ROUTLEDGE LIBRARY EDITIONS:

ARCHAEOLOGY

Volume 15

THE WALKING LARDER

 THE WALKING LARDER
Patterns of domestication, pastoralism,
and predation

Edited by
JULIET CLUTTON-BROCK
First published in 1989

This edition first published in 2015

by Routledge
2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN

and by Routledge
711 Third Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 1989 Juliet Clutton-Brock and contributors
All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by
any electronic, mechanical, or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or retrieval system, without permission in
writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation without intent to infringe.
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library

ISBN: 978-1-138-79971-4 (Set)

eISBN: 978-1-315-75194-8 (Set)
ISBN: 978-1-138-81599-5 (Volume 15)
eISBN: 978-1-315-74645-6 (Volume 15)

Publisher’s Note
The publisher has gone to great lengths to ensure the quality of this book but points out that some
imperfections from the original may be apparent.

Disclaimer
The publisher has made every effort to trace copyright holders and would welcome correspondence
from those they have been unable to trace.
THE WALKING LARDER
Patterns of domestication, pastoralism,
and predation

Edited by Juliet Clutton–Brock

London
UNWIN HYMAN
Boston Sydney Wellington
© Juliet Clutton–Brock and contributors, 1989
This book is copyright under the Berne Convention. No reproduction
without permission. All rights reserved.

Published by the Academic Division of

Unwin Hyman Ltd
15/17 Broadwick Street, London W1V 1FP, UK

Unwin Hyman Inc.

8 Winchester Place, Winchester, Mass. 01890, USA

Allen & Unwin (Australia) Ltd

8 Napier Street, North Sydney, NSW 2060, Australia

Allen & Unwin (New Zealand) Ltd in association with the

Port Nicholson Press Ltd
60 Cambridge Terrace, Wellington, New Zealand

First published in 1989

Paperback edition 1990

British Library Cataloguing in Publication Data

The walking larder: patterns of domestication, pastoralism, and predation.

1. Animals. Relationships with humans
I. Clutton–Brock, Juliet II. Series
591.6
ISBN 0–04–445013–3
ISBN 0–04–445900–9 pbk

Library of Congress Cataloging in Publication Data

The walking larder: patterns of domestication, pastoralism, and predation/edited by Juliet Clutton-
Brock.
p. cm.
Chiefly proceedings of the World Archaeological Congress held in
Sept. 1986 in Southampton, England.
Bibliography: p.
Includes index.
ISBN 0–04–445013–3 (alk. paper)
ISBN 0–04–445900–9 (pbk)
I. Domestication–Congresses. 2. Human ecology–Congresses.
3. Animal remains (Archaeology)–Congresses. I. Clutton–Brock,
Juliet.
 II. World Archaeological Congress (1986: Southampton, Hampshire)
SF41.W35 1988 88–10127
636′.009–dcl9; ff06 03–

Set in 10 on 11 point Bembo by Columns of Reading

Printed in Great Britain
at the University Press, Cambridge
Frontispiece (see p. 252): Llama singing with the emperor (Poma de Ayala 1936, p. 318).
Foreword

This book is one of a major series of more than 20 volumes resulting from
the World Archaeological Congress held in Southampton, England, in
September 1986. The series reflects the enormous academic impact of the
Congress, which was attended by 850 people from more than 70 countries,
and attracted many additional contributions from others who were unable to
attend in person.
The One World Archaeology series is the result of a determined and
highly successful attempt to bring together for the first time not only
archaeologists and anthropologists from many different parts of the world,
as well as academics from a host of contingent disciplines, but also
nonacademics from a wide range of cultural backgrounds, who could lend
their own expertise to the discussions at the Congress. Many of the latter,
accustomed to being treated as the ‘subjects’ of archaeological and
anthropological observation, had never before been admitted as equal
participants in the discussion of their own (cultural) past or present, with
their own particularly vital contribution to make towards global, cross-
cultural understanding.
The Congress therefore really addressed world archaeology in its widest
sense. Central to a world archaeological approach is the investigation not
only of how people lived in the past but also of how, and why, changes took
place resulting in the forms of society and culture which exist today.
Contrary to popular belief, and the archaeology of some 20 years ago,
world archaeology is much more than the mere recording of specific
historical events, embracing as it does the study of social and cultural
change in its entirety. All the books in the One World Archaeology series
are the result of meetings and discussions which took place within a context
that encouraged a feeling of self-criticism and humility in the participants
about their own interpretations and concepts of the past. Many participants
experienced a new self-awareness, as well as a degree of awe about past
and present human endeavours, all of which is reflected in this unique
series.
The Congress, was organized around major themes. Several of these
themes were based on the discussion of full-length papers which had been
circulated some months previously to all who had indicated a special
interest in them. Other sessions, including some dealing with areas of
specialization defined by period or geographical region, were based on oral
addresses, or a combination of precirculated papers and lectures. In all
cases, the entire sessions were recorded on cassette, and all contributors
were presented with the recordings of the discussion of their papers. A
major part of the thinking behind the Congress was that a meeting of many
hundreds of participants that did not leave behind a published record of its
academic discussions would be little more than an exercise in tourism.
Thus, from the very beginning of the detailed planning for the World
Archaeological Congress in 1982, the intention was to produce post-
Congress books containing a selection only of the contributions, revised in
the light of discussions during the sessions themselves as well as during
subsequent consultations with the academic editors appointed for each
book. From the outset, contributors to the Congress knew that if their
papers were selected for publication, they would have only a few months to
revise them according to editorial specifications, and that they would
become authors in an important academic volume scheduled to appear
within a reasonable period following the Southampton meeting.
The publication of the series reflects the intense planning which took
place before the Congress. Not only were all contributors aware of the
subsequent production schedules, but also session organizers were already
planning their books before and during the Congress. The editors were
entitled to commission additional chapters for their books when they felt
that there were significant gaps in the coverage of a topic during the
Congress, or where discussion at the Congress indicated a need for
additional contributions.
One of the main themes of the Congress was devoted to ‘Cultural
Attitudes to Animals, including Birds, Fish and Invertebrates’. The theme
was based on discussion of precirculated full-length papers, covering four
and a half days, and was under the overall control of Dr Tim Ingold, Senior
Lecturer in the Department of Social Anthropology, University of
Manchester, and Mark Maltby, Research Fellow in the Faunal Remains Unit
of the Department of Archaeology, University of Southampton. The choice
of this topic for a major theme arose from a desire to explore, from an
interdisciplinary perspective, the many facets of the varying relationships
that have developed between humans and animals, as these are reflected by
the historical diversity of cultural traditions.
Discussions during the Congress were grouped around four main
headings, each of which has led to the publication of a book. The first,
organized by Tim Ingold, was concerned with ‘What is an Animal?’ leading
to the book of the same title. The second subtheme, on ‘The Appropriation,
Domination and Exploitation of Animals’, lasted for over a day and a half
and was under the control of Juliet Clutton–Brock, editor of this book. A
day was devoted to discussion of the ‘Semantics of Animal Symbolism’ and
the co-ordinator, Roy Willis, is also the editor of the resulting book on
Signifying animals: human meaning in the natural world. Howard Morphy
was in charge of the fourth subtheme on ‘Learning from Art about the
Cultural Relationships between Humans and Animals’, and has edited the
volume on Animals into art.
The overall theme took as its starting point the assumption that there is
no one human attitude consistently maintained towards a particular species
of animal, and that similar human sentiments have been attached to a huge
variety of different animals at different times and in different places. It set
out to investigate the similarities and differences in practices and beliefs
connected with animals, including birds, fish and invertebrates, across both
time and space.
Prior to this century, in the West, animal behaviour was usually portrayed
and interpreted in terms of a contrast with human behaviour. Darwin was
not alone in his frequent adoption of an anthropocentric perspective in
formulating questions and in presenting hypotheses and interpretations. It
has often been claimed that people of non-Western cultures generally view
animals quite differently. Another aim of the Congress theme was to
explore such contrasts and to suggest some of the factors underlying both
anthropomorphic and anthropocentric perceptions of animals which are
currently prevalent at least in Western society.
 Ecological, psychological, cultural, and utilitarian considerations are all
involved in peoples’ attitudes to, and treatment of, other species. These
factors were considered not only from a wide, interdisciplinary point of
view but also, as befits a world archaeological context, especially in an
historical perspective, giving due emphasis to their changes over time.
For example, in the West when those of us who live in towns and cities
think of dogs and cats we usually think of them as companions, although
dogs are also, in other contexts, considered essential for herding, guarding,
and hunting other animals. In ancient Egypt, cats were often shown in
artwork as pets, but they were possibly also used to hunt and catch birds. In
many present-day cultures across the world people think of quite different
animals, such as cattle and pigs, as friends or companions. On the other
hand, the hyena is normally considered by the layman today to be wild and
untrainable, yet an ancient Egyptian representation appears to show one
being handled. Once we move beyond the normal level of trying to
ascertain from any excavation simply what animals were eaten or used for
transportation, we are bound to look again at the nature of the relationships
and interactions between human groups and the animals in their
environments. Another aim of this theme, therefore, was to investigate how
different people think, and thought, about different classes of animals, to
discover the principles of classification involved, and to show how these
principles constituted logical systems of belief and action. The presence of
so many Congress participants from the so-called Third and Fourth Worlds
made it possible to embrace a truly cross-cultural perspective on these
issues.
One point of interest lies in the investigation, on a world-wide basis, of
the reasons why particular animals have been domesticated by humans –
whether for food, such as meat or milk, or for other reasons, such as for
ritual purposes.
Contributors to the theme on ‘Cultural Attitudes to Animals’ adopted a
variety of perspectives for looking at the complex ways that past and
present humans have interrelated with beings they classify as animals.
Some of these perspectives were predominantly economic and ecological,
others were symbolic, concerned with the classification of both the physical
and the social environment, and still others were primarily philosophical or
theological. All these different perspectives are required for a full
interpretation of the artworks of the past, which in their representations of
humans and animals reveal some of the foci and inspirations of cultural
attitudes to animals.
In focusing on the nature of the varying relationships that can develop
between humans and animals, one is led inevitably to the question: what
actually is an animal or a human? By asking such a question, archaeologists
and others are forced to become aware of their own individual and cultural
preconceptions, and to pay attention to a set of problems concerning
attitudes.
The main themes in The walking larder have been detailed in its editorial
Introductions (pp. 1–3, 7–9, 115–18, 279–81). My aim in what follows is to
examine a few of the points which have struck me personally as being of
particular note or fascination.
In this book Juliet Clutton–Brock and her contributors explore the
complex interactions which can exist between humans and mammals, birds,
fish, and invertebrates, in any society, of whatever period, and in whatever
part of the world, within a symbiosis which we often ignorantly assume to
be a simple matter of economic contract and necessity. During the Congress
these topics were discussed under four main headings, ‘Early Husbandry
and the Exploitation of Domestic Stock’, ‘Pastoralism’, ‘Hunting and
Collecting’, and ‘The Interaction of Human and Animal Behaviour and the
Control of Animals’. The unravelling of these complex symbioses from the
past depends on a fascinating interplay of evidence from both archaeology
and zoology which is often difficult to interpret. It is never easy to
determine from subfossil animal bones alone whether or not they have
changed in shape or size sufficiently to warrant the deduction that full
domestication of the animal species had indeed occurred. Similarly, it is
often difficult or impossible to determine from the size of settlement and the
type of tools alone whether the economy of a population was dependent on
domesticated animals or on hunting.
It is not only the technical and detailed problems of how to recognize
animal domestication, and even how precisely the term should be defined
and used, that are presented here. The book also brings to life much of the
actual experience and meaning of the close relationships human beings
have with other species. There is much in this book about humans and their
capacity, or lack of capacity, to form relationships and to live in close
proximity, and in mutual interdependence, with other living creatures. In
this way The walking larder is linked to another of the One World
Archaeology books, What is an animal? The walking larder makes clear
what a human is. It is an organism which not only has an outstanding
capacity to make use of other living organisms to provide itself with food,
but also creates around itself a conceptual and social framework. The
relationship between humans and other living species, as revealed over time
by archaeology, seems rarely, if ever, to have been limited to the domain of
killing and eating, without involving social identity or beliefs. This does not
in any sense deny the economic importance of the domestication process, a
process which used, until recently, to be considered so profound that it was
termed a ‘revolution’ in human affairs, and which many still consider an apt
description of the peak of the domestication process in the Neolithic (and
see Foraging and farming, edited by D. R. Harris & G. C. Hillman).
In 1969 Geoffrey Dimbleby and I wrote (p. xx), in a review of the
development of domestication studies, that ‘domestication was a process
extending over several thousand years and that it had its own special
characteristics in different areas of the ancient world. Domestication did
not, of course, happen only once but has recurred time and time again in
different parts of the world and at different times. Domestication as a
process still continues.’ The walking larder explores this diversity of the
past as well as documenting a range of social and cultural attitudes which
cannot be disentangled from a purely economic domestication process.
Perhaps most important of all, this book moves out of the past into the
present, and demonstrates that, as in the past, the intimate cultural
relationships between humans and other animals continue to be modified,
interwoven with, and altered by the encroachment and innovations of
different cultures.

P. J. Ucko
Southampton

Reference
Ucko, P. J. & G. W. Dimbleby 1969. Introduction: content and development of studies of
domestication. In The domestication and exploitation of plants and animals, P. J. Ucko and G. W.
Dimbleby (eds), xvii–xxi. London: Duckworth.
Contents

List of contributors
Foreword P. J. Ucko
Preface Juliet Clutton–Brock

Introduction Juliet Clutton–Brock

DOMESTICATION
Introduction to domestication Juliet Clutton–Brock

1 Pet-keeping and animal domestication: a reappraisal James Serpell

Introduction
A definition of pets
The extent of the pet-keeping phenomenon
The functions of pet-keeping
Pet-keeping and domestication
Conclusions
References

2 Definitions of animal domestication Sandor Bökönyi

References

3 Defining domestication: a clarification Pierre Ducos

Note
References

4 Some observations on modern domestication processes Sytze

Bottema
 Introduction
Some observations on morphological changes occurring during the
domestication process
The keeping and breeding of geese in captivity: how to start domestic
stock
An analysis of some colour variations occurring in various subspecies
of mallard: an example of isolation
Some observations on the role of temperature in the rearing of
nidifugous birds
Food as a mechanism of control in the domestication process
Acknowledgements
References

5 Feral mammals of the Mediterranean islands: documents of early

domestication Colin P. Groves
Introduction
Material and methods
Results
Discussion
Conclusions
References

6 Escaped domestic animals and the introduction of agriculture to Spain

Iain Davidson
Frontiers between fisher-gatherer-hunters and farmers
The Australian model
The frontier in Spain
The agricultural colonization of Spain
Conclusions
Acknowledgements
References
7 Evidences for the impact of traditional Spanish animal uses in parts of
the New World Elizabeth S. Wing
Acknowledgements
References

8 Osteological evidence for the process of animal domestication

Richard H. Meadow
Introduction
Domestication
Demographic evidence for animal domestication
Zoogeographic evidence for domestic animals
Morphological evidence for animal domestication
Conclusion
References

9 Animal exploitation and the phasing of the transition from the

Palaeolithic to the Neolithic Hans–Peter Uerpmann
Introduction
Terminology of the Palaeolithic–Neolithic transition
Terminology of other economic stages
References

10 A two-part, two-stage model of domestication Frank Hole

Concluding remarks
References

11 The domestic horse of the pre-Ch’in period in China Chow Ben–

Shun
Introduction
The horse of the Chinese Bronze Age
References
12 Utilization of domestic animals in pre- and protohistoric India P. K.
Thomas
References

PASTORALISM
Introduction to pastoralism Juliet Clutton–Brock
References

13 The origins of migration and animal husbandry in the steppes of

eastern Europe Valentin Pavlovich Shilov
Note
References

14 Pastoralism in southwest Asia: the second millennium BC Juris Zarins

Introduction
The early transition (1800–1400 BC)
The later transition (1400–900 BC)
The archaeological evidence
Camel domestication
References

15 Farming to pastoralism: effects of climatic change in the Deccan M.

K. Dhavalikar
References

16 The changing role of reindeer in the life of the Sámi Pekka Aikio
A personal view from my childhood
The origins of reindeer-herding
Reindeer-herding under the pressure of change
Concluding remarks
 Acknowledgement
References

17 The geographical distribution and function of sheep flock leaders: a

cultural aspect of the man–domesticated animal relationship in
southwestern Eurasia Yutaka Tani
Introduction
Flock leader types and their geographical distribution
Behaviour and function of flock leaders in the herding situation
Discussion
References

18 Cattle in ancient North Africa Juliet Clutton–Brock

Introduction
Theories on the origins of cattle in North Africa
The differentiation and dating of cattle remains from North Africa
Cattle in the Nile valley
Discussion
References

19 The development of pastoralism in East Africa Peter Robertshaw

Acknowledgements
Note
References

20 Cattle and cognition: aspects of Maasai practical reasoning John G.

Galaty
Introduction
The logic and experience of pastoralism
Cattle and classification
Cognitive processes: identifying the missing animal
 Some concluding remarks
Acknowledgements
References

21 Prehispanic pastoralism in northern Peru Tom McGreevy

Introduction
The South American camelids
Types of pastoralism
Northern Peruvian Survey data
Conclusions
Acknowledgements
References

22 Andean pastoralism and Inca ideology Gordon Brotherston

Tahuantinsuyu and its antecedents
Models of control and authority
The larger native American context
Conclusions
Acknowledgements
Note
References

23 Origins and development of Andean pastoralism: an overview of the

past 6000 years David L. Browman
Introduction
Distribution and intensification studies
Population and management shifts
Morphological traits
Final remarks
References
24 Are llama-herders in the south central Andes true pastoralists? Mario
A. Rabey
Acknowledgements
References

PREDATION
Introduction to predation Juliet Clutton–Brock
References

25 Did large predators keep humans out of North America? Valerius

Geist
Introduction
When did humans enter North America?
Human expansion along with other Siberian species
Human anti-predator strategies
Bears and native North Americans
The Arctodus problem
References

26 Hunting in Pre-Columbian Panama: a diachronic perspective

Richard G. Cooke and Anthony J. Ranere
Introduction
Chronological and contextual notes on the studied sites
Dietary contribution versus frequency of capture
The dietary contributions of fish and other aquatic resources
Natural environments, habitat modification, and human behaviour
Exotic materials and exchange
Conclusions
Acknowledgements
References
27 Shells and settlement: European implications of oyster exploitation
Derek Sloan
Introduction
Early shell-middens
Denmark
Scotland
Ireland
Discussion
Acknowledgements
References

28 Effects of human predation and changing environment on some

mollusc species on Tongatapu, Tonga Dirk H. R. Spennemann
Introduction
The problem
The changing environment
Proportion of species
Shell size
Conclusions
Implications for further research
Acknowledgement
References

29 Rocky Cape revisited – new light on prehistoric Tasmanian fishing

Sarah M. Colley and Rhys Jones
The Tasmanian fish problem
Revived interest in an unanswered question
Samples from Inner Cave, Rocky Cape South
Preliminary results of the fish-bone analysis
Fish traps
Acknowledgements
 Note
References

30 Mutualism between man and honeyguide Alex Hooper

References

31 Cova Negra and Gorham’s Cave: evidence of the place of birds in

Mousterian communities Anne Eastham
Introduction
Birds in Mousterian communities
Ecological interpretations
Hunting of birds in the Mousterian
References

Index
Preface

When Peter Jewell and I were invited by Peter Ucko in 1983 to organize
one of the sessions of the Congress within the main theme on ‘Cultural
Attitudes to Animals’ we had to avoid overlap with the themes of the
Fourth Conference of the International Council for Archaeozoology held in
London in 1982. We also had to avoid duplication with the Fifth
Conference which was held in Bordeaux in the week following the
Congress in Southampton. We decided to place the emphasis of the session
on behavioural studies, and to try to bring together the past and the present
in discussions that would be focused on concepts rather than on the results
of specialist analyses. In devising a framework for the session we were
helped by talks with Eitan Tchernov from the Department of Zoology at the
Hebrew University, Jerusalem. Later Professor Tchernov resigned from the
Congress over the banning of South African participants. Peter Jewell also
resigned, as described so lucidly by Ucko (1987).
There were 50 contributions to the Congress session on ‘The
Appropriation, Domination, and Exploitation of Animals’ and these
provided a basis for wide-ranging discussions with three themes
predominating: the process and meaning of domestication, studies of
pastoralism today and herding practices of the past, and the relationships
between predators and their prey.
All but three of the chapters in this book were first presented as
contributions to the World Archaeological Congress. James Serpell had
intended to come to the Congress but was unable to attend and so sent his
chapter in at a later date. Serpell’s contribution (Ch. 1) on pet-keeping is
linked to What is an animal? (edited by T. Ingold) in its discussion on
owner-pet relationships and human perception of animals. Many dogowners
would agree that their dog is ‘a person without verbal language’. On the
other hand, perhaps few archaeologists would agree with Serpell that the
domestication of animals for food followed directly from early pet-keeping.
 Pierre Ducos did not attend the Congress but was invited to write his
chapter on defining domestication in response to that of Sandor Bökönyi, as
the two authors have held a long-standing debate on this subject (Chs 2 &
3).
Sytze Bottema’s chapter on the modern domestication of waterfowl was
given to the book by special request of the editor because the observations it
contains are of such particular relevance to discussions on the process of
domestication (Ch. 4).
J. Galaty’s Chapter 20 was presented to the Congress in the section on
‘Indigenous meanings’ within the subtheme ‘Semantics of Animal
Symbolism’.
I am deeply grateful to all the authors of this book who produced their
revised manuscripts in a very short time after the World Archaeological
Congress and who responded so readily to my troublesome demands for
standardization. In particular, I should also like to thank Tim Ingold for his
helpful advice over the planning of the session at the Congress and for
chairing the section on ‘Pastoralism’. Other chairpersons I wish to thank are
Mark Maltby, Richard Meadow, and Elizabeth Wing. But, above all, I wish
to express my gratitude to Peter Ucko for his enthusiastic, if rigorous,
approach to the production of this book, and without whom there would
have been no Congress.
My thanks are due to Kim Dennis-Bryan, Department of Zoology, British
Museum (Natural History), for checking all the bibliographies, and to
Caroline Jones, Department of Archaeology, University of Southampton,
for her help with typing and the solving of innumerable small problems.
Juliet Clutton-Brock
London

Reference
Ucko, P. 1987. Academic freedom and apartheid: the story of the World Archaeological Congress.
London: Duckworth.
Introduction
JULIET CLUTTON-BROCK

History is what you choose to believe. To my father archaeology was the

study of antiquities, pottery, and flint implements, and I was told at school
that I could not be an archaeologist because I had not learned ancient
Greek. Forty years ago animals had little to do with archaeology, except in
artistic representation, and the relationships between humans and animals, if
thought about at all, were classed as anthropology or biology. To the non-
biologist, wild animals were ferocious beasts to be shot on sight, nature was
red in tooth and claw, and natural history was pursued with a butterfly net.
Archaeologists and biologists of my generation have mostly spent their
lives attempting to dispel these beliefs. To me, archaeology is the history of
humans in their environment, and the history of their progressive
domination of a world that appears to be rapidly diminishing in size as we
travel higher and faster around it. I look at the past through the debris of
human settlement and the present through studies of human and animal
behaviour, and I see that the master predator is indeed an animal with
unique consciousness and capabilities, but an animal all the same.
Over the past 40 years great progress has been made in the understanding
of human history and its interlocking with the animal world. Archaeology is
no longer only the study of sherds and flints, and a whole new field of
work, archaeozoology (or zooarchaeology), has become an essential part of
the subject. This began with the simple identification of food remains on
archaeological sites but it now encompasses almost all aspects of the
physical life of humans and animals in the past, and reaches back to the
evolution of the early hominids.
The chapters in this book cover the last 40 000 years but they are not
intended to represent a sequence from the past to the present, nor are they to
be read as a series of research topics in archaeozoology. The three sections
(on domestication, pastoralism, and predation) may appear to be in reverse
order to many readers who see the development of civilization as a
progression from hunting to early agriculture to livestock husbandry; but
the chapters have been arranged intentionally in this order to free them from
this concept of progression. The aim of The walking larder is to present
evidence for the manifold relationships that occur between humans and
animals, and to demonstrate that such relationships continue. The hunting
of some animals and the keeping of others as valued companions was as
much a part of human nature 10 000 years ago as it is today. Just as the
domestic dog has the same behavioural patterns as the wolf, so the modern
human probably differs little in his or her genetically inherited behaviour
from the earliest Homo sapiens. It is only the development of culture and
the ensuing pressures of social systems that change.
This book is called The walking larder because the provision of food is
by far the most important function of animals in all human societies, and
while they are alive the animals are almost always on the move. Like
wolves, humans roam over huge home ranges; they must have protein to eat
but they are compulsive travellers who are always searching to widen their
territories and find new resources. Wild animals are followed and hunted;
domestic animals are driven along with their human owners as a store of
meat on the hoof, indeed as livestock. There are, of course, sedentary
populations of humans and animals, as are discussed in this book. Sloan
(Ch. 27) describes the exploitation of shellfish in northern Europe during
the Mesolithic, and suggests that the abundance of food available from this
source enabled settlements to survive in one place for very long periods. A
similar argument is put forward by Spennemann (Ch. 28) for prehistoric
settlement on Tonga in the Pacific, where molluscs were the main source of
protein. However, sooner or later such food supplies will dwindle and new
resources have to be found.
Apart from the provision of food, the most important function of human–
animal relationships is mutual companionship. The wolf was the first
animal to be domesticated and it was probably not for its meat, but either as
object of affection, as a helper in the hunt, or as a useful scavenger of
human debris – most likely it was for all three reasons.
Besides the obvious ways in which humans differ from animals, they are
unique in their ability to communicate on an interspecific level. This has
meant that humans and animals can have much closer relationships than are
normally found amongst different groups of vertebrates. A human family
can enfold a dog and a cat into the home and the dog can have some sort of
rapport with the cat through their mutual owner, but a wild dog would very
rarely, if ever, be friendly with a wild cat. This ability to communicate is the
basis for the phenomenon of taming and domestication which has been a
nucleus of human cultures for at least the past 10 000 years. There can be
no civilization without beasts of burden (even if these are mechanized as in
many countries at the present day).
For as long as people have travelled around the world they have taken
their dogs, livestock, pets, and parasites with them. This has resulted in
continuous change in the assemblages of animals and plants in every land,
as well as changes in the environment, to an extent that is only just
beginning to be realized. One of the aims of this book is to highlight these
changes and to emphasize their ancient beginnings. People sailing to
Australia took their dogs with them at least 3000 years ago, this being the
earliest radiocarbon date for the remains of dingo. Today, living dingoes are
the descendants of these ancient dogs, which soon established themselves as
feral populations that have had a determining effect on the marsupial fauna.
Neolithic people moving around the Mediterranean region, 7000 years ago,
may have exterminated the last of the endemic fauna of pigmy elephant and
hippo on the islands. The sheep and goats they took with them became part
of the stock from which our present breeds are descended, except for a few
that became feral in the mountains where they have lived ever since. It used
to be thought that the wild sheep and goats of Sardinia, Corsica, Crete, and
Cyprus were relics of the ancient endemic wild fauna, but it is now realized
(from the lack of fossil evidence) that these are feral populations of
Neolithic origin. This topic is reviewed by Groves in Chapter 5.
The rabbit and the fallow deer were both introduced into Europe north of
the Mediterranean by the Romans, and, of course, the horse was taken to
the Americas by the Spanish. Cattle began to move south through Africa at
least 5000 years ago. These are just a few examples of the movements of
large mammals – innumerable others could be mentioned on every
continent, but perhaps the greatest change that has been brought about for
the sake of the walking larder has been the result of deforestation and
burning of the land to provide grazing for livestock and fuel for cooking.
The cutting of thorn scrub and brushwood for making corrals and
enclosures for livestock has also contributed greatly to the destruction of
vegetation in arid regions.
By studying, often in painstaking detail, the remains of animals and
plants that have been found in association with human settlement on
archaeological sites, the changes wrought in the world over the past 40 000
years can be traced in remarkable detail. It is a slow task but a very
important one, and there may not be much time left, for the knowledge
gained must be used to show politicians and economists that modern
technology cannot cure all ills. To dig deeper and deeper wells in the Sahel
will not halt the spread of desertification, indeed it will increase it by
causing more land to be overgrazed and more land to be trampled around
the wells by the ever-increasing numbers of people and cattle. It would be
better to listen to the nomadic camel-herders whose ancestors have lived in
equilibrium with the desert for thousands of years.
DOMESTICATION
Introduction to domestication
JULIET CLUTTON-BROCK

Human beings are the most highly social and gregarious of all the primates,
and it may be postulated that their complicated patterns of social behaviour
evolved from the necessity to provide food for a community or family
group of early hominids that included helpless infants and aged relatives.
The communal hunting of large ungulates for meat could have been part of
the basis for the evolution of much human behaviour during the latter part
of the Pleistocene. To progress from the hunting of wild animals to the
herding of tame ones may seem a small step but it is, in fact, a very large
one, because what separates hunting from herding is the concept of
ownership of the animals.
The problem of what is domestication is of continuing concern to
biologists and archaeologists. During the last century it was a subject that
preoccupied Darwin, and the conclusions he came to about the origin of
species by natural selection were in great part the result of his knowledge
about the breeding of domestic animals. During this century the history of
domestication has been much studied, notably by Hilzheimer, Zeuner,
Herre, and Bökönyi, and yet the process of domestication is still little
understood and we are still arguing about its definition.
In Chapter 2 of this volume Bökönyi presents his views on the meaning
of domestication, and, because he takes up in debate the definition of Pierre
Ducos from an article written in the 1970s, it was only right that Ducos
should be invited to reply, which he does in Chapter 3. Bökönyi states in his
chapter that his definition of domestication is very close to Clutton–
Brock’s, the only difference being that ‘she explains it in a more colloquial
style’.
My definition of a domesticated animal is ‘one that has been bred in
captivity for purposes of economic profit to a human community that
maintains complete mastery over its breeding, organization of territory, and
food supply’. Like Bökönyi and Ducos, I believe that domestication is both
a cultural and a biological process and that it can only take place when
tamed animals are incorporated into the social structure of the human group
and become objects of ownership. The morphological changes that are
produced in the animal follow after this initial integration.
The biological process of domestication may be seen as a form of
evolution in which a breeding group of animals has been separated from its
wild conspecifics by taming. These animals constitute a founder population
that is changed over successive generations by both natural and artificial
selection, and is in reproductive isolation. In wild populations reproductive
isolation will often lead to the evolution of subspecies; in the domestic
animal it leads to the development of breeds. Unlike a subspecies, a breed is
not restricted by geographical locality but, like a subspecies, it has a
uniform appearance that is heritable and distinguishes it from other
populations.
Considering the large numbers of mammals that have been exploited by
humans, it is, at first glance, surprising that so few species have been fully
domesticated. On the other hand, considering the intensity of the
relationship between domestic animals and humans, perhaps it is even more
surprising how many species it has been achieved with and how diverse are
the breeds. The answer lies in the behaviour of the species of animals that
undergo domestication and in the inherent diversity of the genetic
constitution of wild mammals. It is only those species that have inherited
behavioural patterns that correspond to those of humans that can survive the
process, as discussed by Bökönyi in Chapter 2. In Chapter 1, Serpell
considers the closest of all conscious human–animal relationships, that of
pet-keeping, and he summarizes its roles in past and present societies, from
Palaeolithic hunters to modern societies.
Within recent years considerable attention has been given to attempts to
domesticate new species of animals to increase the supply of meat and other
resources, such as antler. Bottema, in Chapter 4, describes his observations
on the breeding of waterfowl in captivity. The birds undergo a process of
domestication and their morphological changes are described.
Chapters 5, 6, and 7 discuss some of the effects of the introduction of
new animals, both wild and domestic, to previously closed ecosystems.
Groves, in Chapter 5, describes the feral status of the sheep, goats, pigs, and
cats on the Mediterranean islands, and discusses the ancient introductions of
many other taxa of small mammals, carnivores, and deer. Davidson, in
Chapter 6, discusses the thieving of newly introduced stock by hunter-
gatherers, and he suggests that this activity, as carried out by the Australian
Aborigines in the last century, could be postulated for the endemic
Mesolithic people of the Mediterranean. This is a most interesting concept,
and it could be an important means by which early domestic animals spread
through the ancient world. There is sound documentary evidence for the
stealing of livestock from European explorers in the Americas and southern
Africa; Cornwallis Harris, for example, in his The wild sports of southern
Africa published in 1839, described in a chapter headed ‘Plundered by
Bushmen Hordes, and left a wreck in the desert’ how all the cattle of their
expedition were stolen by Bushmen. It is of interest to note that these
Bushmen, who had no domestic animals of their own, other than a few
dogs, were adroit at driving cattle in any direction that they wished and at
great speed.
Chapter 7, by Wing, describes the impact in the 16th century AD of the
introduction of domesticates to Florida and Haiti, and this can be seen as a
parallel to the introduction of animals to the Mediterranean, thousands of
years earlier. One particular comment by Wing is noteworthy for its
relevance, for she claims that ‘goats became feral in the mountains of
Jamaica at an early date’, just as they did in the Mediterranean islands, the
British Isles, and many other parts of the world.
Chapters 8–12 deal with the evidences for domestication. Meadow (Ch.
8) discusses the ways in which the process of domestication can be deduced
from the archaeological record, and he continues the discussion, as do
Uerpmann and Hole, on the definitions of domestication. Meadow makes
the important point that with domestication there is a change in focus on the
part of the humans from the dead to the living animal and more especially
to its progeny.
Uerpmann, in Chapter 9, reviews the terminology used for the period of
transition from hunter-gathering to sedentism in western Asia. This chapter,
however, has wider implications than the often contentious subject of
terminology. Uerpmann contrives to explain the beginnings of
domestication through the periods that immediately precede the Neolithic
and discusses the interpretation of socio-economic systems based on
shellfish, which are described further by Sloan in Chapter 27, and he also
examines the economic structure of the so-called Mesolithic of southern
France which had domestic sheep, as described further by Davidson in
Chapter 6.
Hole, in Chapter 10, makes use of his long familiarity with archaeology
in western Asia to review the beginnings of domestication in this core area.
He discusses the structure and timing of the stages of the seasonal reuse of
‘camps’, intensive plant use, the replacement of hunted gazelle by herded
sheep and goats, and the development of a full agricultural economy. Hole
concludes by discussing the environmental distributions of the wild
progenitors of domestic livestock and their herding behaviour which
enabled them to be controlled.
With the exception of the last two chapters, all those in this section of
The walking larder are about the definitions of domestication, its processes
during the prehistoric period and at the present day, and the effects of the
introduction of domestic animals on endemic faunas. Lastly, there are two
chapters that discuss specific examples of early domestication. In Chapter
11 Chow Ben–Shun presents the evidence for the origins of the domestic
horse in China, a subject that is very little known outside China. In Chapter
12 Thomas reviews the first domestication of animals in India and discusses
the beginnings of milking and the use of cow dung, two products that have
been of great importance in the history of India.
1 Pet-keeping and animal domestication:
a reappraisal
JAMES SERPELL

Introduction

Just over a century ago, Charles Darwin’s cousin, Francis Galton, proposed
a new theory about the process of animal domestication (Galton 1883).
Using a dossier of anecdotal material provided by 18th- and 19th- century
explorers and naturalists, Galton noted that the habit of capturing and
nurturing tame wild animals as pets was widespread among what he
regarded as ‘primitive’ people. He then made the assumption that
prehistoric hunting societies indulged in similar practices, and argued from
this that the original domestication of animals arose as a natural
consequence of mankind’s pet-keeping tendencies. ‘Savages may be brutal,’
he said, ‘but they are not on that account devoid of our taste for taming and
caressing young animals.’ It was this taste for taming and caressing pets, he
believed, that led to the birth of animal husbandry and farming.
Although Galton’s idea has been supported by more recent authorities
(e.g. Sauer 1952, Reed 1954, Zeuner 1963, Scott 1968, Clutton–Brock
1981) – particularly in relation to the domestication of dogs and, to a lesser
extent, pigs and poultry – it has lately become less fashionable as a general
explanation for the origins of domestication. More to the point, perhaps, its
existence has done little to stimulate anthropological research into the
meaning or functional significance of pet-keeping in either Western or non-
Western societies (Serpell 1986). Although this trend largely reflects an
increasing emphasis on the ecological and environmental antecedents of
domestication rather than on the processes by which it occurred (Harris
1978), it may also be due in part to long-standing prejudices about pets and
people’s reasons for keeping them. The present review examines some of
these preconceptions, and reconsiders the general validity of Galton’s
theory in the light of recent research on the nature of the human–pet
relationship.
A definition of pets

The Oxford English Dictionary (OED) defines a pet as: ‘Any animal that is
domesticated or tamed and kept as a favourite, or treated with indulgence
and fondness.’ In practice, however, the word tends to be used more loosely
as a blanket description for animals that are kept for no obvious practical or
economic purpose – i.e. pets, as opposed to livestock or working animals.
This has led to a certain amount of confusion in the literature since, clearly,
there are a variety of different reasons for keeping animals that have no
direct economic significance. Animals, for example, are kept for symbolic
purposes; they are used to advertise status and prestige; they are employed
as living adornments, and even as animated playthings or toys. The word
‘pet’ has been applied in each case (Tuan 1984). Yet an important difference
exists between these animals and ‘pets’ as defined by the OED. The former
tend to be viewed and treated essentially as objects or things, whereas the
latter are generally perceived as subjects or quasi-persons – hence the
tendency to indulge and fondle them. Recently, the term ‘companion
animal’ has been widely adopted to distinguish true pets from other kinds of
non-utilitarian animal (see Katcher & Beck 1983), but the present review
will continue to use the original and less cumbersome term according to its
dictionary definition.

The extent of the pet-keeping phenomenon

The keeping of dogs, cats, budgerigars, and other species as household pets
is so widespread in Western societies that it tends to be taken for granted.
Roughly half of the households in Britain contain at least one pet animal,
and per capita pet-ownership is considerably higher in some other countries,
such as France and the United States (Serpell 1986). More often than not,
this level of enthusiasm for non-utilitarian animals is regarded as a
peculiarly Western (and essentially middle-class) expression of material
affluence and bourgeois sentimentality. The prevalence of seemingly
identical behaviour among subsistence hunters and horticulturalists does
not, however, support this point of view.
Individual non-utilitarian animals are (or were until recently) treated with
‘indulgence and fondness’ by a considerable range of different societies
around the world. Indeed, the only notable exception seems to be Africa,
where pet-keeping is nowadays infrequent, but where it may have been
more common in pre-colonial days (Speke 1863, Zeuner 1963). In Australia
the Aborigines kept dingoes, wallabies, possums, bandicoots, rats,
cassowaries, and even frogs as pets (Zeuner 1963, Meggit 1965), and in
Southeast Asia indigenous groups kept dogs, cats, pigs, monkeys, and
various birds (Evans 1937, Leach 1964, Harrison 1965, Cipriani 1966). The
Polynesians and Micronesians favoured dogs, pigeons, parrots, fruit bats,
lizards, and eels (Jesse 1866, Galton 1883, Luomala 1960), and the Indians
of North America kept deer, moose, bison, racoons, wolves, dogs, bears,
turkeys, hawks, crows, and a variety of other small wild mammals and birds
(Galton 1883, Elmendorf & Kroeber 1960). Among the southern
Amerindians, particularly those inhabiting the Amazon basin, the taming
and keeping of wild animal pets was practically a minor industry. The
English naturalist, Bates, recorded a list of ‘twenty-two species of
quadrupeds’ which he found living tame in Indian settlements (Galton
1883), and later observers have specified dogs, cats, deer, tapir, peccaries,
monkeys, sloths, opossums, foxes, coatis, margay, ocelot, jaguar, chickens,
ducks, cormorants, parrots, and an extraordinary variety of small birds and
rodents (Roth 1934, Wilbert 1972, Basso 1973, Fleming 1984, Hugh–Jones
pers. comm.). One author seems to sum the situation up when he says that
‘few, indeed, are the vertebrate animals which the Indians have not
succeeded in taming’ (Roth 1934).
In general, Western observers have nearly always evinced surprise at the
pet-keeping activities of so-called ‘indigenous’ peoples. Not only was the
scale of the phenomenon impressive, but also the intensity of feeling it
seemed to evoke in its adherents. On a trip to South America during the
18th century, the Spanish explorers, Juan and Ulloa, were evidently
astonished at the Amerindians’ affection for pet birds. Such animals were
kept about the house, and the Indians never ate them:

…and even conceive such a fondness for them, that they will not sell
them, much less kill them with their own hands. So that if a stranger who
is obliged to pass the night in one of their cottages, offers ever so much
for a fowl, they refuse to part with it, and he finds himself under the
necessity of killing the fowl himself. At this his landlady shrieks,
 dissolves into tears, and wrings her hands, as if it had been an only son.
(Juan & Ulloa 1760.)

The resemblance to ‘an only son’ was not, apparently, an exaggeration. In a

recent (and rare) anthropological account of pet-keeping among the
Brazilian Kalapalo Indians, Basso (1973) describes the owner-pet
relationship as particularly interesting because its distinctive features: ‘are
also those which define the filiative relationship, or that between human
parents and their children. Children and pets alike are ideally supposed to
be fed, reared and kept protected within the confines of the house. Often
pets are kept secluded like human adolescents “to make them more
beautiful”.’ Although they may belong to species which are otherwise
classified as edible, Kalapalo pets are never killed or eaten, and when they
die they are often buried close to the house or hammock of the former
owner. This manner of burial is one ordinarily reserved for infants that die
at birth or before being named.
In modern Europe or North America, a woman would probably be
charged with indecency if she attempted to suckle a puppy or kitten at her
breast. Yet in many hunting and horticultural societies the suckling of
young mammals is considered perfectly normal and natural. Sir John
Richardson, one of Galton’s informants, not only observed that the ‘red
races’ of North America were fond of pets, such as bison calves, wolves,
and other species, but also that it was not unusual for them ‘to bring up
young bears, the women giving them milk from their own breasts’ (Galton
1883). Early visitors to Hawaii were similarly impressed by the inhabitants’
fondness for dogs, which ‘in spite of their stupidity were in high favour
with the women who could not have nursed them with a more ridiculous
affection if they had really been ladies of fashion in Europe’ (Luomala
1960). Another noted that ‘every woman has a pet animal; and mothers who
are nursing their own offspring will suckle a puppy at the same time in a
rivalry by no means in favour of the strength and number of their own
progeny’ (Jesse 1866). Evans (1937) made similar observations among the
Semang Negritos of Malaysia, where young pigs and monkeys were
commonly suckled by the women, and where he reported seeing a woman
with ‘a child at one breast and a monkey at the other.’ He also noted that
animals reared in this way were not killed.
 Even in societies with relatively strict utilitarian attitudes to domestic
animals, exceptions are sometimes made. Among the Inuit and the Indians
of the Canadian arctic, for example, dogs are ordinarily regarded as
working animals, and treated in a detached and often brutal way.
Nevertheless, childless individuals and couples will occasionally adopt one
particular puppy into the household and rear it as an indulged pet
(Savishinsky 1983). Thereafter, these animals are not expected to engage in
any kind of useful work.
Although women seem to be the main animal-tamers and pet-keepers, the
practice is certainly not confined to women. According to Luomala (1960),
Polynesians of all ages, sexes, and social ranks ‘fondled, pampered and
talked to their pets, named them, and grieved when death or other
circumstances separated them.’ Grief over the death of a pet dog was often
expressed through tears, poetical eulogies, and ceremonial burial.
Amerindian males sometimes travelled around accompanied by their pets.
The 18th-century explorer Hearne encountered one particular Indian who
had two tame moose as pets. When the man made a trip by canoe ‘the
moose always followed him along the bank of the river; and at night, or on
any other occasions that the Indians landed, the young moose generally
came and fondled on them [sic], as the most domestic animal would have
done’ (Galton 1883). Australian Aborigine males displayed comparable
attachments. According to the Swedish explorer, Lumholtz (1884), the
Aborigines reared dingo pups:

…with greater care than they bestow on their own children. The dingo is
an important member of the family; it sleeps in the huts and gets plenty
to eat, not only of meat but also of fruit. Its master never strikes, but
merely threatens it. He caresses it like a child, eats the fleas off it, and
then kisses it on the snout.

It needs to be emphasized that fondness for pets in all societies is largely

independent of the animals’ contribution to the local or family economy.
Lumholtz, for example, attributed the Aborigines’ affection for dingoes to
the fact that these animals were useful as hunting aids. More recently,
Cipriani (1966) and Harrison (1965) used precisely the same argument to
explain the excessive devotion to dogs exhibited by the Andamanese and
the Dyaks of North Borneo. But, clearly, mere economic utility provides no
guarantee of affection. The B’Mbuti pygmies of Zaïre, for instance, rely
heavily on dogs as hunting partners, yet they have a reputation for treating
them with extreme brutality (Singer 1968). Conversely, other cultures make
little practical use of dogs, but nevertheless regard them with overt
affection. Linton’s (1936) observations on the North American Comanche
provide a good example. The Comanche economy was based largely on
buffalo-hunting and raiding neighbouring tribes, both of which depended on
horses and horsemanship. Yet the Comanche possessed a detached and
strictly utilitarian attitude to horses. Comanche dogs, on the other hand,
were of no practical or economic value whatsoever, and were kept purely as
pets. But the average Comanche warrior regarded the loss of a dog as far
more devastating than the loss of several horses, and most of them would
spend hours discussing with fondness all the dogs they had ever owned.
As already indicated, many of the species kept as pets among indigenous
societies were also regularly hunted and killed for food. But once adopted
as pets, these same species were generally exempt from slaughter. Referring
to the Guiana Indians, Roth (1934) is adamant in stating that the ‘native will
never eat the bird or animal he has himself tamed any more than the
ordinary European will think of making a meal of his pet canary or tame
rabbit.’ Similar inhibitions also existed in cultures where the species
involved was raised commercially as an item of food. In Hawaii, for
instance, dogs were reared primarily as livestock, but pet dogs were rarely
slaughtered or eaten, and never without vociferous complaints from the
owner (Luomala 1960). Even when it was pointed out to them by
Europeans, many of these peoples failed to acknowledge the potential
economic value of their pets. When Sir John Richardson tried to buy pet
bear cubs from the Indians, he noticed that ‘in purchasing them there is
always the unwillingness of the women and children to overcome, rather
than any dispute about price’ (Galton 1883). Similarly, Fleming (1984)
found that the Caraja people of Brazil refused to sell some of their parrots
regardless of how much he offered to pay for them. When he suggested that
they train their pet cormorants to catch fish by fastening rings around their
necks, they treated the whole idea as a joke: ‘In conception, rather than in
execution, this project amused them very much; it is clear that they thought
of the birds always as guests, never as servants.’
 In other words, attitudes or former attitudes to pets among subsistence
hunter-gatherers and horticulturalists are not substantially different from
those characteristic of Western societies. Pets are raised, suckled if
necessary, and cherished like children. They are protected, named, and
cared for during life and, after death, they are often mourned. Pet animals
may, in addition, serve practical functions but they are not indulged for this
reason. The mere idea of killing and eating them is typically greeted with
horror.

The functions of pet-keeping

Anthropologists have devoted surprisingly little attention to the possible

functions of pet-keeping, although attitudes to pets and other domestic
animals have featured prominently in ‘structuralist’ discussions about the
origins of dietary and sexual taboos. Some have argued, for example, that
people avoid killing and eating their pets because the animal has been
included in the social world of humans, and its consumption is therefore
tantamount to cannibalism (Lévi–Strauss 1966, Sahlins 1976). Others have
proposed variations on the same theme. Leach (1964) suggests that people
tend to prohibit or taboo things which are difficult to classify, and that pet-
eating is taboo because pets occupy an uncertain, ambiguous territory
between humans and non-humans. The symbolic association between the
act of eating a pet and the act of sexual intercourse between close relatives
has also attracted attention. According to Tambiah (1969), people don’t eat
their pets because it would be metaphorically equivalent to committing
incest. These authors make no attempt to explain why societies keep pets in
the first place. They are solely concerned with people’s reluctance to
slaughter and devour such animals.
The best-known anthropological explanation for the function of pet-
keeping in hunting societies stresses its educational value (Laughlin 1968).
In many hunting cultures, hunters have been observed to catch or collect
small live animals – insects, frogs, lizards, nestling birds, and rodents –
which they later turn over to their children. These ‘pets’ serve as temporary
playthings and, like most toys, tend to be badly treated and short-lived.
Often they end up the objects of target practice or mutilation. According to
the theory, children receiving such gifts, and having the opportunity to play
with them, acquire valuable experience of animals and animal behaviour;
experience which will help them to become more efficient and successful
hunters in later life. The idea makes sense as far as it goes, but it confuses
true pets, i.e. animals treated with indulgence and fondness, with pets as
educational toys. The latter are ephemeral objects which may or may not
serve as useful childhood instruction. The former are cherished companions
and the subjects of strong emotional bonds (Serpell 1986). Indeed, from an
educational standpoint, one could equally argue that treating potential prey
species with indulgence and affection is actually counter-productive.
The widespread human practice of suckling young mammals has also
recently been subjected to functional interpretation. According to Fildes
(1986), rich women in Europe used to suckle puppies in order to relieve
painful distension of the breasts both before, and in the early stages of,
breast-feeding. Since this is evidently a common problem, it could also be
used to account for the suckling of animals in the many other cultures in
which this activity occurs. Fildes takes the argument even further by
suggesting that the breast-feeding of animals could also be used to promote
and extend lactation, prevent conception, and develop ‘good’ nipples in the
latter stages of pregnancy. In view of current interest in the links between
lactation and human fertility (Short 1984), these fascinating ideas would
repay more detailed investigation but, again, they cannot and were not
intended to be used to explain the functions of pet-keeping. For one thing, if
an animal is being employed as a lactational aid, there is no functional
reason why the owner should nurture or cherish it beyond weaning. Yet in
most pet-keeping societies the animal is cared for assiduously until it dies
of natural causes. Secondly, in many societies, such as the Kalapalo of
Brazil, the favourite pets are birds, which cannot be suckled (Basso 1973).
Anthropological reluctance to speculate about the functions of pet-
keeping may stem from misunderstandings about the nature of this activity,
or its impact on pre-agricultural societies. According to the old-fashioned,
ethnocentric view of cultural development, human societies evolved
progressively upward towards an increasingly complex and advanced state
of civilization. Hunter-gatherers and horticulturalists were seen as
occupying the lowest rungs of this developmental ladder, and their lives
were assumed to be correspondingly arduous and uncomfortable (Harris
1978). Conceived of in these terms, hunting economies allowed no room for
non-productive activities such as pet-keeping and, for this reason, the
practice was best ignored, explained away as aberrant, or accommodated
within some form of contrived utilitarian hypothesis. Fortunately, such
ideas can now be substantially revised.
Within the past 20 years, perceptions of hunting and gathering have
changed dramatically. Research on modern hunter-gatherers (e.g. Lee
1969), and the work of palaeoarchaeologists and pathologists (Cohen &
Armelagos 1984), has revealed that subsistence hunters often enjoy more
leisure, and are generally healthier and better nourished than many
agricultural populations. Indeed, according to this view, the Neolithic
switch from hunting to farming was not a cultural advance but rather the
inevitable outcome of the need to intensify food production in the face of
population pressure and diminishing resources (Boserup 1965, Cohen 1977,
Harris 1978). In other words, hunters and horticulturalists may be relatively
affluent in economic terms, and they are probably less restricted about
keeping pets than intensive agriculturalists.
Perceptions of pets have also altered in recent years. Modern theories
about the functions of pet-keeping rest primarily on the assumption that
humans have social as well as material needs, and that social interactions
with pets are able to fulfil at least some of these affiliative requirements
(see Serpell 1986). Loneliness and social deprivation are now known to
produce deleterious effects on human mental and physical health (Schachter
1959, Bowlby 1980, Perlman & Peplau 1981), and evidence has also
accumulated that involvement in positive social relationships can help to
buffer people from the damaging influence of negative life-events and stress
(Lynch 1977, Duck 1983). Analogous psychophysiological processes have
been demonstrated in other social species, and they also appear to operate in
relationships between species (Nerem et al. 1980, Gross & Siegel 1982). It
has recently been shown that social interactions with pets can produce
measurable reductions in heart-rate and blood-pressure in artificially
stressed human subjects (Katcher 1981, Friedmann et al. 1983). Further
research is needed on the potential social and emotional benefits of pet-
ownership, but existing results suggest that people use pets mainly to
complement and augment their social relationships, and so enhance their
own psychological and physical welfare. Pet-keeping is therefore functional
in a broad sense, although one cannot easily evaluate its function in
economic terms (Serpell 1986). This hypothesis does not, of course,
maintain that pet-keeping is universally beneficial, since, like any activity,
the net benefits must be weighed against the costs. It does, however, predict
that, where adequate time and resources are available, pet-keeping will
occur as a beneficial outcome of human social proclivities.

Pet-keeping and domestication

Although there are good theoretical grounds for arguing that Palaeolithic
hunters and incipient agriculturalists reared and nurtured wild animals as
pets, there are a number of objections to Galton’s theory that pet-keeping
led to domestication. The fact that hunters and horticulturalists exhibit
moral inhibitions about killing and eating their pets poses a problem, but it
is one which many societies have found ways of overcoming. Among the
Indians of California, for example, rules existed which required people to
disown pet deer and return them to the wild once they reached adulthood
(Elmendorf & Kroeber 1960). Presumably, these animals then became fair
game without arousing serious ethical conflict. Among the Ainu of Japan,
bear cubs were suckled and reared by women and brought up as members
of the family before, eventually, being sacrificed and consumed. The Ainu
came to terms with the moral contradictions inherent in this relationship by
means of self-justifying myths. According to Ainu legend, the bear was a
temporary visitor from the spirit world whose ultimate objective in life was
to return there. The Ainu believed that they were helping the animal to
perform this transmigration by killing it (Campbell 1984). Similarly, the
Tungus of Siberia shifted responsibility for the slaughter of pet reindeer by
evoking malevolent and capricious supernatural powers who required the
animals’ death as a token of appeasement (Ingold 1980). As already
suggested, the end of the Palaeolithic was probably associated with
population growth in certain regions and the depletion of natural food
resources. Faced with such pressures, it is likely that some Palaeolithic pet-
owners were obliged to convert their pets into livestock by inventing similar
methods of overcoming their scruples about exploiting them for food.
Downs (1960) disputed Galton’s idea on the grounds that, while pet-
keeping is widespread in hunting societies, the phenomenon of
domestication is strictly localized in its early stages. Under the
circumstances, he argues, pet-keeping and animal domestication should
have been coincident in all areas. Again, this objection can be overcome
using modern theories about the causes of the Neolithic revolution. The
postulated intensification of plant and animal exploitation which occurred at
this time was the result of overpopulation and resource depletion in certain
areas. Domestication was therefore localized because only some
Palaeolithic cultures were subjected to the necessary pressure to change
their mode of subsistence. Where these problems did not arise, hunting and
gathering remained profitable, and there would have been no incentive to
exploit pets more intensively. It should also be emphasized that not all wild
animal species are equally amenable to life under domestication (Reed
1954, Clutton–Brock 1981), and it may have been simpler for post-
Neolithic peoples to have adopted the domestic animals of neighbouring
groups as livestock or, indeed, as pets, rather than to have domesticated
existing tame wild animals. Cultural transmission of this kind would
account for the rapid and almost universal spread of species such as the dog
and cat.
The failure of many wild animal pets to breed in captivity has also been
proposed as a reason why pet-keeping could not have led directly to
domestication. Zeuner (1963), for example, accepted that pet-keeping may
have ‘provided one of the bases on which domestication on an economic
scale developed later on,’ but he also states that the form of Palaeolithic
man’s economy ‘prevented him from developing this relationship to full
domestication.’ The statement, however, is based on recent observations of
marginal hunter-gatherers, such as the Australian Aborigines, where pets
are often too undernourished and restricted to be able to breed. There seems
to be little justification for assuming that Palaeolithic pets were similarly
constrained and, unfortunately, no adequate information exists on the
breeding success or failure of pets among less marginal hunting
populations. Indeed, one could turn the argument on its head by asserting
that the pets which were eventually domesticated were precisely the ones
that managed to breed, despite the rigours of captive existence (Galton
1883). It is worth pointing out, in this respect, that most recently
domesticated species – budgerigars, canaries, brown rats, house mice,
hamsters, gerbils, etc. – were originally adopted as exotic pets but have
since acquired economic uses, for example as laboratory animals, as a result
of their ability to breed in captivity.
 The apparent adoption of more sedentary lifestyles towards the end of the
Palaeolithic may have promoted the captive breeding of pets, since species,
such as dogs, pigs, and birds, which rear their young in stationary dens or
nests may have been somewhat limited by the formerly nomadic habits of
their hunting and gathering owners. The earliest remains of a domestic dog,
for instance, have been recovered from a Natufian site in Israel where they
were associated with some of the oldest known permanent or semi-
permanent villages (Davis & Valla 1978). On the other hand, species with
precocious young, such as ungulates, would be unlikely to be adversely
affected by nomadism and, of course, there is no obvious reason to assume
that Palaeolithic pet-owners would necessarily have abandoned puppies,
piglets, or nestling birds whenever they moved.

Conclusions

Recent changes in our perceptions of both hunting and gathering and pet-
keeping reinforce, rather than detract from, Galton’s (1883) hypothesis. On
the basis of present evidence, it is probable that pet-keeping was common,
if not universal, among Palaeolithic hunters and incipient agriculturalists.
Judging from the wealth of different species employed for this purpose by
recent hunter-gatherers, it is also likely that all of our current domestic
species, as well as many which were never domesticated, began their
association with humans in this essentially non-economic role. Since
domestication is invariably associated with some form of captive or
controlled breeding, it is possible that the species we now classify as
‘domestic’ were simply those that bred most readily as pets within the
hunter-gatherer milieu.
Pet-keeping, both in the industrial West and other societies, is perhaps
best regarded as a leisure activity. But this need not imply that the practice
is necessarily without function, any more than it could be said that play or
other recreational pursuits serve no functional purpose. The majority of
hunter-gatherers and horticulturalists appear to possess adequate time and
resources to engage in leisure activities, and the fact that so many choose to
invest these resources in pet-keeping suggests that its social and emotional
rewards are far from negligible. As in the West, the role that pets seem to
occupy in hunting societies is most often analogous to that of infants and
young children; a fact which suggests that pets are used primarily as non-
reproductive outlets for parental behaviour. One hesitates to speculate too
far in this direction, but it is perhaps significant that fecundity is relatively
low in hunter-gatherer populations, as it is in the industrial West (Short
1984). The popular belief that pets are simply ‘child substitutes’ is
doubtless an exaggeration (see Serpell 1986) but, at the same time, it is
difficult to ignore the almost universal similarities between people’s
attitudes to pets and their attitudes to children.
The decision to exploit pet animals as sources of food or labour may have
been forced upon certain Palaeolithic groups by the necessities of survival
in a world of increasing food shortages. In the absence of such economic
and ecological pressures, there would have been little incentive to exploit
pet animals more intensively or to have embarked on the relatively
laborious task of maintaining and breeding them as captive, domestic
populations. On its own, Galton’s theory cannot be used to explain why
animal domestication occurred when and where it did. It does, however,
provide a plausible scenario for the development of more intensive systems
of animal exploitation when and where such systems were required.

References
Basso, E. B. 1973. The Kalapalo Indians of central Brazil. New York: Holt, Rinehart & Winston.
Boserup, E. 1965. The conditions of agricultural growth. Chicago: Aldine Press.
Bowlby, J. 1980. Loss, sadness and depression: attachment and loss, Vol. 3. London: Hogarth Press.
Campbell, J. 1984. The way of the animal powers. London: Times Books.
Cipriani, L. 1966. The Andaman islanders. London: Weidenfeld & Nicolson.
Clutton–Brock, J. 1981. Domesticated animals from early times. London: British Museum (Natural
History) & Heinemann.
Cohen, M. N. 1977. The food crisis in prehistory. New Haven: Yale University Press.
Cohen, M. N. & G. J. Armelagos 1984. Paleopathology at the origins of agriculture: editors’
summation. In Paleopathology at the origins of agriculture, M. N. Cohen & G. J. Armelagos (eds),
585–601. New York: Academic Press.
Davis, S. J. M. & F. R. Valla 1978. Evidence for the domestication of the dog 12,000 years ago in the
Natufian of Israel. Nature 276, 608–10.
Downs, J. F. 1960. Domestication: an examination of the changing social relationships between man
and animals. Kroeber Anthropological Society Papers 22, 18–67.
Duck, S. 1983. Friends for life. Brighton: Harvester Press.
Elmendorf, W. W. & K. L. Kroeber 1960. The structure of Twana culture with comparative notes on
the structure of Yurok culture. Washington University Research Studies, Monograph 2, 1–28.
Evans, I. H. N. 1937. The Negritos of Malaysia. Cambridge: Cambridge University Press.
Fildes, V. 1986. Breasts, bottles and babies: a history of infant feeding. Edinburgh: Edinburgh
University Press.
Fleming, P. 1984. Brazilian adventure. London: Penguin.
Friedmann, E., A. H. Katcher, S. A. Thomas, J. J. Lynch & P. R. Messent 1983. Interaction and blood
pressure: influence of animal companions. Journal of Nervous and Mental Diseases 171, 461–5.
Galton, F. 1883. Inquiry into human faculty and its development. London: Macmillan.
Gross, W. B. & P. B. Siegel 1982. Socialization as a factor in resistance to infection, feeding
efficiency, and response to antigen in chickens. Journal of Veterinary Research 43, 2010–12.
Harris, M. 1978. Cannibals and kings. London: Collins.
Harrison, T. 1965. Three ‘secret’ communication systems among Borneo nomads (and their dogs).
Journal of the Malaysian Branch of the Royal Asiatic Society 38, 67–86.
Ingold, T. 1980. Hunters, pastoralists and ranchers. Cambridge: Cambridge University Press.
Jesse, G. R. 1866. Researches into the history of the British dog. London: Robert Hardwicke.
Juan, G. & A. Ulloa 1760. Voyage to South America, Vol. 1. London.
Katcher, A. H. 1981. Interactions between people and their pets: form and function. In Interrelations
between people and pets, B. Fogle (ed.), 41–67. Illinois: Charles C. Thomas.
Katcher, A. H. & A. M. Beck (eds) 1983. New perspectives on our lives with companion animals.
Philadelphia: Pennsylvania University Press.
Laughlin, W. S. 1968. Hunting: an integrating biobehavior system and its evolutionary importance. In
Man the hunter, R. B. Lee & I. DeVore (eds), 3–12. Chicago: Aldine Press.
Leach, E. 1964. Anthropological aspects of language: animal categories and verbal abuse. In New
directions in the study of language, E. H. Lenneberg (ed.), 170–9. Cambridge, Mass.: MIT Press.
Lee, R. B. 1969. !Kung Bushmen subsistence: an input–output analysis. In Environment and cultural
behavior, A. Vayda (ed.), 47–79. Garden City, New York: Natural History Press.
Lévi–Strauss, C. 1966. The savage mind. Chicago: Chicago University Press.
Linton, R. 1936. The study of man: an introduction. New York: Appleton–Century–Crofts.
Lumholtz, C. 1884. Among cannibals. London: John Murray.
Luomala, K. 1960. The native dog in the Polynesian system of values. In Culture in history, S.
Diamond (ed.), 190–240. New York: Columbia University Press.
Lynch, J. J. 1977. The broken heart: the medical consequences of loneliness. New York: Basic
Books.
Meggit, M. J. 1965. The association between Australian Aborigines and dingoes. In Man, culture and
animals, A. Leeds & A. Vayda (eds), 7–26. Washington DC: American Association for the
Advancement of Science.
Nerem, R. M., M. J. Levesque & J. F. Cornhill 1980. Social environment as a factor in diet-induced
atherosclerosis. Science 208, 1475–6.
Perlman, D. & L. A. Peplau 1981. Toward a social psychology of loneliness. In Personal
relationships in disorder: personal relationships, Vol. 3, S. Duck & R. Gilmour (eds), 31–56. New
York: Academic Press.
Reed, C. A. 1954. Animal domestication in the prehistoric Near East. Science 130, 1629–39.
Roth, W. E. 1934. An introductory study of the arts, crafts and customs of the Guiana Indians. 38th
Annual Report of the Bureau of American Ethnology, 25–745.
Sahlins, M. 1976. Culture and practical reason. Chicago: Chicago University Press.
Sauer, C. O. 1952. Agricultural origins and dispersals. Cambridge, Mass.: MIT Press.
Savishinsky, J. 1983. Pet ideas: the domestication of animals, human behaviour and human emotions.
In New perspectives on our lives with companion animals, A. H. Katcher & A. M. Beck (eds),
112–31. Philadelphia: Pennsylvania University Press.
Schachter, S. 1959. The psychology of affiliation. London: Tavistock.
Scott, J. P. 1968. Evolution and domestication of the dog. Evolutionary Biology 2, 243–75.
Serpell, J. A. 1986. In the company of animals. Oxford: Basil Blackwell.
Short, R. V. 1984. Breast-feeding. Scientific American 250, 35–41.
Singer, M. 1968. Pygmies and their dogs: a note on culturally constituted defence mechanisms. Ethos
6, 270–9.
Speke, J. H. 1863. Journal of the discovery of the source of the Nile. London: W. Blackwood.
Tambiah, S. J. 1969. Animals are good to think and good to prohibit. Ethnology 8, 452–3.
Tuan, Yi–Fu. 1984. Dominance and affection: the making of pets. New Haven: Yale University Press.
Wilbert, J. 1972. Survivors of Eldorado: four Indian cultures of South America. New York: Praeger.
Zeuner, F. E. 1963. A history of domesticated animals. London: Hutchinson.
2 Definitions of animal domestication
SANDOR BÖKÖNYI

Attempts to define the term ‘domestication’ began more than 100 years ago
when research on the nature and origin of domestic animals began. The
stages in the search for a satisfactory definition can be traced rather easily,
for they have always been connected with the development of the natural
and social sciences. However, this chapter is not concerned with history, but
with my own definitions of domestication. My first published definition in
1969 (p. 219) was later extended in 1985 (p. 571) and was based on earlier
discussions on domestication published in German by Keller (1902), Klatt
(1927), Röhrs (1961–62), and others on the one hand and on my own
experience on the other. My definition is as follows: ‘The essence of
domestication is the capture and taming by man of animals of a species with
particular behavioural characteristics, their removal from their natural living
area and breeding community, and their maintenance under controlled
breeding conditions for mutual benefits.’
Since the 1970s, several authors have discussed this definition although
Ducos is the only one who rejected it, on the basis that it contained ‘a priori
propositions on the causes, mechanism and consequences of domestication’
(Ducos 1978, p. 54). Nevertheless, it seems to me that Ducos himself has
confused the process of domestication (which I am defining) with the
owning of fully domesticated animals. Furthermore, he has failed to cover
the keeping of newly domesticated livestock which are not isolated from
their wild forms and are not under human control, say in corrals. In fact,
domestic cattle, pigs, sheep, and goats are kept free in this way in western
Africa today, as Ducos claims, but only because their wild forms do not live
there and because they are at a high level of domestication. Ducos’s own
definition of domestication will be discussed later.
Another author, Hecker, is dissatisfied with the term ‘domestication’ as a
whole and suggests the use of the term ‘cultural control’ instead of it.
However, when the main components of this cultural control are
summarized (Hecker 1982, p. 219) it emerges that they are more or less
identical with the main points of my definition. In his definition, cultural
control does not, however, necessarily mean domestication; wild animals
can be culturally controlled without being domesticated. In this respect
Clutton-Brock’s definition of domestication is very close to mine (Clutton-
Brock 1981, see p. 7 this volume), the only difference is that she explains it
in a more colloquial style because of the popular nature of her book.
Let us approach the term ‘domestication’ again without prejudice or
preconception. Surely, there is agreement that domestication is a highly
developed man–animal relationship that emerged in a rather late phase of
mankind’s history. It certainly had Palaeolithic and Neolithic antecedents in
the form of the isolated taming and keeping of dogs and pigs, two animal
species which could live on the remnants and debris of human food and did
not need large quantities of vegetable fodder (Turnbull & Reed 1974, p.
100, Bökönyi 1978, 38ff., Nobis 1979, p. 610, 1984, p. 74, Altuna &
Mariezkurrena 1985, 11ff.). In this sense the first attempts at domestication
date as far back as the Magdalenian, c. 16 000 BP (Altuna & Mariezkurrena
1985, p. 111). Nevertheless, since dog and pig ate the same foodstuffs as
prehistoric man, they became competitors for food and this may have
hindered their large-scale domestication. A complete and profitable animal
husbandry, which essentially changed man’s economy, in other words
which made the switchover to food production possible, could not develop
from these isolated attempts at domestication. Large-scale domestication
and a more or less complete Neolithic domestic fauna started hand-in-hand
with the beginning of cereal production, which provided the large amounts
of rough fodder necessary for the caprovines, the leading species of the
earliest animal husbandry at the advent of the Neolithic.
The basis of the first large-scale domestication – and for that matter the
later ones as well, because the process did not cease to exist after the
acquisition of the first wave of domestic animals – was specialized hunting.
This certainly was a kind of cultural control, though not every specialized
hunting economy necessarily led to domestication, as clearly demonstrated
by the predominant gazelle-hunting of early Jericho (Clutton-Brock 1971,
Legge 1972, p. 123) or the similar onager- and gazelle-hunting at Umm
Dabaghiyah (Bökönyi 1973, p. 61). In fact, these cases also demonstrate the
complicated nature of domestication and its connection with the
behavioural characteristics of animal species. Hediger stated more than 40
years ago (in a neglected work – 1942, p. 160) that wild species which keep
a large inter-individual distance are not suitable for domestication because
they keep a large distance from man too. Gazelles (and onagers) surely
belong to the solitary type (Clutton-Brock 1978, p. 50). However, the
hunting of these animals was useful for man because in this way he could
increase his biological knowledge, and he undoubtedly used it in Umm
Dabaghiyah, keeping all five Neolithic domestic species, besides the
hunting of herds of wild onagers and gazelles.
But returning to the extremely complex nature of domestication one has
to turn to Ducos’s formula for the definition of domestication. He admits
that in his view, ‘domestication must be defined with reference to human
society’. Consequently, he suggests the following definition:
‘Domestication can be said to exist when living animals are integrated as
objects into the socio-economic organization of the human group, in the
sense that, while living, those animals are objects for ownership,
inheritance, exchange, trade, etc., as are the other objects (or persons) with
which human groups have something to do.’ (Ducos 1978, p. 54).
This definition is only partially correct, for it is over-simplified and
onesided. Domestication is the beginning of a symbiosis that needs at least
two partners, and it is simplistic to view it from the side of one of the
partners alone. It is indeed true that domestication is a special kind of
symbiosis (see also Röhrs 1961–62, p. 8, Herre & Röhrs 1973, p. 9) in the
sense that one of the partners, man, influences the other by isolating,
taming, controlling, breeding, and taking animals into new habitats, etc., but
the animal itself also plays an essential part in this process. As one can see
from the examples of gazelle and onager cited earlier, only animals with
particular behavioural characteristics can be domesticated, and this essential
point is missing from Ducos’s definition. It is important to realize that
gazelle and onager are not the only examples, wall-paintings and reliefs
suggest that the ancient Egyptians tried to domesticate whole series of
animals, from hyenas to antelopes, without much success. These species
certainly have some behavioural barrier that blocks their domestication.
This fact cannot be omitted in a definition because the behavioural patterns
of animals are of crucial importance, and through their study a lot of
questions can be elucidated. Similarly, a satisfactory definition must express
the fact that through domestication animals also influence man and society
as well, though their influence is not as strong as that of man on them.
According to Ducos (1978, p. 54) ‘living conditions are among the
consequences of domestication, not the mark of it’. In my opinion this
contradicts Ducos’s original definition, because if domestication means
integration of living animals as objects into the socio-economic group, one
necessarily has to change the living conditions of the animals by isolating
them, corralling them, etc. At the same time, items such as inheritance,
exchange, trade, etc. are consequences, not components, of domestication.
In practice, the discussion about domestication, its antecedents and
consequences, is a rather academic one, just as is the definition of cultural
control, herding, and animal husbandry, because these factors overlap each
other and clear-cut borderlines cannot be determined. It is the same with the
morphology of the bones of animals, and most probably it was the same
with the appearance and way of life of the living animals themselves in the
early prehistoric period.
Although it is statistically possible to determine the presence of already
domesticated animals in a given assemblage of faunal remains when the
sample is very large, in the early phase of domestication, developed from
specialized hunting, one cannot determine in every case whether a
particular bone represents a domestic or a wild animal. In other words, there
are no exact boundaries between specialized hunting and animal
domestication, and between wild and domestic forms, except in the highly
developed phases of domestication (animal husbandry) and hunting without
local domestication. An example of local domestication providing a large
number of indigenous domestic individuals, which makes possible the
definitive distinction between wild and domestic cattle populations, can be
seen in the late Neolithic sites of Hungary (Bökönyi 1974, p. 112ff). On the
other hand, in Switzerland, for example at the Neolithic site of Seeberg,
Burgäschisee-Süd where it has been claimed there was no evidence for
local domestication, it was much easier to distinguish the wild Bos
primigenius bones from the bones of domestic cattle (Boessneck et al.
1963, p. 160ff.).
Thus domestication can be seen as a gradual and dynamic, though not
always irreversible, process. For example, the ‘wild’ rabbits of Europe east
of Spain and north of southern France are in fact feral animals that escaped
from captivity in ancient times and quickly returned to the wild state. Since
domestication is a complex interaction between man and animal, its
consequences are influenced by society, economy, ideology, environment,
way of life, etc. Any successful definition of domestication must reflect all
these possible aspects of the process. The result of domestication is the
domesticated animal that first culturally and later morphologically differs
from its wild form.
Another question centres on the morphological changes caused by
domestication. Darwin was the first to deal with these in detail (Darwin
1868). In respect of these changes there are two main questions: first, which
are the main types of such changes? Secondly, how quickly do they appear
after domestication? Three selected changes will be considered here: (a)
size decrease, (b) crowded teeth, and (c) the hornless skull.
Regarding size decrease, it is undoubtedly true that under certain natural
conditions, e.g. isolated populations on islands, the size of the animals can
decrease. With animal remains from prehistoric sites, however, decrease of
size can only indicate domestication. As for crowded teeth, which are
common in pigs and dogs, rare in cattle, and which I have seen in only one
single horse skull, this is a reasonably sure proof of domestication. In wild
ruminants, such as horses and pigs, it never occurs, but there are wolf skulls
in which crowded teeth are found (Musil 1974, p. 49), although only in
modern wolves, especially those from zoos. In fossil wolf skulls crowded
teeth have only been seen in specimens later in date than the Gravettian
(Upper Palaeolithic), a fact that suggests early attempts at wolf
domestication. Hornless skulls are never found in wild cattle or goats, but
they do appear as a rare occurrence in populations of wild sheep. However,
it is not clear whether these are truly wild sheep or whether they have
interbred with feral or domestic sheep at some stage in the history of the
population. In general, hornlessness is a reliable indication of domestication
when found amongst animal bones from an archaeological site.
Morphological changes do not appear quickly after domestication.
Recent experiments show that measurable changes need about 30
generations after domestication before they appear. Nevertheless, some
recent authors suppose that such changes do not need more than a couple of
generations. One must not forget, however, that the relevant counting is in
animal generations and not in human generations, so that only 2–3 years is
required for one generation in small species and 4–5 years for large
mammals. It is possible that changes can appear quickly in animals kept in
zoological gardens or in modern experimental stations, but they were much
slower in prehistoric times. If this were not so, there would not be so many
transitional individuals apparent from the metrical analysis of animal bones,
for example from sites with local domestication of cattle.
Another consequence of domestication is the development of animal
husbandry, which itself has two phases. The first is primitive animal-
keeping without conscious, but with unintentional, breeding selection, the
existence of which can be seen in any evidence for castration or the killing
of a high proportion of young males. The second is developed animal
breeding with conscious selection, the main aim of which is the increase of
productivity and the existence of which is seen in the occurrence of
different breeds in a given population.
There is a difference between the term ‘animal husbandry’ as formulated
by Higgs & Jarman (1972, p. 8) and as used here. For Higgs & Jarman,
animal husbandry ‘stresses the important human element in the man–animal
relationships and includes in a single category pastoralists, herders, herd
followers and the like, where some form of intentional conservation was
practised’, whereas, for me, animal husbandry is a developed category
which follows domestication, although it cannot always be separated from
it.
Finally, it should be re-emphasized that animal domestication is a very
complex man–animal relationship, all aspects of which have not yet been
elucidated. This is particularly true for behavioural aspects, not only of the
animal but also of the human.

References
Altuna, J. & K. Mariezkurrena 1985. Bases de subsistencia de los pobladores de Erralla:
macromamiferos. In Casadores magdalenienses en la cueva de Erralla (Cestona, Pais Vasco), J.
Altuna, A. Baldeon & K. Mariezkurrena (eds), 87–117. San Sebastian: Munibe 37.
Boessneck, J., J. P. Jequier & H. R. Stampfli 1963. Seeberg, Burgäschisee-Süd; Die Tierrests. Acta
Bernensia II(3), 5–215.
Bökönyi, S. 1969. Archaeological problems and methods of recognizing animal domestication. In
The domestication and exploitation of plants and animals, P. J. Ucko & G. Dimbleby (eds), 219–
29. London: Duckworth.
Bökönyi, S. 1973. The fauna of Umm Dabaghiyah: a preliminary report. Iraq 35, 9–11.
Bökönyi, S. 1974. History of domestic mammals in central and eastern Europe. Budapest: Akadémia
Kiadö.
Bökönyi, S. 1978. The vertebrate fauna of Vlasac. In Vlasac: a mesolithic settlement in the Iron
Gate, D. Srejovic & Z. Letica (eds), 35–65. Beograd: Serbian Academy of Science and Arts
Monograph DXII.
Bökönyi, S. 1985. Problèmes archéozoologiques. In La protohistoire de l’Europe, J. Lichardus & M.
Lichardus-Itten (eds), 571–81. Paris: Nouvelle Clio.
Clutton-Brock, J. 1971. The primary food animals of the Jericho Tell from the proto-Neolithic to the
Byzantine period. Levant 3, 41–55.
Clutton-Brock, J. 1978. Bones for the zoologist. In Approaches to faunal analysis in the Middle East,
R. H. Meadow & M. A. Zeder (eds), 49–51. Harvard University: Peabody Museum Bulletin 2.
Clutton-Brock, J. 1981. Domesticated animals from early times. London: Heincmann & British
Museum (Natural History).
Darwin, C. 1868. The variation of animals and plants under domestication. London: John Murray.
Ducos, P. 1978. ‘Domestication’ defined and methodological approaches to its recognition in faunal
assemblages. In Approaches to faunal analysis in the Middle East, R. H. Meadow & M. A. Zeder
(eds), 53–6. Harvard University: Peabody Museum Bulletin 2.
Hecker, H. M. 1982. Domestication revisited: its implications for faunal analysis. Journal of Field
Archaeology 9, 217–36.
Hediger, H. 1942. Wildtiere in Gefangenschaft. Basel: Benno Schwabe und Verlag.
Herre, W. & M. Röhrs 1973. Haustiere – zoologisch gesehen. Stuttgart: Gustav Fischer.
Higgs, E. S. & M. R. Jarman 1972. The origins of animal and plant husbandry. In Papers in
economic prehistory, E. S. Higgs (ed.), 3–13. Cambridge: Cambridge University Press.
Keller, C. 1902. Die Abstammung der ältesten Haustiere. Zürich: B.G. Teubner.
Klatt, B. 1927. Entstehung der Haustiere. Berlin: Paul Parey.
Legge, A. J. 1972. Prehistoric exploitation of the gazelle in Palestine. In Papers in economic
prehistory, E. S. Higgs (ed.), 119–24. Cambridge: Cambridge University Press.
Musil, R. 1974. Tiergesellschaft der Kniegrotte. In Die Kniegrotte, R. Feustel (ed.), 30–95. Weimar:
Museum für Ur-und Frühgeschichte.
Nobis, G. 1979. Das älteste Haushund lebte vor 14,000 Jahren. Umschau 610.
Nobis, G. 1984. Die Haustiere im Neolithikum Zentraleuropas. In Die Anfänge des Neolithikums vom
Orient bis Nordeuropa, 9, Der Beginn der Ilaustierhaltung in der Alten Welt’, H. Schwabedissen
(ed.), 73–105. Köln: Böhlau Verlag.
Röhrs, M. 1961–62. Biologische anschauungen über Begriff und Wesen der Domestikation.
Zeitschrift für Tierzüchtung und Züchtungsbiologie 76, 7–23.
Turnbull, P. F. & C. A. Reed 1974. The fauna from the terminal Pleistocene of Palegawra Cave, a
Zarzian occupation site in northeastern Iraq. Fieldiana anthropology 63(3), 81–146.
3 Defining domestication: a clarification
PIERRE DUCOS
(translated by Marie Matthews)

During a symposium in 1975 in Dallas, involving a small number of

participants studying the fauna of the Levant, I formulated a definition of
domestication (1978a).1 It was not meant to be radically new but was the
result of a critical study of several previously formulated definitions,
specifically what may be termed the ‘classic definition’ best formulated by
Bökönyi (1969, and ch. 2, this volume).
In returning to the subject after a period of 12 years, I want first to ask
what is the purpose of defining domestication? If it is purely a semantic
exercise (even in several different languages) then it is a purely academic
discussion. In fact, neither the Bökönyi definition (called ‘definition B’ in
this chapter), nor mine (‘definition D’) are limited to semantics or are
purely academic. Both are meant to be practical.
In my opinion, definition B describes a moment in the evolution of man–
animal relationships and places a milestone in its history. Definition D,
however, strives to justify a methodological approach, being, as it is, in the
domain of logic.
Both definitions have been quoted in full in the previous chapter, and
they represent two different approaches, resulting in what seem to be two
contrasting positions. These positions are located, however, in two quite
different ‘fields’ and cannot be opposed to one another. Between these two
points of view there is a difference in concept and, for the definition itself,
of content.
The difference in concept is epitomized by the beginning of the phrasing
‘The essence of domestication …’ (B) and ‘Domestication can be said to
exist’ (D). Thus, definition D deals with a linguistic problem (‘can be
said’), within the field of logic, while definition B attempts, as concisely as
possible, to isolate the ‘essence’, the core of a phenomenon, an objective
fact, and is thus situated in the field of observation.
It was stated earlier that definition B appears to delimit a moment in the
history of man–animal relationships. For B, domestication is a specific
situation: the capture of a species (presuming, I suppose, a significant
number of individuals of both sexes). Once this capture occurs it produces
several consequences.
Thus, domestication is necessarily the first stage in a process. It is not
itself a process; domestication is the beginning of symbiosis, the result of
domestication is the domesticated animal, and animal husbandry implies a
higher category (in man–animal relationships) which follows
domestication. But on page 25 this volume, Bökönyi writes, ‘domestication
can be seen as a gradual and dynamic, though not always irreversible,
process.’ However, the example that he refers to (the wild rabbits in Spain
and France), clearly demonstrates that we are still dealing with the period of
‘capture’. Also, when he states that animal husbandry is a highly
specialized phase of domestication (thus contradicting the last of his quoted
statements), he does so only to explain the difficulty in differentiating,
within a bone assemblage, domesticated animals from wild ones. Here
again: domestication = capture. Thus, I do not think I am wrong when
stating that for Bökönyi domestication (concretely: ‘capture for profit’) is
the initial and necessary stage from which will evolve, on the one hand,
domestic species and breeds and, on the other, highly modern and evolved
forms of man–animal relationships, such as animal husbandry. This point is
not one that I would necessarily reject and, in fact, definition D does not
exclude it.
In Bökönyi’s view, definition D is necessarily incomplete (p. 24 this
volume), but it is on purpose that definition D excluded all human
behaviour other than that mentioned. It refused to see in the animals’ living
conditions, particularly that of ‘capture’, anything other than the
consequences of this essential relationship with the animal world: ‘living
conditions are among the consequences of domestication, not the mark of
it.’
I have always believed that the idea (popular in the 1950s) that domestic
animals are osteologically distinguishable from wild ones in all periods, is
scientifically unsound. This assumption concealed other possible situations,
where the man–animal relationship was no longer a hunting relationship but
had not, as yet, filtered down to the skeleton and modified the bones. (This
was also the position of Higgs and his school.) It is for this reason that it is
imperative to define a cut-off point within the ‘continuum’ of the man–
animal relationship.
It was consciously for this reason that I wrote: ‘domestication can be said
to exist …’. In French I would have formulated it as: ‘Il y a domestication
lorsque etseulement lorsque …’. I was thus looking for a phraseology to
express the relationship character of the state of domestication. In this case,
domestication does not exclude ‘proto-élevage’ situations or the capture of
isolated individuals without creating domesticated populations or species
that are anatomically different from their corresponding wild counterparts.
For this definition to be useful, it has to permit the creation of a particular
methodology. Since the object of search is located in something preceding
the act of domestication, it is human action that we must examine.
Incidentally, I do not believe that ‘domestication is the beginning of a
symbiosis that needs at least two partners’ (Bökönyi, p. 24 this volume).
These are not partners. I believe (although others may argue against this)
that domestication is not a natural state – it exists because humans (and not
the animal) wished it.
Having said all this, there are surely other, and possibly better, definitions
than definition D. I would like to stress, however, that a method based on
the definition that I gave (Ducos 1978a, pp. 54, 59) helped to identify a
situation of ‘proto-élevage’ in the Middle East, mostly in the pre-pottery
Neolithic: the ox of Mureybet (Ducos 1978b) and later at Catal Hüyük
(Ducos 1988), the Abu Gosh goat (Ducos 1978b), and possibly the Beida
goat (Hecker 1975).
Finally, I would like to point out that definition D does indeed identify a
state of domestication in the Egyptian representations mentioned by
Bökönyi on p. 24 this volume (although he believes that it does not). By
using the method that is based on this definition on large osteological
samples, a state of domestication can also be inferred even in the absence of
artistic representations.
 Archaeozoology is an observational discipline, where experimentation is
not often used and where the most convincing idioms are reputed to be
closer to the truth. We are therefore tempted to state that one of the
definitions is more correct than the other. The truth is that neither expresses
a reality, but rather that both reflect the approaches chosen to study the
complex phenomenon of human–animal relationships. Bökönyi’s ‘classic
definition’ probably applies most successfully to domestication from the
Neolithic and later periods. My 1975 definition seems, however, to open up
useful perspectives for more ancient periods during which humans appear
to have deployed novel methods of exploiting nature.

Note
1 I wrote the original text in what I thought was English but the final version was ‘translated’ into
real English by Richard Meadow and Melinda Zeder, and was approved by me.

References
Bökönyi, S. 1969. Archaeological problems and methods of recognizing animal domestication. In
The domestication and exploitation of plants and animals, P. J. Ucko & G. W. Dimbleby (eds),
219–29. London: Duckworth.
Bökönyi, S. 1989. Definitions of animal domestication. In The walking larder, J. Clutton-Brock (ed.),
ch. 2, London: Unwin Hyman.
Ducos, P. 1978a. ‘Domestication’ defined and methodological approaches to its recognition in faunal
assemblages. In Approaches to faunal analysis in the Middle East, R. H. Meadow & M. A. Zeder
(eds), 53–6. Harvard University: Peabody Museum Bulletin 2.
Ducos, P. 1978b. Le faune d’Abou Gosh. Proto-élevage de la chèvre en Palestine au Néolithique pré-
céramique. In Abou Gosh et Beisamun, deux gisements du VIIe millénaire avant l’ère chrétienne
en Israél, M. Lechevallier, Mémoires et travaux du Centre de Recherches préhistoriques français
de Jérusalem, No 2.
Ducos, P. 1988. Archéozoologie quantitative. Les valeurs numériques immédiates à Catal Hüyük. Les
Cahiers du Quaternaire, 1988, Bordeaux, Ed. du CNRS.
Hecker, H. 1975. The faunal analysis of the primary food animals from pre-pottery neolithic Beidha
(Jordan). Unpublished PhD thesis, Faculty of Political Science, Columbia University.
4 Some observations on modern
domestication processes
SYTZE BOTTEMA

Introduction

In this contribution I present information based upon experience with animal

breeding which may be useful for the archaeozoologist. My experience in
this field has often concerned the propagation of species not kept for
economic reasons, but rather for the sake of interest in their behaviour. These
experiments were not especially connected with archaeological purposes, but
some examples could be selected that are relevant to archaeozoological
research.
My observations are meant to provide the kind of information that is less
often referred to in the discussion of domestication processes in connection
with prehistory, and that could therefore help to broaden the discussion. This
approach is not new, and outlines can be found in Herre & Röhrs (1973).
They stress that the term domestication is difficult to define. Here, the term
domestication is used in the sense referred to by Oldfield (in MacFarland
1981), i.e. that the process of domestication starts as soon as the wild animal
has become habituated to humans. It is well known that wild animals
retained in captivity for more than a few months are sometimes difficult to
re-establish in the wild (Hediger, in MacFarland 1981).

Some observations on morphological changes occurring during

the domestication process

At which moment does breeding wild animals in captivity cause visible

changes? Everybody seems to accept domestication when distinct changes in
the exterior appearance of animals have developed that make them differ
from their wild ancestors. It is easily accepted that such changes have a
genetic basis. An example shows that one cannot generalize this
phenomenon. For instance, the udder in dairy cattle is developed to such a
size that it may easily pass for a domestication characteristic. However, cows
from a dairy breed develop only a small udder, densely covered with hairs,
when left to suckle their own calves. Although the shape of the udder is
hereditary, excessive size is induced by the milking regime, whether milking
is done by hand or by machine.
I will discuss in more detail some observations on various species of
waterfowl. Morphological changes in members of this group can be very
different. It is well known that after a few generations of domestication
greylag geese (Anser anser) become fatter and heavier, losing the power of
flight (Delacour 1964). Besides, after some time, early maturing and loss of
the permanent and monogamous pair-bond occurs. Colour variations such as
white, piebald, and buff appear, and feet turn orange whereas they were
originally pink. The fact that greylag geese become heavier after a few
generations is not a genetic change, but a result of feeding. Next to this
process, a selection on weight took place, resulting in various extraordinarily
heavy breeds. It should be stressed that the greylag and its relatives of the
Anser genus exhibit territorial behaviour during the breeding season. This
contradicts the statement (see, for instance, Garrard 1984) that territorial
animals are usually unsuitable for domestication. Characteristics such as
monogamy and late breeding age soon disappear and more social behaviour
develops. It is likely that the mutual tolerance among members of one family
facilitates the keeping of geese. Offspring of a pair of wild geese in captivity
will only stay with their parents until the next breeding season. Yet the
parents are much more tolerant of these birds than of strangers, especially
when the breeding season is over. Such family ties can be witnessed in many
species of animals, where they exist between mothers and their female
offspring, connecting several generations of female animals.
In white-fronted geese (Anser albifrons) the picture is quite different. This
species has been kept and bred by generations of Frisian goose-catchers who
use an intricate system of a net and a group of decoys. These decoys are not
simply tame white-fronts. The catcher relies on the strong family tie of the
geese, and to make use of this bond he needs a family of two parents with
their offspring of that year. It depends upon the system whether part of the
family is kept tied up on short strings between two nets or, in the case of one
net, upwind of the net (Lebret et al. 1976). The other members of the family,
especially the gander, are kept in the hide, at some distance from the net.
From this hide the catcher pulls the rope that turns over the net. When a
flock of wild white-fronts comes within a critical distance, one or more of
the birds in the hide are thrown into the air. They will immediately fly to
their relatives in the field behind the net, while both groups constantly give
their flight contact call. In this way they may attract the other geese towards
the net. It is necessary to have well-trained decoy birds, which are kept on
the wing but which are so tame that they can easily be gathered after the net
has been pulled. Good white-fronts that breed every year are highly valued
and have been bred by generations of goose-catchers.
There is only one change visible in the exterior appearance of such decoy
white-fronts: some of them have lost the black markings on the lower breast
and belly (Figs 4.1 & 4.2). Compared with the greylag they have not
developed white specimens or heavier birds that have lost the power of
flight.
In the case of the greylag or domesticated forms of mallard (Anas
plathyrhynchos), selection has taken place in favour of heavier animals
valued for meat, for instance the Toulouse breed, these being three times as
heavy as wild geese. In some Anatidae, however, captivity may result in
smaller birds, which have not been selected intentionally. This is so with the
shelduck (Tadorna tadorna) or male pintails (Anas acuta). Food conditions
in captivity may cause such a size decrease.
Figure 4.1 Decoy-breeder, Mr P. Wieland, with typical adult white-front gander, lacking the black
marks on the belly as are found in the wild white-fronts.
 Figure 4.2 Adult white-fronted goose shot at Nylamer (province of Friesland, the Netherlands)
showing the black markings typical for this species.

When looking at another goose species, the bean goose (Anser fabalis),
which is also kept and bred as a decoy, the observer could easily be misled
into concluding that an increase in size has occurred in the captives. Bean
geese are used as decoys for goose-catching in the central and northern parts
of the Netherlands. Geese react specifically to some extent, and it is said that
one cannot catch white-fronts with bean geese, but the reverse seems to
work. If one compares the bean geese used as decoys with the wild ones
present in large flocks in the new polders (reclaimed land) of Flevoland, or
the ones caught for instance in the Eempolder in the Netherlands, the larger
size of the decoys is obvious. Still, this is not a domestication trait. When the
present occurrence of bean geese in the Netherlands is studied, it turns out
that the majority of these geese wintering here at the moment belong to the
subspecies Anser fabalis rossicus, the tundra bean goose. According to van
den Berg (1983), over 100 000 tundra bean geese wintered in the new
polders and other areas in the middle and south of the Netherlands in 1982.
On the Dutch diluvial soils and peat bogs much lower numbers, mainly
small flocks, are recorded that belong to Anser fabalis fabalis. This
subspecies is called the taiga bean goose, originating from Scandinavia,
whereas the tundra form comes from Novaya Zemlya and the Yamal, Gyda,
and Taymyr peninsulas (Delacour 1954). The taiga form reaches the
Netherlands mainly in severe winters. The goose-catchers know this much
bigger subspecies as the ‘geelbek’ or ‘yellow-bill’ (Fig. 4.3), as it often has
more orange-yellow colour on the bill than the subspecies A. fabalis rossicus
(Fig. 4.4) which shows more black. The breeding stock used by the catchers
consists of this larger subspecies. Moreover, A. fabalis fabalis is said to be
tamed more easily than A. fabalis rossicus (van den Berg, written
communication). The distribution and the number of subspecies, if they are
still extant, must be studied to have any value in comparative studies (see
also Herre & Röhrs 1973).

The keeping and breeding of geese in captivity: how to start

domestic stock
When studying the decoy system, the question arises of how the goose-
catchers ever developed a ‘stock’ of tame decoy birds. When asked, the
catchers will always stress the fact that for establishing a stock of birds that
will produce goslings in captivity only first-year birds can be used. This has
been proven in those cases where the wild-caught adults of various Anser or
Branta species had to be kept under optimal conditions for more than 10–15
years to get any breeding success. For instance, Delacour (1964) writes ‘As
far as we know, the only wild-trapped red-breasted goose ever to have laid in
captivity was one at Woburn, Bedfordshire, after fourteen years, and all the
present stock (hundreds) in western Europe and North America are
descended from her.’ Yet the statement of the goose-catchers is not
completely true and it needs further explanation.

Figure 4.3 Head of a taiga bean goose (Anser fabalis fabalis) (photo G. Müskens). The slender bill
shows much orange-yellow. Decoys used by goose-catchers belong to this subspecies.
Figure 4.4 Head of a tundra bean goose (Anser fabalis rossicus) (photo G. Müskens). The bill shows
little orange, the mandibles are heavier. Although much more abundant than the taiga form, it is not
bred as a decoy.

The propagation of members of the goose family (as in many other bird
species) in captivity depends solely upon the females, in the case of birds
caught in the wild. Thus, for any breeding success young female geese have
to be caught, geese that are about 6–9 months old at the time of catching.
When they are brought together with a wild-caught adult male (three or four
years old), they may produce offspring at the time when the female has
matured. In general, this will take 3–4 years in the Anser group. If the
reverse combination is made, with a juvenile gander being brought together
with an adult wild-caught female, no success will be obtained.
Some Anatidae are renowned for very difficult breeding, although they
have been caught in large numbers and have proved to be hardy in captivity.
Breeding success of, for instance, the Brent goose (Branta bernicla bernicla)
and the Baikal teal (Anas formosa) mostly dates from the last five years,
although these species had been kept for at least the past 100 years. Mexican
peasants traditionally collect eggs of the northern red-billed whistling duck
(Dendrocygna autumnalis discolor) which are hatched under chickens. Some
peasants own up to 30 ducks. Keeping red-billed whistling ducks is an old
tradition, but the Mexicans state that no duck ever produced offspring in
captivity. The reason is still unknown (F. Feekes pers. comm.).
The difference in breeding success can be explained by the difference
between sexes, i.e. the anisogamy, the production of different gametes. To
guarantee any success for the investments in gametes made by the parents,
the environmental conditions are much more critical, in terms of selection
pressure, for the female than for the male. The male can produce millions of
gametes whereas the female can only produce a clutch of 4–7 eggs. If the
weight of an egg is compared with that of a sperm cell, the difference of
invested energy is obvious (Krebs & Davies 1982). Thus, the conditions
have to be properly fulfilled, otherwise the female will not waste the energy.
For wild-caught adult females, conditions in captivity are not properly
fulfilled as such geese will have been previously exposed to and imprinted
on other conditions in their natural habitats (see below).
Yet the experienced goose-catchers do not claim their statement without
reason. The advantage of bringing first-year birds together to establish decoy
families is based upon the existence of a rather rigid pair-bond. It is often
stressed that geese form a pair-bond for life and that if one of the pair dies,
the other remains solitary for the rest of its life. That geese form a firm pair-
bond is true, but there is no convincing evidence for the statement that they
remain solitary after losing the partner. In captivity new pair-bonds can be
formed quite easily and sometimes a pair splits up and one of the partners
takes up with another bird. Conditions in captivity are not the same as in the
wild, but solitary adults are not mentioned for the distribution range in the
breeding season of Anser fabalis in Russia (H. Hallander pers. comm.).
If a young female is put together with an adult wild gander, various
problems arise. The young female will have lost her parents and the gander,
caught on migration, will probably have become separated from his partner
and possibly also his own young. The juvenile goose would have broken up
her parental relation altogether some months later, but the gander may not be
charmed at all by the young female that is not sexually mature. If the gander
should finally court her, she may not react adequately because she is too
young. Thus, the goose-catchers are right to advise that the best way to start
off a decoy group is to take juveniles of both sexes, which still have to form
a pair-bond.
I will explain some of the factors that play a role if a wild goose is
suddenly locked up under conditions that differ completely from the wild.
One can imagine that this is a situation that occurred in prehistoric times
with various kinds of animals.
On 10 October 1983 the migration of Anser brachyrhynchos, the pink-
footed goose, was clearly at its peak over the northern Netherlands.
Thousands of pink-footed geese could be seen moving to the west towards
the relatively small wintering grounds between Workum and Gaast in
southwest Friesland. This concerns about half the population that breeds in
Spitsbergen. They arrive in the Netherlands much earlier than the other
Anser species because of the Arctic conditions in their breeding grounds.
They pass Norway, Jutland, and the coast of Schleswig-Holstein to reach
their destination, the wintering grounds. On 12 October, a boy brought me a
first-year bird (in fact about four months old), that had hit a high-tension
cable in the vicinity. According to the peak in migration the bird must have
hit the cable on 10 October. The shoulder of the goose was severely
fractured and the complete wing had to be amputated as the bird constantly
stumbled over it. Within four weeks the bird approached a human being
within a distance of two metres to receive its food.
How can such a change in behaviour be explained, if adult birds caught or
winged by shooting may stay shy for years? For an explanation we should be
informed about the behaviour of geese and the history of this particular
juvenile. The pink-footed geese, like other geese, have a rather fixed
migration pattern. Invariably they will fly the same routes between their
breeding and wintering grounds and they will only move from their
wintering grounds if they are forced to do so by, for instance, long-lasting
snow cover. For the fixation of the route the strong family tie is very
important. The young birds migrating in the year that they are born are
guided by their parents and thus they learn the route. Without their parents
they are forced to join other geese of their own kind. Young birds that do not
have this opportunity will be lost. In this way, we must explain the aberrant
nesting of barnacle geese (Branta leucopsis) outside their normal range, on
Iceland and Gotland. On Gotland it could be ascertained that escaped ringed
birds from the Museum of Skansen near Stockholm settled on the island (H.
Hallander verb. comm.). Birds that were released from captivity could not
establish a migration tradition and chose breeding grounds thousands of
kilometres outside their normal range. Bean geese, caught in Switzerland on
their spring migration, which escaped after two years of captivity, were
recorded back from Archangel and Kaluga (USSR). As these birds had made
the trip at least once they were able to return (Bauer & Glutz von Blotzheim
1968).
The particular pink-footed goose mentioned above had undoubtedly been
accompanied by its parents, covering large distances in a relatively short
time, constantly absorbing new impressions. On Spitsbergen, where it was
hatched and raised, it would hardly have had an opportunity to meet a human
being. During the rapid journey south such contacts were also few. Its
reaction towards humans was purely dictated by its parents and the (sub)
adult members of the group it travelled with. If a person approached, the
experienced adults would take to the air well out of shotgun range.
Nevertheless, our juvenile pink-footed goose will not have been too much
impressed by humans, and the behaviour of its parents was far more
important than the behaviour of the potential enemy. When the bird hit the
high-tension cable it was completely at a loss. It was hardly experienced and,
what is very important, it was still in a state of absorbing new impressions
that differed very much from the situation on the Arctic tundra. In this highly
sensitive state it was learning that humans are dangerous while cattle and
sheep are not, that grass and winter crops are edible, and at the same time it
had to absorb large parts of the map of northern Europe. In judging new
situations it would consciously pay attention to the behaviour of its parents
and other geese in the flock. The difference in behaviour between the freshly
caught youngster and the pink-footed geese bred in captivity, present in the
same fenced meadow, was striking. For instance, a common gull (Lams
canus) soaring over the meadow made no impression on the 5–7-year-old
pink-footed geese also present. However, the newcomer panicked
completely, running for vegetal cover where it crouched and remained
motionless for some time. As a chick it had obviously learnt from its parents,
on Spitsbergen, to fear marauding glaucus gulls (Larus hyperboreus), which
have a flight pattern comparable with the common gull. During the first
weeks the pink-footed goose refused any food and would only graze at some
distance from the threatening human being. It would, however, immediately
try to contact the geese that were already in the fenced meadow, since
belonging to a group is a first prerequisite for survival. In doing so, it
encountered the problem of entering a group of geese only partly belonging
to its own species and with a fixed ‘pecking order’. Captive geese remaining
on their territories (all year round because they cannot migrate) are more
aggressive towards newcomers than those in flocks where all the members
enter the same new situation, mainly feeding and sleeping at the same time.
The pink-footed goose was constantly chased away and was thus forced to
be more or less solitary, whereas the urge for the juvenile to join a group was
very great. After two weeks it learned that the person who entered the
meadow twice a day was not doing any harm and that food, new but edible
material, was brought that always had to be inspected immediately, even if it
was not eaten. It made social contact with a 4-year-old goose of its own
kind, which was also solitary. The bond was formed, as is usual in many
Anatidae, by synchronizing daily behavioural rites such as grazing, sleeping,
and preening. This example shows the behaviour of a freshly captured bird
in relation to a new situation created by man.

An analysis of some colour variations occurring in various

subspecies of mallard: an example of isolation

One of the first changes visible in ducks kept in captivity is the appearance
of white plumage. Many species, such as mallard (Anas plathyrynchos),
wood duck (Aix sponsa), mandarin duck (Aix galericulata), Bahama pintail
(Anas bahamensis), Egyptian goose (Alopochen aegyptiacus), occur in white
varieties. Birds first kept for ornamental reasons or for shooting, such as the
bob white quail (Colinus virginianus), are now available in various colours,
and recently even ‘broiler’ types have been developed for economic
purposes. If a bird, or a group of birds, differ in colour from the wild-
coloured relatives, for instance by being white, the factor that causes white
may imply other changes too. Thus, in Japanese quail (Coturnix coturnix
japonica) kept under the same conditions, white colour seems to be linked
with tamer behaviour and weaker condition compared with wild colour.
Mink breeders know that mink with the gene for ‘colmira’ colour are always
sleepy; animals of ‘pearl’ colour tend to hold their head in a peculiar way
(Crow 1979).
 Colour varieties that appeared in the course of domestication were clearly
selected either for reasons of subsistence or for the sake of exclusiveness.
Such selection may have side effects. A deviating colour such as white may
be linked with other recessive factors, or the gene for such a colour may be
pleiotropic. A pleiotropic gene influences more than one trait (Goodenough
1978).
What caused the white specimens in various members of the Anatidae?
The origin of colour variations in domestic waterfowl is often ascribed to
mutation (Delacour 1964). If large numbers of a species are bred in captivity
one can expect mutations to appear. This depends upon the mutation
frequency of the genes, generally assumed to be in the order of 1 in 105–106
gametes. In this connection, it is striking that mallards obtained from an area
in Sweden where feral ducks can be excluded nevertheless showed white
specimens in the third generation (I. Bossema pers. comm.). The same is true
for a small breeding group of eiders (Somateria mollissima) in captivity in
England that produced a deviating colour (buff) in the third generation.
When European pintail (Anas acuta), normally caught in duck decoys in the
Netherlands, were no longer for sale because of conservation measures, a
breeding stock was soon developed to meet the demand. Within a few years
‘blonde’ specimens appeared. This contradicts the statement that it takes 30
generations to breed deviating specimens (Bökönyi, cited in Meadow 1984).
Such rapid appearance of deviating colours in many species cannot be
explained by mutation during domestication, but it may be due to recessive
factors in the wild population. The colour of wild duck species is generally
dominant over other colours. The wild-colour pattern is caused by many
genes responsible for the various components or for the distribution of the
colours. If a mutant factor is present in a duck in heterozygous form, it will
not show up in the appearance of the bird, because of the dominance of the
wild-colour factors. In practice, the chances of a duck meeting a partner with
the same recessive factor are limited: offspring in which combinations of the
factor have occurred, e.g. white in homozygous form, will therefore be very
rare. Besides, there is strong selective pressure against these white mutants,
as predators can see them from a great distance. For the same reason a
dominant white mutant will have little chance of surviving. On the other
hand, a recessive factor, if present in heterozygous form, is not visible,
cannot be eliminated by selection, and thus survives to produce a colour
variant only if the owner meets a partner with the same genetic combination.
The trait white, a clear negative property in the wild, can be positively
valued in captivity by man. As this is a recessive trait, it will be very easy to
develop a pure breeding stock of white ducks.
One is thus tempted to conclude that colour variations in captivity are
more likely to be the result of recessive factors already present in the wild
population than to mutations. However, the following example shows that
the explanation of the mechanism behind colour variation may be more
difficult. I refer to a situation in the wild which shows clear parallels with
the domestication process of a small group, a situation that may have been
normal in prehistory. On the island of Laysan in the Pacific Ocean, a small
population of endemic Laysan duck (Anas laysanensis) occurs. This small
duck, the size of a teal, is thought to have originated from a group of
straggling mallards that happened to reach the lonely island. As the only
suitable biotope for duck on Laysan (which is c. 5 km long) is a lagoon, the
population was always limited in numbers because of space. In the 19th
century numbers were c. 600 at most. Genetically this is a limited number
with a high inbreeding rate. As can be seen in many island species of
animals, there was a decrease in size compared with its ancestor, the mallard.
In addition, it lost its breeding plumage, male and female both showing
brown feathers all the year round, and the female, especially, shows partial
albinism. Due to the activities of Japanese plumage hunters, the Laysan
ducks were reduced to ten individuals in 1909 (Delacour 1954, Kolbe 1972).
Conservation measures allowed the Laysan duck to increase again, but it is
clear that this species went through a bottleneck, reducing genetic variation.
Obviously, no lethal factors were present in the small population, otherwise
it would have disappeared due to inbreeding. In 1963 the population had
reached its 19th-century level again and some birds were caught for the
purpose of breeding in captivity. In fact this meant a second ‘genetic drift’.
The Laysan ducks were doing very well in captivity and many have been
reared since. After about 15 years other colours appeared in captivity. The
limited number of birds indicates that mutants are unlikely to occur. On the
other hand, it is difficult to accept that the variation in colours developed
from one recessive factor. The rate of inbreeding in the ten individuals left in
1909 was so high that one could expect colour variations to show up soon
after that year. This holds also for the offspring of the few ducks caught in
the sixties from which the breeding stock in captivity was developed. It is
not clear which genetical factors are responsible for the black, blue, and buff
specimens that are present at the moment. It is possible that in Laysan ducks
a colour variation is caused by several factors which have to combine to
have any visible effect.

Some observations on the role of temperature in the rearing of

nidifugous birds

Abiotic factors may strongly influence domestication processes. It is

understandable that animal species kept under conditions that deviate too
much from their ecological amplitude will not survive. A factor that is of
importance in initial domestication is temperature. It is, for instance,
advantageous to rear chicks of various nidifugous species (those that leave
the nest soon after hatching) under higher temperatures than those occurring
in at least part of their natural habitat. I once saw a little girl herding a group
of very young ducklings (about one-week old) near the prehistoric site of
Suberde in Anatolia. The little ducklings had no natural heat source in the
form of a natural mother but because of the high temperature they were
happily running around in search of food. The girl would chase away
possible predators and when a thunderstorm approached she guided the herd
inside the house before the rain poured down. Such a situation would be
impossible in a temperate climate. One can observe that domestic fowl in
farmyards in the Near East always have larger numbers of young than
domestic fowl on farms in temperate regions. This is connected with a
higher survival rate because of higher temperature and not with greater
clutch size.
Some experiments demonstrate that a high temperature may compensate
for deficiency in the daily diet. Teal (Anas crecca) hatched under a bantam
and kept outside are difficult to rear at a time when the April-May
temperatures in the northern Netherlands can still be low. A good bantam
will be a guarantee of warm shelter for the little ducklings but they still may
not survive the first critical weeks of their life. The ducklings will simply
stay under the bantam and die of hunger although food is available. At the
same time, teal ducklings can be seen chasing insects in peat bogs and
ditches in the wild, obviously not harmed by the same relatively cold
weather. If small teal are put in an electrically heated artificial breeder, they
will survive on the same diet that was not sufficient for those raised outside
with a bantam. Where a natural diet seems to compensate for loss of body
warmth, a warm temperature does the same under artificial conditions where
the diet is suboptimal.
I experienced the same phenomenon in the case of lapwing (Vanellus
vanellus) chicks. When kept in an artificial breeder at 30° C they would give
plaintive calls, refusing to eat. If the temperature was raised to 35°C, they
would start to feed. At the same time wild lapwings could be seen in the
field raising chicks at much lower temperatures. Here also, the natural
optimum diet seems to compensate for low temperatures.
Propagating mallard in captivity is far from easy, compared with greylag
or white-fronted geese, a fact stressed by Lepiksaar (cited in Prummel 1983).
One can imagine that domestication took place where the mallard was most
common. The mallard has a widely varying range of habitat and it occurs in
Eurasia, North America, and North Africa. The density of the breeding
population of mallards in the Netherlands is among the highest in its
distribution range. Yet the domesticated duck was introduced into this
country in medieval times only. It is more probable that they were imports
than they were domesticated locally. Early domestication centres are
suggested for Asia. The ancient Egyptians are reported to have kept and bred
ducks a few centuries BC (strangely enough, pintails dominate many of their
frescos). The Romans kept ducks in enclosures especially for fattening. They
are reported from France in the 6th century AD (Bottema 1980). It is rather
easy to raise mallard from eggs collected in the wild. Such birds will nest in
captivity but lose their offspring when the ducklings are a few days old.
It is acceptable that temperature may determine successful breeding when
one starts with chicks. If the possibility of rearing chicks from the egg is
ruled out, one may capture adults, but then specially reinforced housing is
needed otherwise they will escape. It is better to start off with juvenile
ducks, which are caught when they are completely feathered apart from the
primaries. Nevertheless, propagation will be a problem as such ducks will
not be successful in raising ducklings.
Thus, an area with a high spring temperature may have been a locus for
initial duck domestication from which the spread of adapted, or selected,
domesticated ducks could take place into temperate regions.
Food as a mechanism of control in the domestication process

The relation between humans and animals is sometimes explained as

symbiosis. Herre & Röhrs (1973) stress that the term symbiosis used with
reference to domestication does not always correspond to the biological
definition. Tchernov (1984), for instance, uses the term ‘one-way
symbiosis’. The domestication of the fowl is sometimes claimed to have its
roots in some kind of symbiosis that developed on the edge of settlements in
the jungle. The red junglefowl (Gallus gallus) is thought to have been
attracted by food in the settlements. When the inhabitants started to feed the
birds on purpose, they were able to become attached to the village. Wayre
(1969), however, reports that red junglefowl in Bhutan only occur far away
from any human habitation. Practical experience also pleads against
domestication developing from such a proposed form of symbiosis. An
example of a relation between humans and wild birds is provided by
common eiders (Somateria molissima) nesting close to lodgings because of
the protection this affords against egg-preying gulls. The tenant of the house
may gather the down., without harming the sitting bird. Indeed, lonely
houses on Arctic islands may have a complete colony of eiders in front of
them. However, after the eggs have hatched the relationship will come to an
end.
There is a less conspicuous though interesting connection, that often goes
unnoticed, between the magpie (Pica pica) and farmyards in the northern
Netherlands. The magpie is robbed of its eggs by the crow (Corvus corone).
Farmyards in the Netherlands will often have a magpie’s nest in a tree close
to the farm buildings, whereas the nests of crows are found much further
away. The crow is afraid of extending its territory close to the threatening
buildings. The magpie takes advantage of this situation (I. Bossema pers.
comm.). However, the biological term ‘symbiosis’ still cannot be applied
here.
Experience with ducks leads one to think that the reverse is much easier.
Domesticated ducks kept on farms easily revert to a feral state and so a
process of ‘dedomestication’ sets in. In times when duck-keeping for the
production of eggs was still economically viable in the Netherlands,
measures such as regular feeding in combination with locking up at night
were necessary, otherwise the ducks soon disappeared. For this reason duck-
keeping outside suitable biotopes, such as ditches and marshes, is much
easier, because in an unsuitable biotope they have to rely upon food supplied
by the owner. Farmers in the Frisian wetlands keep their ducks under feral
conditions. The wild population mixes with the domesticated ducks to such
an extent that hardly any ‘unspoiled’ mallards can be found. This is in
contrast to the situation on diluvial sandy soils where wild mallard are
restricted to small streams and pingo-ruins (small lakes of periglacial origin).
The few domestic ducks kept on farms have to rely on food offered there and
hardly go astray. Mixing of the two groups has hardly been noticed. The
same is true for mute swans (Cygnus olor) or domestic geese.
Humans have a means of firm control over domesticated animals, in the
form of food. The role of food is pointed out by Oldfield (in MacFarland
1981). Chickens in a farmyard will show at least one sign of
‘dedomestication’ by selecting nest sites carefully hidden in barns or bushes;
by giving them food regularly one can keep them bound to the yard.

Acknowledgements
I am very much indebted to my wife Nicolien and my eldest daughter Fionna for their assistance on
our farm, to my youngest daughter Wytske for driving the tractor, to Mr L. van den Bergh for his
information on bean geese, to Mrs G. Entjes-Nieborg for preparing the manuscript, and to Mrs S. M.
van Gelder-Ottway for correcting the English.

References
Bauer, K. M. & U. N. Glutz von Blotzheim 1968. Handbuch der Vögel Mitteleuropas. Vol. II:
Anseriformes (1). Frankfurt: Akademische Verlagsgesellschaft.
Berg, L. van den 1983. De Rietgans. Vogels 19, 240–4.
Bottema, S. 1980. Eenden. In Zeldzame Huisdierrassen, A. T. Clason (ed.), 191–204. Zutphen:
Thieme.
Crow, J. F. 1979. Overzicht van de Genetica. Groningen: Wolters-Noordhoff.
Delacour, J. 1954. The waterfowl of the world. Vol. I. London: Country Life.
Delacour, J. 1964. The waterfowl of the world. Vol. IV. London: Country Life.
Garrard, A. 1984. The selection of south-west Asian animal domesticates. In Animals and
Archaeology. Vol. 3: Early herders and their flocks, J. Clutton-Brock & C. Grigson (eds), 117–33.
Oxford: BAR International Series 202.
Goodenough, U. 1978. Genetics. London: Holt Rinehart & Winston.
Herre, W. & M. Röhrs 1973. Haustiere – zoologisch gesehen. Stuttgart: Gustav Fischer.
Kolbe, H. 1972. Die Entenvögel der Welt. Neudamm: Neumann.
Krebs, J. R. & N. B. Davies 1982. An introduction to behavioural ecology. Oxford: Blackwell.
Lebret, T., T. Mulder, J. Philippona & A. Timmerman 1976. Wilde ganzen in Nederland. Zutphen:
Thieme.
MacFarland, D. (ed.) 1981. The Oxford companion to animal behaviour. Oxford: Oxford University
Press.
Meadow, R. H. 1984. Animal domestication in the Middle East: a view from the eastern margin. In
Animals and archaeology. Vol. 3: Early herders and their flocks, J. Clutton-Brock & C. Grigson
(eds), 309–39. Oxford: BAR International Series 202.
Prummel, W. 1983. Excavations at Dorestad Vol 2. Early medieval Dorestad, an archaeozoological
study. Nederlandse Oudheden, Kromme Rijn Project.
Tchernov, E. 1984. Commensal animals and human sedentism in the Middle East. In Animals and
archaeology. Vol 3: Early herders and their flocks, J. Clutton-Brock & C. Grigson (eds), 91–116.
Oxford: BAR International series 202.
Wayre, P. 1969. A guide to the pheasants of the world. London: Country Life.
5 Feral mammals of the Mediterranean
islands: documents of early
domestication
COLIN P. GROVES

Introduction

The islands of the Mediterranean are inhabited by a variety of wild

mammals conspecific with widespread domesticates: sheep, goats, pigs, and
cats. In the past these forms were considered to be genuinely wild, vicariant
subspecies of species from which the respective domesticates had sprung.
In recent years, however, it has become clear that the Pleistocene faunas of
many, at least, of the Mediterranean islands were highly differentiated, with
endemic species or genera, and with no traces of Ovis, Capra, Sus, or Felis
(Schwartz 1973, Sondaar 1977, Dermitzakis & Sondaar 1979, Azzaroli
1981, 1982). The implication is clear: the sheep, etc. have been introduced
by human agency at some time during the Holocene, and so are not
naturally occurring subspecies.
It is, therefore, of great interest to know whether these species were
brought to the islands as wild individuals, and released for some reason (to
act as a food source, perhaps), or as domesticates. The fact that they were
previously mistaken for naturally occurring wild forms is an
acknowledgement of how very close they are to truly wild representatives,
so that if they were brought in as domesticates they were clearly in a stage
when domestication had not proceeded very far, giving them an intrinsic
interest as a kind of living museum of the initial stages of the domestication
process. If, on the other hand, they were brought in as wild stock, there is
still considerable interest attached to them as examples of rapid in situ
evolution, since they are not claimed to be precisely identical to any
continental wild form.
 As well as (potential) domesticates, a variety of wild species have been
brought into these same islands: shrews, hares, mice, dormice, spiny mice,
foxes, weasels, martens, badgers, and deer. Again, it is necessary to stress
that there is not a trace of these species in Pleistocene deposits, nor, all
things considered, is there much likelihood of their having introduced
themselves by jumping on to floating logs or other ‘sweepstake routes’.
Why anyone should want to introduce weasels to Corsica is not
immediately apparent; the introduction of badgers to Crete is perhaps less
mysterious, given the wide-ranging interests of the eclectic Minoans. But
introduced they were, and the sheep (etc.) must be considered in the same
context.
The best prospect for determining the feral versus wild status of a wild-
living form is relative cranial capacity. Herre & Röhrs (1973) show that the
relative brain size (compared to body size) of a domestic form is invariably
less than that of its truly wild conspecifics, sometimes massively so;
Hemmer (1983) disputes the extreme degree of reduction, but agrees that it
has occurred. The data of Kruska & Röhrs (1974), in particular, show that
brain size does not increase again in feral forms, over at least 100
generations. Admittedly, this is different from the thousands of years over
which the sheep and other animals have inhabited the Mediterranean
islands; but it is at least worth investigating the proposition that brain size
may have remained small over that period of time.

Material and methods

Skulls of both wild and domestic representatives of Ovis, Capra, Sus, and
Felis were measured. Cranial capacities (to represent brain size) were
measured by pouring birdseed into the braincase (after sealing up the optic
foramina with Plasticine), shaking it at intervals to pack it down, and then
decanting it into a measuring cylinder to measure the volume. Several linear
measurements were taken with calipers, partly to determine which would
act as the most efficient size standard, and partly to act as accessory means
of discrimination.

Results
Wild sheep
Wild-living sheep, known as mouflon, live today on Corsica, Sardinia, and
Cyprus (Figs 5.1 & 5.2). The name Ovismusimon Schreber has generally
been applied to the Corsico–Sardinian mouflon, but Uerpmann (1981) has
shown that the correct citation of this name is Pallas, 1811, and that the type
locality is Transcaspia; hence the name musimon is not available for the
mouflon.
The mouflon of Cyprus has been known as Ovis ophion Blyth, 1840, and
with the reallocation of the name musimon this appears to be the earliest
name for any mouflon. Pfeffer (1967) found Cypriot and Corsico–Sardinian
mouflon to be identical, apart from the former being slightly smaller;
Valdez (1982), however, points out that they can still be distinguished on
average, because the horns of the rams in Cyprus are supracervical, while
those on Corsica and Sardinia are usually homonymous. But in colour and
colour pattern, and in all features of the ewes, they are the same.
Pees & Hemmer (1980) found that, among living wild sheep, the Argali
(Ovis ammon) has a relatively higher cranial capacity than the Urial (Ovis
orientalis vignei); the latter in turn has a slightly higher capacity than the
Corsico–Sardinian mouflon, and domestic sheep are still further reduced.
Some Soay sheep, a long-established feral form from the Scottish Isles, had
cranial capacities equivalent to domestic sheep: increasing confidence that
brain size remains small over very long periods of time.
Figure 5.1 Ovis musimon: mouflon or wild sheep from Corsica/Sardinia, in winter coat, London Zoo.

Figure 5.2 Ovis ophion: mouflon or wild sheep from Cyprus, in summer coat, Hai-Bar Carmel,
Israel.
 Using Pees & Hemmer’s data and basic diagram, I added my own data to
produce Figure 5.3. As in Pees & Hemmer’s findings, Ovisammon has a
very high cranial capacity, followed by O. orientalis (other subspecies,
including the Turkish O. o. gmelini, being now added to O. o. vignei which
Pees & Hemmer had used), followed by Corsico–Sardinian mouflon,
followed by domestic and Soay sheep. There are big overlaps between
mouflon and O. orientalis, on the one hand, and mouflon and domestic
sheep on the other; such that a few domestic sheep capacities fall into the
mouflon polygon and even overlap O. orientalis. From these data, it would
appear that the brain in the Corsico–Sardinian mouflon is somewhat
reduced from that of wild sheep (Urial), but not much. Unless there has
indeed been a reversal of the initial reduction, the explanation that most
immediately impresses itself is that the mouflon is a feral relic of a species
that had not long been under domestication.
A single skull of a Cypriot mouflon was studied. The specimen is
unusually large (in the upper part of the size range for the Corsico–
Sardinian mouflon), but its cranial capacity is very small. It would be
approximately on the domestic/Soay line, extrapolated upwards. Plausibly,
the mouflon of Cyprus has a different, more fully domesticated, ancestry
from that of Corsica and Sardinia.
 Figure 5.3 Double logarithmic plot showing relative cranial capacity in wild, feral, and domestic
sheep.

Wild goats
Wild goats occur on the Aegean islands of Crete, Antimilo (Erimomilos) in
the Cyclades, and Yioura (Jura) in the Sporades. Schultze-Westrum (1963)
considered the goats of Crete and Antimilo to be truly wild, calling them
Capra aegagrus cretensis (Fig. 5.4) and C. ae. pictus respectively, but those
of Yioura to be feral.
I measured the cranial capacities of a large number of wild and domestic
goats (Fig. 5.5). Wild (Bezoar) goats from Turkey and the Caucasus (Capra
aegagrus aegagrus) have the highest cranial capacities; those of C. ae.
blythi from Iran and Pakistan are somewhat lower on average, a finding
which will not be commented upon here. Domestic goats have much lower
capacities, the more highly modified breeds (Alpine, Cashmere, Angora)
having less reduced levels than the primitive Nilotic goats, suggesting an
increase under domestication, after the initial decrease, for some breeds.
Some feral goats from Juan Fernandez island off the Chilean coast, a
population of some 400 years’ standing (Rudge 1984), have capacities of
the same relative size as the ‘higher’ domestic breeds.
The Aegean goats scatter neatly between the wild and the domestic goats,
overlapping marginally with both. There is no difference between those
from the three islands. That there are differences, especially in horn shape,
between them may mean that they originate from different breeds, but their
origin fairly clearly is from domestic goats: as in the case of the mouflon,
their ancestors were ‘only just’ domestic, implying that they derive from
quite an early era in the history of goat domestication.
Figure 5.4 Capra aegagrus cretensis: Cretan wild goat. Hai-Bar Carmel, Israel.
 Figure 5.5 Double logarithmic plot showing relative cranial capacity in wild, feral, and domestic
goats.

Pigs
In a general revision of Sus, I (Groves 1981) called the Corsico–Sardinian
wild pig Sus scrofa meridionalis (Fig. 5.6), and assigned the southern
Spanish wild pig to the same subspecies. A report on cranial capacity
studies was made in a later publication (Groves 1983): cranial capacities are
larger in Eurasian wild pigs than in those from Southeast Asia, with
domestic pigs falling below both. Some known feral forms fell into the
general domestic range, while others of equivocal status could be identified
by this method as feral.
In Figure 5.7, some Corsican and Sardinian skulls have been added to the
picture. They are, as far as brain size goes, simply small specimens of wild,
Eurasian Sus scrofa. Their identity in every other respect with wild pigs
from southern Spain suggests that they are, indeed, wild pigs which have
been introduced from there. This conclusion contrasts markedly with that
for the mouflon, and implies that we cannot simply envisage an enterprising
group of people who migrated from the Levant to Corsica and Sardinia,
bringing their (albeit primitive) domestic stock with them; different species,
of different status, were brought for different purposes (and perhaps by
different peoples!).

Figure 5.6 Sus scrofa meridionalis: wild pig from Sardinia, in summer coat, Tierpark Hellabrun,
Munich.
Figure 5.7 Double logarithmic plot showing relative cranial capacity in wild, domestic, and feral
pigs. Among wild pigs, those from Eurasia (Europe, Siberia, Japan) and those from
Malaysia/Indonesia (Sus scrofa vittatus) are depicted separately.

Cats
Whereas wild sheep, pigs, and goats are found on some islands but not on
others, wild cats appear to occur on them all; at least, they are well enough
known from Sardinia, Corsica, Crete, and even the Balearics to have
received subspecific names on each of those islands. These putative
subspecies are as follows:

Sardinia: Felis silvestris sarda Lataste, 1885.

Corsica: Felis silvestris reyi Lavauden, 1929.
Crete: Felis silvestris agrius Bate, 1905, and F. s. cretensis Haltenorth,
1953.
Mallorca (Balearic Is.): Felissilvestris jordansi Schwarz, 1930.

Haltenorth (1953) doubts that F. s. reyi is really different from F. s. sarda.

The same author considered that the type specimen of F. s. agrius, from
Crete, was actually a feral cat, but that another specimen, collected at the
same time, was a genuine wild cat, so he rejected the name F. s. agrius and
founded a new subspecies, F. s. cretensis, on the second specimen.
Wild-living cats are not known from Cyprus. Recently Davis (1987) has
figured the mandible of a cat from the Neolithic (c. 6000 BC) of Khirokitia
on Cyprus, which he suggests was most probably domestic. Whether wild
or domestic, this find indicates a special human-cat relationship at a much
earlier period than had hitherto been thought probable.
What is remarkable about this picture, and has never as far as I know
been specially commented upon, is that wild cats from North Africa have
also been referred to as F. s. sarda. There are thought (as, for example, by
Haltenorth 1953) to be two North African subspecies, F. s. sarda and F. s.
lybica, in more mesic and more xeric habitats, respectively. I have studied a
number of specimens assigned to F. s. sarda from both the Maghreb and
Sardinia, and it is quite true, there is no difference: they have the same
buffy-grey colour with a dark spinal stripe, and varyingly (but always rather
poorly) expressed striping; head-and-body length averages 58 cm in
Sardinia and 59 cm in North Africa (smaller than most European wild cats,
larger than most from western Asia); tail length averages 57 per cent of
head-and-body length in both places, which is shorter than Arabian races
but longer than most other wild cats; zygomatic breadth averages 78 per
cent of condylobasal length in both, which is the same as the western Asian
and other African wild cats, but more than those from Europe; and the
length of the upper carnassial averages 11.3 mm in North Africa and 11.1
mm in Sardinia, these figures being less than most European wild cats but
greater than those from western Asia and North Africa. There is, in other
words, no room for doubt but that the Sardinian and North African wild cats
really are the same.
What, then, of the domestic or wild status of the Sardinian wild cat?
Hemmer (1976) has illustrated the values of Schauenberg’s Index (skull
length divided by cranial capacity, an isometric relationship in cats), and I
have used this as a basis for Figure 5.8, to which I have added the values of
specimens measured by myself, including some from Sardinia. It is first
noticeable that, as Hemmer found, F. s. lybica has a higher index (i.e.
relatively lower cranial capacity) than F. s. silvestris from Europe, with
North African F. s. sarda, as well as sub-Saharan African and the NW
Indian/western Asian F. s. ornata, falling in between. Domestic cats, as well
as feral cats from the Northern Territory of Australia, have higher values for
the index (but overlapping, even nearly covering, the F. s. lybica range), and
Egyptian mummified cats fall into either the F. s. lybica or domestic range.
The values for Sardinian wild cats are remarkably low: they could not
possibly be feral domestic cats, unless there has been total reversion to the
wild brain-size, but they fit exactly the range of North African F. s. sarda. I
conclude that cats were introduced into Sardinia from the Maghreb region
in the wild state: introduced they certainly must be, but feral (in the sense of
being reverted from a domestic state) they are not.
A single specimen from Corsica, the type of F. s. reyi, fits almost exactly
into the Sardinian range: its head-and-body length is 58 cm, right on the
Sardinian average; its zygomatic breadth is 76 per cent of the condylobasal,
only a little below the Sardinian average but well within the range (73–83
per cent); the skin pattern, as illustrated by Haltenorth (1953, Fig.30) is
similar to that of F. s. sarda; only the tail is apparently shorter (46.6 per cent
of head-and-body length: the F. s. sarda range is 47–68 per cent). I have not
seen the specimen itself, and no cranial capacity data exist, but I believe
that F. s. reyi is a synonym of F. s. sarda, and that Corsican, like Sardinian
cats, were introduced wild from North Africa.
No skulls are available from Crete: only skins. These, as illustrated by
Haltenorth (1953, Fig.31–3), and as seen from the types of F. s. agrius and
F. s. cretensis, are rather densely and clearly striped, and very obviously
different from F. s. sarda, or indeed from any true wild cat known to me. In
the absence of skulls, and of external measurements of fresh specimens, I
can only record my belief that all Cretan wild cats, despite Haltenorth’s
proposed separation of them into two types, are actually feral.
The Mallorcan wild cat is smaller than F. s. sarda, or indeed most other
wild cats: in the type of F. s. jordansi the head-and-body length is 51 cm (at
the bottom end of the F. s. sarda range. 49–62 cm); the tail is short, only 44
per cent of the head-and-body length; the skull is narrow (zygomatic
breadth only 70 per cent of condylobasal length, below the F. s. sarda
range); and the carnassial is 11.0 mm long, within the F. s. sarda and
western Asian ranges but also within (upper end of) the domestic cat range.
As illustrated by Haltenorth (1953, Fig.28–9), the skin is strongly marked,
like Cretan specimens. Figure 5.8 shows the Schauenberg Index of a series
of skulls in the Geneva Museum: the index is above that for Sardinia, with
no overlap, and in the range of domestic cats and of F. s. lybica. I have no
doubt at all that the Mallorcan wild cat is a feral form.

Figure 5.8 Relative cranial capacity in wild, feral, and domestic cats. The ranges are for values of
Schauenberg’s Index (greatest skull length divided by cranial capacity), as described in the text.

Discussion

The results of the investigations recorded above suggest that mouflon on

Corsica and Sardinia are derived from very primitive domestic sheep, those
on Cyprus from more fully domesticated sheep; the wild goats of the
Aegean islands are derived from very primitive domestic goats; the
Corsican and Sardinian wild pigs were introduced, but probably not as
domestic animals, from southern Spain; the wild cats of Corsica and
Sardinia were introduced, again not as domestic animals, from the
Maghreb, while those of the Balearics and (probably) Crete are feral.
Looking at this in another way, the introduced mammals of Sardinia and
Corsica include a very primitive domestic sheep from the Middle East, a
wild pig from southern Spain, and a cat from North Africa; the Balearics
have a feral cat, of uncertain origin; Crete has a very primitive goat from
the Middle East (as do two other Aegean islands) and a feral cat of
uncertain origin; and Cyprus has a primitive but clearly once domestic
sheep.
It was mentioned in the introduction that other wild mammals, of species
in which there is (presumably!) no question of domestication, also occur on
these Mediterranean islands. It is worth asking what the mainland
relationships of these might be, too. As far as I can judge, mainly from
descriptions or opinions expressed by commentators in the literature, the
island by island relationships appear to be as follows.

Balearics
All subspecies of mammals on the Balearic islands appear to be identical or
very close to those on the nearby Spanish mainland. Hemmer et al. (1981)
record the same thing for a Balearic frog, Rana perezi, but find that a toad,
Bufo viridis, is identical to one on Sardinia and in Israel, and postulate a
common Bronze Age origin for them.

Sardinia and Corsica

The two islands have a virtually identical mammalian fauna. The putative
indigenous subspecies of Suncus etruscus, Lepus europaeus (only on
Corsica), Apodemus sylvaticus, Glis glis, Mustela nivalis, and Martes
martes are variously distributed through France, Yugoslavia, and Greece,
but all occur in Italy, which on the basis of the (here possibly spurious)
parsimony principle may be suggested to be their origin. (Interestingly,
Martes martes latinorum occurs also on the Balearics: a possible parallel to
Hemmer et al.’s [1981] toad.) A second group of species has other
relationships: Lepus capensis (Sardinia only; this being one of the few
marked differences between the two islands), Eliomys quercinus, and
Vulpes vulpes of these islands have their closest relatives in southern Spain,
like the wild pig; and the distinctive subspecies of red deer, Cervus elaphus
corsicanus, occurs in both southern Spain and North Africa.
I must admit that there are no direct Middle Eastern affinities in this list.
Whoever brought the primitive domestic sheep to these islands seems to
have brought nothing else (except toads?). At some other time, trade links
involving Sardinia and Corsica, Italy, the Maghreb, southern Spain, and the
Balearics succeeded in spreading around a number of other species, all as
wild animals. It is tempting to hold the Carthaginians responsible for these
transfers (which would, conveniently, provide an ultimate Middle Eastern
connection, the Phoenicians), but the evidence will not at present support
such a surmise, merely raise it as a hypothesis to be tested.

Aegean islands
Most of the mammals of Crete (the only one of the Greek isles to be
tolerably well known faunistically) have detectably mainland Greek
affinities: Erinaceus europaeus, Lepus europaeus, Apodemus mystacinus,
Apodemus sylvaticus, Glis glis, Meies meles. There are, however, a few
species whose closest relatives are in Turkey: Crocidura gueldenstaedti,
Acomys cahirinus, Martes foina. It would, again, be tempting to add
‘primitive domestic goats’ to this latter list, but at present it would be going
beyond the evidence, and I put it forward only as a working hypothesis.

Cyprus
The present-day mammal fauna of Cyprus is little studied.
Archaeologically, sheep as well as goat, pig, cat, and fallow deer occur from
the Neolithic at least to the early Bronze Age (Davis 1984): apparently
occasionally as wild as well as domestic or semi-domestic forms (Schwartz,
1973). The presence of fallow deer, at least, implies a mainland western
Asian origin for part of these imports: if the species is correctly reported as
Dama mesopotamica rather than D. dama, then the Levant/Palestine region,
not Turkey, is implied.

Conclusions
Some of the ‘domesticable’ mammals at present living wild on the
Mediterranean islands are indeed feral, from very primitive domestic stock;
these are the mouflon and the Aegean wild goats, and as such they are
extremely valuable documents of the early stages of the domestication
process. However, the wild cats and pigs of Corsica and Sardinia show no
signs of ever having been domesticated; they would seem to be part of a
complex of trade links, involving transport of wild animals, around the
ancient western Mediterranean. The wild cats of Crete and the Balearics are
almost certainly feral cats, whether particularly primitive or not being at
present uncertain.

References
Azzaroli, A. 1981. Cainozoic mammals and the biogeography of the island of Sardinia, western
Mediterranean. Palaeogeography, Palaeoclimatology, Palaeoecology 36, 107–111.
Azzaroli, A. 1982. Insularity and its effects on terrestrial vertebrates: evolutionary and biogeographic
aspects. In Palaeontology, essential of historical geology, E. M. Gallitelli (ed.), 193–213. Modena:
S. T. E. M. Mucchi.
Davis, S. J. M. 1984. Khirokitia and its mammal remains: a neolithic Noah’s Ark. Fouilles récents à
Khirokitia (Chypre) 1977–1981, A. Le Brun (ed.), 147–62. Paris: Recherche sur les Civilisations.
Davis, S. J. M. 1987. The archaeology of animals. London: Batsford.
Dermitzakis, M. D. & P. Y. Sondaar 1979. The importance of fossil mammals in reconstructing
paleogeography with special reference to the Pleistocene Aegean archipelago. Annales
géologiquesdes Pays helleniques 29, 808–40.
Groves, C. P. 1981. Ancestors for the pigs: taxonomy and phylogeny of the genus Sus. Canberra:
ANU Press.
Groves, C. P. 1983. Pigs east of the Wallace Line. Journal de la Société des Océanistes 39, 105–19.
Haltenorth, T. 1953. Die Wildkatzen der alten Welt. Leipzig: Geest & Portig K-G.
Hemmer, H. 1976. Man’s strategy of domestication – a synthesis of new research trends. Expérientia
32, 663–6.
Hemmer, H. 1983. Domestikation: Verarmung der Merkwelt. Braunschweig/Wiesbaden: Wieweg.
Hemmer, H., B. Kadel & K. Kadel 1981. The Balearic toad (Bufo viridis halearicus (Boettger,
1881)), human bronze age culture, and Mediterranean biogeography. Amphibia-Reptilia 2, 217–30.
Herre, W. & M. Röhrs 1973. Haustiere – zoologisch gesehen. Stuttgart: Gustav Fischer.
Kruska, D. & M. Röhrs 1974. Comparative-quantitative investigations on brains of feral pigs from
the Galapagos Islands and of European domestic pigs. Zeitschrift für Anatomie und
Entwicklungsgeschichte 144, 61–73.
Pees, W. & H. Hemmer 1980. Hirngrosse und Aktivität bei Wildschafen und Hausschafen (Gattung
Ovis). Säugetierkundliche Mitteilungen 28, 39–45.
Pfeffer, P. 1967. Le mouflon de Corse (Ovis ammon musimon Schreber, 1782); position systématique,
écologie et éthologie comparées. Mammalia 31 (suppl.), 1–262.
Rudge, M. R. 1984. The occurrence and status of populations of feral goats and sheep throughout the
world. In Feral mammals – problems and potential, P. N. Munton, J. Clutton-Brock & M. R.
Rudge (eds), 57–84. Morges: IUCN.
Schultze-Westrum, T. 1963. Die Wildziegen der agaischen Inseln. Säugetierkundliche Mitteilungen
11, 145–82.
Schwartz, J. H. 1973. The palaeozoology of Cyprus: a preliminary report on recent analysed sites.
World Archaeology 5. 215–20.
Sondaar, P. Y. 1977. Insularity and its effect on mammal evolution. In Major patterns in vertebrate
evolution, M. K. Hecht, P. C. Goody & B. M. Hecht (eds), 671–707. New York: Plenum
Publishing.
Uerpmann, H.-P. 1981. Ovis musimon Schreber, 1782, oder Ovis musimon Pallas, 1811?
Säugetierkundliche Mitteilungen 29, 59–60.
Valdez, R. 1982. The wild sheep of the world. Mesilla (New Mexico): Wild Sheep and Goat
International.
6 Escaped domestic animals and the
introduction of agriculture to Spain
IAIN DAVIDSON

Frontiers between fisher-gatherer-hunters and farmers

Contact between fisher-gatherer-hunters and agricultural or pastoral people

creates spatial frontiers. Prehistorians may also consider that there is a
temporal frontier (see also Dennell 1985, p. 114). Interaction at either of
these frontier classes (spatial or temporal) has been considered by
Alexander (1976, 1984) and others (Hall et al. 1984, Green & Perlman
1985), with special reference to the implications for our understanding of
the archaeological record. Archaeologists have turned to this theoretical
issue relatively recently, and, therefore, the concerns have been more with
recent issues in archaeology related to social organization, than with
ecological aspects of the record. In particular, most emphasis has been on
the behaviour of people rather than with the ecology of plants and animals
at the frontier. This chapter argues that the documented example of the
behaviour of animals during the European colonization of Australia
suggests that the ecological autonomy of plants and animals was an
important part of the interaction between human groups at the frontier.
Dennell (1985) provides a useful introduction to the history of the study
of the moving frontier between the first agriculturalists and the non-
agricultural indigenous inhabitants of prehistoric Europe. Although there
are some difficulties with the account (e.g. claims of lack of evidence for
violence at the frontier could be countered by some interpretations of the
chronology of the Levantine art of eastern Spain, which depicts opposed
groups of humans using bows and arrows), Dennell makes many good
points about the archaeology of the European frontiers. This chapter,
however, concentrates on an aspect which Dennell only touches on (1985,
pp. 130, 132): natural movement across the frontier of the plants and
animals used by agriculturalists.
 There are, in reality, four categories of people concerned at these
frontiers:

A ‘Hunters’: those who gain their subsistence from plants and animals that
are not cultivated or over which breeding control is not exercised.
B Those who have some mixture of subsistence based on cultivation or
control and subsistence based on other resources:
(i) ‘cultivating hunters’: hunters who cultivate, thus being closer to
category A;
(ii) ‘hunting agriculturalists’: agriculturalists who hunt, thus being closer
to category C.
C Those who gain no important subsistence from such non-cultivated or
uncontrolled resources (’agriculturalists’).

Dennell (1985, p. 114), indeed, goes further and points to the academic
frontier which also separates the study of the behaviour of prehistoric
fisher-gatherer-hunters from the study of prehistoric agricultural societies.
There is, in addition, a curious phenomenon, which has also been noted in
the literature on modern fisher-gatherer-hunters (Ingold 1984, p. 5).
Scholars tend to classify people out of the category of fisher-gatherer-
hunters if they seem to have any attributes of agriculture or pastoralism.
The existence of the four categories complicates the interpretation of the
archaeological record still further as blurring takes place of the criteria for
recognizing which side of the frontier particular archaeological sites may be
on.
In general, we are dealing with a literature which tends to describe all
those who are not in category A (hunters) as having changed in status from
fisher-gatherer-hunters to agriculturalists however little the degree of
dependence on cultivated and controlled resources. Little attention is paid to
understanding the groups which are not in category C (agriculturalists) but
cannot be said to be in category A (hunters) because of the lack of
importance of the subsistence they gain from cultivation or control. This
seems to be the case for archaeological sites as well as in the ethnographic
literature.
Definitions must be very carefully applied here. This is because there are
important status shifts, as people who routinely obtain their important plant
foods from cultivated crops also gather plants from the uncultivated
environment as relishes or luxury foods (truffles, or blackberries, for
example). By the same token, hunting is now a high status activity in some
social groups (grouse shooting, fox hunting) and an important recreation in
others (pig or kangaroo shooting, bird shooting in Mediterranean countries).
In these circumstances, it is difficult to separate the subsistence and social
components of the activity.

The Australian model

When the Australian Aborigines were first seen by agriculturalists from

Europe, they were described by those agriculturalists. Few, if any, other
fisher-gatherer-hunters were described for posterity by the first group of
agriculturalists who saw them. Indeed, Woodburn (1986) has recently
suggested that many of the recent groups of fisher-gatherer-hunters have
determined their behaviour by their encapsulation by surrounding
agricultural societies. In all other cases, there was some local contact
between groups where the contact was prehistoric. In the first European
colonization of Australia, contact was between groups of non-Aboriginal
agriculturalists who wrote of their experiences (see, for example, King
1984) and Aboriginal fisher-gatherer-hunters who did not. In northern
Australia, Aborigines were visited by Macassans (MacKnight 1976) or had
contact with horticulturalists of Papua (Harris 1979). Apart from these
contacts, the Australian Aborigines lived on a continent of fisher-gatherer-
hunters. Unlike, say, the San, the Inuit, the Shoshoni, the Tasaday, or the
Mbuti, they could have had no experience with a radically different way of
life. We can gain some idea from this encounter about how introduced
materials and species crossed the frontier between history and prehistory,
and between the world of agriculturalists and that of fisher-gatherer-hunters.

Materials crossing the frontier

Reynolds (1982) has documented in detail the extent of penetration of the
frontier between Aborigines and non-Aborigines in Australia. Goods
doubtless travelled along traditional trading and exchange routes (Mulvaney
1976) between Aboriginal groups which were in contact with non-
Aborigines as well as with those which were not. Tamar found a button on a
necklace in 1804; Carnegie found an iron tent peg and glass objects;
Warburton a butcher’s knife and a steel axe. Among the first materials
which crossed the frontier were animals.

Escaped domestic animals

The first cattle to land in Australia, five cows and a bull, escaped from the
control of the British settlers on 5 June 1788, a mere 19 weeks after landing
(King 1984, p. 73). They, or their offspring, were not rediscovered until
1795 when an Aboriginal fisher-gatherer-hunter reported them to a
European convict. Clegg (1984) and Lyon & Urry (1979) have shown that
these first escaped cattle were depicted by Aborigines in the paintings at
Bull cave near Sydney.
Rolls (1984, p. 16) has given an account of this and other escapes from
pastoral control. Almost every other useful animal, and some which were
not, went wild (Rolls 1969). The spread of rabbits is not only well known to
Europeans, but remembered by Aborigines (Reynolds 1982, p. 10). Horses
were seen by explorers, such as Leichhardt in 1844 and McKinlay in 1861,
way beyond the regions then settled by non-Aborigines (Reynolds 1982, p.
10). Pigs, foxes, goats, dogs, buffalo, camels, and chickens all have feral
populations in Australia, but, curiously, sheep do not. One reason for this
might be the vulnerability of woolly sheep to fly strike and dingo attack if
left unshorn. This would certainly have been a factor with the fat-tailed
sheep of the First Fleet (1788) and the Gorgon (1791). despite the fact that
their coats were primarily of hair and not wool (Garran & White 1985, pp.
7–9). The second introduction of fat-tailed sheep grew ‘too fat to breed’
(Governor Phillip, quoted by Garran White 1985, p. 13), a problem which
was solved by the import of hairy sheep from India in 1791. Woolly coats
need not have been a problem for the escaping sheep in the earliest phases
of the invasion. If it was some other ecological intolerance to Australian
climates which inhibited the escape of sheep, no such barrier need be
supposed in Europe where the species were dispersing near to the same
biogeographic province in which they originated.
Discussion about the suitability of the vegetational environment for the
introduced species depends on opinion about the extent of environmental
modification by non-agricultural people in Europe, on one hand, and in
Australia, on the other. Clearly this is one area of interpretation in which
understanding is not improved by use of an analogy between recent
Australia and prehistoric Europe. On the other hand, we might also argue
that the successful survival of escaped domesticates to become feral
populations was easier in Australia than it would have been for domestic
species dispersing within a single biogeographic province. The situation for
interactions between introduced and native fauna is much simpler.
Competition with the indigenous fauna would have been greater in Europe,
and there would have been more large predators than in Australia. Amongst
the predators in both cases were indigenous people.

Other considerations
Australian history also suggests some appropriation of animals from tended
herds. Bradley reported the first spearing of a goat on 21 August 1788
(King 1984, p. 89), only 30 weeks after landfall. In a remarkable story
(Campbell 1933), one of the massacres of Aborigines near Armidale, in
northern New South Wales, was justified by the belief that a group of
Aborigines had followed some pastoralists and their sheep down steep
wooded country and stolen their whole flock. The Aborigines, fisher-
gatherer-hunters with at most a couple of decades of knowledge of the
pastoral habits of the non-Aborigines, herded the sheep into the steep gorge
of Kunderang Brook. When the sheep and their Aboriginal rustlers were
found, massacre was deemed to be the appropriate punishment. Similar
stories abound (Reynolds 1982, pp. 156–69). Such theft from pastoralists is,
of course, known in other situations of contact between fisher-gatherer-
hunters and pastoralists, as Schrire (1980) has documented for the San of
southern Africa. Archaeologically, sites of people who stole livestock in
this way would look like the result of activities of people in category B(i)
(cultivating hunters), although in reality they would not have moved out of
our category of hunters.
Jones (1970) documented the rapidity with which Tasmanian Aborigines
adopted European dogs. Despite the fact that the Kunderang story
documents the ease with which Aborigines could learn the physical
manipulation of flocks of sheep, there is no evidence that Aborigines
showed any desire to adopt a pastoral way of life. Of course, the history of
the European invasion of Australia militated against it.
The early days of European colonization of Australia were times of great
hardship for the colonists. Their stock died, were stolen, or escaped. The
crops failed to produce sufficient food or withered in the unaccustomed
heat. The diet of the Europeans had to be supplemented by the hunting of
kangaroos and the catching of fish (see King 1984). In archaeology this
might be interpreted as people in category B(ii) (hunting agriculturalists).
In Australia the historical archaeology of such contact would have some
relatively simple problems of interpretation because of the differences
between the species indigenous to Australia and those which were
introduced. In interpreting the archaeology of prehistoric diffusion, as in
Europe, the problems would be far more difficult, because of the
similarities between species, and because of the possibilities of local
indigenous domestications.

The model
It is inconceivable, in a situation of diffusion of new plants and animals into
a strange environment, that the introduced plants and animals did not have
any opportunities to escape from the control of agriculturalists or
pastoralists. Without even appealing to the conditions in the initial non-
Aboriginal colonization of Australia, we may follow Lewthwaite (1984, p.
26) in pointing to the escape of pastoral animals from loose herding
arrangements. Thus, it is possible to point to an analogy for the escape of
animals either in the circumstances of the introduction of domestic stock
with groups of invading people, or at any stage in their introduction.
Once we postulate the existence of escaped domestic animals, we then
have to assume that these creatures became available as resources for the
indigenous populations of people. If we take an over-simplified model of
Mesolithic fisher-gatherer-hunters and Neolithic farmers, then we should
expect that some of the faunal assemblages of Mesolithic sites should
contain bones of sheep or goats which had escaped from the custody of
their Neolithic herders and had been hunted as game by the indigenous
population.
 Either the escape of the domestic animals to an unprotected state, or the
theft of beasts from tended herds, should lead to the presence of such
animals in Mesolithic archaeological assemblages. Twenty years ago
Helbaek (1966) wrote of ‘cultivated wild barley’ which has since been
shown likely to have been roofing material and unrelated to diet (Dennell
1972). Now we must speak of the likelihood of hunted domestic animals.
People who exploited such animals would be in category A (hunters) and
would be difficult to identify as such, because in identifying the escaped
domestic species we might be inclined to include them in category B(i)
(cultivating hunters). Without these fine divisions there would be a
temptation to identify such people as incipient pastoralists.
In this model there should exist sites of groups in category A (hunters),
which are prior to those of groups in category C (agriculturalists). Diffusion
of agriculture and pastoralism from outside will lead to the existence of
sites which are in reality those of groups of hunters, but because the hunted
prey are escaped domestic animals they appear to be cultivating hunters.
Such sites should be contemporary with those of agriculturalists. The
Australian model suggests that even the material culture will be an
unreliable guide to the true status of these sites because material culture can
also cross the frontier easily.
Australian history might suggest that early agriculturalists should also
hunt and fish to a significant extent, and hence be hunting agriculturalists.
This possibility will be clarified when we understand the process of
agricultural spread in such a way as to take into account the possibility of
hunted domestic animals.
Two consequences follow. The first is a consideration of the methods
which might permit us to recognize such a phenomenon. The second arises
because there do not seem to be sites in Spain which show the presence of
escaped domesticates which were hunted (or, in the case of plants,
gathered). The model would predict that there should be such a body of
evidence, and we need to consider the theoretical implications of such a
lack.

The frontier in Spain

There are two competing interpretations of the beginning of agriculture and
stock-raising in Spain. On the one hand there is the diffusionist model
which postulates that all such economic innovations came from the east,
and on the other is the model of indigenous development (see Geddes 1980,
pp. 56–80 for a summary of the arguments as they apply to southern France,
and Lewthwaite 1986 for a comprehensive synthesis of the cultural
evidence for the whole western Mediterranean). Even extreme proponents
of indigenous development could not account for all of the features which
appear for the first time at Coveta de l’Or (Marti et al. 1980).
Cattle, boars, wolves, and caprines occurred in the Spanish Pleistocene
fauna and, by some bending of the orthodoxy about domestication, could
have been the ancestors of domestic animals which appeared in the early
Neolithic sites. But there was no wild progenitor of sheep. We know next to
nothing about the Pleistocene plant exploitations (see summary in Davidson
1980), but nobody has claimed that there could have been ancestors of
wheat or barley in the Peninsula. There is agreement that the
Epipalaeolithic and the Neolithic stone industries showed some similarities,
but few are happy with the thought that pottery was developed indigenously
in the Peninsula. As things stand at the moment, we would probably need to
postulate some diffusion into the Peninsula of sheep, wheat and barley, and
probably goats and pottery (see Lewthwaite 1986).
It does not matter very much what the conditions of this diffusion were.
In particular, we might postulate whether groups of people carried them into
territory occupied by indigenous fisher-gatherer-hunters or exchanged them
with neighbouring groups whom they instructed in the domestic sciences.
In a small number of cases where determinations are possible, there
seems to be a general absence of remains of the introduced animals in the
sites with the earliest pottery in the Peninsula: Mallaetes (Davidson 1983);
Nino (Davidson 1980, Ch. 10, Fortea et al. 1983); Verdelpino (Morales
1977); Botiqueria dels Moros (Altuna 1978); Cocina (Fortea et al. n.d.);
Nacimiento (Alférez et al. 1981); and Valdecuevas (Sarrión 1980). One
exception to this is Nerja (Boessneck & von den Driesch 1980, p. 24) where
some specimens were found which were described as ovicaprines. It was
further stated that they were indistinguishable from ovicaprines which
would have been called domestic if they had been found in a Neolithic
context. It would seem, from their statement, that Boessneck & von den
Driesch (1980) are not willing to apply their morphological criteria without
taking into account the cultural context, thus begging the question of the
cultural context in which agriculture emerged. The problem here is similar
to the problem of the southern French sites such as Chateauneuf and
Gramari. Even if we accept both the stratigraphic context of the finds, and
the identifications, it is difficult to be clear about the nature of the economy
which was responsible for them, as will be argued below.
Two other Spanish sites are crucial here: Matutano and Fosca. Both of
these are in Castellón province and show evidence for changes in the fauna
through time. These are the sorts of changes which have been identified
elsewhere as those associated with the beginnings of domestication
(Estévez et al. 1983, Estévez pers. comm.). Estévez and his colleagues have
argued for the local operation of the process of domestication. The
argument depends on the application of the same criteria that have been
used in the eastern Mediterranean. These are: the presence of wild and
domesticated animals; presence of transitional animals; difference from the
composition of a wild population; representation of scenes of capture;
survival of aberrant individuals; and different butchery of wild species and
species in the process of domestication (Bökönyi 1977).
Let us accept for a moment that hunted domestic animals would have
existed. Would our methods be suitable for identifying them?
Morphologically and in terms of size the first such animals would have
been indistinguishable from their domestic progenitors (cf. Groves, ch.5,
this volume), although some changes might be expected after several
generations of a more natural selection. Slaughter patterns should follow
those of the hunting strategy applied to animals of similar size or behaviour.
If the introduction of agriculture involved the movement of species into a
similar biogeographical province, one might expect that the previously
existing faunal populations would have been well suited to the existing
conditions. This suitability might have prevented the rapid successful
expansion by the newly escaped domestic species. As a result, populations
of the escaped species might have been low and the killed animals so few as
to make it difficult to estimate the slaughter patterns from an archaeological
assemblage.
Simple interpretation of the fauna suggests the presence of both wild and
domestic species, but this tells us nothing about the relationships between
the people and the animals. In the case of sheep, there would not necessarily
be any transitional animals, although goats might interbreed with the local
Spanish ibex (Galindo 1965, pp. 35–6). Galindo suggests that the offspring
of such crosses seem to resemble their wild sires more than their domestic
dams. Representations of scenes of capture could not enable us to
distinguish between the hunting of domestic species and the normal
activities of pastoralists.
We are left with two only of the criteria used by Estévez and his
colleagues (Bökönyi 1977, Estévez et al. 1983) – the survival of aberrant
individuals and the patterns of butchery.
Aberrant individuals can be found in large samples of ungulates
slaughtered in the wild (Peter Jarman pers. comm. May 1986). They would,
however, normally be scarce, for that is why we recognize them as aberrant.
The assumption is that under normal conditions of natural selection they
would not survive, and that only human protection allows them to reach an
age to be slaughtered by their protectors. Jarman’s observation suggests that
the assumption is a weak one. In addition, they, and their absence, would
not necessarily be a satisfactory clue that the relationship was not one of
pastoralism. They might simply have escaped human predation, or not been
killed at a particular site.
There remains the possibility that hunters would butcher hunted domestic
animals in a manner distinct from that used by the contemporary
pastoralists, but similar to their practices with wild prey. It is certainly true
that we need to pay far more attention to the patterns of butchery, as only
Estévez (in Olaria et al. 1981) has so far published such details for
Mediterranean Spain. Nevertheless, we might appeal to Binford’s (1978)
study of the economic anatomy of reindeer to suggest that the utility of
particular parts of an animal in any given set of environmental conditions is
tied relatively strongly to the nature of the animals and less to the cultural
practices of the people.
It may be that the identification of hunted domestic animals is beyond
our current methods of analysis of the archaeological record.
Of course, Higgs & Jarman (1969) suggested long ago that application of
the conventional criteria for the identification of domestication could
produce some strange results if selectively applied to the interpretation of
partial prehistoric evidence. They, therefore, suggested that a more flexible
approach to ‘man-animal’ relationships (Shawcross 1975 referred to
zoocresis which would avoid the sexist terminology), needed to be adopted.
Consideration of the problems involved in the identification of escaped
domestic animals suggests that their caution still deserves respect.
Moreover, the principle that animals will escape from control in many
pastoral systems brings into question some of the logic by which scholars
have looked for the origins of agriculture in the regions of modern
distribution of the supposed wild ancestors of the domesticates (see, for
example, the papers by Herre & Róhrs, on animals, and Harlan, on plants,
in Reed 1977). It is at least plausible that many of the modern animals and
plants escaped from control at some early stage in the history of
domestication, to form early feral populations in the region of their modern
distribution. On this argument, the modern distributions do not represent the
prehistoric ranges of the wild species concerned, let alone the distribution of
the species which were the wild ancestors of perhaps both the domesticates
and the modern representatives of the non-domestic species.

The agricultural colonization of Spain

How might the expectation that animals escaped beyond the frontier of
agricultural advance affect our interpretation of Spanish prehistory?
In some of the eastern Spanish sites (Parpalló, Les Mallaetes, Nino) the
beginnings of an indigenous process of domestication or the beginnings of
pastoralism cannot readily be identified (Davidson 1980, 1983, Bailey &
Davidson 1983). This leads to a consideration of the conditions which
operate in a situation where agriculture and pastoralism are introduced from
outside.
There is very little evidence for introduced species of plants and animals
in the non-pottery post-Pleistocene sites of eastern Spain. This contrasts
with the situation in southern France where there are at least claims for
sheep, in small numbers, from Chateauneuf (Ducos 1976), Gramari
(Poulain 1971) and other Mesolithic sites (Rozoy 1978). There are
difficulties in chronology and identification: Ducos, for example,
acknowledges that his identifications were made before the publication of
the criteria for distinguishing sheep from goat (Boessneck et al. 1964).
Nevertheless, the finds are such as one would expect from a situation of
escaped animals being hunted by the indigenous non-pastoral human
population. Geddes (1985) has reviewed the old evidence and provided
detailed descriptions of specimens from more recent excavations at the late
Mesolithic sites of Gazel, and Dourgne. Commenting on these and other
remains of domestic animals in Mesolithic sites, he attributes them to
‘isolated survivals of Mesolithic groups which acquired domestic animals
by trade, theft or other social means.’ (Geddes 1985, p. 26). Geddes uses
the phrase ‘incipient animal herding’, and suggests ‘that indigenous
Mesolithic communities could have constituted a principal component in
the emergence of settled farming communities in Mediterranean Europe.’
(Geddes 1985, pp. 44–5). If this was the case, then why are there so few
cases in France, and none in Spain?
One possible explanation is that the early sites with pottery and high
proportions of hunted animals are not the sites of the first farmers, but of
fisher-gatherer-hunters who obtained pots across the frontier and hunted
escaped domestic animals. Some sites, such as Or (Marti et al. 1980)
contain such abundant quantities of pottery and of cereal grains that they
probably represent the initial occupation by people in category B(ii)
(hunting agriculturalists). Other sites, such as Nerja, Parralejo and Dehesilla
(see Murioz 1984) probably fall into category A (hunters).
The site of La Cocina deserves attention. For long regarded as a classic
Mesolithic site (Pericot 1945, Fortea 1971), Fortea’s recent excavations
allow the first interpretation of faunal exploitation by the inhabitants of this
cave (Fortea et al,. unpublished) through the identifications by Pérez Ripoll.
In the layers without pottery there were no remains of species which might
have been thought to be domestic, and analysis of the ages at death of the
Spanish ibex showed that the high proportions of adult and old individuals
contrasted with the high proportions of juveniles at Or (Pérez Ripoll, in
Marti et al. 1980). There was no change in this proportion in the upper
layers at La Cocina, which contained cardial pottery, although small
numbers of bones were identified as Ovis/Capra. This looks like a perfect
example of a site used by people in category A (hunters), who hunted
escaped domestic animals.
Conclusions

The purpose of this chapter has been to draw attention to the continuing
difficulty of recognizing the relationships which groups of prehistoric
people had with animals. If we confine ourselves to recognizing the status
of animals as wild or domestic, then the possibility that people hunted
escaped domestic animals would confuse most interpretations. Similar
problems might exist with plants, although the properties which allow
plants to escape from cultivation to reproduce without human intervention
might actually have direct morphological expression.
If we consider a model of diffusion of agriculture into a new environment
already populated by fisher-gatherer-hunters, then we would predict that
both material culture and animals would cross the frontier between the two
ways of life. Some sites are classified as being created by agriculturalists or
pastoralists because they contain both pottery and the new species which
might be identified as domesticated. In the absence of clear evidence in
Spain that the fisher-gatherer-hunters of the Epipalaeo-lithic or Mesolithic
did exploit the escaped domesticates of the early pastoralists, we should
consider carefully whether some of the sites which are generally regarded
as those of Neolithic agricultural and pastoral groups should be reclassified
as the indications of the hunters of escaped domesticates. Full testing of this
hypothesis will require more detailed re-analysis of individual sites.

Acknowledgements
This chapter has been improved enormously by the generous and thoughtful comments of J. Clegg,
G. E. Connah, J. Driver, R. Fletcher, A. Gilman, L. Godwin, M. Jackes, P. Jarman, D. Lubell, S.
Solomon, D. Witter, and J. P. White. I also acknowledge helpful discussions with I. Plug and E. Voigt
about the prehistoric frontier in southern Africa. G. E. Connah also suggested that the comparison
was worth making – but that would be another chapter. Only I am responsible for the final form of
this one.

References
Alexander, J. 1976. The frontier concept in prehistory. In Hunters, gatherers and first farmers beyond
Europe, J. V. S. Megaw (ed.), 25–40. Leicester: Leicester University Press.
Alexander, J. 1984.Early frontiers in southern Africa. In Frontiers: southern African archaeology
today, M. Hall, G. Avery, D. M. Avery, M. L. Wilson & A. J. B. Humphreys (eds), 12–23. Oxford:
 BAR International Series 207.
Alférez, F., G. Molero, V. Bustos & P. Brea 1981. Apendice II. La fauna de
macromamíferos.Trabajos dePrehistoria 38, 139–45.
Altuna, J. 1978. Fauna del yacimiento prehistórico de Botiqueria dels Moros, Mazaleon (Teruel).
Cuademos de Prehistoria y Arqueologia Castellonense 5, 139–42.
Bailey, G. N. & I. Davidson 1983. Site exploitation territories and topography: two case studies from
Palaeolithic Spain. journal of Archaeological Science 10 (2), 87–115.
Binford, L. R. 1978. Nunamiut Ethnoarchaeology. New York: Academic Press.
Boessneck, J. & A. von den Driesch 1980. Tierknochenfunde aus vier Südspanische Holen. Studien
über frühe Tierknochenfunde von der Iberischen Halbinsel 7, 1–83.
Boessneck, J., H. H. Müller & M. Teichert 1964. Osteologische Unterscheidungsmerkmale zwischen
Schaf (Ovis aries Linné) und Ziege (Capra hircus Linné). Kühn-Archiv 78, 1–129.
Bó kónyi, S. 1977.Animal remains from the Kermanshah Valley, Iran. Oxford: BAR Supplementary
Series 34.
Campbell, J. S. 1933. The Kunderang ravines of New England. Royal Australian Historical Society
journal and Proceedings 18, 63–73.
Clegg, J. 1984. Pictures ‘of bulls and boats. In Under the shade of a coolibah tree: Australian studies
in consciousness, R. A. Hutch & P. G. Fenner (eds), 219–38. Lanham, Maryland: University Press
of America.
Davidson, I. 1980.Late Pleistocene economy in eastern Spain. Unpublished PhD thesis, Department
of Archaeology, University of Cambridge.
Davidson, I. 1983. Site variability and prehistoric economy in Levante. In Hunter-gatherer economy
in prehistory, G. N. Bailey (ed.), 79–95. Cambridge: Cambridge University Press.
Dennell, R. W. 1972. The interpretation of plant remains. In Papers in economic prehistory, E. S.
Higgs (ed.), 149–59. Cambridge: Cambridge University Press.
Dennell, R. W. 1985. The hunter-gatherer/agricultural frontier in prehistoric temperate Europe. In
The archaeology of frontiers and boundaries, S. W. Green & S. M. Perlman (eds), 113–39. New
York: Academic Press.
Ducos, P. 1976. Quelques documents sur les débuts de la domestication en France. In La préhistoire
française, Vol. II, J. Guilaine (ed.), 165–7. Paris: CNRS.
Estévez, J., F. Gusi, C. Olaria, A. Vila & R. Yll 1983. Evolución ambiental y desarrollo de la base
subsistencial hasta el 7000 bp en el Levante Ibérico. In RésumédesCommunications.Premières
Communautés Paysannes en Méditerranée Occidental, p. 27. Montpellier: UISPP.
Fortea, J. 1971.La Cueva de la Cocina. Valencia: Trabajos Varios del Servicio de Investigación
Prehistórica, Diputación Provincial.
Fortea, J., B. Marti, M. P. Fumanal, M. Dupré & M. Pérez Ripoll n.d.Epipaleolitico y neolitizacion en
la zona oriental de la Peninsula Ibérica. (Unpublished.)
Fortea, J., J. M. Fullola, V. Villaverde, I. Davidson, M. Dupré & M. P. Fumanal 1983. Schema
paléoclimatique, faunique et chronostratigraphique des industries â bord abattu de la région
méditerranéenne espagnole. Rivista di Scienze Prehistoriche 38 (1–2), 21–67.
Galindo, F. 1965. La Capra Pyrenaica Hispanica de los Puertos de Beceite (Teruel). Teruel 33, 5–76.
Garran, J. C. & L. White 1985. Merinos, Myths and Macarthurs. Australian graziers and their sheep,
1788–1900. Canberra: Australian National University Press.
Geddes, D. S. 1980 (1983). Patterns of animal exploitation in the late Mesolithic and early Neolithic
in the Aude Valley (southern France). Ann Arbor: University Microfilms International.
Geddes, D. S. 1985. Mesolithic domestic sheep in west Mediterranean Europe. journal of
Archaeological Science 12 (1), 25–48.
Green, S. W. & S. M. Perlman (eds) 1985.The archaeology of frontiers and boundaries. New York:
Academic Press.
Groves, C. P. 1989. Feral mammals of the Mediterranean islands: documents of early domestication.
In The walking larder, J. Clutton-Brock (ed.), ch. 5, London: Unwin Hyman.
Hall, M. J., D. M. Avery, M. L. Wilson & A. J. B. Humphreys (eds) 1984. Frontiers: southern
African archaeology today. Oxford: BAR International Series.
Harlan, J. R. 1977. The origins of cereal agriculture in the Old World. In Origins of agriculture, C. A.
Reed (ed.), 357–83. The Hague: Mouton.
Harris, D. R. 1979. Foragers and farmers in the western Torres Strait islands: an historical analysis of
economic, demographic, and spatial differentiation. In Social and ecological systems, P. C.
Burnham & R. F. Ellen (eds), 75–109. London: Academic Press.
Helbaek, H. 1966. Pre-pottery neolithic farming at Beidha. Palestine Exploration Quarterly 98, 61.
Herre, W. & M. Róhrs 1977. Zoological considerations on the origins of farming and domestication.
In Origins of agriculture, C. A. Reed (ed.), 245–79. The Hague: Mouton.
Higgs, E. S. & M. R. Jarman 1969. The origins of agriculture: a reconsideration. Antiquity 43, 31–43.
Ingold, T. 1984. Time, social relationships and the exploitation of animals: anthropological
reflections on prehistory. In Animals and Archaeology. Vol. 3: Early herders and their flocks, J.
Clutton-Brock & C. Grigson (eds), 3–12. Oxford: BAR International Series 202.
Jones, R. 1970. Tasmanian Aborigines and dogs. Mankind 7 (4), 256–71.
King, J. 1984. The first settlement. The convict village that founded Australia, 1788–1790. South
Melbourne: Macmillan.
Lewthwaite, J. 1984. The art of corse herding: archaeological insights from recent pastoral practices
on west Mediterranean islands. In Animals and archaeology. Vol. 3:Early herders and their flocks,
J. Clutton-Brock & C. Grigson (eds), 25–37. Oxford: BAR International Series 202.
Lewthwaite, J. 1986. From Menton to the Mondego in three steps: application of the availability
model to the transition to food production in Occitania, Mediterranean Spain and southern
Portugal. Arqueologia (Porto) 13, 95–119.
Lyon, K. & J. Urry 1979. Bull shelter: a cow pastures’ conundrum. A.I.A.S Newsletter 11, 39–45.
MacKnight, C. C. 1976. The voyage to Marege’. Macassan trepangers in northern Australia.
Melbourne: Melbourne University Press.
Marti, B., V. Pascual, M. D. Gallart, P. Lopez Garcia, M. Pérez Ripoll, J. D. Acuna & F. Robles 1980.
Covade l’Or (Beniarrés,Alicante). Vol. II. Valencia: Trabajos Varios del Servicio de Investigación
Prehistórica, Diputación Provincial.
Morales, A. 1977. Apendice I. Anâlisis faunistico de Verdelpino (Cuenca). Trabajos de Prehistoria
34, 69–81.
Mulvaney, D. J. 1976. ‘The chain of connection’: the material evidence. In Tribes and boundaries in
Australia, N. Peterson (ed.), 72–94. Canberra: Australian Institute of Aboriginal Studies.
Munoz Amabilia, A. Ma. 1984. La neolitización en Espana: problemas y lineas de investigación. In
Scripta Praehistorica. Francisco Jordd Oblata, J. Fortea (ed.), 349–69. Salamanca: Ediciones
Universidad de Salamanca.
Olaria, C., F. Gusi, J. Estévez, J. Casabo & M. L. Rovira 1981. El yacimiento magdaleniense superior
de Cova Matutano (Villafamés, Castellón). Estudio del sondeo estratigrâfico
1979.CuadernosdePrehistoria y Arqueologia Castellonenses 8, 21–100.
Pericot, L. 1945. La Cueva de la Cocina (Dos Aguas). Archivo de PrehistoriaLevantina 2, 39–71.
Poulain, T. 1971. Le camp mésolithique de Gramari â Méthamis (Vaucluse). III. Étude de la faune.
Gallia Préhistoire 14, 121–31.
Reed, C. A. 1977. A model for the origin of agriculture in the Near East. In Origins of agriculture, C.
A. Reed (ed.), 543–67. The Hague; Mouton.
Reynolds, H. 1982. The other side of the frontier. Ring wood, Victoria: Penguin.
Rolls, E. C. 1969. They all ran wild: the story of pests on the land in Australia. Sydney: Angus and
Robertson.
Rolls, E. C. 1984.A million wild acres. Ringwood, Victoria: Penguin.
Rozoy, J. G. 1978.Les Derniers Chasseurs. Bulletin de la Société Archéologique Champenoise,
Charleville.
Sarrión, I. 1980. Valdecuevas. Estación Meso-neolftica en la Sierra de Cazorla (Jaén).Saguntum 15,
23–56.
Schrire, C. 1980. An inquiry into the evolutionary status and apparent identity of San hunter-
gatherers. Human Ecology 8, 9–32.
Shawcross, F. W. 1975. Some studies of the influences of prehistoric human predation on marine
animal population dynamics. In Maritime adaptations of the Pacific, R. W. Casteel & G. I.
Quimby (eds), 39–66. The Hague: Mouton.
Woodburn, J. 1986.African hunter-gatherer social organisation. Is it best understood as a product of
encapsulation? Unpublished paper given at the Fourth International Conference on Hunting and
Gathering Societies, London, September.
7 Evidences for the impact of traditional
Spanish animal uses in parts of the New
World
ELIZABETH S. WING

Columbus’ celebrated voyage and subsequent ones brought great changes to

the aboriginal way of life and the faunas of the western hemisphere. Even
the very first exploratory ventures were accompanied by European domestic
animals and the commensal rat, as well as by a pattern of subsistence
developed over centuries in the Iberian peninsula. Some of the Spanish
traditions of animal use and exploitation had an impact on the aboriginal
populations, modifying them in a more profound way than any influence
during the previous centuries of stability. Spanish traditions of animal use
were likewise modified. Studies of the animal remains from three pairs of
prehistoric and historic sites in Florida and Haiti are used here to illustrate
the magnitude of the changes and the types of changes which took place.
The sites are listed in Table 7.1. These samples show which European
domesticates first took hold in the New World environment and the cultural
and biological factors determining their success are suggested.

Table 7.1 Names and locations of the prehistoric and historic sites
Sites reviewed Location Dates of occupation Occupants
Hontoon Island central Florida AD 928–1470 aboriginal
8V0202 AD 1540–1758 aboriginal with indirect
influence of Spanish
Fountain of Youth St. Augustine Florida late prehistoric-early aboriginal
8SJ31 historic
Various St. Augustine Florida AD 1565–1700 Spanish partially
provided with food by
aboriginals
En Bas Saline north coast Haiti contact aboriginal and possibly
some sailors from
Columbus’ ship
Puerto Real north coast Haiti AD 1502–1578 Spanish

The two faunal assemblages from Hontoon Island differ in many respects
(Table 7.2). Most of the same species occur in both components of the site
but their relative abundance changes. Aquatic vertebrates predominate
during both time periods. Those terrestrial species which were procured
during historic times were on the average larger fragments and fragments of
larger animals than the prehistoric catches. The most obvious difference
between the two components is the decrease in the gastropod molluscs in the
historic assemblage. During historic times the long-established practice of
gathering snails was discontinued, but mussels were still collected. These
changes in procurement of animals are accompanied by substantial changes
in uses of plants, which suggest clearing and increased agricultural activities
(Newsom 1986). These changes are correlated with the advent of Spanish
influence in Florida. Further research on such sites is needed to determine
how Spanish influence caused these changes.
Further insight into the effect of Spanish traditions may be seen in a
comparison of the faunal assemblages from the prehistoric aboriginal site
and a number of 16th-century Spanish sites in St. Augustine (Reitz 1985).
As at Hontoon, the remains identified from the aboriginal site are
predominantly aquatic, though the species encountered are marine and
estuarine rather than the freshwater species. Differences in the fishes used by
the Indians and the colonists are noted by Reitz (1985). The 16th-century
Spanish faunal samples show a greater dependence on terrestrial fauna, as
measured by the numbers of remains identified as terrestrial animals in the
faunas (Table 7.3). European domestic animals, however, account for only
3–10 per cent of the vertebrates. The Spanish remains include molluscs,
primarily oyster, but these may have been remains of building materials
rather than meals.

Table 7.2 Hontoon Island faunal distribution

Table 7.3 Habitat preferences of species identified from historic and prehistoric sites in St. Augustine
(Reitz 1985, Reitz & Scarry 1985)

The faunal sample from the predominantly aboriginal site of En Bas

Saline in Haiti indicates a focus in the economy on fishing and shellfish-
gathering (Table 7.4). This sample includes only 5 per cent terrestrial
vertebrates, 0.09 per cent of these being animals introduced from Spain. In
contrast, the samples from the colonial site of Puerto Real show far greater
reliance on terrestrial vertebrates which are predominantly introduced
domestic animals (McEwan 1983, Reitz 1986). The aquatic component from
locus 33/35 at Puerto Real is primarily composed of pond turtle (Pseudemys
spp.), which were apparently sought in preference to fishes by the colonists
(McEwan 1983). Shellfish remains are scarce at Puerto Real. The faunal
remains from locus 39 at Puerto Real are primarily cattle (see Table 7.6).
This area of the site is thought to have been a processing area for cattle,
established within ten years of Columbus’ first voyage to the New World
(Reitz 1986).
These three pairs of samples show the same types of differences between
the prehistoric and historic sites or components. The historic samples have
relatively more, or larger, terrestrial species and less shellfish. European
domesticates are represented even in the early historic sites but are not as
relatively abundant as wild species, except at locus 39 at Puerto Real. In
order to understand which domestic species are included in the first colonial
outposts, human cultural characteristics such as refuse disposal practices,
social status, and animal husbandry traditions, as well as the biological
characteristics of the introduced animals, must be understood (Tables 7.5 &
7.6).

Table 7.4 Habitat preferences of species identified from historic and prehistoric sites in Haiti
(McEwan 1983, Reitz 1986)
Table 7.5 Distribution of domestic species in faunal samples f r om 16th-century sites in Spanish St.
Augustine (Reitz & Scarry 1985)

Table 7.6 Distribution of domestic species in faunal samples from 16th-century sites in Spanish Haiti
(McEwan 1983, Reitz 1986)
 Those animals for which there was a tradition of husbandry, which could
be transported easily, have a high reproductive potential, and were
introduced into suitable habitats in the new environment would be expected
to be most successful. On this basis one would expect to find the small
barnyard animals, such as pigs, chickens, dogs, and cats. Of these likely
animals, pigs and chickens are most abundantly represented in the sites
reviewed. The hardy introduced range hogs were so successful in Hispaniola
that licences to hunt the pigs which had become feral were issued as early as
1508 (Sauer 1966, p. 157). Dogs are known to have accompanied the
Spanish. Their scarcity in the sites may reflect a different disposal pattern
reserved for these animals. Cats and rats are known to flourish to the
detriment of native species in the Caribbean. Their scarcity in the sites may
indicate that their remains were disposed of in a different way to the remains
of food refuse. Thus, of the six species which would be expected to flourish
in early colonization, evidence shows that pigs and chickens prospered
above all others. Dogs and cats may have been equally numerous but
possible differences in disposal patterns may result in an underestimation of
their numbers.
The small herd animals, sheep and goats, were important in traditional
Spanish husbandry and would have been easily transported. However, their
remains are rare in the sites reviewed. Goats are reported to have adapted
well in parts of the Caribbean. They became feral in the mountains of
Jamaica at an early date (Sauer 1966, p. 181). They did not adapt as well as
other stock to the humid lowland environments. Sheep do not adapt well to
new environments, or adjust to new diets as well as other livestock
(Williamson & Payne 1965, p. 269). These easily transportable and
traditional animals did not do well in the initial colonization, perhaps for
lack of suitable local habitats around the sites reviewed.
Two highly prized animals, the horse for transportation and cattle for
products such as meat, hides and tallow, present problems for introduction.
They are both large and therefore difficult to transport and they reproduce
slowly. Despite the difficulties of transport and relatively slow herd growth,
cattle prospered remarkably well, particularly in Hispaniola, as evidenced by
documentary reports and the large sample of large-sized cattle remains in
locus 39 of Puerto Real (Reitz 1986). The early history of cattle in Florida is
different. Although cattle were present in the 16th century, ranches were not
established in what is now northern Florida until the second half of the 17th
century (Arnade 1965). This may have been more because of Indian
predation on the herds and their keepers than the absence of suitable
habitats. Horses were particularly difficult to transport. Many succumbed in
transit, particularly in the Horse Latitudes – named for their toll of horses
(Crosby 1972, p. 80). Once safely landed in the New World, they are
reported to have prospered along with cattle in the grazing lands, free of
competitors and predators. Horse remains, however, are seldom encountered
in the archaeological sites. This may be a function of patterns of disposal, in
which remains of beasts of burden which were not usually consumed would
not be incorporated in food or butchering refuse remains.
The impact of Spanish culture and animal introductions into the southeast
and Caribbean was profound. The rapid adaptation of the introduced animals
modified the New World environments. The introduction of Spanish
traditions of animal procurement and uses, and the introduction of domestic
animals changed the western hemisphere almost immediately and
permanently.

Acknowledgements
In presenting these examples I lean heavily on the work of colleagues, friends, and students. I am
particularly grateful for the scholarly production of Elizabeth J. Reitz who did the original research on
the faunas from St. Augustine and locus 39 at Puerto Real (Reitz 1986). I am also indebted to Bonnie
McEwan for her analysis of the fauna from locus 33/35 at Puerto Real (McEwan 1983). Thanks are
due to Karla Bosworth, Laurie McKean, Erika Simons, and several other diligent students for their
work on the identification of the faunas from Hontoon Island and En Bas Saline. I am grateful to Drs
Barbara Purdy and Kathleen Deagan, the archaeologists who entrusted the faunal samples they so
carefully recovered to us for our studies.

References
Arnade, C. W. 1965. Cattle raising in Spanish Florida, 1513–1763. Saint Augustine Historical Society
Publication Number 21, 3–11.
Crosby, A. W. Jr. 1972. The Columbian exchange: biological and cultural consequences of 1492.
Westport: Greenwood Publishing Company.
McEwan, B. G. 1983. Spanish colonial adaptation on Hispaniola: the archaeology of Area 35, Puerto
Real, Haiti. Unpublished thesis, University of Florida, Gainesville, Florida.
Newsom, L. A. 1986. Plant, human subsistence, and environment: a case study from Hontoon Island
(8–VO–202),Florida. Unpublished thesis, University of Florida, Gainesville, Florida.
Reitz, E. J. 1985. Comparison of Spanish and aboriginal subsistence on the Atlantic coastal plain.
Southeastern Archaeology 4(1), 41–50.
Reitz, E. J. 1986. Cattle at Area 19, Puerto Real, Haiti. Journal of Field Archaeology 13, 317–28.
Reitz, E. J. & C. M. Scarry 1985. Reconstructing historic subsistence with an example from sixteenth-
century Spanish Florida. Society for Historic Archaeology Special Publication 3, 1–150.
Sauer, C. O. 1966. The early Spanish Main. Berkeley: University of California Press.
Williamson, G. & W. J. A. Payne 1965. An introduction to animal husbandry in the tropics. London:
Longman.
8 Osteological evidence for the process of
animal domestication
RICHARD H. MEADOW

Introduction

The domestication by humans of a small number of non-human animal

species has taken place in various parts of the world at different times in the
past (for summaries, see Zeuner 1963, Clutton-Brock 1981, Mason 1984).
Records of these successes, as well as of unsuccessful attempts, have been
recovered from the writings, iconographic representations, and material
remains of ancient societies. In this chapter, focus is placed on one kind of
evidence for animal domestication, namely that provided by animal bones
and teeth recovered from archaeological sites. Most of what is discussed
should be seen as referring principally to those social ungulate species that
have become economically important to humans, namely cattle, sheep, goat,
pig, horse, donkey, camel, water buffalo, and South American camelids.

Domestication

To most of us, animal domestication implies the development of special

kinds of human-animal relationships that, intuitively, seem different from
those between hunter and prey. Yet, as some researchers have emphasized,
there is a continuum of conceivable relationships stretching from random
hunting through intentional game-cropping, herd-following, animal-
penning, and pet-keeping to the breeding of genetically isolated ‘domestic’
stock (e.g. Higgs & Jarman 1969, Jarman & Wilkinson 1972, Hecker 1982).
Postulating slow, almost imperceptible changes based on rational,
ecologically sound choices, these investigators contend that no
discontinuity in human-animal relationships is to be expected between pre-
pastoral and pastoral societies. Domestication of hunted species is thus seen
as an elaboration of the predator-prey relationship, the ultimate result being
that herding instead of hunting is used to guarantee a dependable and
renewable source of animal protein with minimum risk and expenditure of
energy.
While such views are compelling, Ingold (1980) has convincingly argued
that they are deficient because they fail to find a place for the important
social factors that help to shape human behaviour. As earlier investigators
(notably V. Gordon Childe) have also maintained, times that see the
beginnings of pastoralism are marked by a transformation in the
infrastructure of society, with the source for this ‘revolution’ to be found in
the area of social relations particularly as they relate to production.
Therefore, it is to manifestations of changing socio-economic relations
within human societies, as well as to changing relations between humans
and various animal populations, that investigators must turn when seeking
to study the domestication of economically important species (Compagnoni
& Tosi 1978, Ducos 1978).
Following from these arguments, animal domestication can be defined as
being a selective diachronic process of change in human-animal
relationships involving, at the very least, a change of focus on the part of
humans from the dead to the living animal and, more particularly, from the
dead animal to the principal product of the living animal – its progeny. This
process, while varying from culture to culture and from species to species,
is manifest in two ways: first, in structural transformations in socio-
economic dimensions of the human societies that interact with the living
animals and, secondly, in changes in the behaviour and eventually in the
morphology and physiology of the animals being domesticated (Clutton-
Brock 1981, pp. 9–25, Meadow 1984, p. 310).
Although the remainder of this chapter is concerned with how animal
(principally ungulate) domestication is reflected in faunal remains from
archaeological sites, it is important to stress that concomitant changes in
society are also manifest in other portions of the prehistoric record. For
example, all three aspects of the ‘Neolithic revolution’ in the Middle East –
the development of settled village life, plant cultivation, and animal
husbandry – reflect a growing concern for property and for its maintenance
through time. The development of this attitude was a prerequisite for the
kind of social stratification and differentiation based on the accumulation of
material wealth that formed the foundation for ‘civilization’ in the region.
One must note, too, that the domestication of bovids and pigs in the Middle
East took place in an already established context of cereal crop agriculture
and permanent villages (see Gebel [1984] for a convenient summary of the
data). One can thus argue that animal-keeping was almost forced upon those
human groups as a means to permit them to maintain a supply of animal
protein in the face of an overhunting of restricted territories, which was
brought on by an absence of periodic movement and by the need to control
crop-robbing herbivores (Uerpmann 1979). Thus, while herding can indeed
be seen as a kind of predator-prey relationship, it reflects not the end of a
continuum but a complete change in human attitudes towards animals.
When evaluating evidence of animal domestication, it is important to
bear in mind that there is a conceptual difference between identifying the
presence of domestic animals at a site and identifying the process of
domestication. The latter implies documenting changes over time (however
short) while the former involves identifying the end results of a process, and
includes an implicit contrast between ‘wild’ and ‘domestic’. Bökönyi, in his
often-cited 1969 essay, recognized this difference and noted that it is first
necessary to determine whether animals were being kept by peoples at a site
and then to determine whether they were domesticated locally. His
statements provide a useful summary of the kinds of evidence commonly
used to support claims for animal domestication:

…there is certain evidence for animal keeping in a prehistoric

settlement if:
(1) the proportion of age groups of a domesticable species is not
the same as found normally in the wild population;
(2) the proportions of the sexes of a domesticable species is not the
same as found normally in the wild population;
(3) domesticated species appear which have no wild ancestors in
that particular region, at least since the Pleistocene;
(4) morphological changes appear in domesticated animals;
(5) there are artistic representations of domesticated animals;
(6) there are objects associated with animal husbandry.
Note that, in all cases, Bökönyi makes at least an implicit contrast between
a normative situation characteristic of wild animals and one characteristic
of kept or domesticated forms. In other words, he is dealing with
dichotomies and is not trying to identify a process. Unfortunately, this sense
of dichotomy remains even in his answers to his own processual questions:

…what is the evidence of local domestication in the material of a

site?
(1) Remains of both wild and domesticated forms on the site.
(2) The existence of transitional forms between wild ancestor and
the domesticated animal.
(3) Changes in the proportions of age and sex groups in the wild
form.
(4) Representations of scenes of capture.

Of principal concern to us in this essay are points 1–4 of evidence for

animal-keeping (Bökönyi 1969, pp. 220–1) and points 1–3 of evidence for
local domestication (Bökönyi 1969, p. 223). It is important to point out,
however, that the significance and meaning of artistic representations and
objects associated with domestic animals, like the faunal evidence itself, are
often open to multiple interpretations particularly when coming from
periods early in the domestication process. Thus, such evidence is best used
as only one line of support for an argument of animal domestication.

Demographic evidence for animal domestication

Age and sex ratios that change or are different from those ‘found normally
in the wild population’ of a taxon are one kind of evidence that has been
widely employed as evidence for domestication or for the keeping of
domestic ungulates (e.g. Dyson 1953, Ducos 1969, 1978, Wright & Miller
1976; for a good summary discussion, see Hesse 1978, pp. 236–56). One of
the most significant applications of the approach was its use to identify the
presence of ‘presumably domestic sheep’ in 9th millennium levels of Zawi
Chemi Shanidar (eastern Iraq) and to help support a claim for the local
domestication of these animals, this on the basis of a high frequency of
bones from sheep in general and from immature sheep in particular (Perkins
1964, revived by Reed 1983).
Considerable doubt has been cast on the validity of the demographic
approach, however (Jarman & Wilkinson 1972, Simmons & Ilany 1975–77,
Collier & White 1976). In the case of those ungulate species for which such
data are even available, there are no such things as age and sex profiles
‘normally found’ in a wild population. Instead, the proportion of young and
juveniles varies widely over time, as does the demography of individual
social units which can even vary seasonally. These factors make average
statistics of questionable value for understanding what sexes and ages might
have been available to a group hunting in a particular area at a particular
time. Furthermore, simulation studies carried out by Uerpmann (1979)
show that, given a closed population of animals (gazelles, for example),
increasing hunting pressure applied non-selectively over time will tend to
create a younger overall population from which the hunters will draw their
prey.
Before using arguments for domestication that rely heavily on age and
sex ratios, therefore, it is necessary to have good modern data on the
demography and behaviour of wild relatives of the taxon or taxa in
question, and to have carefully examined the faunal materials for evidence
of possible seasonal exploitation. Even with the best of information,
however, it is almost impossible to use age and sex ratios to rule out
selective hunting as an explanation of particular faunal configurations.
Furthermore, given the same information, it is equally difficult to prove the
existence of intentional selective hunting (game management), a
phenomenon that some investigators feel must have preceded the
domestication of ungulate species.
If there is no such thing as a ‘typical’ hunting pattern, the same is also
true for herding patterns. A herder will tend to follow different culling
practices for each of the various animal products desired (e.g. meat, milk,
hair, traction, security; see Payne 1973, Redding 1984). Although these
practices may come to be reflected in faunal assemblages, identification of
one or more can be difficult for taphonomic reasons and because of the
presence of mixed economic goals. In addition, social reasons for killing
animals may override economic reasons not to kill them, and disease or
injury can leave the herder with no choice at all.
 Archaeologists deal with probabilities more than with clear-cut proof,
and it is thus necessary to draw upon multiple lines of evidence when
presenting an argument. Age and sex ratios are one kind of evidence that
can provide potentially important information on human-animal
relationships but one that generally should not be used as the sole support
for an hypothesis of animal domestication, or for demonstrating the
presence of domestic animals at a given site. As Dyson (1953, p. 662) has
noted, however, ‘Such objections may be minimized to a large extent by the
repetition of the statistical pattern in a number of sites from a wide area,
provided the sample in each case is of adequate size’.
In sum, documentation of demographic patterns over space, or trends
over time, seems a more productive way to proceed than trying to identify
particular configurations by comparing them with illusionary ‘normal’
patterns. The same is also true for the species composition of a faunal
assemblage, that being another line of evidence that has been used to
identify the presence of domestic animals at a site. A good example of this
latter approach is provided for the southern Levant by Bar-Yosef (1981, Fig.
11, based in large part on the dissertation research of Simon Davis). He
notes that the ‘frequencies of hunted ungulates in each [of 26
Epipalaeolithic and Pre-pottery Neolithic occupations] reflect the available
game in its vicinity…’ and that ‘[d]omestication is reflected in the shift to
caprovines and the almost total abandonment of traditional game. This
occurred with the transition from Pre-pottery Neolithic A to Pre-pottery
Neolithic B in Israel ... It seems therefore that during the Epipalaeolithic the
hunting of available game continued uninterruptedly and no special
man/animal relationship or “hunter’s choice” can be discerned’ (Bar-Yosef
1981, pp. 403–6).

Zoogeographic evidence for domestic animals

One of the potentially most reliable indicators of the presence of a domestic

animal is finding its remains at a site in a region that is beyond the natural
range of its wild relatives. Use of this criterion, however, requires
knowledge of the distribution of past wild populations. More generally, it is
necessary to know what the native wild species of an area were in order to
demonstrate either the introduction of non-native taxa or the possibility of
local domestication. One example is Corsica, where the question of whether
the ‘wild’ sheep and pigs on the island are truly wild or merely feral has
been debated for years (Groves, ch. 5, this volume). The latest opinions are
that the free-ranging animals are feral (Poplin 1979, Franceschi 1980,
Geddes 1985), having been introduced as domesticates on to the island as
early as the 7th millennium BC (Vigne 1984).
Another example is the question of the true horse in Anatolia. Do the 4th
and 3rd millennium remains reported from northwestern and eastern
Anatolia (Boessneck & von den Driesch 1976, Rauh 1981, Bökönyi in
press) represent relict wild populations that were hunted, or imported
domestic stock? The answers to this question, and to others concerning the
late Pleistocene and early Holocene distribution of various mammalian
species, are far from clear (for the Middle East see Uerpmann 1987), with
the result that, in some cases, it is impossible to use the apparent absence of
wild relatives as clear evidence for the domestic status of a given taxon.
Should there be morphological evidence for the presence of bones of wild
stock in faunal collections, however, arguments for local domestication can
be made, but such claims have to be evaluated against the possibility that
local wild stock was being hunted at the same time that domestic animals
imported from elsewhere were being kept.

Morphological evidence for animal domestication

The principal difficulty with the use of morphological evidence has been
not so much in using it to define the presence of fully domestic stock, but
with the conception that it is not possible to use such evidence to identify
early stages in the process of animal domestication because changes would
not have had time to take place. The key to this problem might be to
distinguish between genetically based changes and those resulting from
immediate environmental conditions, including primitive husbandry
practices. A well-known example of the phenotypic manifestation of
genotypic change is in the horncores of sheep and goats, which are very
different in wild and domestic stock (Zeuner 1955, Reed 1960, Hole et al.
1969, Bökönyi 1975, Stampfli 1983). Even here, however, we have very
little understanding of the range of variation occurring in wild populations,
and of the number of generations of relaxed natural selection pressure or of
intentional selection that would have been required to produce observable
results. In addition, there is the problem of possible gene flow from
domestic stock to wild populations. For example, to what degree is the
occasional occurrence of hornlessness in modern wild female sheep due to
such gene flow in the past?
With respect to the time factor in morphological change, Bökönyi (1976)
has made the following observations:
…well defined morphological changes do not occur before about 30
generations, according to modern experiments on domestication. The
length of a generation is 2–3 yr in small species (dog, sheep, goat, pig)
and 5–6 yr in large species (cattle, horse, and so on). Since the earliest
domestication goes back as far as 8,000–14,000 yr, the delay of 60–90 or
150–180 yr in the occurrence of morphological changes is hardly
significant.

While it may be true that, when taken together, the faunal remains from a
multi-component site are unlikely to cover less than 200 years, individual
strata can contain materials deposited in very short periods of time (Wright
et al. 1980). To the degree that it is possible to deal with such deposits
individually, the absence of evidence for morphological change could be an
impediment to identifying the domestication process when it began
occurring.
There are, however, at least two kinds of morphological change that can
begin as early as the first or second generation. The first is the appearance
of skeletal manifestations of pathological conditions brought on by keeping
animals confined. Examples are Tepe Sarab (western Iran, early ceramic
Neolithic) where ‘a lot of chronic arthritis cases’ and some periodontitis are
reported for goats (Bökönyi 1977, p. 38) and, more significantly, Ain
Ghazal (Jordan, PPNB) where a high proportion of morphologically ‘wild’
goat remains show pathologies (Köhler-Rollefson 1986 and pers. comm.).
The second is size diminution, a phenomenon that is known to have
accompanied the domestication of bovids and pigs in the Middle East and
Europe (Bökönyi 1974, Boessneck & von den Driesch 1978, Uerpmann
1979, Stampfli 1983).
 Although biological mechanisms for size change have rarely been
discussed in the zooarchaeological literature, it appears that initial size
decrease need not have been the result of genetic change. Instead, it may be
related to a lower overall level of nutrition and lack of necessary diversity
in the foods made available to early domestic stock whose mobility was
restricted to increasingly overgrazed areas and whose supplementary diet, if
any, was of low-quality forage. Elsie Widdowson has suggested that while
genetics may be of primary importance in determining growth rates in the
early stages of development of a foetus, nutrition had a major influence in
later stages. She notes (1980, p. 7)
…because the growth rate in the last part of gestation has been slow, the
appetite after birth will be ‘set’ at a level appropriate to the size and rate
of growth at the time before birth when the appetite centres in the
hypothalamus were developing. The infant or animal that is small at birth
takes less food than its larger counterpart and shows no sign of the
‘catch-up growth’ which is so characteristic of rehabilitation after
undernutrition at older ages.

This condition resulting from malnutrition of the mother may have been
compounded by the presence of parasitic infestations, or by drawing milk
from the mother for human consumption. Indeed, it is conceivable that
these latter two factors affecting the food intake and metabolism of the
young animal directly could have led to size reduction without the presence
of maternal malnutrition (proposals made by Noddle 1986, Köhler-
Rollefson pers. comm.).
Accompanying nutrition-related changes would have been natural and
even human selection for smaller females bearing smaller young that, just
because they were smaller, would have a better chance of surviving lean
periods in marginal environments (Jarman & Wilkinson 1972, Boessneck &
von den Driesch 1978). At the same time, forces of natural selection that
favoured large males would have been removed and this, combined with
positive selection for smaller body size in the context of restricted
populations, would have led to continuous size diminution until a lower
plateau was reached. Such phenomena would be reflected in the faunal
record by an increase in the variability of bone dimensions followed by an
overall decrease in size, as shown by increasingly smaller extreme and
median measurement values (Uerpmann 1979, Meadow 1984). Only with
selective breeding for size, combined with good nutrition, would the trend
be reversed (Boessneck & von den Driesch 1978).
The above discussion of size diminution is largely an ex post facto
explanation of phenomena observed to have taken place in food species that
were later known to be domestic and that were kept in relatively large
numbers, probably to ensure a dependable and renewable source of animal
products for settled agriculturalists in eastern Europe and the Middle East.
Size change in mammals, however, need not be associated with
domestication. Thus, Davis (1981) has discussed temperature-related size
diminution in fox, wolf, boar, aurochs, wild goat, and gazelle at the end of
the Pleistocene in Israel. Uerpmann (1978) has shown that size decreased
up to the present in Middle Eastern wild sheep, while Pietschmann (1977)
has documented the same pattern for red deer in Europe. Finally, Jordan
(1975) discusses a case of decrease over time in size of red deer on the basis
of remains recovered from prehistoric and early historic levels of Magula
Pevkakia in Thessaly, a decrease likely to be related to the deer being
confined to increasingly marginal areas as the result of the spread of human
settlement. These examples make it clear that, just as with age and sex
ratios, size diminution alone is not sufficient evidence for demonstrating the
existence of the domestication process.

Conclusion

Arguments for animal domestication based on faunal remains from

archaeological sites are likely to be more convincing if they employ
multiple lines of evidence than if they are based on any one feature alone.
Furthermore, examination in isolation of such traits as increasing
proportions of specimens from selected species, morphological change, size
diminution, or changing age and sex ratios, and faunal distribution patterns
can provide little insight into coevolutionary aspects of animal and human
societies. Where possible, all of these features are best examined together
and trends documented and evaluated on the basis of large faunal
collections from single sites covering significant spans of time, or from
multiple sites within a limited region. In addition, interpretations are best
made within the archaeological context of the site and region being
examined, because only then can features of community and settlement
patterning, site structure, and material culture be evaluated and related to
the faunal remains. Such an integrated approach calls for archaeologically
literate faunal analysts and zooarchaeologically literate archaeologists, each
with a good understanding of the possibilities and limitations of the other’s
data. It also calls for an appreciation of biological and behavioural factors
related to both the humans and the non-human animal taxa concerned.

References
Bar-Yosef, O. 1981. The Epi-Palaeolithic complexes in the Southern Levant. In PréhistoireduLevant.
Colloques Internationaux du CNRS no. 598, 389–408. Paris: Editions du CNRS.
Boessneck, J. & A. von den Driesch 1976. Pferde im 4./3. Jahrtausend v. Chr. in Ostanatolien.
Saügetierkundliche Mitteilungen 24(4), 81–7.
Boessneck, J. & A. von den Driesch 1978. The significance of measuring animal bones from
archaeological sites. In Approaches to faunal analysis in the Middle East, R. H. Meadow & M. A.
Zeder (eds). Peabody Museum Bulletin no. 2, 25–39. Cambridge, Mass.: Harvard University.
Bökönyi, S. 1969. Archaeological problems and methods of recognizing animal domestication. In
The domestication and exploitation of plants and animals, P. J. Ucko & G. W. Dimbleby (eds),
219–29. London: Duckworth.
Bökönyi, S. 1974. History of domestic mammals in central and eastern Europe. Budapest:
Akadémiai Kiadö.
Bökönyi, S. 1975. Some problems of animal domestication in the Middle East. In
Domestikationsforschung und Geschichte der Haustiere, J. Matolcsi (ed.), 69–75. Budapest:
Akadémiai Kiadö.
Bökönyi, S. 1976. Development of early stock rearing in the Near East. Nature 264, 19–23.
Bökönyi, S. 1977. Animal remains from the Kermanshah Valley, Iran. Oxford: BAR Supplementary
Series 34.
Bökönyi, S. in press. Late Chalcolithic horses in Anatolia. In Equids in the ancient world II, R. H.
Meadow & H.-P. Uerpmann (eds), Beihefte zum Tübinger Atlas der Vorderen Orients, Reihe A,
Nr. 19/2. Wiesbaden: Dr Ludwig Reichert.
Clutton-Brock, J. 1981. Domesticated animals from early times. London: British Museum (Natural
History) & Heinemann.
Collier, S. & J. P. White 1976. Get them young? Age and sex inferences on animal domestication in
archaeology. American Antiquity 41(1), 96–102.
Compagnoni, B. & M. Tosi 1978. The camel: its distribution and state of domestication in the Middle
East during the third millennium B.C. in light of finds from Shahr-i Sokhta. In Approaches to
faunal analysis in the Middle East, R. H. Meadow & M. A. Zeder (eds), Peabody Museum
Bulletin no. 2, 91–103. Cambridge, Mass.: Harvard University.
Davis, S. J. M. 1981. The effects of temperature change and domestication on the body size of Late
Pleistocene to Holocene mammals of Israel. Paleobiology 7(2), 101–14.
Ducos, P. 1969. Methodology and results of the study of the earliest domesticated animals in the Near
East (Palestine). In The domestication and exploitation of plants and animals, P. J. Ucko & G. W.
Dimbleby (eds), 265–75. London: Duckworth.
Ducos, P. 1978. Domestication defined and methodological approaches to its recognition in faunal
assemblages. In Approaches to faunal analysis in the Middle East, R. H. Meadow & M. A. Zeder
(eds), Peabody Museum Bulletin no. 2, 53–6. Cambridge, Mass.: Harvard University.
Dyson, R. H. Jr. 1953. Archeology and the domestication of animals in the Old World. American
Anthropologist 55(5), 661–73.
Franceschi, P. F. 1980. Essai de caractérisation génétique du pore corse. Thesis 3ème cycle. Paris
VI: l’Université Pierre et Marie Curie.
Gebel, H.-G. 1984. Das Akeramische Neolithikum Vorderasiens. Beihefte zum Tübinger Atlas des
Vorderen Orients, Reihe B. Nr. 52. Wiesbaden: Dr Ludwig Reichert.
Geddes, D. S. 1985. Mesolithic domestic sheep in West Mediterranean Europe. Journal of
Archaeological Science 12(1), 25–48.
Groves, C. P. 1989. Feral mammals of the Mediterranean islands: documents of early domestication.
In The walking larder, J. Clutton-Brock (ed.), ch. 5. London: Unwin Hyman.
Hecker, H. 1982. Domestication revisited: its implications for faunal analysis. Journal of Field
Archaeology 9(2), 217–36.
Hesse, B. C. 1978. Evidence for husbandry from the early Neolithic site of Ganj Dareh in western
Iran. Unpublished PhD dissertation, Faculty of Political Science, Columbia University. Ann
Arbor: University Microfilms International no. 78–9905.
Higgs, E. S. & M. R. Jarman 1969. The origins of agriculture: a reconsideration. Antiquity 43(169),
31–41.
Hole, F., K. V. Flannery & J. A. Neely 1969. Prehistory and human ecology of the Deh Luran Plain.
Memoirs of the Museum of Anthropology, no. 1. Ann Arbor: University of Michigan.
Ingold, T. 1980. Hunters, pastoralists, and ranchers. Cambridge: Cambridge University Press.
Jarman, M. R. & P. F. Wilkinson 1972. Criteria of animal domestication. In Papers in economic
prehistory, E. S. Higgs (ed.), 83–96. Cambridge: Cambridge University Press.
Jordan, B. 1975. Tierknochenfunde aus der Magula Pevkakia in Thessalien. Inaugural Dissertation.
Munich: Institut für Palaeoanatomie, Domestikationsforschung und Geschichte der Tiermedizin
der Universität München.
Köhler-Rollefson, I. 1986. From sedentary animal herding and hunting to semi-nomadic
pastoralism: the faunal evidence from Neolithic Ain Ghazal (Jordan). Unpublished paper
presented at 5th International Conference of the International Council for Archaeozoology,
Bordeaux, France, 25–30 August.
Mason, I. L. (ed.) 1984. Evolution of domesticated animals. London: Longman.
Meadow, R. H. 1984. Animal domestication in the Middle East: a view from the eastern margin. In
Animals and archaeology. Vol. 3: Early herders and their flocks, J. Clutton-Brock & C. Grigson
(eds), 309–37. Oxford: BAR International Series 202.
Noddle, B. 1986. Flesh on the bones. Productivity of animals in the past: speculations based on
historical data and the performance of primitive livestock in Third World countries. Unpublished
paper presented at 5th International Conference of the International Council for Archaeozoology,
Bordeaux, France, 25–30 August.
Payne, S. 1973. Kill-off patterns in sheep and goats: the mandibles from Asvan Kale. Anatolian
Studies 23, 281–303.
Perkins, D. Jr. 1964. Prehistoric fauna from Shanidar, Iraq. Science 144, 1565–6.
Pietschmann, W. 1977. Zur Grösse des Rothirsches (Cervus elaphus L.) in vor-und
frühgeschichtlicher Zeit. Inaugural Dissertation. Munich: Institut für Palaeoanatomie,
Domestikationsforschung und Geschichte der Tiermedizin der Universität München.
Poplin, F. 1979. Origine du mouflon de Corse dans une nouvelle perspective paléontologique: par
marronnage. Annales de Génétique et de la SélectionAnimale 11, 133–43.
Rauh, H. 1981. Knochenfunde von Säugetieren aus dem Demircihüyük (Nordwest-anatolien).
Inaugural Dissertation. Munich: Institut für Palaeoanatomie, Domestikationsforschung und
Geschichte der Tiermedizin der Universität München.
Redding, R. W. 1984. Theoretical determinants of a herder’s decisions: modelling variation in the
sheep/goat ratio. In Animals and Archaeology. Vol. 3: Early herders and their flocks, J. Clutton-
Brock & C. Grigson (eds), 223–41. Oxford: BAR International Series 202.
Reed, C. A. 1960. A review of the archeological evidence on animal domestication in the prehistoric
Near East. In Prehistoric investigations in Iraqi Kurdistan, R. J. Braidwood & B. Howe (eds),
Studies in Ancient Oriental Civilizations (SAOC) no. 31, 119–45. Chicago: University of Chicago
Press.
Reed, C. A. 1983. Archeozoological studies in the Near East: a short history (1960–1980). In
Prehistoric archeology along the Zagros flanks, L. Braidwood, R. J. Braidwood, B. Howe, C. A.
Reed & P. J. Watson (eds), Oriental Institute Publications, Vol. 105, 511–36. Chicago: Oriental
Institute of the University of Chicago.
Simmons, A. H. & G. Ilany 1975–77. What mean these bones? Behavioral implications of gazelles’
remains from archaeological sites. Paléorient 3, 269–74.
Stampfli, H. R. 1983. The fauna of Jarmo with notes on animal bones from Matarrah, the Amuq, and
Karim Shahir. In Prehistoric archeology along the Zagros flanks, L. S. Braidwood, R. J.
Braidwood, B. Howe, C. A. Reed & P. J. Watson (eds), Oriental Institute Publications, Vol. 105,
431–83. Chicago: Oriental Institute of the University of Chicago.
Uerpmann, H.-P. 1978. Metrical analysis of faunal remains from the Middle East. In Approaches to
faunal analysis in the Middle East, R. H. Meadow & M. A. Zeder (eds), Peabody Museum
Bulletin no. 2, 41–5. Cambridge, Mass.: Harvard University.
Uerpmann, H.-P. 1979.Probleme der Neohthisierung des Mittelmeerraumes. Beihefte zum Tübinger
Atlas des Vorderen Orients, Reihe B, Nr. 28. Wiesbaden: Dr Ludwig Reichert.
Uerpmann, H.-P. 1987. The ancient distribution of ungulate mammals in the Middle East. Beihefte
zum Tübinger Atlas des Vorderen Orients, Reihe A, Nr. 27. Wiesbaden: Dr Ludwig Reichert.
Vigne, J.-D. 1984. Premières données sur le début de l’élevage du mouton, de la chèvre et du pore
dans le sud de Corse (France). In Animals and archaeology. Vol. 3: Early herders and their flocks,
J. Clutton-Brock & C. Grigson (eds), 47–65. Oxford: BAR International Series 202.
Widdowson, E. M. 1980. Growth in animals. In Growth in animals, T. L. J. Lawrence (ed.), 1–9.
London: Butterworth.
Wright, G. A. & S. J. Miller 1976. Prehistoric hunting of New World wild sheep: implication for the
study of sheep domestication. In Culture change and continuity, C. E. Cleland (ed.), 293–312.
New York: Academic Press.
Wright, H. T., N. Miller & R. W. Redding 1980. Time and process in an Uruk rural center. In
L’archéologie de l’lraq du débutdeVépoque Néolithique â 333 avant notre ère. Colloques
Internationaux du CNRS 580, 265–84. Paris: Editions du CNRS.
Zeuner, F. E. 1955. The goats of early Jericho. Palestine Exploration Quarterly April, 70–86.
Zeuner, F. E. 1963.A history of domesticated animals. London: Hutchinson.
9 Animal exploitation and the phasing of
the transition from the Palaeolithic to
the Neolithic
HANS-PETER UERPMANN

Introduction

The decision whether plant cultivation or animal husbandry was practised

by a prehistoric population is usually based on surviving plant and/or
animal remains. However, these two sources of information are neither
equally available nor can they be evaluated scientifically to the same extent.
Generally, animal bones are among the most common finds in
archaeological contexts, whereas plant remains are often rare or absent.
Moreover, animal bones are commonly left as a result of animal
exploitation, whereas many plants of economic importance do not produce
waste materials likely to become part of the archaeological record. In
addition, bone remains are closely correlated to the type of use of the
respective species. The skeleton of each taxon has a definitive number of
elements and the survival of these remains is not too different from taxon to
taxon. Thus, contrary to plants, fairly reliable estimations of the relative
importance of different animals can be based on bone finds from
archaeological sites. With plants, on the other hand, meaningful proportions
can only be established between remains that are related in the same way to
the useful parts as bones are related to flesh. This is the case, for example,
with nutshells. However, any plant remains which are dependent on
carbonization for preservation will always reflect their particular
susceptibility to being burnt rather than their relative economic importance.
In the context of the beginnings of cultivation and domestication, there
are morphological responses that enable us to recognize the processes.
These morphological changes probably happen in animals faster than in
plants, or, perhaps more correctly, it is less easy to imagine early forms of
animal husbandry which would not produce morphological changes than it
is to suggest early forms of cultivation which would leave the respective
plants unchanged. Therefore, if an economic definition of the Neolithic
versus the Palaeolithic period is applied, the insight into the animal part of
the economy is potentially more distinct than the information available on
the vegetal side.
It seems appropriate, therefore, to develop a scheme of definitions based
mainly on animal remains through which the various stages of the transition
can be classified from Palaeolithic hunters and gatherers to Neolithic
farmers and herders, at least for the circum-Mediterranean areas and the
adjacent parts of Europe and western Asia.

Terminology of the Palaeolithic–Neolithic transition

Rather than creating another new terminology, existing terms like

‘Neolithic’ and ‘Mesolithic’ will be assigned in this chapter to a more
precise socio-economic meaning. This fairly conservative approach is based
on accepted Childean criteria for the Neolithic. In the 1950s Gordon Childe
took an economy based on plant cultivation and animal husbandry as the
defining character for distinguishing the Neolithic from the Palaeolithic
period. This definition provided a more meaningful subdivision of the Stone
Age than did the former characterization of the Neolithic as the period of
polished stone versus flaked stone tools. However, it would be
unsatisfactory merely to seek evidence of’plant cultivation’ and/or ‘animal
husbandry’ when attempting to characterize the actual processes of
economic change in the Fertile Crescent of Southwest Asia between about
10 000 and 6000 BC. It is now clear that the domestication of meat-
producing animals is only the last major step in the evolution of the
Neolithic economy. There is evidence for plant cultivation prior to animal
domestication, and it is likely that plant cultivation is only an intermediate
step in the development of the early Neolithic. The initial stage of this
process is best described as an economy based on ‘intensive plant
exploitation including long-term storage of vegetal food’. The village facies
of the Natufian culture in the Levant is the prototype of this stage of socio-
economic development.
 Given the problems of the evidence about the exploitation of plants as
outlined above, one cannot exclude the possibility that the Natufian
population had already discovered methods of stimulating the growth of
cereals and vegetables, and thus had begun plant cultivation. The
succeeding culture, which in order to use a consistent terminology should
be called ‘Sultanian’, evidently practised some agriculture. There are cereal
remains from Sultanian contexts which differ sufficiently from their
ancestors to indicate changes in reproductive selection due to manipulation
by early agriculturalists. The still more widespread use of the term ‘Pre-
pottery Neolithic A’ (PPNA) for this cultural unit should be abandoned for
two reasons:

(a) this term is inconsistent with the use of type–site names for the
designation of Stone-age ‘cultures’;
(b) the ‘PPNA’ is not a neolithic culture in the full sense.

Because it is lacking evidence of domesticated meat-producing animals, the

Sultanian and similar cultural units should be grouped with the preceding
Natufian into a stage called ‘Protoneolithic’. Sites like Zawi Chemi, Karim
Shahir, and Asiab in the northern part of the Fertile Crescent are also
included in this term, although evidence for ‘intensive exploitation of plants
including long-term storage of vegetal food’ can sometimes only be
inferred indirectly from structures which suggest some degree of sedentism.
Incipient sedentism is here considered both as a result of, and a
presupposition for, effective long-term storage of larger quantities of food.
Sedentism is also instrumental in the evolution of some of the features of
the fully developed Neolithic, including animal domestication. Therefore,
sedentism based on intensive exploitation and long-term storage of vegetal
food in the absence of domestic animals for meat-production should be the
defining characteristic of the Protoneolithic stage of economic
development.
Among the first civilizations to combine animal husbandry with the
features already present in the Protoneolithic was the ‘Pre-pottery Neolithic
B’ of the Levant. The PPNB is a real Neolithic culture, having both plant
cultivation and animal husbandry as defining characters. However, one
should also apply type-site names for this earliest horizon of the Neolithic
period. The name ‘Tahunian’ has been used in this context (Prausnitz 1970),
but one might also propose to use Beidha as an eponymic site because of
the availability of an extensive description of its flint industry (Mortensen
1970). Good evidence for animal domestication at this chronological level
has been produced for several sites in the southern Levant, including
Beidha (Hecker 1974, 1982, Uerpmann 1979), Jericho (Clutton-Brock &
Uerpmann 1974, Clutton-Brock 1979), and others. In the north of the
Fertile Crescent, the upper part of the aceramic levels of ×ayönü represent a
real Neolithic (Lawrence 1980, 1982). Jarmo (Stampfli 1983) and Ali Kosh
(Hole et al. 1969) are examples from the foothills along the northeastern
border of Mesopotamia, whereas Ganj-e Dareh (B-D) is an early
representative of a Neolithic economy in the Zagros mountains (Hesse
1978).
The first animals to be domesticated for meat production during the 7th
millennium BC were sheep and goat, both in different parts of the Fertile
Crescent (Uerpmann 1979, 1987). The full potential of Neolithic economy
was only reached, however, more than half a millennium later when cattle
and swine joined the domestic livestock. Their domestication could be used
to differentiate between an early and a developed stage of the Neolithic
economy. It is less easy, though, to trace domestication in cattle than in
sheep or goat. Cattle bones are usually even more broken than those of the
smaller ruminants, and the ecological shifts imposed on bovines during
early domestication are potentially less severe than in sheep and goat.
Nevertheless, there is at least a theoretical distinction possible between an
‘early Neolithic’ with an incomplete set of domestic animals, and a ‘full
Neolithic’ with cattle, pig, sheep, and goat forming the complete list of
protein-producing domesticates of this period. The later use of these
animals for other products than meat adds further facets to the Neolithic
economy (Sherratt 1981), but is not marked as a clear-cut ‘revolution’. If
there is evidence for the use of secondary products, however, it can be used
to define a ‘late Neolithic’.
The end of the Neolithic period can also be defined in terms of animal
exploitation: the domestication of special animals for purposes of labour,
like donkey, horse, and camel, is a reflection of the changes in overall
economic patterns at the beginning of the metal ages (the formation of cities
and states in the Middle East). Thus, the end of the Neolithic is marked by a
further broadening of the exploited resources. Like all the economic shifts
described above, it consists of additions of new subsistence technologies
and of new schemes of exploitation of the environment to a previous pattern
which was almost never completely replaced.

Terminology of other economic stages

Within the circum-Mediterranean area the data available at present indicate

that an autochthonous development from the Epipalaeolithic through the
Protoneolithic into the Neolithic economy only occurred in the Fertile
Crescent of southwest Asia. The Epipalaeolithic is found almost
everywhere in the Old World, and the Neolithic spread over most of it in
later times, but the Protoneolithic is confined to the core area where natural
stands of wild cereals and pulses favoured ‘intensive plant exploitation
including long-term storage of vegetal food’ and the subsequent
development of sedentism (Uerpmann 1979). However, a fairly high degree
of sedentism also seems to have been achieved by some ‘Mesolithic’
civilizations without signs of an intensified use of plants. The economic
basis of this sedentism is the exploitation of aquatic animal populations,
mainly in the form of fish and shellfish. The sedentism reached by some
fishing populations actually led to cultural developments diverging from a
Palaeolithic lifestyle. There are similarities with Neolithic features which
actually justify an intermediate term for this stage of development. The
‘Mesolithic’ as a term in economic prehistory should be restricted to groups
with a non-productive economy which nevertheless allowed for some
degree of sedentism based on the exploitation of aquatic resources.
The general hunters and gatherers of the post-Pleistocene, which are also
often included in the term Mesolithic, do not need a special intermediate
term between the Palaeolithic and the Neolithic. From their economy they
are clearly part of the Palaeolithic and might best be grouped under the term
‘Epipalaeolithic’. An economic definition of this term is difficult, although
the use of long-ranging versus medium-ranging weapons (bow and arrows
versus spear) for hunting might provide a basis for its discrimination from
the Upper Palaeolithic, both in the tool kit and in the choice of prey. In
particular, in those parts of the Middle East which had open vegetation,
gazelles only became significantly available to prehistoric hunters after the
invention of long-ranging weapons. The introduction of the bow and arrows
may have been the reason for the apparent expansion of the Epipalaeolithic
into semi-desert areas.
Obviously, Epipalaeolithic groups existed alongside Protoneolithic
villages, and in particular the Natufian had an Epipalaeolithic and
Protoneolithic facies. The limits of typologically defined ‘cultures’ do not
necessarily coincide with economic units such as the ‘Neolithic’ and the
‘Palaeolithic’. This is an important point, for a number of prehistoric
groups, mainly in the deserts of Arabia and North Africa, are called
‘Neolithic’ because of the presence of arrow heads or pottery. So long as the
economic characteristics of the Neolithic have not been demonstrated for
these groups they should really be given a locally defined name which does
not include them in one of the larger periods.
There are, however, prehistoric groups with known economic characters
that do not fit the scheme developed above. For example, there existed so-
called Mesolithic groups in the south of France which had domestic sheep
(Geddes 1985). Apparently, the ascription of these groups to the Mesolithic
is only based on the fact that no pottery was found in their sites. Since this
is not a diagnostic feature, one could easily include the ‘Castelnovien’
within the Neolithic, if there were indications of the other defining
characters of the Neolithic. These examples are most probably groups
which have lost, or only adopted part of, the Neolithic economy during their
distribution and diffusion outside of the central areas of Neolithic
development in southwest Asia. The term ‘Paraneolithic’ seems appropriate
to describe the peripheral situation of these groups which are in the process
of economic change.
Plainly nomadic herders, who apparently did not practise plant
cultivation, also cause terminological problems. Although there is no good
evidence that there was a direct transition from hunters to herders without a
Protoneolithic interlude, such a shift is at least a theoretical possibility. The
term ‘Semi-neolithic’ might be appropriate with regard to the fact that these
groups lack the plants of the full Neolithic. This term should, however, not
be used for derived groups which have given up agriculture while adapting
to particular environmental circumstances. In this case, terms like the
notorious ‘Shepherd Neolithic’ more closely reflect actual developments.
 Another common deviation from the more normal pattern are real
Mesolithic economies with the added use of some domestic animals. In fact,
the exploitation of aquatic resources may well have been much more
significant for the economic status of the respective groups than was the
contribution made by domestication (e.g. Biagi et al. 1984). In such cases,
composite terms such as Meso-Neolithic, could be used and similar names
for other mixtures of Palaeo-, Meso-, and Neolithic features could be
developed.
The importance of having a precise terminology in common usage is to
specify the socio-economic processes and changes which took place. The
particular terminology suggested in this chapter cannot be directly applied
to areas unconnected with Neolithic development in the Near and Middle
East. However, it is hoped that the suggested precision in defining the
relevant processes under consideration may aid comparative studies of the
origins of productive economies in other parts of the world as well.

References
Biagi, P., W. Torke, M. Tosi & H.-P. Uerpmann 1984. Qurum: a case study of coastal archaeology in
Northern Oman. World Archaeology 16, 43–61.
Clutton-Brock, J. 1979. The mammalian remains from the Jericho Tell. Proceedings of the
Prehistoric Society 45, 135–57.
Clutton–Brock, J. & H.–P. Uerpmann 1974. The sheep of early Jericho. Journal of Archaeological
Science 1, 261–74.
Geddes, D. 1985. Mesolithic sheep in west Mediterranean Europe. Journal of Archaeological Science
15, 25–48.
Hecker, H. M. 1974. The faunal analysis of the primary food animals from pre-pottery Neolithic
Beidha (Jordan). Unpublished PhD dissertation, Columbia University, New York.
Hecker, H. M. 1982. Domestication revisited: its implications for faunal analysis. Journal of Field
Archaeology 9, 217–36.
Hesse, B. C. 1978. Evidence for husbandry from the early Neolithic site ofGanj Dareh in western
Iran. Unpublished PhD dissertation, Columbia University, New York.
Hole, F., K. V. Flannery & J. A. Neely, 1969.Prehistory and human ecology of the Deh Luran Plain.
Memoirs of the Museum of Anthropology, University of Michigan no. 1. Ann Arbor: Museum of
Anthropology.
Lawrence, B. 1980. Evidences of animal domestication at Qayönii. In The joint Istanbul-Chicago
universities’ prehistoric research in southeastern Anatolia t, H. Qambel & R. J. Braidwood (eds),
285–308. Istanbul: Istanbul University, Faculty of Letters no. 2589.
Lawrence, B. 1982. Principal food animals at Qayönü. In Prehistoric village archaeology in south-
eastern Turkey, L. S. Braidwood & R. J. Braidwood (eds), 175–99. Oxford: BAR International
Series 138.
Mortensen, P. 1970. A preliminary study of the chipped stone industry from Beidha, an early
neolithic village in southern Jordan. Acta Archaeologica 41, 1–54.
Prausnitz, M. W. 1970. From hunter to farmer and trader: studies in the lithic industries of Israel
and adjacent countries. Jerusalem: Sivan Press.
Sherratt, A. 1981. Plough and pastoralism: aspects of the secondary products revolution. In Patterns
of the past, I. Hodder, G. Isaac & N. Hammond (eds), 261–305. Cambridge: Cambridge University
Press.
Stampfli, H. R. 1983. The fauna of Jarmo with notes on the animal bones from Matarrah, the ‘Amuq,
and Karim Shahir. In Prehistoric archaeology along the Zaqros flanks, L. S. Braidwood, R. J.
Braidwood, B. Howe, C. A. Reed & P. J. Watson (eds), Oriental Institute Publications Vol. 105,
Chicago: Oriental Institute of the University of Chicago.
Uerpmann, H.–P. 1979.Probleme der Neolitkisierungdes Mittelmeerraumes. Beihefte zum Tübinger
Atlas des Vorderen Orients, Reihe B, Nr.28. Wiesbaden: Dr Ludwig Reichert.
Uerpman, H.-P. 1987.The ancient distribution of ungulate mammals in the Middle East. Beihefte zum
Tübinger Atlas des Vorderen Orients, Reihe A, Nr. 27. Wiesbaden: Dr Ludwig Reichert.
10 A two-part, two-stage model of
domestication
FRANK HOLE

It has been customary to conceive of domestication as a ‘revolutionary’

event whose essential elements were in place throughout Southwest Asia as
early as the 10th millennium cal. BC, and were fully established by the end
of the 8th millennium BC. The events leading to domestication have
recently been re-examined, leading to two different points of view (Hole
1984, Moore 1982, 1985). Both Hole and Moore argue for a long period of
development of plant domestication, but Hole considers livestock to have
been brought under control through entirely separate processes, perhaps
without a lengthy period of development, and that it was the conjunction of
these two economic adaptations that resulted in the Neolithic revolution. In
this chapter I shall briefly outline the substance of this argument and then
look specifically at evidence for domestication of sheep and goats.
Essentially, archaeological and zoological evidence suggests that sheep
and goats were domesticated earlier in the Zagros Mountains of eastern
Anatolia and western Iran than in the Levant. Although these species are
present sporadically in the Levant and Anatolia, they do not constitute a
consistently high proportion of remains in archaeological sites in these
western regions until sedentism with cereal agriculture is manifest.
Howe (1983) has outlined the archaeological and zoological finds at the
earliest settlements outside the coastal Levant, with special emphasis on the
Zagros region. Here, a series of sites estimated to be between 8000 and 10
000 cal. BC exhibit the following: semi-sedentary camps lacking traces of
permanent architecture, although some circular huts are present; widespread
use of stones for pounding or grinding vegetable food; a mix of fauna but
with an emphasis on sheep/goat. The locations and physical conditions of
these sites suggest that they were used as seasonal camps rather than as
permanent residences. These camps, in the foothills and mountainous
regions, were probably used during the warmer months when cereals or
other plants would have been ripe. In short, these sites – Shanidar Bl, Zawi
Chemi, Karim Shahir, M’Lefaat, Gird Chai, and Asiab – all suggest some
form of seasonal transhumance, whether for herding and hunting, or
collecting and harvesting of planted fields. At this stage – often called the
Protoneolithic – there are no permanent settlements known.
Because of one radiocarbon date, Ganj Dareh E has often been
considered as a Protoneolithic component in a site that otherwise fits
comfortably into the early Neolithic. However, in a detailed review of the
grounds for dating this and other sites, Hole (in press) prefers the later date
in cal. years BC. This means that Tepes Qazemi and Ghenil are, likewise,
8th millennium. The dates for Shanidar and Zawi Chemi can also be
questioned, particularly as they relate to possible evidence of sheep
domestication, but the relative antiquity of these sites and the others
mentioned seems correct. One hopes that the recent excavations by Stefan
Kozlowski at Nemrik on the upper Tigris will yield dates for this era.
Sites pre-dating the Protoneolithic (generally known as Zarzian and
Baradostian) were very much smaller and often located in caves and
shelters (Hole & Flannery 1967). At none of the Pleistocene sites is there
evidence of either the use of grinding stones or of domesticated livestock.
Thus, the nature of the later Holocene sites and their distributions, coupled
with faunal evidence, suggest that the first step in the process of sheep and
goat domestication had been taken by 10 000 cal. BC, although there is
debate among faunal specialist s whether any or all of these sites actually
evidences the use of domestic stock. The important issue is that the pattern
of life essential to a herding economy had been taken with the seasonal
reuse of these ‘pastoral’ camps. Although the evidence varies from site to
site, it is clear that the processing of vegetable foods went hand in hand
with these seasonal movements.
A critical gap in our knowledge of this era exists for most of Anatolia,
particularly the eastern and southern portions which lie adjacent to the
Zagros, and a similar gap exists on the Soviet side. But perhaps the most
crucial gap is what happened on the rolling foothills at the base of the
mountains in northeastern Iraq and on the adjacent north Mesopotamian
steppe which today provides a transition to the desert. Here, where much
attention has been placed upon mounded sites, one would expect to find the
lowland counterparts to the upland Zawi Chemi type sites (Nemrik is
probably one such site). If other camping debris from this era exists, it has
gone unnoticed in previous surveys, and it may be that most of it is no
longer visible at the surface.
The pattern in the Levant has been summarized by Bar-Yosef (1980) and
Moore (1985). Here, evidence for the introduction of domestic stock,
chiefly goats, appears only in the 8th millennium, during the Pre-pottery
Neolithic B, in spite of a very long history of semi-sedentary and sedentary
sites at which there is abundant evidence of the grinding and pounding of
vegetable foods. Indeed, the appearance of domesticated sheep, goats,
cattle, and pigs heralds a dramatic replacement of the indigenous gazelle,
onager, and deer populations. Sedentism, as manifest at Jericho, Mureybat,
As wad (near Damascus), Abu Hureyra, and Mallaha, preceded the
introduction of domestic livestock by as much as 2000 years. A similar
pattern of abrupt replacement of wild stock is repeated at other sedentary
sites, such as Qayönü in south-central Anatolia (Lawrence 1980).
The archaeological evidence thus suggests a differential timing in the
development of the chief components of domestication: herding is earlier in
the east and plant processing and sedentism are much older in the west.
This assessment of the differences between the two areas is supported by
cytological evidence of a Levantine origin for the domesticated races of
wheat, peas, and lentils and their subsequent dispersal as an agricultural
complex through the Near East and ultimately into Europe (Zohary pers.
comm.). The independent development of these two lines toward fully
effective domestication is Stage One of my rubric, in which the
development of a herding economy is Part One and the slow separate
evolution of an agricultural economy out of a long tradition of plant
processing is Part Two. However, one must not ignore the deplorable state
of our knowledge about these eras. If, as I argue, intensive plant use was a
lowland adaptation, then we would not expect to find early evidence in the
mountainous regions. One might then postulate seed-using economies
developed as early in Mesopotamia as in the Levant, but present evidence is
insufficient to test this hypothesis. In fact, although we know of a long
tradition of grinding-stone technology in the Levant, we are still not able to
state with assurance when cultivation began, so that even in this region the
rapidity of domestication is in doubt. Nevertheless, in each case, patterns of
behaviour – vertical transhumance in the case of sheep and goats, and
storage/sedentism in the context of an intensive plant–gathering economy –
must have been established for some time before a full transition to the use
of domesticates came about.
Stage Two is represented by the 8th millennium in the Near East as the
widespread appearance of permanent settlements which make use of both
domestic animals and plants. It is important to note that sedentism in the
west precedes Stage Two by as much as 2 millenniums, whereas sedentism
(as opposed to seasonal occupancy) in the zone of animal domestication
coincides with the introduction of cereal domestication.
Thus, an outline of the structure and timing of the processes of
domestication is evident even though there are still many problems
concerning the evidence and its interpretation (see Ducos & Helmer 1981).
But perhaps the most important remaining problems are conceptual and
access to relevant sites. Since Braidwood and Childe defined the problems
in archaeological terms (Braidwood & Howe 1961, pp. 1–9), the search for
suitable sites has concentrated on the ‘nuclear’ zones where many species
converged. There are three problems with this:

(a) the environment of southwest Asia changed markedly during the time
of concern;
(b) although there are broad zones where many species overlap in
geographic distribution today, this was not necessarily the case at the
time domestication began; and
(c) each species of animal, and to a lesser extent plant, has its own
characteristics and behaviour which required different methods of
exploitation by humans.

It is not surprising, therefore, to find that the species are differentially

represented in sites, especially during the early stages of domestication.
Thus, we find a preponderance of sheep over goats at Zawi Chemi and of
goats over sheep at Asiab. These occurrences probably reflect different
local environments. But we also find that cattle and pigs, which were
probably distributed throughout the Near East, are rarely present as
domesticates in early sites. Indeed, domestic cattle, Bos taurus, are thought
to have originated in Greece or western Anatolia (but see Meadow 1984 for
Bos indicus), and pigs are absent as demonstrable domesticates before the
8th millennium. These latter omissions are unlikely to have resulted from
environmental considerations; rather, the explanation probably resides in
the behaviours of the species and their interactions with humans. The
picture that emerges, therefore, is one of a series of experiments by people
who took advantage of the locally dominant fauna, which were both
desirable as food and tractable to human control. Of these available
potential domesticates, sheep and goats, both relatively benign, gregarious,
herd animals, lent themselves best to human domination. We must look,
therefore, not for one event – ‘domestication’ – but at a series of events that
brought different animals and plants under control. The problem is thus
more complicated than once conceived, but perhaps more approachable
because we can now focus on separate aspects of the transition rather than
having to deal with all of its parts simultaneously.
Because of environmental changes at the end of the Pleistocene and since
(including man’s dispersal of species), distributions of fauna may have
differed substantially from the present so that today’s maps do not
accurately reflect the biogeographic landscape in which domestication took
place. In particular, changes in the location and proportions of forest and
forage will have had a determining effect upon the distributions of the
browsers (goats) versus the grazers (sheep). In this regard, it is interesting,
in view of the rapid spread of sheep after domestication, that they are
seldom found in either Palaeolithic or early Neolithic sites, whereas goats
are relatively common, along with various other hunted species such as red
deer, gazelles, onagers, and oxen. Sheep occur in the Pleistocene caves of
Shanidar, Palegawra, and Douara. In the early Holocene they are found at
Zawi Chemi and Asiab in the Zagros, Abu Hureyra and Mureybet on the
Euphrates, and in sites in the eastern and southern Levant, including G VIII
in the Negev. These two distributions are strikingly different and suggest
that sheep were not distributed in the past as they are today – that they were
not abundant in the places where we have dug sites. As sheep are grazers
who prefer rolling country, the upper margins of the Fertile Crescent would
appear to have been the most likely places for sheep to have lived.
However, as sheep do not penetrate forested regions, it is possible that their
distribution lay south of what we consider today to be steppe.
According to pollen studies, at the end of the Pleistocene conditions were
drier and cooler and began to grow warmer and wetter after 10 000 cal. BC,
resulting in slow resurgence of forest and grasses over lands that previously
had been arid, shrubby steppe (Van Zeist & Bottema 1982). That some
forest and cereals remained in isolated refuges seems inescapable, and is
supported by the presence of red deer in several Palaeolithic sites in the
Zagros. Nevertheless, for the most part during the late Pleistocene, the
mountainous regions would have been much less suitable to sheep than to
goats, a conclusion that is supported by faunal data from sites in the same
region – only Shanidar and Palegawra Caves hold any evidence of sheep in
the Pleistocene (Turnbull & Reed 1974, Evins 1982).
Following the Pleistocene, it is generally thought that forests stretched
well out onto the lowland steppe to approximately the 250 mm rainfall
isohyet, a zone that today lacks trees owing largely to human interference.
The grassy stretches that would have been preferred by sheep may,
therefore, have been still farther south in the zone that we regard today as
desert-steppe. Evidence to support this is meagre, largely because sheep are
only minor components of early archaeological faunal assemblages. For
example, in the Epipalaeolithic levels at Abu Hureyra 85 per cent of the
fauna is gazelle, and sheep/goat compose only 8 per cent. Interestingly, here
both sheep and goat are present, although one would not expect to find the
latter so far from mountainous terrain (Moore et al. 1986). At Mureybet,
just 35 km north of Abu Hureyra, also on the Euphrates, a slightly later
occupation contains a few sheep but not goats, and the majority of fauna
consists of ass and gazelle. Ducos (1978) offers a behavioural reason for the
relative lack of sheep in these sites: gazelles are easier to hunt, and sheep
were taken only incidentally during gazelle hunts.
It is clearly established that sheep occupied the upper Mesopotamian
lowlands at the time of the earliest Holocene settlements and perhaps the
western Levant as far south as the Negev (Ducos 1978, Payne 1983, Davis
1982); it remains problematic whether they also occupied the mountain
zone to the north. Their entry into this habitat may have been more a
response to the decline of woody and shrubby vegetation following human
uses of the land, than to the ‘naturalness’ ofthat particular habitat for sheep.
That the mountain plains held some forest during the late Pleistocene (in
spite of some pollen indications) is suggested by the ubiquity of red deer
and cattle in Palaeolithic sites (e.g. Shanidar, Gar Arjeneh, and Bisitun).
That much open ground also existed seems indicated by the presence of
onager and gazelles, but the important point is that routes of migration
between grazing areas may have been closed during the early Holocene by
forest and scrub, both of which would impede the travel of sheep more so
than of goats. I argue, therefore, for a spatial separation of the major
concentrations of sheep from goats at the time of early domestication.
According to cytological studies, the wild ancestor of the domestic sheep
is Ovis orientalis (Armenian variety) whose present distribution is west of a
line from the Caspian Sea to the Arabian Gulf (Valdez et al. 1978).
Although I have remarked above on the problems of using modern
distributional data to infer past situations, it is interesting to note that the
cytological data support an origin in northeastern Mesopotamia, as would
the Shanidar and Zawi Chemi fauna. However, on modern environmental
grounds, Ducos Sc Helmer (1981, p. 525) propose that sheep were
domesticated in Central Anatolia, and Hilzheimer (1936), Zeuner (1963),
and Ryder (1982, 1986) have argued that sheep and goats may have
originated in Central Asia. Meadow (1984, p. 324) finds sheep in the
earliest aceramic levels of Mergharh in Baluchistan, but both their absolute
ages and genus are in doubt. Clearly, the issue has not been settled but
present evidence suggests that sheep were much more restricted
geographically than were goats, and that the distributions of the two species
did not overlap much, if at all.
By contrast, there is evidence of goats throughout the mountainous
regions of southwest Asia and eastward into central Asia dating well back
into the Pleistocene, a reflection of their ability to withstand much more
rugged topography and poorer forage, and perhaps an indication that they
were easy to hunt. As browsers they are able to thrive on a variety of foods,
but their density in the Levantine region must have been quite low and the
Capra ibex of the southern Levant was apparently never domesticated.

Concluding remarks

My model is based on evidence of variable quantity and quality, which

presently suggests that plant domestication was essentially a lowland
adaptation that had a long period of technological development, beginning
with grinding–stones from some 18 000 years ago. The species that were
eventually domesticated are native to the grassy steppes and lightly forested
regions. Potentially, adaptations of this type could be found on both sides of
the Mesopotamian plain where rainfall or surface runoff stimulated the
growth of the relevant species. However, present archaeological and
cytological evidence suggests that the development was much older in the
Levant and along the Nile than in Mesopotamia, which it may have reached
through slow diffusion.
In contrast, we find no evidence of the use of domesticated livestock
before about 9000–10 000 cal. BC and it occurs in a narrow zone along the
front of the Zagros mountains in seasonal camps where caprine herders also
engaged in some form of food collecting or cultivation. The picture with
animals is more complicated than with plants in that the four major species:
goats, sheep, cattle, and pigs, were probably domesticated separately and at
different times by different processes. The reasons for this separation may
have to do both with environmental distributions of the animals and with
their different behaviours. The earliest species, sheep and goats, have habits
that place them seasonally in the same locales as the ripening plants that
people wished to harvest, so that propinquity was an important factor.
However, the fact that these species readily lent themselves to herding and
to multiple human uses, as contrasted with gazelles or onagers that also
moved with the ripening plants, may have had more to do with the
domestication than their mere presence in the environment.
In the future, as well as looking into plant use during the late Pleistocene
as Moore has suggested, it will be important to acquire much more precise
data on the changing vegetation of the north Mesopotamian plain, and to
seek possible herding camps as well as seed-processing camps in this
lowland steppe.

References
Bar-Yosef, O. 1980. Prehistory of the Levant. Annual Review of Anthropology 9, 101–33.
Braidwood, R. J. & B. Howe 1961.Prehistoric investigations in Iraqi Kurdistan. The University of
Chicago Oriental Institute, Studies in Ancient Oriental Civilization, no. 31, Chicago: The
University of Chicago Press.
Davis, S. J. M. 1982. Climatic change and the advent of domestication: the succession of ruminant
artiodactyls in the late Pleistocene-Holocene in the Israel region. Paleorient 8(2), 5–15.
Ducos, P. 1978.Tell Mureybet étude archéozoologique et problèmes d’écologie humaine, Vol. 1.
Lyon: CNRS.
Ducos, P. & D. Helmer 1981. Le point actuel sur l’apparition de la domestication dans le Levant. In
Préhistoire du Levant, J. Cauvin & P. Sanlaville (eds), Actes du Colloque Internationale no. 598,
523–8. Paris: Edition du Centre de la Récherche Scientifique.
Evins, M. A. 1982. The fauna from Shanidar Cave: Mousterian wild goat exploitation in
Northeastern Iraq.Paleorient 8, 37–58.
Hilzheimer, J. 1936. Sheep.Antiquity 10, 195–206.
Hole, F. 1984. A reassessment of the Neolithic Revolution. Paleorient 10, 49–60.
Hole, F. in press. Chronologies in the Iranian Neolithic. Colloque International du CNRS,
Chronologies relatives et Chronologieabsolu dans le Proche Orient de 16000â 4000 B P. Lyon.
Hole, F. & K. V. Flannery 1967. The prehistory of Southwestern Iran: a preliminary report.
Proceedings of the Prehistoric Society 33, 147–206.
Howe, B. 1983. Karim Shahir. In Prehistoric archeology along the Zagros flanks, L. S. Braidwood,
R. J. Braidwood, B. Howe, C. A. Reed & P. J. Watson (eds), Oriental Institute Publications, Vol.
105, 23–154. Chicago: Oriental Institute of the University of Chicago.
Lawrence, B. 1980. Evidences of animal domestication at Qayönü. In Prehistoric research in
southeastern Anatolia, H. Cambel & R. J. Braidwood (eds), 285–308. Istanbul: Istanbul
University, Faculty of Letters No. 2589.
Legge, A. J. 1975. The fauna of Tell Abu Hureyra: preliminary analysis. In The excavation of Tell
Abu Hureyra: a preliminary report, A. M. T. Moore. Proceedings of the Prehistoric Society 41, 74–
7.
Meadow, R. H. 1984. Animal domestication in the Middle East: a view from the eastern margin. In
Animals and archaeology. Vol. 3:Early herders and their flocks, J. Clutton-Brock & C. Grigson
(eds), 309–37. Oxford: BAR International Series 202.
Moore, A. M. T. 1982. Agricultural origins in the Near East – a model for the 1980s.World
Archaeology 14, 224–36.
Moore, A. M. T. 1985. The development of Neolithic societies in the Near East. Advances in World
Archaeology 4, 1—69.
Moore, A., J. Gowlett, R. Hedges, G. Hillman, A.Legge & P. Rowley-Conwy 1986. Radiocarbon
accelerator (AMS) dates for the epipalaeolithic settlement at Abu Hureyra, Syria. Radiocarbon 28,
1068–76.
Payne, S. 1983. The animal bones from the 1974 excavations at Douara Cave. In Paleolithic site of
Douara Cave and paleogeography of Palmyra Basin, Syria, K. Hanihara & T. Akazawa (eds),
Museum of the University of Tokyo, Bulletin 21, 1–108. Tokyo: University of Tokyo.
Ryder, M. L. 1982. Sheep-Hilzheimer 45 years on. Antiquity 56, 15–23.
Ryder, M. L. 1983.Sheep and man. London: Duckworth.
Turnbull, P. F. & C. A. Reed 1974. The fauna from the terminal Pleistocene of Palegawra
Cave.Fieldiana Anthropology 63(3), 81–146.
Valdez, R., C. F. Nadler & T. D. Bunch 1978. Evolution of wild sheep in Iran. Evolution 32, 56–72.
Van Zeist, W. & S. Bottema 1982. Vegetational history of the eastern Mediterranean and the Near
East during the last 20,000 years. In Palaeoclimates, palaeoenvironments and human communities
in the eastern Mediterranean region in later prehistory, J. L. Bintliff & W. Van Zeist (eds), 277–
319. Oxford: BAR International Series 133.
Zeuner, F. E. 1963.A history of domesticated animals. London: Hutchinson.
11 The domestic horse of the pre-Ch'in
period in China
CHOW BEN-SHUN

Introduction

Somewhat more than 9000 years ago, certain human populations in China
changed their subsistence patterns from hunting to a more settled way of
life and began to cultivate plants and to domesticate animals. The earliest
Neolithic culture so far discovered in northern China is the ‘Cishan–
Peilikang culture’, of which the radiocarbon dates are around 8000 years
BP. Remains of Sus domesticus and Canis familiaris have been found at
both Cishan and Pei-li-kang (Chow Ben–Shun 1984). The excavations
show that animal husbandry and crop–raising occurred together. The pig
and the dog were the first animals to be domesticated in Neolithic China.
Pigs especially were of economic importance throughout the period, since
they could provide meat rapidly, ensuring a regular food supply for large
groups of people, and organic fertilizer for farming. As a supplier of meat,
the horse was of little importance in the earliest Neolithic cultures.
In the past three decades more than 7000 Neolithic sites have been found
on the mainland of China, and over 100 of them have been excavated.
Thousands of animal bone fragments have been found in the debris of these
sites, but there are almost no reliable reports of horses associated with the
early Neolithic culture in China. The earliest locality that has yielded horse
remains is the Yangshao culture (4800–3000 BC) Panpo Site at Sian, from
which two molars and a first phalanx are regarded by Li You–heng & Han
Defen (1959) as Equus przewalskii. Most investigators have suggested that
the domesticated horse was established during the Luangshan period (3000–
2300 BC), but even at this period osteological remains of horses are rare and
have been identified only as Equus sp., making it impossible to say if the
animals were really domesticated. Horse remains collected from sites of
Machayao culture in Kansu Province (which is a later regional culture, c.
3000 BC, that shares a common origin with the Yangshao) and from a
Chichia culture (2000 BC) cemetery at Chin–Wei–Chia in Yung–Ching
County, Kansu Province (Kansu Archaeological Team 1975) indicate that
the first domestication of horse took place around 2000 BC in northwestern
China. Before the early Shang Dynasty, around 1300 BC, there is no
archaeological evidence for domesticated horse in the Central Plain.

The horse of the Chinese Bronze Age

The Chinese Bronze Age, lasting for at least 1500 years from almost 2000
BC until the 3rd century BC, was a formative stage in Chinese civilization.
The use of the horse in the wars between states and in transport gained it a
special position in the exercise of political power. The horse became an
instrument of power in Eurasia from Europe in the west to China in the east
at almost the same time.
In the past half century, many Bronze-age chariot burials with horses
have been excavated in China, indicating the rapid progress of horse-
raising in this period. During the past few years the author has made an
intensive study of horse skeletons from Bronze-age sites. Nearly 100 horse
skeletons, dating from the Shang Dynasty Anyang phase (13th—11th
century BC), Western Chou Dynasty (11th–8th century BC), down to the
Spring and Autumn Period (770–746 BC), have been studied to trace the
development of Chinese horses, with special attention to the height at the
withers. Complete long bones of these horse skeletons were measured and
the sizes of their withers were calculated. The height of the withers of
domesticated horses of the Shang Dynasty Anyang phase, found in the
sacrificial pits at the northern part of Wuguan village, Yin Hsu, is 133–143
cm. As recent Przewalski’s horses had a withers–size of 134 cm on average,
it would appear that the late Shang horses were related to the Mongolian
wild horse. The horses from a Yen cemetery of the Western Chou Period at
Liulihe, near Beijing (The Joint Archaeological Team & the Municipal
Archaeological Team of Beijing 1984) reached a height of 135–146 cm.
Horses of the Spring and Autumn Period from chariot pits belonging to a
crown prince of the state of Kuo at Shang-ts’un-ling near the city of San-
Men Gorge, Honan Province (The Institute of Archaeology, Academia
Sinica 1959) grew to a withers height of 139–149 cm. Thus, the existing
osteological evidence suggests that in the Chinese Bronze Age, the height
of the horses gradually increased from ponies of 133 cm to medium-sized
horses of 149 cm. They retained throughout this time the characteristics of
the Przewalski’s horse, including relatively heavy heads, thick necks,
upright manes, short bodies, and relatively short legs, and can be regarded
as a homogeneous population. The artists of the Chinese Bronze Age
illustrated these features in their works, as, for instance, in the jade figure of
horses from the tomb of Lady Hao, Anyang period c. 1300–1030 BC
(Institute of Archaeology 1980), and the Western Chou bronze receptacle
(tsun) in the shape of a foal (11 th—8th century BC) unearthed at Meixian,
Shensi Province (Kuo Mo–jo 1957). Down to the Ch’in Period the Chinese
horse was clearly related to the Mongolian pony. The life-size terracotta
war-horses of the First Emperor of China, Ch’in Shih–huang–ti, discovered
at Lintong in Shaanxi Province, near the Emperor’s Mausoleum, indicate
that both chariot and saddle horses belonged to the same stocky breed,
characterized by compact bodies, short legs, and broad necks. After the
Western Han Dynasty the Chinese horses received some inflow of genes
from the Ferghana horses which were derived from the tarpan, and new and
larger breeds of graceful horses with small heads made their appearance.
The Eastern Han bronze horses, excavated from Kansu (The Kansu
Provincial Museum 1974), illustrate perfectly the graceful carriage of the
tarpan breed in which the horse is shown as flying with its hoof on a falcon.

References
Chow Ben-Shun 1984. Animal domestication in Neolithic China. In Animals and archaeology. Vol.
3:Early herders and their flocks, J. Clutton–Brock & C. Grigson (eds), 363–9. Oxford: BAR
International Series 202.
Institute of Archaeology, Academia Sinica 1959.The cemetery of the State of Kuo at Shang Tsun
Ling. Archaeological excavations at the Yellow River reservoirs, Report no. 3, 1–85. Peking:
Science Press.
Institute of Archaeology, Chinese Academy of Social Sciences 1980.Tomb of Lady Hao at Yirtxu in
Anyang. Beijing: Cultural Relics Publishing House.
Joint Archaeological Team of Institute of Archaeology, Chinese Academy of Social Sciences & The
Municipal Archaeological Team of Beijing 1984. Excavation of a Yan cemetery of the Western
Zhou Period in Liulihe in 1981–1983.Kaogu (Archaeology) 200,405–26.
Kansu Archaeological Team, Institute of Archaeology, Academia Sinica 1975. Excavation of Chichia
culture cemetery at Chin-Wei-Chia in Yung-ching County, Kansu Province.Kaogu Xuebao 43, 57–
 69.
Kansu Provincial Museum 1974. The Han Tomb at Lei-Tai in Wuwei County,Kansu Province. Kaogu
Xuebao 41, 87–110.
Kuo Mo-jo 1957. Notes on inscription of bronze vessels made by the Family Li.Kaogu Xuebao 16,
1–6.
Li You-heng & Han Defen 1959. Animal bones from Pien-Po Neolithic site near Sian.
Paleovertebrata et Paleoanthropologia 1, 173–88
12 Utilization of domestic animals in pre–
and protohistoric India
P. K. THOMAS

In India the history of domestication of animals goes as far back as 7000

years BP, from the Mesolithic cultural phases at Bagor, district Bhilwara in
Rajasthan (Misra 1973) and Adamgarh, district Hoshangabad in central
India (Joshi & Khare 1966). A series of radiocarbon dates is available from
Bagor, and the earliest date for the Mesolithic cultural phase is 6245 ± 200
BP, TF 786 (Agrawal et al 1971). At Adamgarh the two enigmatic dates are
2765 ± 105 BP, TF 116, from 1.90 m below the surface on charred bones,
and 7240 ± 125 BP, TF 120 on shells from 0.15–0.20 m depth (Agrawal &
Kusumgar 1968). The discrepancy in the 14C dates of Adamgarh is dealt
with in more detail elsewhere (Thomas 1975). Although these two sites
belong to the same cultural period, the composition of the fauna represented
in them varies considerably. At Bagor, sheep/goat has been identified as the
principal domestic animal (Thomas 1975, 1977); while at Adamgarh cattle
(Bos indicus), buffalo (Bubalus bubalis), Sheep (Ovis aries), goat (Capra
hircus), pig (Sus domesticus), and ass (Equus asinus) are reported as
domestic animals (Joshi 1968).
The humped Indian cattle (Bos indicus) appears to be the most
predominant domestic animal in the early cultures of India, with only a few
exceptions. The Mesolithic site at Bagor and the Late Jorwe phase at
Inamgaon, a Chalcolithic site in Maharashtra, have yielded more sheep/goat
bones (about 65 per cent of the total faunal assemblage) than those of cattle.
From all prehistoric levels, and even up to some of the early historic
cultural periods, cattle were killed for meat and were a major source of
subsistence for the early populations. The age group studies of the cattle
killed in a majority of archaeological sites suggest that young animals
around the age of three years were preferred for food. In order of preference
in the food economy, the next important domestic animals were sheep, goat,
pig, and buffalo. However, in some sites like Navdatoli (Chalcolithic) in
district Nimar in central India, instead of sheep and goat, pigs were the
second most important animals in the food economy (Clason 1977). Dog
(Canis familiaris) is associated with most of the prehistoric cultures and
may have served as watch animal for the settlement and domestic herd, and
also in the hunting pursuits of early man. However, a solitary evidence of
dog being killed for meat is reported in the Late Jorwe phase (c. 1000–700
BC) at Inamgaon (Thomas 1984a).

The role of domestic animals further increased in the economy of the

early populations. Alur identified ossified hock joints of cattle from the
Neolithic sites at Hallur, district Dharwar, and T. Narasipur, district Mysore,
both in the Karnataka State, and suggests that these animals must have been
used for heavy traction or draught purposes (Alur 1971a, 1971b). The
presence of a comparatively large number of aged bulls in sites like
Ramapuram (Chalcolithic) district Kurnool, Andhra Pradesh (Thomas
1981) and Veerapuram (Neolithic) district Kurnool, Andhra Pradesh
(Thomas 1984b) suggests that these animals were probably used in
agricultural operations, such as traction or even for running irrigation
devices. Also, the use of this animal in threshing the harvested crops is a
possibility, as it is practised even today in Indian villages. The evidence of
agriculture comes from the grains, cereals, and pulses excavated from sites,
and also the impressions of some of these on different materials. The
prehistoric agricultural tools made of wood may not have survived in the
Indian archaeological sites as they have not been reported from the vast
majority of sites. The strong antlers with definite shapes found in the
excavations must have also been used in agricultural operations. The large
storage jars and silos found in most of the Neolithic and Chalcolithic sites
may also add to the agricultural evidences. Along with warfare equipment,
the iron hoes, saddle querns, and pounders unearthed from the Vidarbha
megalithic sites such as Naikund (Deo 1982), Mahurjhari (Deo 1973),
Borgaon, and Bhagimohari (Deo in press), all in the Nagpur district of
Maharashtra, point out the agricultural activities of the early inhabitants.
Terracotta wheeled toy carts are reported from some of the protohistoric
sites in India. The earliest record is at Mohenjodaro and Harappa, dating
back to 2300 BC (Allchin & Allchin 1968). The cart from Mohenjodaro is
shown pulled by bullocks. Inamgaon, a Chalcolithic site in Maharashtra,
has yielded an engraving of a cart drawn by bullocks on a storage jar from
the Jorwe phase dated to 1400–1000 BC (Dhavalikar 1974, Sankalia 1974).
These representations suggest the use of cattle as draught animals from very
early times in India.
The cattle by-products must have been of great importance to the people.
There is no direct evidence for the consumption of milk until early historic
times. One of the earliest evidences for milking is depicted in a stone relief
at Mahabalipuram dated to about 7th–8th century AD (Zeuner 1963). That
cowdung was extensively used for plastering the floors and walls of the
early houses, and also as a fuel, is known from the prehistoric settlements.
Neolithic ‘ash-mounds’ consisting of cow dung have been reported at
Kupgal, district Bellary in Karnataka and Utnur, district Mahabubnagar in
Andhra Pradesh (Allchin & Allchin 1968). Cattle dung is the main source
of domestic cooking fuel even today in Indian villages (Harris 1966). This
must have also been used as a manure in agriculture. Another important
cattle by-product is hide, for which archaeological evidences are sparse.
The horse (Equus caballus) was introduced late into the early cultures of
India in an already domesticated form. It was during the Iron Age (c. 1000–
400 BC) that horse-breeding became prevalent (Thomas in press a). Horse
ornaments and equipment retrieved from the megalithic sites in the
Vidarbha region of Maharashtra suggest the use of this animal for riding. A
few scanty representations of horse are noticed in the late phases of the
Neolithic and Chalcolithic cultures (Nath 1968, Alur 1971a, Thomas 1984a)
which may tally chronologically with the Iron Age cultures.
The close man-animal relationship can be further traced from the various
activities conducted by man involving his domestic stock. Terracotta
figurines of humped cattle are reported from a majority of protohistoric
sites, and these animals are also depicted on pottery and rock–shelters. At
present it may not be appropriate to suggest that these animals were
worshipped. One thing that goes against the worship is the profuse killing
of these animals in the pre- and protohistoric periods. However, the
depiction itself suggests the significance of these animals to the early
human populations.
Food offered to the dead is an age-old custom in Indian history. At
Inamgaon, bones of cattle, sheep, and goat are found along with urn and
extended burials. Such offerings are also reported from a number of other
sites. Animal sacrifices were also common in the early cultures. The
important sacrificed animals were cattle, sheep, goat, dog, and horse. At
Burzahom, a Neolithic site in Kashmir, dog/wolf heads were buried with a
human dead body (Allchin & Allchin 1968). The Chalcolithic site,
Ramapuram, in Andhra Pradesh has yielded partial and complete burials of
goat in the human graves. In partial burials, skulls and lower parts of the
limb bones of goat are represented. In some cases complete burial of a goat
has also been reported (Thomas 1981).
The megalithic stone circles in the Vidarbha region of Maharashtra,
belonging to the Iron Age period, have brought to light partial burials of
horses in the form of skulls and lower extremities of limb bones. A number
of megalithic stone circles are being excavated by S. B. Deo of Deccan
College, Pune, and animal remains from these graves and the habitation
sites are being studied by the present author. From these studies it has been
proved that horse sacrifice was not a regular custom, but it might have been
a status symbol among the Iron-age populations of Vidarbha. So far, the
only exception to the sacrifice of another animal in place of horse is in one
of the megalithic circles at Naikund, where cattle bones are identified
(Thomas in press b). From the recent excavations in the megalithic
habitations at Naikund and Bhagimohari (Thomas in press c) the bones of
horse with cut marks are reported. A similar find has been reported from the
Late Jorwe phase at Inamgaon. Probably after the sacrifice and subsequent
offerings to the dead persons, the rest of the body of the horse might have
been consumed by the inhabitants. Another interesting evidence of animal
sacrifice is found at Khanpur, a Harappan site in Gujarat, where a ventrally
pierced atlas vertebra (probably executed with a spear-like instrument) is
reported (Chitalwala & Thomas 1978). This probably suggests the sacrifice
of the animal, involving the collection of hot, spurting blood as in the case
of bull sacrifices performed in Crete (Fabri 1937).
Sharing of cattle meat between people of two different houses belonging
to the same settlement is found at Inamgaon (Thomas 1984a). Two equal
halves of a three-year-old bull are represented in two different houses. The
degree of charring and the metrical homogeneity of bones of the left and
right side of the body have been studied in detail. Probably on some
occasion people of these two different houses shared this animal.
 The economic outlook of the people and the varied uses of cattle might
have been the reasons for the abundance of cattle in the early cultures of
India. Man now felt a closer affinity and emotional feeling towards the
animal which provided almost everything for him. Simultaneously, a taboo
was introduced on the slaughter of cattle and this animal procured a place in
the Hindu religion. An exact date for the ban on cow slaughter cannot be
postulated at present; however, this development in India must have taken
place during Buddhism and Jainism in the beginning of the Christian era.

References
Agrawal, D. P. & S. Kusumgar 1968. Tata Institute Radiocarbon Date list V. Radiocarbon 10, 131–
43.
Agrawal, D. P., S. K. Gupta & S. Kusumgar 1971. Tata Institute Radiocarbon Date list IX.
Radiocarbon 13, 442–9.
Allchin, B. & R. Allchin 1968.The birth of Indian civilization. London: Penguin.
Alur, K. R. 1971a. Skeletal remains. In Protohistoric cultures of Tungabhadra valley (a report on
Hallur excavation), M. S. N. Rao (ed.), 107–24. Dharwar: M. S. Nagaraja Rao.
Alur, K. R. 1971b. Report on the animal remains. In Report on the excavation at T. Narasipur, M.
Seshadri (ed.), 19–104. Mysore: Department of Archaeology, Government of Karnataka.
Chitalwala, Y. M. & P. K. Thomas 1978. Faunal remains from Khanpur and their bearing on culture,
economy and environment.Bulletin Deccan College Research Institute 38, 11–14.
Clason, A. T. 1977. Wild and domestic animals in prehistoric and early historic India. The Eastern
Anthropologist 30(3), 241–89.
Deo, S. B. 1973.Mahurjhari excavations. Nagpur: Nagpur University.
Deo, S. B. 1982.Excavations at Naikund (1977–78: 79–80). Bombay: Government Press.
Deo, S. B. in press. Excavations at Borgaon and Bhagimohari. Bombay: Government Press.
Dhavalikar, M. K. 1974. Subsistence pattern of an early farming community of western India.
Puratattva 77, 39–56.
Fabri, C. L. 1937. The Cretan bull–grappling sports and bull sacrifice in the Indus Valley civilization.
In Archaeological survey of India Annual Report, 1934–35, J. F. Blakistan (ed.), 93–101. Delhi:
Manager of publications.
Harris, M. 1966. The cultural ecology of India’s sacred cattle. Current Anthropology 7(1), 51–66.
Joshi, R. V. 1968. Late Mesolithic culture in Central India. In La préhistoire: problèmes ettendances,
F. Bordes & D. de Sonneville Bordes (eds.), 245–54. Paris.
Joshi, R. V. & M. D. Khare 1966. Microlithic bearing deposits of Adamgarh rock- shelters. In Studies
in prehistory, Robert Bruce Foote memorial volume, D. Sen & A.K. Gosh (eds.), 90–5. Calcutta:
Firme K. L. Makhopadhyay.
Misra, V. N. 1973. Bagor – a late Mesolithic settlement in north-west India. World Archaeology 5(1),
92–110.
Nath, B. 1968. Advances in the study of prehistoric and ancient animal remains in India, a
review.Records of the zoological survey of India 61(1 & 2), 1–63.
Sankalia, H. D. 1974.Prehistory and protohistory of India and Pakistan. Poona: Deccan College
Postgraduate and Research Institute.
Thomas, P. K. 1975. Role of animals in the food economy of the Mesolithic culture of western and
central India. In Archaeozoological studies, A. T. Clason (ed.), 322–8. Amsterdam: North Holland
Publishing Company.
Thomas, P. K. 1977.Archaeozoological aspects of the prehistoric cultures of western India.
Unpublished PhD dissertation, Poona University.
Thomas, P. K. 1984a. The faunal background of the chalcolithic culture of western India. In Animals
and archaeology. Vol. 3, J. Clutton & C. Grigson (eds), 355–61. Oxford: BAR 202.
Thomas, P. K. 1984b. Faunal assemblage of Veerapuram. In Veerapuram: a type site for cultural
study in the Krishna Valley, T. V. G. Sastri, M. Kasturi Bai & J. Vara Prasad Rao (eds), Appendix
C, i-xi. Hyderabad: Bhavani Printers.
Thomas, P. K. in press a.Subsistence and burial practice based on animal remains at Khairwada. In
Excavations at Khairwada , S. B. Deo (ed.). Nagpur: Government Press.
Thomas, P. K. in press b. Animal remains at Naikund. In Excavations in Vidarbha, S. B. Deo (ed.).
Nagpur: Government Press.
Thomas, P. K. in press c. Faunal background of the megalithic habitation at Bhagimohari. In
Excavations at Bhagimohari, S. B. Deo (ed.). Nagpur: Government Press.
Zeuner, F. E. 1963.A history of domesticated animals. London: Hutchinson.
PASTORALISM
Introduction to pastoralism
JULIET CLUTTON-BROCK

Patterns of subsistence based on hunting, herding, farming, pastoralism, and

nomadism all depend on the exploitation of herd animals and were all
established during the prehistoric period yet they still survive at the present
day. The cultural and environmental determinants of these social systems
and the transitions between them have intrigued anthropologists (e.g.
Evans-Pritchard 1940) since the beginning of this century, but it has only
been within the past decade or so that comparative studies have been
carried out on pastoral economies. Notable amongst these have been the
publication (in English translation) of the work on pastoral nomadism by
Khazanov (1984), and the reviews of Ingold (1980, 1984).
As discussed in Chapter 8 by Meadow, it might be logical to assume that
there could be a continuity from the hunting of wild animals to the
following of herds, and hence to pastoralism. Khazanov (1984) and other
authors claim, however, that this has seldom, if ever, occurred. As discussed
by Ingold (1980), the pressures induced by human social systems have
resulted in a sequence of cultural changes that have followed the same
pattern throughout Eurasia. This began with broad-spectrum hunting
towards the end of the Pleistocene, which was replaced by dependence on a
few species of large mammals as resources diminished and the human
population increased. Change of climate, over-hunting, or the immigration
of agriculturalists then resulted in settlement and the cultivation of plants.
The generally accepted thesis is that it is only after agriculture and animal
husbandry became well established that pastoral nomadism developed as a
social system in Europe and Asia. However, as discussed later, plant
cultivation as an essential intermediate stage between hunting and a pastoral
economy is contended for southern Africa and South America, as well as,
with perhaps less evidence, for the reindeer-herders of northern Europe.
Khazanov (1984) maintains that the sources of pastoral nomadism in the
Old World are now clear and that they can be directly linked to a food-
producing economy. He holds, as is now the general belief, that only people
who were leading a relatively sedentary way of life and who had surpluses
of vegetable food could domesticate hoofed animals. These food-producing
economies spread from their core areas into new habitats, some of which
could support settled husbandry but some, especially in arid regions, could
support only mobile pastoralism. The work of Khazanov has been centred
on the East European steppes, and, indeed, the basis for some of his ideas
has been provided by the work of Shilov who discusses nomadism in
Chapter 13. Zarins, in Chapter 14, presents archaeological evidence for the
beginnings of pastoralism in Arabia, which was one outward extension
from the Near Eastern centre of domestication, and he shows how, as the
climate became progressively more arid, sheep and goats were replaced by
camel-herding. Dhavalikar, with a similar theme but from a later period, in
Chapter 15, discusses how successive droughts and famines forced the
settled communities of the Deccan plateau of India to resort to pastoral
nomadism in order to survive.
The traditions of herd management of reindeer in Lapland, together with
some fascinating anecdotes, are described in Chapter 16 by Aikio. While
accepting that the general view of reindeer-herding is that it is of relatively
recent origin, perhaps dating to the end of the Middle Ages, Aikio
personally believes that its beginnings are far more ancient and probably go
back to the end of the Ice Age.
In Chapter 17, Tani describes his exhaustive study of how the shepherds
of the transhumant flocks of sheep and goats in the Mediterranean countries
train certain animals to lead the flocks. These flock-leaders may or may not
be castrated and Tani discusses the origins and functions of castration in this
context as well as drawing parallels with the castration of human males in
the ancient world.
Khazanov (1984) discussed only briefly the origins of pastoralism in
Africa and, indeed, they are only just beginning to be investigated. Cattle-
herding in North and East Africa is discussed in Chapters 18 and 19.
Whereas sheep and goats were undoubtedly brought into Africa from
western Asia, there does seem to be some slight evidence to suggest that
cattle may have been locally domesticated in North Africa (Clutton-Brock
in Ch. 18). Robertshaw, in Chapter 19, gives the evidence for the first
pastoralism in East Africa as being found from the late 3rd millennium BC.
From the diversity of animal remains from archaeological sites and
ethnographic accounts, he discusses the relationships between hunters and
herders, attitudes of cultural superiority by pastoralists, and the question of
stock-thieving by hunters; a topic that has also been discussed by Davidson
in Chapter 6 and mentioned by Clutton-Brock in the Introduction to Section
I of this book.
The insistence by Khazanov (1984) and other anthropologists that
pastoralism can only develop from a society that practises agriculture may
not apply in Africa, perhaps because there was no shortage of wild plant
foods for both animals and people. Robertshaw claims that the pastoralists
of East Africa were reluctant cultivators, and the evidence from the western
Cape of South Africa indicates that the aboriginal Khoi (Hottentots) were
never settled and never cultivated plants, with the exception of a narcotic
(Klein 1986). The Khoi were originally hunter- gatherers who adopted,
sometime after 2000 years ago, the sheep and cattle that had moved south
through the continent with migrating people over the preceding millennium.
Again, as discussed by Klein (1986) the Khoi may have obtained their cattle
by raiding from Bantu immigrants. They used oxen as draught animals and
sheep to provide milk and meat (Smith 1986).
In Chapter 20 Galaty describes the systems used by the Maasai for the
naming and classification of their cattle. The mastery of pastoral cognition
is gained through learning and experience which are today inevitably
decreasing as young people turn towards school education.
The remaining four chapters in this section of the book are about camelid
pastoralism in South America, a subject that has been little investigated
until very recently. Chapters 21 and 22 by McGreevy and Brotherston,
respectively, discuss the role of camelids before the Spanish Conquest and,
in particular, the importance of the llama in Inca society. Browman, in
Chapter 23, presents a review of the earliest evidence for camelid
exploitation and the shift from hunting to pastoralism, which does appear to
have preceded the cultivation of plants. These three chapters demonstrate
how modern data on animal distributions and behaviour, literary sources,
and the investigation of animal remains from archaeological sites can be
combined to elucidate the history of ancient social systems and
palaeoeconomies.
 Finally, in these studies on camelid pastoralists, Rabey in Chapter 24
describes fieldwork that he has carried out amongst llama-herders in the
hills and plains of the south central Andes. With the hill herdsmen Rabey
describes what could be the most primitive form of pastoralism, which is
merely an extension of hunting with very little control or ownership of the
animals. If this is indeed a relic of an anciently established system of
exploitation, it appears to support Aikio (Ch. 16) and to belie the theories of
Khazanov (1984) that pastoralism only develops from true domestication.
On the other hand, like much modern reindeer nomadism in northern
Eurasia, this form of herd-following may involve the secondary use of feral
animals, which were anciently domesticated, within a modern hunting
economy, as discussed by Ingold (1980, p. 123).
At this point it may be helpful to explain some of the terms used in
studies of pastoralism, using the definitions of Ingold (1980) and Khazanov
(1984):

Hunters are food-extractors who are only interested in the dead animal, as
discussed by Meadow in Chapter 8. Like other carnivores, the hunter
interacts with prey only when it is about to be killed. All other social
systems of food-producing protect the living animal until it is ready for
slaughter.

Herd-following may apply to a human population that ranges over the same
area in its annual cycle as the animal population, or it may apply to
particular humans that are associated with particular herds of animals, in
which case it is equivalent to ranching.

The rancher loosely owns herds of animals for exploitation of meat and
other resources that are often marketed. In origin the animals may be wild,
feral, or domestic but they live as wild animals except that their territory is
usually restricted.

Nomads may be wandering hunters and gatherers or mobile pastoralists.

However, the reasons for mobility in these two groups are so different that
Khazanov considers that hunter-gatherers should be termed ‘wandering’
and the term ‘nomadism’ should be reserved for mobile pastoralists.
Pastoralists are divided by Khazanov into a number of different categories
within two main groups. Pastoral nomadism proper, which is characterized
by the absence of agriculture and is exemplified by, say, the pastoralists of
the Sahara, or the ‘pure pastoralists’ of McGreevy (Ch. 21), and semi-
nomadic pastoralism in which there is periodic changing of the pastures
during the year but the cultivation of crops is also practised. This is the
most common form of pastoralism.

Transhumants are agriculturalists living in the Mediterranean and southern

Europe who move their livestock between mountain and lowland pastures.

These somewhat simplistic definitions are given here to clarify some of the
terms used in this section of The walking larder: for further discussion see
Ingold (1980, 1984) and Khazanov (1984). The intention of the chapters on
pastoralism is to present a picture of the complicated and fascinating
relationships between herders and their flocks from many different parts of
the world in the past and at the present day.

References
Evans-Pritchard, E. E. 1940. The Nuer. Oxford: Clarendon Press.
Ingold, T. 1980. Hunters, pastoralists and ranchers. Cambridge: Cambridge University Press.
Ingold, T. 1984. Time, social relationships and the exploitation of animals: anthropological
reflections on prehistory. In Animals and archaeology. Vol. 3: Early herders and their flocks, J.
Clutton-Brock & C. Grigson (eds), 3–12. Oxford: BAR International Series 202.
Khazanov, A. M. 1984. Nomads and the outside world (translated by J. Crookenden). Cambridge:
Cambridge University Press
Klein,R. G. 1986. The prehistory of stone-age herders in the Cape province of South Africa. South
African Archaeological Society Goodwin Series 5, 5–12.
Smith, A. B. 1986. Competition, conflict and clientship: Khoi and San relationships in the western
Cape. South African Archaeological Society Goodwin Series 5, 36–41.
13 The origins of migration and animal
husbandry in the steppes of eastern
Europe1
VALENTIN PAVLOVICH SHILOV
(translated by Katharine Judelson)*

There are two points of view on the origins of nomadic animal husbandry in
the literature. Some researchers (Hahn 1886, pp. 132–3, 1891, p. 484, 1905,
pp. 96, 99–100, Golmsten 1933, p. 89, Gryaznov 1955, p. 24, 1957, p. 2,
Chernikov 1957, pp. 31–2, Rudenko 1961, p. 2) consider that nomads
emerged as a separate group from settled communities engaged in both
arable farming and animal husbandry at the end of the 2nd and the
beginning of the 1st millennium BC. Others (Rousseau 1775, Smith 1776,
Jselin 1786, Schmidt 1915–16, pp. 593, 610, 1924, p. 193, 1951, pp. 213–
17) are of the opinion that nomadic animal husbandry emerged earlier, at
the time of hunting tribes. For the development of the nomadic economy,
there are also two points of view: first that it came into being suddenly
(Gryaznov 1957), and secondly that it took shape over several centuries
(Smirnov, K. F. 1964, pp. 45–7, Smirnov, A. P. 1966, p. 13).
What objective data do we have to resolve these questions? First of all let
us consider evidence from written sources. In the 7th and 6th centuries BC
the territory to the north of the Black Sea was colonized by the Greeks, who
built a dense network of towns, the inhabitants of which established close
links with the local people. From these colonists there have come down to
us quite valuable, albeit fragmentary, observations concerning life in this
territory, which provided the original source for the works of Greek and
Roman historians. The Greek colonists and travellers from this area were
struck first and foremost by the, for them, unusual nomadic way of life in
the steppe, where families and their herds moved from one place to another
‘depending upon where they could find an abundance of grass and water’.
The poet Hesiod (Strabo 1st century AD, p. 7) who lived in the 8th century
BC, pointed out in his Theogony that the Scythians milked horses. Homer,
whose works are usually held to date from between the 10th and 8th
centuries BC, referred to the ‘amazing horse- milkers’ in the steppes
bordering on the Black Sea, who milked horses and drank their milk
(Homer 1936, p. 355). It can also be assumed that the Cimmerites, who
settled the territory adjoining the Black Sea before the arrival of the
Scythians, were already nomads, since they succeeded in carrying out long
expeditions, for instance as far as Asia Minor. Finally, the fact that the
Scythians in the 7th and 6th centuries BC engaged in a highly developed
form of nomadic animal husbandry shows that it must have existed long
before that time. The study of archaeological monuments, in particular that
of Scythian burial mounds, has revealed marked property differentiation.
Side by side with graves for the poorest strata of the population can be
found rich graves, the furniture of which includes costly gold ornaments
(the Litoi burial mound, the burial site on the Kalitva River, the
Kelermessky burial mounds, etc.), and also graves where together with the
deceased were buried hundreds of horses (the Ulsky burial mound and
others). Furthermore, the burial sites of the Scythian nobility were
enormous monuments, the erection of which demanded gigantic manpower
resources. This already reflects a complex picture of the social life of the
nomadic people, and does not in any way point to merely preliminary steps
towards the development of nomadic animal husbandry, but rather to an
advanced stage in that development, when tremendous wealth, in the form
of livestock and expensive luxury objects in gold, was concentrated in the
hands of the leaders.
Another early emergence of nomadic animal husbandry is to be observed
in desert regions, where it can still be found nowadays: for instance, the
ancient population to the west of the Nile was leading a nomadic life in the
neighbouring deserts as early as the 4th century BC. Egyptian hieroglyphic
inscriptions inform us as to the large numbers of livestock the Egyptians
used to obtain from the Libyan deserts, where the people were engaged in
nomadic animal husbandry (Struve 1941, p. 146).
A typical example of a nomadic people was that of the Hyksos who
conquered Egypt in 1710 BC (Struve 1941, pp. 169–70). In the Bible it is
written (Chronicles I v. 39–41) that the tribe of Judah went ‘to the entrance
of Gedor, even unto the east side of the valley, to seek pasture for their
flocks. And they found fat pasture and good, and the land was wide, and
quiet, and peaceable; for they of Ham had dwelt there of old. And these
written by name came … and smote their tents, and the habitations that
were found there, and destroyed them utterly unto this day and dwelt in
their rooms: because there was pasture there for their flocks.’
The date at which herdsmen first moved into the open steppe is a
question that my colleagues and I have studied in the examination of close
on 800 burial sites in the open steppe dating from various periods. In
addition, palaeogeographers and soil scientists working with us have
demonstrated, on the basis of their study of deposits in the lakes of the
steppes and the soils lying beneath the burial mounds that have been
excavated, that the climate and the steppe vegetation found in the Bronze
Age in the steppes of eastern Europe have changed little since then up to the
present day, although there may have been fluctuations in aridity
(Chuguryaeva 1960, p. 282, Neishtadt 1967, pp. 198–9). In 478 Bronze- age
burial sites the animal remains were distributed as shown in Table 13.1. In
so far as there were found in these burial sites the remains of men, women,
children, and infants we can assume that at the time of the Pit- grave culture
there were whole families living in the steppes. It was formerly held that the
people of the Pit-grave culture were hunters and gatherers. Only more
recently has the hypothesis emerged that they had a producer economy
(Latynin 1957, p. 31, Merpert 1961, Lagodovskaya et al. 1962, p. 178). In
40 of the 263 burial sites (15.2 per cent) investigated in the steppes of the
area between the Don, Volga and Ural rivers, bones of domestic animals,
mainly sheep, were found.

Table 13.1 The remains of domestic animals collected from burial sites of Bronze-age cultures in the
lower reaches of the Volga valley
 The custom of placing meat in graves testifies undoubtedly to the fact
that animal husbandry was widespread in the valley of the lower reaches of
the Volga. In the Bronze Age the animals bred were for the most part sheep,
cows, horses, and camels. Dogs were also kept and these were at that time
buried with the humans.
Animal husbandry, hunting, and fishing were the principal activities in
the economy of the Bronze-age population in the lower part of the Volga
valley. Moreover, the role of animal husbandry increased considerably
between the age of the Pit-grave culture (47 per cent) and the Chamber
culture (94.1 per cent).
Sheep and goats predominated in the herds of the Bronze-age population
in the lower part of the Volga valley, an arid land with saline and sandy
soils. In such areas the sheep and goats would graze in the plains during the
winter and in the mountains in the summer.
In pastures where a mixture of grasses predominated there would be
more horses and cows in the herds. When the animal husbandry included
different kinds of animals then the best winter pastures were those where
both steppe and meadow vegetation were to be found, i.e. where there
would be fodder both for sheep and camels, on the one hand, and for horses
and cattle on the other (Geins 1897, pp. 60–1).
Therefore, on the basis of analysis of the bones found in burial sites, it
can be said that in the economy of the Pit-grave culture and subsequent
cultures animal husbandry was the predominant form of economic activity.
Hunting and fishing did not play a prominent part. Up until the present time
it has not proved possible to find permanent settlements of the population of
the Pit-grave culture in the lower part of the Volga valley. So far over 400
settlements dating from the Neolithic Age, Eneolithic, and Early Bronze
Age have been recorded. They are situated in terrain unsuitable for
cultivation, in windswept sandhills (Rykov 1928, pp. 20, 23–4, 26, Minaeva
1929, p. I, Sinitsyn 1931, pp. 81–2, 1933, p. 89, 1948, p. 151–60,
Filipchenko & Kurochkin 1960, p. 272ff.).
It is also impossible to link bone remains with specific cultural horizons,
in so far as they are encountered in a dispersed cultural stratum. This was
borne out by archaeological excavations in the Ryn-Sands near Astrakhan
undertaken by the Leningrad Division of the Institute of Archaeology under
the USSR Academy of Sciences led by A. N. Melentiev. At that site
archaeologists have found dozens of settlements between sandhills in
terrain unsuitable for cultivation. In these settlements pottery and
microlithic tools from the Neolithic, Aneolithic, and Bronze Ages were
found. These settlements resemble in character temporary camps for
herdsmen, rather than sites of permanent habitation. Approximately 300
wooden carts have also been discovered by the members of this expedition.
Thus, the archaeological finds from the steppes of the lower Volga valley
testify to the nomadic way of life of the people from the Pit-grave,
Poltavkin, and North-Caucasian cultures. This was determined by the
natural geographical habitat, which did not favour extensive land-
cultivation. There are large tracts of saline land containing many obstacles
in the form of sandhills, between which there were temporary camps, but no
permanent settlements.
Despite systematic research, nothing has been found so far which would
confirm substantial development of land cultivation before the age of the
Chamber culture. Long-term settlements in the lower Volga valley are in
those areas where chernozem (black earth) predominated, in the northern
part of the Great Irgiz Basin and further north.
Many researchers consider that the predominance of sheep and goats in
herds testifies to a nomadic way of life (Herre 1949, p. 317). This is also
borne out by the small number of graves in the burial grounds of the Pit-
grave culture age (see Table 13.1), and also probably by the emergence of
the custom of making burial mounds over the graves. In the monuments of
the Pit-grave and North-Caucasian cultures covered carts have been found
which constitute houses on wheels (Sinitsyn 1948, pp. 151–60, Kaposhina
1962, pp. 41, 48, Sinitsyn & Erdniev 1963, p. 14, 1966, pp. 32–5).
Ethnological data indicate that animal husbandry in the lower Volga
valley would not have been possible without migration. In order to supply
food for one Kalmyk group, or khoton, (consisting of four tents each with
13 people) it would have been necessary to keep 100 sheep, four cows, two
horses, and four camels. The leader of one such khoton from the Bogutov
Clan of the Yandyk nomad camp, or ulus, was considered the ‘last of the
wealthy men’, i.e. the owner of that minimum below which comes poverty.
Even given such a small quantity of animals, the khoton would have been
obliged to be on the move. A study of a 19th-century khoton revealed that
for about six weeks in the winter and spring the khoton stayed in the
Mashtyk terrain. In mid-May it moved from there to the Khosh terrain some
18 km away, where it spent five days. Here the water had a salty-bitter taste
and the khoton was obliged to move on to the Kiubedin-Ksentse terrain 10
km further on, where it spent 20 days. From there it advanced 30 km to the
Dabst terrain where it spent 15 days. After that it moved 5 km to the Per-
Marzlyk terrain where it spent 10 days and where it was encountered on 10
July by I. A. Zhitetsky (Zhitetsky 1892, PP. 95–7).
Thus the khoton, which began its migration in mid-May, changed
pastures five times and covered a distance of 73 km. Moreover, the animals
fed on the fodder to hand in the local pastures. No stocks of hay were laid
in.
A large-scale Kalmyk economy in 1885 kept 300 horses, 1500 sheep, 50
camels, and 60 head of cattle. The families of the chief, the herdsmen, and
the hands who saw to the wells lived in ten covered tents. This khoton
would spend 60 days in its winter pasture. From 15 February over a period
of 144 days it would move on four times and cover a distance of 80 km.
Moreover, after May all the animals, apart from the cattle, would move on a
further five times and cover a distance of 85 km. The camels and sheep
would have an extra move involving another 25 km. The cattle moved four
times and covered a distance of 80 km, the horses moved nine times over
109 km, the camels 10 times over 194 km, and the sheep ten times over 169
km. In May the khoton would divide into two groups, and during periods of
drought into four groups.
This ethnographic information demonstrates convincingly that in
conditions similar to those of the lower Volga valley it is impossible for
people to engage in animal husbandry as their main economic activity
without migration. The reasons for this are the shortages of grass and water.
In order to feed one cow or horse in the conditions found in the Lenin
District of the Volgograd Region, where the semi-desert steppeland is
covered with tipchak and wormwood, 4.9 ha of degraded pasture, or 8–9 ha
of very degraded pasture, is necessary (Terenozhkin 1937).
So archaeological and ethnographic evidence testifies that the peoples of
the ancient Pit-grave culture, and subsequent cultures found in the lower
Volga valley, led a nomadic way of life, breeding sheep and goats.
Nomadism did not prevail over the whole steppeland, however. Our
research in the lower part of the Dnieper valley has clearly shown that here
there was settlement on the open steppes in the age of the Pit-grave and
Kemiobinsk cultures (the Dmitrievsky burial ground and the Orlov burial
ground with burial mounds). A similar picture was observed in the northern
Caucasus from monuments of the Maikop culture, where bones of cattle
and pigs predominate. Bones of these animals in sites of the Maikop culture
in Meshoko and in the Ukrainian site of the Pit-grave culture at
Mikhailovka point to a settled way of life (Lagadovskaya et al. 1962, pp.
168, 207).
 Thus, it can be shown from study of the animal remains that in the areas
with saline soil in the lower Volga valley and in the open steppes to the
north of the Black Sea (the Dmitrievsky burial ground, the Orlov grave and
others) sheep were kept, together with some goats. In the mountains and the
foothills of the northern Caucasus, rich in alpine meadows and oak groves,
cattle and pigs were kept: in the Don feathergrass steppes (on the eastern
bank of the Don) horses were kept, and in the chernozem (black earth) areas
of the lower part of the Dnieper valley land cultivation and cattle-breeding
were practised. So during the time of the Pit-grave culture we can pick out
regions with all these specialized economies, the development of which was
determined by the natural and geographical environment. Finally, in the
steppes of southern Russia during the time of the Pit-grave culture (3rd
millennium BC) there appeared the first elements of nomadism, which,
according to the testimony of classical writers, followed a long and
complex path.

Note
1 Translation of an abbreviated version of Shilov (1985).

References
Chernikov, S. S. 1957. The role of the Andronov culture in the history of central Asia and
Kazakhstan [in Russian]. Kratkiye soobshcheniya instituta etnografii, Moscow 26, 31–2.
Chuguryaeva, V. A. 1960. Vegetation of the Trans-Volga region in the Bronze Age [in Russian].
Materialy i issledovaniya po arkheologii SSSR, Moscow 78, 282.
Filipchenko, V. A. & Y. V. Kurochkin 1960. Flint tools from the island of Kulaly [in Russian].
Sovietskaya Arkhaeologiya, Moscow 3, 272 ff.
Gcins, A. I. 1897. Collection of literary works, Vol. I, 60–1 [in Russian]. Saint Pctcrsberg.
Golmstcn, V. V. 1933. On the question of ancient animal husbandry in the USSR [in Russian].
Proiskozhdeniye domashnikh zhivotnykh, Moscow 1, 89.
Gryaznov, M. P. 1955. Aspects of the formation and development of early nomadic societies [in
Russian]. Kratkiye soobshchetiiya instituta etnografii, Moscow 24, 24.
Gryaznov, M. P. 1957. Stages of development in the economy of the animal- breeding tribes of
Kazakhstan and southern Siberia in the Bronze Age [in Russian]. Kratkiye soobshchetiiya instituta
etnografii, Moscow 26, 21.
Hahn, E. 1886. DieHaustiere und ihre Beziehungen zur Wirtschaft der Menschen, 132–3. Leipzig.
Hahn, E. 1891. Waren die Menschen der Urzeit zwischen Jagerstufe und stufe Ackerbaus Nomadem?
Das Ausland 64, 484.
Hahn, E. 1905. Das Alter der Wirtschaftskultur der Menschheit, 96, 99–100. Heidelberg.
Herre, V. 1949. Zur Abstammung und Entewicklung der Haustiere. Verhandlungen der deutschen
Zoologen in Kiel 1948, 317. Leipzig.
Homer 1936. Illiad, Vol. XIII, 5, 355. Moscow.
Isaak, E. 1971. On the domestication of cattle. In Prehistoric agriculture, 446. New York.
Jselin, I. 1786. Uber dieGeschichte der Menschheit, Vol. 2. Basel.
Kaposhina, S. I. 1962. Findings from the work of the Kobyakov Expedition [in Russian]. Kratkiye
soobshchetiiya instituta arkheoloqii An SSSR, Moscow-Leningrad 103, 48.
Lagodovskaya, O. V., A. G. Shaposhnikova & M. A. Makarevich 1962. The Mikhailovskoye
Settlement [in Russian], 178. Kiev.
Latynin, V. A. 1957. The level of development of productive forces in the Early Bronze Age [in
Russian]. Kratkiye soobshchetiiya instituta arkheologii, Moscow- Leningrad 70, 3.
Liberov, P. D. 1960. On the history of animal husbandry and hunting in the territory to the north of
the Black Sea [in Russian]. Materialy i issledovaniya po arkheologii SSSR, Moscow 50, 134, Table
6.
Merpert, N. Y. 1961. The Eneolithic Age in the steppe belt of the European part of the Soviet Union
[in Russian]. Lecture delivered at a symposium in Prague, issued as a separate offprint.
Minaeva, T. M. 1929. The flint industry in the lower reaches of the Volga [in Russian]. Trudy Nizhne-
Volzhskogo oblastnogo nauchnogo obshchestva kraevedeniya, Saratov p. 31.
Neishtadt, M. I. 1967. The history of forests and palaeogeography in the USSR in the Holocene
Epoch [in Russian]. 198–9. Moscow.
Rousseau, J. J. 1775. Le discours sur l’origine et les fondements de l’inégalitéparmi les hommes.
Paris.
Rudenko, S. I. 1961. On the question of forms of animal-breeding economies and nomads [in
Russian]. Materialy po etnografii, Leningrad 1, 2. Geographical Society, USSR.
Rykov, P. S. 1928. Archaeological surveys and excavations in the lower reaches of the Volga, carried
out in 1928. Izvestiya Saratovskogo Nizhne-Volzhskogo instituta kraevedeniya im. M. Gorkoqo,
Saratov IV, 20, 24, 26, 33.
Schmidt, W. 1915–16. Anthropos 593, 610.
Schmidt, W. 1924. Anthropos 193, 193.
Schmidt, W. 1951. Zu den Anfängen der Herdentierzucht. Zeitschrift für Ethnoqrafie 76, 213–17.
Shilov, V. P. 1964. Problems relating to the opening up of the steppes in the Bronze Age [in Russian].
Arkheologicheski sbortiik Gosudarstvennogo Ermitazha, Leningrad VI, 102.
Shilov V. P. 1985. Ancient finds in the Kalnyk area of the USSR. Elista, pp. 23–33.
Sinitsyn, I. V. 1931. Flint tools from the dune sites in the Kalmyk Region [in Russian]. Izvestiya
Saratovskogo Nizhne-Volzhskogo instituta kraevedeniya im. M. Gorkogo, Saratov IV, 81–92.
Sinitsyn, I. V. 1933. Ancient monuments of the coastal area of the Kalmy Region [in Russian].
Izvestiya Saratovskogo Nizhne-Volzhskogo instituta kraevedeniya im. M. Gorkogo,Saratov IV, 89.
Sinitsyn, I. V. 1948. Monuments of the Pre-Scythian Epoch in the lower Volga steppes [in Russian],
Sovietskaya Arkheologiya, Moscow-Leningrad 10, 151–60.
Sinitsyn, I. V. & U. E. Erdniev 1963. Archaeological excavations in the Kalmyk ASSR in 1961 [in
Russian]. Trudy Kahnytskogo respublikanskogo kraevedcheskogo muzeya, Elista XVII, 14.
Sinitsyn, I. V. & V. E. Erdniev 1966. New archaeological monuments in the territory of the Kalmyk
ASSR [in Russian]. Trudy Kalmytskogo nauchnoiss- ledovatelskogo instituta yazyka, literatury i
istorii i Kalmytskogo respublikanskogo kraevedcheskogo muzeya, Elista II, 32–5.
Smith, A. 1776. An inquiry into the nature and causes of the wealth of nations, Vols. I & II. London.
Smirnov, A. P. 1966. The Scythians [in Russian], 13. Moscow.
Smirnov, K. F. 1964. The production and nature of the economy of the Early Sarmatians [in Russian],
Sovietskaya Arkheologiya, 3, 45–7.
Strabo, 1st century AD. S VII, 3, 7.
Struve, V. V. 1941. History of the East in ancient times [in Russian], 146. Moscow.
Terenozhkin, I. I. 1937. Wild pastures and hay-yielding grasses in the Stalingrad Region [in
Russian]. Stalingrad.
Zhitetsky, I. A. 1892. Astrakhan Kalmyks [in Russian]. Sbornik trudov chlenov Pctrovskogo
obshchestva issledovatelei Astrakhanshogo kraya, 95–7.

* Modern Languages Department, Totton Sixth Form College, Hampshire.

14 Pastoralism in southwest Asia: the
second millennium BC
JURIS ZARINS

Introduction

The question of pastoral nomadism in the Middle East has been increasingly
re-examined in the light of historical, archaeological, and ethnographic work
(Irons & N. Dyson-Hudson 1972, Rowton 1974, Dyson-Hudson & Dyson-
Hudson 1980, Eph’al 1981, Galaty & Salzman 1981, Lancaster 1981, Briant
1982, Khazanov 1984, Zarins 1988, Chang & Köster 1986). Archaeological
work within a defined arid zone in Southwest Asia over the past decade has
particularly emphasized the complex pastoral pattern which has emerged
since the 7th millennium BC (Fig. 14.1). This chapter seeks to shed light on a
particularly interesting period of this development, namely the pivotal 2nd
millennium BC. It is known that a basic pastoral way of life developed along
certain characteristic lines during the period 6000–1800 BC (Zarins 1988)
and that this society contrasted with a quite different type developed during
the period following 800 BC (Bulliet 1975, p. 76ff., Eph’al 1981, Briant
1982). The earlier stage was characterized by the reliance on domesticated
ovicaprids and bovids and the construction of stone structures which, by
their particular nature, define the associated culture to roughly 4
millenniums. The later culture is characterized by the reliance on
domesticated camels and much more ephemeral archaeological remains. The
principal region in question (Fig. 14.2) covers the northern part of the
Arabian peninsula and the adjoining Levant. The nature of the transition
period will be dealt with here in some detail. It covers the thousand years
from 1800–800 BC, which is the period between the end of the better-
documented Early Bronze Age and the beginning of the Iron Age throughout
much of the Middle East (Edens 1986).
The archaeological evidence for this millennium has long been thought to
be absent or poorly known, particularly in such specialized discussions as
the EB-MB transition in Palestine, the MB-LB Period of Transjordan, the
origins of camel nomadism, the history of Mesopotamia, the nature of the
Second Intermediate Period in Egypt, and other related concerns. Three lines
of evidence will be dealt with here. First, the historical documentation
includes the Middle Assyrian and Kassite cuneiform records and the
Egyptian hieroglyphic texts such as the Execration Texts of the late Middle
Kingdom, the New Kingdom Papyri Harris I, Anastasi VI, and the Tell el
Amarna correspondence. Secondly, we have increasing archaeological
evidence from Eastern Arabia/Bahrain, the greater Nejd, southern Arabia,
Midian (including southern Jordan), and the Negev/Sinai. Thirdly, the status
of camel domestication during this millennium will be reviewed.

Figure 14.1 Principal site localities in the Arabian peninsula and adjoining regions.

The early transition (1800–1400 BC)

The nature of human occupation within areas marginal to the Fertile
Crescent has only been examined within the past two decades. While our
knowledge of the ‘Agricultural Revolution’ has increased dramatically, we
are only now beginning to understand the probable course of human
adaptation in the arid areas south of the primary agricultural zone itself. In
the northern part of the Arabian peninsula the pattern of human occupation
can be divided into two phases: (a) the Early Pastoralist Phase, 6000–3500
BC; and (b) the Mature Phase, 3500–1900 BC. These pastoralists followed a
multi-resource food procurement strategy but principally depended on the
herding of ovicaprids and, to a lesser extent, cattle. The second phase saw
the introduction of equids, such as horses and donkeys, and increased
mobility. By the EB IV period (sensu lato) c. 2200–1900 BC, in the arid
Negev/Sinai, there is a definite change in the number of sites and population
density. Sites become more numerous and smaller and I suggest that this
basic shift may be correlated in large part with deteriorating climatic
conditions. Without exception, the archaeological sites of the technocomplex
‘disappear’ around 1900 BC never to be reoccupied. (For details of this
record, see Zarins 1988. Avner [1984, p. 119] has also commented on the
increase in standing pillars in settled communities only after the decline of
this typical desert feature in the early 2nd millennium BC.)
Figure 14.2 The early pastoral technocomplex (after Zarins 1988).
 This shift may be paralleled in the historic records of late Old
Kingdom/First Intermediate Period Egypt and the Ur III and I/L periods of
southern Mesopotamia (2100–1900 BC). Simpson (1971, pp. 238–40) states
that the Asian Bedouin occupied the Delta nomes at the end of the Old
Kingdom and the well-known Instructions of Merykare describes these
pastoral nomads rather vividly. The involvement of the Amorites in
redirecting the political fortunes of southern Mesopotamia is equally well
known (Edzard 1957). To what extent these pastoral nomads contributed to
bringing about the collapse of the EB civilization in Palestine is still being
debated (Prag 1984). Two exceptions to this overall picture need to be
mentioned here. First, the Amorite pastoral nomads of the Diyala region are
well attested in the Eshnunna documentation of the Ur III period down to c.
1750 BC, but these records have not been fully published (Whiting 1983). In
addition, understanding of the relationship between the pastoral nomads and
the established political system in the Diyala remains vague. Secondly, there
is voluminous data for the Mari nomads of the period 1825–1759 BC
(Küpper 1957, Luke 1965, Prag 1985) but they represent the latest example
of this way of life in the northern tip of the arid region. Neither of these
groups is, in fact, part of the ‘2nd millennium BC’ transition problem.
Historically, the period 1800–1400 BC is rather a curious mixture, and
references to pastoral peoples of the larger Arabian peninsula are virtually
non-existent. For example, the Old Babylonian period is dominated by an
overlapping series of three dynasties (Isin, Larsa, and Babylon) stretching
from c. 1850 to 1600 BC. However, these three dynasties (as well as the
contemporary one from Assyria) are derived from the earlier well-known
Amorite pastoral nomads of the Ur III period — Kinglist A records them as
part of the palû MAR. TU., the Amorite reign (Hallo & Simpson 1971, Fig.
20). Thus, the evidence for pastoral populations in the southern arid regions
adjacent to Babylonia for this period remains essentially unknown. Rowton
(1969, pp. 68–9) suggests that they had vanished from the documentary
sources because the sedentary population had become bi-ethnic during the
Old Babylonian period and distinctions of foreign populations were not
necessary. It may also be plausible that the local Semitic rulers of this
background would not distinguish between the settled populations of
southern Mesopotamia and the nomadic sections or tribes still found in the
adjacent arid zones. Such a distinction appears only in the poorly understood
succeeding Kassite period (see below). Yet even at the time of Ammisaduga
(c. 1600 BC), such honorific titles as Abu Amurrim may have pointed to the
continued presence of a tribal organization of the Amorites (Rowton 1969, p.
69, Matthews 1979, p. 129). Similarly at Sippar, there is the ‘fossilized’
presence of pastoral nomadic terms, such as rabianum and babtum, used in
urban context (Harris 1975, p. 38ff.). Even within the context of Old
Babylonian records, there is a rather abrupt cessation of records in the
southern plain by 1740 BC and 1720 BC in Middle Babylonia (Stone 1977, p.
271). The Assyrian documentation from the north is similar in pattern. There
is abundant documentation from the karum rulers of the Old Assyrian
period, beginning c. 1940 BC, and the Assyrian rule of Shamsi-Adad and
Ishme-Dagan (1740 BC) covering 200 years. But nothing of importance
concerning pastoral nomads appears until Adad-Nirari I (1307–1275 BC).
Concerning the Kassite records, historians generally assign a total length
of rule from 1595 to 1155 BC, covering some 440 years. But even here the
early portion of the period is very poorly attested textually (Brinkman 1976).
Hallo & Simpson (1971, pp. 105–9) remark that ‘for the first two centuries
of their rule we have virtually no contemporaneous documentation’ and label
the early part of the period as a ‘Dark Age’. Economic and historical data of
relevance here only appear at the time of Burnaburias II, c. 1350–1333 BC
(Brinkman 1976, passim).
The archaeological data from southern Mesopotamia itself reflect this
condition. In the Uruk area Adams & Nissen (1972, p. 39) note that ‘the
stable configurations of settlement … of the ED-OB period … drew to a
close in the Warka region with the OB period.’ Widespread abandonment
followed as an outcome of a slow protracted process. In the Kassite period,
sites were smaller and there were a number of disruptive breaks. By the end
of the period, a nadir in settled life had occurred. In the Ur area, commercial
life ceased around 1740 BC and did not revive until approximately 1400 BC,
250 years after the beginning of the Kassite period (Wright 1981, p. 332).
Sites show extensive abandonment both in the late Larsa-OB period and OB-
Kassite times (Ibid. p. 331). Adams (1981, pp. 167–9) suggests that a sharp
economic and demographic entrenchment took place in the OB period in the
entire southern alluvial plain, and in the Kassite period geographical
fragmentation took place.
From Egypt, a somewhat similar analogy can be suggested. The Middle
Kingdom evidence as derived from the Sinuhe account, the Execration
Texts, The Instructions of Amenemhet I, The Instructions of Merykare, and
the Brooklyn Papyrus suggest that Asians (Aamu) were both present in
Egypt and found in a wide range of contexts in Southwest Asia itself. By the
time of Amenemhet IV (1798–1790 BC) they had begun to arrive in Egypt in
large numbers (Van Seters 1966, p. 78, Prag 1985, p. 85). The problem is to
attempt a separation between those Asians who belonged to the pastoral
tribes of the Negev/Sinai and the sedentary populations of the Levant.
Certainly portions of the Sinuhe account and The Instructions of Merykare
refer to peoples of the Sinai, because of geographical proximity to the
eastern delta and defined lifestyle. This interpretation is confirmed by the
following passage from The Prophecy of Nefer-Rohu (in reference to the
fortifications of the northern part of the eastern delta): ‘… ics will not be
permitted to come down into Egypt that they might beg for water in the
customary manner, in order to let their beasts drink’ (Wilson 1958, p. 257).
(For a frontier fort of the new kingdom dated to Seti I in the northern Sinai at
Bir el Abd, see Oren 1973a.) The Execration Texts (Sethe 1926, Posener
1940) are a particular source of controversy. The dating is still debated
(Middle Kingdom c. 1880–1790 BC?), and defining the geographical location
of the names is very difficult. Van Seters (1966, p. 80) is quite adamant that
the mentioned princes and inhabitants are those of cities and ‘that there is no
reason to regard them as nomads.’ However, Brussels texts E50–E51
mentioning ‘chief of the tribes of Kwsw’ may parallel the Amarna
correspondence references to northwest Arabia (Posener 1940, pp. 88–9). In
addition, the Egyptian use of’Aamu’ as a term describing nomads or settled
groups is open to debate.
Similarly, the Pharaonic inscriptions found at Serabit el Khadim in the
eastern Sinai and dated principally to Amenemes III and IV (1842–1790 BC)
(Černý et al. 1955) mention and depict Asians (Černý 1935) in a role
consistent with the Egyptian historical evidence discussed above. We may
also lay to rest the idea that the Asians mentioned in the Middle Kingdom
texts at Serabit el Khadim did not come from the Negev/Sinai (contra Van
Seters 1966, pp. 88–90). These arguments are now rendered obsolete by the
archaeological evidence from the Negev/Sinai. This region was occupied by
a series of peoples from at least 6000 BC, which culminated in a definite
pastoral lifestyle by at least 4000 BC. The archaeological record for this way
of life is both abundant and distinctive, and shows clear ties to both the
southern Levant, Arabia, and Egypt. The disappearance of this
archaeological phenomenon during the MB I period c. 1850 BC, has caused
continuous debate and may be closely related to Asian influx into the delta
by late Middle Kingdom times (Albright 1955, Hayes & Watkins 1955).
The Second Intermediate Period (Hyksos domination, c. 1750–1550 BC)
again spans a critical period of the 2nd millennium BC. Here, as in
Mesopotamia, the Semitic populations of the Negev/Sinai and the southern
Levant dominated the Egyptian state at least in the eastern delta (see
Weinstein 1981, p. 10) and references to Asian populations of the
Negev/Sinai, southern Jordan, and Arabia would of course be nonexistent
and irrelevant. How were the Asians perceived in the delta by the Egyptians
themselves? According to The Admonitions of Ipuwer and other sources, it
appears that the Asians, (1) settled in the delta and took over its affairs, (2)
assimilated Egyptian culture and displaced Egyptians in places of authority,
(3) caused the northeast frontier of the delta to be open to all Asians, and (4)
apparently even became pharoahs in the XIII Dynasty (Van Seters 1966, pp.
115–16, Redford 1970). As Van Seters (1966, p. 126) suggested, the Hyksos
analogy is to the Amorite and Kassite control of southern Mesopotamia.

The later transition (1400–900 BC)

This is the key period in terms of historical data, both from Mesopotamia
and Egypt. It is within this time that we see the appearance of pastoral
nomadic groups in the arid zones which once again threaten the status quo of
established and revitalised urban states, such as New Kingdom Egypt, the
Levantine city states, and Kassite Babylonia/late Middle Assyria. The
relevant historical data are summarized below by ethnic name, beginning
with Mesopotamia.

Ahlamu/Aramu
The earliest occurrence of the term Ahlamu/Ahlame (discounting the earlier
references to Aram) is to be found in the second half of the 2nd millennium
BC. The term seems to refer to tribal confederations found on the fringes of
the Syrian desert. The first question to be asked is if this term refers
generically to ‘bedouin’. If it does, what is the relationship to the earlier
Amorite pastoral nomads decribed in the 3rd and early 2nd millenniums BC?
 In Assyria, the earliest royal reference comes from Adad-Nirari I (1307–
1275 BC) in an apparent mention of his father’s exploits (Arik-din-Ili, 1318–
1307 BC). Here the ‘hordes of Ahlamu’ are mentioned in the context of the
northern Euphrates region (Grayson 1972, p. 58). Shalmaneser I (1274—
1245 BC) mentions ‘an army of the Ahlamu’ who are assisting the Hittites in
the northern Euphrates region (Grayson 1972, p. 82). Tukulti-Ninurta I
(1244–1208 BC) states that the Ahlamu were to be associated with one of the
traditional regions of the Amorite nomads, the hilly arid zone west of Mari
in the Jebel Bishri area. He boasts that he controlled ‘the lands of Mari,
Hana, Rapiqu and the mountains of Ahlamu’ (Grayson 1972, p. 119). The
last royal reference to the Ahlamu is that of Ashur-resha-ishi I (1133–1116
BC) who again mentions the defeat of ‘an extensive army of the Ahlamu’ in
the northern Euphrates area (Grayson 1972, p. 147).
In the Tell el Amarna texts, EA 200 provides the sole example where the
term Ahlamu (lu ah-la-ma-i, or lú ah-la-ma-ú) is used. This text, dated to
Amenophis IV (c. 1389–1358 BC), depicts the Ahlamu as agitating for food
in a context that may mention the king of Babylon (Brinkman 1968, p. 389
& note 2185). If the term denotes ‘bedouin’ the Ahlamu must have been
widespread from the eastern desert of Arabia, well north to the Syrian
steppe. It seems reasonably clear that the Ahlamu are to be identified with
the distinct nomadic group later called Arameans who enter history with the
inscriptions of Tiglath-Pileser I and maintain a distinct ethnic and social
identity well through the 1st millennium BC (for summaries of their
characteristics, see Küpper 1957, p. 117ff., Brinkman 1968, p. 267ff.
Albright 1975b, p. 532ff.). Organized along tribal lines, the units appeared to
be rather small, i.e. segmented social patrilineages, with numerous elders or
sheikhs (nasiku). Of principal concern here is the lack of mention of camels
as domesticated animals.

Shosu/Kashu
The earliest clear reference to the Shosu (Š3św) comes from Thutmose II,
1500–1490 BC. The references reach a peak during the reign of Ramses II
(1304–1237 BC) and the last mention is under Ramses III (1198–1166 BC). A
compilation of the data principally by Giveon (1971, see also the review of
Ward 1973) suggests that they were to be found in the Nile delta, southern
Palestine (Negev), and in the Transjordan. Helck (1968, p. 477) states that
they were located south of the Dead Sea. More specifically, the Shosu are
often associated with Seir, a mountainous region located in southern
Transjordan. The term Seir occurs as early as the Amarna period (1389–
1358 BC) (EA = Tel el Amarna text 288 line 26: a a-di KUR.MEŠ Se-e-ri-ki)
and there are at least four documents to link the Shosu with this region
(Giveon 1971, p. 235). At the time of Ramses II, they were said to inhabit
six districts, among which is Śśw Śạ-’-r-r ‘Shosu in the land of Seir’ (Helck
1968, p. 477). The Tanis stele of Ramses II states that he ‘laid waste to the
land of the Asiatic nomads, who has plundered Mount Seir with his valiant
men.’ (Bartlett 1969, p. 1). The papyrus Anastasi VI associated with
Merneptah, also links the Shosu with Seir (Giveon 1971, pp. 131–4) as does
the papyrus Harris I of Ramses III (Bartlett 1969, p. 2, Giveon 1971, p. 136).
According to Helck, another name for Seir was Ja-ha-wa (one of the
districts in Shosu land at the time of Ramses II) and we see this term also in
an inscription of Amenophis III at the Temple of Amon at Soleb (t3 Šśw yhw
– ‘Yahwe in the land of Shosu’) (Giveon 1971, pp. 26–8).
A number of New Kingdom Egyptian objects and inscriptions found in
Transjordan suggests the local inhabitants were in contact with Egypt during
this period. The best known of these is the Balua Stela which may depict
local Shosu with Egyptians (for the latest summary and references, see
Dornemann 1983, p. 22ff., 153–4). Other objects include an Amenophis III
scarab found near Petra (Ward 1973), two Ramesside scarabs from central
Jordan (Dornemann 1983, p. 27), and an inscription of Queen Twosret at
Deir Alia (ibid.). (For the Wadi Timna material, see below.)
According to Giveon (1971, p. 240), the Shosu were seen principally as
pastoralists. According to the papyrus Harris I quoted above, Ramses III
destroyed the Shosu along with their tents, goods, and herds (Bartlett 1969,
p. 2). The Merneptah papyrus Anastasi VI quotes an Egyptian frontier
official who states: ‘We have finished letting the Shosu tribes of Edom
[IDM] pass the fortress “Merneptah-hotep-her-maat” which is in Tjekou, to
the pools of Pi-tum [Pe-atoum] of M. which is in Tjekou, in order to sustain
them and sustain their flocks through the good pleasure of pharoah.’
(Redford 1963, p. 406, Bartlett 1969, p. 2, Giveon 1971, pp. 131–4, Ward
1973, p. 52). The site of Tjekou is generally identified with the eastern delta
town of Tell Maskhoutah (Redford 1963, p. 406). The titles of the Shosu also
indicate that they were pastoral nomads.
 A direct outgrowth of defining the Shosu is the question of the inhabitants
of the Sinai/Negev in the New Kingdom. During the Old and Middle
Kingdoms there was a wealth of written information about these people from
the Wadi Magharah/Serabit el Khadim localities. However, as researchers
have noted, the New Kingdom inscriptions tend to become almost devoid of
information, heavily ritualized and stylized (Černý et al. 1955, p. 19). Since
archaeological survey in the Sinai has suggested that no new ethnic groups
arrived there during the Late Bronze Age (Beit-Arieh 1984, p. 52), we are
left with the impression that the earlier Middle Kingdom inhabitants
continued to be present in the region, albeit in a changed economic and
social pattern. It thus appears likely that the Shosu are the best candidates for
the inhabitants of the Negev/Sinai in the Late Bronze Age.
If we associate the Shosu with Edom and Moab (for the term Mw-ĺ-b
under Ramses II, see Kitchen 1964, p. 50), which group of people can be
assigned to the Midianites, located further to the south, but unattested in
New Kingdom contemporary accounts? (The term Midian/Madyan is
unknown outside of the Biblical context until classical times, see Knauf
1983, p. 148.) This controversy centres around the term Kaūu. The earliest
possible attestation of the term occurs in the late Middle Kingdom
Execration Texts. The historical clues to Kaūu are derived from the well-
known accounts of Genesis and Judges. According to recent re-evaluation,
the Biblical material suggests the presence of a large confederation or tribe,
sections of which were nomadic (Dumbrell 1975, Payne 1983, Mendenhall
1984). Whether the population functioned in a way similar to later bedouin
tribes is open to debate (Knauf 1983, pp. 150–1). Combining the Egyptian
evidence of Kaūu and the Biblical Midianites, we feel that a powerful
Semitic group in the area of Midian complemented the more northerly group
identified with the Shosu and Edom.

The archaeological evidence

The archaeological record for the Ahlamu will be discussed first, secondly
that of the Shosu/Kashu, and thirdly the evidence from the southern Nejd.
Concerning the archaeology of western Iraq and eastern Syria, virtually
nothing is known from the 2nd millennium BC. Work to date has been
confined to the Euphrates river valley and upper tributaries, and very little in
the way of comparable material of the 3rd millennium BC and earlier is
known from the desert (Zarins 1988). From northeastern Saudi Arabia,
survey work in the tributary wadis to the Euphrates revealed the presence of
3rd millennium BC stone circle complexes, but nothing definitely attributable
to the middle of the 2nd millennium BC (Parr et al. 1978, p. 37ff., Potts et al.
1978, p. 9, Gilmore et al. 1982, p. 16ff.). City III at the Qala on Bahrain
defines the Kassite period (Larsen 1983, pp. 80–1, 249–51 & Fig. 54), but
this distinctive ceramic horizon is also attested at other occupation sites
(Salles no. 2068) and tumuli (al-Hajjar, Diraz, Sar) (McNicoll & Roaf n.d.,
pp. 19–20, Rice 1972, pp. 69–72, Salles 1981, p. 2, Mughal 1983, p. 400). It
should be noted, however, that in each case the number of Kassite
occurrences is minimal. Finally, the record for the Kassite period clearly
suggests a political presence related to shipping and commerce involving the
Mesopotamian state to the north (including Failaka), and not pastoral
nomads (Edens 1986).
For the meagre evidence suggestive of a pastoral nomadic presence such
as the Ahlamu, there are the tumuli found in the Dhahran vicinity on the
adjoining Arabian mainland. Based on the evidence of the Nippur letters,
Cornwall (1952, p. 111) suggested that some of the more modest tumuli at
Dhahran (and Bahrain) were built by the Ahlamu. From the excavations
conducted in the Dhahran area, it appears that the tombs of the 3rd
millennium BC were re-used during the middle to latter part of the 2nd
millennium BC, in a number of which (B-7, B-6, A-5, and B-2) there were
aceramic burials. The associated grave goods consisted of copper and
occasionally iron arrowheads, copper bracelets, finger rings, beads, large
seashells, small stone palettes, ear plugs, and in A-5, an apparent game made
of a small white bowl and five stone balls (four black and one white) (Zarins
et al. 1984, pp. 38–9). A single supporting 14C date comes from the 3rd
millennium BC tomb A-4 which was disturbed sometime in the mid-2nd
millennium BC. The date taken from the fill yielded 1270 BC ± 345 (GX-
9590).
The archaeological evidence from the southern Nejd has filled in a long-
neglected gap in our attempt to delineate the human occupation record of the
Arabian peninsula. It is this record which may throw light on the
archaeological and historical record already presented above. The human
occupation of the Nejd in the Holocene began in the 6th millennium BC. It
seems likely that in the northern part of the Nejd a pastoral way of life was
present which was allied to a larger technocomplex described elsewhere
(Zarins 1988), while in the central and southern Nejd hunting and gathering
were probably dominant. By the end of the 3rd millennium BC radical
climatic and ecological changes took place in the Nejd. With the advent of
the 2nd millennium BC, there is a number of significant structural changes in
the human settlements of the Nejd as well.
The later sites have little in common with the earlier ones. For example,
site location preference is restricted to two basic locations. First, many sites
are situated at the base of small, granitic outcrops in a 50 m arc. Secondly,
sites can be found within small embayments or coves on lowlying
outcropping strata. There are no site clusters and population density was
clearly low. The most numerous structures are hearths, usually constructed
of small slabs and often packed with small cobbles. Living structures are
also common and are radically different from the earlier stone circle
complexes so familiar from northern Arabia, being ‘horseshoe’-shaped,
about 2 m wide and 3 m on each side, and distinctly outlined by small
boulders or cobbles (Zarins et al. 1980, p. 22 & PI. 31B, Whalen et al. 1981,
p. 51). In the Bir Hima region of southwest Arabia, there are 25 such
structures aligned in a single row approximately 95 m long (Zarins et al.
1981, p. 30). Long, rectangular organic superstructures define a second by
an outline in small cobbles (Fig. 14.3). These structures often contain
internal partitions and a well-defined entrance (Zarins et al. 1980, PI. 10A,
31 A, see PI. 8C for re-used example). Troughs are a third type which are
very prevalent (Zarins et al. 1980, PL IB). They are, however, more common
on wadi floors and junctions. These are 3–5 m long, double lines of slabs
about 30 cm apart, filled with small pebbles. In many cases, they are
‘hooked’ on both ends with the extensions about 50 cm long. In some cases,
the structures have an attached extension in the centre protruding about 0.5
m and also filled with pebbles (Zarins etal. 1981, PI. 42; for similar
structures identified as ‘open sanctuaries’ in the Negev/Sinai, see Avner
1984, p.ll9ff.). These structural types appear to be mutually exclusive at the
sites surveyed to date.
Figure 14.3 (a) Cleared habitation structure at 210–3, near Muwayh, Saudi Arabia (after Zarins et al.
1980, PL 10A); (b) plan of site 210–3, near Muwayh, Saudi Arabia (after Zarins et al. 1980, PI. 9B).
 Figure 14.4 (a) Tapered structure site complex at 212–73 near Layla, Jebel Tuwayq, Saudi Arabia
(after Zarins et al. 1979, PL 13A); (b) site plan at Waqir, 210–49, near Taif, Saudi Arabia (after Zarins
et al. 1980, PL 9A).
 Figure 14.5 (a) Outline of enigmatic stone structure at site 211–24 near Bisha, Saudi Arabia (after
Zarins et al. 1980, PI. 8B); (b) Site plan of Safra Huqayl, 206–60B west of Riyadh, Saudi Arabia
(after Zarins et al. 1980, PL 5).

Another group of structural remains may indicate a funerary or ritual

complex. The most noteworthy is the ‘tapered structure’ which looks like an
elongated wedge (Figs 14.4a & 14.5b). Already noted by researchers in the
al Fau area of central Arabia in the 1960s (Field 1971, p. 44 & PI. 21–5),
hundreds of these structures have been examined in the Nejd region. A very
narrow ‘tail’ is built of vertical slabs about 50 cm apart and filled with
rubble. Depending upon the length of the tail, the structure gradually widens
and expands in size until the maximum width is achieved at the ‘head’. Here
the walls taper outward and the centre of the structure remains hollow.
Excavation of these structures has confirmed that they were not utilized for
habitation or burial (Zarins et al. 1980, p. 32). Sometimes the tapered
structures are associated with ‘platforms’ which were first noted on survey
in 1979 (Zarins et al. 1980, p. 19) (Fig. 14.4b).
The platforms are circles up to 10 m in diameter, completely filled with
stones. Two such structures were excavated in 1979 and this confirmed the
fact that they were deliberately built in such a fashion (ibid. p. 32, Gilmore
et al. 1982, p. 16).
Pillars and aligned slabs represent the final structural category. Pillars as
cultic objects are attested from the 6th to the 3rd millenniums in Arabia
(Kirkbride 1969, Rothenberg 1973, Zarins 1979, Avner 1984, pp. 115–19),
but aligned slabs may be manifestations of the 2nd and 1st millennium BC.
Several of the latter in the Nejd have been examined.
Organic remains from the sites are rare since, with a few exceptions, they
are essentially surface sites that have undergone major deflation. Bone awls
have been recovered but animal remains are mostly fragmentary. From a
number of sites ostrich shell is known, and from 211–55 and 211–56
identifiable C. dromedarius fragments have been found (Zarins etal 1980, p.
32 n.6). From 2nd millennium BC Red Sea middens there are equid, bovid,
ovicaprid, and C. dromedarius remains (Zarins Sc Badr 1986 in press). The
best examples of C. dromedarius come from pit excavations at site 217–44
see (see Fig. 14.6).
The ethnoarchaeology of bedouin sites confirms the gap in time between
the modern bedouin and the structures described above. It is unfortunate that
detailed ethnoarchaeological work on the Arabian bedouin is lacking,
although a few preliminary studies have been undertaken in the Negev
(Rosen n.d.), the Sinai (Kozloff 1981), southern Jordan (Banning & Kohler-
Rollefson 1983), Qatar (Montigny 1978, pp. 197–9 & Figs. 8–9) as well as
in the central Nejd (Zarins etal 1980, p. 23 & PI. 11; for Iran, see Hole
1978).
In the Nejd, at the present day there is a minimal use of rock material to
steady major tent poles, outline hearths, and batten down tent sides. In Qatar,
cobbles are used to completely outline the tent structures almost in identical
fashion to that reported from site 210–3A (see above). At Beidha in Jordan,
Banning & Kohler-Rollefson (1983, p. 377) have seen stone hearths in use
and also bed and storage platforms and small stone pens. In the Negev,
Rosen (n.d.) identifies two types of settlements, the ephemeral campsites
characterized by the presence of large stones which serve as benches, lined
or unlined hearths, and black Gaza ceramics; and campsites consisting of
rebuilt stone circles generally of the Roman-Byzantine period or earlier. All
agree that bone, dung, and perishable debris predominate at an encampment.
In contrast to their older archaeological counterparts, stone tools are entirely
absent, ceramics were noted only at the Qatar and Negev sites, and it is
unlikely that recent bedouin build, recognize, or understand the function of
stone platforms, tapered structures, pillars, erect slabs, or troughs.

Figure 14.6 Site plan of 217–44 near Bir Hima, Saudi Arabia (after Zarins et al. 1981, PL 11).

The dating of the ancient sites discussed above presents a number of

difficulties, but probably most of them fall between 1900 BC and AD 650.
Clues to possible dates remain minimal, due principally to lack of
excavation.
The recognition of relevant archaeological data in both northern Arabia
and the southern Levant to complement the findings in the Nejd remains
elusive. There are virtually no 2nd millennium BC sites or structures allied to
those mentioned above. From central and southern Sinai, with the exception
of Serabit el Khadim itself, no sites are attributed to the Late Bronze or Early
Iron I ages (1550–1000 BC) (Beit-Aneh 1984, p. 52 & Map, p. 49). In
northern Sinai, with the exception of the formal Egyptian fort, Bir el Abd
attributed to Seti I, no sites of the Late Bronze or Early Iron Age have been
reported (Oren 1973a, 1973b, p. 200). From the Negev, a perusal of the
Archaeological Emergency Project reveals no Late Bronze Age sites (Cohen
1979, pp. 250–4) although some of the open shrines near Eilat are attributed
to the 13th—11th centuries BC (Avner 1984, p. 124). From southern Jordan,
in the Wadi el Hasa and the Maan-Aqaba areas, again no sites are attributed
to the Late Bronze Age I period (Jobling 1981, p. 109, 1982, MacDonald et
al. 1982, p. 126). However, it is known that pastoral people were present in
the area, from the historical data and from the isolated New Kingdom
Egyptian finds discovered in the Negev/Sinai and as far east as the
Transjordan. This point is driven home particularly well by the surveys and
excavations carried out in the Timna area of the Wadi Arabah. Sites such as
the Hathor shrine and rock engravings attest to the presence of Egyptians
from at least the time of Seti I well into the XXth Dynasty (Rothenberg
1972, pp. 129–207). In the Sinai, what people invented and used the proto-
Sinaitic script? Rothenberg (1980, p. 164) and Beit-Arieh (1983, p. 48)
suggest they were seminomadic native inhabitants, descendants from the
earlier centuries.
Most likely, the answer to why archaeological sites are ‘missing’ in these
areas lies in the nature of the archaeological record because the structural
remains are not diagnostic in themselves, and continued re-use makes
attribution to a specific period difficult (Avner 1984, p. 124). Lithics are
nondescript and ceramics do not readily match known cultural types found
in the settled zones. Thus, in a recent survey of northwestern Arabia, no sites
were found that could be linked to post-Neolithic materials, but ‘the date of
the enigmatic stone enclosures and cairns … may span the second and first
millennia BC’ (Fig. 14.5a) (Ingraham et al. 1981, p. 71 & Tables 1–2).
Similar statements could be made for the adjoining arid parts of Jordan,
Israel, Syria, Iraq, and the Sinai. It appears that in the north, particularly the
northwest, culturally recognizable societies tied to camel-based caravans and
the spice trade did not arise until after 1500 BC (Bulliet 1975). The crucial
period here, then, is the time span between the end of the earlier pastoral
technocomplex and the arrival of the domesticated camel, c. 1500 BC
(Hakker-Orion 1984, p. 209).
Figure 14.7 (a) Phase I ‘outline’ rock art scene at Jebel Kawkab, north of Najran, Saudi Arabia (after
Anati 1968a, p. 138 and Fig. 91); (b) Speared camel of Phase I or II from Bir Hima, site 217–35C
(after Zarins et al. 1981, p. 34).
Figure 14.8 (a) Phase II camel hunting scene in the Bir Hima region (after Anati 1968b, p. 54 & Fig.
4); (b) Phase II camel hunting scene in the Bir Hima region (after Anati 1986b, p. 64 & Fig. 15).

Support for this hypothesis comes from a number of sites. The earliest
bedouin site in the Bir Hima region has a date of 1215 BC; at Zubayda in the
Burayda oasis, excavations yielded a series of 14C dates which begin at 1315
BC (GX 7097, 3265 ± 150 BP for excavation unit III 7; Parr Sc Ghazdar
1980). From northwestern Arabia, in the Midian region, archaeological
investigations at Qurrayya suggest that the Midianite culture began no earlier
that the 13th century BC (Parr et al. 1970). During the 1980 survey,
investigations confirmed that the distinctive Midianite ware was made
locally at Qurrayya (Ingraham etal. 1981, pp. 71–5). Studies of the ware’s
distribution suggest a region which includes northwest Arabia, southern
Jordan, the Wadi Arabah, and more rarely in the adjoining northern Sinai
and southern Levant (ibid., Rothenberg & Glass 1983). Perhaps, in respect to
cultural development, the Midianites paralleled the Nabateans of a
millennium later. Both societies began as pastoral groups but attained
domination by controlling the spice trade with adjoining formal states. For
the Midianites, this can be seen at the Hathor shrine in the Wadi Arabah.

Camel domestication

The final focus of this study deals with the animal which transformed the
pastoral societies of the Arabian peninsula, the camel. Two lines of evidence
can be used, artistic and osteological. Contrary to earlier suggestions, the
camel was known to the early Holocene inhabitants of the peninsula.
Osteologically, the remains are widespread but not common (Ripinsky 1975,
Zarins 1978, Grigson 1983, Hakker-Orion 1984). Artistic evidence,
principally from the southwestern part of the peninsula, indicates that the
camel was hunted as a game animal until sometime in the 2nd millennium
BC. Based on rock art sequences developed by Anati (1962, 1968, 1970,
1972, 1974, 1979) and Tchernov (1974) and modified here, in the
southwestern part of the peninsula a number of depictions can be fitted into
the schema shown in Table 14.1. From southwestern Arabia, there are at
least five camel depictions (from Phase 1 (‘outline style’) Fig. 14.7a) Anati
1968a.,. p. 110 & Fig. 74, (138 & Fig. 91, 1974, p. 234 & Fig. 243). In one
case, the camel underlies a scene with a speared bovid (Zarins et al. 1981,
PI. 36B, site 217–36). In another, direct confirmation of hunting is to be seen
in a depiction of a speared camel (ibid., p. 35 & Fig. 34E) (Fig. 14.7b). From
Phase II, there is a number of additional scenes depicting hunted camels
(Figs 14.8 & 9) (Anati 1968b, p. 11 & Fig. 2, p. 60 & Fig. 29, p. 54 & Fig. 4,
p. 64 & Fig. 15, 1974, p. 132 & Fig. 248, p. 232 & Fig. 271). In one scene,
the hunted camel is attacked by a man using a transverse arrowhead, a type
common in the 5th-3rd millenniums BC (Fig. 14.9a) (Zarins et al. 1981, p. 35
& PL 34F). Rock art from these two periods is known from other areas,
including northern Arabia, Jordan, and the Negev/Sinai. However, rock art
depicting the camel is only rarely attested outside the southwestern zone.
One outstanding example is a Phase I outline camel from Jubba in the Nefud
desert (Fig. 14.10a). From the United Arab Emirates, a camel is depicted in
relief on several cairns at Umm an Nar. These can be dated to the 3rd
millenium BC (Thorvildsen 1962, Figs 7–8).
Table 14.1 The rock art sequence from southwestern Arabia
Phase Period Date
V Islamic AD 650 to present
IV Literate (Thamudic/South 500 BC to AD 650
Arabic)
III Pre-literate 1900–500 BC
II early pastoralists 3500–1900 BC
I early hunters 6000–3500 BC
 Figure 14.9 (a) Camel hunting scene from Phase II at Bir Hima, site 217–23B (after Zarins et al.
1981, PL 34F); (b) Camel hunting scene from the Jebel Kawkab area (after Anati 1968b, p. 11 & Fig.
2).
Figure 14.10 (a) Phase I or II lifesize outline camel from Jubba Lake, Saudi Arabia (view sideways);
(b) Hooked trough from site 207–46 in the Riyadh

Following Phase II, the clearest evidence for camel depictions comes from
Phase IV. The principal theme in this period is riding and hunting or razzia
raids. The rock art evidence for this period is widespread and common. From
a brief analysis of the rock art, it seems clear that camel domestication began
sometime in the late 3rd millennium BC, probably in southern Arabia,
supporting the ideas of several scholars (Mikesell 1955, p. 242ff., Dostal
1959, 1979, Bulliet 1975). The strongest evidence is to be gleaned from the
number of depictions in which human association with the camel is attested.
These are to be found in the southwestern portion of the Arabian peninsula.
To turn now to the osteological evidence: identified remains of camel at
archaeological sites are rare in the early Holocene, but they were part of the
hunted game. For example, their remains have been recovered at Ain al
Assad in southern Jordan in PPNB context (Kohler 1984, p. 201), at the
Pottery Neolithic site of Shar ha Golan, c. 5000 BC (Stekelis 1951, p. 16), at
Early Bronze Age I Arad, c. 2900 BC (Lernau 1978, p. 87), and at the Early
Bronze Age IV site of Bir Resisim in the Negev, c. 2000 BC (Hakker-Orion
1984, p. 209). It is not possible, however, to delineate domestic from wild
camels on the basis of morphological change in the skeleton alone (Hoch
1979, p. 607, Hakker-Orion 1984, p. 209). Thus, the question of camel
domestication is extremely complicated and frustrating. Our best evidence to
date for this process comes from eastern Iran. From the site of Shahr-i-
Sokhta, the excavator recovered not only osteological remains but also hair
and dung, found in a context datable to 2700 BC. This suggests that camel
domestication began in Turkmenia and spread south (Compagnoni & Tosi
1978, pp. 95–9). The domestic camel was apparently known to the
inhabitants of the Indus Valley civilization by 2300 BC, although the species
utilized remains open to question (Meadow 1984, p. 134 and references).
Close trade contacts existed between the Indus Valley civilization and the
eastern Arabian peninsula, and there is tentative evidence that the camel was
merely one of many items traded during the late 3rd millennium BC. At the
site of Umm an Nar, analysis of the osteological remains suggests a step
towards domestication, deduced from the unusual number of camel bones,
the age distribution, and the cultural context (Hoch 1979, p. 613). This
stimulus may well have come from the Indus valley (Zarins 1978).
Another animal involved in this southern trade was the zebu (Bos indicus).
Apparently domesticated by the mid-4th millennium BC on the Indian
subcontinent (Allchin 1969, pp. 318, 322), from the Arabian eastern littoral
the animal is attested both by artistic and osteological remains dating to the
late 3rd millennium BC (Hoch 1979, pp. 567, 614–17, Cleuziou 1982, p. 19
n. 2, Zarins & Badr 1986). As in the case of the domesticated camel, the
zebu is not attested in the Levantine littoral until the 15th century BC (Clason
1978).

Table 14.2 Suggested developmental model for camel bedouin.

 Camel remains from southern Arabia supporting the thesis that the centre
of domestication lay in the south are not common, but this may be due to a
lack of survey and excavation. From Sihi, a shell midden on the southern
Red Sea coast, camel remains have been recovered in a 2nd millennium BC
context. These include large fragments of the left and right maxillae of a
juvenile individual, as well as fragments of ribs, femur, and mandibles of
adults (Zarins & Badr 1986). The likelihood that camel remains will turn up
at other coastal sites of the 2nd millennium BC is great (e.g. see the site of
Subr in South Yemen, Doe 1961, and the Italian Survey of the Yemen Red
Sea coast, Tosi pers. comm.). Several bedouin sites of the 2nd millennium BC
also yielded camel remains. From a series of pits excavated at site 217–44 in
the Bir Hima region, there were a number of articulated camel fragments
(Zarins et al. 1981, PI. 43A–B). Other remains were found at 217–56 (Zarins
et al. 1980, p. 23 n. 6).
From North Arabia and the southern Levant, the occurrence of camel
remains follows the development of cultures involved in the South Arabian
overland spice trade. With direct Midianite association, there is a sherd
depicting a camel (Ingraham et al. 1981; PL 79/14). In the Wadi Arabah, at
Site 2, a copper smelting camp dated to the Ramessid period, c. 1350–1150
BC, ‘several camel bones’ were found with other faunal remains (Rothenberg
1972, p. 105). In a later report, the excavators mention that ‘a large quantity
of camel bones’ was uncovered at the 13th—12th century BC sites of Timna
(Rothenberg & Glass 1983, p. 122 n. 50). From Tell Jemmeh on the Gaza
strip, Wapnish (1982, p. 2, 1984, p. 171) identifies only seven camel bones
from levels attributable to the 14th-10th centuries BC. Similarly, at Heshbon
in the Transjordan, camel remains are very infrequent from the earliest
levels, 1230–1150 BC (Unit E 04–05, Weiler 1981, Table 4).
The osteological remains are supported by the historical evidence. The
earliest attestation of the camel in the cuneiform literature occurs in the
Assyrian context in the reign of Assur-Bel-Kala, 1074–1057 BC (Grayson
1976, p. 55). Camels were rare animals that were herded together and put on
display in Nineveh. From the Biblical context there are the well-known
accounts of the camel-riding Midianites (Judges 6–8), perhaps attributable to
the 13th-12th centuries BC. Finally, from Egypt there seems to be no early
term for the camel. The papyrus Anastasi VI, describing the herds of the
Shosu and other Asians, uses the term I3wt, which usually denotes herds of
small animals like sheep and goat, not larger cattle or camels (C. van Sielen
pers. comm.).
All this information may be combined in a chronology for the use of the
camel in Arabia (Table 14.2), integrating the ideas of Dostal (1959, 1979),
Bulliet (1975), Compagnoni & Tosi (1978), and Zarins et al. (1980, 1981).
In conclusion, it seems that the camel pastoral nomads of the Arabian
peninsula developed along a distinct path over the course of at least 2–3
millenniums. Lexically, the earliest evidence points to groups in the northern
part of the peninsula such as the Ahlamu/Aramu, the Shosu and Kasu. These
peoples must have undergone some type of radical change in cultural
adaptation before the historical context beginning c. 1400 BC. Unfortunately,
the earlier historical record covering the period 1800–1400 BC is largely
irrelevant to the topic of pastoralism. Archaeologically, there was a slow
development in the southern and central parts of the Arabian peninsula that
can be interpreted from a number of distinctive building structures and it
does appear that earlier ideas concerning the spread of camel domestication
from south to north are largely correct.

References
Adams, R. Mc. 1981. Heartland of cities. Chicago: University of Chicago Press.
Adams, R. Mc. & H. J. Nissen 1972. The Uruk countryside. Chicago: University of Chicago Press.
Albright, W. F. 1955. North-west Semitic names in a list of Egyptian slaves from the eighteenth
century BC. Journal of the American Oriental Society 74, 222–33.
Albright, W. F. 1975a. Ch. XX, The Amarna letters from Palestine. In The Cambridge ancient history,
3rd edn. Vol. 11/2, I. E. S. Edwards, C. J. Gadd, N. G. L. Hammond & E. Sollberger (eds), 98–116.
Cambridge: Cambridge University Press.
Albright, W. F. 1975b. Ch. XXXIII, Syria, the Philistines, and Phoenicia. In The Cambridge ancient
history, 3rd edn. Vol. 11/2, I. E. S. Edwards, C.J. Gadd, N. G. L. Hammond & E. Sollberger (eds),
507–36. Cambridge: Cambridge University Press.
Allchin, F. R. 1969. Early domestic animals in India and Pakistan. In The domestication and
exploitation of plants and animals, P. Ucko & G. Dimbleby (eds) 317–22. London: Duckworth.
Anati, E. 1962. Palestine before the Hebrews. New York: Alfred E. Knopf.
Anati, E. 1968a. Rock art in central Arabia. Vol. I: The ‘oval-headed people of Arabia’. Institut
Orientaliste. Louvain: Université de Louvain.
Anati, E. 1968b. Rock art in central Arabia. Vol. II: ‘Fat tailed sheep in Arabia’ and ‘The realist-
dynamic style of rock art in the Jebel Qara’ Institut Orientaliste. Louvain: Université de Louvain.
Anati, E. 1970. The rock-engravings of Dahtami Wells in central Arabia. Bollettino del Centro
Camuno di Studi Preistorici 5, 99–158.
Anati, E. 1972. Rock art in central Arabia. Vol. Ill: Corpus of the rock engravings, sectors A-H.
Institut Orientaliste. Louvain: Université de Louvain.
Anati, E. 1974. Rock art in central Arabia. Vol. IV: Corpus of the rock engravings, sectors J-Q. Institut
Orientaliste. Louvain: Université de Louvain.
Anati, E. 1979. LArte rupestre del Negev e del Sinai. Milano: Jaca Book.
Avner, U. 1984. Ancient cult sites in the Negev and Sinai deserts. Tel Aviv 11, 115–31.
Banning, E. B. & I. Kohler-Rollefson 1983. Ethnoarchaeological survey in the Beidha area, southern
Jordan. Annual of the Antiquities Department of Jordan 27, 375–83.
Bartlett, J. R. 1969. The land of Seir and the Brotherhood of Edom. Journal of Theological Studies 20,
1–20.
Beit-Arieh, I. 1983. Central-Southern Sinai in the Early Bronze Age II and its relationship with
Palestine. Levant 15, 39–48.
Beit-Arieh, I. 1984. Fifteen years in Siriai. Biblical Archaeology Review 10(4), 26–54.
Briant, P. 1982. Etat et pasteurs au Moyen-Orient ancien. Cambridge: Cambridge University Press.
Brinkman, J. A. 1968. A political history of post-Kassite Babylonia, 1158–722 B.C. Rome:
Pontificium Institutum Biblicum.
Brinkman, J. A. 1976. Materials and studies for Kassite history, Vol. 1. Chicago: The Oriental
Institute.
Bulliet, R. W. 1975. The Camel and the wheel. Cambridge, Mass.: Harvard University Press.
Černý, J. 1935. Semites in Egyptian mining expeditions to Sinai. Archiv Orientalm 7, 384–9.
Černý, J., T. E. Peet & A. Gardiner 1955. The Inscriptions of Sinai (2 vols). London: Egypt
Exploration Society.
Chang, C. & H. A. Köster 1986. Beyond bones: towards an archaeology of pastoralism. In Advances
in archaeological method and theory. Vol. 8, M. Schiffer (ed.), 97–148. San Francisco: Academic
Press.
Clason, A. T. 1978. Late Bronze Age-Iron Age zebu cattle in Jordan? Journal of Archaeological
Science 5, 91–3.
Cleuziou, S. 1982. Hili and the beginning of oasis life in eastern Saudi Arabia. Proceedings of the
Seminar for Arabian Studies 12, 15–22.
Cohen, R. 1979. The Negev Archaeological Emergency Project. Israel Exploration Journal 29, 250–4.
Compagnoni, B. & M. Tosi 1978. The camel: its distribution and state of domestication in the Middle
East during the third millennium B.C. in light of finds from Shahr-i Sokhta. In Approaches to faunal
analysis in the Middle East, R.M. Meadow & M. Zeder (eds), 91–103. Cambridge, Mass.: Harvard
University Press.
Cornwall, P. B. 1952. Two letters from Dilmun. Journal of Cuneiform Studies 6, 137–45.
Doe, D. B. 1961. Notes on pottery found in the vicinity of Aden. Department of Antiquities Annual
Report 1960–61, 3–20.
Doe, D. B. 1977. Gazeteer of sites in Oman. Journal of Oman Studies 3(1), 35–57.
Dornemann, R. H. 1983. The archaeology of the Transjordan in the Bronze and Iron Ages.
Milwaukee: Milwaukee Public Museum.
Dostal, W. 1959. The evolution of bedouin life. In LAntica Societa Beduina, F. Gabrieli (ed.), 1–34.
Rome: Studi Semitici.
Dostal, W. 1979. The development of bedouin life in Arabia seen from archaeological material. In
Studies in the history of Arabia, Vol. I, A. al-Ansary (ed.), 125–44. Riyadh: University of Riyadh
Press.
Dumbrell, W. J. 1975. Midian — a land or a league? Vetus Testamentum 25, 323–37.
Dyson-Hudson, R. & N. Dyson-Hudson 1980. Nomadic pastoralism. Annual Review of Anthropology
9, 15–61.
Edens, C. 1986. Bahrain and the Arabian Gulf during the second millennium BC: urban crisis and
colonialism. In Bahrain through the ages, Shaikha A. al Khalifa & M. Rice (eds), 195–216.
London: Routledge & Kegan Paul.
Edzard, D. O. 1957. Die ‘Zweite Zwischenheit’ Babyloniens. Wiesbaden: Otto Harrasowitz.
Eph’al, I. 1981. The ancient Arabs. Leiden: E.J. Brill.
Field, H. 1971. Contributions to the anthropology of Saudi Arabia. Miami: Field Research Projects.
Galaty, J. G. & P. C. Salzman (eds) 1981. Change and development in nomadic and pastoral societies.
Leiden: E.J. Brill.
Gilmore, M., M. al-Ibrahim & A. S. Murad 1982. Preliminary report on the Northwestern and
Northern Region Survey 1981 (1401). Atlal 6, 9–23.
Giveon, R. 1971. Les Bedouins Shosu des Documents Egyptiens. Leiden: E.J. Brill.
Grayson, A. K. 1972. Assyrian royal inscriptions. Wiesbaden: Otto Harrasowitz.
Grayson, A. K. 1976. Assyrian royal inscriptions, Part 2, Wiesbaden: Otto Harrasowitz.
Grigson, C. 1983. A very large camel from the Upper Pleistocene of the Negev Desert. Journal of
Archaeological Science 10, 311–16.
Hakker-Orion, D. 1984. The role of the camel in Israel’s early history. In Animals and archaeology.
Vol. 3: Early herders and their flocks, J. Clutton-Brock & C. Grigson (eds), 207–12. Oxford: BAR
International Scries 202.
Hallo, W. W. & W. K. Simpson 1971. The ancient Near East, a history. New York: Harcourt, Brace &
Jovanovich.
Harris, R. 1975. Ancient Sippar. Istanbul: Nederlands Historisch-Archaeologisch Instituut.
Hayes, W. C. & J. B. Watkins 1955. A papyrus of the late Middle Kingdom in the Brooklyn Museum.
New York: The Brooklyn Museum.
Helck, W. 1968. Bedrohung Palastinas durch Einwandernde Gruppen am Ende der 18. und am Anfang
der 19. Dynastie. Vetus Testamentum 18, 472–80.
Hoch, E. 1979. Reflections on prehistoric life at Umm-an-Nar (Trucial Oman) Based on faunal
remains from the third millennium B.C. In South Asian Archaeology 1977, M. Taddei (ed.), 589–
638. Naples: Istituto Universitario Orientale.
Hole, F. 1978. Pastoral nomadism in western Iran. In Explorations in ethnoarchaeology, R. A. Gould
(ed.), 127–68. Albuquerque: University of New Mexico Press.
Ingraham, M., T. Johnson, B. Rihani & I. Shatla 1981. Preliminary report on a reconnaissance survey
of the North-western Province (with a brief survey of the Northern Province). Atlal 5, 59–84.
Irons, W. & N. Dyson-Hudson (eds) 1972. Perspectives on nomadism. Leiden: E.J. Brill.
Jobling, W. J. 1981. Preliminary report on the archaeological survey between Ma’an and ‘Aqaba
January to February, 1980. Annual of the Antiquities Department of Jordan 25, 105–11.
Jobling, W. J. 1982. Aquaba-Ma’an Survey, January–February, 1981. Annual of the Department of
Antiquities of Jordan 26, 199–209.
Khazanov, A. M. 1984. Nomads and the outside world. Cambridge: Cambridge University Press.
Kirkbride, D. 1969. Ancient Arabian ancestor idols. Archaeology 22, 116–21, 188–95.
Kitchen, K. A. 1964. Some new light on the Asiatic Wars of Ramses II. Journal of Egyptian
Archaeology 50, 47–70.
Knauf, E. A. 1983. Midianites and Ishmaelites. In Midian, Moab and Edom, J. F. A. Sawyer & D. J.
A. Clines (eds), 147–62. Sheffield: J S O T Press.
Kohler, I. 1984. The dromedary in modern pastoral societies and implications for its process of
domestication. In Animals and archaeology. Vol. 3: Early herders and their flocks, J. Clutton-Brock
& C. Grigson (eds), 201–6. Oxford: BAR International Series 202.
Kozloff, B. 1981. Pastoral nomadism in the Sinai: an ethnoarchaeological study. Bulletin de l’équipe
et écologie et anthropologie des sociétés pastorales.
Küpper, J.-R. 1957. Les nomades en Mésopotamie au temps des rois de Mari. Paris: Société d’Edition
‘Les Belies Lettres’.
Lancaster, W. 1981. The Rwala bedouin today. Cambridge: Cambridge University Press.
Larsen, C. E. 1983 Life and land use on the Bahrain islands. Chicago: Chicago University Press.
Lernau, H. 1978. Faunal remains, Strata III—I. In Early Arad, R. Amiran (ed.), 83–113. Jerusalem:
Israel Exploration Society.
Luke, J. T. 1965. Pastoralism and politics in the Mari Period: a re-examination of the character and
political significance of the major west Semitic tribal groups on the Middle Euphrates, ca. 1828–
1758 BC. Unpublished PhD dissertation, University of Michigan, Ann Arbor.
MacDonald, B., G. Rollefson & D. Roller 1982. The Wadi el-Hasa Survey 1981: a preliminary report.
Annual of the Antiquities Department of Jordan 26, 30–41.
McNicoll, A. & M. Roaf n.d. Archaeological investigations in Bahrain, 1973–1975. MS. on file, State
of Bahrain, Ministry of Information.
Matthews, V. 1979. The role of the Rabi Amurrim in the Mari Kingdom. Journal of Near Eastern
Studies 38, 129–33.
Meadow, R. 1984. A camel skeleton from Mohenjo-Daro. In Frontiers of the Indus civilisation, B. B.
Lai & S. P. Gupta (eds), 133–9. New Delhi: Indian Archaeological Society.
Meadow, R. 1984. Qurayya and the Midianites.. In Pre-Islamic Arabia, Vol. II, A. al-Ansary (ed),
137–45. Riyadh: King Saud University.
Mikesell, M. W. 1955. Notes on the dispersal of the dromedary. Southwestern Journal of Anthropology
11, 231–45.
Montigny, A. 1978. Etude anthropologique au Qatar. In Mission archéologique française à Qatar, J.
Tixier (ed.), 181–200. Paris: CNRS.
Mughal, M. R. 1983. The Dilmun burial complex at Sar. Manama: Ministry of Information.
Oren, E. D. 1973a. Bir el-Abd (northern Sinai). Israel Exploration Journal 23, 112–13.
Oren, E. D. 1973b. The overland route between Egypt and Canaan in the Early Bronze Age. Israel
Exploration Journal 23, 198–205.
Parr, P. J. & M. Ghazdar 1980. A report on the soundings at Zubaida (al-’Amara) in the al-Qasim
region: 1979. Atlal 4, 107–17.
Parr, P. J., J. Harding & J. E. Dayton 1970. Preliminary Survey in N.W. Arabia, 1968. Bulletin of the
Institute of Archaeology, University of London 8 & 9, 193–242.
Parr, P. J., J. Zarins, M. Ibrahim, J. Waechter, A. Garrard, C. Clarke, M. Bidmead & H. al-Badr 1978.
Preliminary report on the second phase of the Northern Province Survey 1397/1977. Atlal 2, 29–50.
Payne, E. J. 1983. The Midianite Arc in Joshua and Judges. In Midian, Moab and Edom, J. F. A.
Sawyer & D. J. A. Clines (eds), 163–72. Sheffield: J S O T Press.
Posener, G. 1940. Princes et pays dAsie et du Nubie. Bruxelles: Fondation égyptologique Reine
Élisabeth.
Potts, D., A. S. Mughannum, J. Frye & D. Sanders 1978. Preliminary report on the second phase of
the Eastern Province Survey 1397/1977. Atlal 2, 7–28.
Prag, K. 1984. Review article, continuity and migration in the South Levant in the late third
millennium: a review of T. L. Thompson’s and some other views. Palestine Exploration Quarterly
116, 58–68.
Prag, K. 1985. Ancient and modern pastoral migration in the Levant. Levant 17, 81–8.
Redford, D. B. 1963. Exodus III. Vetus Testamentum 13, 401–18.
Redford, D. B. The Hyksos invasion in history and tradition. Orientalia 39, 1–51.
Rice, M. 1972. The grave complex at Al-Hajjar, Bahrain. Proceedings of the Seminar for Arabian
Studies 2, 66–75.
Ripinsky, M. 1975. The camel in ancient Arabia. Antiquity 49, 295–8.
Rosen, S. A. n. d. Observations on bedouin archaeological sites near Ma’aleh Ramon. Unpublished
MS.
Rothenberg, B. 1972. Were these King Solomon’s mines? New York: Stein & Day.
Rothenberg, B. 1973. Sinai Explorations III. Museum Ha’aretz Yearbook 15/16, 16–34.
Rothenberg, B. 1980. Sinai. Bern: Kummerly & Frey.
Rothenberg, B. & J. Glass 1983. The Midianite pottery. In Midian, Moab and Edom. J. F. A. Sawyer &
D. J. A. Clines (eds), 65–124. Sheffield: J S O T Press.
Rowton, M. B. 1969. The Abu Amurrim. Iraq 31, 68–73.
Rowton, M. B. 1974. Enclosed nomadism. Journal of the Economic and Social History of the Orient
17, 1–30.
Salles, J.-F. 1981. Bahrain: introduction et état des questions. In Fouilles à Umm Jisr (Bahrain), S.
Cleuziou, P. Lombard & J. T. Salles (eds), 1–12. Paris: Edition ADPF.
Sethe, K. 1926. Die Ächtung Feindlicher Fürsten, Völker und Dinge auf Altägyptischen
Tongefasscherben des Mittleren Reiches. Berlin: Walter de Gruyter.
Simpson, W. K. 1971. Egypt. In The ancient Near East, a history, W. W. Hallo & W. K. Simpson
(eds), 185–302. New York: Harcourt, Brace & Jovanovich.
Stekelis, M. 1951. A new Neolithic industry: the Yarmukian of Palestine. Israel Exploration Journal
1, 1–19.
Stone, E. C. 1977. Economic crisis and social upheaval in Old Babylonian Nippur. In Mountains and
lowlands, L. D. Levine & T. C. Young, Jr. (eds), 267–90. Malibu: Undena Press.
Tchernov, E. 1974. A study of the fauna from Sectors A-Q. In Rock art in Central Arabia. Vol. IV:
Corpus of the rock engravings, Sectors J-Q, E. Anati (ed.), 209–52. Louvain: Université de
Louvain.
Thorvildsen, K. 1962. Gravroser pa Umrn-an-Nar. Kuml, 191–219.
Van Seters, J. 1966. The Hyksos, a new investigation. New Haven: Yale University Press.
Wapnish, P. 1982. Camel caravans and camel pastoralists at Tell Jemmeh. Journal of the Ancient Near
Eastern Seminar 13, 101–21.
Wapnish, P. 1984. The dromedary and bactrian camel in Levantine historical settings: the evidence
from Tell Jemmeh. In Animals and archaeology. Vol. 3: Early herders and their flocks, J. Clutton-
Brock & C. Grigson (eds), 171–87. Oxford: BAR International Series 202.
Ward, W. W. 1973. The Shasu ‘Bedouin’, notes on a recent publication. Journal of the Economic and
Social History of the Orient 15, 35–62.
Weiler, D. 1981. Säugetierknochenfunde vom Tell Hesban in Jordanien. Inaugural Dissertation.
München: Detlev Weiler.
Weinstein, J. M. 1981. The Egyptian Empire in Palestine: a reassessment. Bulletin of the American
Schools of Oriental Research 241, 1–28.
Whalen, N., A. Killick, N. James, G. Morsi & M. Kamal 1981. Saudi Arabian archaeological
reconnaissance 1980. Preliminary report on the Western Province Survey. Atlal 5, 43–58.
Whiting, R. 1983. Early Old Babylonian letters from the Diyala. Unpublished PhD thesis, University
of Chicago, Chicago.
Wilson, J. A. 1958. Egyptian prophecy. In The ancient Near East. Vol. I, an anthology of texts and
pictures, J. B. Pritchard (ed.), 252–7. Princeton: Princeton University Press.
Wright, H. 1981. Appendix: the southern margins of Sumer. In Heartland of cities, R. Mc. Adams
(ed.), 295–345. Chicago: University of Chicago Press.
Zarins, J. 1978. The camel in ancient Arabia: a further note. Antiquity 52, 44–6.
Zarins, J. 1979. Rajajil: a unique Arabian site from the fourth millennium BC. Atlal 3, 73–8.
Zarins, J. 1986. MAR-TU and the land of Dilmun. In Bahrain through the ages, the archaeology.
Shaikha H. al Khalifa & M. Pice (eds), 233–50. London: Routledge & Kegan Paul.
Zarins, J. 1988. Archaeological and chronological problems within the Greater southwest Asian Arid
Zone: 8500–1850 BC. In Absolute chronologies in Old World archaeology, R.W. Ehrich (ed.).
Chicago: University of Chicago Press.
Zarins, J. & H. Badr 1986. Archaeological investigations in the Southern Tihama Plain II (including
Sihi, 217–107 and Sharja, 217–172) 1405/1985/. Atlal 10, in press.
Zarins, J., M. Ibrahim, D. Potts & C. Edens 1979. Saudi Arabian archaeological reconnaissance 1978:
the preliminary report on the third phase of the Comprehensive Archaeological Survey Program —
The Central Province. Atlal 3, 9–38.
Zarins, J., A. S. Mughannum & M. Kamal 1984. Excavations at Dhahran South — the Tumuli Field
(208–91), 1403 AH/1983. A preliminary report. Atlal 8, 25–54.
Zarins, J., A. Murad & K. al-Yish 1981. Comprehensive archaeological survey program — a. The
second preliminary report on the southwestern Province. Atlal 5, 9–42.
Zarins, J., N. Whalen, M. Ibrahim, A. Morad & M. Khan 1980. Comprehensive Archaeological
Survey Program — preliminary report on the central and southwestern provinces survey. Atlal 4, 9–
36.
15 Farming to pastoralism: effects of
climatic change in the Deccan
M. K. DHAVALIKAR

It was generally thought by early historians that in the development of

mankind, the hunting-gathering stage was followed by pastoral nomadism
after which man began to produce his own food by domesticating plants and
animals. This was the accepted view until the middle of the 19th century
when Hahn suggested that pastoral nomadism was an offshoot of sedentary
agriculture, and it is now generally agreed that pastoral nomadism came after
crop cultivation (Khazanov 1984, p. 85). One reason for this view is that
nowhere in Eurasia have there been pastoral nomads without the knowledge
of agriculture. An excellent illustration of this view is provided by the
archaeological evidence in the Deccan, which roughly comprises the State of
Maharashtra and parts of Karnatak and Andhra Pradesh. Exploration
followed by selective excavation in the Krishna valley has provided
convincing evidence of flourishing agricultural communities in the latter half
of the 2nd millennium BC, whose successors had to resort gradually to
pastoral nomadism because of the drastic change in climate. This was
confirmed in the course of our intensive excavations at Inamgaon District,
Poona, Maharashtra (lat. 18°35′N, long. 74°32′E). Thirteen seasons’ work
has enabled us to study the culture process and culture change throughout
the nine centuries of occupation at the site, from 1600 BC to 900 BC. This
chapter attempts to explain the culture process during the Late Jorwe phase
of occupation in the Deccan, which marks a shift from farming to
pastoralism.
The first quarter of the 2nd millennium BC marks the appearance of early
farming communities in the Deccan, the region between the Tapi and the
Krishna rivers in the Indian peninsula. This distribution, however, was
sporadic, but from the middle of the 2nd millennium BC the region was
dotted with several self-sufficient villages, some of which, such as Prakash,
Daimabad, and Inamgaon, developed rapidly and became large regional
centres in the Tapi, the Godavari, and the Bhima valleys, respectively (Fig.
15.1). They appear to have been organized into chiefdoms, as the evidence
from excavations at Daimabad and Inamgaon demonstrates (Dhavalikar
1981–83). This prosperity was in a large measure due to the congenial
environment during 1500–1000 BC, which was a relatively wet phase
(Krishnamurthy et al. 1981). The prosperity, however, did not last long and
came to an abrupt end by the close of the 2nd millennium BC.

Figure 15.1 Map showing distribution of Early Jorwe and Latejorwe cultures.

The opening of the 1st millennium BC witnessed large-scale desertion of

settlements by the early farming communities in the northern Deccan.
Everywhere in the Tapi and the Pravara-Godavari valleys human activity
came to a halt (Dhavalikar 1984). We have, as yet, no satisfactory
explanation of this abandonment of habitations, but a possible clue is
furnished by the soil analysis of the sterile layer encountered in the
excavations at Nevasa (District Ahmednagar, Maharashtra) which, according
to Mujumdar & Rajaguru (1965, p. 152), belongs to the brown soil group
and suggests a less humid climate than when the virgin black soil was
formed. They observe:

Our pedological investigations have revealed that the so-called weathered

black layer shows definite soil characteristics and it has been formed by
the weathering of the habitational deposits of chalcolithic period. This
indicates that the place was deserted by the chalcolithic people due to
some calamity and considerable period must have elapsed to allow for the
growth of vegetation and the formation of soil on the top portion of the
habitational deposit. The chemical and physical analyses indicate that the
soil formed here belongs to the brown soil group and as such
comparatively drier climatic conditions must have prevailed with annual
rainfall of about 50 cm accompanied by hot summers and corresponding
scrub forest type of vegetation growing in the locality.

On the basis of further studies, Mujumdar (pers. comm.) thinks that the
rainfall might have decreased to about 300 mm, indicating that the climate
was becoming more and more arid. It is common knowledge that agriculture
in India is a gamble with nature; every third year is a bad year and every
fourth, a famine. Even in our own times, large-scale migrations take place if
there are successive droughts, and it is highly likely that the same may have
happened around 1000 BC. Thus, the early farmers of the northern Deccan
seem to have deserted their settlements at the close of the 2nd millennium
BC, at least in the Tapi and the Godavari valleys, but it appears that the
climate in the Bhima valley was slightly more congenial and, hence, they
continued to live there, as the evidence from many sites in the Upper Bhima
valley indicates. It is this culture which has been labelled as the Late Jorwe
and had been identified first at Inamgaon.
A number of sites was discovered in the Bhima valley, more particularly
in the Bijapur District of Karnatak (Sundara 1968, 1969–70, 1970–71). All
these sites have yielded a typical black-on-red painted pottery, unmistakably
of the Jorwe fabric. But the precise stratigraphical position of this ware was
determined in the course of our large-scale excavations at Inamgaon (District
Poona, Maharashtra) which is located on the right bank of the Ghod, a
tributary of the Bhima (Dhavalikar 1975–76). It is actually the Late Jorwe
which, on the basis of combined evidence of stratigraphy and antiquities, has
been assigned to c. 1000–700 BC. The Late Jorwe occupation is also present
at Songaon (District Poona, Maharashtra), as is clear from the evidence of
pottery which has been described by the excavator as the ‘degenerate’ Jorwe
ware (Deo 1969, p. 5). It appears from the available evidence that the Late
jorwe culture was confined to the Bhima valley and represents the end phase
of the Chalcolithic of the northern Deccan.
Of all the sites of the Late Jorwe culture, perhaps the most extensive is
that at Inamgaon, where the habitation of this phase was spread over an area
of about 5 ha making it one of the largest Chalcolithic settlements in the
Deccan. Upstream in the Bhima valley the westernmost sites are Theur and
Sashtewadi, both near Poona, which were excavated by S. R. Rao of the
Archaeological Survey of India (IAR 1969–70, pp. 27–9, 1971–72, pp. 35–
6). Further southeast, in Karnatak, Sundara (1968) has brought to light over
20 sites of the Late Jorwe culture which he has classified separately (Group
C). All of them are small settlements, extending from 1 to 2 ha in extent. It is
needless to emphasize, therefore, that Inamgaon is the largest settlement of
this culture.
The sparse settlement in the Bhima valley in prehistoric times was due to
the fact that practically the whole basin is a dry area with an average rainfall
varying between 400 and 700 mm. Precipitation is high in the source region,
but the early farming communities never occupied the high altitude areas
with more rainfall. Although the low-lying valley terraces are known for
their black cotton soil in the east (in parts of Ahmednagar and Sholapur
districts), the upland areas have a capping of poor soil and their junction
with low-lying valleys is marked by a reddish loam. Thus, large tracts of
fertile soil were not available and the pioneering colonizers therefore located
their settlements in those areas where patches of arable land existed, as at
Chandoli, Songaon, Inamgaon, etc. Many of these sites are located in areas
where the river takes a sharp meander, as at Inamgaon, or are located on the
confluence of rivers, as at Songaon, and are naturally well protected. Thus
food, water, and security were all available in these areas.
The excavated sites of the Late Jorwe culture include Songaon, Theur,
Sashtewadi, and Inamgaon, all located in the upper Bhima valley. The first
three of these have not yielded anything significant, except pottery, but
Inamgaon has been systematically excavated for 13 seasons, as a result of
which we now have a fairly good idea of the life of the Late Jorwe people.
Over 50 houses of this phase have been exposed and they throw a flood of
light on the micro-settlement pattern of the period. They are all modest huts,
mostly round in plan, and having dwarf mud walls which were probably
covered by the wattle-and-daub construction (Fig. 15.2). The roof, probably
conical, was thatched. Typha grass (Typha latifolia) was used as roofing
material, as is done today. It is waterproof and grows abundantly along the
banks of streams in the region. The impressions of walls and roofs have been
recovered in the course of excavations; they are similar to those of the typha
grass. The floor inside the hut was carefully made; it was composed of a
layer of gravel over which was rammed yellow silt and black clay. It was
frequently plastered with mud and cowdung, as is the case today. It was
probably repaired, or rather relaid, after every one or two years, and in one
hut we could count 14 successive layers of floors. The courtyard of the
house was also similarly well made and plastered with mud. All these huts
were rather small in size, their diameter varying from 2 m to 3 m.
Some of the huts contained a set of four flat stones which were meant to
support either four-legged storage bins, as at present, or ajar with four legs,
the like of which has been recovered in the course of excavations at
Inamgaon. In most of these round huts we came across a hearth (chulah)
which was nothing but a trough-shaped fire pit; very rarely was there a semi-
circular clay chulah identical with modern hearths. Sometimes, however, the
chulah was located outside in the courtyard which, as already noted, was
well rammed and plastered with mud. The hut was usually so small that with
a chulah and a storage jar, there was hardly enough space for people to live
in and one wonders how a family of six could have been accommodated in
it. But it should be stated that in India, the climate being hot and damp, it is
the courtyard where much of the life is lived, while the house or the hut is
used for keeping valuables, cooking, and storage. However, it appears that
each household had more than one hut. This observation is supported by the
artifactual evidence and also by the clusters of three or four huts with only
one hearth and one set of four flat stones for supporting a storage jar.
The house type marks a noteworthy change from the Early Jorwe to the
Late Jorwe, that is from rectangular to circular, and, therefore, an
explanation is in order. Poverty was no doubt the most important cause, and
is reflected also in the coarse pottery of Late Jorwe. It was probably the
result of the decrease in rainfall, which caused the general economic decline.
Dwindling agriculture was also one of the most important consequences.
The people obviously could not have afforded the large rectangular houses
which their predecessors had built and so they constructed small round huts.
Our ethnographic survey in the surrounding region shows that, even today,
poor people, especially migrant labour on sugarcane fields, build small,
round huts mainly because they are easy to build in a short time and, what is
more, they withstand strong winds which are a characteristic feature of arid
or semi-arid regions. Inamgaon today is a semi-arid area falling in the rain
shadow zone with average precipitation around 450 mm. The ethnographic
evidence also suggests that the Late Jorwe people may have lived a semi-
nomadic existence (Flannery 1972).
 Figure 15.2 Inamgaon, Late Jorwe Houses (c. 1000–700 BC)

The poverty of the people becomes extremely marked around 800 BC,
after which we do not come across well-made round huts, but only patches
of flimsy floors and post-holes which do not make a sensible plan. Charred
seeds of cultivated grain become scarce, but the number of animal bones
increases considerably. It was with great difficulty that we could recover a
few houses of the end phase of the Late Jorwe. They are irregular circular or
rectangular in plan and have sunken floors. In one hut the owner was buried
in a crouching posture and an antler was placed near his feet (Fig. 15.3). All
this only goes to show that at the end of the Late Jorwe people were again
resorting more and more to hunting and gathering, and living in flimsy huts
only seasonally.
Along with round huts, a few rectangular houses, probably belonging to
the well-to-do people, have also been recovered. Two such houses were
encountered in the eastern periphery of the principal habitation area on the
river front. One of these was a multi-roomed structure, part of which was
completely burnt by fire, and hence everything inside was found almost
intact. There were several storage jars containing lots of charred grain. The
wooden posts and the roof had been burnt, trapping a three-year-old child
below a wooden post. He died on the spot; his bones and even teeth being
charred. In one of the southern rooms of the house was found the burial of
the owner, which will be discussed later.
In the courtyard of the house were three small oval huts without any well-
made floor. They can better be described as sunken floors. One of these has a
small semi-circular verandah in the front in which were found a stone anvil
and hundreds of finished and unfinished chalcedony blades, indicating that
this was the place where tools were made.
This was the largest house of the Late Jorwe and from the contents it
appears that it may have belonged to a very important person, perhaps the
ruling chief. The rectangular house-plan is said to be indicative of fully
settled life (Flannery 1972), and we may then suggest that, although a
majority of the Late Jorwe people led a semi-nomadic existence, at least a
few enjoyed the luxury of fully settled life. The same situation exists at
present, as will be discussed later.
 Figure 15.3 Inamgaon, a Late Jorwe hut, c. 800 BC.

We have interesting evidence of the subsistence pattern of the Late Jorwe

people from the Inamgaon excavations which have yielded a large number
of charred seeds. The prosperity of the Early Jorwe period was a thing of the
past in the opening century of the 1st millennium BC. The people
nevertheless were farmers, practising subsistence agriculture, cultivating
barley as the principal cereal crop. In the initial stage, wheat was grown as
was done by their predecessors who, however, could cultivate it because of
the artificial irrigation made possible by the diversionary channel through
which the flood-water of Ghod was diverted (Dhavalikar 1975–76, pp. 47–
8). But soon, it appears, the diversionary channel became silted up and fell
into disuse. This also explains the decline in wheat production, which later
completely stopped (c. 900 BC). After this, the people subsisted more on
barley, and perhaps even more on animal foods, as is suggested by the
increase in the quantity of animal bones towards the closing stages of the
Late Jorwe. Fruit (Zizyphus) and fish supplemented the diet.
The middle levels of the Late Jorwe period at Inamgaon mark a
noteworthy change. From here onwards house plans could be recovered with
great difficulty; the quantity of charred grains decreased drastically,
suggesting a total decline in agriculture, but correspondingly there is a
considerable increase in the yield of animal bones. There is indirect evidence
to suggest that the Chalcolithic people subsisted more on animal food than
on plant food. The comparative analysis of tooth size and dental indices of
the Inamgaon skeletal series gives an impression of a relatively and
absolutely large-toothed population indicating that their diet was much
coarser (Lucaks 1985, pp. 820–1). Our statistics show that in the early levels
of the Late Jorwe, the percentage of bones of cattle was much more than the
sheep/goat bones (56 per cent cattle; 25 per cent sheep/goat), but from the
middle levels the percentage of the former begins to decrease and that of the
latter increases considerably (26 per cent cattle, 52 per cent sheep/goat). It
may, however, be stated that the percentage of hunted animals, more
particularly gazelle and antelope, also increases, from 5 per cent to 14 per
cent. It is, therefore, clear that the people were gradually resorting to
sheep/goat pastoralism in the 9th century BC. It is also roughly around this
time that the black-and-red ware appears at the site. This may, perhaps, be
due to the incursions of the megalith builders of the southern Deccan. Very
probably these Iron-age horsemen were to some extent responsible for
driving away the Late Jorwe pastoralists from their settlements. This is not
unlikely, in the light of the existence of megalithic stone circles near
Inamgaon at Pimpalsuti (Ansari & Dhavalikar 1976–77).
The Late Jorwe people worshipped a mother goddess, as did their
predecessors. She was probably the goddess of fertility, as one specimen was
found in close proximity to pit silos of the Early Jorwe period. Such figures
betray crude modelling; their extremities are stumpy and there was no
attempt at depicting the facial features. So, also, is the case of a god who was
worshipped occasionally, as the discovery of two such unbaked figures near
a chulah in the courtyard of a Late Jorwe house suggests. At present, almost
identical figures of wheat flour are made and worshipped at the time of
community feasts, being immersed in the river or a well after the successful
conclusion of the feast (Dhavalikar 1970).
The people believed in life after death. The dead were buried usually
within the house floor in a pit specially dug for the purpose. The custom was
the same as the Early Jorwe burial practice. Adults were buried in an
extended position with head towards the north and legs towards the south,
whereas the children were accommodated in two grey ware urns, placed
mouth to mouth horizontally in the burial pit. Usually a painted bowl and a
spouted jar were also placed in the pit; they probably contained food and
water for the dead. Some burials also contained more vessels, which
doubtless indicate the economic condition of the family. As in the Early
Jorwe period, the portion below the ankle of the dead was, perhaps
deliberately, chopped off. This may have been done with a view to
preventing the dead from turning into ghosts and running away.
A burial of a different class was found in the only multi-roomed house of
this period, which was unearthed at Inamgaon. It was in one of the rooms
inside the house, where, in a pit, were recovered two skeletons, one over the
other, with their heads towards the north and the legs towards the south (Fig.
15.4). The lower skeleton was that of a male aged about 30 years old, and
that over him was of a female aged about 25 years old. The latter has a
concussion mark over her forehead which may have been caused by hitting.
It is therefore highly likely that she was deliberately stunned and made
unconscious and then buried with her consort. Another feature of this double
burial is that the lower extremities of the legs of both the skeletons are intact,
they were not chopped off as was the general rule. A number of pottery
vessels was placed inside the burial pit. All this suggests that he was a very
important person of the Late Jorwe community, living in a large multi-
roomed house which also had three oval huts in the courtyard, probably for
servants. He may, perhaps, have been the ruling chief of the community.
At the end of the Late Jorwe in the 8th century BC the people used coarse
painted jars in place of grey urns for burying children. This reflects the
general deterioration in the potter’s art, which is only a poor survival of the
Jorwe painted ware. Another not eworthy feature is that the black-and-red
ware, which appears in the end phase of the Late Jorwe, also occurs in
burials, but one thing is clear; the people become poor in the latter half of the
Late Jorwe, so much so that instead of two urns some of them used only one
urn for the child burial.
In the 8th century BC the Late Jorwe people completely vanished from the
scene without leaving any trace. The reason for this is not difficult to see.
The Inamgaon evidence demonstrates that, in the closing stages, agricultural
activity had completely dwindled, so much so that, in spite of our best
efforts, very few cultivated grains could be found. The people can therefore
be said to have been again slowly reverting to pastoral nomadism. They
subsisted more and more on animal foods, as is evident from the sudden
increase in remains of animal bones in the middle levels of this period. It is
significant that the oldest surviving text (5th℃3rd centuries BC) of the Pali
Buddhist canon reports an argument between two ascetics, of whom the
more austere censures the other for living as an alms man, begging food
cooked by others, when it would be more righteous to subsist by pure
(vegetarian) food-gathering in the wilderness (Kosambi 1963). This suggests
that food-gathering was an accepted and more respectable mode of
subsistence at the beginning of the early historic period, around the 6th–5th
centuries BC.
 Figure 15.4 Inamgaon, a double burial, c. 1000 BC

It is highly likely that settlement at Inamgaon was rather seasonal, as very

few house plans could be recovered. Their dwellings were flimsy structures
of impermanent nature. All this indicates the deterioration of the agricultural
economy. This was also the time when the megalith builders from the south
were coming into the northern Deccan, as the evidence from Pimpalsuti,
which is hardly 5 km from Inamgaon, shows (Ansari & Dhavalikar 1976–
77). In all probability, the poor Chalcolithic folk were dominated by the
megalith builders, who possessed the horse and effective iron weapons.
There was, however, an overlap between the Chalcolithic and the Iron Age
and this provides the link between the prehistoric and early historic periods
which was thus far missing.
The lifestyle of the Late Jorwe pastoralists was probably not much
different from that of the present-day Dhangars in Central Maharashtra,
which is a semi-arid region. They are semi-nomadic pastoralists who live in
their settlements from May to October, which is the period of the
southwestern monsoon. From October and November they leave their
settlements (vadis) on the plateau and begin their march to the western coast.
They move, with all their belongings loaded on horseback, to the Kolaba
District on the western coast, for grazing the sheep on the stubble in rice
fields where the harvesting is over by October—November. The farmers
invite them so that the sheep grazing in their fields fertilize them by their
droppings. Again, by April they start to return to their settlements on the
plateau, before the onset of the monsoon because the sheep cannot survive in
rain.
Some of the Dhangars also own lands, and in their absence the farmers in
the villages, who lead a settled life, cultivate their lands. Thus, there exists a
symbiotic relationship between the Dhangars and the farmers in a village
(Sontheimer 1975, p. 167). It is noteworthy that the Dhangars today occupy
roughly those areas where the Late Jorwe settlements have been found. This,
however, does not iri any way suggest that the present-day Dhangars are the
successors of the Late Jorwe people, but what is important to note is that,
just as the arid phase in the first half of the 1st millennium BC forced the
Chalcolithic farmers to resort gradually to semi-nomadic sheep/goat pastoral
nomadism, the same phenomenon is being repeated in the second half of the
2nd millennium AD. Pastoral nomads, the Gavali Dhangars, whose main
occupation is rearing water buffalo and whose habitat is the higher rainfall
areas of western Maharashtra, have of late started keeping more and more
goats because of the deterioration of the vegetation (Gadgil & Malhotra
1979, p. 62)
The culture process discussed in the foregoing pages was, indeed, a
remarkable phenomenon in the prehistory of the Deccan. It is, however,
important to know that it was not an isolated phenomenon, but was repeated
throughout the history of the region. Whenever there were successive
droughts and famines, people had no alternative but to resort to pastoral
nomadism. (It is significant that some of the dynasties of the early Medieval
period, such as the Yadavas of Maharashtra, the Hoysalas of Karnatak, and
the Vijayanagar kings of Andhra-Kasnatak, claimed their descent from the
Yadavas, who were pastoralists.)There was a general economic decline in
the early Medieval period from the 7th to the 10th centuries (Sharma 1986)
and again later from AD 1500 to 1900; the effect on human life was the same
as during the first half of the 1st millennium BC. It has been observed that
from the 13th century until the middle of the 15th century there was no
attempt to settle the proper revenue divisions. The land was covered with
grass and the main occupation of the people was cattlebreeding (Gune 1953,
p. 4 f.n.9). This probably continued until the 19th century when the British
conquered Maharashtra from the Peshwas (in 1818). Mountstuart
Elphinstone, the representative of the East India Company who was placed
in charge of the administration, found that there was no lack of arable land
but that there were not enough people to cultivate it, the reason probably
being that they were living as cattlebreeders (Sontheimer 1975, p. 149).
The culture change that occurred in the Deccan during prehistoric and
historic times was doubtless brought about, in a large measure, by the drastic
change in climate. It became more and more arid and, as a result, droughts
became of frequent occurrence. The climate change in the first half of the 1st
millennium BC has been reflected in the archaeological record, whereas that
for the early Medieval (7th–12th centuries) and the Medieval (AD 1500–
1900) has been referred to in the literature (Dhavalikar 1987). It is
significant that it was roughly during these periods, that is c. 1000 BC-500 BC
and AD 1500–1900, that Europe is said to have experienced mini Ice Ages
(Birks 1986). It appears that whenever Europe has periods of intense cold,
the tropical lands suffer droughts. It has been observed that

Climate is a world-wide integrated system. Significant changes cannot

take place in one part of the system without changes occurring in other
places. There are dynamic connections that interlink climatic changes in
various parts of the globe (Bryson & Ross 1977, p. 509).

References
Ansari, Z. D. & M. K. Dhavalikar 1976–77. Megalithic burials at Pimpalsuti. Bulletin of Deccan
College Research Institute 36, 84–8.
Birks, H. J. B. 1986. Late Quaternary biotic changes in terrestrial and lacustrine environments, with
particular reference to north-west Europe. In Handbook of Holocene palaeoecology and
palaeohydrology, B. E. Berglund (ed.), 3–65. New York: Wiley.
Bryson, R. A. & J. E. Ross 1977. Climate variation and implications for world food production. World
Development 5, 507–18.
Deo, S. B. 1969. Sonegaon excavation. Poona: Deccan College.
Dhavalikar, M. K. 1970. A prehistoric deity of western India. Man 5, 131–2.
Dhavalikar, M. K. 1975–76. Settlement archaeology of Inamgaon. Puratattva 8, 44–54.
Dhavalikar, M. K. 1981–83. Chalcolithic cultures: A socio-economic perspective. Puratattva 12–13,
63–80.
Dhavalikar, M. K. 1984. Toward an ecological model for chalcolithic cultures of central and western
India. Journal of Anthropological Archaeology 3, 133–58.
Dhavalikar, M. K. 1987. Cultural ecology oj Maharashtra. Unpublished paper read at the Second
International Seminar on Maharashtra – Culture and Society, University of Poona, Pune, 3–5 Jan.
Flannery, K. V. 1972. The origins of the village as a settlement type in Mesopotamia and the Near
East: a comparative study. In Man, settlement and urbanism, P. Ucko, R. Tringham & G. W.
Dimbleby (eds), 23–54. London: Duckworth.
Gadgil, M. & K. C. Malhotra 1979. Ecology of a pastoral caste: the Gavali Dhangars of peninsular
India. Calcutta: Indian Statistical Institute.
Gune, V. T. 1953. The judicial system of the Marathas. Poona: Deccan College.
IAR 1969–70. Indian archaeology – a review. Annual report of the Archaeological Survey of India,
New Delhi.
IAR 1971–72. Indian archaeology – a review. Annual report of the Archaeological Survey of India,
New Delhi.
Khazanov, A. M. 1984. Nomads and the outside world. Cambridge: Cambridge University Press.
Kosambi, D. D. 1963. Staple grains in the western Deccan. Man 63, 130–1.
Krishnamurthy, R. V., D. P. Agrawal, V. N. Misra & S. N. Rajguru 1981. Palaeoclimatic influences
from the behaviour of radiocarbon dates of carbonates from sand dunes of Rajasthan. Proceedings
of the Indian Academy of Sciences (Earth Planet. Science) 90, 155–60.
Lucaks, J. R. 1985. Tooth size variation in prehistoric India. American Anthropologist 87, 811–25.
Mujumdar, G. G. & S. N. Rajguru 1965. Comments on soils as environmental and chronological tools.
In Indian Prehistory 1964, V. N. Misra & M. S. Mate (eds), 248–53. Poona: Deccan College.
Sharma, R. S. 1986. Historical archaeology and problems of urban history. Unpublished paper read at
the SAARC Archaeology Conference, New Delhi.
Sontheimer, G. D. 1975. The Dhangars – a nomadic pastoral community in a developing agricultural
environment. In Pastoralists and nomads in south Asia, L. S. Leshnik & G. D. Sontheimer (eds),
139–70. Wiesbaden: Otto Harrasowitz.
Sundara, A. 1968. Protohistoric sites in Bijapur district. Journal of Karnatak University (Social
Sciences) 4, 2–23.
Sundara, A. 1969–70. A new type of Neolithic burial in Terdal in Mysore State. Puratattva 3, 23–33.
Sundara, A. 1970–71. Neolithic cultural patterns and movements in north Mysore State, Journal of
Karnatak University (Social Sciences) 6, 3–12, and 7, 1–8.
16 The changing role of reindeer in the life
of the Sámi
PEKKA AIKIO

A personal view from my childhood

In my childhood, the reindeer was the most familiar animal. At home were
also cows, a cat, many dogs, and later horses. But reindeer were the most
common animals of all. This was quite natural because my childhood home
was located far away from other houses in the forest. The nearest neighbour,
the post office, the store, and the school were a 10 km trip, and there was no
road.
Actually, all the reindeer which I knew as a little boy were domesticated
and tame. They were not tied to a tree and were allowed to roam freely in the
coniferous forest close by our home.
All of the members of our family had their own reindeer, which were
marked with the owner’s own mark (Fig. 16.1). This was made by notching
the ears of the reindeer with a knife. My mother owned her own semi-
domesticated draught reindeer; my grandmother, or in the Sâmi language
ahku, owned several; my younger uncles had their own reindeer too, but
their reindeer were half wild and untame. My father owned most of all.
Every draught reindeer was given his own ‘personal’ name. It was
important to know the individual reindeer from the rest of the herd. When
Mum or Grandma (Fig. 16.2) went on a trip, they themselves had to go and
separate their own draught reindeer from the herd. It would have been bad
manners ‘to help’ them. It was also clear that each person used his own
draught reindeer when he made a trip. In Sâmi families, the women were
always economically independent of the men, and this was especially
emphasized by the fact that the women, like everyone else, used only
reindeer that they personally owned.
My youngest uncle, Niilo, helped me to list the draught reindeer we
owned in the 1950s. The list included 66 names (see Table 16.1) of which
only five or six had Sâmi language names. All the rest were Finnish.
My earliest memories concerning the 30 semi-domesticated draught
reindeer which we had at home come from the time when I was four or five
years old (I was born in 1944). My father was extremely overjoyed when a
son (myself) was born to him, and he notched my personal reindeer
ownership mark on the biggest and strongest male reindeer calf that he
owned. Later this reindeer was castrated. When I was five years old, this
reindeer, called Sloagga, was the master of the herd. My special job was to
prepare a treat every day out of rye flour and water which was only served to
Sloagga. This leader of the herd with his big antlers always remembered
this, and every day during winter, he came with his herd to pay us a visit in
our yard. While I served Sloagga his special food, all the other reindeer were
given horsetails, Equisetum sp., which they ate with great relish. The
reindeer passed a couple of hours in our yard and then went back to feeding
on lichen in the nearby forest.
 Figure 16.1 The author aged 12 (1956) with a draught reindeer called Pehkossuivakko.

Even in my childhood in the 1950s, the reindeer was the companion of the
Sarni. The reindeer were, as a general rule, domesticated. Of course, there
were wild reindeer in the herd. But the herd as a whole was not nearly as
wild or undomesticated as it is at present.
 Figure 16.2 The author’s grandmother wearing a reindeer coat and carrying the author’s little sister,
springtime, 1956.

There are many examples of how an individual reindeer showed its

companionship for a man or a woman or for another reindeer. My Uncle
Jouni (Fig. 16.3) told the story of one female reindeer who got up from her
sleeping place in the dead of winter, walked some tens of metres to a draught
reindeer with an injured leg, and began to lick and care for the other
reindeer’s wounds. My uncle watched this and observed that the draught
reindeer’s injury got better.
When the wolf pack howled nearby, the whole herd would press together
near the kota (the conical Sâmi tent) and its fire in order to gain some
protection. The draught reindeer always stuck close to the Sâmi’s home, and
if the reindeer travelled away from this home, it was with reluctance.
However, when they were headed back again home, their hooves really flew!

Table 16.1 List of the names of the draught reindeer of the author’s family in the 1950s
Akselin lainakko Musikki
Aplikki Myrre
Hiljanen Mäenpään nulppo
Hiljanen nulppo Nalhta
Hirrosarvi Njaiti
Hirvensarvi Nulppo
Isohärkä Nylhtä
Isokelosarvi Näppärä
Isopälli Oma pikkuhärkä
Itsellinen Onkisarvi
Jantikka Paha nulppo
Jekke Palonojan merkkinen
Julie Pehkossuivakko
Juovoja Pekan härkä
Jyty Pekan pikkuhärkä
Kartanon Jaakko Pikkukello
Keampa Pikkumusikki
Kelosarvi Pikkusuivakko
Kilkuttaja Pälli
Kumppari Rymysuivakko
Kusilukkari Silmäpuoli
Källi Sloagga
Körri Stuokki
Laiska Suivakko I
Laiskanulppo Suivakko II
Lankakorva Tassukka
Laukki Toinen pikkuhärkä
Laukunkantaja Tonnari
 Loikkari Tulikäpälä
Luosto Valkko
Mitätôn härkä Vanha suivakko
Musta härkä Viikari
Musta nulppo Äidin pikkuhärkä

When a draught reindeer grew old, it was slaughtered and eaten. There
were never any tears. If we didn’t eat it, the wolves surely would, and why
feed the wolves?
A big herd of reindeer always meant more to the Sâmi than an individual
reindeer. The Sâmi’s close relationship with the big reindeer herd was
always brought forth in the oral tradition. A big reindeer herd gave people a
sense of security because a large herd buffered the family from any
catastrophe which might befall the herd (Fig. 16.4). There was a wonderful
balance between the lives of the people and their herds. A large reindeer
herd meant that there would most likely be food on the table and clothes to
wear, even if, in a catastrophic situation, some of the reindeer died. The
emphasis on the economic aspect of large-scale reindeer-herding in
ethnographic research tends to give a rather one-sided picture of the
relationship between the Sâmi and their reindeer.
 Figure 16.3 The author’s Uncle Jouni, with his dog, at a reindeer-calf marking site, in front of a turf
hut, in the 1950s.

Even at the end of the 1950s, the reindeers’ position as companions to the
Sâmi was unshakable. Then in our fifty-family reindeer-herding association,
there were hundreds of semi-domesticated draught reindeer. Traditionally,
the Sâmi herders have owned many draught reindeer. According to statistics,
of all the draught reindeer that were not slaughtered, 55–60 per cent
belonged to Sâmi owners. A little boy like myself didn’t think twice about it;
it was the way it had always been and would always be. But in the 1960s
dramatic changes took place, of which the so-called ‘snowmobile revolution’
was not the least. In the 1970s, there were about 2300 snowmobiles in
Finland, Sweden, and Norway (NKK 1981). In other words, one snowmobile
per 300 reindeer. In addition, a lot of other modern technical aids began to be
used in reindeer-herding, including walkie-talkies, cars, tractors, aeroplanes,
and, in Sweden, even helicopters. Today, our 13-year-old son has only one
semi-domesticated draught reindeer, and in the whole reindeer-herding
association to which belong, there are only ten draught reindeer – not for the
Sâmi’s own use but only for the tourists. During the past ten years, the racing
of draught reindeer has started again. Reindeer are used for driving tourists,
but, first of all, the reindeer races have become very popular. It looks as
though this enthusiasm for racing will save the tradition of racing draught
reindeer and will preserve this very valuable skill. Times have changed.

Figure 16.4 Reindeer in an autumn round-up, 1977.

 Of course, a reindeer-herding family owned more reindeer other than
merely draught reindeer. In general, the entire herd consisted of 10–20 per
cent draught reindeer.

The origins of reindeer-herding

Reindeer-herding originated in Eurasia. Traditionally in Eurasia, reindeer-

herding has taken place in the entire region north of the 0° isotherm. Of the
total 4.5 million reindeer in the world, 3 million are domesticated (Andrejev
1977). Of these 3 million, 77 per cent are in the Soviet Union and 21 per
cent in Finland, Sweden, and Norway. A smaller number of reindeer live in
North America, Scotland, Greenland, Iceland, and on South Georgia Island
in the Southern Hemisphere near the Falkland Islands.
There is hardly any information on reindeer-herding or on reindeer
themselves from ancient times. The small amount of written data and their
direct absence tempt one to assume that reindeer-herding, at least in some
form, is a quite new phenomenon. However, cultural researchers (e.g.
Birket-Smith 1972) warn us not to base too much on written sources or, at
least, the lack of such sources. Reindeer-herding arose as the product of a
widespread Eurasian Arctic culture. The Arctic peoples did not generally
leave copious literary traces. The oldest historical information on reindeer-
herding dates from the year AD 499. The annals of the Liang dynasty of
China tell of the mythical land of Fu-sang where the inhabitants used moose-
like animals (in other words, reindeer) as draught animals and for milk
(Laufer 1917). The Italian explorer, Marco Polo, who visited Mongolia at
the end of the 1200s, also tells of a people who rode mooselike animals. In
1302, the Persian historian, Rashideddi, further tells of a reindeer-herding,
nomadic people living in the area around Lake Baikal. This was a people
who Wiklund (1947) believed were one and the same as the Soyot or Tungus
people, relatives of the Samoyed.
Quite well known is the story of Ohthere’s report to the English King,
Alfred the Great, in the year 892. Ohthere himself owned 600 reindeer, of
which six were decoy animals. Ohthere’s report is considered to be the first
preserved written reference to reindeer-herding in Scandinavia. But Wiklund
estimates (1947), based on old linguistic features, that the Sâmi people had
already engaged in reindeer-herding even before this – perhaps a long time
before this.
Many linguistic and ethnologic factors have pointed to the conclusion that
reindeer-herding originally began in one place and then spread elsewhere.
Birket-Smith (1972) notes Kai Donner’s opinion that reindeer-herding could
have begun with the Samoyed people at the beginning of the Bronze Age
and was fully developed long before our written history. Some common
features regarding reindeer-herding exist throughout the whole of Eurasia,
such as the use of the lasso, notching the ears of the reindeer to make
ownership marks, the use of skis, and the castration of reindeer by biting.
Recently, support has been given to Wiklund’s proposal (1947) that
reindeer-herding developed in several locations independently of one
another.
Man’s earliest contact with reindeer, and in its original form of wild
reindeer, occurred in the form of hunting which took place by means of
traps, pits, and trenches, as well as by fences or corrals. A decoy animal was
used to capture wild reindeer.
Everywhere in Finland are found signs of a wild-reindeer-hunting culture
reaching far back into the distant past. For example, there are systems of
trenches, complete with different types of fences, that were used in hunting.
There are also examples of isolated fences used for the same purpose.
It is not possible to say with certainty how long ago these types of hunting
were used. But it is, in any case, certain that long before our system of
reckoning time began, people were hunting reindeer in these ways. As an
example of this, Helskog (1977) mentions one rock drawing located at
Jiebmaluok’ta near Alta, Norway (Fig. 16.5). The drawing shows a corral
area with the shape of a four-leaf clover. Inside the corral are drawn reindeer,
elk, and perhaps also boats. Many reindeer are coming inside the corral
through an opening. The hunter, with spear in hand, is pursuing the reindeer
into one corner. These rock drawings are situated 23.5–24.5 m above sea
level, and from this elevation, it is possible to calculate that the age of the
drawings is between 5500 and 6000 years. In other words, the ancient
inhabitants of this area had already developed, about 3500 years BC, a
hunting and entrapping culture based on making use of indigenous reindeer.
With the possibility of boats in the drawing, fishing can also be postulated.
 Figure 16.5 The rock drawing at Jiebmaluok’ta, Alta (Helskog 1977).

In the 1970s, in the northwest corner of Finland, near the city of Tornio, a
piece of a reindeer horn was found which was dated as being 34 000 years
old (Fig. 16.6).
The graphic terminology used by the Sâmi people to describe reindeer and
its way of life is also clearly very old. Already in the time of wild-reindeer
hunting, it was possible to use words to describe exactly the structure of the
wild-reindeer herd and the different types of reindeer. Different researchers
have collected close to 1000 terms related to the reindeer (for example,
Itkonen 1948).
Quite often it is emphasized that the completely nomadic way of life
based on reindeer-herding and practised by the Scandinavian reindeer Sâmi
people did not come about until the end of the Middle Ages. The definition
between total and semi-nomadism should not lead, by means of conceptual
obscurity, to uncertainty because long before the appearance of the so-called
large-herd reindeer-herding, there already existed an advanced multiplicity
of ways to herd reindeer. In the polar regions of Eurasia, the wild reindeer
was already an important source of game for the inhabitants, and,
subsequently, the practice of herding tame reindeer appeared. Reindeer-
herding, in so far as the anthropologists are concerned, began in a state of
confusion and inconsistency. Samuli Aikio (1977) has also remarked on the
reasons for the mystery which surrounds research on reindeer-herding with
obscurity, and states that even the history of large-herd reindeer-herding is,
in part, difficult to interpret. Nevertheless, it is certain that even from the
beginning, reindeer-herding has been characterized by many variations in
customs and in the many different ways of herding reindeer. This diversity
can be seen in modern Finnish herding as well.
 Figure 16.6 Therr Views of the Tornio Antler (Siivonen 1975).

The first written evidence of the keeping of large herds is to be found in

the tax rolls of the 1500–1600s. On the other hand, it is known that archers
(in other words, those who paid the bow tax) in Pohjois-Pohjanmaa in the
north of Finland also paid church tithes according to the tithe statute of 1335.
The tithe was every tenth reindeer calf, or one tax unit for each calf
(Tengström 1820–22).
 Figure 16.7 Present-day reindeer-herding areas in Norway, Sweden, and Finland. The shaded area
denotes Sámi settlement. The dotted line represents the southern boundaries of the reindeer-herding
areas.

Nickul (1970) displayed a more uncertain attitude than Wiklund toward

the ability of the tax rolls to give information on reindeer-herding. He found
that conclusions about the abundance of reindeer could not be drawn from
taxes paid in reindeer hides, because the Sâmis needed reindeer hides for
their own use and it was cheaper to pay taxes from the hides of other hunted
animals. Fish could also be used to pay taxes. Nickul continued by saying
that if there was no other way to pay the tax, then people resorted to paying
with reindeer.

Reindeer-herding under the pressure of change

Figure 16.7 shows the reindeer-herding areas in Finland, Sweden, and

Norway. These areas are almost the same as those where the Sâmi people
live. Only in Finland is the Sâmi area smaller than the reindeer-herding area.
During the past 30 years, great technological and economic changes have
affected reindeer-herding. From the beginning of the 1960s, perhaps the
most dramatic change in reindeer-herding occurred – the appearance of the
snowmobile.
The Sâmi language has lost its position as the reindeer-herding language.
Reindeer-herders are becoming Finnicized. Formerly, reindeer-herders used
almost 1000 Sâmi terms describing reindeer. The present-day young herder
masters hardly ten Sâmi terms. At the southern border of the Sâmi home
territory in Finland, the Sâmi language in the 1950s was almost completely
replaced by Finnish, even in reindeer-herding (see Table 16.1). The
surprising fact is that reindeer-herding does not preserve the Sâmi language.
This is in contradiction to the commonly held view (Aikio, M. 1984).
Draught reindeer are no longer needed for transportation. Fortunately, the
skill of how to raise and train a draught reindeer has been preserved because
these reindeer are used by tourists and for racing. Without these modern uses
for the draught reindeer, they would have totally vanished, along with the
skill to train them. Also, the use of the special reindeer- herding dog has
dramatically decreased.
The present-day reindeer-herding legislation is based on the point of view
of cattle-raising and agriculture. Reindeer-herding is viewed as a part of
agriculture, and all plans for the development of reindeer-herding are
modelled after cattle-raising. The primary purpose of reindeer-herding, in
this view, is for meat production. Here lies an obvious contradiction: the
reindeer is a small animal; its carcass weighs under 25 kg (Aikio, P. 1978).
Reindeer-herding is perhaps the most efficient way to utilize the scant
natural resources of the northern terrestrial ecosystem. This can be seen
when comparing the amount of reindeer meat with other meat produced in
three different zones (as defined by the author) of the Finnish reindeer-
herding area (Fig. 16.8). In the northernmost zone (i.e. the Sâmi home
territory), the total amount of reindeer meat produced equals 90 per cent of
total meat production. In the southernmost zone (i.e. the agricultural and
forestry area), this is reversed. When thinking of the entire Finnish reindeer-
herding areas, the reindeer thrives equally well in the northern and southern
areas. However, agriculture and forestry fares best in the south.
During the past 25 years, considerable changes have occurred in the
structure of the reindeer herd. Because the reindeer has become a part of the
meat-producing machinery of the market economy, man has tended to
maximize production. As a result, the proportion of adult females in the
Finnish reindeer herd has increased substantially, as has the proportion of
slaughtered calves (Fig. 16.9). This did not happen in traditional Sâmi
reindeer-herding.
 Figure 16.8 The produced reindeer meat (hatched square) and other meat (dotted square) in 1980 in
three different zones of the Finnish reindeer-herding area. Zone I indicates the Sámi home territory in
the north, which consists of fells and protected forests. Zone II indicates the area where meat
production from reindeerherding and agriculture is equal. Zone III indicates the agricultural and
forestry area. The figures under the squares give meat production in metric tons.

Figure 16.9 The curves symbolize the structure of the Finnish reindeer herd from 1961–86. (a) The
percentage of females in the adult reindeer herd; (b) the percentage of calves born; (c) the percentage
of calves in the entire reindeer population; (d) the percentage of slaughtered calves out of the total
number of reindeer slaughtered.

The reindeer owned by the Sâmi graze on natural lichen pastures. We have
not, to date, developed an extra source of food for the reindeer which would
be economical and useful in the fell areas, and which could be distributed to
large herds. As a result, we cannot increase the number of reindeer above the
present figure in Finland and in the other Nordic countries. This means that
the reindeer is an insignificant source of meat in terms of national
production. The authorities in charge of development actions view the
reindeer only as a meat-producing domesticated animal. This view is not
completely valid. In reality, the reindeer is a half-wild close relative of the
North American caribou. In an agrarian society in the coniferous forest zone,
the reindeer could have a totally different role than in a sub-arctic fell area or
at the forest limit, where the climate is a limiting factor to other human
activities.
Historically, reindeer-herding is part of the Lapp livelihood, along with
hunting and fishing. Reindeer-herding is successfully combined with other
occupations in the sub-arctic environment, which has small primary
production. This can be seen in the research carried out by Eino Siuruainen
(1976). Based on this research, I have made a diagram which plots the
number of reindeer owned against different occupational groups (Fig.
16.10). In Finland at the beginning of the 1970s, according to this diagram,
farmers, fishermen, forestry workers, and those working in the service
industries owned a small percentage of the total number of reindeer.
However, by far the largest share of the reindeer herd was owned by
reindeer-herders.
In the Nordic countries in the 1970s, and especially in Finland, reindeer-
herding developed in the direction of modern deer-hunting. The reindeer as
an animal has become wilder and has significance mainly as an animal to be
slaughtered. In North America, reindeer-herding was introduced as a means
to supplement the diet of the native people. This attempt has encountered
many difficulties because there they do not have the thousands of years of
traditional reindeer-herding experience behind them as is to be found in
Eurasia. In North America, the role of the reindeer was to be an animal for
hunting.
 Figure 16.10 The number of reindeer owned plotted against different occupational groups.
Professional group of the head of the household: (1) farmer; (2) reindeer-herder; (3) fisherman; (4)
forestry worker; (5) contractor; (6) labourer; (7) service-industry worker; (8) businessman; (9) other
occupations; (10) unknown.
Concluding remarks

The traditional Sarni reindeer-herding culture, including all its

characteristics, has been developed over thousands of years to be highly
specialized in its know-how and technology. The modern urban society has
never understood the real nature and significance of this type of reindeer-
herding. Perhaps it is for this reason that the outside world has
misunderstood and misinterpreted the history and development of reindeer-
herding.
Could reindeer-herding be as young as it often is presented in the
scientific literature? Can scientific research be so limited that it only leans on
written sources? There are no written documents on the piece of reindeer
antler which was found near the city of Tornio (Fig. 16.6) and which is 34
000 years old. However, a Sâmi reindeer-herder who saw a drawing of this
antler concluded that it had belonged to a castrated male reindeer. If this is
true, then, this reindeer belonged to a reindeer-herding system and was not
wild. Researchers have almost always seen spears and other weapons in old
rock drawings. What would happen if the human figures in the rock
drawings in the city of Alta (Fig. 16.5) really represent people with
instruments for capturing animals and not for killing them? Why couldn’t
the reindeer, like the caribou, have migrated thousands of kilometres from
the Mediterranean to the north of Europe and back every year during the
latter part of the Ice Age? Perhaps the rock drawings and cave paintings of
middle and northern Europe, together with unprejudiced archaeological
research, can provide new and surprising results in the historical study of
reindeer and the Sâmi people.

Acknowledgement
I wish to thank Nick Gardner for his help in editing and translating this text.

References
Aikio, M. 1984. The position and use of the Sâmi language: historical, contemporary, and future
perspectives. Journal of Multilingual and Multicultural Development 5(3&4), 277–91.
Aikio, P. 1978. Reindeer herding in Finland 1975–76: production and labor. Nordia 12, 273–82. Oulu.
Aikio, S. 1977. Some minor remarks on the economic history of the Sâmi [in Finnish]. Suomen
antropologi 2, 91–4.
Andrejev, V. N. 1977. World reindeer herding: its structure and areal distribution. Summary by Elis
Pâlsson [in Finnish]. Ottar 101, 12–18. The University of Tromso.
Birket-Smith, K. 1972. Kulttuurin tiet (Cultural roads). Porvoo, Helsinki: WSOY.
Helskog, K. 1977. Reindeer round-up in the stone age [in Finnish]. Ottar 101, 25–9. The University of
Tromso.
Itkonen, T. I. 1948. Suomen lappalaiset vuoteen 1945 (Finnish Lapps to 1945). Helsinki: WSOY.
Laufer, B. 1917. The reindeer and its domestication. Memoirs of the American Anthropological
Association 4 (2). New York: Kraus Reprint Corp., 1964.
Nickul, K. 1970. Saamelaiset kansana ja kansalaisina (The Sâmi as a people and citizens). Helsinki:
Suomalaisen kirjallisuuden seuran toimituksia (Publication of the Finnish literature society) 297.
NKK 1981.Reindriftsnaeringen pa Nordkalotten, Nordkalottkommittens publik-ationsserie (Reindeer
economy in the North Calotte, A North Calotte Committee publication) no. 12.
Siivonen, L. 1975. New results on the history and taxonomy of the mountain, forest, and domestic
reindeer in northern Europe. Proceedings of the First International Reindeer and Caribou
Symposium, Fairbanks, Alaska, 33–40.
Siuruainen, E. 1976. The population in the Sâmi area of Finnish Lapland. A regional study with
special emphasis on rates and sources of income.Acta Unwersitatis Ouluensis Series A 40, Oulu,
Geographica 2, 1–138.
Tengström, J. 1820–22. Afhandling om jpresterlinga tjenstgöringen och aflöningen i Abo Erke-Stift,
Utgifven af Domkapitlet i Abo, delen 1, Abo (A study of the tasks of the priests and their
remuneration in the Bishopric of Turku, part 1. Turku: The Judicial Church District of Turku.
Wiklund, K. B. 1947. Lapparna (The Lapps). Stockholm: Bonniers.
17 The geographical distribution and
function of sheep flock leaders: a
cultural aspect of the man–domesticated
animal relationship in southwestern
Eurasia
YUTAKA TANI

Introduction

On a medallion of the 17th century by Zincgreff Julius Wilhelm (Henckel &

Schöne 1967, pp. 537–8), there is a scene of a flock of sheep leaping down
into a river from a bridge. Four sheep are already in the water. Two are
jumping and the others stand on the bridge waiting their turn. The first one,
with large strongly curved horns and a bell around his neck, is in the
forefront in the stream (Fig. 17.1).
The medallion illustrates the ideal relationship between the leader and his
followers with the allegory of the flock of sheep, with the ewes following the
male. From the comments accompanying the medallion we know that this
leading male is a wether: a castrated male.
Sheep have a natural disposition to follow and the flock’s direction of
movement is consequently determined in general by the leading group.
Sheep are timid when they face a gap, or a stream, or something dangerous.
When the leading group stops, the followers jostling with each other begin to
flow sideways like a stream of water intercepted. On such occasions, the
flock sometimes divides into two, or it changes direction and does not follow
the course planned by the shepherd, who may adopt several techniques of
control during his herding activities in order to make the flock travel and
graze along the route he has planned, without dividing or losing members.
The most common and simplest action is to menace a flock by running at
it. By shouting and running at the sheep, however, all that a shepherd can
accomplish is to chase several sheep close to him or make the stray sheep
return to the main group. A sheepdog can chase stray sheep back to the main
group, and it can change the direction of a flock, but unless it is intensively
trained, its effectiveness remains relative. Goats mixed in the flock tend to
run in front, due to their recklessness and rashness, and stimulate the sheep
that follow, but they are too capricious to keep pace with the sheep. Another
way to control a flock on the move is to utilize a natural leader in the flock
itself, if such an individual exists, and with training the shepherd can control
the movements of the whole flock by means of vocal commands to the flock
leader. If the leader is female, there is no problem, but if the leader is male,
sexual activity becomes a problem. Sexually active rams may not obey the
shepherd, and so he separates the breeding rams from the ewe group during
the summer to prevent them from mounting at will. In order to utilize the
male flock leader, it is preferable to castrate him. The medallion mentioned
above shows that the western European shepherd has used a castrated male
as the flock leader for at least the last 300 years.
 Figure 17.1 Medallion by Zincgreff Julius Wilhelm, 17th century (Henckel &Schöne 1967).

It is, however, not only in western Europe but also in the Mediterranean
area and in the Middle East that shepherds utilize castrated males as flock
leaders. In Abruzzo, central Italy, where I carried out field research, this kind
of guide wether (guidarello in Italian) is called manziero. Every year, the
shepherd selects a few breeding males from the male yearlings, most of
whom are slaughtered within a year. From among these selected young rams,
when they reach two years of age, the shepherd picks one as a candidate for
manziero and castrates him.
The castrated male holds a unique position in terms of the man-domestic
animal relationship. So far as the shepherd views the castrated male leader as
a member of the flock, he belongs to the category of the dominated group.
However, being castrated, he is alienated from being an essential element of
the flock’s reproductive process. Moreover, being instructed to respond to
the shepherd’s vocal commands, he works as an agent of the shepherd, and
in this context he belongs to the dominating shepherd’s side. By sacrificing
his sexual capability and learning the shepherd’s vocal orders, he obtains the
role of the leader of the flock (dominated) and at the same time the role of
the agent for the shepherd (dominator). The sacrifice of his sexual power by
castration allows the wether to take the position of mediator between the
dominator and the dominated. If most of the flock is female, it means that he
takes the role of the guardian of the female group. We can see that the role of
the guide wether is functionally analogous to the eunuch.
The eunuch can be defined as the mediator between the emperor and the
people, and transmitter of the emperor’s commands and will to them. The
eunuch was also the guardian of the harem in many imperial courts. The
eunuch took the function of mediator between the dominator and the
dominated.
We do not know where and when the idea of the eunuch originated.
However, the position of the eunuch in human political institutions
corresponded to that of the castrated male flock leader in the shepherd-flock
relationship. Behind the idea of eunuch that had been adopted in the imperial
courts in ancient times, can we suppose an analogy in the techniques of the
guide wether to political institutions?
As shown later, however, the geographical distribution of the utilization of
the guide wether is not world-wide. In some areas, shepherds adopt different
methods to control the flock. They utilize the female sheep instead of the
guide wether (from now on these leader sheep used for herd control,
regardless of sex and other conditions, will be termed ‘flock leader’ or
‘guide sheep’). Any herd control technique can be regarded as a pastoralist
people’s cultural expression of how they look upon and treat their domestic
animals. Moreover, if some of their cognitive or technical features were
analogously applied to the management of man or vice versa, we can
discover their hidden way of thinking on the interrelationship between man
and his domestic animals. Has the eunuch, as apolitical institution, a certain
relevance to the guide wether technique from the cultural and historical point
of view?
In my field survey on the man-animal relationships of pastoralists it was
one of the main purposes to make clear the geographical distribution of the
guide wether techniques. But, during the field survey, I found that other
kinds of flock leader are utilized in certain areas. Moreover, it was known
that some pastoralists place multiple female guide sheep in a flock, and
others utilize both types of flock leader, male and female, in the same flock.
This multiple leadership raises questions of why they put multiple leaders in
one flock and how can the multiple leaders guide the flock harmoniously in
one direction. Questions to be asked are:

(a) What types of flock leader are utilized for herd control?
(b) What is the function of these flock leaders in herd control?
(c) How can we interpret the meaning of the multiple leadership?
(d) What is the position of the guide wether type of herd control, historically
and technically, in comparison with the other types of flock leader?
(e) In which areas and what periods did each type of herd control come into
existence?

Flock leader types and their geographical distribution

The use of a guide wether for flock control, as mentioned above, was seen in
western Europe as well as in central Italy. According to the shepherds in
Abruzzo in central Italy, they train the guide wether (manziero) as follows.
After castration, the shepherd puts a short woollen rope on the wether’s neck
and walks along with him, like walking a dog, and teaches the wether to
respond to his touch. When the wether becomes habituated to him, the
shepherd begins to instruct the wether on how to respond to the vocal
commands: stop (ferma), advance (< avanti), come (vieni). To teach the
wether what each vocal command means, the shepherd shouts the commands
and pulls the rope by hand in a way to make the wether learn how he should
react. The shepherd gently pats him or gives him something to eat whenever
he behaves properly. In the next stage, after the castrated ram has mastered
how to respond to vocal commands, the shepherd puts a long woollen rope
around his neck and puts him in the midst of the flock in the grazing-field,
and continues to train the wether until this novice manziero behaves
obediently even in the flock. Having finished this training the shepherd takes
off the rope. Now, without any physical control, the trained wether (true
manziero) goes ahead or stops or returns upon receiving the shepherd’s vocal
commands, even from a long distance.
The shepherds consider the manziero to be the leader of the flock. They
say that the manziero goes in advance and dares to dash against difficulties,
following the shepherd’s vocal order. In the winter time, when the sheep
must move through the snow, he makes a path in advance of the flock (for
more details, see Tani 1976, 1977).
The shepherd gives the manziero a special personified name, for example
Generale, Capitano, Mussolini, etc. On the other hand, the ewTes do not
have any proper names. Each ewe is nominally referred to according to the
classificatory term based on her body colour patterns. The shepherds set
great value in their guide wether manziero and the theft of a guide wether
leads to a vendetta.
Regarding the pastoralist activities of central Italy, two types of
organization can be found (Tani 1977).

(a) Non-professional – each family in the mountain village usually keeps

20–30 sheep for the daily supply of milk. The owners of these sheep take it
in turns to herd them every few days.
(b) Professional – the professional shepherd keeps in trust other owners’
sheep under contract and takes care of them, forming a joint flock with his
own sheep. Usually the herder has a family partner; his brother or son, who
always attends to the flock which includes his own sheep.

The manziero is usually found in the latter type of herding organization.

In Greece, the Sarakatchani shepherds in Ipiros and Thessalia use a guide
wether for herd control. They call him kriari ghisemia. A ram of more than
two years of age is chosen and castrated. They interestingly adopt the verb
euvuxlşw (to make eunuch) for the act of castration. The shepherd gives him
a proper name and he is trained to react to the shepherd’s vocal orders.
Interestingly, in each flock of 300–500 head, two or three kriari ghisemia are
mixed in. Moreover, as mentioned later, the Sarakatchani use other kinds of
sheep for guiding the flock. Why do they have other kinds of guide sheep
besides guide wethers? Do each of these two categories of guide sheep carry
out different functions corresponding to the distinct herd-control situation?
Apart from these questions, it is certain that the Sarakatchani utilize the
guide wether. All Sarakatchanis visited by the author were professional
pastoralists. They keep their own sheep and migrate with the family.
In Crete as well, the shepherd raises a castrated sheep as a guide wether. A
castrated goat is sometimes used instead of a sheep. Every year, the shepherd
castrates four or five males aged 2–3 months. They do not have any
systematic method for training the leader as observed in central Italy. They
believe that shouting, beating, and throwing stones accompanied by vocal
commands are sufficient to make wethers learn to lead the flock. They also
greatly value their own guide wether, and the theft of a guide wether can be
the cause of vendettas.
Among the nomadic Yöluk of Anatolia (who are of central Asian origin) it
is reported that there is no evidence for the utilization of the guide wether
(Matsubara 1983). On the other hand, the nomadic Kurds near Hakkari in
southeastern Turkey use a wether as flock leader. Usually a two-year-old ram
is castrated and put in the flock after training by the same method as in
central Italy. Sometimes a castrated he-goat (teke) is also used. They said it
was usual to place only one guide wether in a flock. In Iran, according to
Digard, the shepherd of nomadic Baxtyari also raise the guide wether
(Digard 1981, p. 57).
Going further east towards Afghanistan, data were collected on several
pastoralists of different ethnic groups; the Kandahari of Durrani Pashtun, the
Uzbecki, the Arabi, and Shaghni of mountain Tajik, who ascend every
summer to the Shewa high plateau (Dasht Shewa) in Badakhshan (see details
in Tani & Matsui 1980). The first three groups are nomadic and the last, the
Shaghni, are sedentary farmers in the villages along the Pamir River, whose
shepherds come there seasonally for transhumance. The Shaghni do not use
the guide wether, but the other three pastoralist groups do. The only
difference from the Mediterranean , pastoralists is that they castrate not the
ram but the he-goat. The Arabi select and castrate one-year-old he-goats.
They said that such he-goats can be used for about ten years as leaders. The
Arabi and the Uzbecki call the he-goat leader sarkka and the Kandahari of
Pashtun call it mukhi. In the summer quarter, the Kandahari shepherds do the
night-time herding, not individually but collectively. Looking at the
formation of the flock and its direction of movement, two shepherds, one in
the front and another at the back, approach the flock with vocal calling and
whistling. This guiding method reminds us of that described by Baskin as a
primordial herding-control method (Baskin 1974, p. 530). The castrated he-
goat is usually bought by the owner of the flock (tsessten) in the market. The
shepherd (shpun) is in charge of the flock and is employed by tsessten with
an annual contract. His position as a shepherd is not stable. The he-goat is
not given any proper name. Between the shepherd and the he-goat, there is
not any individual command-to-obey relationship.
Further east, in the northeastern territory of India (Kashmir), the
pastoralists of Gujar Bakkalwala (goat-herders) and the Kashmir shepherds
(sheep-herders) did not know about the use of the guide wether, and they do
not care for the flock with as much attention as observed among the
shepherds in the areas mentioned above. It was seldom that their flocks were
followed by any shepherds in the daytime. We have no information about the
Tibetan pastoralists in the Ladakh and the Chinese territory, except for the
northern Zanskar area of the Ladakh which I visited, and they did not use a
guide wether. Most goat-owners of the Bakkalwala, even though they
migrate in the summer quarter with their families, do not attend to the flock,
but hire a servant-shepherd by contract. In the Zanskar of the Ladakhi, on
the other hand, the village women attend, in turn, to the flock collected from
the villagers, as well as having their own sheep amongst the flock.
Though more evidence should be collected from different places, to fill
the gaps between the areas visited, it can be said that the use of the guide
wether or castrated goat for herd control is found along a southern belt from
West Europe to Afghanistan, with the exception of areas where pastoralists
in Turkey have migrated from the north.
Now, turning north from the Mediterranean area, the data on the
Romanian pastoralists will be presented. I visited the following three places:
the southeastern plains of Dobrogea near the Black Sea (around Topalu), the
southern Carpathian mountains in the province of Sibiu (around Tilişca), and
the northern Carpathian mountains in the province of Bistrija and Nasaud
(Ardan, şebis, and Sieu£: see details of my field studies in Tani 1982, 1987).
In Dobrogea, the guide wether used for herd control is called batal. The
shepherd often sleeps using as a pillow a lamb chosen to be a batal in the
future, to form a familiar relationship. In the spring when the lamb is one or
two years old, the shepherd castrates him. A special proper name is given to
him and he is always called by name. By often feeding mamaliga (boiled
maize paste) to the lamb, it is trained to follow the shepherd and respond to
commands. To enhance the appearance of the guide wether, the shepherds
even try to correct the curve of the horns using the hot mamaliga. They said
that only one guide wether is usually placed in a flock. But in the flocks
observed, the shepherds often pointed out three or four guide wethers.
In southern Carpathia (province of Sibiu) also, the use of a batal was
observed, although it was not found in all the flocks. The shepherds said the
batal is only placed in the large flocks, and usually only one in each flock.
The proper names given to the batal are often taken from human names.
According to the etymological dictionary by Hasedeu, the term for the guide
wether batal is derived from the Greek ‘ßaxaX,Xog\ meaning ‘castrated
berber’. The connotation ‘guide wether’ must be added to this term as an
extension of the original meaning. The shepherds at Tilişca, who seasonally
migrated to Dobrogea until the socialist revolution, said that the use of batal
was introduced from Dobrogea. In Dobrogea, trade with Greek merchants
for milk products had been frequent. Taking into account the Greek origin of
the term batal, it is highly probable that the utilization of the guide wether
was introduced from the southern Balkan peninsula.
The shepherd is a professional who joins up the sheep of the village
farmers in his flock, as well as having a large number of his own sheep.
Generally speaking, the herding team is composed of eight shepherds. Half
of them take in turn the charge of herding each week.
Further north in Romania, the utilization of the guide wether batal
completely disappears. In the province of Bistriţa and Nasaud, the shepherds
know of the batal but never use it. Instead of the castrated male, they have
another flock leader that is, interestingly, female. This guide ewe is called
fruntaşa which literally means ‘the female leader that goes in front’.
 There are always some young ewes that approach the shepherds without
fear. The shepherds mark these and begin to tame them by calling their
names and feeding mamaliga to them every morning. After the marked
females have learned their own names, the shepherds teach the meaning of
vocal callings and whistles. The names of address are not any special proper
name but simple classificatory terms corresponding to the colour pattern of
each sheep. During the summer the flock is usually composed of the sheep
owned by several different shepherds, who take care of the flock in turns for
one week. If you ask them to point out the fruntaşa, they indicate about ten
fruntaşas in one flock. Why are so many leaders put in one flock? How do
they use them to manipulate the movement of the flock? Apart from these
questions, which will be discussed in the following section, the main point is
that in northern Carpathia (Romania) they use the female as a flock leader
instead of the castrate1 male.
The labour organization is just the same as that of southern Carpathia.
When I first asked the shepherds if they utilized castrated males as flock
leaders, some denied their use with a disparaging tone, even though some of
them knew that southern Carpathian shepherds use wethers. It is not clear if
this repugnance is derived from a cultural base, such as, for example, the
importance of the feminine or low evaluation of the status of the ‘castrated’.
Or, are these two types of flock leaders, male and female, incompatible for a
technical reason? Evidence against this is found among the southern
Carpathian and Dobrogean shepherds, who use the guide wether, but, at the
same time, raise the fruntaşa type of leader and use her in parallel with the
batal. They call her, instead of fruntaşa, cirmace which means ‘the one to be
followed’. The training method for the female guide sheep is the same as
that in northern Carpathia. The coexistence of two kinds of leadership here,
however, raises the following question. If both guide wether and guide ewe
carry out the same leadership role, which do the other flock members
follow? If both lead the flock, does not the flock divide? It seems better to
have only one leader.
Multiple leadership, however, is not confined to southern Carpathia. Even
the pastoralists of Sarakatchani in Greece, who use the guide wether, also
train a kind of guide ewe and ram. Besides multiple kriari ghisemia (guide
wethers), two kinds of special sheep, manari (male) and manara (female)
are used for herd control by the shepherd. Every morning just before
departure and in the evening after returning back to the camp, manari and
manara are give a handful of barley by the shepherd. When the flock starts
from the camp site for grazing, the shepherd leads the flock with whistles.
These manari and manara follow the shepherd and the rest follow smoothly.
To train these manari, they select a certain male or female lamb from among
the yearlings and keep him or her in the shepherd’s hut in order to be tamed.
Instead of allowing the lamb to suck from its mother, they feed it milk from
a nursing bottle by hand. They treat it like a pet, often holding it in their
arms. After weaning, they feed the lamb barley by hand and call it by name.
After establishing a familiar relationship between shepherd and sheep, the
shepherd begins to teach it the meanings of vocal orders.
Even if the details of the training method of the Sarakatchani are different
from those of Romanian fruntaşa, the main principles of training by feeding
are common to both. The peculiarity of the former, however, is in the use of
the non-castrated male as well as the female. In northern Carpathia, the
shepherd makes use of the guide ewe only. In southern Carpathia, the guide
wether is used with the guide ewe. Even though manari and manara are
categorized in one class by a common word root, we must say that the
Sarakatchani utilize for herd control three kinds of flock leaders: guide
wether, guide ram, and guide ewe. The same kinds of guide sheep are used
by Cretan shepherds.
Further east, in northwestern India, the Ladakhi transhumance shepherds
give milk by cup to establish a familiar relationship with certain lambs. The
Bakkalwala of Kashmir also intervene between the ewe and her offspring,
just after the delivery, and give milk to the lamb. They say that it is possible
to have a close relationship by this feeding, but neither the Ladakhi nor the
Bakkalwala train flock leaders.
The collected data are too scattered to draw a complete distributional map
of the different types of flock leaders. Moreover, the historical evidence is
insufficient to identify the place where these techniques originated and the
process of how they diffused to the regions where they are in use at present.
However, the method for raising the guide wether through castration and
training can be considered more elaborate than that of raising the non-
castrated leader, since the idea of forming a familiar relationship by feeding
and taming sheep like an obedient pet seems to be more natural. Even
though the present geographical distribution of herd control by non-castrated
leaders is very limited in area, this type of leader may have been the first in
the early stage of herding. Subsequently, herd control using the guide wether
may have been initiated in the eastern part of the Mediterranean area and
diffused to the eastern part of the Middle East, and westwards to western
Europe. These are my tentative inferences regarding the historical origins of
herd control techniques by the trained flock leader.

Behaviour and function of flock leaders in the herding situation

In the Bistriţa area of Romania, I asked the shepherds the meaning of

fruntaşa. Many of them answered that in contrast to the codaşa which moves
always in the back of the flock, th efruntaşa is the sheep that runs in the
front of the flock, and is the conductor of the flock. To the question: ‘Is the
offspring of th cfruntaşa also likely to become fruntaşa?’ not all, but many
shepherds said ‘yes’. If we take these answers at face value, the fruntaşa
may be a natural leader with an innate tendency to lead the flock. But some
of the shepherds were reluctant to confirm this. They said that the fruntaşa
does not always lead the flock and denied the existence of any tendency for
the fruntaşa offspring to inherit the leadership from her mother.
Leaving aside the possibility of inheritance, multiple fruntaşas were found
in all the flocks. If the fruntaşas behave similarly to the supposed natural
leader and if the fruntaşas move independently, how could the flock remain
united? Why don’t the shepherds keep only one leader in a flock? These
questions led me to study the actual behaviour of fruntaşas in the herding
situation.
The observations were carried out in the summer on a flock of 500 sheep
owned by eight shepherds at Mt. Câlimani in northern Carpathia. As a rule,
four shepherds, organizing a team, took charge of the herd on alternate
weeks.
It was realized at an early stage that the shepherds gave different answers
on the number of fruntaşas in the flock. At first, to avoid complication, I
asked only one of these shepherds (from now on referred to as A) to point
out all th efruntaşas in the flock. Surprisingly, 20 sheep were pointed out. I
put a strip of differently patterned cloth on each of them to make their
position in the flock obvious.
 The position of each fruntaşa was counted every 15 minutes over three
hours (12 times). The movement of the flock as a whole is as smooth as
flowing water. The positional order of each sheep incessantly changes
because of pausing, grazing, and hurrying up. Though it was very difficult to
count the exact order of the 20 fruntaşas, their approximate position could
be estimated. Although there should have theoretically been 240 scores in
the 12 times of the observation (20 fruntaşas X 12 times = 240), it was
actually possible to count the positional scores for only 197 cases because I
lost the chance to witness 43 cases in the crowd of sheep.
It was quickly noticed that the so-called leader fruntaşas did not tend to
march in the front of the flock. One was around 15th, and another was about
250th. Some of them were even about 10th and 20th from the last and it
often happened that one at 25th at a certain time was found at 90th, 15
minutes later. According to the observations, a fruntaşa was in an advanced
position, within the first ten, in only nine scores among the total number
(197).
If we suppose that these 20 fruntaşas share the same behavioural
disposition with the common sheep, the theoretical probability that at least
one of the 20 fruntaşas is in the first ten should be 0.34. If we take into
consideration the missed cases and make a revision, the actual possibility
that at least one fruntaşa was in the first ten becomes 0.05. The theoretical
probability for at least one of these 20 fruntaşas to be in the first 50, should
be 0.99, while the actual possibility was 0.28. In both of them, the actual
frequency that the fruntaşa was in the advanced position was very low in
comparison to the theoretical values. The data imply that these fruntaşas
might have been clustered together, or the sheep were not randomly
distributed in the flock. Those 20 fruntaşas belong to the same subgroup
owned by one shepherd. Did the tendency to be clustered in the joint flock
depend on it? One day a clue to the function of the fruntaşa was discovered,
when I noticed that another shepherd on herding duty (from now on referred
to as B) took a piece of mamaliga from under his belt while he shouted a
vocal command to urge the flock to start from the camping site in the
morning. He showed it to the sheep who were in front of the flock coming
out from the opened gate. He gave it selectively only to certain sheep who
did not bear any coloured strips. He explained that these sheep were his (B’s)
fruntaşas. Responding to my request to point out his fruntaşas among the
sheep rushing ahead to obtain a small piece of mamaliga from his hand, he
indicated five sheep as his fruntaşas among the 15 sheep in the first group,
while none of shepherd A’s fruntaşas were found within the first 15 heads
rushing ahead towards shepherd B. At this moment 18 of the 20 fruntaşas
marked by shepherd A (those that I could identify immediately) were in the
following positional order: 24, 25, 28, 55, 75, 85, 207, 245, 250, 290, 340,
345, 370, 380, 381, 390, 400, and 405 (average: 232nd among 500 sheep).
This meant that when shepherd B was in charge of the flock, his fruntaşas
promptly approached him and consequently led the flock, while the
fruntaşas of shepherd A were far behind. Urged by the invitation of his
clapping and whistling, his fruntaşas dashed ahead and then the others
smoothly followed after them. To sum up, th e fruntaşas are the sheep who
are tamed by each shepherd through feeding in order to respond to his vocal
orders and are his favourites, and do not respond to the other shepherds.
Even though these fruntaşas do not naturally tend to lead the flock during
grazing, they follow their master without delay when attracted by the
shepherd’s offer of mamaliga and calls, and their movements in turn cause
the rest of the flock to follow. In the latter sense, they can be defined as the
inducers of flock movement. Their function depends on the order-to-obey
relationship established by training through feeding, and on the disposition
of the sheep in general to follow. As regards the problem of multiple leaders,
we can say that the small initial inducement by only one fruntaşa does not
have sufficient effect. The more numerous the inducers, the more effective
the first is in causing the others to move. Since all the fruntaşas make their
common initial move by simultaneously responding to the shepherd’s order,
the plurality of fruntaşas does not bring about any problem in herd control.
As further evidence for the utility of multiple fruntaşas, I often took note
in the field that the shepherd called the name of another fruntaşa when one
fruntaşa reacted slowly and her initial move was not sufficient to urge the
rest of the flock to follow. When the shepherd wanted the flock to turn back
or stop, he whistled with his fingers, called the name of a fruntaşa, and
directed it to come back or stop. But, if the initial move ofthat sheep did not
cause the others to follow, he often repeated the same order, and called other
fruntaşas by name to invoke a more intensified inducing effect.
The disagreement on the number of fruntaşas between the shepherds who
take care of the same flock is not strange, now that we know that each
shepherd raises his favourit efruntaşas from his own sheep in the flock.
Moreover, if we consider the shepherd’s preferential feeding of his fruntaşa,
we can see the offspring of a fruntaşa sometimes become fruntaşas, as some
shepherds recognized. The lamb, being together with her mother fruntaşa,
may have more chances to have close contact with the shepherd. If the
shepherd is likely to feed the sheep that does not fear him, such offspring
would be more likely to be chosen as fruntaşas in future.
I have no quantitative observations on the non-castrated male and female
(manari and manara) of the Sarakatchani shepherd. However, they are
tamed like pets from their infancy and are continuously fed. In the morning,
before departure, they are given a handful of barley and then follow after the
shepherd who walks in front of the flock. In the evening, they walk in the
forefront expecting the offer of barley, even if the shepherd does not
accompany them. The shepherd treats them mostly the same as fruntaşas.
To sum up, the fruntaşa (or cirmace) of Romania, as well as the manari
and manara of Greece, work as mediators between the shepherd and the
flock. Through feeding and training, they become obedient agents reacting to
the shepherd’s vocal order. The plurality of these so-called leaders in a flock
intensifies the effect of leadership.
The actual function of the guide wether will now be discussed. At first,
the data on the guide wether, batal, in Dobrogea will be presented. Three
guide wethers were placed in the flock observed. To observe the positional
order of these batals in the flock, coloured strips were fixed on each. On the
departure from the wooden fence in the morning, these batals tended to
march in the rear half of the group, and it was rare to find them in the front.
Only the goats mixed in with the sheep tended to dash ahead and induced the
others to follow. Later on, while the flock was in the field grazing it was
widespread and slowly moved as a whole in a certain direction, sometimes
being led by several sheep in the lead group. Even in such a situation batals
were seldom found within the first group. The positional order of the batal
was really random.
However, there were occasions when the batal led the flock. For example,
one day the flock was proceeding along the bank of a river, and it happened
to come against a gully forming a dry stream bed. The lead group halted
there without any attempt to jump over. The shepherd immediately called a
batal by his name and approached him. This caused the batal to jump over
the gap. The others followed his initiative and jumped one by one. To
establish this order-to-obey relationship, the shepherd gives salt to the guide
wether while calling his special name. In this sense the batal has the same
role as the fruntaşa. Nevertheless, the shepherds of southern Romania keep
both in the flock. The shepherd emphasizes the bataVs effectiveness in
leading the way in deep snow. His bravery and toughness is the reason why
they utilized the batal with the female guide sheep not only southern
Romania but also in Greece.
Now, the last type of guide sheep found in Greece, the manari (non-
castrated male), has not yet been discussed. Among all pastoralists, it is
common to separate yearlings from the ewe group after they are weaned. As
long as they keep the manari in the yearling group as a guide sheep, there is
no problem. But in the mating season when all categories of sheep, including
the ewes, are joined together into a flock, the manari might be less eager in
his duty. In this respect, the manari cannot be as effective a guide in
comparison to the guide wether.
For the same reason, the fruntaşa and manara as well may be less
effective because in the mating season she will be the objective of the
breeding ram. The guide wether, with the sacrifice of his sexual ability,
should be the best as flock leader.

Discussion

We have described the flock leader as artificially trained and with a close
relationship with the shepherd, even though there are various grades of
elaboration in the training.
If a shepherd does not have his own sheep but is under annual contract,
like the Kandahari nomads, it is impossible for him to foster such a familiar
relationship between himself and the flock leader. Among the Kandahari, the
use of the castrated he-goat was most commonly seen. They did not give any
proper names to these goats. Even though they tame the he-goats by feeding,
they might rely more on the goats’ recklessness and rashness than on control
through training. In central Italy the non-professionals who take care of the
flock of the villagers’ sheep do not use the guide wether. We know that the
he-goat is utilized not only in Afghanistan but also widely in the
Mediterranean and Near and Middle East, where the professional shepherd
often works for the owner of large flocks.
 Now if we classify the guide sheep into two categories: the non-castrated
(female and male) and the castrated (male), direct observation shows that
neither type had any natural inclination to lead the flock. They take the role
of leader only when the shepherd takes the lead in front of the flock or sends
vocal orders to them. Being conditioned by training and instruction, the
guide sheep can understand his own name and understand the vocal order to
stop, or return back, or go ahead, etc. Through this communication they
make the initial move, responding to the shepherd’s orders, and consequently
induce the others to follow. Both types of guide sheep achieve basically the
same function. The difference between them is at first found in the wether’s
toughness and bravery. The guide wether works effectively even in the
mating season and in the winter snow. Nevertheless, in northern Carpathia
where the snowfall is heavy, the guide wether is not used.
In Greece, at least among the Sarakatchani and the Cretan shepherds, both
types of flock leader, the castrated and the non-castrated, were utilized. The
utilization of the guide wether seems to have been recently introduced into
southern Carpathia from Greece.
We do not have any clue which of the types, the non-castrated or the
castrated, was developed first. With respect to the effectiveness as flock
leader, the guide wether is, however, more elaborate than the non- castrated
types, as discussed above.
Now, it is opportune to raise a question. On which domestic animal was
the castration-training technique at first applied: on the ram or on the bull? It
is more probable to have been applied to the bull. We know from a
Sumerologist, Maekawa, that systematic castration and training of the bull
had been applied from the 3rd millennium BC under the Ur III dynasty in
Mesopotamia, to produce plough-animals and to lend them to the allotted
farmers (Maekawa 1979, p. 99).
Is the guide wether the analogical extension of the idea of a trained
castrated bull? I can imagine another way for the beginnings of the guide
wether. With the supposition that the manari and the manara types of flock
leader had been practised in an earlier stage, if shepherds began to castrate
the manari, the guide wether would have become established. The technique
to produce an obedient male by castration and training was perhaps
developed when man first began to use horses and oxen for draught and
riding. But we can also suppose that the eastern Mediterranean shepherd
who already had the non-castrated leader (manari and manara type) learned
this castration-training technique from the civilized world of the ancient
Near East and applied it to their flock control.
Leaving aside the concern about the historical origin of such techniques, it
is opportune to return to the problem of the eunuch. Maekawa (1979, 1980,
1982) has shown from the Sumerian cuneiform tablets of Lagash, of the Ur
III dynasty, that many sons of captured or slave women, who worked in the
specialized weaving camps were used for physical labour and categorized by
the special term for the castrated ploughing bull; amar- KUD. While the
female offspring of the slave women were recruited again as weavers in the
maternal group, their male offspring were castrated and used for physical
labour and separated from the maternal group. This is an interesting example
of the application of cattle management techniques to the management of
slaves.
We do not have any historical data on how and where the institution of the
eunuch was originally developed. Moreover, at present there is no evidence
to suggest that the guide wether was the model of the eunuch. But it seems
possible that this kind of social control, using the castrated male, prepared
the cultural background for the idea of the eunuch. This is, however, just a
sideline. The main purpose of this chapter has been to present information
on the variety of flock leaders, to give the main distribution of each type, and
to define the nature and the function of flock leaders on the basis of field
data. More detailed field data and further discussion of the shepherd’s
intervention into the sheep-flock relationships are published in Tani (1987).

References
Baskin, L. M. 1974. Management of the ungulate herds in relation to domestication. In The behaviour
of ungulates and its relation to management, V.Geist &F. Walther (eds), 530–42. Morges: IUCN
Publications, N.S. 242.
Digard, J.-P. 1981.Techniques des nomades baxtyâri d’Iran. Cambridge: Cambridge University Press.
Henckel, A. & A. Schöne 1967.Emblemata: Handbuch zur Sinnbild Kunst des XVI. und XVII.,
Jahrhunderts. Stuttgart: J. B. Metzlersche Verlagsbuchhandlung.
Maekawa, K. 1979. Animal and human castration in Sumer, I. Zinbun 15, 95–140.Kyoto: Research
Institute for Humanistic Studies, Kyoto University.
Maekawa, K. 1980. Animal and human castration in Sumer, II. Zinbun 17, 1–56. Kyoto: Research
Institute for Humanistic Studies, Kyoto University.
Maekawa, K. 1982. Animal and human castration in Sumer, III. Zinbun 18, 95–122. Kyoto: Research
Institute for Humanistic Studies, Kyoto University.
Matsubara, M. 1983.Yuboku no Sekai (Ethnography of the nomadic Yöluk people in Anatolia). Tokyo:
Chuökoronsha.
Tani, Y. 1976. Bokuchiku Bunkakö (On the pastoralistic culture: interaction between man and
domestic animals and analogical application of the pastoralistic managemental techniques into the
social control of man) [in Japanese] . Bulletin of the Research Institute for Humanistic Studies 42,
1–58. Kyoto: Kyoto University.
Tani,Y. 1977.Italiachübasanson, iboku hitsujino kanri ni tsuite (On the management techniques of the
sheep flock among the transhumance shepherds in central Italy) . Society and culture of Europe-
field research reports, 117–67. Kyoto: Research Institute for Humanistic Studies, Kyoto University,
1976.
Tani, Y. 1980. Man-sheep relationship in the flock management techniques among north Carpathian
shepherds. In Preliminary report of comparative studies on the agro- pastoral peoples in
southwestern Eurasia, Y. Tani (ed.), 67–86. Kyoto: Research Institute for Humanistic Studies,
Kyoto University.
Tani, Y. 1982. Implications of the shepherd’s social and communicational interventions in the flock –
from the field observation among the shepherds in Romania. In Preliminary report of comparative
studies on the agro-pastoral peoples in southwestern Eurasia, Y. Tani (ed.), 1–18. Kyoto: Research
Institute for Humanistic Studies, Kyoto University.
Tani Y. 1987. Two types of human interventions into the sheep flock: intervention into mother-
offspring relationship, and raising the flock leader. In Domesticated plants and animals of the
southwest Eurasian agro-pastoral cultural complex. Vol. 2, Y. Tani (ed.), 1–42. Kyoto: The
Research Institute for Humanistic Studies, Kyoto University.
Tani, Y. & T.Matsui 1980. The pastoral life of the Durrani pashtun nomads in Northeastern
Afghanistan. In preliminary report of comparative studies on the agro- pastoral peoples in
southwestern Eurasia, Y.Tani (ed.), 1–31. Kyoto: Research Institute for Humanistic Studies, Kyoto
University.
18 Cattle in ancient North Africa
JULIET CLUTTON–BROCK

Introduction

The purpose of this chapter is to enter the discussion on the origins of

domestic cattle in North Africa, their phenotypes, and the dates of their first
appearance in the archaeological record. There has been contention about
this subject since the first descriptions of the ancient remains of wild cattle
from Algeria in the late 19th century, but more especially since the
discovery of the wonderful rock art of the Tassili n’Ajjer mountains, which
depicts a diversity of domestic cattle and other bovids (see Fig. 18.1).
Today, the comprehensive reviews of Smith (1980, 1986), Banks (1984),
Wendorf et al. (1984), and Gautier (1987) provide up-to-date accounts of
the archaeology of cattle-keeping in the Sahara and West Africa.

Theories on the origins of cattle in North Africa

The traditional view on the origins of domestic cattle in Africa is that the
animals were brought to the continent across the region of Suez,
approximately 6000 years ago, which is about 2000 years later than cattle
were first domesticated in western Asia and Greece. These long-horned
humpless cattle are said to have spread by diffusion westwards and then
south into North and West Africa and the central Sahara, where they are
represented in the Tassili rock art (Lhote 1959, Epstein & Mason 1984).
Epstein believes that these first long-horned cattle were replaced by a new
wave of humpless short-horned cattle around 4000 years ago, and that these
followed the same diffusion route except that they did not reach the central
Sahara.
Humped cattle, according to Epstein, were introduced considerably later
(c. 3500 BP) across the Horn of Africa. These were cattle with cervico-
thoracic humps and they were the precursors of the true zebu or Indian
humped cattle, which have a hump only over the thoracic vertebrae. The
neck-humped cattle spread across Africa through Sudan and reached West
Africa where they interbred with the primitive longhorns to produce the
Fulani breed. Meanwhile, cattle proliferated in Ethiopia where the neck-
humped cattle were crossed with long-horned animals, around 3000 years
ago, to produce the Sanga breeds. It was from the Sanga that the remarkable
Ankole cattle of Uganda, and the Afrikander cattle of South Africa are
derived.
Later still the true zebu was introduced to Africa, also through Somalia,
and these thoracic-humped cattle rapidly displaced the older Sanga breeds
because they were more rinderpest-resistant and gave a higher milk yield.
These theories have been put forward by Epstein & Mason in their
review of 1984, and are based on deductions from the distributions of
present-day breeds in Africa and on evidence from pictorial representations
in the Sahara, Ethiopia, Sudan, and ancient Egypt. The archaeological
evidence, however, presents a different story.

The differentiation and dating of cattle remains from North

Africa

There is now general agreement amongst archaeozoologists that the

progenitor of all domestic cattle was the extinct aurochs, Bos primigenius,
although the zebu (Indian humped cattle) could be descended from the
subspecies, Bos primigenius namadicus (Clutton-Brock 1981, Epstein &
Mason 1984, Grigson 1985). Formerly, it was believed that in Europe there
were two species of ancient wild cattle, Bos primigenius and a smaller
short-horned form, Bos longifrons or brachyceros. Better dating of
archaeological sites and more rigorous examination of the material has
shown, however, that examples of this small ‘species’ were, for the
chronologically older specimens, the females of the aurochs, whereas the
more recent specimens are the remains of small prehistoric domestic cattle.
Similarly, in North Africa two wild species of cattle were described from
early finds, these being Bos primigenius and a small form, Bos ibiricus or
africanus. Today, it seems clear that the small form was the female of Bos
primigenius (Banks 1984).
 Remains of cattle that are believed to have been domestic from their
rather small size have been retrieved from sites of around 8000 BP in
Greece, Anatolia, and western Asia, while, from Mehrgarh in Baluchistan,
Meadow (1984) has reported the recent finding of a clay model of a
humped ox dating to around 6000 BP. The theoretical view that descendants
of these early cattle first entered Africa through Egypt and that they were
long-horned is not borne out by the subfossil finds. The bases of the horn
cores on a skull from Badari, now in the British Museum (Natural History),
which should be at least 5000 BP, indicate that the animal had rather short,
upstanding horns. In any case, this skull should be directly dated by
radiocarbon before its context is accepted, because the skull of an ass from
the same level at Badari has produced a date of 700 ± 120 BP (OxA-564).
There are also two horn sheaths from a very long-horned ox or cow from
the Fayum in the British Museum (Natural History) collections, but these,
too, should be dated before they can be accepted as predynastic.
The earliest securely dated finds of cattle in a cultural context in Africa at
present come from Capéletti in Algeria, where dates were obtained of 6530
± 250 BP for the lowest levels (where the cattle were not certainly
domesticated), and 5100 ± 150 BP for the upper level, where domestic cattle
predominated (Roubet 1978). Other sites from further south, in the Sahara
and Sahel, where remains of domestic cattle have been identified, are given
in Table 18.1. These sites are also shown in Figure 18.1, together with those
localities cited for pictorial representations. From none of these sites do the
cattle remains show any signs of having been long-horned, and, indeed, the
bones from Kintampo are so small that they resemble those of the modern
dwarf cattle of West Africa, the Ndama breed (Carter & Flight 1972). These
short-horned subfossil remains conflict with the appearance of the cattle
from the rock art of the Tassili mountains, for which Lhote (1959) proposed
four periods. The earliest phase he claimed had pictures of the extinct North
African buffalo, Homoiceras singae. Then came Lhote’s Bovidian phase
from which there are a multitude of paintings showing both long and short
horns and polled cattle. The people who created these paintings were cattle
pastoralists who appear to have had no agriculture, and who may have lived
about 5000 years ago. Later, there was Lhote’s horse phase which was
protohistoric, and finally a camel phase dated to around 2000 BP.
Table 18.1 Sites in the Sahara and Sahel from which remains of domestic cattle have been identified
(from Smith 1980, p. 495).
Site Absolute
date
(BP)
Uan Muhuggiag 5952 5405
Adrar Bous 5780 5740 5130 4910
Karkarichinkat Sud 3960 3640 3310
Karkarichinkat Nord 3950 3710 3620
Dhar Tichitt 3465 3350 3620
Kintampo 3560–
3220
(suggested
dates)
Daima 2400

From the indications of a savanna environment in which a diversity of

wild ungulates could survive as well as humans and their cattle, the artists
of the Tassili mountains must have been living there before the climate
began to be really arid, around 4000 years ago. It could be a matter of
chance that no osteological material from these long-horned cattle has
survived, or it could be that these were rare and valued animals that were
more often painted than seen.
Long-horned cattle were seen to the north of Tassili by Herodotus who
described them in a famous passage, written about 2450 BP : ‘In the
Garamantian country are found oxen which, as they graze, walk backwards.
This they do because their horns curve outwards in front of their heads, so
that it is not possible for them when grazing to move forwards, since in that
case their horns would become fixed to the ground.’
Smith (1980, p. 492) quotes from this passage but infers that the cattle
described by Herodotus were wild Bos primigenius. I disagree with him
over this, for I am sure that these were domestic long-horned cattle that
must have looked very similar to the present-day English Longhorn breed.
These cattle can have exactly the same problem with their long down-
curved horns from which the tips have often to be removed.
 Figure 18.1 Sites and regions in North Africa where either subfossil remains or pictorial
representations of ancient domestic cattle have been found.

Cattle in the Nile valley

The theoretical view claims that cattle first entered Africa through Egypt.
There is, however, no well-dated osteological or pictorial evidence to
substantiate this claim as yet. The cattle remains from the predynastic sites
of Badari and Fayum, as mentioned previously, could be earlier than 5000
BP but they have not been directly dated, and those from Neolithic sites such
as Dakhleh Oasis and the Khartoum sites appear to be from wild Bos
primigenius (Churcher 1982, Smith 1984). The only radiocarbon date
obtained from cattle remains directly (on a skull of a sacrificial ox from
Lahun which had black hide and medium length, upstanding horns) is very
late, being 3420 ± 80 BP (BM-1420). All other information on domestic
cattle comes from the later sites, such as Saqqara, which can be well-dated
historically, of course, but by this time the animals were fully domesticated
and yield nothing new about their early history.
Many of the paintings of cattle from ancient Egypt are very well known.
From Giza there are pictures of spotted and polled cattle (4420–4270 BP)
being carried in boats, which shows that by this period they were already
highly bred, placid animals, and there are many pictures of cattle from
Thebes, dating from around 3411–3070 BP. These are long- horned, lyre-
horned, and some polled. But perhaps the most interesting painting is at
Thebes, dated to c.3400 BP, showing humped cattle together with a herd of
long-horned cattle. This is the earliest known date for humped cattle in
Africa.

Discussion

It is widely believed that the ass, the cat, and the Guinea fowl were first
domesticated in Africa and that these were the only taxa which this
continent contributed to the world’s assemblage of domestic animals (see,
for example, Shaw 1977, p. 110). The Guinea fowl can be allowed as an
original African domesticate, but the cases for the ass and the cat are more
dubious as both could equally well have been first domesticated in western
Asia and only later in the Nile valley. The same may be true for cattle as
there is a definite possibility that these livestock may have been
domesticated locally within the northern Saharan region, albeit about 1000
years later than in western Asia. A number of archaeologists have discussed
this proposition (Shaw 1977, Smith 1980, Banks 1984) and there does not
seem any good reason to gainsay it, just as Meadow (1984) believes that
cattle may have been locally domesticated on the eastern margin of the
Middle East at Mehrgarh, although here the animals would probably have
been humped. There is no evidence to suggest that the earliest domestic
cattle in Africa were humped, or indeed that they were long-horned.
Gautier (in Wendorf et al. 1984) has postulated that the remains of cattle
from early Neolithic sites in the Bir Kiseiba region of the eastern Sahara
could be from domesticated animals derived from wild Bos primigenius in
the Nile valley. The bovid elements are, however, very fragmentary and
their specific identification and domestic status are questionable, as
discussed by Smith (1986). The provisional date for this material is around
9000 BP but, in addition, the specimens should be directly dated before
their context can be substantiated.
After looking at the question of when cattle were first domesticated in
Africa, either locally or arriving there by diffusion, it is then pertinent to ask
why it was necessary to go to the trouble of husbanding livestock, rather
than continuing to hunt wild animals for meat. Perhaps the answer lies in
the additional resource of milk which could be life-saving in a region that
was becoming ever more arid. Shaw (1977) has discussed the movement of
cattle-keeping southwards from the Sahara to the Sahel and West Africa
over a period of 3000 years, and he has linked this diffusion with the
gradual drying of the climate since 4000 BP (see the isochronic diagram in
Shaw 1977, p. 108).
More recently, the changing climate of the Sahara and the Sahel
throughout the early Holocene, together with the early history of
pastoralism in Africa, have been discussed in detail by Banks (1984) and in
many of the papers in Clark & Brandt (1984). David (1982, p. 54) has
briefly mentioned the roles that different breeds of cattle may play in the
changing pattern of pastoralism over time, and it should not be forgotten
that there are many sides to this, with the intermingling of sheep, goats, and
camels with cattle, and, indeed, the herding of sheep and goats may have
occurred earlier than that of cattle in Africa, as it did in western Asia.
Before 8000 BP there was a hunting and fishing economy in the central
and southern Sahara that flourished around numerous large lakes inhabited
by hippopotamus, crocodiles, and fish. This apparently ideal environment
began to dry up shortly after this period, but there was a renewed wet phase
around 7000 BP, and it was during this period that the innovation of herding
began in Africa. By 6000 BP pastoralism was widespread throughout the
Sahara and extended to the Nile valley. Desertification began around 4500
BP, putting pressure on the prehistoric peoples who were forced to follow
the river systems that drained southwards.
Shaw (1977) and Smith (1980, 1984) have suggested that the wetter
conditions of the Sahara in the early Holocene enabled the tsetse-fly belt to
stretch further north into the Sahara, and this formed a barrier beyond
which, to the south, cattle could not survive. With the increasing aridity
humans, cattle, and tsetse flies moved southwards.
 By 3300 BP the pastoralists had moved into the savanna regions, and it
was around this time that it seems that humped cattle made their appearance
in pictorial representations. They were gradually interbred with the local
cattle, or replaced them, and this may have been because humped cattle
were better suited to the long-distance migrations that were essential for
survival in arid regions. This pattern of transhumance, centred on herds of
cattle, camels, sheep, and goats, continued for 5000 years, undoubtedly with
recurrent famines but never on the scale of the present time. In an important
review by Sinclair & Fryxell (1985) the modern disruption to the carrying
capacity of the Sahel has been succinctly put in their abstract:

Migratory pastoralists have traditionally lived with their cattle in balance

with the vegetation. This balance was disrupted in the 1950’s and 1960’s
by (i) the settlement of pastoralists around wells, and (ii) the cash crops
coincident with increasing human and cattle populations. This has
resulted in continuous famine in various parts of the Sahel since 1968. In
addition, widespread soil denudation may be causing climatic changes
towards aridity. Long-term climatic trends in the past 3000 years point to
human interference rather than climatic change as the cause of famine.
The evidence suggests that the Sahel problem is a man-made famine
caused by overgrazing and not by lack of rain.

If we ignore the ancient, adaptive strategies of the pastoralists, deserts and

famines will spread, so that the fragile semi-arid regions of the world will
become totally uninhabitable. What better proof could we have of the need
to study the past?

References
Banks, K. M. 1984. Climates, cultures, and cattle: the Holocene archaeology of the eastern Sahara.
Dallas: Southern Methodist University.
Carter, P. L. & C. Flight 1972. A report on the fauna from the sites of Ntereso and Kintampo rock
shelter six in Ghana; with evidence for the practice of animal husbandry during the second
millennium B.C. Man 7(2), 277–82.
Churcher, C. S. 1982. Dakhleh Oasis project of geology and palaeontology: interim report of the
1981 field season. Journal of the Society for the Study of Egyptian Antiquities 12(3), 103–14.
Clark, J. D. & S. A. Brandt (eds) 1984. From hunters to farmers: the causes and consequences of
food production in Africa. Los Angeles and London: University of California Press.
Clutton-Brock, J. 1981. Domesticated animals from early times. London: British Museum (Natural
History) & Heinemann.
David, N. 1982. The BIEA southern Sudan expedition of 1979: interpretation of the archaeological
data. In Culture history in the southern Sudan, J. Mack & P. Robertshaw (eds), 49–57. Nairobi:
Memoir 8 of the British Institute in East Africa.
Epstein, H. & I. L. Mason 1984. Cattle. In Evolution of domesticated animals, I. L. Mason (ed.), 6–
27. London: Longman.
Gautier, A. 1987. Prehistoric men and cattle in North Africa: a dearth of data and a surfeit of models.
In Arid North Africa. Essays in honor of Fred Wendorf A. E. Close (ed.), 163–87. Dallas: Southern
Methodist University Press.
Grigson, C. 1985. Bos indicus and Bos namadicus and the problem of autochthonous domestication
in India. In Recent advances in Indo-Pacific prehistory, V. N. Misra & P. Bellwood (eds), 425–8.
New Delhi: Oxford & IBH Publishing.
Lhote, H. 1959. The search for the Tassili frescoes. London: Hutchinson.
Meadow, R. H. 1984. Animal domestication in the Middle East: a view from the eastern margin. In
Animals and archaeology. Vol 3: early herders and their flocks, J. Clutton-Brock & C. Grigson
(eds), 309–38. Oxford: BAR International Series 202.
Roubet, C. 1978. Une économie pastorale, pré-agricole en Algérie Orientale: le Néolithique de
tradition capsienne. Anthropologie 82 (4), 583–6.
Shaw, T. 1977. Hunters, gatherers and first farmers in West Africa. In Hunters, gatherers and first
farmers beyond Europe, J. V. S. Megaw (ed.), 69–126. Leicester: Leicester University Press.
Sinclair, A. R. E. & J. M. Fryxell 1985. The Sahel of Africa: ecology of a. Canadian Journal of
Zoology 63, 987–94.
Smith, A. B. 1980. Domesticated cattle in the Sahara and their introduction into West Africa. In The
Sahara and the Nile. M. A. J. Williams & H. Faure (eds), 489–503. Rotterdam: A. A. Balkema.
Smith, A. B. 1984. Origins of the Neolithic in the Sahara. In From hunters to farmers: the causes and
consequences of food production in Africa, 84–92. Los Angeles & London: University of
California Press.
Smith, A. B. 1986. Review article: cattle domestication in North Africa. The African Archaeological
Review 4, 197–203.
Wendorf, F., R. Schild & A. E. Close (eds) 1984. Cattle-keepers of the eastern Sahara: the Neolithic
of bir Kiseiba. Dallas: Southern Methodist University Press.
19 The development of pastoralism in East
Africa
PETER ROBERTSHAW

Mparan amu iyata suami (You have friends because you have animals).
Samburu proverb

Little is known about the prehistory of the pastoral peoples of East Africa
and the development of the pastoral mode of production in the region. This
sad state of affairs is rendered more depressing by the lack of historical
awareness among pastoral development planners. So often the research of
these planners involves ‘reconstructions’ of ‘traditional’ systems of
livestock management practices, which amount, in effect, to the projection
of present systems of animal husbandry on to what can be gleaned from
ethnographers’ accounts of early patterns of land tenure and mobility.
Similarly, with the emotive issue of whether pastoralists cause
desertification, the archaeological and palaeoenvironmental evidence that
might answer the question is very rarely invoked.
The first half of this chapter aims to dispel some of this gloom by
outlining what is known about the introduction and development of
pastoralism in East Africa. The second half examines in some detail the
implications of several archaeological faunal assemblages that seem to
document hunting cum pastoral subsistence. This combination of
subsistence pursuits is extremely rare in East African ethnography and,
furthermore, would appear to involve an internal contradiction within the
social relations of production between the principles of collective access to
hunted resources and of divided access to domestic livestock. Several ways
are explored in which this contradiction may be resolved.
Setting aside extravagant claims for the presence of domestic animals in
eastern Africa at an exceptionally early date, the archaeological evidence
indicates that pastoralism was first practised in East Africa late in the 3rd
millennium BC. The sites that document this are located in the Lake Turkana
basin of northern Kenya (Barthelme 1985). It seems to have taken
approximately 1000 years more for domestic animals to reach the Rift
Valley and Highlands of central and southern Kenya, and Tanzania
(Ambrose 1984). However, ceramic studies (Collett … Robertshaw 1983)
suggest that further fieldwork may lead to a narrowing of this chronological
gap. Cattle, sheep and goats, and probably donkeys, were the animals
involved. While donkeys, and perhaps cattle, were domesticated in North
Africa, sheep and goats are of Near Eastern origin; thus their introduction to
East Africa must be seen as the result of a process of diffusion or human
population movements. Most authorities infer relatively small scale
immigrations of pastoralists from the north, perhaps Cushitic speakers from
Ethiopia (Ambrose 1984). Clarification of the subsequent development of
the early pastoral communities of East Africa is rendered obscure by a
complex culture history, which to the struggling archaeologist appears at
times to verge on the chaotic (Bower et al. 1977, Collett … Robertshaw
1983). Several further immigrations into East Africa by pastoral/agricultural
groups are indicated from historical linguistic evidence (Ehret 1971, 1974)
and receive support, though not unequivocal, from archaeological data
(Ambrose 1982). These pastoralists were stone tool, rather than metal users,
and obsidian was the preferred raw material, being found in substantial
quantities at sites up to 100 km or more from the major sources in the
central Rift Valley (Merrick … Brown 1984).
While we know, obviously, that the early pastoral peoples of East Africa
kept livestock, we have as yet no direct archaeological evidence for grain
cultivation. However, ‘indirect evidence’ (Shaw 1976) for the growing of
cereal crops has been obtained from studies of site locations, ground stone
tools, and historical linguistics. Thus, from the perspective of diet these
peoples were mixed farmers. However, consideration of regional settlement
patterns suggests that pastoralism rather than cultivation was the preferred
mode of subsistence, for areas such as the eastern Highlands of Kenya,
possessing rich, fertile soils, do not seem to have been settled; instead
people inhabited the savanna grasslands which, on the whole, are less
suitable for cereal cultivation, due to thinner soils and less rainfall. Thus, it
seems likely that domestic livestock were of prime importance to the early
food producing communities of East Africa, not necessarily in the sense
that these animals provided the major portion of the diet, but rather in the
sense that there was a cultural preoccupation with livestock to the extent,
perhaps, that people who lost their stock would have chosen to remain in
the grasslands, rather than opt for an agricultural way of life in the adjacent
highlands (Robertshaw … Collett 1983). This preoccupation – some might
call it infatuation – with livestock is prevalent among several modern East
African groups, the classic example being the Nuer: the Nuer social idiom
is said to be a bovine idiom, yet much of their diet consists of millet, the
cultivation of which is regarded as an unfortunate necessity (Evans
Pritchard 1940).
The cultivation of cereal crops may have been an ‘unfortunate necessity’
too for the pastoralists of the 1st millennium BC, as there were no
agricultural peoples from whom they could obtain grain by barter. Survival
entirely from the produce of their herds is a most unlikely hypothesis, since
the ethnographic data make it quite clear that no pastoral society could
survive indefinitely without access to agricultural produce (Monod 1975, p.
134). The need, then, for the early pastoralists to grow crops, as well as tend
their herds, facilitates the construction of a model of settlement patterns.
This model, for which some support can be derived from existing
archaeological data, has been set out in detail elsewhere (Robertshaw …
Collett 1983) and need not be repeated here.
Around AD 200 early Iron age farmers settled the then sparsely populated
highlands adjacent to the savanna grasslands. Thus, from this time on
pastoral peoples would have been freed from the necessity of cultivating
crops themselves; presumably they could now exchange small stock for
grain with their agricultural neighbours, as has been documented, for
example, for Maasai Kikuyu relations at the beginning of this century
(Waller 1976, Berntsen 1979). Given the cultural emphasis placed upon
livestock by the pastoralists, we can conjecture that an economic shift took
place around the mid 1st millennium AD, from mixed farming with some
hunting to specialized herd management with grain obtained through trade
(Robertshaw & Collett 1983).
The model outlined above is complicated by historical linguistic evidence
of a migration of eastern Nilotic speakers into the savanna regions in the 1st
millennium AD, who displaced or absorbed the previous inhabitants. An
archaeological correlate for this event has not been established with any
certainty (Robertshaw 1984 contra Ambrose 1982). Even if population
immigration in this period is eventually proven, it must be emphasized that
it is the coming of Ironage farmers to neighbouring regions, not the identity
of the pastoralists, which is crucial to the development of specialized herd
management strategies in the savanna.
The role of hunting among the early pastoral communities of East Africa
will now be considered. The great majority of the faunal assemblages from
archaeological sites of this period are dominated by bones of domestic
livestock with very few, if any, remains of wild species (Gifford Gonzalez
& Kimengich 1984). This pattern holds true even for areas which support
large numbers of wild ungulates. However, we now know of two, possibly
more, exceptions where sites have yielded significant quantities of both
domestic stock and wild animals, primarily medium to large ungulates.
These sites are Prolonged Drift, located in the Central Rift Valley and dated
to the mid 1st millennium BC (Gifford et al. 1980), and the Elmenteitan
levels at Gogo Falls, situated about 20 km east of Lake Victoria and dated
to about the 1st century AD (Robertshaw 1985). Both are extensive open air
occurrences (Fig. 19.1).
Ingold (1980) has argued that the change from hunting to pastoralism is
not simply a shift in ecological relations from herd pursuit to herd
management (contra Foley 1982), but it involves a major transformation in
the social relations of production and the ideology of prestige. Pastoralism
is based on the principle of divided access to living animal resources, i.e.
the ownership of animals, while hunting rests on the principle of collective
access where an animal becomes someone’s property only with its death.1 A
pastoralist gains prestige by accumulating wealth on the hoof, but for the
hunter it is the demonstration of his skill and his generosity in distributing
the meat from his kills which win him respect. Thus, archaeological sites
which document a combination of pastoralism and hunting appear from this
theoretical perspective as surprising phenomena where the implied
contradictions in the social relations of production and prestige require
explanation. This impression is reinforced by a perusal of the ethnographic
literature on the pastoral peoples of East Africa, in which hunting is, as a
rule, very rarely mentioned. Could it be that the archaeological sites
document not only a system of subsistence no longer extant in East Africa
(Gifford Gonzalez 1984) but also, perhaps, different cultural attitudes
towards animals? To answer these questions we need to look more closely
at the ethnographic evidence.
Figure 19.1 Location of sites yielding bones of both domestic stock and wild animals.
 Many ethnographies of pastoral peoples either completely tail to mention
hunting or remark simply that it is a rare activity, perhaps resorted to more
often in times of severe drought (e.g. Carr 1977, p. 200). For some groups,
wild animals are only of interest in their relationship to domestic livestock:
for the Nandi ‘wild animals are of two kinds: those that injure livestock and
those that do not’ (Huntingford 1953, p. 127). Not only is hunting a rare
activity for many pastoralists, it is also regarded with some contempt: the
Nuer insist that only the absence of cattle makes a man engage in hunting
other than casually (Evans Pritchard 1940, p. 73). Among the Turkana the
word ‘hunter’ (egulokit) is almost synonymous with that for a poor man
(erkeboton). ‘Gazelle are the herds of the poor man’ (Gulliver 1963, p. 33).
Similarly, ‘since Borana Gutu equate wealth with livestock hunters are by
definition poor people’ (Dahl 1979, p. 178). Furthermore, for Borana at
least, among whom reside an endogamous ‘caste’ of Wata hunters, these
people are not only poor but they are also ‘unworthy’ since they lack any
concern for the capital expansion of their resource base (Dahl 1979).
Only one or two East African pastoral peoples appear to hunt regularly
under normal climatic and environmental conditions. These are the Murle
and possibly the Datoga (Barabaig), but no figures are available for the
amount of labour invested in hunting, or the yields achieved. Murle hunting
involves the interception of migrating animals at river crossings; it is
interesting to note that elaborate rules exist among the Murle for the
distribution of all meat, both game killed during the hunt and cattle or sheep
slaughtered at ceremonies (Lewis 1972). Among the Datoga a distinction is
made between the ‘common pastime’ of hunting with bow and arrow and
the pursuit with a spear of dangerous prey, such as lions and Maasai, in
order to demonstrate bravery (Umesao 1969).
Although regular hunting by pastoral peoples seems a rare phenomenon,
we should be aware of the possibility that the pastoralists’ contempt of
hunting may sometimes have rubbed off onto the ethnographer; for example
Spencer (1965) in his monograph on the Samburu makes no mention of
hunting, yet recent fieldwork has revealed that not only do Samburu
warriors regularly hunt to feed themselves at outlying grazing camps, but
also there are Samburu resident in mountainous areas who might best be
described as herder foragers (Louise Sperling pers. comm.).
 If we overlook the suspicion that some ethnographers may have under
reported the occurrence of hunting, it seems fair to generalize, on the basis
of the ethnographic literature on East African pastoralists, that hunting is an
activity of ‘poor’ people – either those whose herds have been seriously
depleted by drought or disease, or ‘castes’ of hunters such as the Wata, who
live in symbiotic fashion among pastoralists. Thus, we may speculate that
the archaeological sites referred to earlier which contain considerable
numbers of bones of both domestic stock and wild ungulates may be the
settlements of ‘poor’ pastoralists attempting to re establish their herds after
some sort of calamity (cf. Robertshaw & Collett 1983, p. 74). This, and the
alternative suggestion of the existence of’castes’ of hunters among the
prehistoric pastoralists, may be tested by further archaeological research, in
which, perhaps, the identification of particular items of material culture as
being symbols of wealth and prestige will be an important goal. The
possibility that the settlements are those of hunters occasionally bartering or
raiding livestock from neighbouring pastoralists may be discounted, at least
for Gogo Falls, because the faunal remains at this site accumulated within a
very large mound of prehistoric animal dung. The diversity of the faunal
assemblage accruing from hunting may indicate whether the hunters were
‘poor’ pastoralists or ‘true’ hunters, for pastoralists, as a general rule, will
eat only those wild species that can be culturally linked to domestic
animals, e.g. eland may be eaten because they are ‘like’ cows but animals
with claws and fur may not be eaten. For the moment at least, it would seem
sensible to set aside the possibility that prehistoric subsistence systems were
so removed from anything known ethnographically that they might even
have embodied different cultural attitudes towards animals.
Finally, we may note that accounts exist which demonstrate how a
‘worthy’ man may re establish himself as an independent herd owner
through hunting:

Ware [a Borana Gutu] had lost all his stock through misfortune, and his
clansmen decided to give him ten cattle. Ware was grateful but decided to
return all the cows and keep only one ox. This ox he bartered for two
doublesized spears (bode). Then he went out to hunt with his spears and
killed two big elephants. With the profit from the sale of two sets of tusks
he was able to invest in a herd of 60 cattle … Ware thus re established
 himself as a herdowner, but he continued to kill game and to increase his
herds until he had hundreds of cattle. (Dahl 1979, p. 179)

Although ivory may not perhaps have been such a precious commodity in
the prehistoric era, the exchange of meat, skins, and other products of the
chase for domestic livestock may have made it feasible for a herd owner to
replenish his stock, and in so doing create a bone midden akin to those of
Prolonged Drift and Gogo Falls. Thus, these faunal assemblages, containing
bones of both wild and domestic animals, should be seen as the refuse
heaps of pastoralists for whom hunting was the means to a pastoral end.

Acknowledgements
I thank Cory Kratz, Louise Sperling, and Donna Klump for sharing information and ideas.

Note
1 However, pastoralists do have collective access to grazing, while hunters may have divided
access to particular resources, e.g. honey among the Okiek (Kratz pers. comm.).

References
Ambrose, S. H. 1982. Archaeological and linguistic reconstructions of history in East Africa. In The
archaeological and linguistic reconstruction of African history, C. Ehret & M. Posnansky (eds),
104–57. Berkeley: University of California Press.
Ambrose, S. H. 1984. The introduction of pastoral adaptations to the highlands of East Africa. In
From hunters to farmers: the causes and consequences of food production in Africa, J. D. Clark &
S. A. Brandt (eds), 212–39. Berkeley: University of California Press.
Barthelme, J. W. 1985. Fisher-hunters and Neolithic pastoralists in East Turkana, Kenya. Oxford:
BAR International Series 254.
Berntsen, J. L. 1979. Economic variations among Maa speaking peoples. Hadith 7, 108–27.
Bower, J. R. F., C. M. Nelson, A. F. Waibel & S. Wandibba 1977. The University of Massachusetts’
Later Stone Age/Pastoral ‘Neolithic’ comparative study in Central Kenya: an overview.Azania 12,
119–46.
Carr, C. J. 1977. Pastoralism in crisis. The Dasanetch and their Ethiopian lands. University of
Chicago, Department of Geography, Research Paper no. 180.
Collett, D. P. & P. T. Robertshaw 1983. Pottery traditions of early pastoral communities in Kenya.
Azania 18, 107–25.
Dahl, G. 1979. Suffering grass: subsistence and society of Waso Borana. Stockholm: Stockholm
Studies in Social Anthropology.
Ehret, C. 1971. Southern Nilotic history. Evanston: Northwestern University Press.
Ehret, C. 1974. Ethiopians and East Africans; the problem of contacts. Nairobi: East African
Publishing House.
Evans Pritchard, E. E. 1940. The Nuer. Oxford: Clarendon Press.
Foley, R. A. 1982. A reconsideration of the role of predation on large mammals in tropical hunter
gatherer adaptation.Man (NS) 17, 393–402.
Gifford, D. P., G. L. Isaac & C. M. Nelson 1980. Evidence for predation and pastoralism at
Prolonged Drift: a Pastoral Neolithic site in Kenya. Azania 15, 57–108.
Gifford Gonzalez, D. P. 1984. Implications of a faunal assemblage from a Pastoral Neolithic site in
Kenya: findings and a perspective on research. In From hunters to farmers: the causes and
consequences of food production in Africa, J. D. Clark & S. A. Brandt (eds), 240–51. Berkeley:
University of California Press.
Gifford Gonzalez, D. P. & J. Kimengich 1984. Faunal evidence for early stock keeping in the Central
Rift of Kenya: preliminary findings. In Origin and early development of food-producing cultures
in north eastern Africa, L. Krzyzaniak & M. Kobusiewicz (eds), 457–71. Poznan: Polish Academy
of Sciences.
Gulliver, P. H. 1963. A preliminary survey of the Turkana. University of Cape Town, School of
African Studies, New Series no. 26.
Huntingford, G. W. B. 1953. The Nandi of Kenya: tribal control in a pastoral society. London:
Routledge & Kegan Paul.
Ingold, T. 1980. Hunters, pastoralists and ranchers. Cambridge: Cambridge University Press.
Lewis, B. A. 1972. The Murle: red chiefs and black commoners. Oxford: Clarendon Press.
Merrick, H. V. & F. H. Brown 1984. Obsidian sources and patterns of source utilization in Kenya and
northern Tanzania: some initial findings.The African Archaeological Review 2, 129–52.
Monod, T. 1975. Introduction. In Pastoralism in tropical Africa, T. Monod (ed.), 1–183. London:
International African Institute.
Robertshaw, P. 1984. The prehistory of pastoralism in Kenya. Unpublished seminar paper, School of
Oriental and African Studies, University of London.
Robertshaw, P. 1985. Preliminary report on excavations at Gogo Falls, South Nyanza. Nyame Akuma
26, 25–6.
Robertshaw, P. T. & D. P. Collett 1983. The identification of pastoral peoples in the archaeological
record: an example from East Africa. World Archaeology 15, 67–78.
Shaw, T. 1976. Early crops in Africa: a review of the evidence. In Origins of African plant
domestication, J. R. Harlan, J. M. J. de Wet & A. B. Stemler (eds), 107–53. The Hague: Mouton.
Spencer, P. 1965. The Samburu. London: Routledge & Kegan Paul.
Umesao, T. 1969. Hunting culture of the pastoral Datoga. Kyoto University African Studies 3, 77–92.
Waller, R. 1976. The Maasai and the British 1895–1905. Journal of African History 17, 529–53.
20 Cattle and cognition: aspects of Maasai
practical reasoning
JOHN G. GALATY

Introduction

In anthropology, the question of pastoralist rationality and irrationality has

long served as an apt case with which to address the more general issue of
how ideas and belief systems influence economic life. Herskovits’ (1926)
notion of a ‘cattle culture’ has been associated with the point of view that
pastoralism was directed and informed by essentially religious rather than
economic precepts, and thus was essentially ‘irrational’ from the economic
point of view. An abundance of more recent research on the economic lives
of pastoralists has made quite clear the forms of rationality underlying
herding systems (Evans-Pritchard 1940), among more specialized
pastoralists for milk-based subsistence, among many agro-pastoral
communities, both for food and for other ends such as traction, manure, and
to provide stores of value (Galaty in press). However, economic studies
have tended to infer forms of rationality from economic practice, rather
than from systems of thought and the assumptions, beliefs, and aims which
underlie pastoralist discourse about their productive activities (Galaty
1984). In this respect, ‘rationality’ implies the logic rather than the
‘practical reasoning’ (Sahlins 1976) underlying pastoralism.
There are, however, cognitive concomitants of herding systems, apparent
in the quasi-formal cultural classification of domestic animals, of the arid
environment, and of forms of labour, and also in the more informal
reasoning implicated in pastoral experience as a dynamic process. These
cognitive concomitants of pastoralism do not simply replicate in mind the
objective properties of livestock, the rangelands and the labour process, but
represent a cultural selection and ordering of the terms by which
pastoralism is carried out. Such cognitive systems maintain a certain
cultural autonomy vis-à-vis the pastoral process per se, the former being
both internally structured and the subject of reflection and discourse.
However, these cognitive capacities, which are in part realized both in the
form of discrete symbolic systems and, somewhat less coherently, in the
form of implicit knowledge, do underpin pastoral practice, which in turn
provides the context within which these systems are acquired (Cole &
Scribner 1974). To what extent do young pastoralists, removed – most often
by schooling – from the routine practice of herding, attain the cognitive
skills and the command of knowledge normally associated with
pastoralism? If they do not, this may imply either that the cognitive
concomitants of pastoralism are embedded in concrete practical experience,
without which this competence is not acquired, or that such individuals are
also likely to be divorced from the more general settings in which they
might gain exposure to more autonomous systems of knowledge (Scribner
1984, Street 1984).
With such questions in mind, this chapter examines certain aspects of the
practical reasoning underpinning Maasai pastoralism: the logic underlying
extensive semi-nomadic specialized animal husbandry, the informal
reasoning associated with the pastoral experience, the more formal
symbolic classification systems through which cognitive apprehension of
the pastoral field of activity cakes place, and the processes of identifying
and remembering individual animals. By concentrating on ‘practical’
reasoning, I do not mean to diminish the significance of more abstract
aspects of’cultural’ reasoning (Sahlins 1976), such as the religious
assumptions made about the relation between animals and people, or the
social values attributed to livestock, in exchange or bridewealth. The case
can be made that Maasai pastoral ideology reflects certain higher order
symbolic notions about the social, economic, and religious value of cattle.
This essay focuses, however, on what might be construed as the ‘first-order’
level of pastoral rationality, the cognitive apprehension, perception, and
classification of domestic animals, the point at which mental and material
categories join and culture constitutes practical life. This domain of primary
cognition is especially pertinent at a time when what Braudel once called
‘material life’ (1973), the daily routines of local and domestic economy, is
being transformed by the spread of commercial production oriented to
demand far from the rangelands and the intimate tending of herds, and
when schooling is progressively altering the nature of the cognitive
experiences of young Maasai, thus transforming the basis of knowledge on
which pastoral practice has for so long rested.
The logic and experience of pastoralism

The logic of specialized pastoralism, as practised in several semi-arid land

communities in eastern Africa, can be distinguished from that of ranching
in several respects pertinent to the cognitive concomitants of animal
husbandry. The most obvious difference is that ranching is a commercial
activity, in which forms of instrumental rationality enter into production
decisions; thus primary production and the market are closely linked
through judgements and assessments of commercial value. While domestic-
level animal husbandry may also involve commercial sale, livestock may
not be produced as commodities per se, but as objects of multiple use and
personal ties. Thus, market calculation is only one, and an imperfectly
developed, mode of valuation of livestock. In contrast to ranching,
pastoralism involves relatively high labour intensity, which in less technical
terms implies that there are relatively more people active in and dependent
on the pastoral process than is true in ranching. This factor of labour also
makes possible the development of personalized ties between people and
animals. Such ties may lie at the heart of the complex symbolic and
emotional aspects of pastoralist ideas and attitudes regarding animals, but
also have implications for production, since ease of herd management,
animal handling, and efficient production of milk depends on a sort of
‘taming’ process enhanced by close personal ties and physical contact
between people and their domestic animals (Ingold 1980).
Cognitive processes, such as classification, description, pattern
recognition, and problem solving, associated with the use of cattle and other
livestock are not simply individual responses to concrete situations of
environmental use and animal production. Such cognitive operations reflect
capabilities which are institutionally shaped, both by the patterns of activity
proper to pastoralism and by the cultural strategies and formal classificatory
systems collectively encoded for carrying out pastoral tasks. This is an
important observation, since ‘psychological’ perspectives are often
theoretically distinguished from the ‘cultural’ as individual to collective
forms of behaviour. And, as previously suggested, ‘cognition’ can be
assessed both in terms of the cognitive practices of individuals and the
collective representations or institutionalized forms of symbolic
classification of a group. But, however spontaneous or individually
generative cognitive activity may be, it reflects the content of social
experience and learned principles of cognitive organization. Pastoralists
reveal an impressive ability to recognize and identify their animals, a skill
which led early observers to attribute to them phenomenal powers of
memory. Contributing to this ability to recognize hundreds of individual
animals is the development over time of personalized knowledge of a herd.
In one interview, it was observed that:

A man gets his cattle when they are few. So he actually comes to know
all the calves that are born into the herd, until they become many. By the
time he has a very large herd, he also knows them by their colours and
their cattle ‘houses’. So you find that the man is in no sense a stranger to
his own herd.

But given the gradual acquaintance of herders with members of the herd,
there are those who are known to be skilled in recognition and
identification, the Abarani, a person with a very good memory, who can
look at a cow very quickly and know which one it is, observing colours and
idiosyncratic features. Describing the able herdsboy, it was suggested that:

One can bring such a young boy to look after a large herd and the boy
can master the cattle very quickly after very few instructions.

The able Abarani is contrasted to the inept Mankalioni, who cannot

‘master’ the herd, who cannot tell the difference between two similar cows.
Young pastoralists who attend school are perforce removed from the
setting in which the cognitive capacities of herding are learned, both
through experience and tutorial. Drawing schoolchildren outside the context
of direct learning and indirect experience of the domain of animals and their
husbandry inevitably influences both the content and organization of
cognitive reasoning. Yet, other skills, such as literacy and arithmetical
calculation, learned through school may also be pertinent to animal
husbandry (Goody 1977). This may be especially true when such skills are
combined with the transformation of needs, values, and social and
economic orientations, often associated in rural schools with the new social
and cultural experience which perforce accompanies the pedagogical
experience. Taste for a diversity of foods, store-bought commodities, access
to an economy of money, and class-inflicted identities correlated with
educational status, indirectly shape new attitudes towards animals and their
production, which are seen more in terms of impersonal commerce than
personal and domestic livelihood. But such a shift also implies an
increasing lack of experience in or declining motivation regarding
pastoralism, which can result in a decline in cognitive capacities, including
the content and organization of the skills and knowledge, entailed by
careful, personal, and productive husbandry. For many educated
pastoralists, the shift in orientation is away from pastoralism altogether, but
for many others the shift is in the logic of animal husbandry from
pastoralism towards commercial ranching (Galaty 1981).
It is clear that intimate and extensive experience with livestock inculcates
the capacity to identify and remember animals in general, but that
personalized acquaintance with the individuals of a specific herd leads to
knowledge and recognition of animals in particular. Such perception of
animals is in part holistic, depending on the recognition of patterns of
visual and behavioural features which represent the uniqueness of each
individual. The personal recognition and identification of up to 1000
individual animals may be remarkable, but when understood as resulting
from years of intimate contact with a herd, and with each individual animal
from birth, no more remarkable than the common capacity to recognize up
to 1000 human faces of people known over a span of years. However, the
holistic nature of personalized pattern recognition, so important to
pastoralism, is in practice underpinned by analytic processes, in part subject
to conscious reflection and deliberation. These conscious and unconscious
processes involve the perceptual encoding of features, not simply based on
visual form but drawn from more formal cultural classification systems.
These classification systems, representing a multi-levelled formal structure
of cognition based on language, provide herders with the means to
comprehend and master the social and natural worlds of pastoralism by
ordering the complex, intimate, and nuanced experience of domestic
animals, productive resources, and labour in terms of a much smaller
number of symbolic attributes and features.

Cattle and classification

In addition to patterns of identification built up through experience,
livestock identification also occurs through systems of symbolic
classification of a cultural order, in particular through earmarks and brands
which signify the owners’ family and descent group, livestock names, and
descriptives. In the case of names, the Maasai endow all cattle in a
particular ‘lineage’ with a single appellation, proper not to the apical bull
but to the bovine ‘matriarch’. Table 20.1 is a list of conventional cattle
names, representing some of the more commonly used. As can be seen,
names tend to be representational rather than arbitrary, being slight
deformations or transformations of other terms, with appellative prefixes
denoting the gender of the animal. For example, Nkeyi is a slight
phonological variant of Keri, meaning ‘speckled’ or ‘splotched’, with a
feminine gender prefix (na). The masculine form would be Lukeyi. The
names thus serve not only as indices of particular cattle lineages or
individuals, but also signify in a representational manner, ‘connoting’ as
well as ‘denoting’. There tend to be five distinct modes of signification in
cattle naming, three utilizing symbols of transactions, two forms of visual
description.
In the Transaction Mode proper, names depict the type of transaction by
which the apical animal was acquired: the return of a debt, an exchange, for
bloodwealth, bridewealth, as a ‘pure’ gift to a friend, from a raid, etc. In the
Donor Mode, the name signifies the giver of the apical animal. In the
Reciprocal Mode, the name signifies the reciprocal presentation for which
the apical cow was exchanged: ‘reminders’ of oxen, donkeys, or other
animals exchanged, the ‘eye’ for which a gift was compensation, the ‘girls’
given in marriage, or the ‘meat’ of the stolen animal, for which the return
was compensation. There are two Descriptive Modes, which use
‘descriptives’ by colour, form, or status to signify either the animal given in
exchange (Reciprocal Description Mode), or the animal received (Reflexive
Description Mode). It is especially interesting that only one of the naming
modes actually signifies the apical animal received and owned, while all the
others signify the transaction itself, the animals given in exchange, and thus
no longer in possession, or the exchange partner. Cattle names, then, are
less about bovine possessions than about social ties and transactions. In
most cases, names which denote animals actually connote – or ‘signify’, in
the form of the images or descriptions which constitute names – not the
animal but the social and economic transactions by which it, or its apical
ancestor, was acquired.
In contrast, ‘descriptives’, being based on attributes, by necessity
connote, but can also be used to denote. In this sense, descriptives serve
many of the functions of names, and, indeed, constitute names in the
Descriptive Modes noted above. In such cases, names presumably began as
descriptives, and thus describe as well as denote. But Maasai livestock
descriptives refer to individuals in a way conventional names cannot, and
most cattle names do not, by evoking the distinctive features of specific
animals rather than serving simply as demonstratives. While names are used
to call cows, especially to their calves, descriptives are used to refer to
them, for the benefit of those who do not share personalized acquaintance
with the herd (Galaty 1982).

Table 20.1 Maasai cattle names.

Name Meaning Comment Signifies
Kusaka debt returned payment, overdue debt TRANS
Nawolli exchanged one received in exchange TRANS
(a-wol)
Meretoi helper gained after a struggle TRANS
Kirror bloodwealth from killer of kinsman TRANS
Torosi bloodwealth from killer of kinsman TRANS
Kiripa for care (a-ilip) given for care TRANS
Reyio by raid stolen in raid TRANS
Noosilan bridewealth bridewealth payment TRANS
Kerreti a blessing given to family TRANS
member
Entotua in friendship gift from dear friend TRANS
Nyorra our love from beloved friend DONOR
Nemwala of Wala received from Wala DONOR
Nenkong’u of the eye payment by bride’s RECIP
father
Noo-Ntoyie for girls bridewealth RECIP
Nor-Mong’i reminder of an ox exchanged for ox RECIP
Nyamu for meat for animal stolen and RECIP
eaten
Mon’go Olkutai reminder of an ox exchanged for ox RECIP
Noosirkon for donkey exchanged for donkey RECIP
Noonkeya for children received for children RECIP
Mong’o kinyi reminder of a small ox exchanged for a small RECIP
ox
Mong’o Sampu for striped ox exchanged for striped DE/RC
ox
Noldere for dusty (enderit) one exchanged for whitish DE/RC
animal
Ngiyo of greyish-brown one exchanged for greyish- DE/RC
brown
Nado Yukunya of red-headed one exchanged for red- DE/RC
headed cow
Nayasha of white-patched one exchanged for N- DE/RC
arasha coloured
Leyaai yellowish (Olerai) named after Acacia DE/RF
seyal
Nkeyi speckled (Keri) white variegation DE/RF
Kerenket in a hole born in valley or hole DE/RF
Nauroyoyi suckler suckled when mother DE/RF
dies
Kutiti nki with small teats describes the animal DE/RF
Sampu brown-striped describes cow DE/RF
Kurtolet slightly-blind describes cow DE/RF
Nenkong’u nabo one-eyed describes cow DE/RF

TRANS, signifies the transaction; DONOR, signifies the giver of the animal; RECIP, signifies the
reciprocal presentation exchanged for the animal; DE/RC, description of the reciprocal animal;
DE/RF, reflexive description of the animal itself.

Table 20.2 Maasai classification of livestock by status.

Males Females
cattle:
calf olashe enkashe
immature olbung’ai entawuo
mature doing’oni enkiteng’
castrate olkiteng’
goats:
kid olkuo enkuo
immature ositima esupen
mature oloror enkine
 castrate olkine
sheep:
lamb olkuo enkuo
immature ositima esupen
mature (ram/ewe) olmeregesh enker
castrate olker

Cattle descriptives tend to utilize five distinct modes of attribution:

status, colour, pattern, horns, and distinctive characteristics. In the status
mode (see Table 20.2), an animal is denoted by its sex (male/female),
maturity (calf/immature/mature), and reproductive status (castrate/whole).
It is worth noting that the principles underlying lexemes in the status
mode make discriminations relevant to animal reproduction, between
offspring, pre-productive yet whole animals, productive animals, and
castrates. Yet, lexical symmetry occurs not between the productive males
and females but between castrates and productive females, with productive
bulls, he-goats, and rams representing the ‘marked’ category. In fact, very
few productive males are kept for breeding purposes, both as a strategy of
selection and as a mechanism of herd control, since whole males are
notoriously ill-behaved and disruptive of the herd. The quintessential
animals, which constitute the bulk of the herd, are thus lexically paired:
castrates and productive females. Symbolically, the ideal male pastoralist
analogue is not the bull but the ox, which exemplifies virtues of size,
strength, beauty, and sociability, as opposed to the wiry, scrawny, and ill-
behaved bull (Beidelman 1966). The lexical parallels between goats and
sheep reflect as well the joint management of the two species in a single
herd (in-tare), in contrast to the separate and more prominent cattle herd
(em-boo).
The second and third principles are not really distinguished by Maasai,
and should be grouped together as a single principle of ‘colour-pattern’ or
‘visual form’. It is only to heighten the distinctiveness of the pattern mode
that I treat it here separately from the colour mode (Table 20.3). Patterns are
more complex, being evoked by both single lexemes and phrases, referring
to the distribution of bodily colours (Table 20.4). Patterns may also be
evoked by compound phrases, composed of nouns and qualifiers, which
signify colour combinations or colours localized on the animal’s body
(Table 20.5).

Table 20.3 Maasai elementary colour classifiers.

Narok ‘black or bluish’
Naibor ‘white’
Sima ‘tannish’ (i.e. ‘colour of eland’)
Sero ‘dark brown’ (i.e. ‘wilderness’)
Elerai ‘cream-like or yellowish’ (Acacia seyal)
Barrikoi ‘reddish’
Nanyokie ‘brownish red’
Nado ‘red’

Table 20.4 Maasai elementary pattern classifiers.

Mugie ‘dark, blackish variegation’ (splotched)
Keri ‘white streaked or lined’ (cf. Oldoinyo Keri, Mount Kenya)
Tara ‘brown spots on the body’
Sintet ‘spotted’ (black)
Arus ‘spotted’ (black and white on stomach)
Owuaru ‘many brown splotches’ (lit. ‘beast-like’)
Sampu ‘striped’ (brown on the body)
Olmotonyi ‘white with a brownish head’ (lit. ‘vulture-like’)
Dere ‘whitish with black around the neck’ (lit. ‘dusty’)
Ngencherii ‘brownish and reddish’
Narasha ‘black, red, or brown with white sides’
Nairimo ‘multiple small spots’
Otialei ‘white eye’

In the horn mode, the presence or absence of horns, their shape, and their
combination with other features provide principles for the genesis of
descriptives (Table 20.6). Other distinctive features regarding physical state
can be used to describe animals, apart from colour, pattern, and horns
(Table 20.7).
 In actual discourse, elements derived from the five principles are
combined to form descriptives, as in Enkiteng’ Odo Lukunya, ‘the cow with
a red head’. Although colour, pattern, horns, and other characteristics are
often mutually exclusive, they can be combined with the basic status of the
animal in an additive way, such as in Enkiteng’ Narok Oomswuarak, ‘black
cow with long horns’, and in cases of ambiguity a further feature can
always be added to distinguish the animal in question from similar ones.
Thus, it is possible to identify cattle not only through personal knowledge
and individual names, but also through descriptives by which they can be
referred to in such a way as to render each animal in a herd or region fairly
distinctive. As can be seen, the actual number of lexemes are relatively
limited, but in combination with other lexemes great specificity can be
quickly achieved. To the unfamiliar, cattle or other animal species may
appear homogeneous, but by using only a few of the five descriptive
principles just mentioned they can be easily distinguished in verbal form,
facilitating actual description and identification.

Table 20.5 Maasai compound colour-pattern classifiers.

Narok Sopia ‘black with some white hair’
Narok Lukunya ‘black head, white body’
Narok Abori ‘black stomach’
Naibor Oshoke ‘black with a white stomach’ (black animal)
Nado Lukunya ‘red head, white body’
Sampu Sintet ‘brownish streaks and black spots’
Sampu Mugie ‘blackish in colour with brownish streaks’
Sampu Kumpau ‘with reddish stripes’ (tobacco colour)
Sampu Lerai ‘cream colour with brown streaks’
Narasha Sampu ‘white sides with brownish stripes’
Narasha Sintet ‘white sides and black spots’
Keri Nasiantet ‘white streak with small black spots’
Keri Ng’iro ‘brown with a white streak on the back’
Keri Barrikoi ‘reddish with white on the back’
Keri Abori ‘white stomach’ (white-streaked below)
Keri Enyokie ‘red body with white line across the body’
Enkeri Sarioni ‘short, incomplete stripe on back’
Keri Ai ‘long white streak on body with bluish spots’
Keri Sampu ‘brownish stripes with white line’
Ol tara Mugie ‘blackish with large brown spots’
Owuaru Leleshwa ‘whitish with brown splotches’
Nairimo Enyokie ‘with many red spots on the body’
Nairimo Erok ‘with many black spots on the body’

Table 20.6 Maasai horn configuration classifiers.

Elementary terms
Ruma ‘without horns’
Arro ‘horns facing down’
Ng’elesh ‘one horn up and one down’
Compound terms
Arro Oirrag Imowuarak ‘horns facing down and close to head’
Ng’elesh Napir ‘one horn up, the other down, and fat’
Narok Ruma ‘black with no horns’
Narok Oomswuarak ‘black with long horns’
Mugie o Mowuarak ‘dark brown with horns’
Ong’ow Imowuarak Siadi ‘horns facing the back’
Mugie o Mowuarak ‘dark brownish with horns’
Imowuarak Siadi ‘horns to the side’
Ogama Imowuarak ‘with horns that almost interlock at front’
Olua Lukunya ‘horns facing the sides’

Table 20.7 Maasai somatic livestock classifiers.

Shompole ‘blind in one eye’
Ng’ojine ‘lame’
Erruk ‘humped’ (i.e. large hump)

The system underlying Maasai cattle descriptives is mastered by most

people involved with the pastoral process, and involves familiarity with
basic terms, the rules of their combination, and the conventional
associations between visual patterns and these linguistic descriptives.
Colour patterning also conveys symbolic meaning and bears cultural value,
but that topic is beyond the scope of this chapter (Fukui 1979).
 As an indirect test of the mastery of descriptives by schoolchildren and
those not attending school (home-people), several boys were requested
simply to enumerate as many cattle descriptives as possible. It could be
assumed either that the responses of schoolboys would reflect less
command of livestock vocabulary and descriptives than those of ‘home’
boys, or that the facility with which boys are able to verbally respond would
increase with schooling, offsetting any differences in knowledge or
cognitive competence in the pastoral domain.
The schoolboys generated only 57 per cent of the number of descriptives
generated by the home boys, the latter appearing able to list descriptives
virtually indefinitely. There was also a strong tendency for schoolboys to
list more elementary terms and fewer complex compound terms than home
boys. Schoolboys listed 79 per cent elementary and 21 per cent compound
terms, while home boys listed 56 per cent elementary and 44 per cent
compound terms. The compound terms listed by home boys also tended to
evoke a greater range of principles and types.
It seems accurate to conclude from this sample that all boys tend to know
some of the basic elements and rules of generating descriptives, but that the
non-schooled have a much larger repertoire of colours and patterns at their
disposal, and are able to generate multi-lexemic combinations with greater
facility than are schoolboys. Girls also demonstrate mastery of the
principles of livestock descriptives, and tend to generate, spontaneously,
more descriptives for small-stock than cattle, probably due to their greater
involvement in sheep and goat-herding. Those removed from direct and
daily involvement in animal husbandry through schooling tend not to
master terms, rules, or associations to the same extent that the non-schooled
do, presumably due to differences in exposure and motivation.

Cognitive processes: identifying the missing animal

There are three quite different cognitive tasks entailed by the identification
or recognition of livestock:

(a) identifying one’s own animal among those of others, a problem of

pattern recognition;
(b) identifying another’s animal among one’s own, a problem of matching
patterns and verbal descriptions; and
(c) recognizing when an animal is absent from one’s own herd.

The third task is quite different from the first two, since what must be
identified is absent rather than present; however, the task also involves
pattern recognition, in that an entire herd forms a complex configuration
which is altered when an individual is missing. This cognitive ability, of
recognizing the ‘missing figure’, is quite important in pastoralism, where a
relatively small number of herders care for a relatively large number of
animals, and where theft and loss to predators is always a threat. Personal
familiarity with one’s herd surely underlies the ability to recognize which
animal out of several hundred is not present, and this holistic ‘sense’ of a
slightly incomplete or imperfect pattern plays a role in this modality of
perception. However, underlying any holistic and synthetic assessment, this
process is essentially ‘analytical’, and depends not on one principle but on
the interaction of several in what proves to be a systematic process of
deliberation and cognitive ‘search’ rather than one of global holistic
apperception.
A herd-owner often meets his herd a distance away from the main kraal,
as it is brought home by the young herders, and helps to drive the herd
along the final stretch to the main gate where the calves await their mothers.
The individual casually examines the animals and looks for signs of
sickness or for disruption of the usual herd pattern. The scrutiny continues
as the herd enters the gate and settles throughout the great central kraal, in
the midst of the circular boma. As evening comes, the herd-owner continues
to ‘ruminate’ on the herd, reflecting on whether all the livestock have
returned. This examination, scrutiny, and reflection involves a fairly self-
conscious evocation of a number of quite different dimensions of
classification of animals; in effect, a ‘search’ can occur down one line of
classification, followed by another cross-cutting ‘search’ down another.
Following the principle of redundancy, the more dimensions of
classification that exist, the greater the likelihood of coming upon the
missing cow.
Maasai elders proved able to discuss in great detail the procedures used
in scrutinizing their herds and checking for missing animals. One elder
reported:

Normally, we elders and younger men who have livestock stand next
to the gate when the animals are returning. You give them a very
thorough look to see which one has gone into the Enkang (kraal) and
which one hasn’t. You also thoroughly observe their colours. So by the
time they have all gone into the home, you know that such-and-such a
cow has disappeared and such-and-such a cow is present.
Then you walk around the cattle yard in order to identify which cow
you haven’t seen. The owner of the herd always knows which cattle
are generally left behind by the rest, and which are always leading the
herd. You can also observe carefully to know which ones always graze
at the sides of the herd, refusing to stay with the others. So when the
cattle come in you look for all those cattle first before the rest.

The dimensions of classification are essentially spatial and genealogical,

but also involve visual signs which distinguish individuals from each other,
and set out especially vulnerable categories. The moving herd has a definite
structure, with certain individuals and cohorts walking in the front, the
sides, and the back of the herd; some lead while others commonly spread
the herd, drawing by example a few neighbours away from the major body
to more peripheral areas of grazing. Often the same spatial configuration
will be maintained as the herd moves towards home, and the herder and the
manager will initially look for the more ambitious animals, who might tend
to get lost or to stray. With their arrival at the main gate, animals often enter
in a roughly similar order, and at that time men and women will search for
individual animals who need special care: those in milk who must be paired
with calves, the diseased or injured, or those which are pregnant. Thus, the
herd is implicitly segmented into special categories, each having a few
individuals. Then the herd settles into the great enclosure, the Boo, and
assumes a spatial pattern which also has stability and regularity over time.
Some individuals stay near the fence, others near their owners’ homes,
while others prefer the centre of the enclosure. As people move through the
herd, checking their animals, they use the location of animals within the
kraal as another dimension of organization, whereby the missing animals
can be identified.
 The structures of animal classification described in the previous section
are explicitly drawn upon in the process of herd assessment. Finally, the
pastoral family sits inside the home, eating supper or talking. In addition to
the topics of the day, a good herd-manager will silently consider the well-
being of the herd, and will deliberate on his animals. A common technique
often mentioned is to think about each genealogical livestock ‘house’, the
descendants of a given female cow classified together by being called by a
single name. A person can quickly review an entire herd through its
genealogical structure, by which it is broken down into a series of
cognitively manageable units. One informant relates:

Another trick we use is that we owners know our cattle ‘houses’

[lineages] and colours very well. If we are doubtful of a given animal, we
can start counting one house of a given cow to the end, and then start on
another one. We are also very keen about the colours. By doing this,
nothing will be overlooked, as you might know that a particular cow has
a certain number of calves and you also know its colour very well. Also,
every woman knows the number of her cattle and their colours. Also the
husband knows which cows belong to which woman. In the process of
counting, he will actually identify that a particular cow which may
belong to one of his wives is not present. The women can quickly tell if
any of their cows are missing.

With somewhat less elaboration, a herdboy mentioned many of the same

principles of scrutinizing a herd:

I count the cattle according to their houses. If a cow has one calf, I count
that. If it has more I count them until I have accounted for them all. I also
note the cattle that are always behind and those that always lead the rest.
And also those that graze at the sides of the herd, and the sick ones. I also
look for those cows that have young calves. During the rainy season,
these cows are quick to sneak back home. When I discover that one is
missing, I will look for it in other herds. I will also describe it [to other
herders] while I look for it, according to its colour, its earmarks and
firemarks [brands] it may have.

Then, it is said, one sleeps and during sleep thinks and dreams about cows,
and in the middle of the night or in the morning the herder or owner will
know if such-and-such an animal is missing. The elder describes:

Another method we use is that when one goes to sleep one counts one’s
herd in one’s sleep. In most cases, you may find out that there is a cow or
an ox you haven’t seen while counting the herd. In the morning, one
checks the herd once more to see if the cow or ox one had not seen in
one’s sleep is in or not. Sometimes it’s likely that it may be in but was not
noticed. It is also true that a person finds the cattle that he hadn’t seen in
his sleep are missing in the morning. Then he will ask the person who
actually tends the herd to check and find out the last time he had seen it.

The association between sleep, dreaming, and cognitive deliberation on

one’s cows was drawn not by one but by every informant consulted. It was
suggested separately by many that by dreaming one could often review
one’s herd and come up with a missing animal more effectively than during
conscious waking review of the herd.

Some concluding remarks

The perceptual process described is not holistic, nor is it iterative, by which

a missing animal would be identified if the count was off. Rather, the herd
is symbolically differentiated by a series of analytical dimensions, some of
which are permanent, such as genealogy, some of which are quite
ephemeral, such as states of well-being. A quite large herd is conceptually
manageable when broken down into a much smaller number of categories,
through a process similar to what has been called ‘chunking’ by cognitive
psychologists. Genealogy represents the most well-defined and
unambiguous principle of classification. One herd of 54 animals described
was subdivided into 13 ‘houses’ or matrilineages, composed of an average
of four animals each, with a range from two or seven members. Of the herd
80 per cent were female, 20 per cent male, and of the females 16 were
identified as having already given birth, 27 presumably being heifers.
Extrapolating, a herd of 150 might be subdivided into around 37 houses,
though such a large herd might actually tend to be composed of fewer
houses with more members. Since these houses are also allocated to
specific matricentric households in the polygynous family, the place of any
given animal within the genealogical structure is quickly narrowed.
 The herd’s use of space outside and inside the kraal represents perhaps
the most direct and accessible system of scanning by the herder to
determine whether the herd is in order. But genealogy (through naming)
probably represents the most reliable mode of cognitive assessment of the
wholeness of the herd. But taken together, the modes of cognitive
classification and processing constitute a structure, by which the herd is
symbolically subdivided in several cross-cutting ways, providing
redundancy in the ordering of animals and for the implicit ‘search’ for
missing members. Any individual animal is classified spatially,
genealogically, and individually; the herder ‘searches’ each dimension, and
if in doubt ‘searches’ another, based on the state of the herd at the critical
moment. The impressive memories of pastoralists are so not because of
their greater potential for perceptual memory, holistic pattern recognition,
or cognitive capacities for ‘storage’ and ‘review’, but because of the nature
of their symbolic organization of domestic animals. Multiple dimensions of
cultural classification provide for cognitive organization and redundancy,
allowing them to efficiently process smaller ‘chunks’ of the herd and to
cross-check one result with another.
These forms of ‘practical reasoning’ represent technical achievements in
pastoral culture, underpinning careful husbandry and facilitating higher
overall productivity, both in terms of labour and aggregate output.
Knowledge of individual animals in a herd, systems of naming and
describing them, and mastery of processes of classification and
identification, are not evenly or randomly distributed throughout pastoral
communities, although there are always those known for their cognitive
skills or lack of them. Rather, pastoral cognitive competence – regarding
both content and process – is gained both through tutorial and experience,
and depends not just on passive participation but also on the sort of
motivation and high valuation of pastoralism that was once taken for
granted. The cultural systems of symbolic classification and cognitive
processing described here represent a critical concomitant of pastoralism
viewed as a highly productive and viable way of life in the semi-arid
rangelands of East Africa. The knowledge and cognitive competence so
represented may well decrease as pastoralist children turn towards formal
education, and the subject matter of the school curriculum, and away from
the pastoral experience through which competence is acquired. It should not
be assumed that pastoral economic life merely represents a set of work
routines, or a simple strategy of subsistence production, which can only be
improved through education or the more elusive process of’modernization’.
This chapter has attempted to demonstrate that there are cognitive
concomitants of the resilient arid-land adaptation represented by specialized
Maasai pastoralism on which successful animal husbandry depends, and the
loss of this basis of knowledge, motivation, and strategy will influence the
viability of this arid-land economy and culture.

Acknowledgements
This research was supported by the joint McGill/Kenyatta University College project on Cognition,
Education and Work, through a Cooperative Grant from the International Development Research
Centre of Canada. For additional research support, I am also indebted to the Graduate Faculty of
McGill University, the Social Sciences and Humanities Research Council of Canada, and the Fonds
F.C.A.R. of Quebec. The assistance of the Bureau of Educational Research of KUC and its Director,
Professor George Eshiwani, is greatly appreciated, as is collegial comment by other members of the
project in Kenya and at McGill. The competent research assistance of Mosinko Ole Tumanka in
gathering much material used here is acknowledged with appreciation.

References
Beidelman, T. O. 1966. The ox and Nuer sacrifice: some Freudian hypotheses about Nuer
symbolism. Man 1(4), 453–67.
Braudel, F. 1973. Capitalism and material life 1400–1800. New York: Random House.
Cole, M. & S. Scribner 1974. Culture and thought. New York: Wiley.
Evans-Pritchard, E. E. 1940. The Nuer. Oxford: Clarendon Press.
Fukui, K. 1979. Cattle colour symbolism and inter-tribal homicide among the Bodi. In Warfare
among East African herders, K. Fukui & D. Turton (eds). Osaka: Senri Ethnological Studies.
Galaty, J. 1981. Land and livestock among Kenyan Maasai: symbolic perspectives on pastoral
exchange, social change and inequality. In Change and development in pastoral and nomadic
societies, J. Galaty & P. Salzman (eds), 68–88. Leiden: Brill.
Galaty, J. 1982. Being ‘Maasai’, being ‘people-of-cattle’: ethnic shifters in East Africa. American
Ethnologist 9(1), 1–20.
Galaty, J. 1984. Cultural perspectives on nomadic pastoral societies. Nomadic Peoples 16 (October),
15–29.
Galaty, J. in press. Pastoral and agro-pastoral migration in Tanzania: factors of economy, ecology and
demography in cultural perspective. In Production and authority, J. Bennett & J. Bowen (eds),
Society for Economic Anthropology, Washington DC: University of Americas Press.
Goody, J. 1977. The domestication of the savage mind. Cambridge: Cambridge University Press.
Herskovits, M. 1926. The cattle complex of East Africa. American Anthropologist 28, 230–72, 361–
88, 494–528, 633–64.
Ingold, T. 1980. Hunters, pastoralists and ranchers. Cambridge: Cambridge University Press.
Sahlins, M. 1976. Culture and practical reason. Chicago: University of Chicago Press.
Scribner, S. (ed.) 1984. Cognitive studies of work. The Quarterly Newsletter of the Laboratory of
Comparative Cognition 6 (1&2), 1–46. San Diego: Center for Human Information Processing.
Street, B. 1984. Literacy in theory and practice. Cambridge: Cambridge University Press.
21 Prehispanic pastoralism in northern
Peru
TOM McGREEVY

Introduction

In the summer of 1982 I conducted an archaeological survey on the high

grasslands around the town of Huamachuco in northern Peru (see Figs 21.1
& 21.2) under the auspices of the Huamachuco Archaeological Project.
Evidence for prehispanic pure pastoralism was sought in this area.
Unexpected results forced a study of Peruvian camelid pastoralism so as to
allow interpretation of the observed patterns from the north highlands in
prehispanic times (McGreevy 1984). This chapter summarizes that study.

The South American camelids

The animals herded in Peru were llamas and alpacas, members of the family
Camelidae, genus Lama. There is debate about whether they should be
separated at a species or a subspecies level (Miller 1979, pp. 1,4) but there is
no doubt that to Andean natives they are functionally distinct (Miller 1979,
p. 7). Furthermore, conditioned by their functions, the range of alpacas is
narrower than that of llamas.
Llamas are used today primarily as beasts of burden. Average loads vary
between 20–40 kg, which can be carried up to 30 km per day (Stouse 1970,
p. 138). Secondary uses include wool and dung production. Llama wool
generally is coarser than alpaca wool and is used for the production of
household goods such as blankets, rugs, and rope (Franklin 1982, pp. 465,
468). It can be used for the manufacture of clothing, but alpaca wool is
preferred. Dung is used both as fuel and as a fertilizer (Flores Ochoa 1975,
p. 307). Llamas are also used for meat and bones; fat, sinews, and pelts are
all used in household products (e.g. Stouse 1970, p. 138). They play an
important role in ritual during individual rites of passage and at times of
collective stress (Stouse 1970, p. 138, Flores Ochoa 1975, p. 307).
Alpacas are primarily wool producers. Their wool is very fine and is used
for the production of textiles (Webster 1973, pp. 121–2). Other uses are
similar to those of llamas, save that alpacas do not carry cargo and are less
important in ritual. Alpacas are the most important animals in Peru today
because their wool can be easily traded or sold, and will become
increasingly so as it penetrates further into international markets (Franklin
1982, pp. 468–9). Llamas become less important as mules and trucks replace
them.
 Figure 21.1 General map of Peru.

In Inca times the relative importance of the animals appears to have been
reversed. The llama’s functions were much as they are today; in addition,
llama wool had greater use and llamas played a larger role in ritual. Their
wool was used for the cloth of the common people where alpaca wool was
absent. However, the Inca state is known to have redistributed alpaca wool
into areas where alpacas were absent, to be woven by the people both for
their own and the state’s use (Murra 1965, pp. 102, 124–5). Thus the degree
of increased importance for llama wool in textile production is complex to
determine. In ritual, llamas played a role in state religion, in addition they
were used in individual and collective ceremonies (Murra 1965, p. 195,
Tschudi 1969 (1885), p. 131). The great importance of llamas to the Inca can
be seen in two facts. First, gifts of llamas by the state were sufficiently
special to help ensure loyalty of subjects (Murra 1965, p. 205). Secondly, the
Inca strongly encouraged the expansion of llamas throughout the empire
(Murra 1962, p. 711).
 Figure 21.2 Map of north central Peru.

There is less information from the Inca period about alpacas. Certainly
they were kept primarily for wool production. Almost all the fine cloth used
by the élite was woven from their wool. These textiles were among the most
highly valued commodities in Peru (Murra 1962). Thus, the importance of
alpaca wool must not be underestimated. Other alpaca uses were similar to
today. In addition, they had an increased role in ritual in Inca times, although
the role is unclear. What is clear is that the Inca did not try to expand the
range of alpacas as they did llamas. This, as will be argued, is due to the
specialized environment alpacas require to produce good wool.
The question of the relative importance of the animals in earlier times
remains unanswered. However, I suggest that alpaca wool has always been
important since domestication of the animal. The acquisition of alpaca wool
by groups without direct access to alpacas is an interesting research problem
(Topic et al. 1987). Llamas became important in northern highland Peru by
the Early Intermediate Period, as they were present in most areas by this
time (see Table 21.1 and summary in McGreevy 1984, pp. 117–19). The
importance of llamas derives only from their day-to-day use.
Alpacas range from central Peru to extreme northern Argentina, rarely
occurring below 4300 m above sea-level (ASL) (Gilmore 1948, p. 441, Gade
1969, p. 341). They are highly dependent upon the succulent pasturage
found in high wet areas, known as bofedales. When raised in these areas
they are generally healthy, but when raised elsewhere they are prone to
diseases, the most serious of which is mange (Webster 1973, p. 121, Orlove
1977, p. 207, and summary in McGreevy 1984, pp. 45–8). When affected by
any disease, alpacas suffer a decrease in wool quality and quantity (Orlove
1977, p. 207). Since they are kept for wool production, their range is
restricted to the zone where they produce the best wool, which is not to say
that they cannot survive in other zones. It has been argued that in prehispanic
times alpacas were found down to 3500 m ASL (Flores Ochoa 1982, p. 69).
However, it has not been conclusively proven that herds living at this
altitude were functioning as wool producers. Rather, several of the sites
appear to be collection points for slaughter.
Table 21.1 Peruvian chronology (prehispanic periods)
Late Horizon AD 1476–1532
Late Intermediate Period AD 1000–1476
Middle Horizon AD 600–1000
Early Intermediate Period 200 BC–AD 600
Early Horizon 900–200 BC
Initial Period 1800–900 BC
Preceramic periods ?–1800 BC

Llamas are found today scattered throughout Peru, Bolivia, Chile,

Argentina, and Ecuador, rarely below 2700 m ASL (Gade 1969, p. 341).
They survive well upon drier forage than alpacas and are less susceptible to
disease (Flores Ochoa 1975, p. 303). In prehispanic times llamas were more
widespread, living on the coast as well as in the highlands (Gilmore 1948, p.
437).
While it is likely that only llamas would have been found in northern Peru
in prehispanic times, the type of pastoralism that would have been practised
in the area is less clear.

Types of pastoralism

There are major physiographic differences between highland northern and

southern Peru. In southern Peru there are large expanses of high grasslands
suitable for herding. Most of the land is over 4000 m ASL (see Fig. 21.3).
Agriculture has limited potential even with crops adapted to high altitude,
due to early frosts, hail, snow, or excessive dryness (Isbell 1978, p. 307);
thus crop failure is not uncommon. In the north, the land is more broken and
lower as well as wetter, and agriculture has much greater potential.
Pure pastoralism and agropastoralism are found in Peru today. Pure
pastoralism is found in the south on the high grasslands as an adaptive
response to an environment largely unsuited for agriculture. Both llama and
alpaca herds are raised with a preference for the more valuable alpacas.
Access to other goods is through trade or barter (Flores Ochoa 1975, pp.
309–11, Custred 1977, p. 119).
 Camelid agropastoralism is found today only on the eastern slope of the
Andes and in deep valleys in highland southern Peru. Agriculture is the main
subsistence source with pastoralism complementary to it. Llamas are the
most important herd animal as they provide labour for transportation
(Webster 1973, p. 121, Yamamoto 1981, pp. 127–8) and dung for fertilizer,
which can be crucial for maintaining soil fertility (Guillet 1983, p. 563).
The antiquity of pure pastoralism is a matter of debate. This adaptation
conflicts with ‘vertically’, a model of land use which presupposes that lands
from different production zones are controlled by a single community which
thereby has direct access to the products of each (Murra 1975). This model
might be relevant only for the last few centuries of prehistory.
In northern Peru in prehispanic times, it is believed that agropastoralism
was the dominant high-altitude land-use strategy. As in the eastern Andes,
high-altitude herding lands are interspersed with lower agricultural lands and
not isolated from them.
 Figure 21.3 General physiography of Peru.

North Peruvian Survey data

Roughly 1200 hectares of land between 3700–4250 m were surveyed for

evidence of prehispanic pastoralism (McGreevy & Shaughnessy 1983). Of
the 29 sites found, only four gave any indication of a pastoralist orientation.
Two of these were undated while the other two contained both Late
Intermediate Period and modern sherds. None of them was a long-term
habitation site and only one had substantial corral space, with seven large
corrals. Thus, the evidence suggests that in prehispanic times this zone was
not used as a long-term habitation area by pastoralists, and that short-term
use occurred only in late prehispanic periods.
An area of 275 hectares between 3300–3600 m was surveyed as a
comparative sample. In this elevation 16 sites were found, all dating from
the Early Intermediate Period (McGreevy 1984, pp. 148–54). The sites were
predominantly agropastoralist, with the layout of some suggesting that
specialist herders may have cared for the animals. This suggestion is
tentative as no excavation was carried out. Sites in this area were situated
along a ridge-top with ready access to higher grazing lands and lower
agricultural lands. Daily herding on the higher lands would leave few traces.
This type of settlement pattern is not unique in this area (see for other
examples McGreevy 1984, p. 166, Shaughnessy 1984, DeHetre 1979, pp.
32, 143–4). Further survey obviously needs to be conducted in the north
highlands to determine the general high-altitude land-use strategy; however
from the present evidence it appears the strategy was agropastoralism.

Conclusions

It is suggested that pure pastoralism was not practised in prehispanic

highland northern Peru. Rather agropastoralism was the norm on high-
altitude lands. Llamas were the herd animals providing labour power, dung,
wool, and meat. Alpacas were absent due to their requirements for good
wool production. Animal numbers increased over time, especially under the
Inca who encouraged breeding as state policy. In terms of settlement pattern
this adaptation suggests that most settlements would be located near the
junction of herding and agricultural lands rather than in the herding lands
proper. Sites found in the herding lands were unlikely to have been
pastoralist, unless they dated from Inca times when the large expansion in
herd numbers occurred. Models to help understand high-altitude land use in
northern Peru should not be derived from southern or central Peru where the
possibility of pure pastoralism was much stronger, but rather from the
eastern slopes where agropastoralism is the dominant subsistence adaptation.

Acknowledgements
I wish to thank the Huamachuco Archaeological Project directed by Drs John and Theresa Topic for
having me do the survey. It was funded by the Social Sciences and Humanities Research Council of
Canada. Permission for the survey was given by the Instituto National de Cultura of Peru. Luis Franco
Quezo acted as my field assistant. Both Dr Theresa Topic and Roxane Shaughnessy have read and
commented on this chapter and I thank them for their assistance.

References
Custred, G. 1977. s in a high altitude Andean environment. In Andean kinship and marriage, R.
Bolton & E. Mayer (eds), 117–35. Washington: Special Publication of the American
Anthropological Association 7.
DeHetre, D. A. 1979. Prehistoric settlement and fortification patterns of La Libertad, Peru: an aerial
photographic analysis. Unpublished Master’s thesis, Department of Anthropology, Trent University,
Peterborough, Canada.
Flores Ochoa, J. A. 1975. Sociedad y cultura en la puna alta de los Andes. America Indigena 35(2),
297–319.
Flores Ochoa, J. A. 1982. Causas que originaron la actual distribucion espacial de las Alpacas y
Llamas. In El hombre y su ambiente en los Andes centrales, L. Millones & H. Timoeda (eds), 63–
92. Senri Ethnological Studies 10. Osaka: National Museum of Ethnology.
Franklin, W. L. 1982. Biology, ecology, and relationship to man of the South American camelids. In
Mammalian biology in South America, M. A. Mares & H. H. Genoways (eds), 457–89. University
of Pittsburgh Pymatuning Laboratory of Ecology Special Publication Series VI 6.
Gade, D. W. 1969. The Llama, alpaca and vicuna: fact vs fiction. Journal of Geography 68, 339–43.
Gilmore, R. M. 1948. Fauna and ethnozoology of South America. In The handbook of South American
Indians, Vol. 6, J. H. Steward (ed.), 345–64. Bureau of American Ethnology, Bulletin 143.
Washington: Smithsonian Institution.
Guillet, D. A. 1983. Toward a cultural ecology of mountains: the central Andes and the Himalayas
compared. Current Anthropology 24 (5), 561–74.
Isbell, W. H. 1978. Environmental perturbations and the origin of the Andean state. In Social
archaeology: beyond subsistence and dating, C. L. Redman, M. J. Berman, E. V. Curtin, W. T.
Langhorne Jr., N. M. Versaggi & J. C. Wanser (eds), 303–13. New York: Academic Press.
McGreevy, T. H. 1984. The role of pastoralism in prehispanic and modern Huamachuco. Unpublished
Master’s thesis, Department of Anthropology, Trent University, Peterborough, Canada.
McGreevy, T. H. & R. E. Shaughnessy 1983. High altitude land use in the Huamachuco area. In
Investigations of the Andean past, D. H. Sandweiss (ed.), 226–42. Ithaca: Cornell University Latin
American Studies Program.
Miller, G. R. 1979. An introduction to the ethnoarchaeology of the Andean Camelids. Unpublished
PhD dissertation, University of California (Berkeley). Ann Arbor: University Microfilms.
Murra, J. V. 1956. The economic organization of the Inca state. Chicago: University of Chicago.
Murra, J. V. 1962. Cloth and its functions in the Inca state. American Anthropologist 64(4), 710–28.
Murra, J. V. 1965. Herds and herders in the Inca state. In Man, culture and animals, A. Leeds & A. P.
Vayda (eds), 185–215. Washington: AAAS Publication 78.
Murra, J. V. 1975. El control vertical de un maximo de pisos ecologicos en la economia de las
sociedades Andinas. In Formaciones economicas y politicas del mundo Andino, J. V. Murra (ed.),
59–115. Lima: Instituto de Estudios Peruanos.
Orlove, B. S. 1977. Alpacas, sheep and men: the wool export economy and regional society in
southern Peru. New York: Academic Press.
Shaughnessy, R. E. 1984. High altitude land use, past and present, in the Huamachuco area, north
highlands, Peru. Unpublished Master’s thesis, Department of Anthropology, Trent University,
Peterborough, Canada.
Stouse, P. A. D. 1970. The distribution of llamas in Bolivia. Proceedings of the Association of
American Geographers 2, 136–40.
Topic, T. L., T. H. McGreevy & J. R. Topic 1987. A comment on the breeding and herding of llamas
and alpacas on the north coast of Peru. American Antiquity 52 (4), 832–5.
Tschudi, J. J. von 1969. (1885) La Llama (translated by G. Custred). In Mesa redonda de ciencas
prehistoricas y antropologicas, Vol. I, 123–38. Lima: Pontifica Universidad Catolica del Peru.
Webster, S. S. 1973. Native pastoralism in the south Andes. Ethnology 12(2), 115–33.
Yamamoto, N. 1981. Investigacion preliminar sobre las actividades agropastoriles en la Distrito de
Marcapata, Departmento de Cuzco, Peru. In Estudios etnograficos del Peru meridonal, S. Masuda
(ed.), 85–137. Tokyo: University of Tokyo.
22 Andean pastoralism and Inca ideology
GORDON BROTHERSTON

In the long history of man’s relationship with other animal species a critical
difference occurred with his domestication of certain of them, in order to
make of them pastoral creatures. This is the truer the closer we hold to a
definition of the pastoral process which respects Ducos’s emphasis:
‘domestication can be said to exist when living animals are integrated as
objects into the socio-economic organisation of the human group, in the
sense that, while living, those animals are objects for ownership,
inheritance, exchange, trade, etc., as are the other objects (or persons) with
which human groups have something to do’ (see Bökönyi, ch. 2, Ducos, ch.
3, this volume). Historically, most evidence on the matter has been gathered
in the Old World rather than the New, and has given rise to major debates
on the role of pastoralism within the ‘Neolithic Revolution’, in particular its
relation to hunting economy on the one hand and to agriculture on the other.
In this Old World tradition one thing, however, emerges with clarity. And
that is how thoroughly pastoralism has been inscribed in the twin
ideological supports of western culture: the Graeco-Roman classics, and the
Bible. Through the former, the pastoral has contributed first premises to
political science and whole modes to art and literature, while the latter
strenuously privileges flock-keeping morality through such key examples as
Cain, the demeaned agriculturalist, and Abel, the preferred shepherd;
Abraham and the ram that substitutes for Isaac; the commandment that lists
wife after ox and ass; the parable of the sheep and goats, and of the lost
sheep; and the role of Christ himself as both pascal lamb and almighty
shepherd-lord. So pervasive, indeed, has this ideology been within the
western tradition that it has proved most hard to isolate and define. A signal
case is that of Jean-Jacques Rousseau, whose highly influential essay ‘On
the Origins of Inequality’ provides the hinge between the classical and
Biblical traditions and the new era of social enquiry typified by Marx.
Asking why it should be that man, born free, is everywhere in chains,
Rousseau repeatedly compares the exploitation undergone by human
masses with that of animal herds and draws an explicit parallel between the
fall of ‘natural’ man and the domestication of such animals as horses. Yet
such was the power of the pro-pastoral prejudice that he found himself
holding other culprits responsible for this degeneration: metallurgy (in a
reworking of the world-age story found in Hesiod), and agriculture, Cain’s
calling.
Something of the prejudice that afflicted Rousseau can perhaps be felt
still in approaches to pastoralism in western scholarship even today; in any
case, the links between the economy and the ideology of pastoralism have
hardly been over-explored, not least when the New as well as the Old World
is taken into account. Indeed, the New World offers in this respect a
welcome term of comparison since its ideological patterns, being less
familiar to western eyes, are likely to be more readily detectable. Moreover,
the single focus of pastoralism within the New World, the Andes, and the
former Inca empire Tahuantinsuyu (‘Four Districts’, Fig. 22.1), holds a
special and distinctive place in that continent, making yet more purposeful
comparisons possible; for while Tahuantinsuyu exhibits the usual native
American traits in such practices as weaving, curing, agriculture,
metallurgy, and so on, in other respects, not least its social and political
organization, it emerges as no less unique than it does on account of its
animal herding.
 Figure 22.1 Map of the ancient empire of Tahuantinsuyu

Fundamental to the present enquiry into pastoralism and the New World
is the existence of evidence of not just an archaeological or anthropological
order but of primary texts, classics of the native American tradition. From
Tahuantinsuyu itself come two, both dating from c. AD 1600, which are of
especial value. One is Guaman Poma’s Nueva coronica y buen gobierno, an
encyclopaedic survey of Peru sent as a letter to Philip III of Spain (edited
by Murra & Adorno 1980, Pease 1980). The first part of this work or Nueva
coronica, with its tabular arrangement of chapters and illustration-sets,
bears out its author’s claim (p. 367)1 that it was in part transcribed from
quipu sources: this knotted-string device, whose first use appears to have
been pastoral, served as the means of official communication and memory
in Tahuantinsuyu (see Ascher & Ascher 1981). The other work, composed
entirely in Quechua, the language of Tahuantinsuyu and of millions of
Andeans today, stems from Huarochiri, on the road between the capital
Cuzco and the Pacific coast (Trimborn 1939–41, Arguedas 1975). It affords
many insights into the role of animals in Andean cosmogony and religion
while offering an internal critique of metropolitan Inca practice (Spalding
1984). As editions and translations have become available over recent
decades, the Nueva coronica and the Huarochiri Narrative (see Trimborn
1939–41, Arguedas 1975) have been increasingly drawn on by scholars,
notably Murra (1980), who devotes a whole chapter to ‘Herds and herders’.
Yet there is much that these sources still have to offer on this subject. This
is especially true when they are related to their quipu precedents (which
even Murra derogates with the telltale epithet ‘pre-literate’); when they are
put together with other examples of Tahuantinsuyu literature, like its liturgy,
poetry, and drama; and when this corpus is placed in turn in its larger native
American context.

Tahuantinsuyu and its antecedents

On the high island of the Andes, four types of camelid have their natural
home: the llama and alpaca, and the guanaco and vicuna (or the short-
necked llama, see Shimadu & Shimadu 1985). Having in common with the
Old World camels a remote and long-vanished ancestor in the North
American Ohio Valley, these animals differ somewhat in size, the llama
being the largest; in quality and colour of wool, the vicuna having the finest
and the alpaca the most; and in range, the guanaco extending down on to
the Argentine pampa. Most important, only the first two have been
successfully domesticated, the llama possibly deriving from the guanaco.
Using bone samples, Browman (ch. 23, this volume) has put the
beginnings of domestication back to about 4000 BC and has identified
centres at Junin northwest of Cuzco, and at Lake Titicaca to the southeast,
suggesting a third possible site at Ayacucho in between. He further detected
an important difference in emphasis between Junin and northerm
Tahuantinsuyu generally, where camelids were used mainly for meat, and
Titicaca, where they served mainly as suppliers of wool and for transport,
uses which appear to have spread north as late as AD 500, that is at dates
comparable with the spread of other more generally recognized traits of
Tiahuanaco culture, which likewise had its base on the shores of Lake
Titicaca.
At any event, long before the rise of the Inca, Tiahuanaco and the cold,
high plateau around Lake Titicaca, where agriculture is restricted, was
reputed to be the richest in llama herds, and caravans plied regularly from
there to the Pacific coast. Moreover, this area, known as the Collao or Colla
country, is consistently referred to as the one from which the Inca drew
support when they came to establish their own centre of power at Cuzco, a
century or two before the European invasion. One account, stemming from
the Aymara-speakers of the Collao, says simply that the Inca made off with
Colla animals in order to build up herds of their own (Matienzo 1567,
quoted by Murra 1980, p. 52). From the Inca side it was more a question of
deference to the llama typical of the Collao during royal initiation and other
ceremonies. That the Inca, from the time of the first emperor Manco Capac,
acknowledged their origin to have been in Titicaca, is made quite clear by
Poma de Ayala (1936, pp. 84, 265); they favoured the white llama that the
Collas themselves had gone so far as to revere as a tribal ancestor, and they
would deck out such llamas, known as napas, with red shirts and necklaces
and gold earrings. In the chapter of his Nueva corönica which he devotes to
the festivals typical respectively of the Inca and of the four suyus of the
empire, Guaman Poma also shows how pastoral songs of the Colla and
Collasuyu were held in particular esteem at the royal court of Cuzco, to the
extent of being sung in Aymara, the Colla language (1936, pp. 129, 319).
The Inca connection with Titicaca is also supported by certain architectural
parallels, while in his Royal Commentaries (I, xviii) Garcilaso el Inca notes
that the political arrangement of a Tahuantinsuyu, or ‘four districts’, of
empire had been previously elaborated at Tiahuanaco. A striking indication
of how powerful the Colla with their massive llama herds had become as
the Inca’s predecessors lies in the fact that of all four suyus theirs was the
one which retained most rights to local herd ownership under Inca rule;
after the European conquest and the collapse of Cuzco, the Colla even
recovered some of the llama wealth that had been alienated by the Inca.
 Turning now to the pastoralism of the Inca and following Poma de
Ayala’s chapter on their rise (1936, pp. 79–193), we can trace a trajectory
that runs from small beginnings in Cuzco to the latter-day Tahuantinsuyu,
which spread thousands of miles between the present-day states of
Colombia in the north and Chile in the south; and in so doing we can detect
a turning-point in the 15th-century emperor Pachacuti, conqueror and
acquirer of territory especially at the expense of the Colla, and Tupac
Yupanqui, the great consolidator under whom herd units ran into many
millions. Yet for present purposes, Inca practice, which in certain details no
doubt echoed Colla and other precedents, can more conveniently be
considered as one, especially insofar as it constitutes a system which went
quite beyond any previous system in the Andean area in organization and
scale.
In Tahuantinsuyu the domesticated Andean camelid, from now on
referred to by the general term ‘llama’ (the alpaca not being separately
noted except where the difference matters), can be seen to have combined
an impressive array of functions. For the Inca and their subjects ate its flesh,
fresh or as charqui; wore its skin as sandals, or cut it into thongs to secure
the foot-plough (taclla) or into bottles to carry water across the
desert;turned its fat into tallow; spread its dung as manure or gathered it for
fuel; made its tendons into slings for the scarecrow, the herder, and the
soldier; shaped its bones into weaving instruments; and span and wove its
wool into cloth both coarse (auasca) and fine (cunbe), and into the threads
and main cord of the quipu. As well as singly providing these organic and
commodity resources, and in the absence of any analogue except the human
being itself, the llama was also widely exploited by the Inca as transport,
along the roads for which Tahuantinsuyu is famous, and in particular to
support the crop-carrying farmer (Fig. 22.2) and the campaigning soldier.
For all these and yet other purposes, the Inca instituted a programme of
breeding, distinguishing and counting types and ages of beast down to the
minutest detail by means of the quipu, and conducting a thorough census in
the month Aya Marcay (November; Poma de Ayala 1936, p. 256). Also
through running and other athletic trials (which conjoin the two meanings
of ‘race’) they prized the strongest and fittest beasts. The Huarochiri
Narrative further tells us (Ch. 10) that the prowess revealed by such trials
was associated with the enhanced penis displayed by golden and silver
llama statuettes that have survived from Inca times. By means of this
controlled reproduction, the llama came to acquire another order of value,
one of permanency more like that of the precious metals with which it was
equated and into which it was cast ceremonially. Beyond this again, through
multiplication and increase, the llama became literally capital, perhaps its
most distinctive role of all in the plan of Tahuantinsuyu, as we shall see, and
one presided over by Yacana, the celestial llama at the centre of the sky
(Huarochiri Narrative, Ch. 29, Zuidema 1983).
A prime Incaic use of the llama was the military one. By mobilizing
troops and ranks of llamas, for both transport and food on the hoof, little
affected by season and harvest, the Inca disposed of a state army
unparalleled in America, whose campaigns related less to ritual than to
policies of permanent territorial gain, and which in its day proved largely
irresistible. In his chapter on the world-ages that preceded the Inca (1936,
pp. 48–78), Poma de Ayala makes much of the military potential that went
with the llama breeding characteristic of the latter two ages of the Purun
and the Auca (whose name means ‘war-like’). Then, after military conquest
and as part of Inca policies of pacification and colonization, the llama had
no less critical a role. In the case of people already in possession of herds,
these and all their reproductive goods, in Murra’s words: ‘became the
property of the Inca crown, which then reissued some of it back to the
inhabitants and set public boundaries’; again, after conquest ‘all llamas
were defined at least in theory as state property’ (1980, pp. 94, 96). In
practice, no other interest group was permitted to keep herds which in any
way might rival those of the Inca, the former being linguistically
distinguished from the latter by the respective terms huaccha (poor) and
capac (mighty).
Figure 22.2 Llamas carrying potatoes (Poma de Ayala 1936, p. 1149)
 This alienation of ethnic property facilitated, in turn, the distinctively
Inca policy of granting llamas as capital to settlers in conquered territory
where previously there had been few or no llamas; known as mitima, these
grantees were further encouraged to migrate by being made exempt from
labour tribute and other state obligations (Murra 1980, p. 178). Through the
device of the llama grant, the Inca secured their hold on the coastal valleys,
filling in and completing Condesuyu as an imperial quarter; and between
the larger and repeatedly extended Colla and Chinchasuyu they removed
populations over thousands of miles – no less than 4000 mitima families
were seen journeying up to occupy former Canari lands in Ecuador and
Chinchasuyu when Pizarro was already at Caxamarca. Herding was
specified as the first of the skills required of a mitima, and the head-
shepherds among them and the longest standing came from the Collao. In
return for their official generosity, the Inca expected to gain from the
multiplication of their grantees’ animals, taking, as it were, interest from
capital. They also expected the supply of cloth woven from llama and
alpaca wool which was universally recognized as ‘one of the main bonds
and symbols of citizenship’ (Murra 1980, pp. 52, 55, 156, 174). They might
even be seen to have created a market-demand for the mitima and llama-
producers generally in the strict laws that they instituted, which, according
to Poma de Ayala (1936, pp. 272–3), obliged every community however
small regularly to sacrifice these animals in order to consume their blood
and meat.
Above all, through the mitima (those who ‘leave’ their first homes) and
their interest, in both senses, in their own displacement, the Inca achieved
stability for Tahuantinsuyu as a continuous territory, within its frontier, or
outer fence of pasture. In this respect it is highly significant that of the four
suyu the one which most resisted conquest, the Antisuyu of the montana
and valleys of the upper Amazon, was also the one known to be least
adaptable to llama herding.
In all these respects, at the time of the European invasion nothing like
Tahuantinsuyu existed elsewhere in America, this difference being directly
attributable to the resource inherited by the Inca as Andeans and heirs to
Tiahuanaco, and exploited by them as architects of Tahuantinsuyu: the
domesticated camelid. In the tribute lists of Mesoamerica we find an official
catalogue of commodities supplied in Peru by the llama; clothing as cotton
capes, food as bushels of maize and beans, durables as precious metals
(Brotherston 1979, pp. 233–6): besides having all these assets in one, the
llama, as we have seen, supplied transport and, above all, increased and
multiplied its value through breeding. This key economic difference
doubtless relates, in turn, to the different ways in which the four-quarter
model of tribute was developed in Mesoamerica and Tahuantinsuyu. In the
former, the four quarters mapped on the title-page of the Aztec Mendoza
Codex yielded only commodity tribute, which was collected via chains of
head towns and along routes that often ran through neutral or hostile
territory within the quarter. In the latter, despite an even less tractable
terrain, the four quarters, tahuantin-suyu, were consolidated entire, like
pasture within its fence.

Models of control and authority

So fundamental was herding to the Inca enterprise that in Tahuantinsuyu the

social relations it implies as an activity were transferred ideologically to
become a model for the state itself, indeed it arguably provided the enabling
concept of ‘state’ in the first place. This is so for the two dominant social
models of the Inca state, that which relates ruler to ruled, and that which
relates ruler to authority. For both, a wealth of evidence is provided by the
native sources used so far; in addition we have the 11 hymns or prayers,
recorded in the 16th century in Quechua by Molina, which make up the
liturgy of the Situa, a ceremony of cleansing and purgation held in the
month Coya raimi, which was September (Rowe 1953, Lara 1969, pp. 179–
86).
First, the Situa hymns repeatedly equate flock with folk, both as subjects
of the Inca ‘who founded Cuzco’. Requests are made, in the same terms and
the same phrase, that under the Inca both people and animals (runa llama)
should enjoy peace and safety; should increase and multiply; and should not
fall into enemy hands or stray into sin.

Oh dew of the world Let me live in peace and in safety,

Viracocha Father Viracocha,
inner dew with food and sustenance,
Viracocha with maize and llamas,
you who dispose by saying with all manner
‘Let there be greater and lesser of skills.
gods’
great Lord Abandon me not,
dispose that here Remove me
men do multiply from my enemies
fortunately. from danger
from all threat
of being cursed, ungrateful
or repudiated.

The parallels here with Semitic liturgy and the logic of, say, ‘The Lord is
my Shepherd’ (Psalm 23) are so strong that an influence via the Spanish
Christian mission might be suspected were it not for numerous independent
testimonies to the nature of the Situa. What is more, while obviously a piece
of spiritual rhetoric, the flock-folk equation proves to have a firm material
and economic basis in Tahuantinsuyu. For the sheer tally of the two orders
of unit in question, animal and human, was consigned to the recording
device used initially for the former: the quipu. This prime piece of the
herder’s equipment (Poma de Ayala 1936, p. 351) registered animal units
with decimal-place value notation; displayed colours as semantic variables
which corresponded in the first instance to those of the actual llama wool;
and had a structure of cord and dependent thread that even replicated the
main-cord custom of llama tethering.
The competence of the quipu as a human, as opposed to just an animal,
tally in Tahuantinsuyu can be judged not least from the fact, recorded in the
Nueva corönica, that the llama census of the month Aya marcay was also a
human census; Guaman Poma even makes an implicit comparison between
the selection of males and females from both species for particular
purposes, like male troops for warfare, and chosen virgins for wool
production (1936, p. 257). In another chapter (pp. 193–234), the Nueva
corönica details the categories of age and usefulness, ten for male and ten
for female, according to which the human census or ‘visit’ was conducted.
Extended to the minutest element of value in the state, reliably and
retrospectively over the years, the quipu accounting system had, as a major
feature, the noting of absence and non-performance. The Quechua term for
this failure in conduct, hucha, was noted on the quipu as greater (hatun) or
lesser (huchuy) (Poma de Ayala 1936, p. 361). For its part, âpropos
ceremonies that accompanied irrigation work, the Huarochiri Narrative
(Ch. 31) notes darkly how absences of goods and of personnel were
recorded on official quipus. By these means, and through a matching
apparatus of police whose initially pastoral function is patent from the
official title (llama michic; michic), the state was able precisely to gauge
quantities not just of commodity but of labour tribute rendered.
Outraged by Spanish lawlessness and lamenting the demise of
Tahuantinsuyu, Guaman Poma in his day significantly appealed to the
notion of the michic as a last means of preventing disaster and of restoring
order to society. At the same time he underpinned the traditional flock-folk
equation of Tahuantinsuyu most succinctly when complaining to Philip III
that in vice-regal Peru the Indians bore the burden of tax payments like
domesticated animals, while mestizos and other mixed-bloods exempt from
tax were allowed to remain wild like the vicuna (1936, pp. 215, 890, 1153).
From the non-performance and the non-compliance monitored by the
quipu, it was but a step to the rhetoric of disobedience, crime, and sin, and
corresponding retribution in the name of the state. Another of Poma de
Ayala’s chapters is devoted precisely to orders and types of official
punishment (Nueva corönica, pp. 301–14); one such was reserved for those
who simply moved without permission from their allotted place in the realm
(Murra 1980, p. 110). In other words, like their flocks, the subjects of
Tahuantinsuyu could be considered contained and penned, pastured
elements of the great Pax incaica, safe as such from the threat of enemies
and the barbaric wild beyond its rim.
When it comes to relations not so much between ruler and ruled as
between ruler and authority, the Situa hymns further highlight the pastoral
model. For here, in what emerges as a truly monotheistic impulse,
monarchy is endorsed by the supreme spiritual principle known as
Viracocha, and as the ‘creator’ (camac) of earth and men, and so on.
 Invoked in most of the hymns, this figure is asked to guard the Inca just
as the Inca guards his flocks. Throughout Tahuantinsuyu the rites of this
supreme shepherd can be shown to have been imposed over local deities
and ‘ huacas’ a process examined by Poma de Ayala in another chapter of
his Nueva coronica (pp. 261–73). Irreverence beneath the Inca imposition is
shown up in the Huarochiri Narrative which basically remains loyal to the
lightning god, Pariacaca, and other shamanist huacas of the region; this
source also tells how the priests imported and appointed by the Inca left
their posts on hearing about Pizarro’s advance (Ch. 18). This last detail is
significant also because it indicates how far religion, or ‘the Church’ as it is
often called, had become subject to and regulated by the state, exposing the
universally imposed Viracocha or divine shepherd to have been, in practice,
a back-projection of and from the secular power of the Inca themselves.
Similarly, though formally distinct and guarded by special herders, who
possibly included the aclla, or chosen virgins of the sun (another
exclusively Inca institution), the Church herds relied on the state for
allocations of pasture, just as ritual llama sacrifice imposed by the Inca
served mitima and state-herding interests.
On the same subject of Inca-appointed priests, the Huarochiri Narrative
specifies, in passing, their period of service as 15 days, or half the official
Inca month. This indicates, in turn, how the appropriation of divine
authority coincided in practice with the institution of a state year-calendar,
one which could encompass in a single standardized whole the various
rhythms of agriculture, curing, and the myriad other rituals of society, as
well as the demands of material tribute. According to Poma de Ayala’s
chapter on the subject (1936, pp. 236–60), between the solstitial and
equinoctial celebrations in honour of the divinely sanctioned Inca and his
queen, this calendar deferred thematically to the tasks of the pastoral year
with its llama census in Aya marcay (November) and intervening one in
Aimoray (May), and with its regular sacrifices of llamas and alpacas
throughout, like that actually depicted for Pacha pucuy (March; see Fig.
22.3).
From these further sets of evidence, Tahuantinsuyu appears to have been
as distinctive ideologically within native America as it was economically.
The sort of entreaty made in the Situa hymns, to a single god that can
guarantee monarchic power and guide its course like a shepherd, goes
beyond anything that can be found in comparable native American religious
texts, which likewise are entirely devoid of equations in principle between
human and animal herds that are faithful to their keeper. In Mesoamerican
sources, such as the Twenty Sacred Hymns of the Aztecs and the Aztec
address to the Franciscan missionaries of 1524 (Brotherston 1979, pp. 63–
9), the divine, the ruler, and the ruled interrelate quite differently, a typical
model being the tripartite one which places the archetypally opposed social
groups of planters and hunter-warriors under the aegis of the aristocrat-
priests; and that these last were no mere appointees of the emperor is clear
from a whole range of evidence, like the generic difference between priestly
and secular texts, the plural calendrics of the divine tonalpohualli and the
tribute-year, and the sheer layout of Tenochtitlan’s pyramids.
Figure 22.3 Llama sacrifice in the month Pacha pucuy (Poma de Ayala 1936, p. 240)
 In Tahuantinsuyu, how far flock-economy shaped ideology is well caught
in these lines from another prayer to Viracocha, in which learning
obedience to State and Church is directly equated with animal
domestication (Lara 1969, p. 192):

the vicuna of the wilds

the viscacha of the rocks
become domesticated
in his presence
so too my heart
with each dawn
renders you its praise
my father and creator

(purun wikuna / qaqa wiskacha / uywaman tukun / paypaj qayllapi /

sunqoypas kikin / sapa paqarin / anayniskuni / yayay kamaqey)

The larger native American context

Exploring Tahuantinsuyu texts in their New World context helps us to

recognize further principles of difference which impinge directly on our
assessment of the nature of the Inca state and, by extension, of pastoralism
generally. For Andean and Inca herding evidently modified whole cultural
paradigms that were otherwise common to America, by virtue of their
common source in shamanism, and in agricultural practice whose antiquity
is universally testified to, not least by the Popul vuh in Mesoamerica and
Poma de Ayala in Tahuantinsuyu (1936, p. 48); and just as evidently it
produced whole new modes and genres of expression otherwise absent from
the continent.
In formulating human-animal relations as such, a wealth of evidence
suggests that Andean society continued to share certain beliefs and codes of
behaviour with other parts of the New World, despite its own advance into
pastoralism. In the cosmogonical context of the world-ages or suns found in
much of native America, for example, we may detect in Tahuantinsuyu
texts traces of the animal-other that is typically invoked in the hunter’s
propitiation, and which is best known in the Mesoamerican philosophy of
the ’nahuaV. In the case of the Flood, the catastrophe that ends the first
world-age, the Huarochiri Narrative (Ch. 3) tells us that humans were
warned of its onset by none other than a llama, one possessed of just the
sixth sense, and that psychic intimacy, if not solidarity, that characterizes
the nahual. In the subsequent catastrophe of the eclipse, Tahuantinsuyu’s
consistency with native American belief is yet more striking, since here we
are involved specifically with domestication, and the domestic contract that
binds humans to creatures who have left the wild to fall under their control.
In fact, the eclipse in question is said to have been occasioned by heartless
human exploitation of domestic creatures: regaining their wildness these
turn on their masters and help to destroy them. In Mesoamerican texts, such
as the Popol vuh, the upheaval is led by the mortars and kitchen utensils
tired of unfeeling use, and not least by the dog and turkey, the two creatures
whose long-standing alliance with American man is elsewhere celebrated in
rituals that survive even today, for example in the southwest of the United
States (Brotherston 1979, pp. 111–13). That in Tahuantinsuyu this same
contract was perceived to have extended to llamas is clear from the
Huarochiri Narrative (Ch. 4), where during the eclipse the revolt of the
utensils is accompanied by one on the part of these creatures, who turn on
their masters in savage herds; the same moral is then reinforced in the
parable of the few wild vicuna who help the poor Huatyacuri to outdo the
teams of llamas enlisted by his rich and insensitive antagonist (Chs 4 & 5).
Consistent with the terms of this domestic contract is the inclusion of the
tamed creature as a ritual ally and even social companion. The actual
process of inclusion is well illustrated in Poma de Ayala’s report of the
ceremonies proper to the month Uma Raimi (October) (p. 254), in which
supplicants for rain included not just human children and dogs, whose tears
and howls were elicited as sympathetic magic (as in Mexico), but a black
llama as well.
Taken together, all this suggests that at one level of its culture, the Andes
shared a native American recognition of the moral problem raised by the
use and exploitation of animals, and of the need to regulate and ritualize it.
And, in practice, the Inca would refrain from eating llama meat at certain
periods, out of that contractual respect which in Mexico extended even to
maize, doctrinally the source of human flesh. Yet, as we have seen, as
objects of Tahuantinsuyu’s pastoral economy, llamas were none the less
bound to be treated in ways that could not be contained within these
traditional contractual limits. For they were obliged to function as mere
units of value, items of exchange devoid of particular status or rights, and
were transacted on the grand decimal scale of the quipu as the indispensable
capital-plus-interest of the state. This is certainly the role they are assigned
in Guaman Poma’s account of the world-ages, which lies closer than does
the shamanist Huarochiri Narrative to official Inca thinking, and
emphasizes the material and statistical value of llamas and their wool as the
factor that distinguished the age of the Purun and the beginnings of state
power.
One response to this discrepancy or moral dilemma appears to have been
the selection of individuals on which to bestow, in a sacred-cow logic of
compensation, as it were, the regard that could not be practicably accorded
to the species generally. In this respect, the single black llama who sang at
Umu raimi, while hundreds of his lesser brethren were slaughtered and
consumed by state order, resembles the royally chosen and lavishly attired
napa of the Colla ceremonies; or the red llama which the emperor took as a
singing companion and musical guide in court performances of the
Quechua yaravi (see Frontispiece). Indeed, only interms of such privileging
could individual llamas have come to serve humans as their representatives
in their deeper spiritual needs, which is what they undoubtedly did as
bearers of guilt, a paradoxical and terminal privilege. In the Situa ceremony
human sins and failings were transferred on to eviscerated llamas which
were washed out of the capital via its river. Once again, no other American
religion offers anything like this parallel to the well-known scapegoat and
sacrificial-lamb practices developed in the Old World by the Semitic
pastoralists.
Finally, besides modifying cosmogonical and other paradigms found
widely in native America, Inca herding led to a literary pastoralism that was
otherwise quite absent from the continent. While the courts at Tenochtitlan
and Cuzco alike cultivated poetry in modes that derived from social
practice, planting songs from the farmer, battle songs from the warrior,
funeral songs from the mourner, and so on, only in the Inca case were the
pastoral songs of the herder also heard, for reasons that by now will be
obvious (Brotherston 1979, pp. 260–5). Themes characteristic of this
pastoral mode included the nameless yearning inspired by the landscape of
the high puna, lonely and sublime (unruly sons of royal families were
customarily exiled to pastoral life on the puna); love between male and
female herders, in which the desired one is compared with the elusive
untamed vicuna, or is revealed as the inaccessible princess of the aclla who
guarded the Church’s flocks; and the herders’ flutes whose plaintive sound
could presage the suicide of one or both victims of an impossible love.
So strong are these Andean pastoral conventions that they survive even
today in Quechua poetry, despite the influence of their more sexist
European analogues (Yaranga Valderrama 1986). They likewise inform a
whole series of legends (‘Hirtenmärchen’ in Keim 1968) and even dramas,
which tend, however, to allude more directly to the social conditions and
constraints under which herders actually worked. A pastoral love-story
recounted by Murua in this way reveals the pastures beyond the edge of
town and agriculture as the place of transformation and fantasy: the
metamorphosis of shamanism is here made to serve the particular needs of a
lonely or alienated individual, who changes into a staff, a snake, a bear, in
the cause of sexual fulfilment (Basadre 1938, pp. 31–8, Arguedas 1949,
Keim 1968). The drama Apu Ollantay, where the rebellious Antisuyu hero
strays from the right path like a disobedient llama lamb, foregrounds Inca
policies of confining high-born daughters to the aclla (Brotherston 1986).
Like the encyclopaedic Nueva corönica and Huarochiri Narrative, and
like the Situa hymns, these poems, legends, and dramas offer their own
perspective on the complex workings of Andean pastoralism, and on those
grounds deserve more attention than they have received. Indeed, as more
self-consciously literary products, they offer special insights into Inca
pastoral ideology as it involved such concepts as obedience, alienation, and
love. Yet for now, the point must simply be to indicate their consistency in
turn with the economics of Tahuantinsuyu, and their consequent uniqueness
in the New World.

Conclusions

Privileging native texts as evidence and respecting their New World

context, this enquiry started by suggesting Tahuantinsuyu as a valuable term
of reference by which to assess the ideology, no less than the economics, of
pastoralism. In the first place, there can be little doubt about the huge
significance of pastoralism for and within Tahuantinsuyu: the practices
associated with it, particularly the mitima, radically impinged on the lives
and movements of beast and human alike, to an extent which it is,
incidentally, hard to reconstruct, in the absence of the Inca system itself,
from llama herding as it has survived in the Andes today. From this base,
pastoralism was shown to have permeated principal aspects of Inca
ideology. It patently shaped models of political control and authority, along
with their extension into the divine; it modified traditional codes of
interaction between humans and animals, such as the domestic contract of
American world-age cosmogony; and it inspired a gamut of literary
pastoralism. The internal consistency of this evidence, its distinctiveness
within native America, and its wider socio-political resonance, surely
endorse it as consequential for any comprehensive analysis of pastoralism.

Acknowledgements
In preparing this chapter I have been helped by Olivia Harris, who read through an early draft, and by
the discussion at the World Archaeological Congress 1986.

Note
1 The page-numbers given for Poma de Ayala’s Nueva Coronica are his original ones (Poma de
Ayala = Guaman Poma).

References
Arguedas, J. M. 1949.Canciones y cuentos quechuas. Lima: Huascaran.
Arguedas, J. M. 1975.Dioses y hombres deHuarochiri. Mexico: Siglo XXI.
Ascher, M. & R. Ascher 1981.Code of the Quipu. A study in media, mathematics, andculture. Ann
Arbor: University of Michigan.
Basadre, J. 1938.Literatura inca. Paris: Desclée de Brouwer.
Bökönyi, S. 1989. Definitions of animal domestication. In The walking larder, J.Clutton-Brock (ed.),
ch. 2. London: Unwin Hyman.
Brotherston, G. 1979.Image of the New World. The American Continent portrayed innative texts.
London Sc New York: Thames & Hudson.
Brotherston, G. 1986. The royal drama Apu Ollantay. Comparative Criticism 8, 70–94.
Browman, D. L. 1989. Origins and development of an Andean pastoralism: anoverview of the past
6000 years. In The walking larder, J.Clutton-Brock (ed.), ch. 23. London: Unwin Hyman.
Ducos, P. 1989. Defining domestication: a clarification. In The walking larder, J.Clutton-Brock (ed.),
ch. 23. London: Unwin Hyman.
Keim, A. 1968.Vom Kondor und Vom Fuchs. Berlin: Gebr. Mann.
Lara, J. 1969. La literatura de los quechuas. La Paz: Editorial Juventud.
Murra, J. V. 1980. The economic organization of the Inka state. Greenwich: JAI Press.
Murra, J. V. & R.Adorno (eds) 1980. Guaman Poma, El primer nueva corönica y buen gobierno. 3
vols. Mexico: Siglo XXI.
Pease, F. (ed.) 1980.Guaman Poma, El primer nueva corönica y buen gobierno. Caracas: Ayacucho.
Poma de Ayala, F. G. 1936.El primer nueva corönica y buen gobierno. Facsimile edition. Paris:
Musée de l’Homme.
Rowe, J. H. 1953. Eleven Inca prayers from the Zithuwa Ritual. Kroeber Anthropological Society
Papers nos 8 & 9.
Shimadu, M. & I. Shimadu 1985. Prehistoric llama breeding and herding on the north coast of Peru.
American Antiquity 50, 3–26.
Spalding, K. 1984.Huarochiri. An Andean society under Inca and Spanish rule. Stanford: Stanford
University Press.
Trimborn, H. 1939–1941.Dämonen und Zauber in Inkareich. Berlin: Gebr. Mann.
Yaranga Valderrama, A. 1986. The Wayno in Andean civilization. In Voices of the first America,
G.Brotherston(ed.), 178–95. Santa Barbara: New Scholar.
Zuidema, R. T. 1983. Towards a general star calendar in ancient Peru. In Calendars in Mesoamerica
and Peru, A. F. Aveni & G. Brotherston (eds), 235–61. Oxford: BAR.
23 Origins and development of Andean
pastoralism: an overview of the past
6000 years
DAVID L. BROWMAN

Introduction

Substantial advances have been made in the 1970s and 1980s in deciphering
the complex history of the domestication of Andean camelids. Four closely
related taxa are involved: the wild vicuna (Vicugna vicugna) and guanaco
(Lama guanicoe), and the domestic llama (Lama glama) and alpaca (Lama
pacos).
Franklin (1978, 1983), Raedeke (1979), and Jefferson (1980) have done
much to refine the work by Koford (1957) and others detailing the ethology
of the wild species. As one or both of the wild camelids have been
suggested as ancestors of the domestic varieties, such knowledge is
indispensable to our theories of domestication. While some contemporary
researchers (such as Cardozo 1975, Craig 1985) argue for Pleistocene
ancestors for these domestic species (such that the guanaco and vicuna
would not be involved as progenitors), I follow the majority who believe
that the llama and alpaca are directly descended from the wild species.
Vicuna are sedentary, obligate drinkers of water, and altitudinally
restricted to between 3700 and 4900 m above sea-level, while guanaco can
be either sedentary or migratory, are both grazers and browsers, are periodic
drinkers of water, and range from sea-level to over 4000 m in altitude. In
addition, guanaco have a longer period of parental care of their young,
extending over two growing seasons, compared with one season for the
vicuna (Franklin 1983, pp. 573–4). The greater flexibility of the guanaco
allows it to occupy a wider array of zones, including the semiarid puna and
altiplano grasslands, where the archaeological sites, discussed here, have
been excavated.
 In the Old World, pastoralism is often seen as a secondary spin-off of
agriculture, as a specialization resulting from farmers being forced to adapt
to environments on the margins by population pressures (as described, for
example, by Lees & Bates 1974). In contrast, the pattern that is being
recovered from the Andes suggests that llama and alpaca pastoralism, and
what Wing (1973, 1977) calls the ‘pastoring’ of guinea-pigs, developed
much earlier than plant cultivation.
Studies of domestication in the Andes have four major foci:

(a) distributional studies (target animal occurs outside its previous normal
range, implying human intervention);
(b) utilization studies (a sudden shift or increase in the importance of the
animal, suggesting management of the stock by human curators);
(c) population studies (a shift in the mortality, suddenly more young or old
animals being cropped; and
(d) morphometric studies (occurrence of distinctive morphological features
that result from human selective pressures).

Distribution and intensification studies

Historical records as well as present-day distribution indicate the Andean

highlands to be the primary habitat for domestic camelids and thus research
focused there. Gade (1969, p. 341) is among several noting that the majority
of herding centres for alpaca today are in an area within a c. 150 km radius
of Lake Titicaca, and thus suggested the Titicaca basin as the centre of
alpaca domestication. To test this, and other hypotheses, we need
appropriate faunal analyses of prehistoric periods. It has been in this area
that major contributions have been made in the past two decades.
The basic presumption is that a high degree of dependence on one animal
is correlated with domestication and herding. A significant difference exists
in the methodologies employed to assess species importance. For example,
Wing uses MNI (minimum number of individuals) while Wheeler and I use
NISP (number of individual specimens present). NISP skews as follows:
(a) it ignores the difference in number of bones in the skeletons of different
species;
(b) it is affected by cultural practices such as fragmenting bones for marrow
extraction;
(c) it results in an over-emphasis of species butchered on site; and
(d) it reflects collection techniques (e.g. more bone fragments recovered
through finer mesh screens, etc.).

MNI skews as follows:

(a) it is very sensitive to aggregation procedures, i.e. how the collection unit
is defined and what its size is;
(b) it does not provide information on the relative abundance of fauna at the
site as a whole;
(c) there are intrinsic problems in how to match or pair elements; and
(d) it overestimates the contributions of rare species.

MNI is more appropriate if one wants to calculate estimates of biomass or

meat contribution, and if the diversity in the number of species represented
is high. However, it is my argument that in the case of domestication
studies, where species diversity is low, and dependence on a single species
is high, that MNI is less desirable than NISP as MNI tends to place too
much emphasis on rare species, and masks the relative abundance of the
target species (the potential domesticate).

Table 23.1 Chronology for camelid dependence at Telarmachay (sources: Lavallee et al. 1984,
Wheeler 1984)
Date Camelid percentage
7000–5200 BC 65
5200–4800 BC 78
4800–4000 BC 82
4000–3500 BC 87
3500–2500 BC 86
2500–1750 BC 89
 The percentages that follow are, where possible, calculated from the
numbers of individual specimens. In some cases, however, original data are
only available as MNI percentages.

Central highlands
Three sites near Lake Junin, Panaulauca, Pachamachay, and Uchcumachay,
and one site slightly further north in Cerro de Pasco, Lauricocha, provide
evidence that the pampas of Junin was one centre of domestication.
Faunal analyses at Panaulauca by Wheeler (Wheeler et al. 1976 [1977],
Pires-Ferreira et al. 1977) indicated a shift in species utilization there at
around 5500–5000 BC. Camelids contributed 26 per cent of the assemblages
dating to 7000–5500 BC, but increased to 86–87 per cent of the assemblages
dating between 5500 and 2500 BC. Subsequent work by Moore (1982,
1985a, 1985b) has confirmed this general pattern at Panaulauca. Initial
analyses of Pachamachay by Wheeler indicated 96–98 per cent of the
collections were camelid in the phases between 4200 and 1750 BC.
Subsequent analysis by Kent (1982a) indicated camelid percentages in more
recent levels of the site, ranging from 93 per cent at 2200 BC to 82 per cent
by 400 BC. Recent work at Telarmachay by Wheeler (1984) and Lavallee et
al. (1984) provided a finer-tuned chronology for increasing camelid
dependence (Table 23.1).
For Lauricocha, Wing (1972) placed domestication in the 6000–3000 BC
period. Cardich (1983) argued that the shift to domestication occurred
between Lauricocha II and Lauricocha III. Wheeler and colleagues were
able to re-examine some of the Lauricocha collections, and estimate a
proportion of some 59 per cent for Lauricocha I, before 6000 BC, and an
increase to about 85 per cent by Lauricocha III, around 3000 BC.
The pattern for the Junin area thus appears to be one of increased
dependency upon camelids, beginning at least as early as 7000 BC. The
recovered assemblages prior to 7000 BC are too small to provide good
statistics, but suggest little dependence upon camelids. By 4500–4000 BC,
up to 80–90 per cent of the faunal assemblage in some highland sites may
be made up of camelid species, and there appears to be little change in the
emphasis on quantity of camelids for the next 4 millenniums.
North highlands and north coast
As one moves further north, remains of domestic camelids occur later. At
two sites near Pampa de Lampas, just north of Junin, camelid percentages
are still high (listed at 64–80 per cent at c. 3000 BC, Wing 1978, p. 169).
But further north at Guitarrero Cave (the locus of the earliest claimed
occurrence of the domestic Phaseolus bean), the importance of camelids is
dramatically less. Initial estimates ranged from 10 to 17 per cent for all
levels, but later re-analyses indicated no more than 10 per cent for levels
earlier than 5000 BC, and no more than 33–35 per cent for the next 6
millenniums (Wing 1978, 1980).
Studies of the Chavin contemporary temple sites in Ancash and
Cajamarca have allowed some refining of our understanding of the later
spread of camelids. At Chavin de Huantar, there is a sharp shift in
dependence on camelids c. 500 BC. For the Urubarriu phase (850–450 BC),
camelids made up about 70 per cent of the faunal remains, but there was an
increase in the subsequent Chakinani and Janabarriu phases (450–200 BC)
to 96–97 per cent camelid (Miller 1979, 1981). At Huaricoto, there is a shift
in camelid utilization from 26–31 per cent for levels pre-700 BC up to 56–69
per cent in post-700 BC levels (Burger 1985). This shift is more dramatically
reflected in the site of Huacaloma near Cajamarca (Shimada 1982, 1985)
where the average contribution per phase can be assessed as in Table 23.2.
Northern highland sites indicate an increase in the number of camelids
about 700–500 BC, indicating a shift in human and camelid interaction. The
precise nature of the shift is debated: was it evidence of southern
pastoralists moving into the area, of local herding adaptation, or of new
trade simply bringing in meat animals?
The occurrence of domestic camelids becomes progressively later further
north. While the Chavin contemporary site of Pacopampa exhibited a shift
in camelid dependence after 800 BC (Rosas & Shady Solis 1974, p. 24),
unpublished work by A. Meyers and U. Oberem suggests that it is not until
c. AD 1000 that substantial numbers of camelids occur at Cochasqui,
Pinchincha, near Quito.

Table 23.2 Camelid content of faunal remains at Huacaloma (source: Shimada 1982, 1985)
Date Phase Camelid percentage
1200–900 BC Early Huacaloma 15
900–600 BC Late Huacaloma 10
600–500 BC EL 41
500–300 BC Layzon 88
post –600 BC various Cajamarca 95

North-coast sites reflect this late northward expansion of domestic

camelids. At Cholupe, Shimada (1981) first suggested introduction of
camelids by the Peruvian Initial Period (1750–1000 BC), although later
(Shimada et al. 1983) she indicated that domestic camelids did not occur at
Huaca Lucia-Cholupe until 700 BC. Pozorski (1976, p. 101) suggested an
Initial Period/Early Horizon period (c. 1000 BC) for introduction of
camelids at Cabello Muerto, with subsequent extensive dependence upon
camelids during the Moche phases (beginning 200 BC). Recently, Pozorski
(Lynch 1986, p. 174) suggested that an invasion of new peoples brought
maize, new ceramic and architectural styles, guinea-pigs, and camelid
husbandry to the north-coast Casma Valley about 900 BC. Both highland
and coastal evidence suggests a northward expansion of domestic camelids
sometime between 1000 and 500 BC.

Southern highlands
Ayacucho appears to represent an area transitional between possible
domestication centres in the Junin and Titicaca grasslands. In his initial
analysis, MacNeish (1969, pp. 26, 38) believed that he had isolated
evidence for the llama at Jaywamachay Cave between 6300 and 5000 BC,
with domestication clearly evident in the subsequent phase from 5000 to
3800 BC. This initial assessment was subsequently revised, with the first
appearance of domestic camelids not seen until 3100–1750 BC (MacNeish
et al 1975, p. 46). Both initial calculations by Wing (1978, p. 169), and my
subsequent computations based on the final report (MacNeish et al. 1983)
suggest relatively little dependence on camelids, especially as contrasted
with either Junin to the north or Titicaca to the south.
One of our major problems in terms of clearly identifying the Titicaca
basin as a second centre of domestication is that we have almost no good
data from sites prior to 1500 BC. Minaspata and Marcavalle are
representative of data from the Cuzco area. The initial analysis of
Minaspata assemblages, for a unit c. 1000 BC to AD 500 (Wing 1973, 1978),
indicated an MNI estimate off. 45 per cent camelids. Subsequent analysis
with finer chronological units, provided the NISP estimates given in Table
23.3. The apparent shift in camelid percentage actually reflects a drop in
guinea-pig dependence at the site; the camelid percentage might be argued
to be rather constant if guinea-pig-‘corrected’. The sequence from
Marcavalle is much finer in time-scale, with four phases from 1000 BC to
600 BC (Mohr-Chavez 1982, p. 244); MNI-based camelid percentages in
these phases range from 84 to 97 per cent.

Table 23.3 Camelid content of faunal remains at Minaspata (based on Dwyer & Wheeler 1985)
Date Original report % Guinea-pig-‘equalized’ %
1000–0 BC 55 73
AD 1–500 79 79
AD 500–1000 83 83

Table 23.4 Camelid content of faunal remains of Pikicallepata (source: Wing 1973, 1978)
Date Camelid percentage
1350–1150 BC 56
1150–950 BC 48
950–750 BC 46
750–250 BC 40

Table 23.5 Camelid content of faunal remains at Chiripa

Date Camelid percentage
1350–1000 BC 98
1000–850 BC 91
850–600 BC 92
600–350 BC 91
350 BC–AD 100 88
 In the Cuzco sites, there is good match between MNI and NISP
estimates, further south, in the Pucara vicinity, the MNI and NISP estimates
are not in as good concordance. For the site of Pikicallepata, the number of
guinea-pigs, as well as the MNI computations, result in lower estimates of
camelids (Table 23.4).
For the period 1450–1050 BC the site of Qaluyu had 41 per cent camelid
(corrected for human and fish), and the site of Q’ellokaka had 28 per cent,
using MNI estimates. But for Pucara, using NISP estimates, Wheeler &
Mujica (1981) computed 96 per cent camelids in the pre-Pucara levels, 93
per cent in the 850–500 BC levels, and 97 per cent in the 500–200 BC levels.
I suspect that the disparity in percentage is an artefact of the NISP versus
MNI approaches, rather than a sudden shift in camelid importance at the
site.
At the southern end of Lake Titicaca, the Chiripa NISP-based sequence
also covers no earlier than 1350 BC (Table 23.5). The gradual decrease in
camelid proportions is due to an increase in numbers of birds from the lake,
which also correlated with an increase in fish bone, suggesting a probable
increase in exploitation of lacustrine resources.
The evidence from Ayacucho, Cuzco, and Titicaca indicates a shift to
herding prior to 1500 BC; the revised Ayacucho data indicate that herding
can be identified no earlier than c. 3000 BC which is roughly in line with the
Junin data.
The evidence from further south indicates a later spread of herding into
Argentina and Chile. The data here are much more tentative. Jensen and
Kautz (1974, p. 46) reported domestic camelids at Tarapaca 2A, which they
estimated to be c. 2000–1500 BC. Subsequent analyses by Simons (1973,
1980) indicated that the Tarapaca 2A materials were clearly guanaco, but
there might be some domestic camelids at Tarapaca 12 & 18, c. 2700–1800
BC. Hesse (1982) reports identification of domestic camelids from the
Atacama area by 3000–2500 BC. Clear evidence of domestic camelids can
be found in Rio Loa by 500 BC. In Argentina, Yacobaccio (1979) reports
domestic camelids at Huachichona, Humahuaca, dating to 1450 BC and
further south at Las Cuevas, Salta, by 500 BC. The Chilean and Argentine
evidence is sketchy, but appears to support a southerly diffusion of domestic
camelids by the 2nd millennium BC.
Population and management shifts

The idea that shifts in patterns of mortality are indicative of human

management or domestication has been current in studies of herding for
some time, and is one presently emphasized in the Andes. Collier & White
(1976) have challenged the idea that a shift to many immature animals in an
archaeozoological sample indicates domestication. They show wide
variation in population structures in non-domestic herds, in vulnerability of
age/sex classes to predators, and in selective hunting impact. While these
are valid criticisms, herding practices in the Andes make it appropriate to
consider the mortality argument seriously. Contemporary herders in the
Andes experience high mortality amongst neonate animals of both camelids
and introduced European stock, due to climatic variables and disease.
Wheeler (1982, 1984) and Lavallee et al (1984) have argued that neonate
mortality is principally due to enterotoxaemia diarrhoea caused by
Clostridium sp. bacteria. Risk of this disease is very high in modern herding
practices, as the disease is spread in corrals, and in specialized grazing areas
called ‘bofedales’. Hence Wheeler argued that a shift to a high number of
neonate deaths can be interpreted as evidence of these herding practices,
and thus evidence of domestication. Similarly, Kent (1982a) has pointed out
that occupation of a site during the calving season can result in high
numbers of neonate remains. Wheeler has suggested that the occurrence of
a single burial of 13 neonate camelids from Telarmachay is clear evidence
of diarrhoea-related mortality; but Franklin (1982) and Malo Anccasi
(1982) point out that among both wild and domestic camelids, climate is a
major cause of mortality. In my experience with Bolivian herders,
enterotoxaemia did not kill in a single catastrophe, rather deaths were
spread out over several weeks. On the other hand, an unexpected snowfall
during February resulted in the immediate catastrophic death of nearly 80
per cent of the camelid offspring in one herd. Obviously, caution must be
exercised in interpreting the cause for shifts in neonate numbers in mortality
curves.
The generally accepted interpretation of a substantial number of juvenile
camelids in an assemblage is that the primary purpose of herding was for
meat animals. However, if there are substantial numbers of adult domestic
animals it is assumed that a shift in management has occurred, with an
emphasis on keeping older animals for both wool production and for use as
caravan animals. Sites in the Andes provide evidence for both types of
herding. The Pucara and the Chiripa patterns indicate a clear shift toward
keeping older animals for caravan and wool purposes in the Titicaca basin.
Contrasting the patterns from Kent’s (1982a) analyses of a Junin pampa site
(Pachamachay) and a Titicaca basin site (Chiripa), it appears that two
different management patterns are being employed. The Junin focus on
pastoralism seems primarily meat-oriented, while that of the Titicaca area
seems primarily wool- and caravan animal-oriented.
The meat-orientation emphasis appears to have been maintained in the
northern part of Peru for an extensive period. When camelid utilization
expanded north c. 1000–500 BC, it was the meat aspect which appeared to
be important; juvenile animals dominated at Chavin after 400 BC, with 40–
50 per cent of the animals slaughtered at 2–3 years of age; and a similar
pattern is observed for Huacaloma, where the shift from a meat orientation
to a cargo/wool orientation does not occur until the Middle Cajamarca
phase c. AD 500 (Shimada 1985).

The ‘charki’ factor

The well-known ‘schlepp’ effect of Perkins & Daly (1968) describes the
practice of bringing only the edible parts of a carcass back to the home
base. Dried and salted camelid meat is known as ‘charki’ in the Andes.
Miller (1979, 1981) observed an absence of foot elements at the
archaeological site of Chavin; he also noted that modern herders near Cuzco
only traded charki made from body and limb parts, and never used foot or
head elements. Thus, he hypothesized that an absence of foot elements
could be correlated with meat traded as charki, and coined the term ‘charki
effect’, which is synonymous with the ‘schlepp effect’ of the Near East.
However, there is a major problem with this charki effect : it simply does
not exist uniformly. In markets of Bolivian Aymara, charki is often
composed exclusively of flattened heads (broken, dried and stacked in 0.5
m diameter sheets) and foot elements. Stanish (1985) and Williams & Clark
(1986) report a number of Late Intermediate Period (AD 1000–1450) tombs
in the Moquegua coastal zone, immediately west of the Titicaca basin,
which include large numbers of once-articulated llama feet (mainly
phalanges 1, 2, and 3) but no other llama bones. Similar offerings of llama
feet are reported for tombs of the Maitas culture (AD 750–1200) in the
Azapa valley of Chile (Munoz Ovalle & Focacci Aste 1985, p. 26), just
southwest of the Titicaca basin. Knowing this Bolivian pattern for trade in
charki, where foot and head elements dominated, Miller’s assertion that
head and foot elements are ‘never used’ for charki is not true, and caution
must be exercised in arguing that a shift in foot/limb elements implies either
a management shift to meat production, or to cargo/wool orientation.
Some reassessment might be needed in the argument for the appearance
of coastal llama herds by 700 BC. Shimada et al. (1983) and Shimada &
Shimada (1985) assume that since the charki effect pattern is not observed
on the north coast sites, therefore trade is not responsible for the presence of
llama in these sites, and that the bone distribution is evidence of coastal
herding. However, as the charki effect is not universal, this argument
collapses. Moreover, today in the southern coastal areas, live animals
brought in by highland herders are a very important trade item. Faunal
assemblages from recent coastal sites would include not only bones from
entire animals, but the sites would also include the corrals where the
caravan animals were kept. Thus, the occurrence of corrals and entire
animal skeletons on the coast is not evidence against trade and for local
herding, but might be seen as prima facie evidence for trade.
The limb/foot ratio, and the neonate/juvenile/adult ratios do provide us
with techniques to identify shifts in utilization and management. Together
with other evidence they can be utilized to demonstrate certain aspects of
domestication.
Kent (1982a, b) suggests that one ought to be able to distinguish highland
from coastal camelids based on osteon (Haversian system) densities, but
this idea has not been tested.

Morphological traits

The most appropriate measure of domestication might be argued to be the

occurrence of discrete morphological traits only present in domestic
animals. There are some size distinctions between the llama and alpaca, but
unfortunately the same is true of the guanaco and the vicuna. Wing (1972)
distinguished small camelids from large camelids on the basis of the
measurements of femur, humerus, and scapula, and later refined this (1973)
to measurements of the astragalus, calcaneum, and distal widths of the
humerus and tibia. While this provided a mechanism for determining the
number of small versus large camelids on a site, it did not permit
differentiation of wild from domestic (alpaca and vicuna are both small;
guanaco and llama are large). Independently, Miller (1979) and Kent
(1982a) worked out a series of metrical criteria to differentiate all four
species. Kent’s set is the most complete, and apparently the most
successful.
Rick (1980, 1983) proposed that Pachamachay had been inhabited by a
settled population of vicuna-hunters, living year around at the site, with
seasonal utilization first occurring in Period 7, when he suggested a dry-
season period of hunting. Kent’s (1982a) re-analysis of the fauna, using his
morphological parameters, indicated no evidence for vicuna being the
major camelid, and that, if anything, after Rick’s Period 5(c. 1500 BC),
domestic camelids, particularly alpaca, dominated; and that if there was any
seasonality in Period 7 (800–400 BC), it was likely wet-season occupation,
not dry-season. Thus, a faunal analysis using better diagnostic criteria
contradicted most of Rick’s model for sedentary vicuna-hunting. For those
of us believing the domestication of alpaca to have taken place in Junin,
possibly as early as 4000 BC, Rick’s model was uncomfortable, as it flatly
contradicted the domestication postulates; thus I prefer Kent’s hypothesis.
Apart from the relative sizes of the bones, another morphological trait
that can be used to identify camelids is the shape of the incisors. Alpaca
incisors are thought to be intermediate in form between vicuna and
guanaco/llama. Vicuna incisors are parallel-sided, with open roots; guanaco
and llama incisors are spatulate-sided with closed roots. Alpaca incisors
have an enamel distribution like that of the vicuna, but the cross-section is
more rectangular than it is in the vicuna incisor, where it is square. Wheeler
(1984) has used this trait to argue for alpaca domestication as early as 4000
BC at Telarmachay. While this seems to be clearly an alpaca trait, not all
alpaca exhibit such incisors (Kent 1982a, p. 142, Shimada & Shimada 1985,
p. 18), so it is only useful in identifying the presence of alpaca, not in
quantifying their numbers.
Final remarks

Our current understanding suggests that there was at least one centre of
domestication in the Lake Junin area, and possibly a second in the Lake
Titicaca area. Consideration of all parameters indicates a probable period of
intensification of utilization of camelids c. 7000 BC, with sufficient
management taking place that domestic animals may be recognized by c.
4000 BC. The centre for alpaca domestication may be the Lake Junin area.
In measures of small versus large camelids, the ratio of small camelids in
the Lake Junin archaeological sites frequently goes as high as 9 or 10 to 1,
while, in contrast, large camelids frequently dominate the Lake Titicaca
environs, being 3 or 4 times as frequent in some sites as small camelids. A
shift in the management of camelids from meat to wool and cargo-bearing
appears to occur first in the Lake Titicaca region. The two zones are for a
time distinguishable in terms of focus on large versus small domestic
camelids, and on meat versus wool/cargo-bearing attributes.

References
Burger, R. L. 1985. Prehistoric stylistic change and cultural development at Huaricoto, Perú.
National Geographic Research 1(4), 505–34.
Cardich, A. 1983. A propósito del 25 aniversario de Lauricocha. Revista Andina 1(1), 151–73.
Cardozo, A. 1975. Origen y filogenia de los camélidos sudamericanos. La Paz: Academia Nacional
de Ciencias de Bolivia.
Collier, S. & J. P. White 1976. Get them young? Age and sex inferences on animal domestication in
archaeology. American Antiquity 41(1), 96–102.
Craig, A. K. 1985. Cis-andean environmental transects: late Quaternary ecology of northern and
southern Peru. In Andean ecology and civilization, S. Masuda, I. Shimada & C. Morris (eds), 23–
44. Tokyo: University of Tokyo Press.
Dwyer, E. B. & J. C. Wheeler 1985. Animal utilization in the southern highlands: Early horizon to
early intermediate at Minas Pata, Lucre Valley, Cuzco. Paper presented at the 50th annual
meetings, Society for American Archaeology.
Franklin, W. L. 1978. Socioecology of the vicuna. Unpublished PhD dissertation, Wildlife Sciences,
Utah State University, Provo, Utah.
Franklin, W. L. 1982. Biology, ecology, and relationship to man of the South American camelids.
Mammalian Biology in South America. Pymatuning Symposia in Ecology 6, M. A. Mares & M. H.
Genoways (eds), 457–90. Pittsburgh: University of Pittsburgh Press.
Franklin, W. L. 1983. Contrasting socioecologies of South America’s wild camelids: the vicuna and
the guanaco In Advances in the study of mammalian behavior, J. F. Eisenberg & D. G. Kleiman
 (eds), 573–629. Special Publications of the American Society of Mammalogists 7.
Gade, D. W. 1969. The llama, alpaca and vicuna: fact vs. fiction. Journal of Geography 68(6), 339–
43.
Hesse, B. 1982. Archaeological evidence for camelid exploitation in the Chilean Andes.
Säugetierkundliche Mitteilungen 30(3), 201–11.
Jefferson, R. T. Jr. 1980. Size and spacing of sedentary guanaco family groups. Unpublished MS
thesis, Animal Ecology, Iowa State University, Ames, Iowa.
Jensen, P. M. & R. R. Kautz 1974. Preceramic transhumance and Andean food production. Economic
Botany 28(1), 43–55.
Kent, J. D. 1982a. The domestication and exploitation of the South American camelids: methods of
analysis and their application to circum-lacustrine archaeological sites in Bolivia and Peru.
Unpublished PhD dissertation, Anthropology, Washington University, St. Louis, Missouri.
Kent, J. D. 1982b. Osteon population density and age in South American camelids. Unpublished
paper presented at the 47th annual meeting, Society for American Archaeology.
Koford, C. B. 1957. The vicuna and the puna. Ecological Monographs 27(2), 153–219.
Lavallee, D., M. Julien & J. Wheeler 1984. Telarmachay: niveles precerámicos de ocupación. Revista
del Museo Nacional 46, 55–133.
Lees, S. H. & D. G. Bates 1974. The origins of specialized nomadic pastoralism: a systemic model.
American Antiquity 39(2), 187–93.
Lynch, T. 1986. Current research: Andean South America. American Antiquity 51(1), 171–6.
MacNeish, R. S. 1969. First annual report of the Ayacucho archaeological-botanical project.
Andover: Robert S. Peabody Foundation for Archaeology, Phillips Academy.
MacNeish, R. S., T. C. Patterson & D. L. Browman 1975. The central Peruvian prehistoric interaction
sphere. Papers of the Robert S. Peabody Foundation for Archaeology 7. Andover: Phillips
Academy.
MacNeish, R. S., R. K. Vierra, A. Nelkin-Terner, B. Lurie & A. Garcia Cook 1983. Prehistory of the
Ayacucho Basin, Peru. Vol. 4: The preceramic way of life. Ann Arbor: University of Michigan
Press.
Malo Anccasi, M. 1982. Causas de mortalidad en crías de alpaca. Allpak’a 1(1), 31–4.
Miller, G. R. 1979. An introduction to the ethnoarchaeology of Andean camelids. Unpublished PhD
dissertation, Anthropology, University of California, Berkeley, California.
Miller, G. R. 1981. Subsistence and social differentiation at Chavin de Huantar: some insights from
the preliminary analysis of the faunal remains. Paper presented at the 46th annual meetings,
Society for American Archaeology.
Möhr Chavez, K. L. 1982. The archaeology of Marcavalle, an Early Horizon site in the valley of
Cuzco, Peru, Pt. 1. Baessler-Archiv 28, 203–329.
Moore, K. M. 1982. Prehistoric animal use in Junin: preliminary results of the 1981 season at
Panaulauca Cave. Paper presented at the 10th annual meeting, Midwest Conference on Andean
and Amazonian Archaeology and Ethnohistory.
Moore, K. M. 1985a. Current research in the animal economy of the central Andes: teeth and bones
from Panaulauca Cave, Junin. Paper presented at the 13th annual meeting, Midwest Conference
on Andean and Amazonian Archaeology and Ethnohistory.
Moore, K. M. 1985b. Hunting and herding economies on the Junin puna: recent
paleoethnozoological results. Paper presented at the 50th annual meeting, Society for American
Archaeology.
Munoz Ovalle, I. & G. Focacci Aste 1985. San Lorenzo: testimonio de una comunidad de
agricultores y pescadores Post-Tiwanaku en el Valle de Azapa (Arica, Chile). Chungara 15, 7–30.
Perkins, D. & P. Daly 1968. A hunter’s village in Neolithic Turkey. Scientific American 219(5), 96–
106.
Pires-Ferreira, E., J. Wheeler Pires-Ferreira & P. Kaulicke 1977. Utilización de animales durante el
período precerámico en la cueva de Uchcumachay y otros sitios de los Andes centrales del Perú.
Journal de la Société des Americanistes, Paris 64, 149–54.
Pozorski, S. G. 1976. Prehistoric subsistence patterns and site economies in the Moche Valley, Peru.
Unpublished PhD dissertation, Anthropology, University of Texas, Austin, Texas.
Raedeke, K. J. 1979. Population dynamics and socioecology of the guanaco (Lama guanicoe,) of
Magallanes, Chile. Unpublished PhD dissertation, Zoology, University of Washington, Seattle,
Washington.
Rick, J. W. 1980. Prehistoric hunters of the high Andes. New York: Academic Press.
Rick, J. W. 1983. Cronología, clima y subsistencia en el precerámico peruano. Lima: Instituto
Andino de Estudios Arqueológicos.
Rosas, H. & R. Shady Solis 1974. Sobre el período formativo en la sierra del extremo norte del Perú.
Arqueológicas 15, 6–35.
Shimada, I., C. G. Elera & M. J. Shimada 1983. Excavaciones efectuadas en el centro ceremonial de
Huaca Lucia-Cholope, del Horizonte Temprano, Batan Grande, costa norte del Perú: 1979–1981.
Arqueológicas 19, 109–210.
Shimada, M. J. 1981. Ethnozooarchaeology of north Perú: highland-coast comparisons. Paper
presented at the 46th annual meetings, Society for American Archaeology.
Shimada, M. J. 1982. Zooarchaeology of Huacaloma: behavioral and cultural implications. In
Excavations at Huacaloma in the Cajamarca Valley, Perú, 1979, K. Terada & Y. Onuki (eds), 303–
36. Tokyo: University of Tokyo Press.
Shimada, M. J. 1985. Continuities and changes in patterns of faunal resource utilization: Formative
through Cajamarca periods. In The Formative Period in the Cajamarca Basin, Perú: excavations
at Huacaloma and Layzon, 1982, K. Terada & Y. Onuki (eds), 289–310. Tokyo: University of
Tokyo Press.
Shimada, M. J. & I. Shimada 1985. Prehistoric llama breeding and herding on the north coast of
Perú. American Antiquity 50(1), 3–26.
Simons, D. D. 1973. Faunal remains from northern Chile. Paper presented at the 38th annual
meetings, Society for American Archaeology.
Simons, D. D. 1980. Man and guanaco at an early site in northern Chile. In Prehistoric trails of
Atacama: archaeology of northern Chile, C. W. Meighan & D. L. True (eds), 189–94. Monumenta
Archaeologica 7. Los Angeles: University of California at Los Angeles.
Stanish, C. 1985. Mortuary architecture and interregional elite alliance in the Post-Tiwanaku south
central Andes. Unpublished PhD dissertation, Anthropology, University of Chicago, Chicago,
Illinois.
Wheeler, J. C. 1982. Lamoid domestication and early development of pastoralism in the central
Peruvian Andes. Paper presented at the 47th annual meetings, Society for American Archaeology.
Wheeler, J. C. 1984. On the origin and early development of camelid pastoralism in the Andes. In
Animals and archaeology. Vol.3: Early herders and their flocks, J. Clutton-Brock & C. Grigson
(eds), 395–410. Oxford: BAR International Series S202. (Spanish version [1984]: La
domesticacíon de la alpaca (Lama pacos) y la llama (Lama glama) y el desarrollo temprano de la
ganadería autóctona en los Andes centrales. Boletín de Lima 36, 74–84.
Wheeler, J. C., E. Pires-Ferreira & P. Kauhcke 1976. Preceramic animal utilization in the central
Peruvian Andes. Science 194(4264), 483–90. (Spanish version [1977]: Domesticación de los
camélidos en los Andes centrales durante el período precerámico: un modelo. Journal dela Société
des Americanistes, Paris 64, 155–65.
Wheeler, J. C. & E. J. Mujica 1981. Prehistoric pastoralism in the Lake Titicaca basin, Perú, 1979–
80 field season. Final report to the National Science Foundation.
Williams, S. R. & N. R. Clark 1986. Investigations in the burial areas of the late prehistoric
Estuquina site near Moquegua, far southern Peru. Paper presented at the 14th annual meeting,
Midwest Conference on Andean and Amazonian Archaeology and Ethnohistory.
Wing, E. S. 1972. Preliminary report on the prehistoric use of animal resources in the Peruvian
Andes. Paper presented at the 37th annual meeting, Society for American Archaeology.
Wing, E. S. 1973. Utilization of animal resources in the Peruvian Andes. In Andes 4: Excavations at
Kotosh, Peru, S. Izumi & K. Terada (eds), 327–52. Tokyo: University of Tokyo Press.
Wing, E. S. 1977. Caza y pastoreo tradicionales en los Andes Perúanos. In Pastores de Puna:
Uywamichiq punarunakuna, J. Flores Ochoa (ed.), 121–30. Lima: Instituto de Estudios Perúanos.
Wing, E. S. 1978. Animal domestication in the Andes. In Advances in Andean archaeology, D. L.
Browman (ed.), 167–89. The Hague: Mouton.
Wing, E. S. 1980. Faunal remains. In Guitarrero Cave: early man in the Andes, T. F. Lynch (ed.),
149–72. New York: Academic Press.
Yacobaccio, H. D. 1979. Arte rupestre y tráficode caravanas en la puna de Jujuy: modelo e hipótesis.
Paper presented at the Jornadas de Arqueológia del Noroeste Argentino, Universidad del Salvador,
Buenos Aires.
24 Are llama-herders in the south central
Andes true pastoralists?
MARIO A. RABEY

Llamas (Lama glama) are currently classified as ‘domestic’ animals, the

human societies that interact with them as ‘pastoralists’, and the set of
interactions between these camelids and such human societies is often
named ‘pastoralism’. In this chapter, I will present evidences that suggest
that these classifications may be not fully appropriate. The information used
here was obtained from my own fieldwork among the llamas’ herders in the
farthest part of northwest Argentina, between 21°S and 23°S. I have used
data published by Flores Ochoa (1968, 1975) and Custred (1977) about
herders in Peru, and by Aldunate et al. (1981), Castro et al. (1982), and
Gunderman (1984) with reference to Chile.
It must be pointed out that alpaca (Lama pacos), the other Andean
domestic camelid, does not exist in the Argentine Andes, where the herds are
only llamas. But in most parts of the Andes, present-day herding practice is
to breed llamas and alpacas together. Inhabitants of punas – the land situated
above 3000–3500 m, in the zone where llamas are found – also breed sheep
and goats, but these are managed separately. Unmixed llama breeding is of
great theoretical interest, because the llama is a more generalized animal
than the alpaca, and is probably more similar to the wild ancestor.
Consequently, unmixed llama-herding provides a model which may be more
appropriate in understanding the domestication process and the relationship
between humans and animals in that process, than the mixed herding of
llamas and alpacas.
Although there are several ways of breeding llamas, depending on the
type of habitat involved, in the Argentine punas there are two main systems,
called by the local herders: del cerro (hill breeding) and del campo (plains
breeding).
Hill breeders (Figs. 24.1 & 24.2) live between 3900 and 4600 m above
sea-level (ASL). These heights, and those which follow, are approximate and
can vary in the order of 300 m, according to the specific ecological features
of different areas. Their social unit is a family, or a coalition of kindred
families, which usually has possession of the territory surrounding a small
stream, and whose boundaries are the peaks separating it from neighbouring
small basins.
The hill breeders’ main homestead is situated on the lower part of their
territory, at about 4100 m, since they can thus control the fringe between
3900 and 4300 m. They remain there during the rainy season, from late
November to the beginning of April. Their llamas make use of two foraging
zones in that area:

(a) the small plains and light slopes covered with a sparse steppe of shrubs,
the tolar, which in the rainy season is covered with a carpet of small plants;
(b) the ciénegos (bogs) bordering the streams, where there is an unbroken
vegetation cover.

When the rains come to an end, the seasonal pastures become exhausted, and
the ciénego pastures cannot support the herds of llamas and other domestic
animals.
Figure 24.1 The two herding systems.
 Figure 24.2 The hill people’s herding system.

At that time of year, the peasants and their flocks carry out one of the
changes of ground in the annual cycle currently named ‘transhumance’.
From the grazing grounds near the main homestead, they move to the zone
situated between 4300 and 4800 m, near the steep mountain peaks which are
covered by eternal snow. The animals here again make use of two grazing
areas:

(a) the highest slopes covered with a scattering of grass, and also carpeted by
small plants during the rainy season and immediately afterwards, the iral;
(b) the vegas, on the permanently moist ground formed by daily thawing of
ice, an area where llamas prefer to forage.

Sheep and goats often remain on the lower levels, because they cannot
withstand the cold as the llamas do. The llamas go up alone to the high
mountains, around 15 April, without any driving by the herders but, on the
contrary, followed by them.

The herders return to their main homestead at the end of July, to make
chayacos, a basic ritual in their indigenous religion and adaptive strategy
(Merlino & Rabey 1983, Rabey & Merlino 1985). They take some llamas
with them: those they can drive easily, usually young animals. In fact, it is
the only moment of the annual cycle in which llamas do not move
spontaneously, but must be driven by the herders. The herders do not worry
about animals which remain in the uplands because all are adults, including
a high percentage of males. Pumas (Felis concolor), the only adult llamas’
predator, cannot easily make such animals their prey. The herders and their
young flocks remain a month at the main homestead, in whose neighbouring
bogs there is sufficient pasture for the occasion. Then control over the llamas
is relaxed, and they afterwards return to the high area, followed by the
herders. A new descent is made in December, when the rainy season
generally starts. All the llamas now go down, and they do so without any
human driving. Animals and herders then remain in the lowland area, until
the beginning of the following dry season.
 The hill herders’ control over their llamas is very lax. With the exception
of pregnant females and those with offspring, the llamas forage alone over
well-defined natural territories, migrating periodically without any human
driving. The herders’ care is restricted to:

(a) selecting males as breeders, habitually those that protect females and
young without a great amount of aggression;
(b) protecting young llamas from predators;
(c) feeding orphans and baby llamas whose mothers have no milk;
(d) periodically keeping the herd together.

The herding system is very different among the plains breeders (Fig. 24.1).
The annual cycle of seasonal movements is reversed here (Rabey et al. 1985,
p. 25). During the rainy season, the herders live on the mountainous fringe of
shrubby steppe that surrounds the puna’s fluvial and lacustrine plains. By the
beginning of April, when the dry season starts, they move to their main
homesteads on the bottom slopes. Llamas then graze on the temporarily
moist ground, as the waters dry out. Some family members occasionally
settle on the lower fringe at the edge of rivers or lakes, where vegetation is
similar to that of the ciénegos. Finally, at the end of the dry season, the
animals graze on the bottom of the depressions. When the dry season
finishes and the rains begin, the herders and their animals go up again to the
highlands.
Contrary to what happens in the hills, in the plains breeding system the
movements of llamas are guided by the herders. This difference is clearly
apparent in their going down at the end of the rainy season. The llamas then
try to go up to higher pasture grounds. The plains herders must then exercise
a great deal of effort to make their llamas go to the lower pasture grounds.
With regard to the relationship between the llamas’ behaviour and that of
the human population linked to them, some conclusions can be reached from
comparing the two herding systems. First, it seems that there is an innate
behavioural pattern that guides the seasonal movements of llamas according
to a basic rainy season-lower areas and dry season-higher areas scheme. This
pattern was probably fixed early in the behaviour of llamas’ wild ancestors,
even before their first contact with human populations; although some
research must be done in order to eliminate the alternative hypothesis of a
behavioural pattern learned from humans. The control in hills breeding is
only directed towards: (a) obtaining a product; (b) diminishing the mortality
in young animals; and (c) strengthening the territorial features of llamas’
behaviour, like the demarcation of feeding territories. The structural
elements in the more generalized hills pattern are two of the llama’s
behavioural features:

(a) the selection of a feeding and sleeping territory by a single family group;
(b) the annual cycle of movements.

The first feature is, according to Franklin (1983, p. 573), common with the
contemporary American wild camelids; the second one is characteristic of
llamas, but it may be also characteristic of the (still not well-known)
behaviour of wild guanacos (Lama guanicoe) which, according to Raedecke
(1978), Franklin (1983, p. 573) and Cajal (1985, p. 90), are either migratory
or sedentary. The seasonal ground-changing cycle of the punas’ inhabitants
must have been built around their camelids’ behavioural pattern. The
Andean ground-changing cycle could be a cultural pattern basically adapted
to the innate behavioural pattern of llamas or of their wild ancestors, and
based on a proper knowledge of this.
On the contrary, on the plains surrounding the rivers and lakes, herders
must make a strong effort to use the natural resources available there for
their llamas. The inversion of the annual cycle of movements which they
must practise results in a much stricter human control over the llamas’
activities than that practised by their neighbouring herders in the hills.
Likewise, in Peru and northern Chile, where mixed herds of llamas and
alpacas are the normal system, the herding scheme is strongly modified by
the presence of the alpacas. In fact, the alpaca is an animal with much more
strict ecological, nutritional, and management needs than the llama. For
instance, as Gunderman (1984, p. 107) says, herders must keep the males
and females separated from each other. Therefore, herds containing alpacas
require intense human care, which includes management of mating and
continuous herding over good pasture grounds.
It is, therefore, evident that in the more generalized herding system of the
hills, llamas and human societies do not constitute a pastoralist system, at
least in the conventional meaning of the word, because llamas forage almost
freely and – except in the case of young individuals and their mothers – they
require almost no care. The herders’ family activity consists of
accompanying, but not leading, them on their annual grazing cycle, and
practising a set of minor controls designed to reduce natural pressures.
Therefore, the relationship between llamas and human societies in the upper
Argentine punas should not be termed ‘pastoralism’.
Furthermore, the set of selective pressures on llama populations created
by human control in the environment of the high punas cannot be named
‘domestication’ if we use the most current definitions. For instance, Bökönyi
(1969, 1985, ch. 2, this volume) stresses that the essence of domestication
includes basically the removal of the animal species from their natural
habitat and natural breeding community, and their maintenance under
controlled breeding conditions. Zeuner (1963) defined domesticated species
as those whose breeding is completely controlled by humans; based on this,
Eisenberg (1986) says that domestication ‘reaches a point where the gene
pool of a species is under the total control of humans’. In the punas, llamas
are not removed from their natural habitat, neither are they maintained under
controlled breeding conditions, except the above-mentioned selection of
males; even less is their gene pool under the total control of humans.
Moreover, the behavioural changes postulated by Meadow (1984) as a
consequence of the domestication process are not apparent in the llama’s
case. The man–llama relationship can only be included in the
‘domestication’ category if we use a broader definition, such as that
proposed by Ducos (1978, p. 54, ch. 3, this volume), for whom
domestication is linked with human appropriation of living animals as
economic objects of ownership, inheritance, and exchange.
The distinction between hunting and pastoralism, or between wild and
domesticated animals, may therefore become non-operative. Ingold (1980,
1986) has presented data about the man–reindeer relationship which also do
not enter into the above-mentioned definitions; reindeer are domestic
because they belong to the Lapp households, this domestication being a
consequence of taming, i.e of a man–animal social relation, and not of a
biological process. From my own point of view, this set of problems in the
classification of man–animal relationships will require much stronger
theoretical discourse than that made to date, at least on the ‘domestication’
side of the problem.
Acknowledgements
The fieldwork on which this paper is based was possible thanks to CONICET (National Council for
Scientific and Technological Research, Argentina) financial aid. An important part of this fieldwork
and the subsequent theoretical analysis has been carried out in co-operation with the biologist R.
Tecchi. I am also grateful to R. Merlino with whom I discussed various subjects developed here. J.
Garcia Fernandez had the idea for the figures and C. Colarich converted them into their final form.
Pedro Edmunds and Norberto Méndez contributed to the English version of this chapter.

References
Aldunate, C., J. Armesto, V. Castro & C. Villagrán 1981. Estudio etnobotánico en una comunidad
precordillerana: Toconce. Boletín del Museo de Historia Natural, Chile 38, 183–223.
Bökönyi, S. 1969. Archaeological problems and methods in recognizing animal domestication. In The
domestication and exploitation of plants and animals, P. J. Ucko & G. W. Dimbleby (eds), 219–29.
London: Duckworth.
Bökönyi, S. 1985. Problèmes archéozoologiques. In La protohistoire de l’Europe, J. Lichardus & M.
Lichardus–Itten (eds), 571–81. Paris: Presses Universitaires de France.
Bökönyi, S. 1989. Definitions of animal domestication. In The walking larder, J. Clutton–Brock (ed.),
ch. 2. London: Unwin Hyman.
Cajal, J. L. 1985. Comportamiento. In Estado actual de las investigaciones sobre camélidos en la
República Argentina, J. L. Cajal & J. M. Amaya (eds), 87–100. Buenos Aires: Secretaría de Ciencia
y Técnica.
Castro L., M. C. Villagrán & M. K. Arroyo 1982. Estudio etnobotánico en la precordillera y altiplano
de los Andes del Norte de Chile (18–19°S). In El ambiente natural y las poblaciones humanas del
Norte Grande de Chile 2, A. Veloso & E. Bustos (eds), 133–203. Montevideo: UNESCO.
Custred, G. 1977. Las punas de los Andes Centrales. In Pastores de Puna, J. Flores Ochoa (ed.), 55–
85. Lima: Instituto de Estudios Peruanos.
Ducos, P. 1978. ‘Domestication’: definition and methodological approaches to its recognition in faunal
assemblages. In Approaches to faunal analysis in the Middle East, R. H. Meadow & M. A. Zeder
(eds), 53–56. Harvard University: Peabody Museum Bulletin 2.
Eisenberg, J. F. 1986. The selection of mammalian species for domestication. Precirculated paper. In
Cultural attitudes to animals including birds, fish and invertebrates. Vol. 1, World Archaeological
Congress (mimeo).
Flores Ochoa J. 1968. Los pastores de Paratía: una introducción a su estudio. Mexico: Instituto
Indigenista Interamericano.
Flores Ochoa, J. 1975. Pastores de alpacas. Allpanchis 8, 5–23.
Franklin, W. L. 1983. Contrasting socioecologies of South America’s wild camelids: the vicuña and
the guanaco. In Advances in the study of mammalian behaviour. Special Publication, American
Society of Mammalogists, Vol. 7. J. Eisenberg & D. Kleiman (eds), 573–629.
Gunderman K. H. 1984. Ganaderia aytnara, ecología y forrajes: evaluación regional de una actividad
productiva andina. Chungará 12, 99–124.
Ingold, T. 1980. Hunters, pastoralists and ranchers. Cambridge: Cambridge University Press.
Ingold, T. 1986. Reindeer economies and the origin of pastoralism. Anthropology Today 2(4), 5–10.
Meadow, R. H. 1984. Animal domestication in the Middle East: a view from the eastern margin. In
Animals and Archaeology. Vol. 3: Early herders and their flocks, J. Clutton–Brock & C. Grigson
(eds), 309–37. Oxford: BAR International series 202.
Merlino, R. & M. A. Rabey 1983. Pastores del altiplano andino meridional: religiosidad, territorio y
equilibrio ecológico. Allpanchis 21, 141–79.
Rabey, M. A. & R. Merlino 1985. El control ritual-rebaño entre los pastores del sur de los Andes
Centrales. Presented at the 45 Congreso Internacional de Americanistas, Bogotá. América Indígena,
in press.
Rabey, M., R. Rotondaro & R. Tecchi 1985. El ecosistema laguna de Pozuelos. Ambiente 45, 18–25.
Raedecke, K. J. 1978. El guanaco deMagallanes, Chile: distribución y biología. Santiago, Chile:
Corporación Nacional Forestal.
Zeuner, F. E. 1983. A history of domesticated animals. London: Hutchinson.
PREDATION
Introduction to predation
JULIET CLUTTON-BROCK

Predators feed on prey; that is they are normally carnivores which kill other
living animals for food, but a parasite may also be termed a predator on its
host. It has been argued that all human activities involving the exploitation
of animals are predation and, in the social sense, parasitism. But it is only
when humans are subsisting as hunters, in fact like carnivores, that there is
no element of protection in the interactions between predators and prey. As
soon as animals are herded or domesticated humans become protectors
rather than predators. The relationship ceases to be one of simple predation
and becomes a form of symbiosis in which individual animals benefit from
the association until the moment of their death, and the species benefits
genetically by becoming much more widespread than it was in the wild.
In the development of cultural systems it is generally agreed that humans
were hunter-gatherers in the Palaeolithic period. In the ensuing Mesolithic
they began to change towards the cultivation of plants, the storage of grain,
and the keeping of dogs and maybe a few herd animals. The full
domestication of livestock animals followed, on a world-wide scale, during
the Neolithic period, around 7000 years ago. Hunting did not end with the
Palaeolithic, however, and people in almost all societies past and present
have been and are hunters, fishers, and gatherers of wild foods. The
behavioural patterns of the hunter-gatherers, which evolved in hominids
over, say, 3 million years, are still present in humans today, in the same way
that the Pekingese still has the same predatory behaviour as the wolf.
In all interactions between predators and their prey there is an attack on
the part of the predator and a defence on the part of the prey. The defence
may be active, as when a moose is attacked by a wolf, and it will fight back,
sometimes even killing the predator, or it may be passive and achieved by
the concentration of numbers, as in a colony of tightly packed shellfish or
the bunching together of herd animals. Wilson (1980, p. 23), in his, now
classic, work on sociobiology, which is the study of the biological basis of
social behaviour, quotes the Ethiopian proverb that says, ‘When spiders’
webs unite, they can halt a lion.’ Wilson gives a succinct summary of the
tenets that are now well known on the evolution of social behaviour,
territoriality, and defence against predation. From these tenets it can be seen
that humans, unlike most other animals, have evolved all the strategies of
both predators and prey, and indeed in the Pleistocene they did play both
roles. As hunters they developed the physical ability to range over large
territories and the co-operative skills to kill prey much larger than
themselves, but their gregarious behaviour and tolerance of crowding also
provided a defence against predators such as wolf, bear, and the big cats.
The need for this is explained by Geist in Chapter 25 who suggests a new
scenario for predator–prey interaction in the Pleistocene of North America.
Hunting may appear to the anthropologist to be a straightforward activity,
a matter of applying the necessary skills to searching for prey, killing it, and
then sharing the meat and other products amongst the nuclear family or
band. This so-called contented world of the hunter-gatherers has been
described by Sahlins (1974) as ‘the original affluent society’. From the
biological point of view, however, the ecological relationships of predators
and their prey are very involved and interdependent. With animal predators,
a balance is normally maintained by the controls of environment and
population dynamics so that prey species are not exterminated. With human
hunting at the end of the Pleistocene, these controls appear to have broken
down so that ‘overkill’ became a common occurrence in many parts of the
world.
Although some will argue against the supposition that humans were
responsible for the extermination of the mammoth, the mastodon, North
American equids, and other large mammals, it is certain that hunter-
gatherers do deplete their resources. Evidence for this is discussed by
Cooke & Ranere in Chapter 26 for prehistoric Panama where they have
carried out a comprehensive study of broad-spectrum hunting as well as
predation on invertebrates.
Sloan, in Chapter 27, postulates how the massive exploitation of
shellfish, and in particular the oyster, in the Mesolithic of northern Europe
postponed the need for the walking larder and made possible the
development of sedentary hunter-gatherer systems. The terminology of
these shellfish economies is also discussed by Uerpmann in Chapter 9.
Even the resource of shellfish could not last indefinitely, however, as is
demonstrated, on the other side of the world, by Spennemann in Chapter 28
on the over-exploitation of molluscs in Tonga, around 1500 BC.
Chapter 29 by Colley & Jones demonstrates how renewed study of the
fish remains from Rocky Cape has revealed a much greater diversity of
species than was identified in the preliminary work on this prehistoric site
in Tasmania. The question still remains, however, of why did the
Tasmanians stop eating fish?
The obvious partner for humans in the hunt is the dog, the first animal to
be domesticated, but Hooper in Chapter 30 puts forward the intriguing
suggestion that the honeyguide has had a much more ancient symbiosis in
leading early hominids to nests of bees. Another ancient interaction with
birds is described by Eastham in Chapter 31 who gives an account of the
large number of species killed by the Mousterians 40 000 years ago in
Spain. The killing of large numbers of small birds, ostensibly for food, still
continues today in Mediterranean countries, although it has become a
subject of great concern to conservationists.
As with the chapters on domestication and pastoralism, those on
predation cover a wide variety of species in many parts of the world in
ancient times and at the present day. The intention of bringing them
together in this book is to demonstrate the great diversity of interaction
between humans and animals, and the continuity of this diversity through
time. Even today, a wolf may be looked on as a dangerous predator, or as a
much-loved companion, or as a prey to be killed to make a fur hat.
The anciently formed ecological and behavioural links between humans
and animals have not changed in the modern world and it is essential for
our survival that we conserve them. It is not so much that, ‘if we do not
study the past we shall be doomed to live it’, but that if we do not study the
past we may be doomed to live like battery chickens, enclosed in small,
clean buildings, protected from predators, and with an enormously reduced
perception of the living world.

References
Sahlins, M. 1974. Stone age economics. London: Tavistock Publications.
Wilson, E. O. 1980. Sociobiology. Cambridge, Mass. & London: Belknap Press of Harvard
University Press.
25 Did large predators keep humans out of
North America?
VALERIUS GEIST

Introduction

The notion of human supremacy is so deeply rooted in our culture that it

colours our views of prehistory, including how humans colonized America.
The most popular conception of this is Martin’s (1967) overkill hypothesis,
upgraded (1973, 1984) and discussed in the affirmative by a number of
authors in Martin & Klein (1984), the latest attempt to deal with the
problems of megafaunal extinctions. Martin’s conception is one of appealing
simplicity and great attractiveness in that one can make a most plausible case
for it. Human colonization and extinctions of numerous large, huntable, and
edible mammals did go hand in hand in many localities, so why not in North
America? As Martin depicts it, humans entered North America probably 12
000–13 000 years BP, then expanded rapidly southward, colonizing all of
North America by 11 000 BP and South America by 10 000 BP. In the process
they brought, to use Martin’s term, a ‘blitzkrieg’ upon the megafauna and in
a very short time, too short to be adequately recorded in the fossil record,
exterminated the Rancholabrean megafauna.
Using much the same evidence, plus some data based on more realistic
assumptions than Martin made, one can turn those conclusions on their head:
humans, and other members of the Siberian Pleistocene fauna failed to gain
entry into North America because of the Rancholabrean megafauna, and
colonized North America only after the megafauna had begun to collapse.
The Siberian fauna is made up of cold-climate generalists. They faced in the
Rancholabrean fauna a densely packed fauna of specialists, characterized by
large-bodied herbivores and carnivores. The largest carnivore was a bear
(Arctodus simus), much larger in size than the largest brown bear (Stock
1953, Kurtén & Anderson 1980), a long-legged cursorial species with a
foreshortened, ‘bull-dog’ face. The predator fauna was diverse, containing in
the latter part of the Wisconsinian glaciation a huge lion (Panthera leo
atrox), various sabre-toothed cats (Smilodon fatalis, S. gracilis,
Homotherium serum), a cheetah-like large cat (Miracinonyx studeri), the
mountain lion (Felis concolor), the jaguar in the warmest latitudes (Panthera
onca augusta), several short-faced bears (Arctodus pristinus, A. simus,
Tremarctos floridanus), black bear (Ursus americanus), and dire wolf (Canis
dims). Man, as a super-predator, would have faced stiff competition, and
would have faced at his kills direct confrontations with at least the largest
predators. Since humans have had historically such great difficulties dealing
with the much smaller, omnivorous brown or grizzly bear (as shown below),
how could they have handled the much larger carnivorous Arctodus,
particularly in an open landscape without trees to climb? How is it that
various Siberian species appear in Alaska early in the Pleistocene, but – with
few exceptions – do not colonize lower North America until the
Rancholabrean fauna had begun to collapse?

When did humans enter North America?

There is little dispute that after 12 000 BP humans colonized lower North
America (Haynes 1967), nor do I contend that human hunters after
successful entry hastened the extinction of the remnants of the
Rancholabrean herbivores (Martin 1984). I accept Guthrie’s (1984)
ecological hypothesis of megafaunal extinctions as most likely; Guthrie
(1984) and King & Saunders (1984), point out that the proboscidians killed
by Palaeo-Indians were small, indicating a decline in net-energy and
nutrients for ontogenetic growth. That is, the mammoth and mastodons were
declining in size at the time they were hunted, an indication that ecological
conditions were probably not favourable. Other mammals also became
dwarfed (Edwards 1967, Wilson 1980, Guthrie 1984). That extinctions
coincided with a pause in deglaciation has been proposed by Ruddman &
Duplessy (1985). Also some extinctions preceded the entry of humans (i.e.
Glyptotherium, 13 970 ± 310; Eremotherium, 15 900; Miracinonyx, 12 770 ±
900; Bootherium and Equus complicates, 17 200 ± 600; Mammuthus
primigenius in Alaska, 15 380 ± 300, see Kurten & Anderson 1980, Table
19.6 & p. 364, Anderson 1984).
 However, the controversial dates indicating an earlier entry of humans
into North America are of special interest. It is as if humans pecked around
the edges of North America, but failed to colonize the continent. There are
early, disputed dates in Alaska, the Yukon Territory, and Venezuela (Bryan
1973; Rouse 1976; Stanford et al. 1981, Bischoff & Rosenbauer 1981, see
also Kurtén & Anderson 1980). Linguistic evidence also suggests an early
intrusion of humans in the south (Rogers 1985). Since there was a major
expansion of upper Palaeolithic people 40 000–30 000 years BP, and the
Alaska dates approach that range, it may well be that humans came to
America’s door in an early wave, but failed to gain entry.
If humans did indeed come earlier than 12 500 BP, they could have
travelled in boats along the west coast (boats may have been used to
colonize Australia some 40 000–35 000 years ago, see Geist 1978). If so,
then we expect the faunas of the Channel Islands to show extinctions before
12 500 BP. There were dwarf mammoth on the Channel Islands (Cushing et
al. 1984); such island dwarfs should readily fall victim to human hunters, as
so well documented by Martin (1984). From Haynes (1967) we know that
the frequency of human finds was 11 sites before 12 000 BP, ten in the
following millennium, 15 in the next, 26 in the next, and 31 in the
millennium 9000–8000 BP. Megafaunal extinction spans from about 14 000–
8000 BP (Kurtén & Anderson 1980). Thus human colonization of North
America is inversely related to the presence of the megafauna, with human
sites increasing after megafaunal collapse. Human sites keep pace with
megafaunal extinction. This is not expected if humans came and conquered.
Occupation should not pace extinction, but should shoot up following
colonization. It does not. The pace of human colonization and megafaunal
extinction is nearly perfectly correlated over 6000 years, and the rate of
increase in human sites is extremely slow (assuming occupation sites from
13 000–8000 BP are as likely to be found) (Table 25.1).

Table 25.1 Human sites and megafaunal extinctions 14 000–8000 years BP

Human sites, time to doubling approx. 3000 years, annual growth of human population 0.69/3000 ×
100 = 0.02% per annum.
Percentage human sites vs. percentage species alive, r = 0.956, r2 = 0.914, t = 6.5, P < 0.01.

One site only, that from Lubbock Lake in Texas, indicates that humans
met the large Arctodus (Fig. 25.1). The site is dated about 11 100 BP
(Johnson & Shipman in press; the 12 650 ± 350 BP date given by Anderson
[1984] is in error). The remains indicate human cut-marks. Whatever the fate
ofthat bear, and the remains indicate a large specimen (Johnson pers.
comm.), humans and Arctodus did co-exist for a few centuries in lower
North America. At that time the domestic dog was already in America
(Anderson 1984), which has some bearing on how humans and Arctodus
could have co-existed.

Human expansion along with other Siberian species

Humans colonized North America as a late part of a Siberian megafauna,

members of which had existed in Alaska for hundreds of thousands of years.
With megafaunal collapse, expanding populations of Siberian colonizers
mixed with expanding populations of a few, eurytopic American species to
form the new, Holocene American fauna.
 Figure 25.1 A comparison of approximate body sizes of Arctodus simus (crosshatched) and Ursus
arctos horribilis. Large, male grizzly bears measure about 1.2 m in shoulder height. Mounted
skeletons of California grizzly and Arctodus from Rancho La Brea measure about 1.0 m and 1.3 m
respectively (Stock 1953). In Alaska and the Yukon Arctodus reached gigantic proportions (Kurtén &
Anderson 1980).

Humans entered lower North America along with grizzly bears (Ursus
arctos), moose (Alces alces), caribou (Rangijer), and glutton (Gulo luscus).
The wapiti (Cervus elaphus canadensis) and extant Bison bison are probably
both recent Siberians. I cannot find a single unequivocal and soundly dated
find of wapiti prior to megafaunal extinction in lower North America. A
Sangamon find turned out to be a jaw from a young bison; so called Cervus
teeth are not acceptable evidence in the absence of more diagnostic parts
(see Churcher 1984). The North American bison is today a Holocene dwarf
(Wilson 1980), and most likely of recent Alaskan origin; one would expect
Bison to colonize along with wapiti and thinhorn sheep (Ovis dalli). Prior to
about 11 000 BP Bison b. antiquus is found; it is quite different from B. b.
occidentalis /bison found after that date (Wilson & Churcher 1984). Iso-
enzyme and DNA analyses show that all wapiti in North America are not
only closely related, but also closely related to wapiti from the Altai in
Siberia. If the hypotheses pertaining to wapiti and bison are valid, then
bison, be they the plains or woodland forms, should be virtually identical.
 Another Siberian species that expanded with megafaunal extinction is the
timber wolf (Canis lupus), which may well have colonized along glacial
margins in late glacial times (see Martin & Gilbert 1978). This species
appears sporadically in the Rancholabrean tar pits (Kurtén & Anderson
1980), but expanded roughly with the demise of Canis dirus, the large ‘dire
wolf. An old Siberian with limited success during the Rancholabrean, the
bighorn sheep (Ovis canadensis) expanded just before megafaunal extinction
(17 000–14 000 BP, Geist 1985).
As the Rancholabrean specialists vanished, so primitive members of the
fauna expanded in range and numbers (i.e. Odocileus sp., Tayassu tajacu,
Antilocapra americana, and Ursus americanus). Human colonization was
part of a larger faunal event that shaped today’s impoverished, almost
certainly poorly adapted fauna of North America (see Geist 1985).
However, if humans only hastened the extinction of America’s
Rancholabrean fauna, and if the findings of earlier entries by humans into
America are valid, then we must ask what kept the Beringian or Siberian
fauna out of lower North America? Even if the earlier dates for humans are
disputed, there is no dispute that wapiti arrived in Alaska as early as the
Irvingtonian (Guthrie & Matthews 1971) and existed at least sporadically in
Alaska until late glacial times (Guthrie 1966, Kurtén & Anderson 1980).
Brown bears existed in Alaska in the Wisconsinian period, but not in lower
North America (Kurtén & Anderson 1980); they did exist in Alaska along
with Arctodus, maybe via ecological segregation since Arctodus finds are in
the unglaciated areas. Post-glacially, the brown bear occupies the range once
held by Arctodus simus. The Wisconsinian fauna of Beringia contains other
Siberian species which appear in lower North America only in late
Wisconsinian/Holocene times (i.e. Gulo gulo, Alces alces, Rangifer
tarandus). However, the Alaskan fauna does contain a good number of
typical Rancholabrean species (see Kurtén & Anderson 1980).
Sporadically, Siberian species did gain entry during mid-Pleistocene times
into lower North America (i.e. Panthera leo atrox, Bison latifrons,
Oreamnos sp.), while elephants (Mammutus sp.) and black bear (Ursus
americanus) came in early Pleistocene times. These became successful
members of the Rancholabrean fauna; others did not. The dhole (Cuon
alpinus) came and went extinct (Kurten & Anderson 1980). The bighorn
sheep (Ovis canadensis) lingered a long time before becoming successful
(see Geist 1985).
On the Siberian side of Beringia Upper Palaeolithic people appear about
14 000 BP (Vereshagin 1967, Rouse 1976) and shortly thereafter in America
(Haynes 1967). In Europe, the first expansion of Upper Palaeolithic people,
40 000–35 000 BP, coincides with a glacial recession and a brief interstadial
(Geist 1978). Members of this early dispersion could have reached Beringia
before the last glacial advance, but if so, they failed to establish themselves.

Human anti-predator strategies

At this point it is useful to turn to a subject addressed by very few

investigators. How do humans prevent themselves from becoming prey?
What did we do in our long evolutionary history to avoid predation? In
Australopithecus, arboreal adaptations indicate that it probably depended on
climbing abilities, particularly when escaping from a predator. In dire straits,
humans climb, a throw-back to ancient adaptations.
Since little remains of arboreal adaptations in Homo’s morphology,
excepting maybe in enigmatic Neanderthals who hardly used that capability
for climbing (Geist 1981), we suspect that anti-predator strategies other than
those of Australopithecus were employed. How does one safely spend the
night on the ground without fire, guns, or axe? Kortland (1980) provides an
answer: in Africa, use a packaging of thorns! Lions will not penetrate
through thorns to get at prey. In short, Homo erectus, using a thorn-covered
ground-nest, a corral of thorns, could spend the night safely away from trees.
And that method is still employed to this day (Post 1974). It can be
improved by fire, as well as the use of a stabbing weapon, and probably the
use of a loud voice that mimics a predator’s threat call (Geist 1978). From
the perspective of a predator, being confronted by a visual wall, behind
which sounds a peculiar, angry conspecific, mixed with incongruous smells
and sounds may be awe-inspiring. Maybe such suffices even for a thin wall
without thorns, but we do not know. Our safety currently relies, in the wild,
on destruction of carnivores that kill humans, so that the habit of eating
humans cannot be passed on. That is an important point: if food habits are
indeed initiated as a tradition, then stopping that tradition is important to
human safety. We do kill man-eating carnivores (Corbett 1946) and we may
well have acted like that in the past if the behaviour of native Americans is
an indication (Storer & Tevis 1955), provided, of course, we could kill
carnivores at an acceptable cost in injuries and lives. The upper limit in this
ability may well have lain with the brown bears (and polar bears – provided
sufficient dogs were at hand).

Bears and native North Americans

The relationship of the great brown and polar bears to humans that coexisted
with them has been one of unease. A study of interactions between bears and
men such as undertaken by Herrero (1985), or as one can pursue oneself in
such accounts as are available (i.e. Wright 1909, Storer & Tevis 1955,
Nelson 1969, 1973, Kurtén 1976, Craighead 1979, Young & Beyers 1980),
or in reflecting on one’s own encounters with grizzlies, allows one to
unbraid certain patterns in the behaviour of grizzlies. One also notes that in
pristine North America, grizzly bears were locally very common west of the
Mississippi. And it was indeed possible, in season, to travel along rivers and
see hundreds of bears, as did Pattie (1831) along the Arkansas River. He
counted 220 in one day, eight of which his party killed as these bears
attacked, while two more were killed in camp that night. In spring these
rivers were lined with the carcasses of drowned buffalo, but also elk (see
Roe 1951, Prince Maximilian von Wiede, in Thomas & Ronnefeldt 1982)
which was one reason why grizzly bears would be so numerous along rivers.
Travellers in early California reported seeing 30–50 bears per day (Storer &
Tevis 1955). In 1871 alone, 750 grizzly-bear hides were bought by the
Hudson’s Bay Company in the Cypress Hills of what came to be southern
Alberta: an area of only 2500 km2 (Ondrack 1985).
That native people co-existed with the great bears, often tenuously, is
amply recorded from the American west (Storer & Tevis 1955). The
evidence was in the form of some natives killed (David Thompson 1787, in
Hopwood 1971) and many natives and some voyagers with lacerated,
deformed faces, missing ears, nose and eyes as reported by Ross (1831) in
his journals from British Columbia, and others (Storer & Tevis 1955). There
is further evidence to be seen in the awe in which grizzlies were held by
natives, their reluctance to hunt the bears then as now, the honours that
accrued to one who had killed grizzlies, the fact that a grizzly was not
usually attacked except by going ceremoniously on a war-path against it,
with all the preparations such a decision entailed, and never except in
company of 4–10 warriors, or – depending on tribe and locality – after great
ambush preparation, council, arid shamanistic rites (Wright 1909, Storer &
Tevis 1955, Prince Maximilian von Wiede 1833, in Thomas and Ronnefeldt
1982). Californian Indians avoided good bear habitat as much as possible.
Some would not allow young men to go hunting alone, for fear of bears;
they delighted in the fact that the Spaniards had the capability to readily kill
grizzly and expressed their pleasure at this with gifts; they rapidly colonized
land cleared of grizzly bears by Spanish soldiers; some tribes would not hunt
grizzlies or trap them but readily sought the aid of white hunters to do the
task for them; tribes might join to avenge the killing of one of their members
by a grizzly, hold dance festivals when a grizzly was killed or even erect a
cairn on the spot where a grizzly expired at their hands. Yet at least to one
tribe, the Monos, an offshoot of the Shoshoneas, grizzlies were beasts to be
hunted and killed, a task at which they excelled (Storer & Tevis 1955).
However, in aboriginal times California was very sparsely settled by Indians
(Kroeber, in Storer & Tevis 1955), a reflection possibly of the fact that
grizzlies were not only avoided, but were also direct competitors of man for
food.
The central valleys of California had the right characteristics to sustain an
advanced agricultural culture, similar to the Aztecs. These valleys contained
fertile, flooding rivers traversing open landscapes, essential ecological
determinants of all great civilizations, occidental, oriental, or American
(Carneiro 1970). Today’s California is a densely settled land with thriving
agriculture. Yet the Spanish found it a land with a low native population, but
many large grizzly bears (Storer & Tevis 1955).
The accounts of many witnesses indicate that the great brown bears were,
in all but exceptional cases, avoiding interactions with humans, excepting
the female with cubs if surprised, or the occasional starving bear, be it in
Spring after hibernation, in Fall after berry crop failures, from debilitating
old age, or the occasional wounded bear (Nelson 1973, Young & Beyers
1980, Herrero 1985). The manner in which the bears acted indicated that
they did not regard humans as prey, although they might investigate humans
and their activity probably from sheer curiosity, a curiosity which some
observers misinterpreted as attacks. When wounded or provoked and
attacking humans, the brown bears treated humans as if they were some kind
of bear. That is, they tended to rise on their hind legs – as against a
conspecific – and tried to ‘disarm him’. To disarm a conspecific, a grizzly
bear bites and holds the opponent’s snout, disabling him from biting. They
do the same to humans, crushing face and jaw. Many bears after punishing
with bites left the offending humans. A few covered him with branches, as if
to ‘cache him’. Since bears do eat conspecifics, it does not surprise me that
an occasional bear prepares a human, whom he would treat like a
conspecific, for food. Standing still and upright, facing the grizzly, is
advocated by the Kutchin people of sub-Arctic Alaska (Nelson 1969) and by
Zenas Leonard (1804–57) a mountain man and trapper (in Froncek 1974, p.
280), which implies that the lack of flight may intimidate the grizzly; flight
provokes attack. Once aggressively pursued by humans, the great brown
bear becomes very secretive and, except for his signs, virtually invisible. The
accounts of the stories of the stock-killing bears that inhabited the west
between about 1880–1920 present tragic evidence of this (Young & Beyers
1980). The brown bear gives every indication of willingness to avoid and
learn, as if it had been second to some larger bear in the past, and indeed it
had, to the cave bear, Ursus spelaeus in Europe, and to the great Arctodus
simus in Alaska.
Killing a bear safely with weapons tipped with stone or bone points is a
very difficult task. The first technical problem is that flint and obsidian
points on arrows or spears shatter when they hit a bone, while bone points
are likely to chip and also fail to penetrate (see David Thompson in
Hopwood 1971, pp. 193–293, Guthrie 1983). Stone points do cut very well
through soft tissues, as good as iron points or better, but if the projectile is
aimed at the heart, it is not at all certain to reach it. From the front, the heart
and lungs of a large mammal are so well protected that a projectile would
strike bone 90 per cent of the time; from the side, it is still some 50 per cent
or so protected (Fig. 25.2). Half the spears thrown will not penetrate to their
mark – if such is the heart.
If arrows or throwing-spears are the prime weapons, then even if one does
penetrate the animal’s chest, the narrow wound channel of such a weapon is
not likely to disable a bear, nor kill it quickly. In fact, grizzly bears wounded
with narrow cavalry lances remain long capable of sustained attack and die
very slowly. Coronado’s soldiers lanced a grizzly, pushing the shaft to half
its length into a bear. This bear still caught the rider’s horse, and while
mauling it was run through with a second lance, after which it was
apparently lassoed and finally dispatched (Thomas 1935). A similar incident,
in which three Mexicans lanced a grizzly, was reported by Lieutenant Z. M.
Pike in 1808 (Wright 1909, p. 33); two of the Mexicans were killed by the
grizzly and the third was wounded.
Captain Clark of the famous Lewis and Clark expedition used a flintlock
rifle on grizzly bears. His largest bear received five balls through the lungs
and five balls in other parts of the body, and took 25 minutes to die after this
wounding (Lewis 1804–1806). Tales such as this are legion (Wright 1909,
Young & Beyers 1980). A projectile through the brain or the spinal chord
will anchor a grizzly, but not a body wound inflicted by lance, arrow, knife,
or small-bore, low velocity rifles. The broad-bladed, thick, hand-held boar-
spear of medieval European hunters, with a blade about 10–12 cm wide and
2–4 cm thick, mounted on an eight-foot pole, cross-wound with leather, is a
weapon used on bears. Such bears were distracted by a pack of hounds and
then speared by a team of hunters. That even so formidable a weapon as a
boar-spear, which opens a wound 12 cm wide by about 40 cm in depth does
not kill instantly, is attested to by the damage the speared animal inflicted on
the shafts of hunting spears exhibited in European hunting museums today.
A group of determined men, accompanied by brave and equally determined
large dogs, and armed with stout, broad lances could probably dispatch
brown bears if not in safety, then with reduced risk. Inuits in pre-firearm
days apparently did just that, but still considered such hunting very
dangerous (Nelson 1969). Indians in the west expected injury and death on
grizzly hunts (Lewis 1804–1806, Wright 1909). They did have lances as well
as dogs, but I cannot find whether they used the latter; Indian people in
Alaska do use dogs to warn of bears (Nelson 1973).
 Figure 25.2 Distribution of bones in a large mammal, in lateral and frontal view, as illustrated by a
domestic horse (after Ellenberg et al. 1956). The lungs and heart are well protected by bone. Few light
throwing-spears are expected to reach vital organs from the side, let alone from the front.

The experimental work on the performance of reconstructed throwing-

spears tipped with bone points fashioned after late Palaeolithic material by
Guthrie (1983) shows both an appalling lack of penetration on moose
carcasses, and the great fragility of the bone tips. Upper Palaeolithic hunters
used light throwing-spears on bears, as revealed in cave art (Kurtén 1976),
but the bear they hunted was small compared to the grizzly, and possibly
more timid, judging from what is known of today’s European brown bears.
Kurtén (1976) studied Pleistocene bear remains in great detail, searching
among other things for evidence of interactions between men and bears. He
concluded that whereas the European brown bear had been hunted by
Neanderthal and Upper Palaeolithic hunters, no evidence exists that the
much larger cave bear had shared a similar fate. Kurtén (1976) was critical
of earlier interpretations of cave-bear remains, showing how such
misinterpretations arose. It appears that Palaeolithic hunters stayed away
from bears larger than brown bears, that is, from cave bears.

The Arctodus problem

Had humans reached North America when it had an intact Rancholabrean
fauna, they would have met, among a number of large carnivores, the large
Arctodus simus. Large specimens of this bear, such as were typical of Alaska
or the Yukon (Kurtén & Anderson 1980) dwarfed even the giant among
extinct brown bears, the large Alaskan coastal brown bear (Kurtén 1976).
Skeletons of Arctodus simus from California are a foot taller at the shoulders
than the skeletons of large California grizzly bears (Stock 1953). Moreover,
Arctodus simus differed in several important respects from the grizzly bear:
it was carnivorous. Arctodus simus possessed large, functional carnassials, a
snout which shortened in evolution into a tool for grabbing and holding onto
prey, and a cursorial body form. It had long legs and was yet massive in
build as behoved a ‘bulldog bear’ (Kurtén & Anderson 1980).
Also, Arctodus apparently differed behaviourally from brown bears: like
sabre-toothed cats, dire wolf, and American cheetah, it is found in relatively
high numbers in natural traps, in which its prey also came to grief. Such we
find in Natural Trap cave in Wyoming (Martin & Gilbert 1978), and in the
tarpits of Rancho La Brea (Stock 1953). One does not find similar
depositions of brown bears or black bears, a possible suggestion that
Arctodus simus was more ready to go after prey or carrion in chance
circumstances that would tempt neither the brown nor the black bear. This
also means that, as the largest Rancholabrean carnivore, it was probably
much more ready to approach potential prey, the kills of other carnivores or
of human hunters. Naive polar bears do as much, and even hunted ones may
become belligerent if hungry (see Nelson 1969).
The distribution of grizzly in America overlaps in good part that of the
extinct Arctodus; grizzly bears expanded into lower North America with the
extinction of Arctodus, although in Alaska they apparently were
contemporaneous (Kurtén & Anderson 1980). Lower North America proved
to be superb grizzly bear habitat. Since grizzly exclude black bears in open
plains, except where the latter have recourse to trees, one has reason to
suspect that Arctodus excluded grizzly bears, except in the coldest
periglacial regions. The grizzly could not outclimb Arctodus, and in areas
with few trees, neither could Upper Palaeolithic hunters. Such hunters,
facing an approaching Arctodus, an eager, carnivorous bear twice the mass
of a grizzly, would be faced with a difficult problem. Their weaponry, good
for caribou, would not have been of much use, but would only enrage the
giant bear. There is no evidence for broad lances in the Beringian late glacial
record, and one questions the use of narrow-bladed spears in view of
Cornados’ experience of lancing bears. Whether dogs would have been an
aid in dispatching this giant bear is unknown, but dogs probably were
available to Upper Palaeolithic hunters arriving in Siberia 14 000 BP (see
Kurtén & Anderson 1980).
That predation on America’s megafauna was historically quite heavy is
indicated by several factors: the diversity of predators was high, and they
were individually much larger than their Eurasian or African counterparts.
The prey grew large and cursorial, and early horned and antlered artiodactyls
grew huge horn-like organs. The best examples of the latter are Bison
latifrons, Cervalces, and Ovis canadensis. Large horns and antlers are a
function of the security strategy for the neonate: if the young must grow
rapidly to survivable size then the female must produce a milk rich in solids,
that is she must be capable of saving a large fraction of nutrients from
growth towards production. In males this capability is reflected in large
antler-size, which is proportional to the female’s perinatal investment (Geist
1986, 1987). Ovis canadensis in its expansion, following the peak of the
Wisconsinian maximum, was closely associated with both Rancholabrean
predators and Siberian ones (Martin & Gilbert 1978). In Bison latifrons, the
relatively small molars and premolars indicate a deviation from grazing
towards softer foods. Grasses have a lower protein and mineral
concentration than foliages (Vogt 1948), and small teeth suggest foliage and
forbes feeding. That is, high milk solid production requires ‘concentrate
feeding’ (Hoffmann 1973). Such a diet is also indicated in the dentition of
Cervalces, which carried antlers rivalling the largest antlers of the Alaskan
bull moose. In short, predation appears to have been heavy in the
Rancholabrean megafauna, and I suggest that Siberian hunters moving into
Alaska would have found the large carnivores an insurmountable foe.

References
Anderson, E. 1984. Who’s who in the Pleistocene: a mammalian bestiary. In Pleistocene extinctions, P.
S. Martin & R. G. Klein (eds), 40–89. Tucson: University of Arizona Press.
Bishoff, G. L. & R. G. Rosenbauer 1981. Uranium series dating of human skeletal remains from the
Del Mar and Sunnyvale Site, California. Science 213, 1003–5.
Bryan, A. L. 1973. Paleo-environments and cultural diversity in late Pleistocene South America.
Quaternary Research 3, 237–56.
Carneiro, R. L. 1970. A theory of the origin of state. Science 169, 733–8.
Churcher, C. S. 1984. Sangamona: the fugitive deer. In Contributions in Quaternary Vertebrate
Paleontology, H. H. Grenoway & M. R. Dawson (eds), 316–31. Pittsburgh: Carnegie Museum of
Natural History, Special Publications No. 8.
Corbett, G. 1946. Man-eaters of Kumaon. Oxford: Oxford University Press.
Craighead, F. C. Jr. 1979. Track of the Grizzly. San Francisco: Sierra Club Books.
Cushing, J., M. Daily, E. Noble, V. L. Royh & A. Wenner 1984. Fossil mammoth from Santa Cruz
Island California. Quaternary Research 21, 376–84.
Edwards, W. E. 1967. The late Pleistocene extinction and diminution in size of many mammalian
species. In Pleistocene extinctions, P. S. Martin & H. E. Wright Jr. (eds), 141–154. New Haven:
Yale University Press.
Ellenberg, W., H. Bau & H. Dittrich 1956. In An atlas of animal anatomy for artists, L. S. Brown
(ed.). New York: Dover.
Froncek, T. (ed.) 1974. Voices from the wilderness. New York: McGraw–Hill.
Geist, V. 1978. Life strategies, human evolution, environmental design. New York: Springer.
Geist, V. 1981. Neanderthal the hunter. Natural History 90, 26–36.
Geist, V. 1985. Pleistocene bighorn sheep; some problems of adaptation, and its relevance to today’s
American megafauna. Wildlife Society Bulletin 33, 351–9.
Geist, V. 1986. The paradox of the great Irish stags. Natural History 95 (3), 54–64.
Geist, V. 1987. On speciation in Ice Age mammals with special reference to deer and sheep. Canadian
Journal of Zoology 65, 1067–84.
Gould, S. G. 1974. The origin and function of ‘bizarre’ structures: antler size and skull size in ‘Irish
elk’ Megaloceros giganteus. Evolution 28, 221–31.
Guthrie, R. D. 1966. The extinct wapiti of Alaska and Yukon Territory. Canadian Journal of Zoology
44, 45–7.
Guthrie, R. D. 1983. Osseous projectile points: biological observations affecting raw material
selection and design among Paleolithic and Paleoindian peoples. In Animals and archaeology. Vol.
1: Hunters and their prey, J. Clutton–Brock & C. Grigson (eds), 273–294. Oxford: BAR
International Series 163.
Guthrie, R. D. 1984. Mosaics, allochemicals and nutrients. In Quaternary extinctions, P. S. Martin &
R. G. Klein (eds), 259–98. Tucson: University of Arizona Press.
Guthrie, R. D. & J. V. Matthews 1971. The Cape Deceit fauna – early Pleistocene mammalian
assemblage from the Alaskan Arctic. Quaternary Research 1, 474–510.
Haynes, C. V. 1967. Carbon-14 dates and early man in the New World. In Pleistocene extinctions, P. S.
Martin & H. E. Wright Jr. (eds), 267–86. New Haven: Yale University Press.
Herrero, S. 1985. Bear attacks. New York: Lyons Books, Winchester Press.
Hoffman, R. R. 1973. The ruminant stomach. East African Monographs in Biology. 2, 1–354.
Hopwood, V. G. (ed.) 1971. David Thompson. Travels in western North America, 1784–1812.
Toronto: Macmillan of Canada.
Johnson, E. & P. Shipman in press. Scanning electron microscope studies of bone modification.
Current Research in the Pleistocene 3.
King, Y. E. & J. J. Saunders 1984. Environmental insularity and the extinction of the American
mastodon. In Quaternary extinctions, P. S. Martin & R. G. Klein (eds), 315–339. Tucson:
University of Arizona Press.
Kortland, A. 1980. How might early hominids have defended themselves against large predators and
food competitors? Journal of Human Evolution 9, 79–112.
Kurtén, B. 1976. The cave bear story. New York: Columbia University Press.
Kurtén, B. & E. Anderson 1980. Pleistocene mammals of North America. New York: Columbia
University Press.
Lewis, M. 1804–1806. Reprinted 1980 in Man meets grizzly, F. H. Young & C. Beyers (eds). Boston:
Houghton Mifflin.
Martin, L. D. & B. M. Gilbert 1978. Excavations at Natural Trap Cave. Transactions of the Nebraska
Academy of Science 6, 107–16.
Martin, P. S. 1967. Prehistoric overkill. In Pleistocene extinctions, P. S. Martin & H. E. Wright Jr.
(eds), 75–120. New Haven: Yale University Press.
Martin, P. S. 1973. The discovery of America. Science 179, 969–74.
Martin, P. S. 1984. Prehistoric overkill: the global model. In Quaternary extinctions, P. S. Martin & R.
G. Klein (eds), 354–403. Tucson: University of Arizona Press.
Martin, P. S. & R. G. Klein (eds) 1984. Quaternary extinctions. Tucson: The University of Arizona
Press.
Nelson, R. K. 1969. Hunters of the northern ice. Chicago: University of Chicago Press.
Nelson, R. K. 1973. Hunters of the northern forest. Chicago: University of Chicago Press.
Ondrack, J. 1985. Big game hunting in Alberta. Edmonton: Wildlife Publishing.
Pattie, O. G. 1831. Reprinted 1980 in Man meets grizzly, F. H. Young & C. Beyers (eds). Boston:
Houghton Mifflin.
Post van der, L. 1974. A story like the wind. London: Penguin.
Roe, F. G. 1951 (2nd edn 1970). The North American buffalo. Toronto: University of Toronto Press.
Rogers, R. A. 1985. Glacial geography and native North American languages. Quaternary Research
23, 130–7.
Ross, C. 1831. Reprinted 1980 in Man meets grizzly, F. H. Young & C. Beyers (eds). Boston:
Houghton Mifflin.
Rouse, J. 1976. Peopling of Americas. Quaternary Research 6, 597–612.
Ruddman, W. F. & J. C. Duplessy 1985. Conference on the last deglaciation: timing and mechanism.
Quaternary Research 23, 1–17.
Stanford, D., R. Bonnichsen & R. E. Morlun 1981. The Ginsberg experiment: modern and prehistoric
evidence of a bone-flaking technology. Science 212, 438–40.
Stock, C. 1953. Rancho La Brea. Los Angeles County Museum, Series no. 20. Paleontology no. 11
(5th edn).
Storer, T. J. & L. P. Tevis Jr. 1955. California grizzly. New York: Promontory Press.
Thomas, A. B. 1935. Reprinted 1980 in Man meets grizzly, F. H. Young & C. Beyers (eds). Boston:
Houghton Mifflin.
Thomas, D. & K. Ronnefeldt (eds) 1982. People of the first man. New York: Promontory Press.
Vereshagin, N. K. 1967. Primitive hunters and Pleistocene extinctions in the Soviet Union. In
Pleistocene extinctions, 369–98. New Haven: Yale University Press.
Vogt, F. 1948. Das Rotwild. Vienna: Osterreichischer Jagd und Fischerei Verlag.
Wilson, M. 1980. Morphological dating of late Quaternary bison on the northern plains. Canadian
Journal of Anthropology 1, 81–5.
Wilson, M. C. & C. S. Churcher 1984. The late Pleistocene Bighill Creek formation and its
equivalents in Alberta: correlative potential and vertebrate palaeofauna. In Correlation of
Quaternary chronologies, 150–75. Norwich: Geo Books.
Wright, W. H. 1909. The grizzly bear. Reprinted 1977, Lincoln: University of Nebraska Press.
Young, F. H. & C. Beyers (eds) 1980. Man meets grizzly. Boston: Houghton Mifflin.
26 Hunting in Pre-Columbian Panama: a
diachronic perspective
RICHARD G. COOK and ANTHONY J. RANERE

Introduction

Linares (1976) has demonstrated that even small-scale human modification

of the tropical rain forest enabled Pre-Columbian hunters to concentrate on
medium-sized caviomorph rodents and other mammal species, whose
biomasses were enhanced by horticulture and its attendant disturbances.
Berlin & Berlin (1983) report the same pattern of ‘garden-hunting’ among
the contemporary Aguaruna and Huambisa in the Peruvian ‘montaña’. In
this particular situation, the procurement of animal protein does not impinge
upon carbohydrate production; rather, the two are complementary.
Linares’ study site was Cerro Brujo, a small hamlet located on a steep–
sided peninsula on the northwestern Caribbean coast of Panama (Fig. 26.1).
Here year-round heavy rains permit continuous plant-cropping, but prevent
effective burning. The hamlet’s shifting cultivations could fallow their forest
plots for long periods of time, ensuring the replenishment of soils and of the
mammal populations (Linares 1976, 1980a, Linares & White 1980).
On the opposite (Pacific) side of the isthmus, the climate is very different.
Dry seasons are long. Around Parita Bay (Fig. 26.1), 4–6 months can pass
without effective precipitation. Evaporation is intensified by strong katabatic
winds. Dry periods during the rainy season can ruin crops. Uncontrolled
fires quickly destroy forest cover (Linares 1977, Cooke 1979).
In this region, human cultural development and demography have
followed patterns dissimilar to those of the Caribbean coast. The recently
surveyed Santa Maria drainage (Fig. 26.1) was settled by horticultural
peoples as early as Period 2B (Preceramic B: 5000–2500 BC) (see Cooke &
Ranere 1984 for a summary of the regional chronology). During this period,
human populations, though small and scattered, were distributed across all
the major ecozones from the Cordillera to the coast (Weiland 1984, 1985).
The cultivation of maize (Zea mays) and the intensive collection of tree
crops, particularly palm-nuts, are documented for some sites (Ranere &
Hansell 1978, Piperno 1984, Piperno & Clary 1984, Piperno et al. 1985).
By the beginning of the Christian era, permanent nucleated settlements,
for which maize was now a staple crop, are in evidence along the narrow but
fertile coastal plan (Hansell 1987). At the time the Cerro Brujo hamlet was
first occupied (about AD 600), these had become organized into antagonistic
village federations, which vied for prestige, scarce resources, and, probably,
arable land (Linares 1977, Cooke 1979, 1984a). Towns of over 1500 people
were visited by the invading Spanish between 1515 and 1529. Their
chronicles describe in detail the open and, in many parts, treeless habitats
adjacent to the population centres, such as Natá. They refer, too, to hunting
techniques appropriate for capturing the prevalent prey species, such as
netting doves (Columbidae) and driving deer (Cervidae) by firing the grass
(Andagoya 1913, p. 197, Espinosa 1913, p. 178; see also Cooke 1979,
1984b).
 Figure 26.1 Map of the Parita Bay region of central Panama, showing the location of 13
archaeological sites which have provided faunal materials. Inset: map of the Republic of Panama,
showing the location of Cerro Brujo and La Pitahaya, and the Tonosi valley.

How did faunal exploitation on the Pacific slopes differ from that of the
under-populated Caribbean coast? Is it possible to distinguish between
environmental and cultural influences, i.e. between the effects of a
seasonally arid environment and human social and hunting behaviour? Did
strategies develop to offset the depletion of terrestrial game resources? In
this chapter, we will address these questions in a preliminary form by
reference to samples of vertebrate bone (other than fish) from ten sites
located within 55 km of the present-day coastline of Parita Bay.

Chronological and contextual notes on the studied sites

The central Panamanian sites whose faunas we consider are listed in Table
26.1 together with information on dating, precipitation, and excavation and
recovery techniques. For comparative purposes, we also refer graphically
and textually to the published bone samples from Cerro Brujo and from La
Pitahaya, a large village located on the coast of Chiriqui (see Fig. 26.1),
which was occupied from about AD 500–1200 (Linares 1980b, Linares &
White 1980).
Four basic site types are represented in the Parita Bay area:

(a) three small, ‘one-family’ rockshelters: the Cueva de los Vampiros –

hereafter ‘Vampiros’ (Cooke & Ranere 1984), the Aguadulce Shelter –
hereafter ‘Aguadulce’ (Ranere & Hansell 1978), and Carabalí (Valerio 1985,
1987);
(b) one large, ‘multi-family’ rocksheiter: the Cueva de los Ladrones –
hereafter ‘Ladrones’ (Bird & Cooke 1978, Cooke 1984a),
(c) three coastal shellmounds, occupied by several families, but not
necessarily permanent settlements: Cerro Mangote (McGimsey 1956, Ranere
1980), Monagrillo (Willey & McGimsey 1954, Ranere & Hansell 1978), and
Zapotal (Willey & McGimsey 1954, Giausserand in prep.); and
(d) three large and permanent villages: La Mula-Sarigua (Willey &
McGimsey 1954, Hansell 1987), Sitio Sierra (Cooke 1972, 1979, 1984a, and
Natá (NA-8) (Cooke 1972).

Two of the small rockshelters – Vampiros and Carabalí – have light

occupations that date to Period 2A (Preceramic A: 8000–5000 BC). The only
non-fish vertebrate taxon that can be identified in the deposits that date to
this period is the nine-banded armadillo (Dasypus novemcinctus), at
Carabalí.
The shelters were most intensively utilized during Period 2B (Preceramic
B: 5000–2500 BC), Period 3A (Early Ceramic A: 2500–1000 BC) and Period
3B (Early Ceramic B: 1000–300 BC). Carabalí and Aguadulce (during Period
2B) were probably intermittently occupied hunting and gathering stations.
Ladrones and Aguadulce (during periods 3A and 3B) seem to have been
more permanent settlements for families growing maize and rootcrops in
neighbouring plots.
 Cerro Mangote is a shellmound overlooking Parita Bay. At the time of its
occupation (5000–3000 BC), it was between 3 and 1 km from the sea (Clary
et al. 1984, and references therein). The depth and composition of the refuse
and the presence of a cemetery (McGimsey et al. 1966) suggest that it was a
regular place of residence, being used by several families. The samples from
Ranere’s test pits comprise the basal red zone (‘R.Z.’) [4900–4000 BC] and
the overlying brown zone (‘B.Z.’) [4000–3000] (Ranere 1980).

Table 26.1 Parita Bay faunal samples

PSM members of the Proyecto Santa Maria (Cooke & Ranere 1984).

Monagrillo is a smaller shellmound than Cerro Mangote. It was used by

humans as early as 2400 BC, but witnessed its most intensive occupation
between about 1700 and 1000 BC (Ranere, quoted in Cooke 1984a, p. 273).
At this time, it was close to the active shore and at the mouth of a small river
(the Parita). The faunal sample was collected by Ranere in 1975 in two small
test pits and was analysed in preliminary fashion by E. Wing (Ranere &
Hansell 1978).
Zapotal seems to have been a temporarily or seasonally occupied
settlement located along an ancient beach line. A series of small and
apparently ephemeral dwellings are being excavated by Giausserand. The
site was occupied between about 2000 and 1500 BC. The faunal sample
reported here comes from a 2 × 1 m test pit dug in 1984. It has only been
partially analysed (only bones identifiable to the generic level are included
in the figures).
La Mula–Sarigua was a sizeable settlement, probably a permanently
occupied village, by the beginning of the 1st millennium BC. The sample
here presented was recovered in a 2 × 1 m test pit excavated at the site by R.
Cooke and A. Blanco in 1986. It has a single 14C date 220 ± 90 BC (Beta–
18863; shell, corrected for 13C/12C fractionation). At this time, the site was
on a low ridge overlooking the sea, at the north of the Parita river.
Sitio Sierra, located on a flood-free knoll 12 km upstream from the mouth
of the Santa Maria river, attained a maximum extent of 45 ha, probably
between AD 500 and AD 700. Excavations conducted here in 1975 identified
houses made of palm thatch and cane walls. The faunal samples we consider
come from house features and rubbish-dumps whose 14C dates fell between
65 ± 80 BC and AD 475 ± 110 (Cooke 1979).
The sample of nucleated settlements is completed by Natá (NA-8),
arguably the largest town on the Pacific watershed at the time of the Spanish
conquest (Cooke 1979). It is situated about 10 km from the coast, on the
banks of a small river (the Chico). The bone sample was recovered from a 2
× 1 m test trench excavated in 1970 (Cooke 1972). Associated ceramics
suggest a date of between AD 1300 and AD 1520.
The samples listed above give us information on the non-fish vertebrate
taxa that were either used for food and other domestic purposes or were
commensals discarded on middens. To complete the regional picture of
faunal utilization, we also offer brief comments on materials from grave
contexts which reflect the hunting of animals for ritual rather than diet.

Dietary contribution versus frequency of capture

In archaeofaunal analysis it is important to distinguish between dietary

contribution and frequency of capture. Recent compendia of archaeofaunal
techniques (e.g. Klein & Cruz-Uribe 1984) put undue emphasis on the use of
large mammals as food, and ignore smaller fare. It is true that many
Neotropical hunting groups that live close to productive forests acquire most
of their meat through the hunting, by males, of large- and mediumsized
mammals and birds (e.g. Hill & Hawkes 1983); smaller animals are simply
ignored. Where mammal biomasses are low, however, or where an
unpredictable environment leads to periodic shortages of terrestrial
organisms, small and apparently unpalatable fare (such as frogs and toads,
snakes, mud turtles, rats, and passerine birds) can be important dietary
elements, especially when they are seasonally abundant and/or gregarious.
Weiss’ recent work on the Campa of Peru has shown that this group
regularly consumes several species of toxic amphibians (Cooke in press).
Berlin & Berlin (1983, p. 306) report that 30 of the 85 species of reptiles and
amphibians that are recognized by the Aguaruna and Huambusa of Peru, are
actually eaten by them. In terms of the long-term dietary well-being of a
community, the regularity of animal meat and fat consumption is more
important than feasting on the products of occasional hunting trips (see
Hugh-Jones 1979, pp. 170–80 for a succinct discussion of the
unpredictability of male food acquisition versus the stability of female
production in tropical Colombia).
Figure 26.2 The relative frequency of reptiles, anurans, birds, and mammals in dietary faunal samples
from eight central Panamanian sites, calculated as percentage of skeletal elements (E), Body mass
(BM) and MNI (I). The figures are percentages of the non-fish vertebrate fauna only.
Totals are as follows:
CERRO MANGOTE R.Z.: E: 117, I: 17, BM: 2.5 kg
CERRO MANGOTE B.Z.: E: 953, I: 71, BM: 24.1 kg
LADRONES (P.2B & 3): E: 279, I: 13, BM: 4.8 kg
AGUADULCE (P. 2B): E: 818, I: 40, BM: 8.4 kg
AGUADULCE (P. 3): E: 1134, I: 47, BM: 9.0 kg
ZAPOTAL: E: 21, I: 10, BM: No data
CARABALÍ (P. 2B & 3): E: 272, I: 19, BM: No data
LA MULA: E: 778, I: ND, BM: 1.4 kg
SITIO SIERRA (one feature E: 1231, I: 105, BM: 53.2 kg
only):
NATÁ: E: 68, I: 17, BM: No data

In order to give a broad, if somewhat crude, impression of the discrepancy

between dietary contribution and frequency of capture in the Parita Bay
samples, we have presented, in Figure 26.2, the relative abundance of
reptiles, anurans, birds, and mammals in the non-fish vertebrate bone
samples at eight sites, expressed in terms of (a) number of skeletal elements
(E); (b) body mass (BM) (calculated by skeletal mass/body mass allometry,
Wing & Brown 1979); and (c) minimum number of individuals (I). All
fragments attributable to each class have been included, except in the case of
Zapotal where only bone identifiable to the generic level has been included.
The body mass conversions give the impression that mammal meat
represents a far more important contribution to the regional diet, at all
periods of time, than do reptiles, anurans, or birds. The minimum numbers of
individuals, on the other hand, indicate that these three groups together make
up a higher percentage of individuals taken than mammals, at the three sites
with the largest samples: Cerro Mangote, Aguadulce, and Sitio Sierra. In a
single large refuse-dump at Sitio Sierra, for example, reptiles, anurans, and
birds each make up over 20 per cent of the non-fish vertebrates. At
Aguadulce, the most frequently taken non-fish taxa are small turtles of the
genera Kinosternon and Chrysemys.

The dietary contributions of fish and other aquatic resources

Any evaluation of the harvesting of terrestrial organisms should take into

account the availability and productivity of fish, and of coastal and aquatic
resources in general. Fishing and the collection of molluscs and crustaceans
were undoubtedly profitable enough at several Parita Bay sites to have
affected the scheduling, duration, and geographical range of hunting trips.
Unfortunately, the data on the dietary contribution of shellfish and
crustaceans have not yet been fully quantified (though see Hansell 1979 for a
detailed analysis of the mollusc fauna at Monagrillo). In Figure 26.3, we
present the proportions of fish skeletal elements (E) and body mass (BM)
relative to those of the non-fish vertebrates. Quite understandably, the
general trend is for fishing to decline in importance the further one goes
from the coast. Thus, at Vampiros, the site nearest the coast, 99.6 per cent of
the elements are fish, while at Carabali, the site furthest from Parita Bay, the
value is a mere 5 per cent. This last site is located near several streams and a
small river (the Gatú); nevertheless, the cool, fast-flowing water supports a
low fish biomass and sustains few species (Cooke 1986).
The one site where the trend seems to be reversed is Sitio Sierra, located
near freshwater 12 km from Parita Bay. Here fish comprise 97 per cent of
the elements and 59 per cent of the body mass. By way of comparison, the
Preceramic B Cerro Mangote site, which was situated much nearer the active
shore, was less reliant upon fish: in the basal red zone, fish contribute 46 per
cent of the body mass.

Natural environments, habitat modification, and human

behaviour

We will now consider some aspects of the taxonomic composition of the

central Panamanian non-fish vertebrate fauna, in an attempt to identify the
distributions that might have been affected by natural environmental
conditions, those that imply anthropogenic disturbances of some kind, and
those that could have been influenced by less tangible cultural factors, such
as socially induced restrictions upon hunting, and weapons technology.
In Figure 26.4, we present the relative abundance of 12 mammalian taxa,
calculated according to their percentage minimum numbers of individuals
from seven Parita Bay sites, La Pitahaya, and Cerro Brujo. Figure 26.5 gives
the percentages often taxa in the herpetofaunal samples from three Parita
Bay sites and Cerro Brujo. Table 26.2 summarizes the avian fauna from
Cerro Mangote. (See Cooke 1984b for a discussion of the Sitio Sierra
avifauna and Cooke 1984a and Cooke & Ranere in press for complete
mammalian, avian, and herpetofaunal species lists.)

The white-tailed deer (Odocoileus virginianus)

The white-tailed deer is the only mammal species which occurs in all the
Panamanian samples, both Atlantic and Pacific. The archaeological
distributions show that it maintained its population levels along the Pacific
watershed, from Cerro Mangote’s occupation until the Spanish conquest,
thus confirming ethnohistoric sources, which indicate that it was still locally
abundant in the 16th century. At Cerro Brujo, it was taken less regularly (5.7
per cent of the MNI). The brocket deer (Mazama americana), which shuns
open conditions, is present at this site, but is absent from all the Pacific-side
samples.
Linares’ scenario of ‘garden-hunting’ could be applied to the Parita Bay
deer populations: in spite of rising human population densities (Weiland
1984) and increasing hunting pressure and deforestation, this species would
have been favoured by the exponential disturbance of the primary forest
cover (which it is unsuited to) and by the expansion of agricultural fields in
different stages of regrowth (Bennett 1968). No more complicated an
explanation may be necessary to explain the persistence of Odocoileus. In an
unpublished paper, however, Cooke (1978) suggested that some kind of
cultural management might have influenced the deer populations in the
Parita Bay region. There is documentary evidence for deer not being
consumed in one chiefdom (Parita),while in an adjacent territory (Natá)
large supplies of dried and salted deer meat in special storehouses were
encountered by the Spanish (Cooke 1979 and references therein).
 Figure 26.3 The proportions of fish to non-fish remains in the dietary vertebrate faunal samples of
eight central Panamanian sites, calculated as the percentage of the body mass and skeletal elements.
(Fish: the two bars to the left; non-fish: the two bars to the right.) The Sitio Sierra figures refer to the
same feature as that in Figure 26.2 (date: ± AD 300). The Ladrones and Carabalí samples combine
Period 2B and 3 levels.
 Figure 26.4 The relative frequency of mammalian taxa at seven Parita Bay sites, La Pitahaya, and
Cerro Brujo, expressed as the percentage of the mammalian MNI (I) calculated for each context. CM 1
(left): Cerro Mangote, 1979 excavations, three features; CM 2 (right): Cerro Mangote, 1956
excavations, all features; LA: Ladrones, Period 2B & 3 levels; AG 1 (left): Aguadulce, Period 2B
levels; AG 2 (right): Aguadulce, Period 3 levels; MO: Monagrillo; ZA: Zapotal; SS: Sitio Sierra, four
features; LP: La Pitahaya; NA: Natá; CB: Cerro Brujo. Total I values: CM 1 = 80, CM 2 = 74, LA =
10, AG 1 = 14, AG 2 = 15, MO = 9, ZA = 6, SS = 97, LP = 12, NA = 11, CB = 108. ‘Rat/mouse’
represents the following taxa: CM: Liomys adspersus, AG: L. adspersus, Zygodontomys brevicauda,
Oryzomys sp., Sigmodon hispidus; ZA: Z. brevicauda; SS: L adspersus, Z. brevicauda, O. cf. concolor;
NA: undetermined; CB: Hoplomys & Oryzomys.
 Figure 26.5 The relative frequency of ten taxa of reptiles and anurans from Cerro Mangote (CM),
Aguadulce (AG), Sitio Sierra (SS), and Cerro Brujo (CB), expressed as the percentage of MNI (I)
herpetofaunal samples only. CM 1: 1979 exc., all features; CM 2: Brown Zone, three features; AG 1:
 Period 2B levels; AG 2: Period 3 levels; SS: single feature (same as Figs 26.2 & 26.3); CB: all
features. Total I values: CM 1: 67, CM 2: 29, AG 1 = 23, AG 2 = 27, SS = 50, CB = 50.

In this competitive and politically unpredictable society, potlatching meat

supplies is understandable behaviour: social controls over hunting the one
available large mammal would protect stocks to ensure sufficient meat for
periodic feasts. Another explanation could be that, in Parita’s chiefdom, deer
were tabooed in deference to cognitive prohibitions that are common among
South American Amerindians today (Redford & Robinson 1987) and which
were, until recently, followed by the Kuna of eastern Panama.
In contrast to Odocoileus, the distribution of the other mammals in the
Parita Bay samples is more erratic. In some cases, a particular species’
abundance can be explained simply as a function of the availability of
habitats; in others, human predation would seem to be responsible for low
representations.
One species, for example, whose presence and absence would have been
conditioned by the proximity of a particular habitat, is the raccoon, Procyon
lotor, which is the second commonest mammal at Cerro Mangote, where it
represents over 25 per cent of the MNI. Raccoons are especially abundant in
Panama along the coast, where they roost in mangroves and feed on
mudflats and along beaches.
Other examples of ‘serendipitous’ hunting associated with local habitat
availability can be seen in the herpetofaunal and avian samples. Ctenosaura,
the black iguana, which in Panama prefers coastal habitats, is predictably
more frequent at Cerro Mangote than elsewhere. The bird taxa recorded at
this site (Table 26.2) can all be captured along the shore, in mangroves or in
xerophytic scrub woodland. At Sitio Sierra, the archaeological avifauna
includes several species, such as the white-tailed nightjar (Caprimulgus
cayennensis), the Aplomado falcon (Falco femoralis) and the crested bob
white (Colinus cristatus), whose present-day distribution throughout tropical
America coincides with seasonal aridity and open habitats (Cooke 1984b).
In the above cases, the tenets of optimal foraging can be applied to explain
the archaeological distributions: the Pre-Columbian hunters simply focused
on those taxa which were easiest to obtain near their settlements.

Table 26.2 The avifauna from Cerro Mangote (5000–3000 BC). The figures have been aggregated
 from 18 different contexts identified by Ranere in the 1979 excavations (identifications by S. Olson
and R. Cooke*)
Taxon English name 1 %I
Egretta c.f caerulea or tricolor little blue or tricoloured heron 1 4
Eudocimus albus white ibis 8 32
c.f Trinca melanoleuca* greater yellowlegs 1 4
Catoptrophorus semipalmatus willet 5 20
Calidris canutus knot 1 4
Calidris mauri or pusilla western or semipalmated sandpiper 1 4
Calidris sp. indet. +
c.f Numenius phaeopus whimbrel 1 4
Geotrygon montana ruddy quail-dove 1 4
Columbina talpacoti ruddy ground-dove 1 4
c.f Zenaida asiatica* white-winged dove 1 4
Amazona ochrocephala yellow-lored parrot 1 4
Passeriformes passerines 3 12
Total 25 100

Caviomorph rodents
The diurnal agouti (Dasyprocta punctata) is either very rare or absent in the
Parita Bay samples, except at Monagrillo. It is notoriously absent from the
large samples at Sitio Sierra. At Cerro Mangote it is represented by one
equivocal bone. The nocturnal paca (Cuniculus (Agouti) paca) is present at
more sites, though it exceeds 10 per cent of the mammalian minimum
number of individuals only at the Aguadulce Shelter.
Agoutis are most abundant in forests with plentiful supplies of fruiting
trees, such as Astrocaryum palms and Gustavia (Smythe 1978). On Barro
Colorado, they represent 6.1 per cent of the present-day mammalian
population (Glanz 1982, Table 2). Around Parita Bay, where dry seasons are
longer and more intense than on Barro Colorado, the agoutis’ preferred food
trees are scarce; hence this species may have had primevally low natural
populations in this region. The paca, on the other hand, withstands drier
conditions and can be very abundant in gallery forests. Linares (1976)
showed that at Cerro Brujo the abundance of these two species was
enhanced by the presence of man-made habitats associated with agricultural
plots. Their distribution in the Parita Bay samples does not coincide with this
pattern. It is likely, then, that human hunting pressure adversely affected
these species, with the agouti coming off worse due to its lower population
densities.

Peccaries
All the tayassuid bones that preserve distinguishing characters in the central
Panama samples are from the collared peccary (Tayassu tajaçu). This species
has one of the largest ranges of any living wild ungulate and will thrive in
habitats as different as primary rain forest and cactus scrub. Being a herd
animal, however, its home range is large (60–800 ha) and its food
requirements are prodigious. Its abundance, then, tends to be conditioned by
the availability of suitable fruiting plants and by the unbrokenness of cover
(Sowls 1984, Donkin 1985). At Monagrillo and Zapotal, both located to the
south of the Santa Maria river, T. tajaçu represents 22.2 per cent and 16.7 per
cent, respectively, of the mammalian minimum numbers of individuals. Just
north of the river, however, peccaries have a different distribution: they are
absent at Sitio Sierra and are very rare at Cerro Mangote. At Ladrones, T.
tajagu represents a quarter of the mammalian fauna. This site is situated at
the edge of the foothills, where forest cover probably survived longer than
on the coastal plain.
At Cerro Brujo, Tayassu represents 10.5 per cent of the minimum number
of individuals and can be considered the third species in Linares’ garden-
hunting scenario. Peccary bones are also well represented at Mayan sites
(Pohl & Feldman 1982 and references therein). The Miskito of Nicaragua
take more white-lipped peccaries (T. pecari) than any other mammal on their
long treks into the forest (Nietschmann 1973, p. 166). Various South
American hunter-gatherers and horticulturalists prey heavily upon peccaries
(Hames & Vickers 1983, Redford & Robinson 1985). Around Parita Bay, the
collared peccary probably suffered heavy hunting pressure nearest the largest
population centres and in areas where forest cover was least continuous.
Even so, the incongruencies between the distributions of this species north
and south of the Santa Maria river are difficult to explain.

Primates and other arboreal species

At all Panamanian sites, including Cerro Brujo, arboreal taxa are extremely
scarce in midden bone samples. Primates and sloths are absent. Some
facultatively arboreal species, like the coati (Nasua nasua), which can be
extremely abundant in forest-edge habitats, are also absent. Squirrels are
represented only at Cerro Mangote, Aguadulce, and Ladrones. The green
iguana (Iguana iguana), which spends a large part of the day browsing in
trees, is absent at Cerro Brujo where suitable habitats must have prevailed
near the hamlet. At Cerro Mangote, on the other hand, the green iguana
represents more than 20 per cent of the herpetofauna. Later in time, it
becomes proportionately scarcer around Parita Bay, probably in response to
human hunting pressure.

Dogs and technological limitations

Two aspects of hunting behaviour should be considered when the above
distributions are analysed: the use of dogs for hunting and the kinds of
hunting technologies that might have been employed by the inhabitants of
Parita Bay.
Pre-Columbian communities of the Neotropics possessed dogs, however
mangy and feeble they may have seemed to the hound-keeping Spanish
(Gilmore 1950). In Panama, canid post-cranial elements are very rarely
found in middens and can be usually attributed to the grey fox (Urocyon
cinereoargenteus), a native species that is common in semi-open areas
(Cooke 1984a). Their scarcity implies that they were not used as food, as
they were in Mexico (Wing 1978, Hamblin 1984). Nevertheless, large
quantities of dog teeth were used to make the necklaces that adorned some
Panamanian cadavers (see following section).
At Cerro Mangote, a large canid humerus recovered from the 4th
millennium BC ‘brown zone’ is probably from a hound-like dog (Cooke &
Ranere in press). It is difficult to imagine that dogs of this size were not used
for hunting. Certainly, hunting with dogs was practised by the Maya (Pohl &
Feldman 1982). The specialization of certain Central and South American
groups, such as the Boruca of Costa Rica and the Waiwai of Guyana, in the
breeding of hunting dogs for exchange suggests the continuation of a pre-
European custom (Fock 1963, pp. 5, 239, Fernández Guardia 1969, pp. 12–
13).
The use of such efficient predators from the Preceramic B onwards could
have had a drastic effect on mammalian and large reptile populations
(Diamond 1984). The agouti and iguana for example, are very susceptible to
dogs because of their diurnal habits and self-destructive escape strategies
(Cooke 1979, Smythe pers. comm.). Collared peccaries can be effectively
corralled by packs of dogs. In contrast, the fleet-footed white-tailed deer is
efficient at escaping from dogs in broken terrain where dense undergrowth is
available.
With regard to the scarcity of arboreal organisms, and especially primates,
we find it unlikely that the archaeological distributions are due solely to
habitat destruction or overhunting. Howler monkeys (Alouatta alouatta)
survived in forest tracts near the Santa Maria river until recently (Bennett
1968) and there is still a remnant population at the southern edge of the
watershed (Cooke, personal observation). Technological deficiencies may
have been partly responsible: blowguns, which are the forest hunter’s
weapon par excellence (Yost & Kelley 1983), might never have been used
for hunting in central and western Panama, making arboreal species less
worthwhile to hunt (Linares 1976). The warriors of the Parita Bay chiefdoms
fought the Spanish with shark-tooth studded wooden clubs, stones (with
slings?), and spears with spear-throwers. They were not renowned as
bowmen. Blowguns were used in Colonial times by the Talamancan groups
in the forests of eastern Costa Rica for hunting birds; but their projectiles
were clay pellets rather than wooden darts. The three species of highly toxic
Phyllobates frogs, whose poison is used by the Emberá Chocó in Colombia,
are not indigenous to Panama (Myers et al. 1978). Hence it is possible that a
lack of rapidly acting dart poisons curtailed the effectiveness of the blowgun
in prehistoric central Panama.

Exotic materials and exchange

The faunal samples from Parita Bay that we have considered so far come
from kitchen middens and house floors. Artifacts made of bone are rare and
are limited to deer metapodial needles – probably for net manufacture –
Ariid catfish-spine awls, and, at Natá, a single jaguar (Felis onca) polished
phalanx, which was probably used as an amulet (Cooke 1972).
The dietary and industrial use of animals, of course, only reflects a part of
the total human impact upon native faunas. To complete the picture, we
should consider, albeit briefly, possible ritual associations between animals
and humans, especially as the recorded distributions in burial and other
ceremonial sites are somewhat different from those found in domestic
contexts (for a more detailed discussion, see Cooke & Ranere in press). The
oldest burials, those of Cerro Mangote and Sitio Sierra’s early cemetery
(250–25 BC; Cooke 1979), contain only shell beads, sting-ray spines, and a
human skull (a trophy?). In the later, Sitio Sierra cemetery (c. AD 1100;
Cooke 1972, 1984b), the burial goods are no more extravagant and represent
locally obtained animals: shark teeth, a flute made from a brown pelican
(Pelecanus occidentalis) humerus, a scarlet macaw (Ara macao), a quail
skeleton, and the calcined tibia of a paca.
On the other hand, the animal materials associated with the burials at
Sittio Conte (Lothrop 1937) and other Pacific-side cemeteries (Ladd 1964,
Ichon 1980) emphasize the importance of exotic or scarce items as symbols
of social status (Linares 1977). Sitio Conte was used for high-rank
individuals between AD 500 and AD 900, for whom the tusks, the teeth,
bones, feathers, and skins of certain species were part of a metaphorical
attire that could not be easily obtained within the territorial confines of the
village.
Some males were adorned with necklaces made of hundreds of dog teeth
and peccary tusks. The low frequencies of Tayassu bones in the local dietary
samples and the sheer quantities of tusks used for the necklaces and aprons
at Sitio Conte suggest that peccary teeth were traded into the Pacific plains
from the forested regions of the interior. Manatee bone, which is ideal for
carving, must also have been imported: sea-cows have not been recorded on
the Pacific coast of tropical America during the Holocene. At Cerro Brujo,
they were commonly hunted (Linares & White 1980). Sea-turtle remains are
also rare in Parita Bay middens; few beaches appropriate for nesting are
available along this section of the coast. Hence, it is likely that the whole
shells placed in some graves were acquired outside the village territory.

Conclusions

In this chapter we have been careful not to stress a single pattern of causality
to explain the proportions of non-fish vertebrates in archaeological bone
samples from central Pacific Panama. Some of the samples have low counts
and should be interpreted with extreme caution (Grayson 1978). More
information is also needed from inland sites.
 We have noted that the medium and large mammal species that are
archetypically hunted by Neotropical forest peoples – such as tapirs
(Tapirus), capybaras (Hydrochoerus), white-lipped peccary (T. pecari),
spider- and howler monkeys (Ateles, Alouatta), and coatis (Nasua) – are
conspicuously absent. So too, are large forest birds like curassows, guans,
and toucans (Rhamphastidae).
We have followed Linares (1976) in suggesting that the lack of an
appropriate hunting technology – such as the blowgun – could account for
the absence of some arboreal taxa. In spite of this possible limitation,
however, the species composition of the samples suggests that the local
hunters did not operate in unbroken tracts of forest, nor did they make
periodic treks into distant sylvan habitats (perhaps because the constant
threat of abduction or ambush by hostile groups made it too dangerous for
them to do so). When exotic materials were needed to satisfy the whims of
the local élites, they were probably imported through the tribal kinship
networks from outlying villages.
The pattern of ‘garden-hunting’ described by Linares for the Caribbean
slopes is not as easy to identify around Parita Bay as at Cerro Brujo. It is true
that the white-tailed deer, naturally abundant in disturbed and semi-open
areas, must have benefitted from the clearing of the forest and the expansion
of agricultural plots. The erratic distributions of the caviomorph rodents and
the collared peccary, however, indicate that the beneficial symbiosis between
Pre-Columbian farmers and these prey species that was possible in the
wetter, less heavily populated areas of Panama, was not achieved around
Parita Bay, where the climate is seasonally arid and where human hunting
pressure was more prolonged and intense than on the Caribbean slopes. We
have suggested that these taxa may have been naturally scarce in this semi-
arid environment.
The consumption of species which have restricted environmental
requirements, such as raccoons and shore-birds, in addition to the frequent
capture of food of small and less palatable items, such as passerine birds,
amphibians, small lizards, and commensal mammals, vouch for an
opportunistic element in regional hunting and collecting patterns, and
suggest that women and children were responsible for acquiring certain
types of meat. We have also proposed (Cooke 1986, in press) that the
incorporation into the diet of toxic species like the toad Bufo marinus, and
improved fishing techniques for small shoaling species with high seasonal
biomasses, might have been internal responses to an increasingly
depauperate terrestrial mammal fauna. Nevertheless, when one bears in mind
the obvious productivity of the coastal biomes of Parita Bay, it is still
dangerous to assume that the procurement of animal protein was in crisis.
Caution is the better part of valour. We must await larger and geographically
more varied bone samples before we can be confident that we have
identified correctly the different cultural and environmental phenomena that
act upon archaeological faunal assemblages.

Acknowledgements
We would like to acknowledge the contributions of the following archaeozoologists and
palaeontologists to the analyses of the Parita Bay faunal samples: Storrs Olson (birds), Elizabeth Wing
(the initial study of the Sitio Sierra fish and the Monagrillo samples), James Berry (the Sitio Sierra
turtles), William Duellman and Arnold Kluge (anurans), Charles Handley (mammals, especially
teeth), and R. Medlock (the 1956 Cerro Mangote deer bone). Laboratory assistance has been provided
by: Marcela Camargo, Milton Collazos, Linda Cunningham, Julio Jaen, Gina Maduro, Annette Rolin,
Carlota Rios, and Aureliano Valencia. We are indebted to all for their constancy and patience. Finally,
we wish to thank Olga Linares, Stanley Rand, Nicholas Smythe, Neal Smith, and many other
colleagues at the Smithsonian Tropical Research Institute for their helpful comments about the
Neotropical fauna.

References
Andagoya, P. de 1913. Relación de los sucesos de Pedrarias Dávila, en las provincias de Tierra-Firme
y de lo ocurrido en el descubrimiento de la Mar del Sur-&-(fragment). In El Descubrimiento del
Océano Paafico: Vasco Núñez de Balboa, Hernando deMagallanes y sus Compañeros Vol. II, J. T.
Medina (ed.), 191–207. Santiago de Chile: Imprenta Universitaria.
Bennett, C. F. 1968. Human influences on the zoogeography of Panama. Berkeley: University of
California Press.
Berlin, B. & E. A. Berlin 1983. Adaptation and ethnozoological classification theoretical implications
of animal resources and diet of the Aguaruna and Huambisa. In Adaptive responses of native
Amazonians, R. B. Hames & W. T. Vickers (eds), 301–28. New York: Academic Press.
Bird, J. B. & R. G. Cooke 1978. La Cueva de los Ladrones: datos preliminares sobre la ocupación
formativa. Actas del V Simposium Nacional de Antropologia, Arqueologia y Etnohistoria de
Panama, 283–305.
Clary, J. H., A. J. Ranere & P. Hansell 1984. The Holocene geology of Parita Bay, Panama. In Recent
developments in isthmian archaeology, F. Lange (ed.), 55–83. Oxford: BAR International Series,
212.
Cooke, R. G. 1972. The archaeology of the western Code province of Panama. Unpublished PhD
dissertation, London University, London.
Cooke, R. G. 1978. Maximizing a valuable resource: the white-tailed deer in prehistoric central
Panama. Unpublished paper presented at the 44th Annual Meeting of the Society for American
Archaeology, Tucson.
Cooke, R. G. 1979. Los impactos de las comunidades agrícolas sobre los ambientes del Trópico
estacional: datos del Panama prehistórico. Actas del IV Simposium Internacional de Ecologia
Tropical, Vol. III, 919–73.
Cooke, R. G. 1984a. Archaeological research in central and eastern Panama: a review of some
problems. In The archaeology of lower Central America, F. Lange & D. Z. Stone (eds), 263–302.
Albuquerque: University of New Mexico Press.
Cooke, R. G. 1984b. Birds and men in prehistoric central Panama. In Recent developments in isthmian
archaeology, F. Lange (ed.), 243–81. Oxford: BAR International Series, 212.
Cooke, R. G. 1986. Some social and technological correlates of in-shore fishing in Formative Central
Panama. Unpublished paper presented at the 9th Annual Chac–Mool Conference, University of
Calgary, 7–9 November.
Cooke, R. G. in press. Anurans as food in tropical America. Archaeozoologica.
Cooke, R. G. & A. J. Ranere 1984. The ‘Proyecto Santa María’: a multidisciplinary analysis of
prehistoric human adaptations to a tropical watershed. In Recent developments in isthmian
archaeology, F. Lange (ed.), 3–30. Oxford: BAR International Series, 212.
Cooke, R. G. & A. J. Ranere in press. Precolumbian influences on the zoogeography of Panama: an
update. Proceedings of the International Symposium on the Zoogeography of Mesoamerica, Merida,
Yucatán. New Orleans: Tulane University.
Diamond, J. M. 1984. Historic extinctions: a Rosetta Stone for understanding prehistoric extinctions.
In Quaternary Extinctions, P. S. Martin & R. G. Klein (eds), 824–62. Tucson: University of Arizona
Press.
Donkin, R. A. 1985. The peccary – with observations on the introduction of pigs to the New World.
Transactions of the American Philosophical Society, 75.
Espinosa, G. de 1913. Relación hecha por Gaspar de Espinosa, alcalde mayor de Castilla de Oro, dada
à Pedrarias Dávilla. In El Descubrimiento del Océano Pacífico: Vasco Núñez de Balboa, Hernando
Magallanes y sus Compañeros, J. T. Medina (ed.), Vol. II, 154–83. Santiago de Chile: Imprenta
Universitaria.
Fernández Guardia, R. 1969. Reseña histórica de Talamanca, 2nd edn. San José de Costa Rica:
Imprenta Nacional.
Fock, Niels 1963. Waiwai: religion and society of an Amazonian tribe. National-museets Skrifter,
Etnogafisk Raekke, no. VIII. Copenhagen: National Museum.
Giausserand, M. In preparation. Excavations at the Zapotal site in central Panama. (Fieldwork for
PhD dissertation, Yale University).
Gilmore, R.M. 1950. Fauna and ethnozoology of South America. In Handbook of South American
Indians, 6, J. Steward (ed.) (Bulletin of the Bureau of American Ethnology, 143), 345–464.
Glanz, W. E. 1982. The terrestrial mammal fauna of Barro Colorado island: censuses and long-term
changes. In The ecology of a tropical forest: seasonal rhythms and long-term changes, E. Leigh, A.
S. Rand & D. Windsor (eds), 455–68. Washington DC: Smithsonian Institution Press.
Grayson, D. K. 1978. Minimum numbers and sample size in vertebrate faunal analysis. American
Antiquity 43, 53–65.
Hamblin, N. L. 1983. Animal use by the Cozumel Maya. Tucson: University of Arizona Press.
Hames, R. B. & W. T. Vickers (eds) 1983. Adaptive responses of native Amazonians. New York:
Academic Press.
Hansell, P. 1979. Shell analysis: a case study from Panama. Unpublished MSc dissertation,
Department of Anthropology, Temple University, Philadelphia.
Hansell, P. 1987. The Formative in central Pacific Panama: La Mula–Sarigua. In Chiefdoms in the
Americas, R. D. Drennan & C. A. Uribe (eds), 119–38. University Press of America.
Hill, K. & K. Hawkes 1983. Neotropical hunting among the Aché of eastern Paraguay. In Adaptive
responses of native Amazonians, R. B. Hames & W. T. Vickers (eds), 139–88. New York: Academic
Press.
Hugh-Jones, C. 1979. From the Milk River. Cambridge: Cambridge University Press.
Ichon, A. 1980. L’archéologiedu Sud de la Péninsule d’Azuero, Panama. Mission Archéologique
Française au Méxique, Série III, Mexico.
Klein, R. & K. Cruz-Uribe 1984. The analysis of animal bones from archaeological sites. Chicago:
University of Chicago Press.
Ladd, J. 1964. Archaeological investigations in the Parita and Santa Maria zones of Panama. Bulletin
of the Bureau of American Ethnology, 193.
Linares, O. F. 1976. Garden-hunting in the American tropics. Human Ecology 4, 331–49.
Linares, O. F. 1977. Ecology and the arts in ancient Panama. Studies in Pre-Columbian art and
archaeology, no. 17, Dumbarton Oaks.
Linares, O. F. 1980a. Ecology and prehistory of the Aguacate Peninsula in Bocas del Toro. In Adaptive
radiations in prehistoric Panama, O. F. Linares & A. J. Ranere (eds), 57–66. Peabody Museum
Monographs, no. 5. Cambridge, Mass.: Harvard University Press.
Linares, O. F. 1980b. Ecology and prehistory of the Chiriqui Gulf sites. In Adaptive radiations in
prehistoric Panama, O. F. Linares & A. J. Ranere (eds), 67–77. Peabody Museum Monographs, no.
5. Cambridge, Mass.: Harvard University Press.
Linares, O. F. & R. S. White 1980. Terrestrial fauna from Cerro Brujo (CA-3) in Bocas del Toro and
La Pitahaya (15–3) in Chiriqui. In Adaptive radiations in prehistoric Panama, O. F. Linares & A. J.
Ranere (eds), 181–93. Peabody Museum Monographs, no. 5. Cambridge, Mass.: Harvard University
Press.
Lothrop, S. K. 1937. Coclé: an archaeological study of central Panama, Part 1. Memoirs of the
Peabody Museum of Archaeology and Ethnology 7. Harvard University.
Lothrop, S. K. 1942. Coclé: an archaeological study of central Panama, Part 2. Memoirs of the
Peabody Museum of Archaeology and Ethnology 8. Harvard University.
McGimsey, C. R. III 1956. Cerro Mangote: a preceramic site in Panama. American Antiquity 22, 151–
61.
McGimsey, C. R. III, M. B. Collins & T. W. McKern 1966. Cerro Mangote and its population.
Unpublished paper presented at the 37th International Congress of Americanists, Mar del Plata.
Myers, C., J. Daly & B. Witkop 1978. A dangerously toxic new frog (Phyllobates) used by the
Emberá Indians of western Colombia with discussion on blowpipe fabrication and dart poisons.
Bulletin of the American Museum of Natural History 161, Article 2.
Nietschmann, B. 1973. Between land and water: the subsistence ecology of the Miskito Indians,
eastern Nicaragua. New York: Seminar Press.
Piperno, D. 1984. A comparison and differentiation of phytoliths from maize and wild grasses: use of
morphological criteria. American Antiquity 49, 361–83.
Piperno, D. & K. H. Clary 1984. Early plant use and cultivation in the Santa Maria Basin, Panama:
data from phytoliths and pollen. In Recent developments in isthmian archaeology, F. Lange (ed.),
85–121. Oxford: BAR International Series, 212.
Piperno, D., K. H. Clary, R. G. Cooke, A. J. Ranere & D. Weiland 1985. Preceramic maize in Panama.
American Anthropologist 87, 871–8.
Pohl, M. & L. H. Feldman 1982. The traditional role of women and animals in lowland Maya
economy. In Maya subsistence: studies in memory of Dennis E. Puleston, K. Flannery (ed.), 295–
311. New York: Academic Press.
Ranere, A. J. 1980. Nueva excavación y re-interpretación de Cerro Mangote, un conchero
precerámicoen la Región Central de Panamá. Unpublished paper presented at the ‘III Simposium
Nacional de Antropologia de Panamá’, December.
Ranere, A. J. & P. Hansell 1978. Early subsistence patterns along the Pacific coast of central Panama.
In Prehistoric coastal adaptations, B. L. Stark & B. Voorhies (eds), 43–59. New York: Academic
Press.
Redford, K. H. & J. G. Robinson 1987. The game of choice: patterns of Indian and colonist hunting in
the Neotropics. American Anthropologist 89, 650–67.
Sowls, L. K. 1984. The peccaries. Tucson: University of Arizona Press.
Smythe, N. 1978. The natural history of the central American agouti. Smithsonian Contributions to
Zoology 257.
Valerio, W. 1985. Excavaciones preliminares en dos abrigos rocosos en la Región Central de Panamá.
Vinculos (Costa Rica) 11, 17–29.
Valerio, W. 1987. Analisis funcional v estratigrafico de 5F-9 (Carabalí), un abrigo rocoso en la
Región Central de Panamá. BA thesis, University of Costa Rica.
Weiland, D. 1984. Prehistoric settlement patterns in the Santa Maria drainage of central Panama. In
Recent developments in isthmian archaeology, F. Lange (ed.), 31–53. Oxford: BAR International
Series, 212.
Weiland, D. 1985. Preceramic settlement patterns in the Santa Maria drainage in central Panama.
Unpublished paper presented at the 45th International Congress of Americanists, Bogota, July.
Willey, G. R. & C. R. McGimsey, III. 1954. The Monagrillo culture of Panama. Papers of the Peabody
Museum of Archaeology and Ethnology, 49(2), Cambridge, Mass.
Wing, E. 1978. The use of dogs for food. In Prehistoric coastal adaptations, B. Stark & B. Voorhies
(eds), 43–59. New York: Academic Press.
Wing, E. & A. Brown 1979. Paleonutrition. New York: Academic Press.
Yost, J. A. & P. M. Kelley 1983. Shotguns, blowguns and spears: the analysis of technological
efficiency. In Adaptive responses of native Amazonians, R. B. Hames & W. Vickers (eds), 189–224.
New York: Academic Press.
27 Shells and settlement: European
implications of oyster exploitation
DEREK SLOAN

A feature of the food-quest which first becomes obtrusive in the north

European archaeological record during the 4th millennium BC was the
consumption of shellfish and more particularly of Ostrea edulis, Mytilus
edulis, Cardium, Nassa reticulata and Littorina littorea on a scale sufficient
to result in the accumulation of substantial shell-mounds. Why this should
have happened at this particular period and not for instance during the
earlier phases of marine transgression is one of many problems left
unsolved (Clark 1975, pp. 192–3).

Introduction

Shell-middens are a world-wide feature of the archaeological record. They

are the logical product of any society which exploits the margins of the sea
as a major economic resource. Marine molluscs are abundant and easily
replenished, and their products have many potential uses (Ceci 1984),
although it is probably reasonable to assume that in prehistory their value as
a human food outweighed all other uses. Sadly, the early evolution of
archaeological thought meant that shell-middens were associated with the
concept of ‘strand-looping’, in which miserable and degenerate societies
scavenged a poor living from the seashore in degraded circumstances, and
this view of shell-middens exists to this day (Sloan 1985).
Much research on shell-middens has been undertaken in the past twenty
years, and ideas have changed radically. The interpretational barometer has
swung wildly as shell-middens and their enveloping ‘cultures’ have been
analysed: at the time of writing, it is fair to say that European shell-middens
may or may not represent:

(a) seasonal camps of inland hunter/gatherers;

(b) processing stations of essentially coastal societies;
(c) the miserable strand-loopers of tradition;
(d) any combination of the above.

However, these studies have tended to concentrate on individual sites, or

long-known groups of sites, and the results of such studies are inevitably
insular. This chapter takes a broader view of one group of shell-midden
sites, and demonstrates that there may be more important considerations to
be evaluated than some researchers have allowed.
The sites under consideration have one common feature – they are
massive mounds of the common oyster, Ostrea edulis. Three assemblages
of such sites have so far been identified: in Denmark, in Ireland, and in
Scotland. This chapter concentrates on the Scottish sites, as the other two
areas have been fully discussed in other publications.

Early shell-middens

As briefly mentioned above, a traditional view of shell-middens has gone

down into archaeological lore. One of the aspects of this has been the
assumption that shell-middens belong to the Mesolithic period – that they
are the product of essentially nomadic hunter/fisher/gatherer societies. The
problem of interpretation is increased by the knowledge that early sites will
have been lost in sea-level rises; this is a confusing factor as in some areas
isostatic uplift has raised early shorelines well above modern sea-level,
while in the majority of maritime Europe almost all of the prehistoric
shorelines have been inundated. Generalizations on this topic are
dangerous, but it would seem (certainly from Scotland and Ireland) that
earlier (‘Mesolithic’) shell-middens were fairly small, and based on the
exploitation of a fairly wide range of resources (Mellars 1978, Coles 1971,
Woodman 1978). These sites may or may not have been seasonal coastal
camps or, conversely, part of a truly coastal-based economy (Mellars &
Wilkinson 1980). Of the molluscan resources exploited, oysters were rare in
the site assemblages – although it must always be remembered that the
species exploited is largely dependent on the nature of the shoreline under
consideration – and the individual sites seem only to have been occupied
for periods of, at most, a few hundred years. However, at around 4000 BC, a
new phenomenon emerges – massive middens of oyster shell.

Denmark

The classic Ertebølle sites have been well documented, and there is much
discussion as to their exact status. Only a brief summary of the physical
natures of these sites will be given here.
‘Ertebølle’ sites come in a range of sizes, and not all ‘Ertebølle’ sites are
shell-middens; they are classically assumed to be the product of a
‘Mesolithic’ culture. (This is because, when first identified, they lacked the
polished stone axes which were taken to denote ‘Neolithic’; if recognized at
a later period, they would almost certainly have been regarded as
‘Neolithic’ because they contain pottery.) Only the larger shell-midden sites
are of interest to this chapter – those of 2000 m3 or more of deposits, the
main bulk of these being the valves of Ostrea edulis. These sites also
contain sizeable assemblages of fish, bird, and mammal bone, which in two
fairly recent studies (Clark 1975, Bailey 1978) have been calculated as the
dominant resources, with oysters representing a fairly minimal level of
nutritional value. Other interesting aspects of these middens are that they
contain large numbers of fireplaces, many artefacts (although microliths are
absent) and pottery; it has also been recorded (Lubbock 1865) that the
oyster valves were noticeably larger than modern specimens from the area.
The Ertebølle midden sites can be demonstrated to be part of a large and
complex economy, and much work bas been done on their analysis and
interpretation. Many smaller sites can be linked to the major middens, as
shown in the analysis recently undertaken by Rowley-Conwy (1983). The
interesting aspect of these sites has little to do with these minutiae,
however; it is that the advent of the Ertebølle middens seems to arrive at
4100 BC (and to last at least until 3200 BC), to coincide with a major marine
transgression and that the period of the sites’ use is also the period of the
climatic optimum. Most interestingly, although this period is well within the
chronological ‘Neolithic’, there is no convincing evidence from any of
these sites to show that they had a ‘Neolithic’ economy – in particular, no
evidence of arable farming – although pottery and ‘lamps’ of fired clay are
among the type artefacts of this culture.

Scotland

It has long been known that there were shell-midden sites in Scotland; as a
result of isostatic uplift, Scotland’s early shorelines have been well
preserved. Two major middens of oyster shell have been investigated in the
past, both in the Forth Valley – Inveravon (Grieve 1874, Mackie 1972) and
Polmonthill (Stevenson 1946), and both were presumed to be typical
‘Mesolithic’ hunter–gatherer sites. The radiocarbon dates recovered by
Mackie (1972) of between 4060 ± 180 BC and 2250 ± 120 BC seemed to
support this interpretation, although, as Mackie noted at the time, the later
dates, for a site without visible evidence of any hiatus in occupation,
suggested serious falsities in the standard Three Age system (ibid).
Recent work has identified a total of 18 middens of oyster shell in the
Forth Valley. Similar sites exist around Inverness (Gourlay 1980), around
Elgin (Lubbock 1865, Sloan 1986), and possibly under Glasgow (Sloan
1982a (Fig. 27.1) Only one of these sites (Nether Kinneil) has been
extensively excavated, and even this excavation represents no more than a 4
per cent sample of the site (Sloan 1982b): the neighbouring sites of
Polmonthill and Inveravon have been investigated on a superficial level
(above), and one further site of this group – Cadger’s Brae – has been
radiocarbon dated. A small trial excavation of one of the Inverness sites
(Muirtown) was undertaken in 1979 (Gourlay & Myers unpublished).
Despite the limited amount of excavation, the sites have an impressive
number of common qualities. Almost all the the material in the middens
consists of oyster valves, although there are small representations of other
common littoreal species of mollusc (Littorina littorea, Cerastoderma
edule, Patella vulgata, Mytilus edulis), and crabs. All the sites contain
copious burning-levels and many stone-built hearths; in addition the Nether
Kinneil site produced vast and enigmatic stone structures with suggestions
of a complex and planned layout (Sloan 1986). Animal bone, including
apparently domesticated cattle and possibly sheep, was present at Nether
Kinneil, although only in very small quantities (about one piece for every
cubic metre excavated), and there were small quantities of other materials –
a very few stone artefacts, a few small sherds of coarse pottery, one
elaborate bone pin, and one shell bead. Most unexpectedly, there is no
evidence from any of the sampled sites for the exploitation of either
wildfowl or fish; although the remains of hazel-nut and edible weeds have
been recovered from Nether Kinneil, no trace of cereals has been found.
 Figure 27.1 Location of ‘Forth Valley’ type shell-middens. Forth Valley sites: 1 Mumrills; 2
Cadger’s Brae; 3 Millhall; 4 Little Kerse; 5 Piggery; 6–8 Polmonthill; 9 Inveravon; 10 Nether
Kinneil; 11 East Kerse 2&3; 12 East Kerse 1; 13 Deil’s Burn; 14–17 Torryburn. Loch Spynie sites: 1
Bennet Hill; 2 Easterton; 3 Findrassie; 4–5 Spynie; 6 Meft. Inverness sites: 1 Clachnaharry; 2
Muirtown; 3 Bank Street.

As a result of geomorphological factors and the limited nature of the

investigations so far carried out on these sites, it is very hard to calculate the
exact sizes of these middens. The three main Forth Valley sites (i.e. Nether
Kinneil, Inveravon, and Polmonthill) probably contain in excess of 3000 m3
of deposits, and the site of Meft on Loch Spynie (originally attributed to the
historic period on the basis of very flimsy evidence, see Lubbock 1865, p.
234) also seems to have been very large. (This site may not, strictly
speaking, belong to the same grouping, as it does not appear to have been
so predominantly comprised of oyster shells; only a reinvestigation can
make its status certain.) Nineteen 14C dates from four sites are available; all
fall within the period 4060–2200 BC, although it is possible that this date
range may be extended in both directions with further excavation. (Note: of
the dates, only two, Muirtown and one from Inveravon, are on charcoal; the
rest are on marine shell.) No matter what the exact significance of these
dates in absolute terms, the important consideration is that, on the evidence
available, these sites seem to have been continuously occupied. In the case
of Inveravon, at least 1900 years of uninterrupted occupation is suggested,
despite a transgressive episode which inundated part of the site.
The available data on vegetation and climate are, unfortunately, sparse;
but it is possible to say that the inception of this economy – as with that of
the Danish sites – seems to coincide with the climatic optimum and a major
marine transgression (they date to the period immediately after the post-
glacial maximum transgression), and to have been based on broad and
shallow estuarine environments. Although analysis has been hampered by
lack of funds, and it is still a long way from completion, it is possible to
suggest that at Nether Kinneil the oyster valves show a decrease in size
during the period of occupation of the site, and that they were originally of
a very large size. This would argue for a selectivity in the process of
collection of the shells, and possibly hint at a gradual over-exploitation of
the resource. Another interesting observation is that there is evidence to
suggest that the end of the use of Nether Kinneil may coincide with large-
scale land clearance and the inception of arable farming; there are plough
marks cut into the top of the midden at one point, and some parts of the site
are covered with thick, burnt deposits, which also coat the attendant buried
beach levels. There is a possibility that a similar event may have followed
the abandonment of the Muirtown site (Gourlay & Myers unpublished).

Ireland

Recent investigations (Burenhult 1984) of a series of massive shell-middens

in Ballysadare Bay, County Sligo (Fig. 27.2) provide interesting parallels
with, in particular, the Scottish sites. The analysis of this work is still in
progress, and, unfortunately, the earliest evidence would seem to have been
lost to marine erosion. It is possible (ibid, p. 41) that the beginnings of a
massive exploitation of Ostrea edulis may date from 3700 BC, although the
earliest date from the excavated site (Culleenamore) is 2700 BC. The
Culleenamore midden is some 100 m long, 30 m wide and up to 3 m deep,
and forms part of an almost 3000 m long, more or less unbroken, chain of
kitchen middens (ibid, p. 68).
The striking aspect of the Culleenamore site is its similarity to the
Scottish middens. The vast majority of the shells in the midden are Ostrea
edulis, with some representation of other common littoral species. The
oyster valves have been convincingly demonstrated to decrease in size
during the occupation of the midden, and there are copious burning levels
and hearths within the bulk of the site. Despite the abundance of charred
material, no cereals have been demonstrated to belong to the earlier parts of
the site (Monk 1984, p. 212). Although one partial cereal grain was found
in the later levels, this is most probably reflective of a much later, Iron-age,
event. As with the Scottish sites, the bones of fish and wildfowl are absent,
although a small sample of the bones of both domestic and wild animals
was recovered (cattle, pig, sheep/goat, red deer), together with some
artefacts, including pottery.
Other interesting parallels between this site and the Scottish middens lie
in the possibility of structures within the mound, an apparently continuous
occupation (Burenhult 1984, p. 133), and an association with a period of
falling sea-level combined with a warm, wet climate (ibid, pp. 42, 338,
Osterholme & Osterholme 1984). The occupation seems to have bridged the
conventional ‘Neolithic’ and ‘Bronze Age’ periods, and it has been
suggested that these middens formed ‘home base’ sites (ibid, p. 133).

Discussion

All the sites discussed here seem to be associated with periods of falling
sea-level following major transgressions, all occur within the ‘climatic
optimum’, and all are situated in shallow estuarine environments.
Obviously they have the major attribute of being largely composed of
oyster shell, a resource little seen in the archaeological record of earlier
periods; they also all have plentiful evidence for activities taking place on
the middens, in the form of burning layers, structures, etc. There is also a
constant suggestion of over-exploitation of the basic resource, and
uninterrupted occupation. All have evidence of ‘Neolithic’ activities, in the
form of pottery and domesticated animals (although it can be argued that
the ‘domestic’ faunas from the Ertebølle sites are intrusive, this may be a
case of wish-fulfilment; concerted efforts were made to deny the existence
of such faunas from the Forth Valley sites until the evidence became
overwhelming).
 Figure 27.2 Location of Culleenamore and Ballysadare Bay (after Burenhult 1984, p. 23).

Yet these groups of sites are widely separated geographically, and there
are differences between them: the Ertebølle middens seem to represent
more complex economies than the other two groups; however, it must be
remembered that the Ertebølle complex has been studied in much greater
detail than the Scottish or Irish sites, and that further fieldwork and
excavation could fill in the gaps in the record in the other two areas, the
fishing and wildfowling camps, etc., all of which are very hard to locate.
It has been variously suggested (e.g. Clark 1975, Woodman 1984, Sloan
1985) that each of these midden complexes represents a mere seasonal
aspect of a mobile or semi-mobile economy. However, this may not be the
real answer to the question of the status of the sites.
Rowley-Conwy (1983) has produced a powerful argument for regarding
the large middens of the Danish Ertebølle as permanent sites, the home-
bases of complex (i.e. sedentary) hunter-gatherers. These bases would,
logically, be situated near the source of the most reliable food supply, and
oysters could be most readily exploited during the lean periods of the year.
Dating evidence for other Danish ‘Mesolithic’ cultures, suggesting that the
coastal and inland sites may not be so closely related as previously thought,
lends support to this theory (Jensen 1982, p. 64). There is a most attractive
proposition here, if only we are willing to accept it; it is simply that we are
used to thinking of ‘hunters’ as nomadic, and only farmers as ‘sedentary’,
yet hunters can be sedentary, given sufficient resources. Another of
Rowley-Conwy’s points (1983, p. 112) that sedentary hunters have a more
developed technology, with pottery, etc., also seems apposite, as does the
contention that sedentism allows for a greater population. A greater
population carries its own inertial force, possibly accounting for both the
long occupations of the midden sites and the eventual over-exploitation of
resources noted in all these midden complexes; yet a seasonally
concentrated and non-mobile marine resource such as oysters is less
vulnerable than, say, red deer, which might be driven away by the presence
of a large and permanent human population.
This is an interesting theoretical approach, but it has only been applied in
detail to the Danish sites. However, there is evidence from both Ireland and
Scotland to fit this thesis. Recent research has shown a total lack of
evidence for any sites contemporary with the Forth Valley shell-middens in
the hinterland of the area (Sloan 1987); a similar observation has been made
in respect of the sites in Ballysadare Bay (Burenhult 1984). Although little
weight can be placed on such negative evidence, it seems a little odd that
there should be no late ‘Mesolithic’ or early ‘Neolithic’ sites known, even
from chance finds, in the intensely exploited Forth Valley, especially as
there are many sites of Early Bronze Age and later dates known. So it does
seem reasonable to suggest that the shell-middens themselves were the
main foci of settlement during the 4th and 3rd millenniums BC.
It can be tentatively suggested that there is a common link here, and that
a massive exploitation of Ostrea edulis made possible the development of
sedentary hunter-gatherer systems in these three areas. This is apparently
linked to more favourable marine and climatic conditions which began at
around 4000 BC, making oysters a richer, more easily available, and more
reliable resource. These complexes are restricted in geographical spread
because it is only on temperate coasts that the optimal conditions could
exist (Rowley-Conwy 1983, p. 118).

Acknowledgements
I am indebted to Paul Mellars and Peter Woodman for discussion of the ideas contained within this
chapter. Also to Mrs D. W. Sloan for unfailing support in fieldwork, research, typing, for drawing
Figure 27.2, and too many other areas to mention. To all volunteers, staff, and specialists who worked
on the excavations and in post-excavation; to Frances Murray, David Devereux, and Patrick
Ashmore; and to the SDD (Ancient Monuments), DES, the Russell Trust, the Royal Commission on
Ancient and Historic Monuments of Scotland, and the British Archaeological Research Trust for their
financial support. Not least to Dr Ted Luxon and Mr John Smith for their unfailing efforts to
publicize and raise funds for this research.

References
Bailey, G. N. 1978. Shell middens as indicators of postglacial economies: a territorial perspective. In
The early postglacial settlement of northern Europe, P. A. Mellars (ed.), 37–63. London:
Duckworth.
Burenhult, G. 1984. The archaeology of Carrowmore. Theses and Papers in North-European
Archeology 14, 23.
Ceci, L. 1984. Shell midden deposits as coastal resources. World Archaeology 16(1), 62–74.
Clark, J. G. D. 1975. The earlier Stone Age settlement of Scandinavia. Cambridge: Cambridge
University Press.
Coles, J. M. 1971. The early settlement of Scotland: excavations at Morton, Fife. Proceedings of the
Prehistoric Society 37, 284–366.
Gourlay, R. 1980. Muirtown, Inverness. Proceedings of the Prehistoric Society 46, 363–4.
Gourlay, R. & A. Myers unpublished. A report on the preliminary investigation of a shell-midden,
Muirtown, Inverness.
Grieve, D. 1874. Notes on the shell-heaps near Inveravon, Linlithgow. Proceedings of the Society of
Antiquaries of Scotland 9, 45–52.
Jensen, J. 1982. The prehistory of Denmark. London: Methuen.
Lubbock, Sir J. 1865. Prehistoric times. London: Frederic Northgate.
Mackie, E. W. 1972. Radiocarbon dates for two Mesolithic shell heaps and a Neolithic axe factory in
Scotland. Proceedings of the Prehistoric Society 46, 412–16.
Mellars, P. A. 1978. Excavation and economic analysis of Mesolithic shell middens on the Island of
Oronsay (Inner Hebrides). In The early postglacial settlement of northern Europe, P. A. Mellars
(ed.), 371–96. London: Duckworth.
Mellars, P. A. & M. Wilkinson 1980. Fish otoliths as evidence of seasonality in prehistoric shell
middens: the evidence from Oronsay (Inner Hebrides). Proceedings of the Prehistoric Society 46,
19–44.
Monk, M. 1984. Charred plant remains, other than wood charcoal, from the Carrowmore
excavations. In The archaeology of Carrowmore, G. Burenhult (ed.), Theses and Papers in North-
European Archaeology 14, 210–13.
Osterholm, I. & S. Osterholm 1984. The kitchen middens along the coast of Ballysadare Bay. In the
archaeology of Carrowmore, G. Burenhult (ed.), Theses and Papers in North-European
Archaeology 14, 326–45.
Rowley-Conwy, P. 1983. Sedentary hunters: the Erterbolle example. In Hunter-gatherer economy: a
European perspective, G. Bailey (ed.), 111–26. Cambridge: Cambridge University Press.
Sloan, D. 1982a. A prehistoric site under Duke Street? Glasdig 2, 7.
Sloan, D. 1982b, Nether Kinneil. Current Archaeology 84, 13–15.
Sloan, D. 1985. Shell-middens and chronology in Scotland. Scottish Archaeological Review 3, 73–9.
Sloan, D. 1986. Shell-middens of Scotland. Popular Archaeology 7 (1), 10–15.
Sloan, D. 1987. The puzzle of the shell mounds. Scots Magazine, 127 (4), 383–9, new series.
Stevenson, R. B. K. 1946. A shell-heap at Polmonthill, Falkirk. Proceedings of the Society of
Antiquaries of Scotland 80, 135–9.
Woodman, P. 1978.The Mesolithic in Ireland. Oxford: BAR British Series 58.
Woodman, P. 1984. Discussion. In the archaeology of Carrowmore, G. Burenhult (ed.), Theses and
Papers in North-European Archaeology 14, 389.
28 Effects of human predation and
changing environment on some mollusc
species on Tongatapu, Tonga
DIRK H. R. SPENNEMANN

Introduction

The Tongan Islands, being part of the western Polynesian triangle Fiji–
Samoa–Tonga, were initially settled by people belonging to the Lapita
cultural complex. The earliest radiocarbon dates show that Tongatapu was
settled around 1500–1300 BC. The earliest sites, pottery-bearing shell-
middens, are situated in a narrow band along the accessible shore (Groube
1971, Kirch 1978). As known from the midden deposits, the subsistence of
these settlers relied predominantly on the exploitation of molluscs and fish
offered by the lagoon and fringing coral reef. Although present, horticulture
seems to have played a minor role only. At later times, by the first centuries
AD, a change had taken place: pottery production was abandoned and
horticulture and domestic animal-breeding (pigs, chicken) increased. In
addition the settlements were shifted inland, apparently to be nearer to the
gardens.
Most of the archaeological research carried out in Tonga has concentrated
on the Lapita period (Spennemann 1986). Detailed studies on the
exploitation of molluscs, however, have not been undertaken so far in Tonga,
although some approaches have been made.

The problem

Onlookers at one of Jens Poulsen’s excavations commented on the size of

the excavated Gafrarium sp. shells and mentioned that such shells no longer
occur in modern Tongatapu (Poulsen 1967, p. 299). This observation
resulted in a series of measurements of the main mollusc species Gafrarium
tumidum, G. gibbiosum, and Anadara antiquata (Table 28.1). If the Tongans
were right in their statements, then two possible reasons for this change in
overall size could be presumed: over-exploitation of the resources or change
of the local environment. An initial graphical analysis of the length
measurements failed to prove any significant reduction in shell size among
the Gafrarium sample (Poulsen 1967, pp. 299–300), but a statistical re-
analysis showed a decrease in size beyond expectation (Poulsen 1984,
Spennemann 1985a 1985c.
Table 28.1 Species of shellfish from Tongatapu.

This chapter discusses the effects of both human predation and changing
environment on the predominantly exploited mollusc species Gafrarium and
Anadara, as well as the consequences following this decrease in size for the
subsistence of the early Tongans.
Shellfishing in ethnohistoric and modern Tonga is done by the women,
who walk at low tide to the reef and mudflats or in the lagoon, search the
ground with their toes, and dig out or pick up the shells with their hands.
Due to overall size, Anadara shells offer more meat than Gafrarium and are
therefore preferred. Gafrarium shells, however, are considered to be more
tasty.
Under discussion are the sites excavated by Poulsen in 1963–64 and
recently analysed shell samples from other sites. All are situated along the
shore of the lagoon. The relative chronology of the pottery, derived from rim
forms and decoration, allows clear sequencing of Poulsen’s sites and major
categorization into early, middle, and later Tongan Lapita (Poulsen 1983).
Since the pottery series from the other sites are too small or not sufficiently
worked on, the general division into early to late has to be adopted.

The changing environment

Covering 245 km2, Tongatapu is the largest island of the Tongan archipelago
(Fig. 28.1). It is a flat, tilted coral limestone block with only minor
elevations, rising to a maximum height of some 65 m. While the northern
shore slopes gently into the sea, the southern coast consists of inaccessible
cliffs. Modern Tongatapu is dominated by the large Inner Lagoon, which
consists of two pockets, Fanga ’Uta and Fanga Kakau.
Figure 28.1 Map of Tongatapu, Tonga. Solid black dots: Lapita sites mentioned in the text; circles:
other Lapita sites; A: Fanga ’Uta Lagoon; B: Fanga Kakau Lagoon; C: Nuku’alofa–Ma’ofanga
peninsula.

As shown by Poulsen, based on an analysis of soil samples from site TO–

Pe–1 (Crook 1967), the Fanga ’Uta Lagoon extended further inland during
Lapita times than it does today. Geological research has shown that
Tongatapu was either uplifted as a whole or tilted in its northern part (Taylor
& Bloom 1977, Bloom 1980). Assuming that the ancient shorelines are
represented by the Lapita shell-middens, that is the water level was 1.5–2.0
m higher than today, then the shoreline at the time of the arrival of the
humans must have looked roughly like the sketch drawn in Figure 28.2a.
Prehistoric Tongatapu was an island with a wide open bight and scattered
islets within. This reconstruction is confirmed by some Lapita shell-middens
which lie today within the Nuku’alofa-Ma’ofanga peninsula, and in Lapita
times were on small islets (TO–Nu–2, –8, –12, –16, –18, –19). Taylor &
Bloom (1977) argued for a former lagoonal entrance in the Nuku’alofa area,
as coral heads found in situ in the streets of Nuku’alofa were dated to 4000
BP. This described environment closely resembles that typically preferred for
Lapita settlement (Jennings 1980, p. 3).
During the tectonic changes, the present Nuku’alofa–Ma’ofanga peninsula
was formed and half of the bight cut off from sea water. This part of the
lagoon (Fanga ’Uta) turned into a inner lagoon with brackish water and
almost no intake of pure sea water from the reef area. This can be
documented by the tidal range, which is currently some 1.2–1.4 m at the sea
side of Nuku’alofa, but at the inner lagoon is a mere 0.2 m (Braley 1979,
Belz 1984).
Figure 28.2 Tongatapu, Tonga: (a) coastline at about 1500 BC, (b) coastline at AD 1986.

This brackish water favoured species preferring this environment, like

Gafrarium, while putting other species, which are dependent on pure sea
water (i.e. Anadara) at a disadvantage. Theoretically this closure of one half
of the lagoon had the following effects on the two major mollusc species
Anadara and Gafrarium:
(a) while Anadara shells became scarce, Gafrarium shells increased in number;
(b) the Gafrarium shells increased in size, while the Anadara shells became smaller.

Proportion of species

The ratio between the lagoon species (Gafrarium) and the reef species
(Anadara) within the shell-middens adjacent to the inner lagoon are set out
in Tables 28.2 and 28.3. In general, the trend of changing environments
seems to have started prior to the very beginning of Lapita settlements on
Tongatapu, since the percentage of Gafrarium in the subsoil was already at
45–65 per cent.
As can be judged from the ratio between these two genera in the lagoon
entrance site (TO–Nk–2) and in those sites at the inner lagoon where data on
the underlying subsoil are presently known (i.e. TO–Pe–1, TO–Pe–5), the
living conditions for Gafrarium shells were already rather favourable at the
beginning of the Lapita occupation and were becoming more favourable
during time.
Comparing ratios within the midden horizons of the individual sites (Table
28.2) it is clearly observable that the number of Gafrarium shells is steadily
increasing. This is not only due to the environmental change, since
contemporary adjacent sites show different ratios. It seems that the
exploitable resources of Anadara shells were quickly depleted in the area. If
this is true, then we can argue that neighbouring populations (like TO–Pe–3
and TO–Pe–5, roughly 1 km apart) exploited different shell beds. A detailed
analysis and further sampling of the site might provide additional data and
enable reconstruction of exploitation areas.
Table 28.2 Percentages of Gafrarium sp. and Anadara sp. shells in various shell-middens from
Tongatapu, Tonga with the chronology (after Poulsen 1983)
 Table 28.3 Percentages of Gafrarium sp. and Anadara sp. shells in various shell-middens from
Tongatapu, with the modern location of the sites (see Fig. 28.1)
Gafrarium : Anadara n
Reef islet sites
TO–Vi–1 5.9 : 94.1 34
Lagoon entrance sites
TO–Nk–2 39.6 : 60.4 409
Open lagoon sites
TO–Mu–2 75.7 : 24.3 37
Inner lagoon sites
TO–Pe–3 66.0 : 34.0 3383
TO–Pe–5 73.1 : 26.9 1270
Innermost sites
TO–Nu–8 82.5 : 17.5 240
TO–Pe–1 93.2 : 6,8 1455
TO–Pe–6 79.9 : 20.1 2922

Shell size

The greatest length of all Gafrarium and Anadara shells derived from shell
columns sampled by J. Poulsen was measured (see Fig 28.3). His data had
been recorded to the nearest 0.25 cm and therefore vary slightly from the
recently gathered data taken to the nearest millimetre. In addition to the data
from archaeological contexts, some spits at TO–Pe–1 were sunk into the
subsoil, providing information on the natural death assemblage before the
arrival of humans. The other end of the sequence is represented by a market
catch of Gafrarium shells bought in summer 1985 by the author
(Spennemann 1985b). The recent background sample for Anadara was
provided by a sample from a modern shell-midden (1970s) on Makaha’a
Islet, offshore of Tongatapu (Spennemann unpublished fieldnotes 1985).
The interpretation of the data for the Gafrarium is straightforward as there
is a steady decrease in size from early Lapita layers onwards. While the
decrease from early to middle, and from middle to late Lapita is highly
significant, the decrease from late Lapita to the recent sample is not. This
shows that the major change in size had already taken place during Lapita
times. In addition to the overall reduction of length, the modern Gafrarium
shells are not as thick-walled as those found on the Lapita sites. Human
activity – predation – is clearly observable. Instead of becoming larger, as
might be expected from the favourably changed living conditions, the shells
become smaller. A comparison of the natural death assemblage with the
early Lapita layers shows that the shells from the archaeological layers are
significantly larger. Human selection for the bigger specimens is clearly at
work. The recent sample is very significantly smaller than the natural death
assemblage, showing that human predation had a long-lasting impact on the
Gafrarium population.
 Figure 28.3 Greatest length taken. Left: Gafrarium sp.; right: Anadara sp.

While these trends work out perfectly for the broad chronological steps, a
detailed spit to spit analysis of individual sites gave no clear results. The
shell size varies, alternately increasing and decreasing throughout the
sequence.
A more complicated pattern arises for Anadara. The sequence shows no
trend, either in the broad chronological steps, or in the detailed spit to spit
analysis, but alternately increases and decreases in shell size.
Comparison of the modern sample and the natural death assemblage
beneath site TO–Pe–1 shows that the modern shells are much bigger. Since
living conditions for Anadara on the reef islet have not changed, this
observation reflects the fact that some environmental change had already
taken place, affecting the natural Anadara population around the area of
TO–Pe–1.
That the size of the Anadara sp. shells is dependent on the location of the
site in the bight/lagoon can be documented by data from an as yet undated
but definitely post-Lapita shell sample from Velitoa Hahake (TO–Vi–1), a
small islet off Tongatapu. These shells are comparable in size to those from
Makaha’a, and the environmental conditions at both sites are similar.
 Shells originating from the early Lapita site TO–Nk–2, situated at the
modern lagoon entrance, are very significantly larger than both the pooled
early Lapita and the natural death assemblages, but are significantly smaller
than the samples from Makaha’a and Velitoa.
The lower layers of site TO–Pe–1, also belonging to the early Lapita
phase, are almost identical in size to the natural death assemblage derived
from the underlying subsoil, but are very significantly smaller than those
from site TO–Nk–2.
Another set of chronologically contemporary, but geographically distant,
sites shows the same pattern: the late Lapita site TO–Mu–2, situated at the
Fanga Kakau lagoon, shows larger shells than site TO–Pe–6, situated at the
Fanga ’Utu lagoon. If the Anadara shell size is a highly sensitive tool for
observing the changing environment, then the size pattern observable within
three sites located near to each other allows for dating of the closure of the
Nuku’alofa peninsula: the natural death assemblage, the early Lapita layers
from TO–Pe–1, and the middle Lapita series from TO–Pe–3 show all
similarly big shells. Those from the late Lapita site TO–Pe–6, however, are
very significantly smaller. This implies that the Fanga ’Uta lagoon was
completely cut off by the end of the middle Lapita period, i.e. about 700–500
BC.

Conclusions

Summarizing these findings, we can conclude that the Lapita population had
to face a dwindling supply in large Anadara sp. shells, which was caused by
both predation and changing environment. The other main species,
Gafrarium, became continuously smaller in size.
This implies that the same number of shells collected contained less meat
than before. If we assume that the Lapita population was growing in number
– as can be documented to some extent by the high number of middle Lapita
and especially late Lapita sites along the lagoon – then these resources must
have become less and less sufficient. In this case, the meat supply had to be
topped up by exploiting other shellfish species, in particular those given in
Table 28.1. Some clue for this is given by comparing the percentages of
Gafrarium and Anadara shells to all shells in the sample. The amount of
other shells increases within the individual middens as well as between the
chronological steps. In addition Poulsen (1987) observed an overall decrease
of shells compared to the total midden debris (stones, earth, etc.).
The subsistence economy of the later Lapita times suffered from
diminishing shellfish returns both in quality and quantity. It has to be asked
whether this lack was made good by intensified exploitation of lagoon and
reef fish, or whether the horticultural aspect was increased. The giving up of
pottery and the introduction of a new food-preparation method, the ’umu, as
well as a significant increase of both pit-digging for presumed food storage,
and pig bones as another indicator for horticulture, would seem to support
the latter solution, which is further underlined by the relocation of
settlements into the inland areas. However, it has to be mentioned that, to
date, studies on the exploitation of fish are insufficient to give conclusive
evidence.

Implications for further research

The data available to date are not geographically equally distributed

throughout the lagoon, as can be seen in Figure 28.1. Further sampling and
midden analysis is necessary to provide a broader basis of data. It is
necessary to understand the regional component of the shellfish exploitation,
as hinted by the ratio between Gafrarium and Anadara, as well as by the
representation of the other species in the later sites.
In particular, data from sites on the Nuku’alofa-Ma’ofanga peninsula are
needed for a better understanding of the environmental conditions at the time
of the arrival of humans and the environmental change thereafter.

Acknowledgement

I am gratefully indebted to J. Poulsen (Aarhus) who readily allowed

reworking of the data gathered by him, and for discussion of the matter.

References
Belz, L. H. 1984. Nuku’alofa sanitation and reclamation. Nuku’alofa: Ministry of Health.
Bloom, A. L. 1980. Late Quaternary sea level change on south Pacific coasts: the study in tectonic
diversity. In: Earth rheology, isostasy and eustasy, N.-A. Moerner (ed.), 505–516. Chichester: J.
Wiley.
Braley, R. D. 1979. Penaeid prawns in Fanga’uta Lagoon, Tongatapu. Pacific Science 33, 315–321.
Crook, K. A. W. 1967. Appendix III. Analysis of soil samples from To.1. In A contribution to the
prehistory of the Tongan Islands. Unpublished PhD thesis by J.I. Poulsen, Australian National
University., Canberra.
Groube, L. M. 1971. Tonga, Lapita pottery and Polynesian origins. Journal of the Polynesian Society
81, 278–316.
Jennings, J. D. 1980. Introduction. In Archaeological excavations in western Samoa, J. D. Jennings &
R. N. Holmer (eds), 1–4. Pacific Anthropological Records 32. Honolulu: Bernice P. Bishop
Museum.
Kirch, P. V. 1978. The Lapitoid period in West-Polynesia: excavations and survey in Niuatoputapu,
Tonga. Journal of Field Archaeology 5, 1–13.
Poulsen, J. I. 1967. A contribution to the prehistory of the Tongan Islands. Unpublished PhD thesis.
Australian National University, Canberra.
Poulsen, J. I. 1983. The chronology of early Tongan prehistory and the Lapita-Ware. Journal de la
Société des Océanistes 39, 46–56.
Poulsen, J. I. 1984. Analysis of length measurements on Gafrarium shells. MS on file, Department of
Prehistory, Research School of Pacific Studies, Australian National University.
Poulsen, J. I. 1987. Early Tongan prehistory. Terra Australis 12. Canberra: Australian National
University.
Spennemann, D. H. R. 1985a. A comparative analysis on some Gafrarium sp. and Anadara sp. shells
from shellmiddens on Tongatapu. Osteological Report DRS 33 (1985). Ms on file, Department of
Prehistory, Research School of Pacific Studies, Australian National University, Canberra.
Spennemann, D. H. R. 1985b. Length measurements of a modern comparative sample of Gafrarium
sp. shells from Tongatapu, Tonga. MS on file, Department of Prehistory, Research School of Pacific
Studies, Australian National University, Canberra.
Spennemann, D. H. R. 1985c. A re-analysis of some shell-measurements on Gafrarium sp. and
Anadara sp. shells from Tongatapu, Tonga. Osteological Report DRS 37 (1985). MS on file,
Department of Prehistory, Research School of Pacific Studies, Australian National University,
Canberra.
Spennemann, D. H. R. 1986. Zum gegenwärtigen Stand der archäologischen Forschung auf den
Tonga–Inseln. Ergebnisse und Perspektiven. Anthropos 81. 469–495.
Taylor, F. W. & A. L. Bloom 1977. Coral reefs on tectonic blocks, Tonga Island arc. Proceedings of
the Third International Coral Reef Symposium. Vol. 2:Geology. Miami: Rosentiel School of Marine
and Athmospheric Science, University of Miami.
29 Rocky Cape revisited – new light on
prehistoric Tasmanian fishing
SARAH M. COLLEY and RHYS JONES

The Tasmanian fish problem

Ethnographic accounts indicate that at the time of European contact the

Tasmanians did not eat fish. Jones’ excavations in two sites, North Cave and
South Cave at Rocky Cape, northwest Tasmania, carried out in the 1960s
(Fig. 29.1), revealed abundant fish remains in midden layer dated before c.
3800 BP (Jones 1971). More recent layers contained no fish bones. A similar
pattern was subsequently discovered elsewhere in Tasmania (Lourandos
1970, Jones 1978, Bowdler 1984, Vanderwal & Horton 1984). It was clear
that the Tasmanians had previously eaten fish but had stopped sometime
between c. 3800 and 3500 BP.
After careful consideration of the evidence, Jones proposed a cultural
explanation for this behaviour. In his view, the decision to drop fish from the
diet did not make economic sense. He saw it as a cultural aberration which
survived only because the Tasmanians were an isolated population cut off
from competing societies on the Australian mainland.
Jones’ views on prehistoric Tasmanian society prompted a storm of
discussion which attracted world-wide archaeological attention (for example
see Vanderwal 1978, Allen 1979, Horton 1979, Bowdler 1980, 1984, Walters
1981, Sutton 1982, White & O’Connell 1982, pp. 157–70). Fierce criticism
was levelled at the political implications of some of Jones’ conclusions, such
as his contention that Tasmanian society was ‘doomed to a slow
strangulation of the mind’ (1977, p. 203). Other critics questioned his
interpretation of the archaeological data, in particular arguing that loss of
fish from the Tasmanian diet did in fact make economic sense. Following an
initial surge of interest, discussion of the Tasmanian fish problem subsided
somewhat. However, reference to the question as an important
archaeological issue in two recent publications (Mellars 1984, White 1984)
confirms its continuing relevance.

Revived interest in an unanswered question

Currently the question ‘Why did the Tasmanians stop eating fish?’ remains
unresolved. For several reasons we believe it now merits further attention. It
impinges on a key area of archaeological interest – the question of ‘cultural’
versus ’ecological’ explanations of human behaviour. The material itself is
of great archaeological value due to the nature of the site, which according to
Bowdler (1984, p. 2) is ’still unparalleled in Australian archaeology’.

Figure 29.1 Map to show the location of Rocky Cape.

All discussion of the Tasmanian fish question to date has been based on
Jones’ original analysis of faunal samples from the Rocky Cape cave sites.
The work was carried out at a time when little archaeozoology had been
done in Australia and when the zoological expertise and comparative
collections necessary to identify and fully analyse the whole range of the
Rocky Cape fish fauna were not available. Particular problems with the fish
bones also arose because the soil was mostly sieved through a 5 mm mesh,
now considered too large to retain small bones.
Almost all contributors to the Tasmanian fish debate have taken Jones’
results at face value, ignoring any associated methodological problems.
White even commented that: ‘The facts of the Tasmanian case now seem
beyond dispute …’ (1984, p. 13). On the contrary, new analysis of fish
remains from Rocky Cape shows that there is still much to learn.

Samples from Inner Cave, Rocky Cape South

Preliminary results suggest that further analysis of Jones’ original faunal

samples from the North and South Caves are likely to produce valuable new
results. Of even greater potential interest is the material from an undisturbed
living floor, with excellent preservation of environmental remains, which
was sealed off soon after 6700 BP – the Inner Cave or Enclosed Chamber at
Rocky Cape South (Jones 1971, pp. 558–84, Jones 1980). Recognizing the
great importance of the site, and that contemporary archaeological methods
would soon become obsolete, Jones conducted only limited excavations,
preferring to leave the site for future investigation.
One hundred soil samples were taken from transects c. 0.3 m wide and 50
mm deep across the living floor (Fig. 29.2). Three-quarters of this soil was
sieved through 10 mm, 5 mm, and 1 mm meshes, and preliminary sorting
was carried out. One-quarter of excavated soil was retained as bulk soil
samples, together with all sieved material and a sample of the residue, as an
archive for future analysis. Jones’ preliminary study of the site is presented
in his thesis (1971), but detailed analysis of faunal remains is yet to be done.
The Inner Cave samples contain many types of fish remains which are
absent in the samples recovered by Jones from the main midden sequences.
Findings from such a unique site as the Inner Cave are interesting in
themselves. They can also provide a check on what may have been lost
elsewhere on the site due to worse preservation conditions, and sampling and
excavation techniques.
Preliminary results of the fish bone analysis

Seventeen families or types of fish have been identified so far (Colley &
Jones 1987). At least 14 more are present but not yet identified owing to lack
of comparative skeletons (Table 29.1). Commonest are bones from all areas
of the skeleton of wrasses and leatherjackets. These occur in most samples
from both the main midden sequence and the Inner Cave. Most of the
porcupinefishes and boxfishes originate from the Inner Cave, represented
almost exclusively by dermal spines or dermal plates. Bones of other fishes,
represented by vertebrae and a few jaw bones, are found occasionally in both
the main sequence and the Inner Cave.
Full discussion must await detailed quantitative data. However,
preliminary results show that many more taxa are represented than were at
first identified. Jones described fishing based almost exclusively on wrasses
(1978, p. 27). Leatherjackets, which are common across the site, were not
recognized at the time. Taking differential preservation into account,
leatherjackets were probably as common in prehistoric Tasmanian fishing
catches as wrasses. The relative numeric importance of other fishes is
difficult to assess at this stage. Some types, such as Australian salmon and
cartilaginous fishes, have very fragile skeletons and are almost certainly
under represented. The fish listed in Table 29.1 represent a variety of
habitats and feeding behaviours, suggesting several different fishing
strategies. Species not yet identified may well indicate yet further variety in
fishing methods.
Figure 29.2 Plan of Inner Cave, Rocky Cape South, showing sampled transects.

Table 29.1 Fish identified from preliminary analysis of Rocky Cape samples
Plus at least 14 other types of fish, all represented occasionally, and all recognized solely by vertebrae.
Ja = jawbones; He = other head bones; Sp = spines and fin rays; Sc = scales, scutes and dermal
denticles; Ve = vertebrae.

The second conclusion involves differential recovery. Spines of the

toothbrush leatherjacket, Penicipelta vittiger, are common in the Inner Cave
1 mm residues and bulk samples. They do not occur in the main midden
sequence, although other leatherjacket bones are common. The spines, which
are extremely small, were presumably missed during excavation and fell
through the 5 mm sieves. A similar explanation may account for the
differential spatial distribution of boxfish and porcupinefish dermal plates
and spines.
Much of the Tasmanian fish debate centres on the relative importance of
fish compared to other food resources. Obviously, answering this question
will require further study of evidence for other subsistence activities.
Likewise, the Tasmanian fish question is only part of a group of
archaeologically visible phenomena (e.g. changes in stone-tool types, the
disappearance of bone points) which may be related, and are currently
receiving renewed study.
 ‘Relative importance’ can refer to the relative contribution of each type of
food to the diet, which involves calculating relative dietary contribution
from relative proportions of food remains in archaeological samples, taking
into account factors of site formation, differential survival of evidence, and
excavation bias. Some of the difficulties are already apparent in the above
preliminary discussion of the Rocky Cape fish remains.

Table 29.2 Fish taken in simple baited-box traps (based on information supplied by P. Last, C. Turner,
and I. Whitehouse).
Fish type Identified in Rocky Cape samples to
date
leatherjackets (Monacanthidae) Yes
wrasses (Labridae) Yes
porcupinefishes (Diodontidae) Yes
cowfishes (Ostraciontidae)* Yes
morid cods (Moridae) Yes
lings (Ophidiidae) Yes
conger (Congridae) Yes
marblefish (Aplodactylidae) Yes
sergeant baker (Aulopodidae) ?+
small sharks ?‡
sweeps (Scorpidae)§ No
pufferfish (Tetraodontidea) No¶
* Very small specimens would escape unless trap was closely woven.
+ Could be present but not yet identified, due to lack of a comparative skeleton.
‡ Vertebrae of cartilaginous fish have been identified but not yet separated into shark and skate or
ray
§ An active swimmer which requires a well-built trap to retain it.
¶; Highly poisonous.

A related definition of ‘relative importance’ involves trying to reconstruct

how and when different foods were taken, to try to understand the relative
costs, returns, risks, and social implications involved in their exploitation.
The reconstruction of past fishing methods is important to this approach.

Fish traps
Table 29.2 lists fish most likely to be taken in a simple baited-box trap (Fig.
29.3, P. Last pers. comm.). Stockton suggested this as a likely method for
catching the wrasses originally identified by Jones from the Rocky Cape
excavations (1982, pp. 112–13). The suite of fishes most likely to be found
together in simple baitedbox traps (e.g. wrasses, leatherjackets,
porcupinefishes, cods, etc.) are also fish which occur most often in the
Rocky Cape samples. Cowfishes or boxfishes, also common in the
archaeological material, can be taken in baited box traps provided the mesh
size of the trap is small enough to retain them.
The absence of certain fish from the samples is also significant. Two very
common species in the Rocky Cape area today – magpie perch
(Cheilodactylus nigripes) and herring cale (Odax cyanomelas) – are never
taken in such traps, but are very easily speared while diving (P. Last pers.
comm.). The Rocky Cape archaeological fish remains have been checked for
these fish, using modern comparative skeletons, but neither has been
identified in the samples to date. Had spearing been an important fishing
method these species would be highly likely to occur at the site. Sweep,
Scorpis aequipinnis, a strong swimmer which can escape from a very simple
trap, has also not yet been identified in the samples, despite careful checking
against a modern skeleton. Current evidence therefore strongly favours the
use of simply constructed baited-box traps for prehistoric fishing at Rocky
Cape.

Figure 29.3 Diagram of a simple baited-box trap. Fish swim into the funnelshaped entrance (1)
attracted by the suspended bait (2) and are unable to escape through the narrow exit (3).
 Some fish identified in the Rocky Cape samples, e.g. barracouta
(Gempylidae), trevally (Carangidae), mullet (Mugilidae), and flatfish (Order
Pleuronectiformes) could not be taken in baited-box traps (P. Last pers.
comm.). It seems likely that some sort of tidal fish-trap was also used at
Rocky Cape.
A stone walled tidal fish-trap was initially noted near Rocky Cape by
Jones (1971). Several more in the area have since been described by
Stockton (1982). All have existed longer than any local resident can
remember, and have been used in recent times, but they are otherwise
undated. It is most unlikely that any stone fish-trap existing in the area today
was used in prehistoric times by Tasmanian Aborigines. However, traps
based on similar principles could have been used around Rocky Cape then.
Tidal traps in the area of Jacob’s Boat Harbour and Sister’s Beach (Figs
29.4 & 29.5) were used in recent times to catch a few specified fish: cocky
salmon (young Arripis trutta up to about 0.5 kg); black-back salmon (adult
Arripis trutta); warehou (Seriollela brama)1 and black bream
(Acanthopagrus butcheri) (I. Whitehouse & C. Turner pers. comm.).
Stockton (1982, p. 112) claimed that similar traps near Penguin, a few
kilometres east along the coast, ‘retained all the sizes and types of fish that
come into the area’, and his informant (G. Paine) named at least 19 types of
fish caught in these traps, which are, unfortunately, not listed by Stockton.
Stockton himself caught three types of fish in a tidal trap near Penguin:
garfish (Hemiramphus melanochir), yellow eyed or freshwater mullet
(Aldrichetta forsten), and yellowtail scad (Trachurus mcchullochi).
Figure 29.4 Sketch map showing the approximate location of five stone-walled fish traps in the
Jacob’s Boat Harbour and Sister’s Beach area, NW Tasmania, and main catch taken in each.

Table 29.3 Fish taken in stone-walled tidal fish-traps (based on information supplied by P. Last, C.
Turner, I. Whitehouse, and published in Stockton 1982).
Fish type Identified in Rocky Cape samples to
date
Australian salmons (Arripidae) Yes
barracouta (Gempylidae) Yes
flatfishes (Pleuronectiformes) Yes
porcupinefish (Diodontidae) Yes
leatherjackets (Monacanthidae) Yes
mullets (Mugilidae) Yes
whitings (Silaginidae) Yes
wrasses (Labridae) Yes
trevallys (Carangidae) Yes
sting rays (Dasyatidae) ?*
black bream (Sparidae) ?+
black bream (Sparidae) No
garfishes (Hemiramphidae) No
* Vertebrae of cartilaginous fish have been identified but not yet separated into shark and skate or
ray.
+ Could be present but not yet identified, due to lack of a comparative skeleton.

Figure 29.5 A stone-walled fish-trap at Jacob’s Boat Harbour, NW Tasmania (photo Richard
Cosgrove).

Table 29.3 lists fish which could be taken in tidal fish-traps. The bones of
many of these fish occur in the Rocky Cape samples, albeit in small
numbers. However, it should be remembered that since the bones of many of
these fish are relatively fragile, they are unlikely to survive well in
archaeological contexts, and are therefore probably grossly under
represented.
Discovery of a wide range of fish in the Rocky Cape samples has altered
the view that prehistoric Tasmanian fishing was restricted to a few species.
Ecological and historical information has proved to be useful for suggesting
‘suites’ of fish likely to be taken together by different fishing methods. The
absence of some fish on the site may be as significant as the presence of
others. Preliminary results suggest that at least two fishing methods (simple
baited-box traps and some sort of tidal trap) may have been used by the
Aboriginal inhabitants of Rocky Cape.

Acknowledgements

We wish to thank the following people: Ann Nicholson, Neil Sanderson, and
Kindi Smith for sorting soil samples; Barry Blain, Richard Cosgrove, Peter
Last, Betty Meehan, Don Ranson, Fran Pease, Eric Staedler, Charles Turner,
and Ivor Whitehouse for help in Tasmania; Bill Orders, Alan Scrymgeor, and
Nicholas Thorne for supplying fish; Andrew McNee for cleaning fish
skeletons; Winifred Mumford for preparing the illustrations.

Note
1 Warehou, which belongs to the Family Centrolophidae, is confusingly called ‘trevally’ in
Tasmania, although it is not a true trevally (Family Carangidae).

References
Allen, H. 1979. Left out in the cold: why the Tasmanians stopped eating fish. The Artefact 4, 1–10.
Bowdler, S. 1980. Fish and culture: a Tasmanian polemic. Mankind 12, 334–40.
Bowdler, S. 1984. Hunter Hill, Hunter Island. Terra Australis 8. Canberra: Department of Prehistory,
Research School of Pacific Studies, The Australian National University.
Colley, S. M. & R. Jones 1987. New fish bone data from Rocky Cape, north west Tasmania.
Archaeology in Oceania 22, 41–67.
Horton, D. R. 1979. Tasmanian adaptation. Mankind 12, 28–34.
Jones, R. 1971.Rocky Cape and the problem of the Tasmanians. Unpublished PhD dissertation,
Department of Anthropology, University of Sydney, Australia.
Jones, R. 1977. The Tasmanian paradox. In Stone tools as cultural markers, R. V. S. Wright (ed.),
189–204. Canberra: Australian Institute of Aboriginal Studies.
Jones, R. 1978. Why did the Tasmanians stop eating fish? In Explorations in ethnoarchaeology, R.
Gould (ed.), 11–48. Albuquerque: University of New Mexico Press.
Jones, R. 1980. Different strokes for different folks: sites, scale and strategy. In Holier than thou, I.
Johnson (ed.), 151–71. Canberra: Department of Prehistory, Research School of Pacific Studies,
The Australian National University.
Lourandos, H. 1970. Coast and hinterland: the archaeological sites of eastern Tasmania. Unpublished
MA dissertation, Department of Prehistory & Anthropology, Australian National University,
Canberra.
Mellars, P. 1984. Review of J. Flood: the story of aboriginal Australia and its people’. Antiquity 58,
232–3.
Stockton, J. 1982. Stone wall fish-traps in Tasmania. Australian Archaeology 14, 107–14.
Sutton, D. G. 1982. Towards the recognition of convergent cultural adaptation in the Subantarctic
Zone. Cultural Anthropology 23, 77–97.
Vanderwal, R. 1978. Adaptive technology in South West Tasmania. Australian Archaeology 8, 107–
27.
Vanderwal, R. & D. Horton 1984. Coastal southwest Tasmania. Terra Australis 9. Canberra:
Department of Prehistory, Research School of Pacific Studies, The Australian National University.
Walters, I. 1981. Why did the Tasmanians stop eating fish: a theoretical consideration. Artefact 6, 71–
7.
White, J. P. 1984. Absence of evidence? The Quarterly Review of Archaeology 5, 13.
White, J. P. & J. F.O’Connell 1982.A prehistory of Australia, New Guinea and Sahul. Sydney:
Academic Press.
30 Mutualism between man and
honeyguide
ALEX HOOPER

While early man was certainly a hunter, fisher, and scavenger, the story of
his relationship with animals in sub-Saharan Africa is by no means fully
known. There is evidence that in Egypt successful attempts were made to
tame and control gazelle, bubal and other antelopes, Barbary sheep, ibex,
and hyenas, and, at Jebel Uweinat, giraffes and ostriches, but none of these
species seems to have become fully domesticated in the way that the only
known purely African domesticates – the ass, the cat, and the Guinea fowl –
did (Shaw 1975, but see also p. 204, this volume). It has been suggested
that because domesticates like cow, sheep, goat, and pig, introduced from
southwestern Asia and perhaps elsewhere, adapted well to tropical
conditions, there was little need to find local substitutes (Ajayi 1971). Also,
an abundance of game and plant food in Africa might go towards
explaining this lack of domestication of aboriginal animals. Further
enlightenment on African approaches to animals might be found by looking
at other forms of man–animal interaction, such as man’s association with
the honeyguide.
The honeyguides, the Indicatoridae, are a family unique among
vertebrates in that its members have the ability to digest beeswax. The
greater, or black-throated, honeyguide (Indicator indicator) has, among
other specialities, the behavioural characteristic of attracting man’s attention
by its calling and movements – it chatters loudly and persistently and
flutters a short distance – and, if followed, it will repeat this performance
until it has led the man close to a bees’ nest. When the man breaks into the
nest to take the honey, the bird is rewarded by having access to the broken
honeycomb. Apparently, the bird cannot easily break into the bees’ nest by
itself because it has only a small beak. Nevertheless, species of honeyguide
that do not guide have been found to contain beeswax in the stomach. The
main food of honeyguides is insects, especially bees and wasps, and it is not
known exactly what contribution the beeswax makes to the birds’ diet
although it has been established that a bacterium (Micrococcus cerolyticus)
helps the bird to digest the wax.
A mutual-assistance relationship, or mutualism, of the kind just outlined
also obtains between the honeyguide and the ratel (sometimes known as the
honey-badger or honey-weasel), Mellivora capensis. Friedmann reported
that baboons also sometimes respond positively to the bird’s guidance, but
that monkeys and mongooses, although known to be solicited by the bird,
have never been reliably observed to co-operate in the venture (Friedmann
1955). No instances of mutualism between honeyguide and chimpanzee or
gorilla have been recorded, although it is reasonable to assume that the bird
would approach these animals from time to time. Formerly, it was supposed
that a discrepancy existed in the ratel–honeyguide association in that the
former was known to be active by night while the latter is diurnal. However,
Friedmann established that the ratel will operate diurnally and also that it
will climb suitable trees in search of honey, although it normally prefers to
proceed on the ground.
It is not known whether, in the beginning, man observed the honeyguide–
ratel association and then imitated the ratel, whether the bird instigated the
partnership with man by its soliciting activities, or whether some other
course of events brought about the mutualism between them. Whatever the
exact origin of the relationship may have been, some weight is lent to the
man-emulating-ratel possibility by the fact that modern man on the honey
quest may, in order to stimulate the honeyguide into action, mimic with his
voice the ratel’s guide grunts and, by knocking on trees, the noises the ratel
makes breaking into bees’ nests (see Friedmann 1955, whose Zulu guide
performed this mimicry and described his intentions). The bird tends to be
solitary in its daily habits, not often mixing with others of its own kind, and
this together with the brood-parasitism of its reproduction process (i.e. the
honeyguides do not construct their own nests or other incubation devices
but leave their eggs in the nests of other selected species, and on hatching
the infant honeyguide kills any other infants in the nest) suggests that its
behaviour in its relationships with the other animals is instinctive rather
than learned from its parents or other honeyguides. This instinct may be not
specifically directed at man, ratel, baboon, or the other known recipients of
the honeyguide’s attentions, but rather at any animal the bird meets with at
bees’ nests, to be reinforced with positive responses by the collaborating
species.
A further interesting phenomenon reported by Friedmann is that while
the honeyguide may frequent areas of permanent human settlement, such as
villages or the edges of towns, it does not try to apply its guidance routine
to persons in such settlements. Normally, it approaches individuals or small
groups in the open but will also make overtures to them in temporary
camps. This suggests that the honeyguide’s guidance of man probably was
more frequent in times past, before permanent settlement became the
normal mode of living for so many peoples in Africa.
Man’s relationship with the honeyguide seems complete. There is no
need for restraint (such as cages or leashes) or training and continual
retraining of the bird, as in the case of man’s association with the birds of
falconry or with the cormorants involved with the well-known fishing
method practised in China (Forde 1950).
It is likely that a very long time would be needed for a bird species to
evolve such a complex pattern of behaviour (Ezealor 1981). Man’s taste for
honey is known to be ancient, being documented in rock art from early
times in Europe and Africa, and it is more than probable that the
predecessors of Homo sapiens also favoured honey. In view of the low
number of millions of years that have elapsed since speciation from the
Australopithecines, it could even be conjectured that the honeyguide’s
mutualism with hominids might pre-date the emergence of the genus Homo.
However, this remains conjectural. A counter-hypothesis, equally lacking
in material basis, could be put forward that it might be a species of Homo
that originally activated the relationship with the bird, in that it would not
be until man’s emergence from the Australopithecines, concomitant with
the increase in the size of his brain and intelligence, that he would be able
to recognize that the honeyguide (presumably, already in possession of the
intestinal bacterium needed for the digestion of beeswax) was an excellent
indicator of bees’ nests. Thus, man would seek out the bird and, in some
unknown way, the bird’s current behavioural set became instinctive. Then,
mutatis mutandis, the bird would seek partners (not necessarily man) for its
raids on bees’ nests.
 Whatever the exact mechanisms may have been, it is clear that man’s
relationship with the honeyguide is a very ancient one. It is perhaps his
oldest surviving partnership in predation.

References
Ajayi, S. S. 1971. Wild life as a source of protein in Nigeria. Nigerian Field 36(3), 115–27.
Ezealor, A. U. 1981. The honeyguide: a partner in an early symbiotic relationship with man. Zaria
Archaeology Papers III, 55–7.
Forde, C. D. 1950. Habitat, economy and society. London: Methuen.
Friedmann, H. 1955. The honey-guides. US National Museum Bulletin, no. 208, Smithsonian
Institution, Washington.
Shaw, T. 1975. Why ‘darkest’ Africa? Archaeological light on an old problem. Ibadan: Ibadan
University Press.
31 Cova Negra and Gorham’s Cave:
evidence of the place of birds in
Mousterian communities
ANNE EASTHAM

Introduction

Cova Negra de Bellus at Jativa in Alicante province and Gorham’s Cave,

Gibraltar, are both sites which have produced a sequence of Mousterian
industries. No other sites belonging to the first two phases of the Würm have
been so extensively excavated in. southern and southeastern Spain, and,
although the locations of the two caves appear to be totally dissimilar,
cultural comparisons have been made (Waechter 1964, Villaverde 1984)
regarding the human occupation of the caves.
In this chapter I will attempt to make a comparison of the remains of birds
recovered from the caves and draw some conclusions regarding indications
about the environment, local habitat, and exploitation of the birds.
Cova Negra is situated on the Rio Albaida some 3 km south of Jativa and,
at the present day, 30 km as the crow flies, or 40–45 km up valley, from the
sea (Fig.31.1). At the maximum glacial periods of Würm I this distance may
have been doubled as the sea-level was lowered. The cave faces east and is
in the side of the gorge through which the Albaida flows at this stage.
Fourteen levels of Mousterian occupation have been identified.
Gorham’s Cave is situated at the base of the cliffs on the east side of
Gibraltar (Fig.31.2) (Eastham 1968). The earliest Mousterian deposits rest
on the Monastirian II beach. Above this are eight levels of Mousterian
industry, interspersed with sterile layers (Waechter 1951, Zeuner 1953).

Birds in Mousterian communities

The specific identifications of the birds in Würm I show that at Cova Negra
there was a variety of duck, all freshwater dabbling species, a number of
small hawks and owls, a proportion of partridge, some quail, several
thrushes, swifts, and chough. The commonest species throughout was the
rock dove, whose bones numbered nearly 56 per cent of the total.
The other most frequent species were mallard, partridge, Alpine swift, and
chough. All the birds recovered from these levels may be described as
‘temperate’ rather than ‘cold’ species.
Figure 31.1 Map of Cova Negra.
 Figure 31.2 Map of Gorham’s Cave, Gibraltar.

At Gorham’s Cave, during Würm 1, the species are different but the
pattern remains the same. There is a group of aquatic birds, but these include
cliff and coastal species such as the shag, gulls, great and little auks, red-
throated diver, coot, and ducks, all of which are able to live in estuarine,
lagoon, and coastal waters and some which are more common in temperate
and boreal regions today. The predators include Bonelli’s and white-tailed
eagles, black kite, griffon vulture, and falcon, all of which are compatible
with the different localities on the rock. Partridge, rock dove, and chough are
the most common species throughout all levels, following the pattern of
Cova Negra.
Plant remains from the cave include pine-cones, probably Pinus pinea,
Erica, and box. There are no so-called ‘cold’ inland species, except a single
femur of snowy owl (Nyctea scandiaca), which is also found on a number of
sites during the Middle and Upper Palaeolithic of south and southwest
France. There is also a humerus and ulna of an owl very much larger than
the tawny owl (Strix aluco) and close in size to that of the Ural owl (Strix
uralensis).
During the Würm I–II interstadial, because of the rise in sea-level, human
occupation ceased at Gorham’s Cave; the two birds, raven and chough, could
be intrusive or have been sedimented naturally. But at Cova Negra the
sequence continues and, in fact, there is no change in the bird population:
duck, partridge, pigeon, swift, and chough are the predominant species and
there is little to indicate any change of climate. The pattern is the same as
during Würm I.
It is in the Würm II deposits of Cova Negra that a single coracoid of shag
(Phalocro coraxaristotelis) was recovered. Since Cova Negra is quite a
distance from the sea, and the shag, unlike the cormorant, is essentially
marine, it is difficult to reconcile this species with the environment of the
area. To do so demands acceptance of a hunting range of up to 70 or 80 km
which implies an expedition of 2–4 days, depending on whether the party
was male members only or included women and children. There is always
the possibility that the shag at Cova Negra and Gibraltar were individuals
kept in captivity and trained for fishing, a supposition that can only be
speculative. Alternative possibilities are to question whether the present
marine habitat of the species was always its only environment, or whether
this specimen was transported back to the cave as an unusual find. For
whichever reason it was deposited at Cova Negra, it is interesting to find a
shag bone at such an inland location.
The onset of Würm II meant a lowering once more of the sea-level at
Gibraltar. But the range of bird species in these deposits is reduced; fewer
waterfowl, no waders, and a reduction in the number of predators. The
predominant species are partridge, rock dove, and chough, with a few eagle,
vulture, kite, and falcon. Even though there are fewer species, the pattern of
Würm I is retained, a pattern which is repeated in similar levels at Devil’s
Tower (Garrod & Bate 1928).
The same may be said in general terms for the Würm II deposits at Cova
Negra, but here waterfowl remain a significant element. There are fewer
partridge, in terms of numbers of bones, than in Würm I but a few snipe and
woodcock appear as well. There are a number of owls, two bones of which
are problematic. One is from level III where a single tibiotarsus of Strix sp.
appears. Like the Gorham’s Cave specimens, it is very much larger than
Strix aluco and closely matches the dimensions given for Strix intermedia by
Mourer Chauviré (1975). The other owl bone, a distal end of a humerus from
level V, has been identified as eagle owl, Bubo bubo, but it is also larger than
any recent example of this species. The measurements of this humerus match
those given for Bubo bubo davidii from the site of L’Escale St. Esteve
Janson, also identified by Mourer Chauviré (1975). However, all the material
from this site is dated much earlier, to the Mindel glaciation, so on dating
grounds the bones are difficult to equate, and the large size of these owls still
poses an interesting problem. The other species in these levels present few
difficulties. Rock doves, a number of species of thrush and corvines remain
abundant and the passerines, swift, and shrike act as seasonal indicators.
At Cova Negra it is the swifts, mainly the Alpine swift (Apus melba),
which give the clearest indication of a season when humans were certainly
present at the cave, although they may have lived there at all seasons. Swifts
are extremely useful in this respect, since their passage through an area is
limited, within a few days, to dates of arrival during the last weeks of April
and of departure during the third and fourth week in August.
Seasonal factors at Gorham’s Cave point to a possible winter occupation
on this site during the period of Würm I. Here there are no swifts or summer
migrants. The best indicators are the wildfowl, among which pochards
(Aythyini) and scoters (Melanittini), though not truly seasonal, have a
tendency to move south by sea during the late autumn, returning to northerly
inland waters to breed. The pine-cones and kernels recovered during
excavation (Waechter 1964) also point to an autumn/winter date. There is no
such firm evidence for Würm II. The isolated bone of a long-tailed duck
(Clangula hyemalis) suggests ‘cold’ conditions but there are no other
elements to support it.
The evidence from these two caves and from other sites in southern Spain
gives some indication that certain species may have had a slightly more
extensive range in Palaeolithic times than is normally ascribed to them at the
present, or has been recorded from ornithological observations. For example,
the great auk (Aha impennis) which was found in Layer K of Gorham’s
Cave, in Devil’s Tower on Gibraltar, in later Magdalenian levels of the
Cueva de Nerja (near Malaga), and at Urtiaga in Guipuzcoa, was never, in
historic times, recorded in the Mediterranean. Brittany seems to have been
its recent southernmost area of colonization in western Europe (Symington
Grieve 1895).
The snowy owl (Nyctea scandiaca), although not common in Würm II, is
found to be far more widespread in southerly latitudes during the middle
Pleistocene. It occurs in the Mousterian of Gorham’s Cave and on a number
of sites in France, mainly in the Rhône valley, Pyrenees, and in Aquitaine. It
is normally regarded as a ‘cold’ species, and in ecological terms should
reflect lower summer temperatures. However, there is always the possibility
that the frequency of the snowy owl in Palaeolithic deposits was due, in part,
to larger populations of the species, which caused their range to be extended
over a wider area.
Again, supposing that comparable migration patterns existed, there is
some indication that summer temperatures were not significantly lower
during the Pleniglacial phases of the Würm. In Spain, Alpine swifts, which
were found in equally large numbers in all levels of Würm I, the I–II
interstadial, and Würm II at Cova Negra, are also present in contemporary
deposits at the Cueva de Zafarraya, near Malaga. In France, likewise, the
Alpine swift is found on a number of sites belonging to different phases of
Würm: in Würm II deposits at L’Hortus at Valflaunes, Herault, with the red-
rumped swallow (Hirundo daurica). In the Würm IV deposits of La Balme
des Grottes, Isère, on the left bank of the Rhône, the Alpine swift is found in
association with large numbers of reindeer and a bird population dominated
by ‘cold’ species. It also occurs as far up the Rhône valley as la Grotte des
Romains à Pierre Chatel, Commune de Virignin, Ain, once again in
association with ‘cold’ species (Mourer Chauviré 1975, p. 179). And yet
today the northern limit for the Alpine swift is that of the July isotherm of
21.1°C.
Hence there is the possibility that the presence of migrants like the swift
can be doubly significant: suggesting not only a seasonal occupation but also
indicating a minimum summer temperature for the locality.

Ecological interpretations

The particular taxonomic groupings of birds found in each of these caves

depends upon suitable habitat provided by the topography, vegetation,
mammalian, and insect resources being available in the vicinity of the cave,
each species having its preferred niche within the local pattern.
At Cova Negra (Fig.31.1) the habitat types near the cave at the present
day fall into four general categories:

(a) Light woodland and scrub, with Quercus, Pinus, Olea, Certonia, Ficus,
and Myrtus; suitable habitats for jay, magpie, shrike, owls, excluding
barn owl and Tengmal’s owl.
(b) Open land with grasses and herbs, Dactylis, Thymus, Rosmaria,
Jasonium, and Erica; feeding-ground of partridge, quail, thrush,
bunting, crow, and chough.
(c) Rocky hillside and cliff, providing a roosting habitat for accepiters,
rock dove, thrush, and chough.
(d) River and river-bank at the bottom of the gorge with Phragmites and
other reeds, and tree cover, in which all waterfowl and many of the
woodland and grassland species would be found.

It would not appear to have been necessary to walk more than 5–8 km from
the immediate vicinity of the cave in order to capture all or any of the
species of bird r[illegible text]red in the Mousterian levels.
At Gorham’s Cave (Fig 31.2) there are essentially three kinds of habitat
within 5–10 km of the cave:

(a) The pine and ericaceous scrub on the summit of the rock, which would
provide a suitable habitat for the predators, partridge, snowy owl,
thrush, and corvid.
(b) The cliff-face and rocks, descending to sea-level around the cave
mouth, providing roosting and nesting facilities for shag, gulls, auks,
starling, and chough.
(c) The waters of Algeciras bay surrounded by mud flats, marsh, and reed,
and connected during the periods of low sea-level to the cave entrance
by a broad stretch of beach. Here would be found the waterfowl, divers,
heron, stork, oyster-catcher, and here the white-tailed eagle would catch
its prey, and gulls would feed.
Hunting of birds in the Mousterian

It appears that the birds found on a number of Middle and Upper Palaeolithic
sites were, in general, hunted at a distance from the cave within a maximum
radius of 5–6 km. This is certainly true, as may be seen from Figures 31.1
and 31.2, of both Cova Negra and Gorham’s Cave. It also holds good for the
cave of El Salt at Alcoy, and for Upper Palaeolithic and later occupation
sites – la Cueva de Nerja near Malaga, in eastern Spain at Volcan de Faro,
and in the northwestern Cantabrian belt – where studies of caves like La
Riera, Ekain, Eralla, Atrialda, Urtiaga, and Amalda showed that there were
suitable habitats for most of the species found in the deposits within a
relatively short walking distance of the cave, 5–6 km being a maximum
radius.
The question remains whether all the bird material in the cave was
brought in by humans. At Gorham’s Cave there is clear evidence that at
times when man was in residence, bird remains are relatively abundant and
varied. In levels where there is no sign of human occupation, the only bird
bones are those of chough. Since this cave was excavated many years ago
(1951–54), the precise location of finds was not individually recorded and
there is no indication of any area of concentration of microfauna which
might demonstrate the activities of an avian predator. At Cova Negra, by
contrast, bird bones were recovered from all levels, and the recording of
finds was carried out sufficiently precisely to be able to locate the source of
all bone material. However, this has revealed no concentrations to suggest an
owl roost inside the cave; and only one of the owl species, the barn owl (Tyto
alba) from the upper levels of Würm II, would have roosted inside the cave.
On balance, it seems that an argument favouring man as the principal
hunter on these two Mousterian sites is a valid one. Some species – swift,
swallows, choughs, and doves especially – are subject to accidental death, by
collision with the cliff-face and other hazards, and small passerines die as a
result of other circumstances, but despite these factors there are signs that
humans exploited fully the bird resources within an easy distance of their
habitation.
Comparative material from other sites on the Iberian peninsula is limited
but similar. Devil’s Tower, Gibraltar (Garrod & Bate 1928) has an avifauna
very close to that of Gorham’s Cave. The Cueva de Zafarraya, Malaga, with
a Mousterian industry, has an eagle, kestrel (Falco tinnunculus), red-legged
partridge (Alectorisrufa), Alpine swift (Apus melba), swallow (Hirundo sp.),
blue rock-thrush (Monticola solitarius), and chough (Pyrrhocorax sp.). In
the Cova Negra region there are a number of small Mousterian sites –
Cochino, Petxina (Villaverde Bonilla 1984), and El Salt – in the cliff-face
beside the steep waterfall of the River Serpis descending into the swampy
valley below the town of Alcoy. The birds of these places would have had
access to a variety of habitats, very like Cova Negra, and the very limited
excavation by Pascual produced bones of mallard, partridge, dove, and
chough.
In the north, in Cantabria, no bird studies have yet been done on
Mousterian sites, except for Amalda Cave, and it is to the important work of
Mourer Chauviré in France one must refer for the nearest comparisons.
Studies of bird faunas from Würm I–II have been carried out at Grotte du
Prince Grimaldi (Boule 1910), the Grotte de l’Observatoire, Monaco (Boule
& Villeneuve 1927), in Herault L’Hortus and La Grotte du Salpêtre, and in
Aude La Grotte Tournal. In Ariège there is the Grotte d’Aurensan, the Grotte
de Balazuc in the Ardèche, Pech de l’Aze and Combe Grenal in the
Dordogne.
Most of these sites contained species of mixed climate with some
regarded as ‘temperate’ and some ‘cold’; a mixture in which the influence of
habitat, mountain, river valley, wetland, forest, or steppe would appear to
have had a greater effect on the range of birds than strictly climatic factors.
The checklists from Pech de L’Aze and Combe Grenal show a particular
similarity to that of Cova Negra, especially in respect of the small birds and
passerines, unusual species such as the Calandra lark, the blue rock-thrush,
and shrikes appearing at Combe Grenal and Cova Negra.
However, it may be even more important to note that many of the sites
which show intensive occupation during Würm I and II, and are rich as
regards cultural remains, are also rich in bird remains in terms of quantity
and of variety of species and habitat. This is true of Cova Negra, of
Gorham’s Cave, and of Combe Grenal. And the rich bird fauna argues that
the human inhabitants had a considerable knowledge of bird behaviour and
the best means of exploiting it, even though material evidence for the
methods of exploitation has not been found. Netting in various forms seems
the most likely means of capturing many of the more difficult species such
as the passerines, the game birds, and waterfowl, although missiles such as
sling stones or bolas have been suggested; yet, above all, knowledge of the
birds and expertise in pursuing them must have been the most effective
weapon available to the Mousterian hunters of these caves.
There is certainly some suggestion at G-orham’s Cave that the Mousterian
hunters were making more use of avian resources and transporting bird
carcasses from further away than their Upper Palaeolithic successors. The
waterfowl, the crane, heron, stork, and other marshland species disappear
from the scene altogether after Würm II. Levels B and D, with their
Aurignacian- and Magdalenian-type industries, show a much more restricted
range of species, and none of those is strictly marshland.
One is drawn, therefore, to the conclusion that despite the fact that birds
formed only a minimal part of the diet of these people, the avian material of
Cova Negra and Gorham’s Cave, along with other sites of the period, does
appear to show extensive exploitation of all the bird resources which they
could find within easy reach of the homebase.

References
Boule, M. 1910. Les grottes de Grimaldi: géologie et paléontologie. Vol. I, Fasc. III: Les oiseaux,
299–331. Monaco: Institut Océanpgraphie.
Boule, M. & L. de Villeneuve 1927. La grotte de l’observatoire à Monaco. Archives de l’institute de
paléontologie. Humaine Memoire I.
Eastham, A. S. 1968. The avifauna of Gorham’s Cave, Gibraltar. Bulletin of the Institute of
Archaeology 7, 37–42.
Garrod, A. E. & D. M. A. Bate 1928. Excavations of a Mousterians rock shelter at Devil’s tower,
Gibraltar. Journal of the Royal Anthropological Institute 58, 33–113.
Mourer Chauviré, C. 1975. Les Oiseaux du Pleistocene moyen et supérieurdeFrance. Unpublished
PhD dissertation, University of Lyon.
Symington Grieve of Edinburgh 1895. The great auk or gare fowl, its history, archaeology and
remains. London: Thomas Jack.
Villaverde, B. V. 1984. La Cova Negra de Xativa y el Musteriense de la region central del
Mediterraneo Espanol. Obra Editada con la Colaboración del excelentisimo Ayuntamiento de
Xativa.
Waechter, J. D’A. 1951. Excavations at Gorham’s Cave, Gibraltar. Proceedings of the Prehistoric
Society 17(1), 83–92.
Waechter, J. D’A. 1964 Excavations at Gorham’s Cave, Gibraltar. Bulletin of the Institute of
Archaeology 4, 189–221.
Zeuner, F. E. 1953. The chronology of Gorham’s Cave, Gibraltar. Proceedings of the Prehistoric
Society 19(2), 180–8.
Index

Abarani 217
Adams, R. M. 131
Africa
cattle-herding 116
domestication of animals 22, 200–5, 347
Maasai pastoralism 215–30
pet-keeping 11
rock art 348
thorn defences 287
wild cats 53, 54
Afrikander cattle 200
agouti 307
Ahlamu/Aramu 133, 135–6, 149
Aikio, M. 179
Aikio, P. (Contribution) 169–84
Aikio, S. 177
Ainu 17
Ajayi, S. S. 347
Alaska 283, 286, 288, 292
Alberta 288
albinism 41
Albright, W. F. 132, 133
Aldunate, C. 269
Alfred the Great, King 175
Algeria 201
Allchin, B. 109, 110
alpaca 231–5, 237, 242, 244, 269, 274 See also llama
Altuna, J. 23, 65
Alur, K. R. 109, 110
Amazon Indians 11
Ambrose, S. H. 207, 208, 209
Amerindians
local deer taboo 306
pet-keeping 11, 12
Amorites 130, 131, 133
amphibians 301, 311
amulets 310
Anadara spp. 326, 327, 329, 330–3
Anati, E. 144
Anatolia 42, 98
Andagoya, P. de 296
Andaman Islands 13
Anderson, E. 283, 284
Andes 117, 240–68, 269–75
animal (s)
bones in archaeological sites 91, 122, 264–5, 284
exploitation 19, 91–5, 252, 279
husbandry 24, 26, 31, 91, 93, 105, 115, 121, 123, 216, 218
introduced 64
protein 80, 81, 293
sacrifices 110, 249
worship 110
Ankole cattle 200
Ansari, Z. D. 163, 166
antelopes 24
anthropology 15, 115, 215, 280
antler size 292
Arabia
Asian populations 132
bedouin 140
camel domestication 127, 128, 144–9
camel use chronology 148 14.2, 149
pastoralism 128, 130–1, 132
Arameans 133
archaeofaunal analysis 299
archaeozoology 1, 31, 86, 201
Arctodus sinus 282, 283, 284 25.1, 286, 289, 291–2
Argentina 269, 274
armadillo 297
Armelagos, G. J. 16
Arnade, C. W. 78
arrow-heads 95
ash-mounds 109
Assyria 130, 131, 133, 149
auk, great 353
aurochs 87
Australia(n)
Aborigines’ pets 11, 13, 18
ancient immigrants 2
archaeozoology 337
European colonization 59, 62–3
feral animals 54, 61, 62
massacres of Aborigines 62
salmon 338, 342, 29.3
sheep 62
Australopithecines 287, 348, 349
Avner, U. 130, 140, 141
Aymara 243
Aztecs 247, 249, 288
Azzaroli, A. 46

baboons 347
badgers 46
Baikal, Lake 175
Bailey, G. N. 66, 318
Balearic Islands 56. See also Mallorca
Ballysadare Bay 321, 322 27.2
Baluchistan 201
Banks, K. M. 200, 204
barley 63, 64, 163
barn owl 355
Bar-Yosef, O. 84, 98
Bate, D. M. A. 352, 356
Bates, H. W. 11
Bauer, K. M. 39
bear(s)
black 282, 286, 291
in California 287–8
curiosity 288–9
European brown 291
grizzly 283, 285, 286, 287, 288, 291 -hunting 288
interaction with human beings 287–91
killed by Spaniards 288
as pets 17
Rancholabrean 282, 284, 285 25.1
weapons against 289–91
bedouin 130, 133, 140, 143
bees 347–9
Beit-Arieh, I. 141
bell-wether 185
Beiz, L. H. 329
Bennett, C. F. 306
Berg, L. van den 35
Berlin, B. 295, 301
Berlin, E. A. 295, 301
Berntsen, J. L. 209
Biagi, P. 95
Binford, L. R. 66
birds
ancient interaction 280
honeyguides 347–9
migration 38–9, 354
Mousterian hunting of 355–7
passerine 301, 355–6
 as pets 12, 14, 15, 16
Bischoff, G. L. 283
bison 285, 292
Blanco, A. 299
Bloom, A. L. 328
blowguns 309, 311
B’Mbuti pigmies 13, 61
boar-spears 290
Boessneck, J. 67, 84, 86
Bökönyi, S. 7, 8, 65, 66, 81–2, 84, 85, 274. (Contribution) 22–7
bolas 356
Bolivia 235
bone artefacts 309
Bos africamts 201
Bos indicus 100, 108
Bos primogeuius 201
Bos taurus 100
Bossena, I. 40
Bottema, S. 8. (Contribution) 31–45
Bower, J. R. 208
Bowlby, J. 16
bow tax 179
Braidwood, R. J. 99
Braley, R. D. 329
Brazil 14
Briant, P. 127
Brinkman, J. A. 131, 133
Bronze Age 106, 120, 121, 141
Brotherton, G. (Contribution) 240–55
Browman, D. L. (Contribution) 256–67
Brown, A. 301
Buddhism 111, 166
budgerigars 18
buffalo 14, 61, 108, 202
Bulliet, R. W. 127, 147, 149
bull sacrifices 110
Burenhult, G. 321, 323
burial (s)
chariot 106
goods 309, 310
mounds 120, 123, 124
of surviving spouse 164
Bushmen 8
butchery patterns 65, 66

Cain 240
Cajal, J. L. 274
Californian Indians 17
camelids 80, 117, 231–5, 242–5, 260, 261, 262, 269
camels 61, 116, 121, 134, 140, 144–9
canaries 18
cannibalism 15
carbohydrates 295
carbonization 91
Careja Indians 14
caribou 181
carnivores 282, 287
Carr, C. J. 211
Carter, P. L. 202
Carthaginians 56
castration 26, 169
Castro, L. 269
cat
domestic xi, 2, 204
feral 53, 77
sabre-toothed 282, 291
wild 2, 53–5
cattle
in Africa 3, 8, 200–4, 215–30
bones 93
 descriptives 222–5
draught 109
dung 109
and geese 39
hide 109
Hindu slaughter taboo 111
humped 108, 200
long-horned 200, 202
milk 201, 204
names 219–22
in Nile valley 203–4
origin of domestic 100, 200
recognition and identification 217–30
shell-midden evidence 320, 321
udder size 31
Ceci, L. 316
cereals
in Indian sites 109
Neolithic 23
shell-midden evidence 321
in Sultanian sites 92
wild 94
in Zagros region 97
Chalcolithic sites 108, 109, 110, 158, 163, 166
Chamber culture 121 13.1
Chang, C. 127
Channel Islands 283
charki 263–4
chernozem 122, 124
Childe, V. Gordon 81, 92, 99
Chile 50, 235, 243, 269, 274
China 9, 105–7, 175, 348
chough 355, 356
Chow Ben-Shun 9. (Contribution) 105–7
Chuguryaeva, V. A. 120
‘chunking’ 228–9
Churcher, C. S. 203, 285
Cimmerites 119–20
Cipriani, L. 11, 13
Clark, J. G. D. 316, 318, 323
Clary, K. H. 295
Clason, A. T. 148
Clegg, J. 61
Cleuziou, S. 148
climatic changes 100, 156–67
Clutton-Brock, J. 10, 17, 22, 23, 80. (Contributions) 1–3, 7–9, 115–18, 200–5, 279–81
cognitive processes 225–8
Colla 243, 246, 252
Colley, S. M. (Joint contribution) 336
Colombia 243, 301, 309
Compagnani, B. 81, 147, 149
concentrate feeding 292
Cooke, R. G. (Joint contribution) 295–315
cormorants 12, 14, 348, 352
Corsica 3, 47, 51, 54, 55, 56, 84
Costa Rica 309
Cova Negra 350–7
Craighead, F. C. 287
cranial capacity 47–57
Crete 3, 46, 50, 56, 57
crocodiles 205
crows 42
cuneiform records 127, 149, 198
Cuzco 242, 243, 247, 253, 261
Cyprus 3, 47, 49, 57

Dahl, G. 211, 212

Dallas 28
dart poisons 309
Davidson, I. (Contribution) 59–71
Deccan
British conquest 167
burials 164
desertion of farming settlements 157, 158
Jorwe house-types 159–61
location 156
megalith builders 163, 166
religion 163–4
decoys 32, 33 4.1 35
deer 3, 46, 56, 296
deglaciation 283
Delacour, J. 32, 35, 41
Denmark 317–18, 323
Dcnnell, R. W. 59–60, 63
dental indices 163
Deo, S. B. 109, 110, 158
desertification 3, 205
de Villeneuvc, L. 356
Dhangans 166
Dhavalikar, M. K. 109, 116. (Contribution) 156–68
dhole 286
dietary taboos 14, 252, 306
diffusion 63, 64
dingo 2, 13, 61
dogs
adopted by Tasmanians 62
Andamanese and Dyak devotion to 13
Comanche 14
domestic 1, 2, 23
feral 2
Hawaiian 12–13
hunting 13, 108, 290, 308–9
as meat 108
in Mesolithic 279
reindeer-herding 179
 sheep- 185
as watch animals 108
working 13
domestication
alpaca 24
ass 204, 207
camel 17, 128, 144–9
cat 53
cattle 93, 201–5, 207
cultural and biological process 7–9
definitions 2–6, 22–4, 28–30, 31, 81, 240, 274–5
demographic evidence 82–4
food as control mechanism 43–4
Galton’s theory 10
goat 97, 207
horse 94
mallard 43
morphological changes 25–6, 31–5, 85–7, 91, 147, 264–5
Neolithic achievement 93, 279
osteological evidence 80–7, 147, 148–9
and pastoralism 240
reindeer 169, 275
selective breeding 26, 86
shell-midden evidence 320, 321, 323
two-part, two-stage model 97–103
zoogeographic evidence 84–5
Donner, Kai 175
Dostal, W. 147, 149
dove 355
dove-netting 296
Duck, S. 16
ducks 40–2, 43
Ducos, P. 7, 22, 23, 24, 67, 81, 83, 101, 240, 274. (Contribution) 28–30
Dumbrell, W. J. 135
dwarf cattle 202
Dyson, R. H. 83, 84

Eastham, A. (Contribution) 350–7

East India Company 167
Egypt
Ahlamu agitation for food 133
ancient attempts at domestication 24, 347
Asians’ presence attested 131–2
conquest by Hyksos 120
domestication of ducks 43
Middle Kingdom 131, 132
New Kingdom 133
no early term for ‘camel’ 149
Old Kingdom 130
representations of cattle 201, 203
Second Intermediate Period 127, 132
Shosu references 134–5
Transjordan contacts 134
elephant 286
Elgin 318
elk 176
Elmendorff, W. W. 11. 17
Elphinstone, Mountstuart 167
Equisetum spp. 170
Equus przewalskii 105, 110
Ertebølle sites 317–18, 323
Estévez, J. 65, 66
ethnoarchaeology 140
eunuchs 187–8, 198
Evans, I. H. N. 11, 13
Evans-Pritchard, E. E. 115, 208, 211, 215
Evins, M. A. 101
Execration Texts 127, 131, 132
Ezealor, A. U. 348
Fabri, C. L. 110
falconry 348
fallow deer 3
Feldman, L. H. 308
feral animals 2, 25, 53, 77
Ferghana horses 106
Fertile Crescent 93, 94, 128
Field, H. 139
Fildes, V. 15
fish
eating 301–2, 336–7
exploitation 94
payment of taxes 178
traps 341–5
fisher–gatherer–hunters 59–60, 68
fishing
possible use of shag 352
postulated in Alta rock-drawing 176
use of cormorants 14, 348
Flannery, K. V. 161
Fleming, P. 12
flint industry 93
Flipchenko, V . A. 122
flock leaders
Afghanistan 190, 197
Central Asia 190
Crete 189, 192
Greece 189, 1%, 197
Italy 187, 188–9
Romania 190–1, 192, 193–6
Turkey 189–90
Flores Ochoa, J. 231, 269
Florida 8, 72
fodder 23
Foley, R. A. 209
food
domestication control mechanism 43–4
habits 287
taboos 14, 252, 306
forest fires 295
Fortea, J. 64, 65, 67
Forth Valley shell-middens 318, 319 27.1, 320, 323
France
Alpine swift habitat 354
bird studies 356
Mesolithic sites 67
pet-keeping 11
snowy owl habitat 353
Franklin, W. L. 273, 274
Friedmann, H. 347, 348
frogs
Balearic 56
in diet 309
as pets 11
Fulani cattle 200

Gadgil, M. 166
Gafrarium spp. 326, 327, 329, 330–3
Galaty, J. G. 127. (Contribution) 215–30
Galton, F. 10, 11, 12, 13, 17, 18, 19
game management 83
‘garden-hunting’ 295, 308, 311
Garran, J. C. 61
Garrod, A. E. 352, 356
gazelles 23, 83, 87, 94, 101, 163
Geddes, D. S. 64, 67, 84, 95
geese
barnacle 38
Brent 37
domestic 44
 Egyptian 40
greylag 32
pink-footed 38, 39
taiga bean 35, 36 4.3 white-fronted 32
Geist, V. (Contribution) 282–94
genetic drift 41–2
Gibraltar 350–7
Gifford-Gonzalez, D. P. 209, 210
Gilmore, R. M. 135, 140
Glasgow 318
glaucous gulls 39
goats
castrated 190
domestic 50, 51
domestication in Fertile Crescent 93
feral 8, 50, 77
herding 205
horncore changes 85
in Neolithic 2
pathology 85–6
in sheep flocks 186, 190
shell-midden evidence 321
wild 50–1, 87
Golmsten, V. V. 119
Gourlay, R. 318
grasses 292
Greece 189, 192, 196, 197
Greenland 174
Grigson, C. 144
grinding-stones 99
Groube, L. M. 326
Groves, C. P. 8, 65. (Contribution) 46–58
Gryaznov, M. P. 119
guanaco 242, 274
Guiana Indians 14
guide wether 187
Guinea fowl 204, 347
Gundcrman, K. H. 269, 274
Gune, V. T. 167
Guthrie, R. D. 283

habitat(s)
Cova Negra 354–5
man-made 307
modification 302–9
open 306
Hahn, E. 119, 156
Haiti 8, 72
Hakker-Orion, D. 143, 147
Hall, M. J. 59
Hallander, H. 38
Hallo, W. W. 130
Haltenorth, T. 53, 54
Hames, R. B. 308
hamsters 18
Harappa 109
harem 187
hsres 46
Harlan, J. R. 66
Harris, Cornwallis 8
Harrison, T. 11, 13
Hawaii, 12–13, 14
Hiyes, W. C. 132
Hlynes, C. V. 283–4
Hocker, H. M. 22, 80, 93
Hediger, H. 23, 31
Helbaek, H. 63
Hoick, W. W. 134
Helskog, K. 175
Hemmer, H. 47, 54, 56
herd-following 117
Herodotus 202
heron 357
Hrrre, W. 7, 24, 31, 35, 43, 47, 66, 123
Herrero, S. 287, 288
Hesiod 119
Hesse, B. C. 83, 93
Higgs, E. S. 26, 29, 66, 80
Hill, K. 301
Hilzheimer, J. 7, 101
Hinduism 111
Hittites 133
Hoch, E. 147
Hoffman, P. R. 292
Hole, F. 9, 85, 93. (Contribution) 97–104
Homer 119
honey 213 n. 1
honeyguide 347–9
Hooper, A. (Contribution) 347–9
Hopwood, V. G. 288
Horse Latitudes 78
horses
in China 105–7
Comanches’ attitude 14
on Don steppes 124
in Haiti 77 7.6, 77–8
in India 163, 166
milked by Scythians 119
horticulture 61, 295, 308, 326
Horton, D. R. 336
howler monkeys 309, 311
Hoysalas 167
Hudson’s Bay Company 288
Hugh-Jones, C. 12, 301
human
 anti-predator strategies 279, 286–7
co-existence with Arctodus 284
colonization of Alaska 283
colonization of Australia 2, 283
encounter with Siberian megafauna 282–6
interaction with bears 287–91
notional supremacy 282
suckling of pets 12, 13, 15–16
Hungary 24
hunter-gatherers 16, 121, 156, 279, 280, 308 See also Fisher–hunter–gatherers
hunters and hunting
definition 117
pet-keeping 12, 16
possibly sedentary 323
post-Pleistocene 94
pre-Columbian 295–301, 302–6, 308–11
Upper Palaeolithic 291–2
hyenas xi, 24
Hyksos 120, 132

ibex 65, 102

Iceland 174
iguana 306, 308
Inamgaon 158, 159, 163
inbreeding 41–2
Incas 117, 234, 240–54
incest 15
India 108–112, 156–67, 190
Indus valley 147
Ingold, T. 17, 115, 117, 209, 275
Ingraham, M. 143, 144
interbreeding 65
interpretation
of rock drawings 183
of Wilhelm medallion 185
Inuits 13, 61, 290
Inverness 318
Iran 190
Ireland 321, 322 27.2, 323
Iron Age 110
Irons, W. 127
Israel 18, 56, 84, 143
Italy 56, 187, 188–9
Itkonen, T. I. 176

jade figures 106

jaguar 282
Jainism 111
Jamaica 77
Jarman, M. R. 80, 83, 86
Japan 17
Jensen, J. 323
Jericho 23, 98
Jesse, G. R. 11, 13
Jobling, W. J. 141
Johnson, E. & P. 284
Jones, R. 62. (Joint contribution) 336–346
Jordan, B. 87
Jorwe cultures 108, 156–64
Joshi, R. V. 108
Juan, G. 12
Juan Fernandez 50

Kalapalo Indians 12, 16

Kalmyks 123
kangaroos 63
Kaposhina, S. I. 123
Kassite period 130, 136
Kenya 207, 208
Khazanov, A. M. 115, 127, 156
khoton 123
King, J. 60–1, 63
Klein, R. 282, 299
Köhler-Rollefson, I. 86, 147
Kortland, A. 287
Kosambi, D. D. 166
Kozloff, B. 140
Krishnamurthi, R. V. 156
Kruska, D. 47
Kuo Mo-jo 106
Kupper, J.-R. 130, 133
Kurtén, B. 286, 287

laboratory animals 18
Lagodovskaya, O. V. 121, 124
Lancaster, W. 127
Lapita cultura 326–34
Lapps 169, 275
lapwing 42–3
Larsen, C. E. 135
lasso 175
Latynin, V. A. 121
Laysan duck 41
Leach, E. 11, 15
leatherjacket 338, 340
Lebret, T. 32
Legge, A. J. 23
Leonard, Zenas 289
Lernau, H. 147
Lévi-Strauss, C. 15
Lewis, B. A. 211
Lewthwaitc, J. 63, 64
Lhotc, H. 200, 202
lichen 170, 181
Linares, O. F. 295, 296, 310, 311
Linton, R. 13–14
Li You-heng 105
llama
as beast of burden 231, 235
behaviour 273–4
breeding 269–74
domestication questioned 269, 274–5
dung 231, 235
functions 244
herding system 234, 269, 270 24.1, 271 24.2, 273–4
human interaction 251–2, 259
in Inca culture 232–4, 237
military use 244
puma predation 272
revered as tribal ancestor 243
ritual use 231, 232–4, 252–3
territorial range 234–5
wool 231–4
See also alpaca
Lothrop, S. K. 310
Lubbock, Sir J. 318, 320
Lucaks, J. R. 163
Luke, J. T. 130
Lumholtz, C. 13
Luomala, K. 11, 12, 14
Lynch, J. J. 16

Maasai 116–17, 215–30

MacDonald, B. 141
MacFarland, D. 13
McGreevy, T. (Contribution) 231–9
Mackie, E. W. 318
MacKnight, C. C. 61
McNicoll, A. 136
Maekawa, K. 197
magpies 44
Maikop culture 124
maize 246, 252, 295, 296, 297
mallard 32, 40–2, 44
mammoth 280, 283
Manco Capac 243
man-eaters 217
mange 234
mangroves 306
Marti, B. 64, 67
Martin, L. D. 285, 291
Martin, P. S. 282, 283, 292
mastodon 280, 283
maternal malnutrition 86
Matthews, V. 131
Maya 308, 309
Meadow, R. H. 102, 115, 201, 274. (Contribution) 80–90
Mediterranean islands 46–57
megafaunal extermination 280, 286
megaliths 110, 163, 166
Melentiev, A. N. 122
Mellars, P. A. 317, 336
Merlino, R. 272
Merrick, H. V. 208
Mesolithic culture 2, 317
Mesopotamia 130, 131, 133, 136
mice 18, 304 26.4
Micrococcus cerolyticus 347, 349
microliths
absence in Ertebølle sites 318
in Ryn Sands sites 122
Micronesia 11
Middle East 127–49
Midian(ites) 127, 135, 148, 149
milk 83, 109, 116, 119, 292
mink 40
Minoans 46
Moab 135
Mohenjadaro 109
molluscs 316–24, 326–34
Monaco 356
Mongolian wild horse 106
Monod, T. 208
Montigny, A. 140
Moore, A. M. T. 97, 98, 101
Morales, A. 64–5
mother goddess figures 163–4
mouflon 47–9, 55
Mourer Chauvire, C. 353, 356
Mousterian culture 280, 350–7
Mulvaney, D. J. 61
Muñoz Amabilia, A. M. 67
Musil, R. 25
mute swans 44
mutualism 347–9
Myers, A. 318

names and descriptors

Maasai cattle 219–25
Romanian sheep 191
Sàmi reindeer 169, 170, 172 16.1
Natufian culture 92
Neanderthal man 291
Negev/Sinai 127, 132, 135, 141
Nelson, R. K. 287, 288
Neolithic revolution 17, 81, 240
Nerem, R. M. 16
Netherlands 35, 42
Nietschmann, B. 308
Nobis, G. 23
Noddle, B. 86
nomadism 18, 95, 115, 116–17, 119–24, 127–31
Norway 173, 174, 175
Novaya Zemlya 35
Nuer 208, 211

O’Connell, J. F. 336
Ohthere 175
onagers 23, 98
Ondrack, J. 288
Oren, E. D. 132
Osterholm, I. 321
Ostrea edulis 316, 317, 324
ostriches 140, 347
overgrazing 3
overkill 280, 282
oysters 280, 316–24

pair-bonding 32
palaeoarchaeologists 16
Palaeolithic–Neolithic transition 91–5
palm-nuts 295
Panama
deer taboo 306
pre-Columbian hunting 295–311
predation on invertebrates 280
Spanish conquest 296, 299
Panaulauca 258
Panthera leo atrox 282, 286
Papua 61
papyri 127, 134, 149
parasitism 279
Parr, P. J. 135
parrots 11, 12, 14
pastoralism
 and agriculture 156, 240
Andean 240–54, 256–65
Biblical resonances 240, 247, 253
cognitive concomitants 215–16
definition 118
logic and experience 216–18
pre-Hispanic Peruvian 231–7
‘pure’ 118, 235, 237
Payne, S. 83, 101
peat bogs 35
peccary 307–8, 309
pecking order 39
Pees, W. 47–8
Peplau, I. A. 16
Pericot, L. 67
Perlman, D. 16
Peru 231–7, 295
Petra 134
pet(s)
anthropological research 10, 12
as child-substitutes 19
definition 10–11
emotional benefits 16–17
function 14–17
Galton’s theories 10, 17
human suckling 12, 14, 15
in Palaeolithic 17, 18
sacrifice and slaughter 17
in West 19
Philip III, King 242
Phoenicians 56
pigs
feral 61, 77, 84
in Neolithic China 105
Palaeolithic and Neolithic farming 23
 shell-midden evidence 321
wild 51–3, 84
Pimpalsuti 163, 166
Piperno, D. 295
Pit-grave culture 121–2, 123, 124
Pizarro 246, 249
plant cultivation 91, 92
Pohl, M. 308
Polo, Marco 175
Poltavkin culture 122
Polynesia 11, 13, 326
Poma de Ayala, Guaman 242, 243, 244, 248, 252
Poplin, F. 87
Posener, G. 132
Post, L. van der 287
potlatching 306
pottery
Jorwe 158, 161, 163, 164
shell-midden evidence 320, 323, 326, 333
Tongatapu 326
Potts, D. 135
Poulain, T. 67
Poulsen, J. I. 326, 327, 331
predation 279–81
proto-Sinaitic script 141
Prummel, W. 43
pulses 94, 109
puma 272

Qatar 140, 141

quails 40
Quechua 242, 247, 248
Quechua yaravi 252–3
quipu 242, 244, 248
rabbit 3, 61
Rabey, M. A. (Contribution) 269–76
raccoon 306, 311
Raedecke, K. J. 274
ranchers 117, 216
Ranere, A. J. (Joint contribution) 295–315
ratel 347
rats 301
Rao, S. R. 158
Rauh, H. 84
red deer
in early Neolithic sites 100
shell-midden evidence 321
temperature-related size diminution 87
Redford, D. B. 132, 134
red jungle fowl 43
Reed, C. A. 85
reindeer
carcass size 179
castration 169, 175
domesticated 169, 275
ear-notching 169, 175
economic analysis 66
large herds 177
individual names 169, 172 16.1
meat 179, 180 16.8
origin of herding 174
as pets 17
racing 174
Sámi terminology 176
in Scotland 174
semi-domesticated draught 168
slaughter 172, 173
taxes 177–8
in Würm IV deposits 354
reptiles 301, 305 26.5
return to wild 31
Reynolds, H. 61
Richardson, Sir J. 14
rinderpest 201
Ripulsky, M. 144
rites of passage 232
Robertshaw, P. T. (Contribution) 207–14
Robinson, J. G. 308
rock art 142–3, 144 14.1, 176, 183, 200, 202, 348
rockshelters 110, 297
Rocky Cape 336–45
Roe, F. G. 287
Rogers, R. A. 283
Rörs, M. 22
Romans 3, 43
rootcrops 297
Rosen, S. A. 140, 141
Ross, C. 288
Roth, W. E. 12, 14
Rothenberg, B. 140, 141, 149
Rouse, J. 283, 286
Rousseau, Jean-Jacques 119, 240
Rowley-Conwy, P. 318, 323, 324
Rowton, M. B. 130, 131
Rozoy, J. G. 67
Ruddman, W. F. & J. C. 283
Rudenko, S. 119
Rykov, P. S. 122

saddle querns 109

Sahlins, M. 15, 215, 280
Sámi 169–84
Samoyeds 175
Sanga cattle 200, 201
Sankalia, H. D. 109
Saqqara 203
Sardinia 3, 47, 53–4, 55–6
Sauer, C. O. 10
Savishinsky, J. 13
scarabs 134
Schauenberg’s index 54, 55 5.8
schlepp 263
Schmidt, W. 119
Schrire, C. 62
Schultze-Westrum, T. 50
Schwartz, J. H. 46, 57
Scotland
reindeer 174
shell-middens 317, 318, 319 27.1, 320–1, 323–4
Scythians 119, 120
sedentism 94, 98, 99
selective breeding 26, 86
Semang Negritos 13
Serpell, J. A. (Contribution) 10–21
sexual taboos 14–15
shag 352
shamanism 252, 288
sharks 343 29.3 n.
Shaw, T. 204, 347
Shawcross, F. W. 66
sheep
Anatolian domestication 97
bighorn 286
domestication in Fertile Crescent 93
escaping 61–2
fat-tailed 61
feral 8, 25, 47, 84
fertilization of arable 166
flock control 185–98
 herding 189–96, 205, 272
horncore changes 85
Inamgaon evidence 163
landscape preference 100
milk 116, 189, 191
sacrifices 110
Soay 49
wild 47–9, 84, 87
sheikhs 134
shell-middens
Denmark 317–18, 323
Ireland 321, 322 27.2, 323
modern 331
Red Sea coast 148
Scotland 317, 318–21, 323–4
traditional view 316, 317
shellmounds 297, 299
Shilov, V. P. (Contribution, trans. K. Judelson) 119–26
Short, R. V. 15, 19
Shoshonee Indians 288
Shosu 134–5, 149
Siberia(n)
pet-keepers 17
Pleistocene fauna 282–3
Singer, M. 13
Situa 247, 248
size diminution 86
skis 175
slaves 198
sling-stones 309, 356
Sloan, D. 9. (Contribution) 316–25
Smith, A. B. 119, 200, 203, 204
Smythe, N. 307
snowmobiles 173
soils 157–8
Sondaar, P. Y. 46
Songaon 158, 159
South Georgia 174
Sowls, L. K. 308
Spain
agriculture 59–68
avifauna 356
butchery patterns 66
domestication evidence 67
Mousterian communities 350–7
Pleistocene fauna 64
wild pigs 51, 55–6
Spanish invaders of South America 243, 246, 296
spears 309
Speke, J. H. 11
Spennemann, D. H. R. 280. (Contribution) 326–35
Spitzbergen 39
Stampfli, H. R. 85, 93
Stockton, J. 342, 343 29.3, 29.4
stone circles 110
stone tools 92
Storer, T. J. & L. P. 287, 288
Strabo 119
‘strand-looping’ 316
structuralism 14–15
Sultanian culture 92
Sutton, D. G. 336
swallow 355
Sweden 38, 173, 174
swift 355
Switzerland 24–5
symbiosis 24, 41, 43, 279, 280
sympathetic magic 252

Tahuantinsuyu 241–7, 248–54

Tambiah, S. J. 15
taming 7, 12, 275
Tani, Y. (Contribution) 185–99
Tanzania 207
Tasmania 280, 336–45
Taylor, F. W. 328
Tell el Amarna 127, 132, 133
temperature 87
terracotta horses 106
Thessaly 87
Thomas, D. 287, 288
Thomas, P. K. (Contribution) 108–112
tourists 174
transhumance 97, 99, 118, 190
trevally 348
n. 1
truffles 61
tsetse fly 205
Tungus 17, 175
Turkey 50, 189–90
Turnbull, P. F. 23
Turner, C. 342
turtles 301, 310
typha grass 159

Uerpmann, H. P. 9, 47, 83, 86, 87, 280 (Contribution) 91–6

Uganda 200
Ulloa, A. 12
United States of America 11
Ur III 130, 197, 198
urn burials 164
USSR 174

Valdez, R. 47
Valla, F. R. 18
Vanderwal, R. 336
vegetable fodder 23
Venezuela 283
Vereshagin, N. K. 286
verticality 235
Vickers, W. T. 308
vicuna 242, 248, 251, 252, 253, 265
Vigne, J.-D. 84
Vijayanagar kings 167
Vogt, F. 292

Waechter, J. d’A. 350

Walters, I. 336
wapiti 285, 286
waterfowl 8, 357
Weiland, D. 295, 306
Weiler, D. 149
Wendorf, F. 200
wether 185
White, J. P. 336
White, R. S. 295, 310
Whitehouse, I. 342
white-tailed deer 302–6
Widdowson, E. M. 86
Wiklund, K. B. 175
Wilbert, J. 12
Wilkinson, M. 317
Wilson, E. O. 279
Wilson, J. A. 132
Wilson, M. 285
Wing, E. 301. (Contribution) 72–90
Woburn 37
wolf
dire 282, 286, 291
diversity of human response 281
 domestication 2
fossil skulls 25
as pet 11, 12
reaction of reindeer 171
temperature-related size diminution 87
timber 285–6
Woodburn, J. 60
Woodman, P. 317, 323
wrass 338
Wright, H. T. 83
Wright, W. H. 287, 288, 289
Würm deposits 350–7

xerophytic scrub woodland 306

Yadavas 167
Young, F. H. 287, 288, 289
Yukon Territory 283

Zaire 13
Zarins, J. (Contribution) 127–55
zebu 200, 201 See also Bos indicus
Zeuner, F. E. 7, 10, 80, 85, 101, 109, 274, 350
Zhitesky, I. A. 123

zooarchaeology 1, 31, 86, 201

zoocresis 66

Table and figure numbers are printed in italics after the number of the page
on which they occur.