approximates 500 persons. This tendency is independent of regional variations in density. Since the data cover mean annual rainfall variations from 4 inches to more than 160 inches, the size of the dialectical tribal unit is insensitive to regional variations in both climatic and biotic factors. Its primary determinants are competence in speech, and mobiE ty on foot.

Dialectical homogeneity at thesimple level of hunters may profitably be viewed as a consequence of the pattern of density in face-to-face communications. Involved are (I) frequency of interaction, (2) intensity of interaction, (3) duration of interaction, and (4) facility of communication. The density of communication defines the clustering of groups that will show a dialectical homogeneity. These variables in turn are mediated by the basic geometry of the local groups. It has been shown (Birdsell, 1958) that as the number of local groups increases, greater internal cultural and linguistic homogeneity should be expected, for alien contacts are reduced in a systematic fashion. Thus, there is a positive force acting to increase the number of local groups who speak the same dialect’ as a consequence of the pattern of communication density. But the effect of isolation by distance opposes this integrating force. It can be shown that given con

I. Speech is homogeneous within the dialect tribe only relative to other tribes. Very slight differences can be expected between all of the local groups that comprise the tribe, as a simple function of distance as an isolating mechanism within the tribe. stant limits on personal mobility, in time the dialect unit frequently becomes too large to maintain its tribal homogeneity. Speech differentiation proceeds further, and in time will produce tribal separation through dialectical fragmentation. Since there is a central tendency for the dialect tribe to approximate 500 persons among widely different groups, such as Australians, Shoshoni, and Andamanese, among others, it is tempting to consider that this value is relatively independent of both the local biota and the technology upon which its extractive efficiency is based.

Deviations from the modal value of 500 persons in themselves provide material for investigating those factors which may change the balance point of the communication equilibrium system. In Australia, on parts of the Darling and Murray rivers systems, there were several tribes regionally adjacent to each other in which the size of the dialectical unit greatly exceeded the normal value. Among the Kami- laroi, the Wiradjuri, and the Worgaibon, tribal numbers are estimated to have ranged from 2,000 to 5,000. In this same area, there is a clue as to why dialectical integration may have resulted in these larger groups of people. Unlike most regions of Australia, men in these tribes could and did inherit territorial rights to special food resources in their mothers’ territory. The data suggests that this simple modification in jural behavior modified the pattern of the density of communications so as to result in the linguistic integration of unusually large numbers of local groups. In a comparable fashion, among the Kalahari Bushmen, variations in residence have been reported in terms of the universal bride-service (Marshall, 1960). When combined with an incipient tendency toward political authority vested in headsmen, this may have resulted in the development of larger dialect groups there. These examples suggest that the size of the dialect units, and their equilibrium value, may respond sensitively to seemingly minor differences in behavior which they infringe upon the pattern of communication density.

The Evidence

There are a variety of data bearing upon the existence of the dialectical tribe as an expression of the communication equilibrium system. Its widespread existence, as in cases mentioned above and others, gives an empirical verification. It is a commonplace in Australia that when tribal numbers expand to 1,000 or more individuals, then dialect fragmentation regularly occurs. At the time of contact the Aranda were visibly differentiated into five dialect subgroups (Strehlow, 1947) with a combined population estimated at 1,500 persons. The Walbiri according to Meggitt (1962), numbered somewhere between 1,000 and 1,400 aborigines, and had differentiated into four dialect subgroups. Those data suggest that isolation by distance begins to produce recognizable dialect fragmentation when tribal size approaches 1,000 persons and that the differentiating dialect subgroups range from about 250 to 375 in numbers, values well below the equilibrium value of 500 persons. In the Kimberleys of Western Australia other large tribal groups, such as the Walmadjari and Njamal, were in the process of linguistic differentiation at the time of contact. The model of band interaction, referred to above, suggests that a pair of main and opposing forces, involving cultural cohesion and isolation by distance, tend to maintain the equilibrium numbers in a fairly rigid manner.

Finally, in an experiment in nature (Birdsell, MS-I), it can be shown that after the size of the Australian tribe has been abruptly reduced in numbers (but with no reduction in density), there follows in time a regular reintegration, presumably in terms of local groups, until the full and normal number of 500 persons is restored for the dialect tribe. The social dislocation in this case involves tribal names and identity rather than dialect differentiation. The form of the restoring curve is heavily damped, with no tendency to overshoot the equilibrium value. This is what one would expect in terms of a system whose values are primarily determined by the overall density of communication.

Predictions for the Pleistocene
  1. Pleistocene populations will ordinarily be expected to be arranged in tribal units marked off by dialect differences. If exceptions are to be postulated, they should be based upon some influence which tends to minimize distance as an isolating factor, such as a concentration of food resources.

  2. In a statistical sense, the central tendency for the size of the dialect tribe in the Pleistocene can be estimated as 500 persons, although the variance will be considerable. This is based upon the value among a variety of living hunters and gatherers showing a wide range of variations in extractive technology, biota, and other ecological variables. There is no reason why this estimate should not be pushed as far -back into the Pleistocene as the development of fully symbolic speech and foot mobility would allow.

  3. While much individual variation in the size of the dialect tribe will occur in the Pleistocene, as in the present, those which become too numerous will be subject to linguistic differentiation and division, while those which decline below a critical number will be forced into amalgamation with others, since they would reach a position of instability in the equilibrium system.

Research Needed

A linguistic analysis of the degree of difference between the dialects which form tribal groups is badly needed. While the native informants themselves recognize varying degrees of difference, these will not become materials for further analysis until they are linguistically standardized. There is even some question whether dialect tribal groups are always defined at the same level of difference. Are the much larger groups found among Kalahari Bushmen real, meaning that the linguistic differentiation is at the same level, or has a different measure of language differentation been used there than in Australia and elsewhere? Linguistic controls. are obviously needed to answer questions like this. I t would be helpful further to determine the degree of linguistic differentiating normal among the local groups which comprise the dialect tribe.

There exist, in the voluminous ethnological data upon Australians, almost no details which can be used to investigate and elaborate the variables which make up the overall density of communication. While most writers report that initiation ceremonies exist, they never report the interval in years between them, the recruitment, comp o sit ion and origin of the visiting parties, how far they had traveled, or other factors of importance. Models serve as guides for this sort of investigation, but they can ollly be tested and refined against good empirical data. It is too late to hope for the critical data from the Australia of today, so they must be recorded for other economically simple groups which are still functioning.

Local Group Eqiltillbriltm Systems

The System

Just as most terrestrial vertebrates have evolved patterns of behavior which result in their efficient distribution in spare to maximize population survival probabilities, so have men at a hunting and gathering level of economy. Human behavior is both generally adaptive and cOIl\’(‘niently flexible so that the composition of the local territorial group and the details of its relations with neighboring groups can he expected to vary. There is no need to COnsider the rigidly ideal model constructed by Hiatt (1962) and criticized with complete competence by Stanner (19(5). Local group functioning and structure have seldom been observed in undisturbed density equilibrium systems, although a few data do exist. In general, their characteristics must be reconstructed through a detailed and conscientious evaluation of precontact food resources, surface water availabilities, and group compositions. This is a most difficult task, for local group structure and function are the first things to disappear in the contact situation.

Using Australian aborigines as models, some general attributes of the modal typc of local group can be defined. It will tend to consist of a suitably sized group of genealogically related families. Anthropologists have noted and models confirm (Birdsell, MS-3) that the variety of possiLle relationships is both great and will change with time. The type of food resources is perhaps the most important single determinant of local group size and composition. \’\There plant and animal resources are largely raised locally and comprise wide variety of forms, the size of the exploitive groups will be small, and figures tend to oscillate around an equiliLrium value of 25 persons. Such groups are almost universally exogamous, although T in dale (l g62b) has fouhd an interesting exception, which is to be explained by the unique circumstances in which it occurs.

Wherefood resources are more concentrated, the dietary becomes more specialized and local group structure frequently changes. It should be stressed that this is not a consequence of increasing density, but rather a function of concentration. In such circumstances the average size of local groups may range between 50 and 100 individuals, and in some exceptional cases reach several hundred persons. But it is to be noted that the reality of such groups can be reconstructed o n l y where sufficient food and water resources can be systematically demonstrated by detailed data. This has not been done, for example, by either Meggitt (1962) or Elkin (1962) for the Walbiri of central desert Australia. This does not deny that transient groups of several hundred individuals may not congregate where food and water resources are available, for evidence indeed suggests this is an almost universal human impulse. But a temporary aggregation cannot be considered as representing local groups or communities who demonstrate their territoriality in some consistent fashion. Where ecological resources allow these larger groups for some portion of the annual cycle, it is to be expected that both social and genealogical relations, as well as forms of recruitment, will differ from smaller groups. Forms of descent, marital restrictions, and permanence of membership can be expected to vary within the reasonable limits of adaptive human behavior.

The first type of local group, characterized by small numbers supported by a local biota which generally requires greater group mobility for its use, does represent a rather tight equilibrium system. But the system’s forces which maintain its central tendencies are not easy to discern. I t is tempting, regarding such local groups in Australia, to consider that territorial dcfense, which occured frequently, set the limit of group size by its demand for a minimal number of competent spearmen. But the same size group prevails among the Kalahari Bushmen, where territorial defense is exceedingly rare. I t is also found among the

Birhor of India, where Williams (Chapter 14, this volume) shows that rigid territoriality does not exist. Williams’ explanation, that local groups of this size and character have size limits determined primarily by such social arrangements as are involved in the pattern of marital exchanges of women, is attractive and more likely will prove the primary determinant. Stochastic variations in the periodicity and sex of births can be visualized as imposing a minimum size for successful survival on such local groups. The upper limitations on size will be determined by a complex of factors involving human mobility, the nature and concentration of food resources, and the form of the technologies by which the biota is exploited. I t is revealing that using netting techniques for the primary pursuit of game, the Birhor consider that three to four families comprise the efficient work group in their environment, whereas Turnbull (1961) indicates that four or five families are too few to constitute a viable netting team among the Bambuti of the Congo rain forests. There can be no doubt that the type of local work group will represent an important determinant in group size, but more data are needed to define its impact. That reconstructions of basic ecology, local residential group, and work groups can be achieved, in some cases long after they have disappeared functionally, has been demonstrated by vVata- nabe (1964a) in his elegant work upon the Tokapchi Ainu.

The Nature of the Evidence

Local groups dependent upon a generalized dietary have been well censused for three different peoples. In each case the average size approximates 25 individuals and this seems to represent a minimum, modal value for viability for this type of group organization. Among the Birhor, Williams shows that in 25 bands the average group consisted of 26 individuals. Marshall (1960) found that for eighteen bands of !Kung Bushmen, the average size was again 25 persons per group. Unpublished data from Tindale (personal communication) indicates that among ten preeontact desert Australian local groups counted by him, the average number is again 25 persons. In the last case, the range is very tight and suggests that the forces which maintain such an equilibrium of numbers act powerfully. Data detailed by all three authors validates the nature of the equilibrium system by the demonstration that bands that are either unusually small or large in their numbers find an explanation within the forces of the system. Local groups of this type must be large enough to function successfully in a social sense, and small enough to minimize strain on food and water resources.

Local groups of larger size have been widely documented in the ethnological literature and in a general way their occurrence is based upon concentrations offood rather than high popu,la- tion densities per se. The nature of such food concentration varies and may be partly regional, as in the cases where acorns and salmon are major food resources, or highly localized as in the case of sago swamps, elephant-bog traps, and the ancient sea mammal resources as at Iputiak. The very availability of large masses of food in a concentrated form, combined with methods of food preservation, suggests that the local groups will be maintained by a much looser system of equilibrium forces and that these will show a wide variety of forms. Behavioral preferences may override energetics. There is no profit in attempting to model such situations until much more detailed data have been provided from field research.

Local group territoriality manifests itself adaptively in a wide range of definitions. These are well discussed by Stanner (1965) where he introduces a terminology involving (I) estate, (2) range, (3) domain, and (4) regime. These terms focus attention upon the ecologically varying nature of territoriality among such generalized hunters as the Australians.

Predictions for Pleistocene Hunters
  1. Hunters dependent upon a generalized and locally raised biota should fall into the equilibrium system found among desert Australians, Kalahari Bushmen, and Birhor. Their modal numbers should approximate 25 persons, marriage should be exogamous, and the probabilities are high that the group was patrilineal and patrilocal.

  2. Where large food resources are eoneen’- trated either regionally or seasonally, local groups take on a very different nature. They will be larger in size, different manifestations of territoriality can be expected, and their principles of descent, marriage, and recruitment can be expected to show wide variety. Data among living hunters in such situations show such a wide range of adaptive behavior that Pleistocene reconstructions will largely depend upon archeological evidences. While ecology will determine the limiting factors in such cases, it will also permit a wide variety of social solutions.

Required Research Jor the Future

The most urgent need is to detail or reconstruct the functioning of local groups. How many people, of what ages and sexes, in what specified biological and social relationships, reside or camp together for what defined periods, and for what reasons? Travel distances between camps in terms of seasonal rhythms, and defined in terms of exploitation both of varying food resources and water supplies, are further imperatives. Much of the current confusion in the literature stems from a failure to define the ecological base which makes possible larger groupings. Australian ceremonial gatherings held at intervals of several years in locally favored spots are not comparable to the more or less permanent residential groups of similar size elsewhere which depend upon concentrations of food and some form of food preservation. Although these extremes can be conceived of as existing along a kind of complex continuum, they are near its poles and the ecologically necessary conditions require specification.

Equilibrium in Family Size

The System

In a population stabilized at the operationally defined carrying capacity of its given environment, some limitation on procreative activities naturally filter down to the level of the biological family. These may be examined most profitably in terms of the requirements which affect the spacing of the natal survivors. Generalized hunters with their requirements of high mobility, present the most exacting model. Australian data (Birdsell, MS-3) indicate that the inability of a mother to carry more than one child at a time, together with her female baggage, impose the first insurmountable barrier to a large number of children. This factor, strongly reinforced by an equally limiting incapacity to nurse more than one child simultaneously, imposes a minimum of a three-year spacing upon children designed for survival. Since human female reproductive physiology does not reliably prevent conception while still nursing, children who cannot be reared are frequently conceived and horn. The solution is systematie infanticide. There is every reason to believe that it would have been present throughout human groups in the Pleistocene period. I t seems to represent a dislocation in human reproductive physiology following entrance into the bipedal niche and rapid brain expansion.

The Evidence

Among 194 matings based upon precontact genealogies for desert Australians (126 collected by Birdsell, 68 by Tindale), data show a sex ratio at adulthood of approximately 150 per cen t (150 males for 100 females). Other data show that the sex ratio at birth among Australian aborigines is 100 per cent. The discrepancy in the sex ratios among adults indicates that at least 15 per cent of the children born to aboriginal women are eliminated by infanticide. Ethnological evidence for Australia documents that infanticide was preferentially practiced upon female babies. This was true for desert groups, and also for those living in better endowed areas. The value of IS per cent demonstrated by these data is a minimal one, and the observations of observant, early, postcontact Europeans places the figure as high as 50 per cent. Unfortunately, medical science has not yielded accurate information on the rela tionships between nursing and ovulation needed to fix the figure more accurately. Overall considerations incline a judgment toward the higher values.

That the Australian evidence is not isolated is provided by data from other groups. Birket- Smith (1929) documented sex ratios equally warped among the Caribou Eskimo, and his values were strikingly consistent among the five groups he censused. But Marshall (1962) found no sex preference in the systematic infanticide practiced by the!Kung Bushmen. There is reason to believe that the limiting forces involved in the necessary spacing of children to be saved would apply broadly to all hunting groups, although sex preferences might vary.

The factors which produce compulsory spacing of the viable children among hunters limits the total number that can be saved and hence affects the variance statistically. Among the 194 sibships contained in the precontact desert genealogies referred to above, the range of children surviving to adulthood extends from o to 7 offspring. And 185 of the sibships included between 1 and 5 children each. In a hunting population in which total numbers are remaining essentially constant, the standard deviations for the surviving offsprings approximates 1.4 and this when squared gives a variance of about 2.0. These values not only represent equilibrium figures, but exist in so tightly structured a system that they may be considered a constant for people living under these conditions. Such values should characterize human populations far back into the Pleistocene period.

Predictions for Pleistocene Populations

Systematic infanticide may be assumed to have characterized human populations throughout the Pleistocene. Its probability of being preferential female infanticide is strengthened by data from recent hunters. The variance in surviving adult sibs would approximate the constant value of 2.0, with a standard deviation of 1.4 and a range from 0 to 7.

Research Needed

Demographic and genealogical data extending into precontact situations are urgently needed. Disturbed sex ratios at birth or among adults provide indirect evidence of systematic infanticide. But their absence does not deny the practice. Medical research is urgently needed to define the relationship between the resumption of ovulation in terms of the prolongation of nursing for periods up to three years after birth.

Equilibrium in the Size of the Effective Breeding Population

The System

In terms of evolutionary processes, one of the most important population factors involves the effective size of the breeding population. Its value is determined from the following equation given by Wright (1938):

r-b-richard-b-lee-irven-devore-man-the-hunter-59.png

The equation states in essence that for evolutionary processes the effective size of the breeding population is a direct function of the number of persons who contribute their genes effectively to the next generation and an inverse function of the variance of the offspring so produced. Offspring to be counted must in their own time reach the midpoint of their reproductive span. Thus, both for parents and children very special types of data are required and these are difficult to obtain in practice.

The actual number of parents who reproduce effectively will be some proportion of the total of adults in the breeding isolate. Since in Australia, and among other hunters generally, the genetic isolate is the equivalent of the dialectical tribe, it becomes clear that the number of effective breeders will be related to tribal size. As shown above, the size of a dialectical tribe is in itself an equilibrium system and it follows that this will hold true for the number of effective breeders as well.

The rate of intertribal marriages acts in such a fashion as to further reinforce the equilibrium system represented by the number of effective breeders. Tindale (1953) showed that over wide stretches of aboriginal Australia the rate of intertribal marriage averaged 14 per cent per generation. But this rate may be expected to vary with the size of the dialectical tribe involved. In tribes which number much fewer than the 500 persons which comprised the modal tribe, more women will be brought in from outside tribes than is customary. This will tend to extend the size of the breeding population and so restore its numbers toward the higher modal value. In contrary circumstances, where the dialectical tribe considerably exceeds 500 persons, there will be less need to go outside of tribal boundaries to obtain the wives so that the rate of intertribal marriage should be less than average. In very large tribes it is to be expected that distance will also operate as an isolating factor within the bands of the tribe. These tendencies will somewhat reduce the number of effective breeders and again act as a restoring force toward the equilibrium values.

The Evidence

In spite of having worked intensively with Australian aboriginal material for a quarter of a century, the author has not yet calculated the effective size of the breeding population. This is not due to any difficulty in calculating the variance of the effective offspring in Wright’s equation, but lies in the difficulty of estimating the number of effective breeders in the population. As indicated earlier they can only be included when in turn their children reach the midpoint of the reproductive span. For man, the problem is complicated by the fact that generations overlap in real popula- tiom.Genealogical evidence, as useful as it is for many purposes, contains the built-in artifacts that originate from its origins in the memories of men. The errors contained in genealogies will often be systematic and hence effect the calculation of the size of the breeding population.

As a figure of magnitude, tht. size of the effective breeding population for aboriginal Australia should approximate 175 persons where the size of the dialectical tribe is 500. As explained above, this figure will rise a little less rapidly than the number of individuals in the dialectical tribe increase; it will diminish a little more slowly than will a comparable decrease in numbers of tribal members. But further data are needed for an accurate estimation of the real value. The size of the effective breeding population is affected by polygamous marriages, and these data for Australia are available. But biological realities require the estimation of biological paternity insofar as it differs from social paternity. This will require the analysis of blood group data to calculate a reasonable rate of illegitimacy and so to correct for the number of males effectively contributing gametes to the next generation.

One of the methods available for refining the estimate of the number of effective breeders depends upon the availability of detailed population pyramids. These have never been published for precontact Australians, but data in two areas will become available in the near future. These will help control the complication introduced by the fact that in a society where males outnumber females in the ratio of 150: I 00 it is still possible to practice polygamous marriage. The solution, of course, is to be found in the delayed age at which males obtain their wives. The quantitative data provided by Meggitt (1962) and several others will help estimate the effect of a sequence of husbands upon the effective reproduction of a given female.

There is another further loose end to be tied in before proper estimates can be made of the number of effective breeders in a genetic isolate. Given the stringent requirement that parents can only be counted as effective when their own offspring in their proper turn produce effectively, genealogical data do not provide adequate time depth. This can be found in a model of the proper construction. The author (MS-3) has explored the long-range effects of this phenomenon in a patrilineal society. Using the range of effective sibship size determined by the precontact genealogies referred to earlier in a model in which stochastic reproductive processes are assigned by random numbers, an unexpected result was obtained. Patrilineages die out at the rate of approximately 30 per cent per generation. This is of course one of the great sources of errors in genealogical materials. It has not yet been possible to determine what the exact impact of this high rate of patrilineage extinction would be upon the number of effective breeders. It is likely to reduce the size of the effective breeding population well below the value of 175 persons as estimated in an earlier and incomplete analysis.

A final problem involved in calculating the size of the effective breeding population turns upon the occurrence of periodic reductions in isolate numbers due to catastrophes in nature. Neither the spacing nor the effect is determinate, but the result would be to further reduce N.

Predictions for Pleistocene Populations

With the exception of those special instances in which concentrated food resources allow for larger human societies, it can be predicted that throughout that portion of the Pleistocene in which man lived as a successful hunter the size of the effective breeding populations will be small in numbers. The equilibrium system will be somewhat tighter than that which structures the dialect tribes through its communication system. The evolutionary processes which are to be associated with small breeding populations should have been operative during the entirety of this period. While random genetic drift has been indicated by the mathematical models to be anticipated under such conditions, empirical data from Australia aboriginal isolates suggest that the processes of evolution are in fact more complicated than have been predicted.

Research Needed

Granted that the effective size of the breeding population does constitute its own equilibrium system, there is a great need for detailed data at a precontact level for the factors which influence its magnitude. This includes not only the overt aspects of marriage sequencing but the covert aspects that may influence procreation outside of marriage. The definition of effective parenthood in terms of evolutionarily effective offspring imposes difficulties which do not promise a direct solution. Even so, empirical values from real populations and detailed genealogies are needed to assist in the modeling of effective estimates bracketing the real values.

Conclusions

Systematic infanticide has been a necessary procedure for spacing human children, presumably beginning after man’s entry into the niche of bipedalism, and lasting until the development of advanced agriculture. It involved between IS and 50 per cent of the total number of births. Among recent hunters it tends to be preferentially female in character and probably was in the Pleistocene. The variance among effective surviving sibs approaches 2.0 as a constant.

The necessary mechanisms for spacing human societies are primarily invested in local groups. Among generalized hunters, dependent upon local biota for their food sources, the local group averages about 25 persons as a basement figure. Smaller populations would be handicapped socially in a variety of ways, but primarily in those affecting martial exchanges of girls. Larger local groups would be at a disadvantage in terms of food and water resources except in unusually rich environments. Local groups of this character may be projected backward in human evolution to that point which the observer estimates as involving local group exogamy as a source of social regulation.

For generalized hunters and gatherers, the density of communications normally defined by foot mobility and fully developed symbolic speech results in a language community known as the dialect tribe. Its numbers tend to stabilize around a population involving 500 individuals, quite independently of population density, the richness of the biota, and the teclmo- logical development of extractive efficiency. Since the evidence for competent bipedal mobility, as defined by the structure of the foot, goes back into the beginning of the Pleistocene, the development of dialect tribal groups can be best associated with the evolution of the capacity for a rich oral communication. This would presumably push these linguistic units back into the middle of the Pleistocene period.

The size of the effective breeding population, a parameter of considerable evolutionary significance, has its own equilibrium system, a modification of that based upon the density communication system. For dialectical triiJes with the modal value of 500, it would be less than 175 in numbers. While its exact determination is not possible, Pleistocene populations were effectively small.

There are numerous situations in which concentrated food resources, such as migrating big game, spawning salmon, or sago swamps provide a different basis for existence. When accompanied by either artificial or natural means of a preservation, these concentrated sources of energy characteristically result in the development of much larger groups involving various kinds of recruitment, descent, marital relations, and seasonal residence. Such local population will reflect the adaptive nature of human behavior in their variety. With man’s gregarious tendencies limited largely by food resources, this type of local group should be projected well back into the Pleistocene in localities which are suitable for its development. It will appear in the archeological record with a biased frequency since generalized hunters leave but scanty remains for recovery. Since it can be shown that the size of the dialect tribal group varies sensitively with those factors which alter the pattern of the density of communication, it should be expected that large local groups based upon concentrated food supplies would also increase its size.

Since the intrinsic rate of increase in man allows it, human populations should be considered to reach the culturally effective carrying capacity of a given environment within very short time periods. For most Pleistocene situations, the population density should be considered to approximate that of operational saturation. Regional densities will be determined by variations in biota, climate, and the extractive efficiency of the men who exploit the environment. These can be determined for given zoogeographical regions only from detailed analyses using living hunters and gatherers as analogue types of models. In all cases population attributes must be effectively reconstructed to approximate precontact conditions.

25. Discussions, Part V

25a. The Demography of Hunters: An Eskimo Example

Speaker: Laughlin

LAUGHLIN: I should like to use population reconstruction, primarily of the Aleut and. Eskimo stock, as a means of demonstrating one kind of systematic approach to the analysis of a population and its history. To begin with, we ought to appreciate that the concept of the Eskimo is enormously distorted, since their distribution is quite at variance with geographical areas to which we have formerly paid most attention. The distribution of the population makes sense in terms of the history, for which there is a considerable antiquity- some 8,500 years documented in the west- and in terms of resources available in an accessible form. The polar Eskimos, who are most talked about, constituted 250–300 at the time of their discovery. The literature published about them must surely exceed in actual poundage the total body weight of all the polar Eskimos living since the year 1000. Much the same thing applies for the some 7,000 central Eskimos, who are exceeded in number by the single population of Konyang Eskimos living on Kodiak Island. The population distribution very clearly centers in the southwest, where the greatest amount of cultural diversity occurs in language, material culture, art styles, etc. Another major population center is around the Bering Strait, and a third is found on the west (particularly the southwest) coast of Greenland.

The first ordinate that we can see is the contrast between the zones in which winter ice deprives people of access to the intertidal area, and the open water zones where there is easy access to the intertidal area. If we eschew talking about a community as such and instead break it into components-such as the kinds of people and varieties of habitat pre- sent-we find that we can easily chart kinds of people and the particular habitats that they can exploit. This, in turn, makes it clear that there are more habitats available to people in certain areas than in others, which suggests one reason why there are more people living in the former areas.

If we take an Aleutian or Kodiak Island example, we can categorize the population in terms, roughly, of mobility in the following (ascending) order:

  1. Old, infirm females, who cannot do much beyond collecting shellfish;

  2. Old, infirm males, who can do the same things, and can also go fishing in boats;

  3. Pregnant women, who can do a variety of things, subject to local taboos (not touching salmon or whales, not going near streams, etc.) and the advancement of pregnancy;

  4. Children, who actually produce a great percentage of their own food supply in the subarctic zone—including, especially, shellfish;

  5. Young and middle-aged females; and

  6. Young and middle-aged males. On both of these last groups we have a good deal of data, which in some cases tends to bias our views about subsistence.

We can see immediately that as one moves north of the winter ice line, the proportion of the population actively engaged in procuring food must decline. Since the aged and infirm simply cannot drive a sled over the ice, punch a hole through two feet of ice, harpoon a seal, and bring it home, members of this group are all potential candidates for abandonment and exposure. Overall, we can count some four or more different habitats available to limited segments of the population, which keeps members of these groups from being a burden to the rest. So population density reflects not just the natural resources, but the natural resources keyed to certain kinds of individuals (and to differential technological development).

Another factor in the matter of ecological resources is, of course, the annual distribution of the resources. When bears, walrus, and other exotic forms are not available, the populations of the southwest can get by on fish, shellfish, octopi, and other denizens of the intertidal zone. They eat less well, but they survive. In contrast, people in the central Arctic may undergo final starvation, as witness the Sadler- miut Eskimo skeletons in our collections.

I would like to say something about population “bottlenecks,” since these are prominent all over the world. The people anticipate these, and there is a considerable feedback process. They know that they have to adjust to the leanest months of the year and to areas of lesser resources or greater rigor. In such situations there is an increase in intentional and explicit population controls such as prolonged nursing, abortion, infanticide, abstention, and increased adoptions. Infanticide, in particular, involves decisions based on careful consideration of a number of factors. The choice of the child to be left out to die depends upon the sex and health of the child and his siblings, the time of year, and the parents’ estimate of the likelihood of having a child of the desired sex. The anticipation of population bottlenecks is as prominent in the thinking of the Aleuts as it is among peoples further north. It influences their movements and, as described, calls forth intentional population controls.

Turning now to the topic of longevity, one of the variables about which we have reliable information is the large contrast in longevity between isolates around the world, and we can demonstrate them most neatly in the Aleut and Eskimo area. One of the two best samples that we have is the collection of 187 skeletons —collected by Dr. Charles Merbs and me—of the now-extinct Sadlermiut Eskimo. In this particular population (treating the collection like one population although this is not actually the case), we find that no one lived to an age greater than 55. Something on the order of 40 per cent died before age 4. Undoubtedly, as with all archeological series, certain ages are underrepresented, and the youngest are most underrepresented. But Eskimos are required to confess stillbirths and miscarriages, and they tend to bury the infant dead, so the series is still pretty good.

In marked contrast to this is a series for the Fox of the eastern Aleutian islands, a series compiled by Bishop Veniaminov, who spent ten years there. Fortunately, he recorded age at death, and here we find a number of people between ages go-ioo and quite a few over age 55. Infant mortality was on the order of 20 per cent, roughly half that in the central Arctic. This situation can be generalized for the area as a whole. Longevity is obviously and considerably greater in an area of good, accessible resources and high technological development.

With an increase in longevity, other changes take place in the composition of the population, some with genetic and some with cultural implications. I will point out that the development of intellectual knowledge depends upon the retention of older people who are, in a sense, libraries. There is actually some: feedback in the documented fact that the older people were the sophisticated doctors who intervened in the delivery of offspring and did a substantial amount to reduce infant mortality. So it is not just a question of more people with more food.

Pathology has told us very little about the causes of death. We can learn a certain amount about other things from gross abnormalities in the skeletons. Many of the Arctic peoples periodically suffered from malnutrition, as shown by cases of enamel hypoplasia on the teeth of the Aleuts and the Arctic Eskimos. We also know that in this area the people had a high frequency of spinal disorders-separate neural arches, and spondylitis. Yet these and some other gross pathologies apparently did not affect their behavior as much as one would expect. The frequency of fractures over the entire area is rather low, in spite of the fact that we know that they often fall and occasionally have encounters with animals having ulterior motives.

The pathological or disease load that these people carry is difficult to understand today, because they have been put through several selection filters. After the first smallpox epidemic went through the area, the population was genetically restructured. Apropos of this, one thing that we ought to keep in mind is the fact that a population can maintain its distinctive marker genes-those that deal with the external characteristics by which we identify them-and yet undergo an enormous and significant restructuring as regards certain other genes.

25b. Population Control Factors: Infanticide, Disease, Nutrition, and Food Supply

Speakers: Alland, Birdsell, Deevey, Isaac, Turnbull, Washburn, Williams, Woodburn

Birdsell: I would like to talk briefly about family size and birth spacing-subjects on which we have some very good data. If you arrange precontact genealogies and adult sib- ships of Australian aborigines according to completed family size, the sex-ratios form a very interesting systematic curve varying from about 130 males per 100 females for the modal families to about 260 males per 100 females for families in which five children have survived. Statistically, this shows an enormous amount of family planning that is not indicated in the ethnology.

Doctors who have done the best work at present on the relationship of ovulation and conception to lactation have still never done it under optimal circumstances. For example. one team did endometrial biopsies on roughly 200 women for a year; they were able to get good data on ovulation, but since the duration of lactation over this period was not known, it is of little use for our purposes. However, it is clear that in humans lactation only incompletely and/or sporadically suppresses ovulation. That is, I take it, a post-bipedal phenomenon, and I see no reason not to suppose that it characterizes all human groups. On the basis of this, I project the hypothesis that difficulties of nursing and mobility in the Pleistocene may have made necessary the killing of 1 5–50 per cent of children born, since lactation alone would not have provided sufficient spacing of births to maintain equilibrium.

WOODBURN: In talking about birth control and infanticide, it is important to remember that these decisions appear to be made right within the family, often by the mother alone. This has important implications; chief among these is the simple fact that the making of the choice is relative to the circumstances of a particular mother and not necessarily to the circumstances of the whole population.

DEEVEY: As you know, it is a moot question both in ecology and in evolutionary biology as to how it is possible for the system to exert pressure independently of or in conflict with an individual component of the system. An aspect of the same problem is the question whether there can be group selection in contrast to individual selection.

WOODBURN: I think that the N etsilik are particularly interesting in this respect. It looks as though the female infanticide rate works to the disadvantage of the population in general. I would be interested to know about the effects of infanticide on other populations, taken as wholes.

Birdsell: The Caribou Eskimo make a good case in this regard. I think that they do show the individual and the system working together. Birket-Smith (1929) had five groups- subtribes, as we might call them. He only dichotomized the population into children and adults. All five showed an average sex ratio of 145 for children. In the adults, the average ratio had fallen to 80. Those data suggest that infanticide was done as a primary family spacing, and was so programmed in terms of population effects that there were just about enough hunters that could hunt for two families as one would expect in a population of this sort. In other words, I take it that the number of competent hunters in a future context was predicted.

WILLIAMS: Birdsell once suggested in an article that populations of man have fertility such that they can fill up an ecological niche in fairly short order. Is there any evidence that the populations of the Hadza and Mbuti are growing now? And if not, why not?

WOODBURN: There is no evidence that the Hadza population was ever any greater or smaller than at present. I think the population could be a good deal larger than it is without causing hardship for the Hadza or damage to the environment. I really cannot understand the widespread emphasis on the ties between population and physical environment to the exclusion, say, of medical factors. One should not overlook the fact that when good health services are brought to an area, there are likely to be sudden and great population increases in that area, especially if malaria or any equally important disease is eradicated.

DEEVEY: Regarding population size and limits on it, I would like to ask Dr. Woodburn about mineral nutrition, especially salt.

WOODBURN: When I was in the field in 1960 I was visited by pediatricians, from Makerere College Medical School, who examined the children and described them as one of the best- nourished groups they had seen in East Africa. In the ordinary course of things, Hadza do not take any salt as such. A salt lake is nearby, but no one bothers to gather salt except those living right at the lake. They occasionally obtain a little through trade, but they do not consider it as worth much effort or expense. I assume, therefore, that they must get sufficient quantities of minerals from their animal and vegetable foods.

ALLAND: Dubos, in his recent book Man Adapting (1965), has suggested that disease does not play an important factor in population control. I am not sure that I wholly agree with him, but I do think that disease (excepting, perhaps, malaria and trypanosomiasis) is not a factor in population control among hunters and gatherers, and for the following reasons: ( I) Isolated populations are less likely to pick up epidemic diseases—an epidemic likes large number of people. (2) The parasites themselves tend to accommodate to populations in which they have existed for a long period, of time, so that there is a genetic sieve operating both to make humans more resistant to parasites and to make the parasites less noxious to humans. In hunting and gathering groups, the most common infection is helminthic, and helminths accommodate particularly well to their hosts. Furthermore, there is good evidence that many of the helminths may have been in human hosts for a considerable amount of time and may, in fact, have switched from nonhuman primates to humans a long time ago during the evolution of hominids. (3) One of the factors that affects the severity of disease, particularly in agricultural peoples, is the nutritional factor—especially a high-starch, low-protein diet, where the immunochemical system is unbalanced. In hunters and gatherers, where there is an adequate supply of game animal protein of high quality, one expects the nutritional substratum would be good. (4) Density-dependent factors are more likely to be related to large, stable populations than to migratory peoples. (5) Many of the huntergatherers that we see now have been pushed into areas that are marginal, not only in terms of productivity, but also in terms of their being unhealthy and uncomfortable.

Isaac: It seems to me that in discussing population limits and in postulating a mechanism, say, disease, we must look at it in relation to density. That is, if the Hadza are indeed in an equilibrium situation and disease is taken as the mechanism, then clearly the incidence of disease must be in balance as a result of density-dependent factors.

Washburn: In discussing population numbers, it is well to have a model in mind, for example, one could check the birth rate and sec how fast the population would expand if no one died, and then see how many die of snakebite, how many from disease, etc. One of the reasons for looking at hunting peoples is to throw light on the changes that came with agriculture, when the human population soared. One finds monkeys in areas with plenty of available food, where one has reason to expect rapid population increases—I 6 per cent per year in the rhesus, for example—and yet the increase does not occur under natural conditions. Yet, if food is put out for them, as has been done on Cayo Santiago Island, then the population increases in accordance with the model; even though the natural food supply seemed not to be holding down the population. So the relation between food supply and^ birth rate appears to be a tricky matter.

We should note also that the size of these hunter groups relative to their food supply and habits is very small by primate standards. A group of 80 monkeys is not large, and groups of 180 baboons are observed frequently. So I think we must look for social and economic factors that make small hunter groups of 30–50 efficient.

The data that we are now getting on monkeys show that we have to be very careful in speaking of a species as territorial or as defending a territory. For instance, Steve Gartlan (in press) studied small monkeys on Lolui Island, where they are overcrowded. Here they are territorial and boundaries are sharply marked; hence, this is a territorial species if you look at it only on Lolui Island. Howey-er, a few miles away in Chobe the same species of monkey is not territorial at all. According to Phyllis Jay’s data, rhesus monkeys in the forest are not territorial (Jay, 1965), but in the city-where their ranges are restricted and overlapping-they fight and act territorially. So if a species of subhuman primate can vary so ‘jTeatly in territorial behavior, we may certainly expect a wide variation in man.

TURNBULL: I am surprised by the assumption that any environment should be filled to its economic capacity. In studying a community, the economic aspect cannot be separated from others; and particularly in thinking of the social unit in relation to the environment, we have to think of political and religious aspects as well. Economically, an area in the Ituri forest could support double the number of Pygmies that hunt there, but they need extra space for political reasons-for example, to avoid conflict with neighboring groups.

Let me add that the only years in which the Ik in northern Uganda and the Sudan starve to death are the years when they are cultivators rather than hunters.

ALLAND: I agree with Turnbull that people need space and that this accounts for fission, but to say that there is a religious or economic reason for ecological stability is to beg the question. We have to know how they maintain a population size without infanticide, geronto- cide, disease, or other biological mechanisms.

BIRDSELL: We have frequently discussed hunters as living in marginal areas, and I would like now to single this out as pseudoconcept. These areas are marginal for agriculturalists, but not for hunters. I think we have had enough data here to show that if you took a density value near the maximum of 1 person in I or 2 square miles, such density can be found in almost any environment. In fact, I suggest that these scattered occurences on several continents destroy the notion of an optimum environment, since human hunters do not increase in an environment as we think they should. It may well be that we are facing the problem raised by Turnbull’s Pygmy examples, where density controls are invisible and social, rather than economic.

25c. The Magic Numbers “25” and “500”: Determinants of Group Size in Modern and Pleistocene Hunters

Speakers: L. Binford, Birdsell, Damas, Freeman, Hiatt, Sahlins, Washburn

Hiatt: There has been some discussion in Dr. Birdsell’s paper of the mystical number 25, and the number 500 appears to be another such. We are told that there is an amazing constancy of numbers from the dialect tribe in Australia, centering around 500. However, the range given by Elkin (1954) is about 100 to 1,500 or 2,000. I recorded numbers in nine tribes near the Liverpool River of Arnhem Land. The largest of these—the one I concentrated on and for which I have full genealogies — is about 300. There was a good deal of intermarriage, particularly around the edges of this largest tribe. And it seems to me that in talking about the tribe as a dialectal unit, the important thing is not a mystically determined number but, rather, communication through,regu — lar association. The 1,500 or so Walbiri were divided into four communities, and there were dialectal differences among them. Each community lived separately, but there were contacts between them. Contact and separation in these circumstances are obvious factors in the persistence of common language and the formation of dialectal differences.

BIRDSELL: The basic data were published some years ago (Birdsell, 1953). It was explained very carefully at that time that for Australia, 500 was a central tendency, and a rather wide-frequency polygon showing the range of variation was illustrated. The number 500 was derived from taking early observers’ estimates and breaking them into three categories (anthropologists, careful amateur observers, and seemingly untrustworthy ones). They scattered about as one might expect, and that diagram was also illustrated in the paper. The sample was too small to justify calling the number 500 anything but an indication of magnitude, and I am now inclined to think that the number is probably a little too high. To show what magic there is in numbers, however, Krzwicki (1934) in his book on primitive vital statistics—using different methods and with erroneous tribal assignments—got exactly the same figure. This, of course, is too good to be true, and I do not press the figure.

The tribe as an isolate is based upon a sliding definition. A genetic isolate consists of a group of people who tend to breed more among themselves than with others..:-ow, this is a very fluid genetical situation, since 5I per cent endogamy would make an isolate. Actually, the Australian tribe is a much better isolate. Tindale’s basic datum for precontact intertribal marriages was 14.7 per cent, greater in some areas and less in others (Tindale, 1953). This is ample semantic assurance that the tribe does act as a genetic unit. On the other hand, as numbers decline--even in precontact situa- tions-one would expect the number of intertribal marriages necessarily to increase. Hiatt’s enumerations are postcontact in time.

SAHLINS: Having never lived in Australia, I am somewhat confused by the notion of tribe. Even Radcliffe-Brown, with his Aristotelian mind, admitted that neither he nor the natives could tell where one tribe ended and another began. Perhaps someone would explain how this concept can be used with such statistical rigor.

BIRDSELL: It was pointed out earlier that Radcliffe-Brown never saw a bush native. He was dealing in a very dramatic case of salvage ethnology. The tribal boundaries are a closed system. There are, of course, occasional small errors in the placement of one to the other, and they shift a little. The statistics, however, are good, and in most areas the concept of the tribal boundary is clear.

DAMAS: In regard to the unity of the dialect tribe, I wonder how far that would be congruent with a genealogical pool or marriage universe unit, and just how sharp the edges of these types of units might be in this area.

Birdsell: I have tried to generate interest among linguists in doing dialectical transects, but it would be very hard to tell where to find local groups still functioning in a band context. The layman’s evidence is not very good. Larger tribes do begin to show dialect differentiation. The largest ones arc quite clearly in the process of fragmenting. The Aranda would be one of these, and there are others. So, as you approach 1,000 persons per tribe, distance becomes an isolating factor. Foot mobility and density of communications patterns remain the same, and they simply cannot maintain dialect homogeneity over too large a group. Interestingly enough, this does not seem to be a density- determined function at all. All Australians can walk all the places that they need to walk during the annual cycle.

There are a few cases alleged in which bands on the margins of dialectical groups speak languages that may be intermediate or may indicate bilingualism in some of the older members of these bands. Roger Owen (1965) has pointed out one or two cases reported by Radcliffe-Brown, and I have one or two cases of my own. The natives usually say “These are all-mixed-up ones.” But evidence suggests another type of explanation involving changing band territoriality.

Dr. Damas has asked about the genealogical reality of the tribe. I do not know of anyone who has ever covered a complete tribe, although Professor Rose may have something pretty close to this for Groote Eylandt. In any case, genealogies or no, the tribe does function as a genetic isolate. So if you are concerned with relationships, it is the biological evolutionary unit.

Binford: I think it is interesting in the genetic material that a linear distribution may give slightly different patterns of gene flow, with resultant greater restriction along the linear distribution; while the model used by both Williams and Birdsell is a continuous spaet’ model with hexagonal exposure on all sides. During the Pleistocene, if you take a combination of Leibig’s law (Odum, 1959) and the law of the optimum (Dansereau, 1957), you find that many kinds of adaptations are those that move from one major environmental zone to another. During the time that people are distributed in any one environmental zone, they tend to be distributed optimally within that zone, so that there is a hexagonal distribution of persons. On the other hand, near the terminal Pleistocene—when you begin to find tremendous exploitation of unearned resources along the coastal regions and of resources that allow sedentism at the edges of major biotic zones—this is, in fact, a linear distribution. So we might expect to find a major change in th(> size of dialect groups and in the way that genetic changes might occur during the adaptation to linear or edge-oriented adaptations.

To turn to the figure of 25 for local group size, my wife and I have been concerned with minimal population figures, as estimated from size of life space in bounded areas like the caves of the Paleolithic. We have found a correlation between the presence of tools used to make tools and an average floor space of over 250 square meters. According to cross-cultural correlations developed by both Cook and Heizer (1965) and Naroll (1962), an average living space is 10 square meters per person. This suggests that a limiting factor in choosing a cave as a base camp was the size of the cave, which had to have at least 250 square meters of floor space, or about enough for 25 people.

WASHBURN: Binford has mentioned the changes in population that took place at the terminal Pleistocene. Now, looking back towards the non-human primates, I would like to comment on the two mystic numbers. We frequently assume that as man becomes a more efficient tool-user, the density in any particular area goes up. This is certainly not the case. There are many species of monkeys that live at a much higher density than does man right down until the origin of agriculture. Two hundred animals per square mile is a normal figure for many of the forest monkeys. So we might consider whether, in fact, the hunting way of life may greatly have reduced the density of our ancestors, which is the opposite of the traditional view.

A second point is that the number of human beings estimated for the Pleistocene—ten million is a common t’stimate-is also much less than for any of the non-human primates. We tried to get an estimate for baboons and came up with something on the order of forty million baboons in the savannah of Africa alone—competing, you see, with only a small part of the estimated ten million human beings. Again, there is the suggestion that for some reason the particular way of life of man leads to many fewer people than is the case with the nonhuman primates.

As regards local group size, 25 is much less than for many of the non-human primates. So one needs an explanation for why 25-granting a great deal of variation —is a relevant and fairly constant figure. In the non-human primates, there is, of course, no correlate for the number 50() or for the tribe. There really is something new in this.

I suggest that the adaptation of man is such that it determines the number of small groups, determines mobility between them (involving language, which is also really new), and leads to one species living over an area so large as to preclude adaptation to local disease and diet in the sense in which monkeys are adapted to them.:Man has moved through areas that are so diverse in diet and disease that he has had to adapt culturally, rather than biologically. In terms of diet especially, I suspect that there are very few areas of the world that will support many human beings unless they cook food, carry water, and use other such cultural devices. So my suggestion, in general, is that the success of the mobile hunting adaptation is paid for in non-adaptation to local disease and diet, if we take the monkeys and apes as standards of efficient local adaptation.

BIRDSELL: I think that if we visualize the bipedal niche as involving essentially a partially carnivorous diet, then we can see that such a relatively expensive diet will cause a reduction in numbers. But it may be that this is a good niche because in a number of critical areas in the Old World tropics a little meat will carry one across an otherwise dangerous period of the year. In other words, man gave up density for plasticity and, consequently, remains a single species.

Mr. Gardner was good enough this morning to give me three more groups of generalized hunters in India whose groups are said to be of the magnitude of 25. I do not know whether to ask Mr. Freeman if he still stands by the idea that his butchering sites suggest groups of 25 in the Mindel, but perhaps he will comment on that later. Nonetheless, there is no reason why this group size should not have very great antiquity, since it is presumably a basic minimum social structure. And I suggest that both this and the dialectical tribe can be projected backward in imagination a long distance. The dialectical tribe depends on speech as well as on feet, so one must limit its projection to the range within which one thinks symbolic communication existed.

FREEMAN: The evidence from the site of Torralba, to which Dr. Birdsell refers, is derived from the analysis of materials from a single occupation level. The figure of 25 people was my tentative estimate of the possible minimum number commensurate with the patterned distribution of materials in that occupation, and of course involves numerous assumptions about duration of occupation and other factors. I would not want to go on record as suggesting that during the Acheulian-or even at Torralba alone-the mean band size was 25 active people. But r would definitely say that figures of that order of magnitude and even larger ones are not excluded by the evidence at hand.

25d. Pleistocene Family Planning

Speakers: Deevey, Whiting

DEEVEY: Let me note first the previously unsuspected importance to anthropology and archeology of quantitative data on age at death. It is the most easily observed and, in some respects, the most informative single statistic that one can get in studying ancient or primitive demography. My own treatment of this, published in Scientific American (Deevey, 1960) some years ago, was not very professional, but it is the kind of calculation that is borne out by the papers we have heard here. If one is prepared to assume that over long times and wide areas during the Pleistocene human populations were in some sort of equilibrium, and Dr. Birdsell makes this point seem defensible— and if it is also assumed that about half the children died before sexual maturity, then to maintain the population at a steady state requires that the number of children born per parous female be of the order of four—two to maintain the population and two to make up for the infant mortality rate. Now, the available data indicate that for a large population, under any ordinary circumstances, the mean number of children born during a female’s lifetime is of the order of eight. So it would appear that over a long period of the Pleistocene there was a degree of population control and family planning that one would not have attributed to primitive man until very recently.

We have heard that female infanticide among particularly disadvantaged hunting’ populations like the Arctic Eskimo is, perhaps, the most important mode of limiting family size. r doubt that that is a universal method. There are other devices, including certain drugs that decrease the ovulation rate. And it is not evident on the face of it that, to use Dr. Birdsell’s figures, 15–50 per cent limitation by female infanticide would make up the difference between eight and four.

Given this steady-state system, in which the family size is maintained in approximate equilibrium, one may ask what determines this. Here I think anthropologists have been misled by a professional argument within ecology on the dependence of population density upon food supply. Those who make use of Wynne- Edwards’ book (Wynne-Edwards, 1962) are, I suspect, unaware that it centers around an argument between British ornithologists and a group of Australian economic entomologists as to whether or not climate is the main control on the population density of insects. Much of this seems irrelevant to anthropology, since insects are not vertebrates. r would recommend that anthropologists looking for comparative information on. limitation of population size among vertebrates, especially territorial vertebrates, look at David Lack’s books (1954, 1960). The population sizes maintained are not atypical for primates generally.

It certainly looks as though a modal band size of 25 is reasonable, and I might cite Star Carr as an instance (Clark, 1 954) in which archeological finds support figure for the Mesolithic. And the common tribal size of 500, mentioned repeatedly in this discussion, matches Livi’s data for the mean size of medieval Apennine hill villages (1949). Now, it seems unlikely-in view of what we have heard about the Pygmy, the Hadza, and others-that it was purely the food supply that set these size limits. Social forces (whatever that means) seem to be much more relevant both Jar man and for other vertebrates. The environment, of course, sets upper limits, but these are, in fact, almost never reached. Instead, the limitation falls at a much lower level through such bottlenecks as social interaction, territoriality, and antagonisms.

After a remarkably long period during which we can assume populations to have been in equilibrium and well below the carrying capacity of the environment, something quite extraordinary happened. Food-producing was invented, and since that time the huntergatherers have come to see themselves as disadvantaged relative to others, bringing all sorts of psychological and cultural compensations into existence. This means that it is impossible, strictly speaking, to study Pleistocene ecology by anthropological methods. However, I suspect that most of the anthropologists working with primitive peoples around the world are not working with hunter-gatherers but, rather, with agricultural peoples at various levels of technology. So great opportunities exist for applying much more quantitative thinking to the time and stage of the origin of agriculture.

We seem to know a great deal all of a sudden about the amount of land under cultivation at one time in slash-and-burn agriculture. But none of this is published in any form that an ecologist can use. At least I have been unable to find more than a handful of relevant data to indicate, even roughly, the average length of the fallowing cycle in slash-and-burn practice. The reason, of courst.’, is that anthropolo- gists-although they have learned to count and weigh things-have not yet learned to pay attention to land areas and time cycles. So the perfectly elementary question as to how much land is supporting people is a difficult one to answer, even though it is just a first approach to the question of ecology. This applies with even greater force to those studying huntergatherers. Regardless of what the people say about their boundaries, the plants and animals that they collect do have an areal distribution. And we can never get very far with inputoutput economics or estimates of productivity until we can get caloric data on particular land areas.

WHITING: I want to ask Dr. Deevey about the figure of eight children per mother. What assumptions are made about the spacing of children in arriving at this average? I ask this in order to compare it with the average spacing among hunters, which comes to about four years.

DEEVEY: The figure is an empirical finding for French Canada and large segments of the Brazilian population. Both sets of data are from Lorimer’s original Unesco publication. This takes into account all women subject to bearing children, multiple births, etc.

Whiting: I understand that the average spacing in sedentary populations is about two years, so I assume that the hunter’s average of four years would cut down the total number of children born.

Deevey: That was my original point; i.e., that there are spacing devices at work in hunter populations.

Part VI: Prehistoric Hunter-Gatherers

26. Traces of Pleistocene Hunters; An East African Example

Glynn L. Isaac

Anthropological interest in peoples that subsist solely on hunted and foraged foodstuffs has long been heightened by the notion that these people are in some sense survivors from the primitive condition of all mankind. Today this hypothesis-like the hunting peoples them- selves—is so hedged about by reservations and restrictions, that it no longer has much value. However, the conviction remains that comparative anthropological information concerning hunters is of relevance to prehistorians as an aid to the interpretation of archeological evidence. The range of social and economic patterns observed among the “foraging” peoples of historic times certainly provides a valuable source of data for assessing the degree of contrast between this way of life and that of farming peoples (see Whiting, Chapter 35a, this volume). Comparisons of the behavior of modern “foraging” peoples with that of other primates have the advantage that the subsistence base is reduced as nearly as possible to a common factor for all species. Thus, although simple analogies have in the past generally proved more misleading than useful, broad comparative studies certainly have great importance for establishing perspective In prehistory.

It scarcely requires emphasis that neither the behavioral science of primatologists, nor that of anthropologists can replace archeology as a means of discovering the actual evolutionary pathways that human behavior has followed during the course of its development. However, prehistorians have passed through oscillating phases of attachment to, and divorce from, these other disciplines, and there is real need to consider these oscillations, and determine afresh what kind of interrelationship will be most fruitful for all disciplines.

Darwinian evolutionary theory created an initial phase of uncritical exuberance among prehistorians and anthropologists. Contemporary hunter-gatherers were viewed as “living fossils, “ surviving from more or less remote periods, while Pleistocene archeological evidence was commonly seen as a manifestation of behavior analogous in detail with that of some convenient contemporary “primitive” group. Thus Sollas, writing a general tract in 191 I on “Ancient Hunters,” felt it appropriate to intersperse chapters dealing with Tasmanians, Australian aborigines, and Eskimos with the chapters dealing with the Lower, Middle, and Upper Paleolithic respectively. This kind of thinking has resulted in a strong reaction away from oversimplification and misleading analogy. During much of this century, prehistorians have confined their researches to the study of stone tools and the factual content of the Pleistocene prehistory of the “Old World” has come to be expressed almost exclusively in the technical terms of stone-tool morphology (Sackett, 1965, 1966).

Considerations of activities other than toolmaking tended to become a minor, semi- disrespectable adjunct to the interpretation of selected artifacts. Clearly the results of highly technical artifact studies have only a limited contribution to make to our understanding of the evolution of human behavior as a whole. Hence there has been in the last decade a revival of interest in obtaining, by rigorous archeological field work, actual evidence to be used as a basis for reconstructing total patterns of human behavior during the Pleistocene. Some ofthe requirements and problems of such archeological investigations are outlined in J. D. Clark’s contribution to this symposium (Chapter 30). Lewis Binford and L. G. Fretman, Jr. (Chapter 28 and 27, this volume), have given valuable comments on the role of ethnographic data in modern behavior-oriented archeological studies.

It appears to me that the contributions of ethnography and primate ethology can be usefully divided into three components: (I) Comparative study helps to establish the terms of reference for an archeological inquiry, i.e., specific hypotheses can be formulated for testing. (2) Ethnographic data often provides inspiration for the interpretation of archeological evidence, the significance of which might otherwise be obscure. (3) The application of archeological methods to ethnographic material can test the validity and limitations of archeological inferences.

Elsewhere in this symposium, L. Binford has stressed the point that archeological deductions must stand on their own merit as verifiable hypotheses, regardless of the existence of close modern analogies. Nonetheless, it remains true that the most useful and economic hypotheses will normally be suggested by a knowledge of situations that are more completely documented than prehistoric ones ever can be. Further, there are certain types of semiquantitative information obtained by archeological methods which can only be used as a basis for behavioral reconstructions if they are “calibrated” by reference to contemporary observations. Relationships between size of occupation area and size of occupying group are of this kind, but have only rarely been systematically observed.

It must be emphasized that many topics of interest to anthropologists and primatologists are in general not amenable to archeological investigation. Kinship systems and marriage arrangements, which are prominent subject matter in this symposium, can scarcely be considered in relation to the archeological evidence for the Pleistocene. I t is necessary in planning archeological investigation to consider what questions it is legitimate to ask from the evidence available and then to attack these questions incisively, while remaining alert to other possibilities that may emerge. Washburn (1965) and DeVore and Washburn (1963) discussed certain questions which arose from observation of apparently consistent differences between primate and human social and ecological patterns. These questions were specifically chosen as being potentially answerable from the archeological record. They included such matters as the existence of “home bases,” the size of cohabiting groups, the magnitude of territorial ranges, and population densities. These arc all legitimate meaningful questions and their explicit formulation has helped to clarify some of the issues with which contemporary prehistory must concern itself.

The remainder of this contribution is devoted to an outline presentation of some aspects of a specific investigation, given in order (I) to illustrate some of the features of archeological evidence concerning Pleistocene hunters; (2) to provide examples of the use of comparative data in archeological inference; (3) to indicate some of the archeologist’s special requirements for ethnographic data of various kinds not generally recorded. The example presented is based on the author’s work on the archeological evidence stratified in a “Middle Pleistocene” sedimentary formation at Olorgesailie in Kenya.

Preservation, Samples, and Hypotheses

Detailed investigation of archeological evidence for hominid behavior during the earlier portion of the Pleistocene is only possible in rare localities where favorable geological circumstances have preserved primary patterned associations of fossil artifacts and food remains. Such preservation occurs mainly in relatively moist topographic depressions where continual deposition of fine-grained sediments has occurred. Olduvai Gorge, Kalambo Falls, Isimila, Torralba, and Oloi’gesailie are all similar in this regard (Howell and Clark, 1963). Thus it can be expected that situations with satisfactory preservation of evidence might constitute a very biased “sample” of the totality which they are used to represent. For the Upper Pleistocene subperiod, cave sites probably constitute an analogously biased sample. It is dimcult at p,‘esent to assess the importance of such environmental bias in the evidence, but it has to be kept in mind. The material which we have at present relates directly only to specific topographic situations. To take an example on a large scale-the prominence of East Africa as a sou,‘ce of evidence concerning early hominids is at least in part a consequence of this selective preservation bias. Numerous suitable Lower and Middle Pleistocene sedimentary formatiolls are exposed there and have yielded a wealth of fossils and artifacts in primary contexts. But the hypothesis that the region was the “center” of hominid evolution remains untested because corresponding opportunities for preservation occurred much more rarely in most other parts of the Old World.

The problem of “sampling” at the locality level of investigation can be illustrated by reference to the Oiorgesailie instance. The evidence which is outlined in this contribution was all drawn from study of a stratified series of sediments, the Olorgesailie formation, which were deposited over an area of about fifty square miles in a small basin of internal drainage within the Gregory Rift Valley, latitude 1° 35’S, longitude 36° 26’E (Cole, 1963; Isaac, in press, c). Erosion has exposed the beds and their archeological contents over an area less than five square miles in extent. No archeological material of comparable age has been found on the rocky hills that constitute the dominant topography of the area around the basin. Hence the sites investigated at Olorge- sailie represent exclusively a sample of activities within the special environment of the alluvial fiats surrounding a small lake, and even the sample within this restricted situation is drawn from a small, possibly non-representative segment of the lake basin.

The Olorgesailie sedimentary formation attains a maximum thickness of about 200 feet (60 meters), and represents a considerable time span within the East African equivalent of the late Middle, or possibly early Upper, Pleistocene. However, the cultural material appears to fall entirely within a single cultural- stratigraphic unit (see Bishop and Clark, 1967 for terminological discussion) which is a segment of the “Acheulean Industrial Complex.” For the purposes of this paper, evidence from all levels is combined to constitute, however unequally, the basis for deducing behavior in the area during part of the Middle Pleistocene.

The Occurrence of Archeological Evidence

A very diffuse scatter of artifacts and fossil bone occurs throughout the alluvial and sub-aerial deposits of the sedimentary formation, estimated at less than one piece per 1,000 cubic meters. These presumably represent the wanderings and sporadic activities of Acheulean men. Much higher degrees of artifact concentration are assumed to result from localized activities of men; this we may term an occupation. Three types have been recognized (Howell et al., 1962) at Isimila, and these apply equally to Olorgesailie:

  1. Horizontally and/or vertically diffuse.

  2. Horizontally concentrated, but vertically diffuse (rare at Olorgesailie).

  3. Horizontally and vertically concentrated (an occupation floor).

Types (1) and (2) seem likely in many cases to represent sporadic, intermittent occupations of great duration. Such occupation patterns may not in themselves denote patterns of settlement and movement distinct from the casual wanderings of primate troops (DeVore and Washburn, 1963). Type (3) can probably be interpreted as fairly stable “home bases.” The most spectacular instance at Olorgesailie is a campsite with stone and bone debris covering an area of 200 square meters, and weighing more than a ton (1,000 kilograms).

Table 1. Locations

r-b-richard-b-lee-irven-devore-man-the-hunter-54.jpg

Location and Character of Occupation Sites

Such information is likely to prove a useful basis for deductions about the movements and general natural history of earlier Pleistocene hominids. The Olorgesailie evidence on these points is summarized in Table I.

The choice of sandy substratum in or near the course of a seasonal braided channel is a strongly recurrent feature of the location of most dense accumulations of occupation debris at Olorgesailie. It has been recorded also for a Lower Acheulean site at Peninj, Tanzania (Isaac in press, a and b), and has been reported by J. D. Clark at Latumne in Syria (1966). Some of the Isimila and Kalambo Falls camps may also be examples.

The occurrence of prodigious numbers of handaxes from some fluviatile deposits in western Europe may well relate to similar preferences for sand and gravel patches on the flood plains of rivers. Concentrations of several thousand handaxes in the small area of a gravel pit seem more likely to have been brought about by intensive localized occupation than by natural agencies concentrating initially dispersed material.

At Olorgesailie this choice of sites may be partly accounted for in the following ways: either the sandy channels were bare patches in otherwise swamp-grass covered plains or the lake was two or more miles distant, and water may have been obtainable by shallow scoops in the sandy runnels. The camps may have been situated away from the actual lake shores to avoid the heavy infestation of mosquitoes common near fresh water, and to avoid disturbing game. Systematic ethnographic observation of the location of camp sites in relation to topography and substratum is urgently needed, together with inquiry into the explicit and implicit factors determining choice in each case.

Table 2. Occupation Site Data

r-b-richard-b-lee-irven-devore-man-the-hunter-55.jpg

Size of Occupation Sites

The horizontally diffuse sites may cover considerable areas: for example, approximately 120 X 80 meters for a site at the foot of the mountain; IOO X IOO meters for the thirteen- rock peninsula site; ISO X 120 meters for the LHS scatter on clay soil substratum. These scatters are all assumed to be palimpsests of unknown numbers of occupations, localized only vaguely in a particular area. Their size depends presumably on the interaction of numerous factors-numbers of persons in different occupying groups, number and duration of occupations, extent of overlap with previous occupation areas, etc. At present there is little prospect of getting evidence relating to group size or duration of occupation from such diffuse scatters.

A limited number of “occupation floors” has been excavated, principally on the “Main Site” where eight occur within a strip 120 meters long, 60 meters wide, all stratified within Member 7, a layer of alluvially deposited volcanic silt 2 meters thick. It was found that the “floors” in this complex did not show gradational boundaries, but were discrete patches of stone and bone occupation debris with sharp limits. Such marked site boundaries were not anticipated, and one can only surmise that they coincide with some former natural or artificial limit to the occupation areas. Natural confines might be caused by, for instance, the bank of the sandy runnel in which the occupation was situated, or the edge of the shade afforded by a large tree. A thorn hedge is a possible artificial barrier that would leave no direct archeological traces but would have limited the spread of material. Systematic determination of phosphate concentration patterns over both modern and Pleistocene camp sites may eventually provide the best basis for estimating group size/camp area relationships. The data on the occupation sites are summarized in Table 2.

If we assume that the distribution of stone artifacts, rubble, and cobbles fairly represents the minimum area of a camp site, then it can be surmised that the smaller sized floors (5–6 meters) would result from occupation by a group comprising no more than four or five adults, while the larger ones may have held twenty or more. The problem of the size of prehistoric social aggregates was discussed among the participants of the symposium, and is one that requires more archeological and ethnographic data before firm conclusions are drawn. However, it seems possible that the size range of sites observed at Olorgesailie is consistent with the pattern outlined by Birdsell in his treatment of the demography of precontact Australia: that is, the small, 30-square-meter sites possibly represent occupation places of single family groups, while the larger sites, 200–300 square meters, may represent “home bases” of “local groups” of the predicted order of magnitude, twenty to thirty adults. Lee (personal communication) reports a range of camp sizes from 230 to 790 square meters for groups of !Kung Bushmen between 25 and 30 in number. The mean is 440 square meters (n = 6). There are too many variables for precise equivalence to be assumed.

We have no direct evidence for deducing the duration of occupation in sites of the kind referred to above. However, it is clear that the accumulation of a ton or more of stone artifacts and manuports, as in the largest examples, requires either a continuous occupation of some length (perhaps not less than two or three months), or repeated reoccupations amounting to an equivalent period. The sharpness of the margins and the low degree of rearrangement by natural forces favors continuous rather than intermittent occupation. Lee reports that Bushman groups reoccupy the same area, but seldom exactly the same site.

Food Remains

No traces of vegetable foods have survived on any site at Olorgesailie excavated to date. It might be expected that stones used in the preparation of vegetable food could give indirect evidence. Lee and Woodburn have mentioned “nutting stones” and seed-pounding stones. Similar stones are recorded for Australia (D. Thomson, 1964). The Olorgesailie sites commonly contain an abundance of subspherical stones suitable for use as hammers and pounders, but flat preparation platform stones are not usually obvious. More detailed ethnographic information on the morphology of stones in current use would help investigation of this point.

Bone fragments occur on all occupation sites, suggesting that hunting and/or scavenging was generally practised. The density of bone on the sites is generally low, consisting of small scattered splinters and a few teeth. However, experimental observations suggest that dispersal of bone by scavengers and disintegration by weathering proceed at a sufficient rate to account for this. M. Posnansky, who was warden at Olorgesailie 1957–58, laid out part of the skeleton of a goat on the ground and covered it with wire netting to protect it from scavengers and footfalls. By 1965 most of the thinner bones, scapulae, ribs, and parts of the skull had disintegrated. The remaining bones are so friable that only small fragments would have survived if it were not for the protective cover. In addition, during part of my stay in East Africa, bone dispersal was experimentally observed. Domestic bones and bone fragments were recorded and dumped at a peg out in the “bush.” Scavenging animals removed most of the bones beyond a radius in which is was possible to search (30–50 yards); only relatively small splinters and a few isolated larger fragments remained near the peg (see also Leakey, 1965).

Thus we are not in a position to assess general level of hunting success by Acheulean men camping at Olorgesailie. We do however have some instances where interesting evidence has been preserved. Specialist reports on the bone assemblages are not yet completed, so that this account can only be in general qualitative terms.

Species Represented

Most of the excavated sites have yielded identifiable scraps of the teeth of the following:

Hippopotamus sp.

Equus, and/or Stylohipparion.

Bovidae (relatively scarce).

Some sites have yielded fragmentary remains of the following groups:

Simopithecus, an extinct robust baboon (superabundant on one site).

Suidae.

Elephas recki.

Rhinocerotidae.

Rodent bones, bird bones, and reptile bones including crocodile are present on some sites, but are scarce and poorly preserved. Fish and amphibian bones were found on a few sites, but were not in concentrations sufficient to indicate that they formed part of the diet. Even on those sites where bone is comparatively abundant, it mostly consists of finely comminuted splinters. It seems clear either that the occupants had cause to pulverize bones on the camp site, or that scavengers moved in afterwards and did so.

Some examples of sites on which evidence is comparatively well preserved may serve to illustrate the character of the evidence:

  1. 13, “Peninsula Site”: An excavated area of about IOO square meters yielded forty equid teeth, eighteen bovid teeth, and five hippo teeth, together with about thirty podials and bone fragments which are possibly large enough for anatomical identification. The remainder consisted of 19.2 kilograms of fossilized bone splinters with an average weight per splinter of 1.25 grams. The excavation is a sample of a total area of a diffuse occupation scatter estimated to be nearly 100 times larger. Extrapolation from these figures is unreliable but could indicate that a fair number of carcasses were broken up in the vicinity.

  2. “Hippo Banda Site”: A former warden, Mr. Andrews, excavated in an area 10 X 5 meters a portion of a disarticulated hippo carcass and preserved his finds in situ by covering them with a roof. The bones consist of a hind leg, teeth, vertebrae, and rib fragments. Among the bones are a small number of Acheulean artifacts. No records of the excavations are preserved, but there seems no reason to doubt that this exhibition excavation exposed a hippo butchery site.

  3. The largest and most completely preserved occupation floor to be excavated was that of DE/89 Horizon B (Land Surface 7 of L. S. B. Leakey, 1952), in Member 7 on the Main Site. The occupation concentration is oval in plan, about 19 X 13 meters in extent. I t has yielded more than 1,000 kilograms of stone artifacts, rubble, and unmodified cobbles. Many of the 500 cleavers and handaxes show signs of chipping and blunting by use. Among the stones comprising the floor there was a large quantity of bone splinters and relatively few intact or identifiable pieces, mainly teeth and podial bones. Most of these belong to numerous individuals of an extinct robust species of baboon, hitherto classified as Simopi- thecus oswaldi. The material is at present being studied under the direction of Dr. Leakey. Field records and a preliminary sorting done by Dr. Leakey and S. Mitchell indicate that more than 800 teeth were recovered in my own and Dr. Leakey’s excavations. A preliminary estimate indicates a minimum of fifty adult and subadult baboons, together with at least thirteen juveniles. In addition to more than 150 shaft and articular fragments of long bone and very numerous phalanges, tarsals, and carpals, there were 12.7 kilograms of comminuted bone splinters, most of which is probably Simopithecus bone. With the Simopithecus bones are a few scattered teeth and bones of equids, bovids, hippos, and an elephant; but these probably constitute much less than 10 per cent of the bones.

This site thus attests either continued extremely successful, specialized hunting of baboons, or perhaps the massacre of a single baboon troop. J. Woodburn (Chapter 5, this volume, and personal communication) reports that the Hadza people of north Tanganyika normally hunt singly but occasionally band together to surround a baboon troop at night, while the animals are asleep. The baboons are dislodged by arrow shots and clubbed to death as they attempt to break out. This record provides a plausible explanation of the special abundance of baboon remains on the site, without involving the postulation of elaborate hunting or trapping techniques; but the hypothesis cannot at present be critically tested. Woodburn and Lee have also gathered valuable observations on the procedures for cutting up carcasses, selection of parts for transport to camp, and the choice of bones broken up for the extraction of marrow, but more quantitative data on these processes would be of great interest.

Hunting Methods

There is no direct evidence regarding the nature of hunting methods employed by Acheulean men at Olorgesailie. Recent excavation has revealed nothing definite in support of the hypothesis that the bolas was known to these people. It is hard to imagine how hunters on foot and without bows could kill large numbers of baboons except by surrounding them at night. Drugged waterholes is another possibility that comes to mind.

Positive evidence of fire is absent from the Olorgesailie occupation sites, as it is at all “Middle Pleistocene” sites in East Africa. Microscopic carbon is present in silts of the formation, and it may be that the negative evidence on the excavated sites is not conclusive.

Limited territorial range of the groups occupying sites at Olorgesailie can be guessed from the scarcity of stone exotic to the vicinity. Chert and metamorphic rocks are generally present on the sites only as small worked-out stubs or splinters. The metamorphic rocks do not outcrop closer than twenty miles away.

The absence of any human remains from Acheulean culture occupation sites at Olorge- sailie and Olduvai is in contrast to several occurrences on earlier Oldowan culture sites. It can tentatively be concluded that corpses were removed from occupation sites, and that cannibalism was not an important practice.

As the discussions of this symposium have indicated, changes in demographic patterns during prehistory are of considerable interest to anthropologists and human biologists. Washburn (1965) has deduced from the evidence of a comparative primatologist that hominid population densities declined from a relatively high number (ten to 400 or more per square mile), characteristic of non-human primates, to the relatively low number characteristic of most non-agricultural human societies for which data exists: for example, 1 per square mile maximum for Australia (Yengoyan, Chapter 20, this volume). Most prehistorians (for example, J. D. Clark, 1965) conclude that the record shows a gradual increase in hominid population density through the Pleistocene. Both points of view require more factual documentation, but because of the nature of Pleistocene archeological evidence and the bias of samples, archeology is unlikely ever to be able to contribute unambiguous, direct demographic evidence for the Lower and Middle Pleistocene subperiods. However, there are features of the archeological record which may conceivably give indirectly the required indications: artifact assemblages may be found to show patterns of differentiation which are related to demographic factors and population density, in a fashion analogous to linguistic dialects (see Birdsell, Chapter 24, this volume). Preliminary treatment of this problem for some North American archeological occurrences has been given by Binford (1963). Individual campsite assemblages at Olorgesailie show marked stylistic idiosyncrasies, perhaps indicating stability of group composition. When the patterns have been worked out at this and other sites, tentative demographic deductions may become possible.

Archeologists are frequently under pressure to abandon the shelter of the technicalities and indirectness of their evidence and to risk speculative total reconstructions of the behavior of prehistoric peoples; but uncontrolled speculation is of no value, or is even misleading. Therefore, in accepting such challenges, the archeologist should firmly omit aspects of culture for which he can find no prospect of material evidence to use as a test of validity. Yet, a bold outline reconstruction of the situation actually indicated by the evidence may be attempted, to conclude this example:

Acheulean men at Olorgesailie appear to have lived in groups varying in size from approximately four to approximately thirty adults, the smaller groups possibly being the result of temporary splitting of relatively stable “bands” with twenty to thirty adult members. The “bands,” or individual specialist members, appear to have shown real, perceptible stylistic variation in their artifacts. This, and other observed variation (Kleindienst, 196Ib) might also in part relate to temporary, perhaps seasonal, or sectional task differentiation, which may be reflected in diversity of composition among tool kits.

The people subsisted at least in part on hunted, or scavenged, animal food, but the evidence is consistent with an opportunistic subsistence pattern similar to those observed among the Bushmen, Hadza, and other contemporary “foragers,” where meat is a socially important prestige food, but the diet consists predominantly of gathered plant food. In one instance at least, the evidence suggests systematic communal hunting of baboons. Territorial range was quite possibly rather restricted. The general scatter of artifacts indicates that man was present persistently in the area, but ecologically he was probably no more significant than any other predatory species.

Generalizations from such evidence as that of Olorgesailie to wider hypotheses about a “stage” in behavioral evolution should probably not be attempted yet. Other comparable site complexes in Africa have been investigated and are reviewed by Howell and Clark (1963), though none are published in detail. However, the results of each hold out promise that meaningful paleoethological and paleoecological interpretations are possible where sufficient care is taken in the investigations.

The example given should also have shown some of the uses of ethnographic data both at the level of interpreting site features, and for general hypothetical reconstructions. It should also be clear that the prehistorian often needs a specific type of ethnographic information which can only be recorded by a worker with an “inside” knowledge of archeological evidence. Close field collaboration between archeologists and ethnologists is thus urgently needed.’

I. Much ofthe field and laboratory work ofthis study has been done jointly with my wife. Dr. and Mrs. Leakey kindly gave us the opportunity to work at Olorgesailie, and helped and encouraged us throughout its execution. TheGovernment of Kenya gave permission for excavation, while the National Parks and National Museum provided many facilities during my employment there. Research grants have been generously provided by the following organizations: Wenner-Gren Foundation for Anthropological Research; Boise Fund, Oxford University; British Academy; British Institute for History and Archaeology in ‘East Africa.

27. A Theoretical Framework for Interpreting Archeological Materials

L. G. Freeman, Jr.

This essay discusses the proposition that the most serious failings in present models for interpeting archeological evidence are directly related to the fact that they incorporate numerous analogies with modern groups. This has prevented the development of frameworks of theory which might lead to an understanding of the sociocultural significance of archeological residues based directly on the comparison of those residues. The use of analogy has demanded that prehistorians adopt the frames of reference of anthropologists who study modern populations and attempt to force their data into those frames, a process which will eventually cause serious errors in prehistoric analysis, if if has not done so already. It is unnecessary, because it is possible to develop models for the interpretation of archeological evidence which minimize analogy. It is unscientific, because if we utilize models which are only sensitive to the elucidation of parallels with modern groups, the discovery of parameters of sociocultural structure unique to prehistoric time periods is impossible. Unless we can discover those parameters where they exist, evidence from prehistory will contribute very little to the understanding of ranges of variation in cultural systems, the nature of the interrelationships between elements of culture, or processes of cultural development.

Recent Developments

In the last decade, prehistoric research has attained a new level of sophistication in the gathering and interpretation of archeological materials. The revolution that has taken place is a twofold one, involving the development of new methodological approaches to the gathering and simple description of data (most of which owe a great deal to other disciplines such as physics, statistics, paleontology, paleobotany, geomorphology, geography, climatology, and pedology) and the construction of new theoretical approaches to the interpretation of those data.

While all prehistorians agree that the materials with which they deal represent only a small proportion of the materials used in and altered by human behavior, many would now reject the view Hawkes expressed only thirteen years ago, that without the aid of written records little information except that dealing with past economies can be extracted from archeological evidence (Hawkes, 1954). A brief survey of only some of the modern research that illustrates this trend shows studies exploring the ramifications of White’s (1959) view of culture as man’s extrasomatic means of adaptation and Steward’s (1955) concept of cultural ecology by Binford (1962) and Struever (1966); studies involving consideration of the nature of the socialization process (Deetz, 1960; Whallon, 1 965); examination of the process of cultural drift (Binford, 1 963); studies concerned with the nature of stylistic differentiation of sociocultural groups and subgroups (Binford and Binford, 1966b; Cronin, 1962; lketz, 1960; Longacre, 1964b, Whallon, 1(65), and attempts at definition of the num- her and nature of tasks undertaken by prehistoric groups (Binford and Binford, 1 966b; Freeman and Bro\-Vn, 1964; Freeman, in press). The best of these studies have oeen directed to the isolation and examination of the fUllctional and processual dimensions of cuI turai systems. Much less effort has been spent in the construction of frameworks for viewing the structure of such systems. Even where attention has been given to this aspect of cultural studies, research has involved attempts to determine the existence, in the prehistoric record, of structural principles observable in (especially “related”) model’ll societies. The method takes fill’ granted that it is possible to derive, from the study of a sample of modern societies, elements of sociocultural structure (including whole institutions and corporate groups) which are homologous with those of the prehistoric period. Although this approach may be an especially fruitful one when applied to recently extinct cultural systems, it is likely to yield misleading results when applied to the study of cultural materials produced by more ancient societies, especially societies more than 40,000 years extinct.

The Use of Analogy

In part, the use of analogy in archeological interpretation has been due to a desire to construct categories of cultural development— “levels” of economic organization or social complexity-under the assumption that such constructs are the goal of evolutionary studies, and that the principles of the classification are derivable from our knowledge of the evolutionary process. However, the construction of such categories, which has been called “general evolution” by Sahlins and Service (1960) is really Ilot “evolution” at all, but taxonomy. Multitudes of classifications of the same items, be they objects, organisms, or sociocultural systems, are possible (Simpson, 1961). Some of those are “evolutionary” in the sense of being derived from the developmental history of the items classified, and some are not. To establish the relevance of such a classification to the evolution of the items concerned, one must base it on the historical record of development of the items. As yet, studies of the “fossil record” of cultural evolution are inadequate to serve as the basis for any evolutionary classification that is detailed enough to be useful. I t is impossible to classify as yet ungathered data.

Rut are the data really ungathered? It is often assumed that this is not the case. Admittedly, it is said, the “fossil record” is incomplete, but we can substitute for missing elements in the record studies of the behavior of “modern representatives” of those elements, As Service says, “Certainly aboriginal Arunta culture is not younger than western civilization; it is obviously a great deal older, and precisely thert’in lies one of the virtues of studying that kind of culture” (1962, p. 8). The assumption that modern representatives of past stages of cultural devciopmt’nt exist is a major justification for the use of analogy. Curiously, that justification is a derivative of the view that culture is an adaptive system. As Service goes on to say: “the aboriginal culture of the Arunta ... is ... a form of adaptation to a particular kind of (total) environment made long, long, ago and preserved into modern times because of its isolation” (1962, p. 8; the parentheses are his). This kind of reasoning is misleading.

I t is based, of course, on the hypothesis that like environmental stimuli produce like cultural responses. Jn a very general way, this is true. (There are a limited number of methods of working stone by percussion. Elements not present in an environment cannot be utilized.) Nevertheless, if the sta tement is examined in detail, it is false. Each society exercises some degree of control over the influence of its environment by exploiting some aspects of environment at the expense of others. No society utilizes all it could of the offerings of its surroundings. In addition, the differences in the manipulation of the same resource by two distinct cultures are often great. Two “distinct” cultures from exactly similar environments, both of which are affected by exactly the same aspects of those environments, and both of which utilize identical resources in identical ways, would be part of exactly identical ecological systems. This is really the same as saying they would be one and the same culture. The validity of making inferences based upon general principles of adaptation discernible among modern populations is not denied; on the contrary, such inferences are necessary. But that is not the same as the inferential process I am attacking. In fact, it leads to contrary results.

It is known that modern populations of higher animals and their distributions are the result of a complex historical process involving long sequences of changes in adaptation to changing environments, including other animal populations. The present diversity of such animal forms is the end product of a series of developments involving numerous transitions from old to new environmental situations, either by population spread or environmental change, and numerous consequent readaptations. In addition to this process, the complementary development of a variety of new “ways of making a living ... examplified in the phenomenon of adaptive radiation” (Simpson, 1961, pp. 14–15) also played a large part. Competition for resources resulted either in differentiation of forms, often involving increasing specialization in the utilization of specific resources, or in the disappearance of all but one form from the environmental locus of competition (Simpson, 1961, pp. 16–17).

Sociocultural systems, like animal populations, have tended to regional-and-resource specialization during the course of human history. New ways of making a living have occurred at the same time: one can certainly speak of the dispersal of food-production as an example of an adaptive radiation. Any such radiation alters the interaction between members of the invaded natural community in some way (Simpson, 1961, p. 10). In the case of the spread of food-production, the process of clearing land for planting, among other factors, altered the size and nature of animal communities, and thus altered the possibility for hunter-gatherers in competition with agriculturalists to survive. At the present time, hunting and gathering adaptations tend to exist in situations which are undesirable to food-producing peoples. Where hunter-gatherers survive in environments utilized by food producers, they have usually had to specialize in the extraction of kinds of resources least affected by food-production. They must, in fact, be totally unrepresentative of the sorts of hunting-gathering adaptations that existed before the advent of food production.

Another line of reasoning that militates against Service’s hypothesis is based simply on the logical limits to prediction from a limited sample. Hunting-gathering adaptations of the present are extremely diverse. From a detailed analysis of Bushman cultural systems, it would be possible to predict very little about the social structure of the Kwakiutl. The cultures do have elements in common, of course, but those elements are of such a general nature that information gleaned from one group is not particularly useful in interpreting the behavior of the other in any detail. (It is true that in another sense a great deal can be learned from the comparison. It illustrates the diversity of forms of structural elements among huntinggathering peoples of the present, and the dangers inherent in reasoning from one or a few such systems to all.) Now, useful and detailed analyses of sociocultural systems have really only been made among peoples who lived during the last hundred years. The total length of time during which hunting-gathering adaptations have existed, on the other hand, is on the order of two million years or more. It would seem logical that Bushmen are many thousands of times more likely to be representative of all modern hunting-gathering groups than all such groups of the present are to be representative of the total range of hunting-gathering adaptations past and present. This is especially so because most past groups were composed of beings biologically so different from present humanity that we simply cannot assume continuities (other than such broad ones that they are relatively useless in interpretation) between their behavior and our own.

I have not meant to imply that the comparison of past and present sociocultural adaptations can reveal no important similarities or identities. However, such parallels must not be assumed to exist before it has been demonstrated that they do. The use of assumed similarities with modern behavior in the explanation of the behavior of extinct groups is not only fallacious, it is also deleterious to research since it prevents the discovery that the postulated similarities do not exist.

A Model Minimizing Analogy

! have attempted to establish the fact that analogical reasoning from modern behavior must be kept to a minimum in the construction of models of past cultural systems. I intend to show in the remainder of this paper that the construction of a workable model of the structure of culture, for use in interpreting archeological materials in which a minimum of analogical reasoning is involved, is feasible, and that its application avoids the pitfalls I have outlined.

The Nature of Culture

Any model of cultural structure which is to be of utility to the prehistorian must Consider the material aspects of culture, since those include the observational data upon which he must base inferences about human behavior. 1 t must be assumed, for the purposes of such a definition, that patterned occurrences of the elements the prehistorian studies can be discovered, and that when they are derived from undisturbed contexts they indicate that patterned human behavior was responsible for their existence. It must also be assumed that patterned behavior due to biological factors can be isolated from culturally conditioned behavior, at least potentially. Last, although ideas and values are important to the prehistorian as they influence behavior, values which do not become observable through some effect on behavior need not be considered part of culture. A definition of culture which satisfies these restrictions is the following: culture consists of both the total configuration of patterned activities (which are not >simply referable to the biology of the actors) performed by a society, including the materials used in or produced by those activities, and the social units responsible for activity performance. This definition resembles that of Malinowski (1960) except that the focus of attention is on the end products of his institutions, and the “charter” of the institutions is equated with their “function.”

I stress that the prehistorian cannot reconstruct any activity undertaken by a given society unless that activity produced some preserved material evidence. Binford, on the other hand, has claimed that it is possible to “recover, both from the nature of the populations of artifacts and from their spatial associations, the fossilized structure of the total cultural system” (Binford, 1964, p. 425)’ This statement would seem at first glance to contradict what I have just said. I do not really think it does. Binford does not mean to imply that we can reconstruct an extinct linguistic system, for example, from prehistoric materials. However, the linguistic system as part of the general system of communication in a given society is also part of the mechanism of socialization, and the nature of the process of socialization certainly cannot be denied to influence the patterning of activities in the society, right down to the form of the tools made and used by social units.

The Nature of Social Units

While it is relatively easy for the prehistorian to discern patterned occurrences of elements and to infer from them some of the parameters of the activities which produced them, and at least some of the norms governing their performance, it is much harder to determine the nature of the social units which performed those activities. In this stage of analysis, prehistorians have tended to refer to the patterned materials they observe as the end products of activities undertaken by corporate groups like those observable in one or another modern society. Once more, caution is necessary. In the first place itis unfortunately fair to say that few significant advances have been made from the study of modern peoples, in the ascertainment of “the extent to which the behavior patterns entailed in exploiting the environment affect other aspects of culture” (Steward, 1955, p. 41) since Steward’s formulation of the method of “cultural ecology”; there is really no body of data available in analogy with moderns that can be applied to this problem without numerous intervening assumptions. Were usable data available, even if all extant social groups were found to exhibit a given correlation between social structural type and activity patterns, I am not prepared to admit that it is justifiable to assume that past social groups with many or even most of the same activity patterns necessarily also had the social structural type that is their modern correlate. I would expect to find among extinct cultural systems at least some relations between social structural type and activity pattern that are totally unrepresented among modern societies.

Another criticism of the equation of archeological materials with the activity patterns of corporate groups can be directed at a general confusion about the nature of social groups that is manifest in that equation. Social anthropologists have long recognized that not all social groups are corporate. A corporate group can be defined as one which has a body of collective rights and duties, an “estate,” vested in all members and “activated in diverse situations,” so that it can be said to be a “multipurposive” group (Fortes, 1953; Nadel, 1951, p. 160). In addition it may have longer existence than the life span of any member. AH members of a corporate group may act as a body on occasion for the performance of some activity, or, on the other hand, only some of the members may cooperate as representative of the group as a whole. In contexts where they act as group representatives, they are recognized and recognize themselves as such, and their way of acting and their organization then follow from the rules of organization of membership in the group and its way of acting (Nadel, 1951, p. 161). However, members of a corporate group may cooperate in contexts in which that membership is irrelevant. The structure of a hunting party need not be based upon the same principles as the structure of a composite family, even where all members of the hunting party are members of the same composite family. Some of the dimensions of group organization must vary, at least in the relative intensity with which they are stressed, as the group performs different functions.

Even though social anthropologists have tended to focus their attention on the corporate groups in society, those groups need not be the only important groups, or even the most important ones for the day-to-day survival of society as a whole. Special-purpose groups made up of members of one or more corporate groups, cooperating to perform specific tasks and, perhaps, immediately dissolving after a very briefexistence are the basic units of action in society. They are also the units responsible for the accumulation of archeological materials. While such parties may, in fact, frequently coincide with corporate groups, the prehistorian cannot assume that they do; he must prove that they do, where possible, and this involves distinguishing the two conceptually for analytic purposes. To be of utility to the prehistorian, the definition of the “social unit” must inelude special-purpose groups or “parties” as well as corporate groups. (Since the culturally patterned activités of individuals can be as important to group survival as those of multi-person groups, the most utilitarian definition of the “party” is: allY number of individuals [from I to n], who contribute to the performance of a given activity.)

A General Illustration of the Use of the model

The application of this theoretical framework to the study of prehistoric materials does not produce any spectacular insights about their significance. In fact, its results are not nearly as interesting or emotionally satisfying as the probably greatly misleading caricatures of prehistoric life-ways which have often been derived by the misuse of analogy. I t necessitates the slow and painstaking isolation of regular types of associations of materials, and their formal equation with activity types. Only much later may an attempt at functional definition of those activities, based on the characteristics of artifacts and contexts, be made. Each activity type must be first be assumed to be the result of the behavior of a distinct party type. Next, detailed examinations of the formal characteristics of the artifacts which indicate the techniques of their manufacture and reflect motor habits involved in that process (Binford and Quimby, 1963), combined with microscopic study of variations in their wear-characteristics (Semenov, 1964), and analysis <of the distribution of associated materials in the

clusters may aid in the discovery that ranges of variation of these characteristics overlap for some clusters and an: distinct for others. This will hopefully permit the identification of parties which are multipurposive, or involved in multiple activities. Perhaps membership characteristics of a party may in future be determinable from the recognition of individual idiosyncracies in artifact manufacture and use. These studies in conjunction with an examination of the -configuration of between-cluster spatial relationships and cluster size (the “proxemic” pattern of each occupation [Hall,: (661) may be expected to lead eventually to the discovery of the boundaries of identityconscious social groups.

Conclusions

The system just outlined affords a systematic, objective method for the control of selected culturally significant aspects of archeological evidence for the purpose of intra- and interoccupation comparisons. It makes possible control over activity type, as an example, permitting eventual study of variations in party makeup or size, or of the variation between functionally equivalent units indicative of activity performance by different identity-conscious social groups. Starting with evidence from one site, we may hope to extend these comparisons first to a few other sites, then gradually over the totality of the prehistoric period, as more excavations conducted to recover comparable evidence are completed. This method, an extension of the technique of “controlled comparison” (Eggan, 1 954) to archeological evidence, offers the only secure means of acquiring an understanding of the nature of the types of prehistoric institutions and the mechanisms which contributed to their maintenance or transformation (Eggan, ‘954, p. 748).

It is certainly desirable for all of us, as anthropologists, to work toward increased communication, and to make our findings as intelligible as possible to each other. But no anthropological subdiscipline has yet elicited the laws governing the structure and operation of cultural systems. The idea that prehistorians must interpret their evidence solely in terms of inferences derived from social and cultural anthropology is as fallacious as the idea that interpretations of the behavior of modern groups must be derived from prehistory. Each of the subdisciplines of anthropology studies but one part of the total spectrum of cultural behavior. No segment of the spectrum is any more important than any other. All must be combined if we ever hope to understand the nature of culture in all its dimensions, and, hopefully, from that understanding to derive general laws regulating the structure of cultural systems, their interrelationships and the processes whereby they are transformed.’

I. I am aware that many of the ideas expressed here are the results of the genius of others, especially F. C. Howell, L. R. Binford, ]. D. Clark, ]. Sackett, R. Klein, and, more recently, ]. Deetz and D. Damas. I suspect that the ones I consider original are also secondhand, and that I have simply forgotten where I borrowed them. The total configuration is my own, however. It was helpfuily criticized by R. Klein, C. Merbs, and S. Tax, and students at the University of Chicago, while I was writing the drafts. I am grateful for the advice I followed, and apologize for having ignored the rest.

JI/ote: Sackett’s unpublished paper. entitled “Archaeological Interpretation in the Upper Paleolithic,” delivered to the AAA annual meeting in 1965 also incorporates some of the same elements in a model of cultural systems. This paper did not come to my attention until it was circulated for the Man the Hunter Symposium, after my essay was completed.

28. Methodological Considerations of the Archeological Use of Ethnographic Data

Lewis R. Binford


Any consideration of the implications for archeological interpretation of new ethnographic data on hunter-gatherers requires an examination of the general relationship between ethnographic observations .and archeological reasoning. Only then can the particulars of any new ethnographic data be placed in proper perspective.

It is frequently stated that one of the main tasks of the archeologist is the interpretation of the past and that the primary means available is reconstruction based on analogies to living peoples. Such a view presupposes that our knowledge of the past is only as good as our knowledge of the present and that our reconstructions are valid only insofar as we are justified in projecting knowledge:Jf living peoples into the past. I vv’ould like to take exception to the above-stated position and set as the task of the archeologist and the anthropologist the explanation and explication of cultural differences and similarities. Ethnologists may, by virtue of their particular field of observation, explicate certain cultural forms not directly observable in the archeological record. The archeologist, on the other hand, may explicate forms of cultural phenomena not generally discussed by the ethnologist, although these phenomena may be available to the ethnologist for observation. However, some phenomena, particularly those of a processual nature covering considerable periods of time, may be unavailable for direct observation by the ethnologist.

Adequate explanations are in no way dependent upon the data having been collected within any particular frame of reference; ethnologists can and do use archeological data, and vice versa. Cultural systems function within an ecological field whose structural changes are most frequently not synchronously phased with the life span of an individual. It follows that even if an individual were to devote his life exclusively to observation, he could not be expected to give an accurate or replicable account of the operation of cultural evolutionary processes. We would expect that some explanations for phenomena explicated through ethnographic observation to be found in archeological data, and vice versa. Further, archeological research might be expected to yield explanations for some observations of cultural phenomena made exclusively through archeological data.

If we define an hypothesis as the statement of a relationship between two or more variables, and if both variables are observable in the archeological record, then the hypothesis formulated is testable. It is only through the testing of hypotheses logically related to a series of theoretical propositions that we can increase or decrease the explanatory value of our propositions. The same procedure is the only means available to the ethnologist for generating adequate explanations. Methodologically, then, the archeologist is in no way dependent upon the ethnologist. Archeologists arc dependent for building models upon the knowledge currently available on the range of variability in form, structure, and functioning of cultural systems. Much of this information has, of course, been provided by ethnographic investigation. I t is this background information which serves the archeologist in offering ex- piallatory propositions for some of the differences and similarities observed in the archeological record, many of which may not necessarily reveal differences or similarities between distinct cultural systems.

For example, some observed differences and similarities may be explained by differences in preservation; others may reveal differences in functioll between sites occupied by the same social unit; others may document different occupatioll histories by social units at separate locations. l\{aIlY adequate and accurate explanations may refer to functional relationships between locations, material items, and classes of human activities, etc. However, as I have argued elsewhere (L. R. Binford, 1967) the “interpretation” of the archeological record by the citation of analogies between archeologi- cally observed phenomena and phenomena from a known behavioral context simply allows one to offer his postulate that the behavioral context was the same in both cases. In order to increase the probability that the postulate is accurate, a number of testable hypotheses must be formulated and tested.

Archeologists are not limited to analogies to ethnographic data as the sole basis for offering explanatory postulates; models can be formulated in a theoretical calculus some of which may deal with forms without ethnographic analogs. Archeologists are certainly indebted to ethnographers for providing sources which can be used as inspirations for model-building. The crucial point, however, is that our understanding of the past is not simply a matter of interpreting the archeological record by analogy to living societies as has been commonly asserted (cf. Thompson, 1956, P. 329). Our knowledge is sound to the degree that we can verify our postulates scientifically, regardless of the source of their inspiration. Scientific verification for archeologists is the same as for other scientists; it involves testing hypotheses systematically.

If archeologists and ethnologists are to overcome the limitations of their observational fields and contribute to the general field of anthropology, they must develop methods which will allow explanatory propositions regarding the operation of cultural systems to be tested by both archeological and ethnographic data. The archeologist, independent of the ethnologist, must search for order in the data available to him. After the recognition of classes of order in the record, the archeologist must then develop models that will allow him to relate the archeologically observed phenomena to variables which, although observable in different form among living peoples, are thought to have explanatory value. The only other alternative is to advertise his findings as indicating a new explanatory variable previously not isolated which might still be operative in the cultural systems of living peoples.

For example, archeologists and ethnologists may observe different phenomena to gain knowledge of a common variable. Ethnologists may interview informants and determine verbally the jural rules regarding postmarital residence. In addition they may take a census to determine the degree of conformity between the stated jural rule and actual decisions. These data may then be descriptively summarized and later treated as a case in testing various hypotheses regarding the conditions determining form of postmarital residence.

For many years it would have been argued that knowledge of postmarital residence rules of prehistoric communities was unobtainable by archeologists since it was not “material cuiture.” A Soviet archeologist, Tretyakov, was the first to my knowledge to suggest that there was evidence for the forms of postmarital residence preserved in the archeological record (Tretyakov, 1934, p. 141). His work was summarized and commented upon favorably by Childe (r 943, p. 6). Tretyakov’s argument was fairly straightforward: the form of fingerprints on the inside of vessels indicated that it was females who manufactured pottery. In societies where matrilocal residence was the rule, there would be less formal variability expected in the execution of ceramic designs within any single community than under conditions where patrilocality was the rule, since patrilocality brings about a mixed population of female potters. Quimby (1956) suggested a similar argument to account for observed differences in variability between samples of pottery from Huron and Chippewan historic sites in the Great Lakes. The development and refinement of this particular argument and the perfection of reliable analytical methods have been accomplished by several recent workers (Deetz, 1965; Freeman and Brown, 1964; Hill, 1965, 1966; Longacre, 1 964a, 1 964b; McPherron, 1965; Whallon, 1965). As a result of this work, we can today state within definable confidence limits the postmarital residence patterns of prehistoric communities. Given the discussions at this conference over the nature of the determinants of residence rules and the degree to which patrilocality characterizes “pristine” hunter-gathers, we may find that some of the answers will be found in archeological data.

This case provides a good example of a study in which different forms of information were relevant in the elucidation of a single variable. It also exemplifies another important point- the interpretive model for the archeological data was not based on simple analogy to ethnographically known societies. The model was drawn deductively from several assump- ti0ns and propositions:

  1. Females were the potters.

  2. Homogeneity of cultural expression within a group varies directly with the homogeneity of the group’s composition.

  3. Many formal characteristics of pottery are stylistic and tend to vary with tradition rather than utiliarian or mechanical factors.

The model developed from these propositions states the nature of the expected relationship between two variables: formal variability in items produced at a given location by females and variability in premarital residence of these females. While the points of premarital residence of females of an archeologically known community are not directly observable, this proposition remains a postulate. However, among contemporary groups, or groups whose residence patterns are known through documents, the proposition may be stated as a testable hypothesis. This illustrates one of the functions of ethnographic data in archeological reasoning. These data may be used for testing hypotheses for which information on one or more of the relevant variables is not obtainable through archeology. While models for the interpretation of archeological data may be tested and verified on other than ethnographic sources, it may often be more impressive or scientifically more efficient in obtaining high levels of confidence to make such tests with ethnographic data.

This latter kind of investigation was strongly advocated by Kleindienst and Watson (1956) in urging a kind of inquiry they termed “action archeology.” Watson actually did carry out a study of a living community from the perspective of an archeologist while she was a member of the staff of Braidwood’s 1959 60 Iranian project. Since then there have been increasing numbers of workers studying archeologically relevant data among living peoples. Ascher ( 1 962) studied the Seri; Leshnik (1964) lived with village agriculturalists in India; Richard Gould of the American Museum of .:’\atural History is currently engaged in research among the Australian aborigines. Margaret Hardin, a graduate student at the University of Chicago, is currently studying functional variability in ceramic styles among Mexican potters. Longacre and Ayres recently reported their analysis of an abandoned Apache wickiup (1966) .

From the preceding discussion, we st’e that ethnographic data can play two basic roles in archeological investigation: first, they serve as resources for testing hypotheses which seek to relate material and behavioral cultural phenomena; second, they may often (but need not always) serve as the basis for models of particular social relations which are postulated to have been the context for an observed archeological structure. In the former case, “action archeology” studies are relevant; in the latter, model building and testing can be related to ethnographic facts, but verification of propositions would remain a problem to be solved by the formulation of hypotheses testable by archeological data.

Given this relationship between ethnographic data and archeological inquiry, how can cooperation between the two specialists be maximized for the solution of common problems? In the first place, if archeologists and ethnologists are to work with common problems, their observations must be geared toward gathering data on the same variables, despite the obvious differences in their fields of ob- s(Tvati6n. Second, they must work in terms of comparable sociocultural units. Finally, there must be a free exchange of information between archeologists and ethnologists to achieve the first two aims.

The kind of profitable feedback that can occur through such interdisciplinary exchange is nicely illustrated by this conference. Archeologists have worked too long without the benefits of the understandings of ethnology as to the operation of sociocultural systems. Further, some of the work discussed at this cor:ference was directly relevant to the perfecting of tech- niq ill’S for gaining information on identical yariables. I am referring to the “action archeology” studies of Richard Lee among the Bushmen and Bob Williams among the Birhor. These two studies add tremendously to our knowledge and are two of the most comprehensive in seeking to document the relationships between behavior and the spatial structure of artifacts which would be observable in the archeological record. Such information will be very useful in testing some ofthe propositions of Cook and Treganza (1950), Cook and Heizer (1965), and Naroll (1962) as to the relationships between population size and site size as well as population size and amount of enclosed space required. The site maps and structural details illustrating the internal spatial-formal structure of settlements, documented by Lee and Williams, will be valuable in testing a number of propositions regarding spatial disposition and correlates of social status. With the data now available it would be possible to measure the degree to which proximity between living areas in a settlement is correlated with social distance as measured in kinship affiliations. Exceptions to such correlations and the kinds of contingencies which intervene can be more easily spotted with data like Lee’s and Williams’ available.

On a slightly more critical note, and writing from only a superficial knowledge of the data gathered by Lee, some ofthe observations made and some of the emphases stressed at this conference bear witness to the need for freer communication between ethnographers and archeologists. For example, Lee’s site maps are informative, yet he uses them to illustrate how little of the physical remains of a living people are left for the archeologist to observe. However, it should be noted that Lee made only surface observations and did no excavation. In the discussion of the relative importance of meat versus plant food, Lee’s exact measurements of intake of the!Kung Bushmen is good to have, but it does duplicate many of the points raised and resolved fourteen years ago and reported in An Appraisal ojAnthropology Today (Tax, 1953)’ Linton discussed many of the points raised here and stated:

there are very few places in the world where people live entirely by hunting ... you may remember that the occupation layers of Peking Man at Choukoutien have hackberry seeds (1953)-

I suspect that an archeologist actively engaged in research might have made many different observations than those made by Lee. In his site maps Lee does plot hearths and fire areas but treats them all as the same. Detailed descriptive data on the formal differences in discrete and metrical attributes of hearths used in lighting, heating, and cooking of various kinds of food might have contributed greatly to our understanding ofthe functional variables which must be considered in dealing with archeologically observed hearths. This kind of information might help to prevent the simplistic kind of archeological interpretation recently made by Movius in which the size of the hearth is taken as an index of the size of the social unit occupying the site (Movius, 1966, p. 321).

This critical note is not meant to discourage “action archeology” studies, but it is to be hoped that ethnologists in making such observations might do well to put their observations in the framework of archeological questionasking.

The emphasis by the participants in this symposium on questioning certain propositions which have generally been accepted as truisms has far-reaching implications for archeological research. To my knowledge all archeological theorists who have considered the role of subsistence technology in evolutionary change have used a rather simplistic Malthusian model for population dynamics. The traditional view, superbly dealt with here by Sahlins, has been termed by Boulding (1955, p. 197) the Dismal Theorem. It holds that the ultimate check on population is misery; population will grow until the nutritional level falls and disease brings about population equilibrium. The fresh and challenging viewpoint put forth by Sahlins, and less playfully by Birdsell and Williams certainly brings into question many anthropological cliches about the origins of agriculture and animal domestication.

One of the most encouraging aspects of this conference is that archeologists and ethnologists are moving in the direction of dealing with comparable units. For example, archeologists assume that the size, composition, and spatial structure of an assemblage are jointly determined by: first, the size and composition of the social unit responsible; and second, the form of differential task performance carried out by individuals and segments of the occupying social unit. Such assumptions allow us to analyze in structural terms the’ contents of sites and permit the definition of the different tasks represented (see Binford and Binford, 1 966b). I n all his work the archeologist has available information on the behavior of persons making up either task-specific work groups or local residential groups. Any social unit larger than these are known archeologically only through comparative analysis of differences and similarities in form, composition, and distribution in a generally unbounded universe of sites yielding archeological data. Local groups and task forces are the social units about which archeologists can get information without having to work their data through a fairly elaborate body of culture theory.

In view of this, it is very heartening to hear at this conference considerable discussion on form and cyclically varying composition of local groups. The work of Helm and Watanabe is especially useful in this regard and will undoubtedly serve archeologists as a basis for model-building.

But perhaps the major contribution of this conference for me is the stimulation it has provided; the specific data as they were presented were literally food for thought. At the risk of presenting some of these ideas before—to continue the metaphor—they have been properly digested, I would like to offer an argument which is the direct result of on-the-spot linkages made between data presented here and archeological problems. The following observations are drawn largely from archeological data:

  1. Judging from the scanty information on settlements and from the inferred functions of recovered tools, man’s adaptations during the Pleistocene were accomplished almost exclusively through the use of implements. Implements have been defined by Wagner (1960) as tools which serve to translate or enhance energy exchanges; examples would be spears, knives, digging sticks, atlatls, etc.

  2. Near the close of the Pleistocene and during the immediately post-Pleistocene period there was increased elaboration in the use of facilities (this is Wagner’s term also). Facilities are objects which serve to prevent motion and/or energy transfers—that is, fish weirs, nets, pottery.

Let us add to these observations some points made at this meeting. Hunters and gatherers whose subsistence is largely obtained through the use of implements (Bushmen, Hadza, forest Pygmies, and during the winter the Central Eskimo) are said to be somewhat casual about death and nonchalant in their treatment of the dead or dying. An exception was the Birhor discussed by Williams who hwnt with facilities-nets.

Let us now add one more archeological observation: Systematic burial of the dead and elaborate mortuary ritual are greatly increased on a worldwide basis at the close of the Pleistocene. True cemeteries appear first in the Mesolithic in the Old World and the Archaic in the New World; the Archaic and the Mesolithic are further characterized by a heavy dependence on facilities in their subsistence activities.

It is tempting to make functional linkages between the structure of technological adjustment, nature of status definition, and hence of attitudinal involvement with persons occupying these statuses as expressed in mortuary ritual. Facilities to be efficient require precise placement in space; their effectiveness is dependent upon energy flow and they must be placed so as to maximize the interruption of this flow. Fish weirs are an excellent example.

Facilities further require cooperative labor for their construction and maintenance.

All of this implies that the responsibility for coordination of effort will be assumed and, more important, that stewardships involving maintenance will be established. There would also be a necessary development of rules govcrning access to the facility and the distribution of its yields. Many other elements of role content would become part of newly defilled status positions arising out of the use of facilities at the close of the Pleistocene. It is also suggested that emotional involvements would thus have been linked to interstatus dependencies of an order unknown in societies in which implements dominated the technology. In this new kind of facility-dependent society the death of an individual occupying a structured status position would necessitatc the reallocation of his position to others, the retirement of “debts” to his descendants, and many other kinds of socially defined obligations which may well have been effected through mortuary rites.

If some of the ideas briefly outlined above should prove upon testing to have explanatory value, the coincident appearance of cemeteries in the New and Old Worlds at the close of the Pleistocene (as well as some of the “aberrant” features of the Birhor) might be elucidated simultaneously. The development and refinement of idcas stimulated by this conference will be one of the chief profits of having been in attendance.

29. Ethnographic Data and Understanding the Pleistocene

Sally R. Binford

The following quotation is taken from the paper circulated by DeVore and Lee as part of the announcement of this conference:

It should soon be possible for eq,thnographers and archeologists to meet together and, as an exercise in the deduction of extinct life patterns, realistically compare the occupation sites of living hunter-gatherers with those of prehistoric huntergatherers. It should be clear that this is a “logical exercise, “ and that there is no assumption that living hunter-gatherers are somehow living relicts of the Pleistocene.

At the risk of sounding a discordant note in a harmonious meeting of people working in various subdisciplines, I would like to remark that the pertinence of ethnographic data to prehistoric research involves more than a “logical exercise. “ All of us are here for more than tightening up our intellectual muscle tone and have some rational beliefor irrational faith that what we are saying is relevant to common problems. It is the problem of the translatability of ethnographic data back into earlier Pleistocene times that I would like to discuss briefly.

It is stated in the quotation above that there can be no valid assumption that living huntergatherers are relicts of the Pleistocene. This is unquestionably true. However, I would like to issue a corollary caveat—that Pleistocene hominids cannot be assumed to be simply earlier representatives of living hunter-gatherers. This means that the use of ethnographic data from the Arunta or the Bushmen to elucidate a Lower or Middle Pleistocene way of life must take into account certain differences in the biological and psychological constitution of the hominids involved which have been suggested by recent work in paleoanthropology.

Our typologies of lithic remains from the entire Pleistocene time range have undergone considerable refinement in recent years (see Bordes, 1953, 1961; Kleindienst, 1961a; deSonneville-Bordes and Perrot, 1954–56), and the use of comparable typological units has allowed comparisons over broad geographical areas. Most of the typologies currently used for Paleolithic ‘ assemblages deal with functional classes of tools—for example, scrapers versus points versus handaxes. The further back in time one goes, the fewer clearly recognizable functional classes of tools exist. This might be the result of fewer kinds of tasks or more multipurpose tools in the earlier time ranges, or both.

Besides function, another analytical tool used by archeologists is that of style. Stylistic- attributes are those formal qualities of artifacts that are not directly explicable in terms of the nature of the raw material or of the technology of production (L. R. Binford, 1962, p. 220).

In the archeological assemblages associated with anatomically modern man (Homo sapiens sa/liens) we can discern clearly definable style zones bounded areas in which stylistic attributes are seen to cluster. Such style zones characterize the terminal Pleistocene assemblages of both the Ncw World (Mason, 1962) alld the Old World (Sackett, 1966). The function of style is thought to be a means of either gmup or individual identification with a product or class of products (L. R. Binford, 1963).

No such style zones can be seen in earlier Pleistocene assemblages. I t is a truism that handa.xes from South Africa, northern France, and southern England are remarkably similar. J n the words of an anonymous archeologist, “You’ve seen one handax, you’ve seen ‘em all”; the absence of stylistic differences within major classes of artifacts during the Acheulian is striking, regardless of the wide variety of raw materials used in their manufacture. It has been suggested (J. D. Clark, 1959; Kleindienst, 1961 h) that the selection by an Acheulian group of a subclass of tool within seme functional class of tool (for example, limande versus cordiform handaxes) may yield some information (In style. However, this is very different from style in later assemblages.

The stylistic differences observable in earlier assemblages would appcar to fit the model of cultural drift-the accretion of small changes or variations which give the appearance through time of directional change. Stylistic differences in assemblages from the Upper Paleolithic and later, on the other hand, appear to be the result, at least in part, of self-conscious symbolic behavior.

The point in time and the context in whieh such self-conscious symbolic behavior appears is a present subject of investigation by the writer and L. R. Binford. It is our present as- yet-undemonstrated hunch that the first evidence of style clines appear in the Mousterian. Bordes’ type list of Mousterian artifacts, whieh is essentially a functional one, can be applied broadly-to Mousterian a.ssemblages of the Near and Far East, North Africa, as well as France. But in using this type list in widely separated areas, the investigator is beset by a feding of unease, since attributes not covered by Bordes’ list appear to vary between broad geographical zones. The construction of an attribute list designed to inform about stylistic differences in the Mousterian is a present concern.

The distinction between style zones and clines is important, since a zone is a bounded area formally distinct from adjacent regions. Since Mousterian stylistic variability has not yet been systematically measured, we do not know if the differences we impressionistically perceive cluster in the discrete units which have been demonstrated fi)!’ later asscmblages.

The purpose here of bringing up style and its function, and of differences observable through time, is to raise the question of the implications of new kinds of ethnographic data for archeology. In dealing with pre-sapienJ populations, observations from living groups cannot be directly used. (The exception would be observations on the function of specific artifact forms in similar environments.) O bserva- tions of living groups, to be usable for the early time ranges, must serve as a base line against which we measure differences in the structure of cultural systems and the psychological capacities of the culture-bearers.

30. Studies of Hunter-Gatherers as an Aid to the Interpretation of Prehistoric Societies

J. Desmond Clark

It has become clear to me whilf’ listening to the debates of the two previous days that a better understanding of precontact behavior, social grouping patterns, population density, and distribution in Australia and, probably, also in North America and Africa would be possible by making a closer use of archeological data. A critical evaluation of the archeological evidence that can be obtained from late or historic settlements, supplemented by oral tradition, provides a promising means of resolving some of the problems of conflicting data as between present-day and early-contact times and also for the immediately previous period.

If archeological evidence has hitherto been little used by the ethnographer, the reverse is also true of the archeologist in his use of ethnographic data to help in interpreting the way of life and behavior of prehistoric populations. However, the evident flexibility of the social unit and the lack of any consistent composition or general grouping patterns among huntergatherers, as demonstrated at this conference, makes the archeologist’s task even more difficult than it already appeared. Whereas the dangers of adducing any direct archeological and ethnographic comparisons are generally recognized and avoided, the basis for reconstructing prehistoric behavior must lie in a systematized use of the whole corpus of information on human behavior at the hunter-gatherer and other subsistence levels, and the selection and critical application of what is relevant in this for archeology.

The more complete the evidence for length and area of occupation and the associations of the full range of cultural remains, the greater the validity for any attempted interpretation of the site and its contents. The older the site, the more complex any reconstruction becomes.

Only during the last ten years or so has systematic excavation and study of artifact distribution on Lower Paleolithic living floors been undertaken in the Old World. Such studies have shown that it is quite possible to find sealed occupation sites that have suffered little or no natural disturbance before or after burial. In such cases it is presumed that the distribution scatter of the artifacts and any other artificial features of the site is as it was w’hen abandoned by the occupants. Theoretically, therefore, it should be possible to go a long way toward reconstructing the activities of the occupants and obtaining some knowledge of their general behavior. In practice, however, what remains unknown and the yariables among what is known, make any attempted reconstructions at present highly tenuous.

Prehistoric Settlement Sites and Survival Patterns

Leaving aside butchery sites, I know of only one (Olorgesailie, Land Surface 8)-perhaps two (Olduvai, FLK I)—excavated Lower Paleolithic occupation areas that can claim to represent complete occupation units. All others comprise excavation of only a part of more extensive units, the exact areal distribution of which remains unknown. In not a few cases this can never be known since it was the natural erosion of the site that revealed it in the first instance. Clearly, therefore, the first requirement is the excavation of complete units of settlement, both those where there are no natural features limiting the spread of occupation debris (open sites) and those where the area of occupation is naturally confined (cave and rock shelter sites).

Once a sufficient number of distribution plots of complete occupation units are available, it will be possible to see the extent to which the association patterns of the components of a site vary or are consistent-the associations between different classes of tools, of certain tools with food debris, hearths, working areas, ctc. Such different activities as, for example, required the use of the La Quina scraper, denticulated scrapers, handaxes, cleavers or core-axes, may sometimes all have been carried on contemporaneously on one site, as at the Kalambo Falls during the late Upper Acheulian. At other sites, the range of types is restricted, suggesting a special rather than a general-purpose occupation. At present the interpretation is hampered by the narrow lateral spread of most excavations and permits of more than one explanation for the several limited but distinctive tool kits of Paleolithic times.

In nearly every case, also--and this is most important-it is impossible at present to give an accurate estimate of the duration of time that is represented by one of these concentrations of tools and lithic debris. Since there is rarely, if ever, any accumulated midden waste present, either the sites were occupied for a short time only or their antiquity has resulted in the complete disappearance of organic remains, except bone. There is reason to believe that both explanations are pertinent. It seems probable that a site such as the Upper Acheulian occupation at Broken Hill represents only temporary occupation. The same seems likely for the sand bank sites of similar age at the Kalambo Falls, whereas thirty feet or so of stratified succession at the Cave of Hearths represent interrupted but recurring occupation over a long period of time. We are, however, still without the means of determining whether days, weeks, or years elapsed before most of the sites we have were finally sealed over. In this connection it is heartening to know that Bushmen and Hadza today rarely reoccupy exactly the same camp site unless this be a cave or rock shelter.

Again, at most Paleolithic sites only the stone artifacts survive, though sometimes, especially in temperate latitudes, bone is plentiful. More rarely, wood is preserved, but there are very few sites where stone, bone, and wood all occur together. Only a very small part of the material culture of a Paleolithic group is available, therefore, as a basis of reconstruction of group behavior. However, when this is added to the complete or partial floor patterns, it amounts to considerably more data than were generally available ten years ago.

As a result of coordinating work in a number of different disciplines, it is now possible to reconstruct with some accuracy the paleoenvironment of the occupation site and it is only once this has been determined that interpretation of the cultural data becomes at all meaningful.

Obviously, therefore, the living sites selected for investigation need to be very carefully chosen with the reasonably certain knowledge that much of the ecological data will be forthcoming and that the occupation areas are undisturbed and in primary context. Such a combination of prerequisites is not readily come by, and this indicates the necessity for systematic survey of likely regions in the main environments occupied by early man.

With the best possible conditions of preservation, therefore, the prehistorian starts with the same kind of evidence as is preserved on the settlement sites of present-day hunting and gathering groups after these have been abandoned for a sufficient length of time for the more perishable remains to have disappeared. However, if we can reconstruct the environment, we can then build models for the potential food and other resources, vegetable, animal, water, etc., available to the human population. Then, by applying the adaptive data from extant hunter-gatherer groups to the specific archeological occurrences, we can attempt a hypothetical estimate or estimates of behavior, group size and density.

So far as Upper Paleolithic and later sites are concerned, since they represent the living places of Homo sapiens, we are in a better position to interpret the behavior of the occupants. For the various “paleoanthropoid” populations, however, meaningful interpretation of the cultural data is dependent also upon a knowledge of the physical and intellectual capabilities of the makers, and these are but imperfectly known.

These are some of the disabilities under which the paleoanthropologist has to work.

For historic and late prehistoric sites, good results can be expected for interpreting social behavior if the excavated evidence can be directly related to oral tradition and to the behavior of an extant population. From this it may prove possible to establish for different habitats a consistency between archeological patterns and behavior. If such’ consistency exists, attempts will be made to project this back into the more distant past. The validity of such extrapolation for interpreting human behavior at the Paleolithic level will depend on a demonstration of significant consistency in distribution and association patterns. At the same time, there must be a related investigation into the various agencies responsible for the distribution and disappearance of campsite debris. Probably the best way to start is by studying the abandoned settlements of extant populations and observing what happens to these over set periods of years. In some regions such as South-West Africa, where the habitat is dry or semiarid, reasonably good results can be obtained since it has been my experience that in such areas stone is used for many more purposes connected with the construction of the camp and with hunting practices than is the case in tropical regions of higher rainfall where heavier timber is very often used to the exclusion of stone and where, in a few years, organic and inorganic agents will have removed practically every trace of the settlement. Generally, however, evidence of dwelling floors, hearth stones, enclosure walling, pits, etc., is preserved and if, moreover, the population is still a stone-using one, much extremely valuable information must result since the existence of lithic material can provide a direct basis for comparison with the prehistoric settlement.

This emphasizes the urgent need for “ethno- archeological” studies of such extant “Stone Age” groups while they still exist. When studies of the Baines Mountains OvaTjimba groups in South-West Africa or of the Bindibu and other groups in Australia are completed, they should provide a wealth of new data which can be used as a basis for establishing a more precise reconstruction of behavior at prehistoric sites. There cannot be many such groups surviving, but there must still be several in Australia and South America, more than one in Africa and, perhaps, in the Arctic also. Studies of peoples who are known to have used stone until fairly recent times can also be expected to produce not a little data regarding techniques-for example, the method of manufacture of honematting needles and arrowheads by the MaKwango Bushmen, or the punch technique behind the gun-flint industry among the Chokwe and Lunda in northeastern Angola.

Where the group studied is no longer using stone, there is so much less permanent debris remaining on an abandoned site that it may be douhted how much direct use the settlement data are going to be to the prehistorian. Personally, I think the fact that most of the evidence disappears after a few years does not matter too greatly if we know the basis of the pattern while the site is in use and the agents and manner of dispersal of what remains after the site is abandoned. The sharply defined limits of the sleeping area, the spread and limits of h€arths, of the waste resulting from special activities-the peeling of water lily bulbs, the cracking of rnongongo nuts, the shaping of a digging stick or a bow stave, for example-will help to give an approximate indication of “spread” of debris which can be compared with the spread of factory waste and tools, hearths, sleeping areas, etc., on a prehistoric living site and with the dispersal of debris by natural agencies. It is most important to have controlled data from observation and experiment here.

There is a large number of questions to which the prehistorian hopes for an answer from sllch studies as those on the Kalahari Bushmen by Lee and DeVore, on the Hadza by Woodburn, 01” the Mbuti by Turnbull. Is there a correlation between habitat, the dimensions of the settlement, human activity and season? Can we, for example, distinguish dry-season from wet-season camps? On this basis can one compute the approximate size of the occupying group? What alteration to the habitat results from man’s presence at the site over different periods of time? What is the rate of disappearance of perishable occupation debris in different ecosystems and is it possible to obtain any indication of age of an abandoned site from the differential weathering of bone or other materials? Does an abandoned occupation area consistently show an increase in phosphate and calcium content of the soil? If so, how long does this persist? To what extent is artifact and food debris trodden or otherwise carried down into the soil of the occupation area? If settlements are reoccupied, is there any discernable and, however slight, any vertical separation of the artifacts of later and earlier seasons? Is there any consistent pattern of weathering or absence of weathering of bone, ivory, or wood that is quickly buried or suffers long exposure? Is there a consistent pattern behind the different activity areas? Does each family or individual social unit work within the confines of its own habitation area, or are activity areas randomly or sometimes intentionally selected? For example, is bone waste regularly piled and are there regular butchery places? What is the rate of accumulation of debris that can be expected at such sites and what is the rate of dispersal by natural agents, man, and animals? What activities are associated with hearths (stone working, shaft straightening, cooking, sleeping, etc.) and what kind of disturbance or modification of the hearth results? Significant hearth pattern differences must exist as between the cold temperate and subarctic regions and the tropics, so that it is necessary to know these and also the rate of dispersal of charcoal on different kinds of settlement areas under specific conditions.

Some of these queries have been in part or in whole answered at this symposium, but whether it is practicable to obtain answers to others and similar questions will depend, I suggest and would particularly stress this, on the cooperation of ethnographer and archeologist working closely together in the field. It may be that many of these questions cannot be resolved, but until the attempt has been made we are not in a position to know one way or the other. In due course, it should be possible to combine the special knowledge of ethnographer and archeologist in a single individual, but I believe this needs to be tackled as a special training project that has yet to be formulated.

Technology

Much can also be learned regarding the technique of working different raw materials and the reasons behind the selection of some materials to the exclusion of others. Especially valuable are the data from groups still using stone tools. In these cases it is possible to know how many tools are made for a particular purpose, how many individuals cooperate in this, how many of the tools that are made are actually used, what the variation is in shape or form of a single class of tool, and what is the intensity of use. Are the tools left on the site or are they carried away? Are tools made in one year reused on reoccuIJation in another year? What are the methods of use of different tools -how are they held, what muscles are used, and where and when is the force applied? Are tools used for more than one purpose or in more than one way and, if so, is this a regular or an unusual practice?

Such inquiries inevitably lead on to a study of the wcar patterns that result from use of the tools. Tools on prehistoric sites in primary context show fracture and damage that is the result of use. Although as yet it is only the more obvious damage that has received investigation, a great deal more is likely to be learned from systematic studies of edge damage -for example, it should be possible to isolate particular patterns of wear. Comparison of these with the wear patterns on stone tools, or even on iron or bone tools, in an ethnographic context and with the damage on specimens resulting from controlled experiment, should throw appreciably more light on the functional nature of Paleolithic tools. After the initial disastrous essays of the late nineteenth century, archeologists have rightly fought shy of functional interpretations, but the time is now ripe for new attempts to be made.

Historical Sources

There are many other matters having a more general bearing on the interpretation of the behavior of prehistoric hunting and gathering populations that the new “ethnoarcheological” studies may clarify. Besides those having to do with the relationship of the human group to the animal and vegetable resources of the habitat, both for food and utilitarian purposes, and the degree to which the human population is independent of the seasonal availability of these, it would be particularly valuable to know what are the changes that can be seen in the technology and the settlement pattern when a hunting and gathering group accepts domesticates, thus changing its economic basis, as some of the Botswana Bushman groups are doing today. What are the factors that bring about permanent settlement at the hunting and gathering level and what is the importance of freshwater and sea foods in this connection?

Most of the hunting and gathering groups existing today are living in some of the least favorable habitats and have for long been in contact with more complex societies and technologies. They can, therefore, no longer be considered “typical” or useful for any direct comparison with prehistoric populations of optimum and favorable, or even of the marginal, ecological zones; yet a knowledge of their technology and behavior is of general significance for an understanding of human behavior at this economic level. Moreover, it should still be possible to supplement the data obtained from direct field studies by searching the historical sources,jf these are re-examined with the specific problems in mind that such studies are trying to answer. Much exists in the literature on Indian populations in the New World, and the same can be said also for Africa so far as the Bush, Hottentot, BaTwa, and Pygmy peoples are concerned.

Coordination of the results from fieldwork among hunters and gatherers undertaken by ethnographers and archeologists working together, and supplemented also from historical sources as well as from comparative study of edge-damage patterns on tools, should broaden considerably the basis for a more precise interpretation of the behavior of prehistoric societies as revealed by the excavation of selected living places. If and when a critical comparison of these results, with the archeological data from undisturbed living places establishes consistencies in patterning and grouping that are common to both, the framework for reconstruction will exist. The operation, as I see it, is twofold: first, to obtain this “ethnoarcheo- logical” data by field work; second, to apply it to the archeological evidence with sufficient precision that a reasonable measure of accuracy in reconstruction is ensured.

31. Discussions, Part VI

31a. Hunters in Archeological Perspective

Speaker: Deetz

DEETZ: This statement probably represents a viewpoint somewhat different from those of the other participants in this symposium, since I am neither a Paleolithic archeologist nor have I had field experience among living hunting peoples. I suppose there are few subjects within anthropology as far removed from the study of hunting and gathering peoples as European Colonial North America, the area in which my major research interests lie. Yet as an archeologist, I feel that there are certain things which I might profitably say that are germane to the subject simply because they are based on archeological and ethnological theory in general, as well as being concerned with hunters in particular. Many of the remarks which follow should be taken as attempts to tighten up our view of hunting peoples; they are not meant to be as critical as they might seem. In many instances the position taken is that of devil’s advocate.

The topic of the symposium-Man the Hunter-suggests that there is a utility in the concept of hunting man as opposed to man involved in other activities. While I am certain that all of us use the idea of hunters and gatherers to signify a certain cultural level, both in archeology and ethnology, I am equally certain that none of us would deny that grouping hunting peoples under one head creates some strange bedfellows. Ifhunting and gathering is taken to indicate the total absence of any type of productive economy, then we can include Kwakiutl, Asmat, Eskimo, Chumash, and even Arapaho under one heading; certainly these five examples exhibit a great variety of cultural types and share in few characteristics other than a common economic base. The grouping excludes the Wampanoag, who even fertilized their corn but whose social organization is similar to most northeastern hunting peoples; the Ge, who cultivate, but who are more than casually similar to the Australians in many aspects of their society; the Siriono, who are admittedly a refugee group and not exactly typical of anything; and most of the peoples of the Chaco, who cultivate, but place great emphasis on hunting and gathering, and share many aspects of social and ritual culture with those groups usually thought of as hunters.

Since hunting, even among those peoples who we normally think of as hunters, is perhaps of less importance than was earlier thought (see Lee’s remarks, Chapter 4, this volume) it really becomes reduced to the status of a single occupational activity, such as gathering, pottery making, or house construction. It might then be equally profitable to consider Man the Gatherer or Man the Potter, and in the latter case, grouping those cultures which possess ceramics just possibly might produce sets with more culture traits in common. Of course, this is an over-literal interpretation of the title of the symposium, since hunting is taken as a convenient label to attach to a group of cultures which do share in many things, as can be seen from looking at the program and reading the titles of the various papers. However, I think if we reduce this to a semi-absurdity such as that above, a purpose is served in that we do not tend to let the label over describe the subject. The numerous suggestions for alternate terms for hunters, such as foragers, indicate this concern among many of us for a more descriptive category.

The symposium is divided into a number of sessions devoted to specific problems; I will comment on these from a perspective which is largely archeological, though not entirely so.

The Subsistence Base

Determination of the basis of subsistence from archeological material is perhaps the most satisfying of all of the aspects of society under consideration here. In this case, we can excavate the remains of food, as well as the tools used to obtain it, and make quite specific and quantitative statements about who was eating what, and when.

As a word of caution, however, it has occurred to me, as it probably has to many others, that there is a peculiar difference between the evidence of hunting and the evidence of gathering in terms of what is preserved. Bone and stone are the normal materials which survive from the Paleolithic. These materials also form the bulk of the evidence of game pursuit at that time; combined with the cave paintings of the Upper Paleolithic they give us a clear picture of hunting activities and technology.

If ethnographic example is in any way trustworthy, and in this case I think it is, the tools used in gathering, and the plant foods gathered, are quite perishable, and except for rare cases of carbonization or excellent preservation of organic substances, would not be present in midden deposits of the period. In view of recently gained understandings of the ratio between hunting and gathering in the subsistence base of living hunters, might we be overemphasizing the role of hunting during the Paleolithic? Perhaps not, but the caution voiced here may be in order. It can be argued that the cave paintings of the Upper Paleolithic show a great concern for animals, and suggest in some way a high interest in the hunt. But if I were faced with the choice between drawing a stag or a mammoth on the one hand, and a seed or a root on the other, I am certain what decision I would make-a subjective, ethnocentric remark to be sure, but perhaps relevant.

The Social Unit

Until lately, statements regarding the specific nature of prehistoric social units were almost impossible to formulate. There has been recent interest in social reconstruction which has resulted in our ability, or at least potential, to make such specific statements about certain aspects of prehistoric social organization in a way not available earlier. The symposium at the national meeting of the American Anthropological Association in Denver, which addressed itself to problems of social organization in the past, focused attention on these innovations in archeological method.

In a Paleolithic context, where the most pressing and important problems concerning the social organization of hunters probably exist, these new methods hold promise of developing models of the sort discussed by Lewis Binford to detail the structure of Paleolithic societies. The caution ‘which I would suggest in this case is that of making certain that we have very precise chronological control before making too many statements about patterning in assemblages as it reflects social patterning in a synchronic way. As Roger Owen suggested to me in a conversation last fall, it would be most desirable to know that the assemblages which we were comparing were within 25 years of each other in time, since this period represents roughly one generation. If hunting populations were as fluid in the past in terms of membership and sharing of ideas, and the limits of such sharing, more than one generation could well introduce differences which might be seen as a function of synchronic variation but which are in reality a matter of chronological separation. Now I don’t think we can ever get ’25-year accuracy with regularity, if at all, in the Paleolithic, but we must be careful that we do not compare assemblages which might in fact be hundreds or thousands of years apart in time.

I heartily agree with Binford’s emphasis on viewing prehistoric cultures as integrated systems, but such integration automatically assumes a large measure of contemporaneity. Perhaps there are already at our disposal means of determining that a single Paleolithic stratum, for example, represents a short interval of time, or at least that there is some depositional unit which can be seen in such a way. I honestly don’t know, but from what we know about trash-deposition rates based on studies such as Lee’s and DeVore’s of the Bushmen, it takes time for material to accumulate. Of course, caves have a certain built-in “trash container” function which does not prevail in open-air situations.

In this connection, I would suggest that features, especially structures, with their associated artifacts are considerably more dependable in assuring the archeologist of contemporaneity than stratified middens. Deflation, for example, can combine separate midden deposits into a rather homogeneous whole. Structures, on the other hand, such as the spectacular houses at Kostienki and other south Russian Paleolithic sites (Klein, I g6sa, Ig6Sb), are behaviorial units wi th a set of contemporary objects within. Thus in inspecting the contents of a single structure, we can be fairly confident that the associated assemblage was all in use at one time, if not made at the same time.

In any case, all inferences drawn regarding social units in the Paleolithic whieh are based on comparing patterns from assemblage to assemblage must ultimately rest on the contemporary nature of these assemblages for their accuracy. I do not mean that such an approach is not advisable; quite to the contrary, this method promises exciting advances in the direction of sharpening our understanding of social organization in the Paleolithic. I am only concerned that we are as careful to control the dimension of time as we are in our control of form and space.

The problem addressed in a part of the Binfords’ statements (Chapters 28 and 2g), that of the validity of using ethnographic data to assist us in our reconstruction of earlier hunters’ social units, is a genuine one, and they have made explicit some of the ramifications and possible pitfalls involved. I would like to call attention to another aspect of this same problem which might have implications that go beyond those involved in archeological reconstruction. Even if we were to agree that there was great value in using the Bushman, Eskimo, or Pygmy as typical hunters, and possible survivors from an earlier time, we must ask how typical these people actually are of the type of life which existed before the initial years of the Neolithic revolution. To put it another way, of the hunting peoples in existence today, which might be thought of as most typical of pre-Neolithic society?

I think that there is a possibility that the closest approach to Paleolithic social units which exist in the ethnographic present is to be seen in the Australians, the Ge, and to a somewhat lesser extent, aboriginal Californians. This rather strange selection is based on certain logical considerations. If we consider the hunting peoples of the world in the time immediately before widespread European influence (sixteenth through nineteenth centuries), there are three areas in the world which were almost absolutely immune from influence and pressure from cultures based on a productive economy -areas where hunters had lived for millenia in a stable relationship to their environment. These areas-California, eastern Brazil, and Australia-are the largest such continuous regions where the influence of the Neolithic or formative simply did not penetrate. I can not think of any other comparable area which has had such long isolation, with the possible exception of the Arctic. The Eskimos are a rather specialized case in many ways, however, and in any event, are relative newcomers to their area.

All other hunting peoples of which we have ethnographic knowledge were subject in some way or another to influences from developing Neolithic peoples from as early as the eighth millenium b.c. in many cases. The Bushmen, for example, were under the direct influence of other Africans prior to the arrival of Europeans, and the marginal and encysted nature of the other enclaves of hunters around the world probably came into being prior to their contact with European culture. It is only in the three areas mentioned above that we can confidently postulate long-term undisturbed development of hunting peoples, free of any disruptions from without. The Californians were cut off by the Basin and deserts; their nearest farming neighbors were encysted themselves. The Ge seem to have been quite free from major influence from tropical forest farmers to the north; while manioc is cultivated by some of the Ge groups, the main cultivated crops include maize, beans, and squash, a very different complex from that found among their nearest neighbors. Furthermore, Ge mythology includes origin myths for these latter three plants, but none for manioc. The additional cultivation of Cissus by these people, unique in the world, points to their long isolation, and independent activity in matters pertaining to cultivation. The Australians of course, have been isolated since the end of the Pleistocene.

There are certain similarities seen in all three of these groups in the form of their social units which seem to be more than coincidental. Most striking is the existence and strong development of moiety organization in all three cases. The correspondence of this feature in Ge and Australian culture has often been commented on, and moiety development is marked among many California Indian groups. It is assumed that moieties serve the purpose of providing the individual with a clearly defined sense of social orientation wherever he might find himself over a broad geographic area. Thus in the case of thinly distributed populations, and small social groups, a person can immediately align himself with this or that social unit even when he enters a strange community for the first time.!-vloieties are also found in more complex cultures, but their function is somewhat different in these cases. But the only instances of strong development of moieties among hunting peoples that I have been able to discover are the Ge, AlIstralian, and Californian cases.

Is this a coincidence? Could it be that these three large isolates have retained a form of social organization which was much more widely distributed before the Neolithic? Hunting populations living in the same environment for millenia might be expected to evolve such a structure, given a large amount of stability, with no outside influence, since it serves a valuable social integrative purpose among thinly distributed populations. All other hunting groups, having been more or less moved about and influenced by expanding Neolithic culture, might not have been able to retain such a complex structure. In one sense, such an organization is a spatial phenomenon and depends on a rather formal and fixed spatial relationship between the people and the environment. Should frequent movement become common, with some environmental change, such a delicate and complex institution might be expected to weaken and disappear.

If the Ge, Australians, and Californians are indeed more typical of pre-1\‘eolithic social structures than those groups which we conventionally think of as hunters, with some form of “patrilocal band” organization, orsome similar type, then we might consider shifting our concept of a modal hunting culture away from the Bushmen, Pygmy, and Eskimo and in the direction of these somewhat more complex societies. While Service suggests that the composite and family-band types defined by Steward are the result of pressures brought to bear by Europeans, perhaps the patrilocal band is also such a contact phenomenon, although the superior cultures would be different in this case. If this is so, most of the hunting peoples of the world are very little like those of the Paleolithic, and only a very few of them retain even a vestige of what may have been a very complex social form in the time before the Neolithic.

Population Reconstruction

The determination of population density from archeological data is less precise than the reconstruction of at least a portion of prehistoric subsistence bases of early hunters, but is still a relatively productive and valuable approach to understanding these cultures. This is particularly so when one has available archeological information concerning houses, and some idea of household sizes within a community. In many ways, some of the most stimulating results, to me at least, of the recent work by Lee and DeVore among the Bushman are those which indicate just how many people are needed to create a recognizable archeological unit. In this case, it would seem to be more people than the average Bushman band contains.

I was talking to Richard Lee about the problem of “archeological visibility”—that is, how much activity is needed to produce a site which could be found a long time later. We decided that there might be a value in talking about some kind of a threshold value which would represent the amount and type of activity needed to produce a visible pattern on and ultimately in the ground. It seemed that two factors are involved at least. In the first place, a quite large community would make less of an imprint on the archeological record if it were more mobile than a smaller populatioji which maintained occupancy of one location over a longer period, or reused it periodically. The second factor is absolute community size, and the two combine to create an archeological impression which is either ahove or below this threshold. The threshold would then represent that combination of community size and use of a given locus of sufficient intensity to yield a recognizable dfect on the locus after final abandonment. The Bushmen seem well below this threshold, and many other hunters are probably below it as well. If in our studies of known hunters in the ethnographic prescnt we were to consider how they relat<: to this threshold value, we might be ahle to come up with some statement of minimal “archeological visibility,” which could then be used with advantage in our archeological interpretation. The Californians are certainly ver-y visible, as are the Upper Paleolithic inhabitants of much of Europe. Less visible, except at their kill sites, are the pedestrian hunters of the Plains prior to the coming of the Europeans with their horses, or the Shoshonean peoples of the Great Basin. Presumably these groups are well above the threshold, and the Bushmen are below.

I am concerned about those peoples in the past who are invisible in this sense, and whether many modern hunters are more like them, or closer to those cultures who left their imprint on the landscape in a wide and deep way. If they are, then those cultures which have been “invisible” through time could well have been extremely similar to their modern counterparts, but aligning many modern hunters with archeo- logically visible groups is in part a potential error which could have serious effects. The often asked question, “Are the Bushmen typical of culture in the past?” would then be answered in the affirmative, but with the qualification that we cannot determine how these earlier cultures functioned from the archeological record, since they would be below the “visibility threshold.” The Bushmen thus would be typical of some past cultures, but not those which we are able to excavate and study. If the above remarks concerning the Ge and similar cultures have any merit whatsoever, perhaps there are remnants of archeologically visible cultures with us today, but they are few in number.

There is yet another factor contributing to the degree of utility which any archeological configuration might possess for inferential purposes. This factor might be termed “sharpness” or focus; that is, how distinct is the pattern which is visible? A very visible assembly might not be too sharp in the patterning which it exhibits, while a much less visible unit might be very clear in its patterning. At least two factors would seem to afTect the sharpness in focus of an assemblage.

Groups of the kind we have heard discussed at this session, characterized by very fluid, changing behavioral patterns, might be expected to produce more diffuse patterns in their assemblages than highly structured societies in theirs.

31b. Archeological Visibility of Food-Gatherers

Speakers: L. Binford, Clark, Deevey, Owen, Pilling, Sackett, Strueuer, Suttles, Washburn

WASHBURN: I have two brief points on the matter of visibility in respect to the tremendous problems of reconstructing the vegetative components of an archeological site. I think it is frequently forgotten that animal bones found at a site often give us some rough notion about the plants on which the animals fed. The other point, one that Robert Heizer has made, is that the trace elements in the site are altered by the particular kinds of plants that grew there. It can be proved, for example, that there was corn at a certain site, even though there are no actllal plant remains whatsoever. So I do not think that we should think of visibility in an archeological context simply in the usual sense of the word.

STRUEVER: There are other recovery techniques, such as flotation techniques, that archeologists do not yet use. By such techniques, one can recover a remarkable quantity of plant remains invisible to the eye, even though they are macroplant remains. These techniques could be used to extend the record a lot further.

PILLING: I might also stress the importance of minute fragments of charcoal, which normally allow species identification. Also, we have some pretty tricky ecological questions to settle for many grassland areas of the world. If a prehistoric culture participated in fire drives for game, I think this will have a significant effect on the floral remains. The remains may last until the present time in what we would interpret to be a fairly good representation of prehistoric floral remains, and yet not be an accurate representation of this at all.

CLARK: I would like to join the others in emphasizing the importance of studying vegetable foods in the prehistoric context. As Jim Deetz has said, archeologists have grossly underestimated their importance up to now, mainly because lhe data are not a part of the typed series of artifacts that they have ana- lyzed—or, rather, they have felt that the data were not there. There are certain sites that clearly produce this evidence, such as waterlogged sites, which are well known in New Zealand and in Mesolithic and Paleolithic sites in Europe. At another such site, Kalambo Falls, we have actual seeds and nuts of about half a dozen species, which tell us Ibesides the fact that they were collected) the time of year at which the site was occupied. Also, there are certain short sticks, pointed at both ends; and I would be interested to know from ethnographers whether anything of this kind is used for digging up vegetable materials.

One of the main pieces of evidence that occur on these sites with shaped artifacts are rough blocks and chunks of stone. These must be looked at very carefully and compared with ethnographic evidence. James Woodburn collected a hammer used for breaking bones among the Hadza. Now this is a perfectly ordinary chunk which has nothing more revealing on it than a few bash marks and blunted edges. So I would like to know from ethnographers what, in addition to shaped tools, is used for breaking up and preparing vegetable foods.

Finally, something further to Dr. Pilling’s remarks about minute pieces of charcoal: there is a danger in dealing with them. These occur in certain of the mushitu, relict forest patches in central Africa, and it was thought that they might well give an indication of the age of the original, unburned forest patches. But the evidence suggests that burr?wing animals have almost certainly complicated the matter of interpretation. So one must be careful in dealing with microscopic particles.

DEEVEY: In view of the high probability that pebble tools of Paleolithic age were almost invariably used for chopping up plant food, and in view of the ease with which gas chromatography extracts microgram and millimicrogram quantities of hydrocarbons from meteorites, I would strongly recommend that archeologists stop washing and handling the pebble tools that they find.

SUTTLES: If I may return to that anomalous area, the Northwest coast, I might say that among the Salish the pattern for exploiting the habitat consisted in their going to favorite places at particular times of the year. For example, at the lowest tides in June people came from several villages to a favorite clam bed, dug clams in great quantities, steamed and dryed them, and took them back-leaving the shells behind. To the archeologist, this might look like long occupation by a small group or, perhaps, seasonal occupation by a much larger group than, in fact, was involved. To cite another example, a professional hunter might go into the hills, spend the summer’ hunting and process the kills at a campsite, taking the dried meat back and leaving the bones. An archeologist digging into such a site might not even recognize it as belonging to the same culture that had left the pile of shells on the beach.

BINFORD: This very thing has occurred. There are cases that could be documented for eastern North America, where “cultural units” have been defined for sites that actually represent either seasonal or task-specific occupation. I refer particularlv to some of the sites along the Tennessee River. However, if one follows an operational pr{)cedure based on theoretical considerations of the type outlined hy Frec- man, one’s first concern is to account for the observed variations within a region. This at least gives us a chance of set’ing the shell heap and the isolated campsite as task-specific sites within a variable region, rather than as necessarily separate basic units. Hopefully, by anticipating such variability, w(c may be able to develop means of testing alternative hypotheses and coming out with an approximation of the adaptive system present in the region.

, OWEN: I wonder whether someone would elaborate on the techniques that make it possible to tell whether a living floor was occupied for a short or a long period of time.

CLARK: This sort of thing is a problem for the geologists as well as for archeologists, but in any particular case it will be clear whether a horizon has been exposed for a long time or not. If it has been exposed for a long time there will be erosion, formation of soils, differential wearing of the artifacts, artifacts of different ages associated, natural accumulations of rubble, and other such signs.

SACKETT: You can also tell a great deal from maturation evidence in bones. For example, in the Upper Paleolithic of France you find reindeer teeth and antlers showing every stage of growth, and you know that you are dealing with year-round occupation. Likewise, bone material that indicates year-round occupation is usually found in association with strictly local raw materials; while evidence of periodicity in animal bones is often associated with exotic raw materials.

31c. The Use of Ethnography in Reconstructing the Past

Speakers: Clark, De Vore, Howell, Sharp, Turnbull

HOWELL: I want to direct the attention of anthropologists concerned with existing hunter-gatherers to some of the situations and problems that confront the pal(‘oanthropolo- gist concerned with Quaternary hominids. Until this afternoon, there has been littk said about cultural evolution beyond an occasional mention of how unsatisfactory are precontact cultural reconstructions for some of the groups discussed.

Paleoanthropologists seem to fall into two camps —I hesitate to say “bands” or “hordes”: (i) those who feel that intensive study and deep understanding of the present may otkr guidelines or even solutions to problematic situations of the distant past, and (2) those, less so inclined, who arc ever hopeful that new methods, procedures, and formulations will be developed which might provide insights into the adaptation and behavior of hominid populations, often long extinct and biologically, at least in part, quite unlike ourselves. I stand among the latter, and I trust that after the discussions of the past few days, our Ilumbns may increase markedly. To list a number of the points on which we have heard new and/or conflicting data here—all of them used more or less unquestioningly by some paleoanthropologists—I might mention: the importance to be assigned to scavenged meat; the extent and significance of game procurement among so-called “hunters”; the importance of vegetable food sources; the permeability of a well- delineated “territory”; the fluid size and semistructured character of human groups; the frequency of bilaterality; the variability of and exceptions to fixed-cousin marriage rules; the frequent and perhaps universal lack of immediate environmental determinism of group size, spacing, or mobility; the diversity and significance of subcommunity units, whether dyads, suhfaetions, etc.; and tbe low degree of predai:or-dctermined mortality. The paleoanthropologist who is still set on reconstructing the past on the basis of existing hunter-gatherer evidence has, obviously, an unusually wide range of choices from which to draw analogies about social structure and huma