Wendrich y Barnard 2008-The Archaeology of Mobility

The Archaeology of Mobility

COTSEN ADVANCED SEMINAR

Volume 5 The Construction of Value in
the Ancient World
Edited by John K.
Papadopoulos and Gary
Urton

Volume 4 The Archaeology of Mobility:
Old World and New World
Nomadism
Edited by Hans Barnard and
Willeke Wendrich

Volume 3 The Archaeology of Ritual
Edited by Evangelos
Kyriakidis

Volume 2 Agricultural Strategies
Edited by Joyce Marcus &
Charles Stanish

Volume 1 Theory and Practice in
Mediterranean Archaeology:
Old World and New World
Perspectives
Edited by Martin Peilstöcker
and Aaron A. Burke

For a complete list of our titles, please visit the Cotsen Press website.
T he A rchaeology
of Mobility
Old World and New
World Nomadism
Edited by
Hans Barnard and Willeke Wendrich
Cotsen Institute of Archaeology

University of California, Los Angeles
The Cotsen Institute of Archaeology at UCLA is a research unit at the University of California, Los
Angeles that promotes the comprehensive and interdisciplinary study of the human past. Established in 1973,
the Cotsen Institute is a unique resource that provides an opportunity for faculty, staff, graduate students,
research associates, volunteers and the general public to gather together in their explorations of ancient human
societies.
Former President and CEO of Neutrogena Corporation Lloyd E. Cotsen has been associated with UCLA
for more than 30 years as a volunteer and donor and maintains a special interest in archaeology. Lloyd E. Cotsen
has been an advisor and supporter of the Institute since 1980. In 1999, The UCLA Institute of Archaeology
changed its name to the Cotsen Institute of Archaeology at UCLA to honor the longtime support of Lloyd E.
Cotsen.
Cotsen Institute Publications specializes in producing high-quality data monographs in several different
series, including Monumenta Archaeologica, Monographs, and Perspectives in California Archaeology, as well
as innovative ideas in the Cotsen Advanced Seminar Series and the Ideas, Debates and Perspectives Series.
Through the generosity of Lloyd E. Cotsen, our publications are subsidized, producing superb volumes at an
affordable price.
The Cotsen Institute of Archaeology at UCLA

Charles Stanish, Director
Elizabeth Klarich, Assistant Director
Shauna K. Mecartea, Executive Editor & Media Relations Officer
Eric C. Gardner, Publications Coordinator
Editorial Board of the Cotsen Institute of Archaeology

Jeanne E. Arnold, Christopher B. Donnan, Shauna K. Mecartea, John K. Papadopoulos, James Sackett,
Charles Stanish, and Willeke Wendrich
Editorial Advisory Board

Chapurukha Kusimba, Joyce Marcus, Colin Renfrew, and John Yellen
This book is set in 10-point Janson Text, with titles in 29-point OPTI Forquet Oldstyle.
Edited by Joe Abbott
Designed by William Morosi
Index by Robert and Cynthia Swanson
The picture on the cover was taken in 1998 by pastoral nomad Mohamed Eid, with one of the cameras provided
to the Ababda tribe by the Eastern Desert Antiquities Protection Project. It shows his family living in the Eastern
Desert, between the Nile and the Red Sea in the border area between Egypt and Sudan.
Library of Congress Cataloging-in-Publication Data

The archaeology of mobility : old world and new world nomadism / edited by Hans Barnard and Willeke
Wendrich.
p. cm. -- (Cotsen advanced seminars ; v. 4)
Includes bibliographical references and index. ISBN 978-1-938770-38-8 (eBook)
ISBN 978-1-931745-49-9 (pbk. : alk. paper) -- ISBN 978-1-931745-50-5 (cloth : alk. paper)
1. Nomads. 2. Human beings--Migrations. 3. Land settlement patterns, Prehistoric. I. Barnard, H. II.
Wendrich, Willemina. III. Cotsen Institute of Archaeology at UCLA. IV. Series.
GN387.A73 2008
305.9'0691--dc22
2008015350
Copyright © 2008 Regents of the University of California

All rights reserved. Printed in the USA.
C ontents
1 The Archaeology of Mobility:

Definitions and Research Approaches . . . . . . . . . . . . . . . . . . . . . . 1
Willeke Wendrich and Hans Barnard
Definitions: Moment, Movement and Motivation . . . . . . . . . . . . . 3
Terminology Related to Hunter-Gatherers. . . . . . . . . . . . . . . . . . . 7
Terminology Related to Pastoral Nomads. . . . . . . . . . . . . . . . . . . . 7
Relations Between Mobile and Settled Groups. . . . . . . . . . . . . . . 10
Material Culture, Landscape and Fieldwork. . . . . . . . . . . . . . . . . 11
Ethno-Archaeology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Discussion and References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Part I: The Past at Present

2 Things to Do with Sheep and Goats:
Neolithic Hunter-Forager-Herders in North Arabia. . . . . . . . . . 25
Alison Betts
The End of the PPNB. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Models of Steppic Economic Strategies . . . . . . . . . . . . . . . . . . . . 28
Evidence from the Harra and the Hamed . . . . . . . . . . . . . . . . . . . 29
Jebel Naja. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Wadi Jilat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Qasr Burqu’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Qa’ al-Ghirqa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Dhuweila. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3 An Archaeology of Multisited Communities . . . . . . . . . . . . . . . . 43
Reinhard Bernbeck
The Mobility-Sedentary Dichotomy. . . . . . . . . . . . . . . . . . . . . . . 45
Archaeological Indicators of Mobility. . . . . . . . . . . . . . . . . . . . . . 49
The Case Study of Fıstıklı Höyük. . . . . . . . . . . . . . . . . . . . . . . . . 52
Phytoliths and Stone Tools. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Anchoring and Dispositional Practices . . . . . . . . . . . . . . . . . . . . . 61
Making Practices Visible . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Future Fieldwork and Its Problems. . . . . . . . . . . . . . . . . . . . . . . . 63
v
vi     The Archaeology of Mobility
4 Archaeology and the Question of Mobile

Pastoralism in Late Prehistory . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Abbas Alizadeh
Complexities of Seasonal Migration . . . . . . . . . . . . . . . . . . . . . . . 88
Agricultural Activities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
External Investment of Tribal Pastoral Economy. . . . . . . . . . . . . 90
Political and Military Potentials. . . . . . . . . . . . . . . . . . . . . . . . . . . 91
The Zagros Mountains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Lowland Susiana. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
The Central Plateau. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5 Desert Pastoral Nomadism in the Longue Durée:
A Case Study from the Negev and the Southern
Levantine Deserts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Steven A. Rosen
Achieving Pastoral Nomadism. . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Technological Developments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
The Impact of the Outside World. . . . . . . . . . . . . . . . . . . . . . . . . 126
6 The Origin of the Tribe and of ‘Industrial’ Agropastoralism
in Syro-Mesopotamia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Giorgio Buccellati
Historical Background:
The Urban Revolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
The Geographical Horizon:
The Steppe as a Perceptual Macroregion. . . . . . . . . . . . . . . . . . . 148
The Concomitant Economic Development:
The Nature of Industrial Nomadism. . . . . . . . . . . . . . . . . . . . . . 152
The Nature of the Evidence:
The Role of Archaeology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
7 Pastoral Nomadism in the Central Andes:
A Historic Retrospective Example . . . . . . . . . . . . . . . . . . . . . . . . 160
David L. Browman
Junin Seminomadic Agropastoralism. . . . . . . . . . . . . . . . . . . . . . . 161
Andean Pastoralism Patterns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Origins of Central Andean Pastoral Lifestyles . . . . . . . . . . . . . . . 169
C o n t e n t s      vii
8 Colonization, Structured Landscapes and Seasonal Mobility:

An Examination of Early Paleo-Eskimo Land-Use Patterns
in the Eastern Canadian Arctic. . . . . . . . . . . . . . . . . . . . . . . . . . . 174
S. Brooke Milne
Paleo-Eskimo Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
Environmental Knowledge and Landscape Learning. . . . . . . . . . 180
Mobility, Social Interaction and Lithic Procurement. . . . . . . . . . 183
The View from Southern Baffin Island. . . . . . . . . . . . . . . . . . . . . 185
9 The Emergence of Cultures of Mobility in the
Altai Mountains of Mongolia: Evidence from the
Intersection of Rock Art and Paleoenvironment. . . . . . . . . . . . . 200
Esther Jacobson-Tepfer
Location of the Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
The Rock Art at Aral Tolgoi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
The Rock Art in the Upper Tsagaan Gol Complex. . . . . . . . . . . 214
The Rock Art at Tsagaan Salaa/Baga Oigor . . . . . . . . . . . . . . . . . 218
10 Nomadic Sites of the South Yergueni Hills on the Eurasian Steppe:
Models of Seasonal Occupation and Production . . . . . . . . . . . . . 230
Natalia I. Shishlina, Eugeny I. Gak and Alexander V. Borisov
Geomorphologic and Geobotanic Descriptions. . . . . . . . . . . . . . 230
Archaeological Excavations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
Archaeozoologic Identifications. . . . . . . . . . . . . . . . . . . . . . . . . . . 239
Archaeobotanic Determinations . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Analysis of the Paleosoil. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
Chronology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
11 Trogodytes = Blemmyes = Beja?
The Misuse of Ancient Ethnography . . . . . . . . . . . . . . . . . . . . . . 250
Stanley M. Burstein
The Trogodytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
The Evidence Reconsidered. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
viii     The Archaeology of Mobility
12 Is the Absence of Evidence, Evidence of Absence? Problems in

the Archaeology of Early Herding Societies of Southern Africa. . . 264
Andrew B. Smith
Material Culture and Archaeological Visibility. . . . . . . . . . . . . . . 267
The Earliest Herders in Southern Africa. . . . . . . . . . . . . . . . . . . . 269
Theories of Colonization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
The Transition of Herding. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
13 The Social and Environmental Constraints on Mobility
in the Late Prehistoric Upper Great Lakes Region. . . . . . . . . . . 280
Margaret B. Holman and William A. Lovis
Theoretical Framework. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
Ethnographic Analogy and Late Prehistoric Mobility . . . . . . . . . 283
The Environment of the Study Area. . . . . . . . . . . . . . . . . . . . . . . 286
The Chippewa and the Mackinac Phase People. . . . . . . . . . . . . . 289
The Ottawa and the Juntunen Phase People. . . . . . . . . . . . . . . . . 293
14 Nomadic Potters: Relationships Between
Ceramic Technologies and Mobility. . . . . . . . . . . . . . . . . . . . . . . 307
Jelmer W. Eerkens
Mobility and Pottery Technology . . . . . . . . . . . . . . . . . . . . . . . . . 309
Resolving Conflicts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Part II: The Present and the Future

15 Mobility and Sedentism of the Iron Age
Agropastoralists of Southeast Kazakhstan . . . . . . . . . . . . . . . . . . 329
Claudia Chang
The Study Area. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
16 Crossing Boundaries: Nomadic Groups and Ethnic Identities�� 343
Stuart T. Smith
Ethnicity, Boundaries and the Nation State ��344
Ethnicity and Archaeology��346
Ethnicity in the Archaeological Record��348
Askut. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
Tombos. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
Nomads and Ethnicity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
Berenike . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
C o n t e n t s      ix
17 Variability and Dynamic Landscapes of Mobile

Pastoralism in Ethnography and Prehistory. . . . . . . . . . . . . . . . . 366
Michael D. Frachetti
Ethnography of Nomadism as a Study of Variation . . . . . . . . . . . 368
The Archaeology of Mobile Pastoralism. . . . . . . . . . . . . . . . . . . . 372
Dynamic Pastoral Landscapes. . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
The Pastoral Archaeology of Eastern Kazakhstan . . . . . . . . . . . . 377
Variation in the Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
Burial and Settlement Geography and Forms. . . . . . . . . . . . . . . . 381
Modeling the Landscape Dynamics. . . . . . . . . . . . . . . . . . . . . . . . 388
18 Mobility and Sedentarization in Late Bronze Age Syria . . . . . . . 397
Jeffrey J. Szuchman
Were Aramaeans Nomadic?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
Aramaeans and Assyrians in the Late Bronze Age. . . . . . . . . . . . . 401
Models of Sedentarization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
Towards an Archaeology of Sedentarization. . . . . . . . . . . . . . . . . 405
19 Suggestions for a Chaîne Opératoire of Nomadic
Pottery Sherds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
Hans Barnard
Eastern Desert Ware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
Experimental Pottery Production . . . . . . . . . . . . . . . . . . . . . . . . . 419
20 History of the Nomadic Architecture of the Hadendowa
in Northeast Sudan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441
Anwar A-Magid
North African Nomadic Tent-Dwellings. . . . . . . . . . . . . . . . . . . . 444
The Hadendowa Tent-Dwelling . . . . . . . . . . . . . . . . . . . . . . . . . . 446
Skin Tent-Dwellings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
Hair Tent-Dwellings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452
Mat Tent-Dwellings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454
Reed Huts, Caves and Rock Shelters. . . . . . . . . . . . . . . . . . . . . . . 456
Euphorbia Stem-Dwellings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
Litters and Palanquins. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
The Packsaddle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
21 The Bedouin Tent: An Ethno-Archaeological Portal
to Antiquity or a Modern Construct?. . . . . . . . . . . . . . . . . . . . . . 465
Benjamin A. Saidel
The Bedouin Black Tent. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467
x     The Archaeology of Mobility
Coffee, Tobacco, and Pottery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470

The Bedouin Tent in Archaeological Context. . . . . . . . . . . . . . . . 473
The Ethno-archaeological Value of the Bedouin Tent. . . . . . . . . 475
22 Naming the Waters: New Insights into the Nomadic
Use of Oases in the Libyan Desert of Egypt. . . . . . . . . . . . . . . . . 487
Alan Roe
Approaches to Old World Nomadic Pastoralism . . . . . . . . . . . . . 488
The Physical Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
The Human Environment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
Pastoral Migration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493
Pastoral Ecology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496
Relations with the Oaseans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498
Archaeological Signatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500
Nomadic Use of the Egyptian Oases. . . . . . . . . . . . . . . . . . . . . . . 502
23 From Objects to Agents: The Ababda Nomads
and the Interpretation of the Past. . . . . . . . . . . . . . . . . . . . . . . . . 509
Willeke Wendrich
The Ababda . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511
Social Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512
Land and Resource Ownership . . . . . . . . . . . . . . . . . . . . . . . . . . . 514
The Overnight Bag and the Portable Residence. . . . . . . . . . . . . . 517
Foodways and Cooking Utensils. . . . . . . . . . . . . . . . . . . . . . . . . . 527
Personal Care, Clothing and Adornment . . . . . . . . . . . . . . . . . . . 528
Gender Priorities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528
Immateriality of Ababda Culture. . . . . . . . . . . . . . . . . . . . . . . . . . 530
Mobility, Distance and Social Life. . . . . . . . . . . . . . . . . . . . . . . . . 534
Impact on the Landscape. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535
Ababda Material Traces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536
24 No Room to Move: Mobility, Settlement and Conflict
Among Mobile Peoples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
Roger L. Cribb
Conflict Among Contemporary Aboriginal Populations . . . . . . . 544
Horizontally and Vertically Integrated Societies. . . . . . . . . . . . . . 544
Tolerance Thresholds Relating to Conflict. . . . . . . . . . . . . . . . . . 545
Responses to Settlement Density. . . . . . . . . . . . . . . . . . . . . . . . . . 548
The Impact of Fixed Housing. . . . . . . . . . . . . . . . . . . . . . . . . . . . 550
Contemporary Communities on Cape York Peninsula. . . . . . . . . 550
The Camp at Chinaman Creek. . . . . . . . . . . . . . . . . . . . . . . . . . . 552
C o n t e n t s      xi
Public Space and Conflict . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553

25 NOMAD: An Agent-Based Model (ABM) of
Pastoralist-Agriculturalist Interaction . . . . . . . . . . . . . . . . . . . . . 557
Lawrence A. Kuznar and Robert Sedlmeyer
Pastoral Nomad-Sedentary Agriculturalist Dichotomy . . . . . . . . 558
Cycles of Conquest and Pastoral Nomads. . . . . . . . . . . . . . . . . . . 559
Ethnography and Sedentarization. . . . . . . . . . . . . . . . . . . . . . . . . 561
Simulating Pastoralist-Agriculturalist Interactions. . . . . . . . . . . . 563
The Human Dimension. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
The Rules of the Game . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 568
Expectations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571
Model Runs and Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571
List of Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584
List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592
Ch a p t e r 1
T he A rchaeology of
Mobility:
Definitions and Research
Approaches
Willeke Wendrich and Hans Barnard1
T hroughout history, groups and individuals have traveled the face of

the earth for manifold reasons and along a multitude of paths, both in space
and in time. Such mobile people, be they hunter-gatherers, pastoral nomads
or otherwise, left archaeological traces distinctly different from settled popula-
tions. These traces are usually simply characterized as ‘ephemeral campsites.’
It is frequently stated that mobile people obtain their material culture from
neighboring settled populations rather than producing their own and that they
do not leave recognizable archaeological traces (Finkelstein and Perevolotsky
1990). From the 24 chapters in this volume, however, and from many more
studies that have appeared elsewhere in the recent past (for instance Irons and
Dyson-Hudson 1972; Bar-Yosef and Khazanov 1991; Cribb 1991; Chang and
Koster 1994; Bar-Yosef and Rocek 1998; Khazanov and Wink 2001; Veth et al.
2005), it must be concluded that there is indeed an ‘archaeology of mobility.’ By
using specific and well-defined methods, which take into account the low density
of artifacts and concentrate on regional studies, it is eminently possible to come
to a better understanding of mobile people in archaeological contexts.
The archaeology of mobility encompasses more than tracing ephemeral
campsites. Much like any other group, mobile people produce, use and discard
a distinct material culture that includes functional objects, art and architecture.
The latter can be found at the campsites, or at locations in the landscape, where
graves or ritual structures have been erected (Betts, Chapter 2; Rosen, Chapter
5; Browman, Chapter 7; Shishlina et al. Chapter 10; Frachetti, Chapter 17;
Saidel, Chapter 21). In an archaeological context the traces we encounter are
occupation debris but also objects or assemblages that have been left behind,
1.
We would like to thank Jelmer Eerkens and Steven Rosen for their comments on
earlier versions of this chapter; and William Lovis for the final paragraphs, dedicating
this volume to the memory of Margaret Holman.
1
2     W i l l e k e W e n d r i c h and Hans Barnard
stored or cached on purpose (Alizadeh, Chapter 4; Milne, Chapter 8; Eerkens,

Chapter 14; Barnard, Chapter 19; Wendrich, Chapter 23). Art or written
records occur in the form of petroglyphs, rock drawings, landscape signs, group
symbols and written texts (Jacobson-Tepfer, Chapter 9). However, most of the
written accounts on mobile people have reached us through the eyes and pens
of outsiders (Burstein, Chapter 11).
Similar to its predecessors,2 the Fourth Cotsen Advanced Seminar aimed
to stimulate and facilitate the discussion and interaction between scholars of
Old World and New World archaeology. As archaeology in the Old World is
usually associated with history, languages and the sciences, while archaeologists
in the New World have more affinity with social sciences, especially anthro-
pology, this often requires the building of bridges between relatively insular
disciplines and, occasionally, the breaching of historical barriers between
different archaeological fields of research. Lifting specific subjects out of their
customary regional or temporal context, as well as discussing individual studies
in a multidisciplinary setting, enable us to approach our research with fresh
considerations and to discuss methodologies and results in a wider interpre-
tative framework. That such a multiregional and multidisciplinary approach
can be extremely stimulating and lead to surprising new insights is clear from
similar initiatives (such as Bailey and Parkinson 1984; Veth et al. 2005), as well
as from the chapters presented here.
This volume combines the proceedings of two meetings. The first was a
symposium (number 82) during the 69th Annual Meeting of the Society for
American Archaeology (Montreal, 2 April 2004); the second was a workshop
that took place in the Cotsen Institute of Archaeology at UCLA (Los Angeles,
22–24 June 2004). Presenters were asked to consider three themes: the defini-
tion of mobility, the material culture of mobile peoples, and, when applicable,
the relations between mobile and settled groups sharing the same geographical
area. The discussion centered on the best field methods to record the mate-
rial traces of mobile groups and the interpretative framework to understand
the ancient remains in their cultural context. In addition, contributions are
included from scholars who were involved in the discussions, partly through a
video conference that took place in June 2004, but could not attend either of
the meetings. The chapters in this volume consequently cover a large range
of subjects and periods (Figure 1.1; Table 1.1).
These seminars are Theory and Practice in Mediterranean Archaeology: Old World and
2.
New World Perspectives (Papadopoulos and Leventhal 2003); The Archaeology of Ritual
(8-9 January 2004, convened by Evangelos Kyriakidis); and Ritual Economy: Untethered
by Space, Time, or Economic Form (2-3 March 2006, co-organized by E. C. Wells and P.
A. McAnany).
The Archaeology of M o b i l i t y      3
Figure 1.1. World map showing the location of regions discussed in this volume.
This volume is divided into two parts: Part I, ‘The Past at Present’ provides
an overview of the state of the research on a certain area, people or period; Part
II, ‘The Present and the Future’ emphasizes ethnoarchaeology or experimental
archaeology and provides outlines or suggestions for further research. Considering
the objective of both the meetings and this volume, to stimulate discussions
between archaeologists working on different research topics in different areas of
the world, the explicit choice has been made to not present the material by region.
Instead the focus is on multiregional common themes, as well as obvious differ-
ences; thus, the placement of the chapters in the book may invite users to read
accounts dealing with different regions and periods. Within each part the chapters
are presented in chronological order of the subject matter. A comprehensive table
of contents and an extensive index will allow readers to find sections of special
interest quickly; an overview of the subject matter of each chapter is presented in
Table 1.1; their geographical distribution is illustrated in Figure 1.1.
Definitions: Moment, Movement and Motivation

The etymology of the word ‘nomad’ goes back to the ancient Greek nomades,
pastoral tribes, related to the verb nemo, which means to distribute, to pasture
or to graze flocks (Liddell et al. 1958). The term is, therefore, originally linked
to pastoralism but has achieved a much broader meaning in which the most
Table 1.1. Overview of the Subject Matter of the Chapters in this Volume:
Chapter/Author(s) Geographical Subject matter Time period

focus
Part I: The Past at Present
2 Betts Jordan Overview of several sites and 7th millennium BCE
archaeozoological data
3 Bernbeck Turkey Theoretical considerations based 6th millennium BCE
on the case of Fıstıklı Höyük
4 Alizadeh Iran Combining ethnographical and 5th millennium BCE
archaeological data for Southwest
Iran
5 Rosen Israel Theoretical considerations based 6th millennium BCE
on the case of the Negev to 1500 CE
6 Buccellati Syria State of the research on the 3rd–2nd millennium
Amorites BCE
7 Browman Peru, Bolivia State of the research on camelid 2nd millennium BCE
pastoral nomadism in the Andes to 500 CE
8 Milne Canada Models based on lithic finds from 2nd–1st millennium
several sites on Baffin Island BCE
9 Jacobson-Tepfer Mongolia Survey and interpretation of 1st millennium BCE
petroglyphs
10 Shishlina, Gak Southern Russia Report on several sites and 1st millennium BCE
and Borisov archaeobotanical data
11 Burstein Egypt, Sudan Analysis of historical sources 500 BCE to 500 CE
12 A. B. Smith South Africa Combining historical and 0–1500 CE
archaeological evidence for the
Khoekhoen
13 Holman and United States Theoretical considerations based 500–1500 CE
Lovis on the case of the Great Lakes
Region
14 Eerkens United States Models based on ceramic and 800–1800 CE
lithic finds in the Great Basin
Part II: The Present and the Future
15 Chang Kazakhstan Models based on ethnographical 8th century BCE to
and archaeological data 1st century CE
16 S. T. Smith Sudan, Egypt Theoretical considerations based 1800 BCE to 600 CE
on Askut, Tombos and Berenike
17 Frachetti Kazakhstan Models based on several sites in 2nd–1st millennium
the Koksu River Valley BCE
18 Szuchman Syria State of the research on the 2nd–1st millennium
Aramaeans BCE
19 Barnard Egypt, Sudan Ceramic analysis and 4th–6th century CE
experimental archaeology
20 Magid Sudan Ethnohistorical and 3rd century BCE to
ethnographical description of the present
Hadendowa
21 Saidel Southern Levant, Ethnohistorical and 17th–19th century
northern Arabia ethnographical description of the CE
and the Sinai Bedouin black tent
22 Roe Northwest Egypt Resource use by pastoral nomads 20th century CE
23 Wendrich Southeast Egypt Ethnohistorical and 20th century CE
ethnographical description of the
Ababda
24 Cribb Australia Conflict management among the 20th century CE
Aboriginals
25 Kuznar and World Computer models for nomadism 21st century CE
Sedlmeyer
important aspect is that of a (residentially) mobile existence. This meaning

urges us to further delineate what variations of mobility occur. Here we define
‘mobility’ as the capacity and need for movement from place to place. This
definition takes into account that even though a person’s or group’s activities
may concentrate or depend on movement, there will always be stationary
periods, if only to sleep. This is an intentionally broad definition, much more
so than, for instance, the one proposed by Holman and Lovis (Chapter 13). It
focuses on the movement of humans in the landscape but leaves their organiza-
tion and motivations to be determined. A further specification of the type of
movement is based on the mobile versus settled segment of the group and on
the motivation, in the most abstract sense.
Figure 1.2 illustrates four basic types of mobility: the entire group travels
from resource to resource (Figure 1.2a); segments of different groups travel to
and from specific resource areas (Figure 1.2b); segments of the group gather
resources for a base camp (Figure 1.2c); the entire group travels, following a
distinct and fixed pattern (Figure 1.2d). These schematic representations focus
on the movement of groups in the landscape but leave their motivations, the
composition of the population segment, the distance and the time frame to be
specified. The length of stationary periods compared to the time of travel is
Figure 1.2. Schematic representation of mobility patterns. Black dots indicate the main group; gray
dots indicate segments of the group.
equally undefined. Many activities, varying from the slaughtering of animals

to the production of textiles or pottery, require such stationary periods. Other
activities, such as procuring chert, finding water or collecting mushrooms, are
linked to specific locations and seasons. The length of these stationary periods
may vary from one night (rest, finding water) or a number of hours (slaugh-
tering, flint knapping) to several years, when the lakes are full of fish or the
pastures are lush and green enough to sustain a population for a long period.
We can summarize the definition of mobility as a combination of the moment
in time, the type of movement and the motivation for mobility.
In his groundbreaking work Nomads in Archaeology, Roger Cribb (1991) outlined
the differences between hunter-gatherer and pastoral societies (Table 1.2).3 Cribb’s
comparison focused on the motivation for mobility, which is either centered on
the humans in the group or on the flocks or herds. He also outlined differences in
the movement pattern and the resulting archaeological scatter of artifacts. While
this comparison, with its clear juxtaposition of features, is probably true for some
hunter-gatherer and pastoral-nomadic groups, both archaeological and ethno-
graphic studies have shown many variations that clearly deviate from this pattern
(Bernbeck, Chapter 3; Rosen, Chapter 5). By insisting on the forced separation
of hunter-gatherers and pastoral nomads, we risk overlooking similarities and
overlap. Even the terminology of the two fields of research deviate to the extent
that communication about mobility patterns is rendered almost impossible.
Table 1.2. Differences Between Hunter-Gatherers and Pastoral Nomads (after Cribb 1991):
Hunter-gatherers Pastoral nomads

Procurement and consumption Production (determined by consumption pattern
of livestock)
Moves toward resources or gathers resources for Moves toward resources for flocks (independent
humans of human resources)
Mobility to bring variation in resources Mobility to maintain access to one resource
(grazing)
Consumption immediately after procurement Human consumption independent of herd
Migration follows complex pattern Migration track follows simple pattern
Risk minimization for humans Optimization of pastoral production, minimize
risk for herds
Archaeological remains scattered widely Constrained scatter, less functional variability
Stable and localized migratory patterns Instability and dramatic shifts in migration tracks
3
In August 2007, when this volume was in its final stages of production, Dr. Roger Cribb
passed away after a protracted battle with cancer. He was only 59 years old and holding
an honorary position as Research Fellow at James Cook University (Cairns, Australia) at
the time. Roger not only participated in our workshop in Los Angeles (June 2004) and
contributed a chapter to this volume, but his seminal work on the archaeology of nomads
is quoted in most chapters. He will be sorely missed by his family, friends and colleagues.
Ter minology Related to Hunter-Gatherers

Binford (1980) introduced the term logistical mobility to express the movement
of part of a hunter-gatherer group to secure provisions (Figure 1.2c). Logistical
mobility can consist of a one-day trek or require multiple nights away from
the base camp or settlement, which in itself is not necessarily permanent.
Usually only part of the group, defined by gender, age, social status or health,
embarks on such an expedition (Milne, Chapter 8). In contrast to logistical
mobility, the term residential mobility has been introduced to indicate a pattern
where the entire group leaves and moves the base camp (Figure 1.2a; Kelly
1992). Tethered mobility expresses specific limitations to the range or time of
movement (Figure 1.2d; Ingold 1980), by dependence on particular resources
or by social circumscription. Examples of tethered mobility are found in very
different social and environmental circumstances. Caching of pottery by
hunter-gatherers can be taken as an indication of tethered mobility (Eerkens,
Chapter 14; Kelly 1995).
Ter minology Related to Pastor al Nomads

Pastoral nomadism is the general term for mobility centered on the maintenance
and welfare of flocks or herds (Figure 1.2a), while semi-nomadic pastoralism
denotes a situation where part of the group is settled (Figure 1.2c), or the entire
group is settled for part of the year (Figure 1.2b). Agropastoralism specifies a
combination of agricultural and pastoral activities (Figure 1.2c or 1.2b), and the
term focuses on the motivation, not on the mobility pattern (Buccelati, Chapter
6). Tethered nomadism indicates a dependence on particular resources, other
social groups, or features in the landscape (Figure 1.2d). Enclosed nomadism
describes the close relation between a nomadic group and the surrounding
settled population (Rowton 1974), while peripheral nomadism is used for groups
that live at the fringes of a settled society. These two latter terms tacitly refer
to the relation between the settled and nomadic population rather than to a
mobility pattern. They also imply that the studies of mobility are written from
the viewpoint of the settled population.
Several terms are used for seasonal migration with flocks or herds, which
is determined primarily by weather conditions or scarcity of resources in
certain periods of the year (Figure 1.2b). The most general term for this is
transhumance, but specifications occur, such as vertical transhumance (or vertical
mobility) for seasonal movements in mountainous areas. In this volume authors
describe such a process in several different regions: Iran (Alizadeh, Chapter
4), Kazakhstan (Chang, Chapter 15; Frachetti, Chapter 17) and the Andes
(Browman, Chapter 7). High-altitude pastures are used in summer, whereas the
snow forces the herds to low-lying meadows in winter. Horizontal transhumance

(or horizontal mobility) describes movement at approximately the same elevation
over a large area, mostly to find resources in spite of heavy snow cover (Milne,
Chapter 8) or drought (Johnson 1969). In the same country, herds are known
to follow the winter rainfall (Wendrich, Chapter 23) or relocate in summer to
a region where water can be found in wells (Roe, Chapter 22).
The purpose of developing a set of definitions, and a related terminology, is
to describe our subject matter in a concise way and to enable communication
among scholars. Such definitions and terminology are, in a way, abbreviations
and, at the same time, generalizations of detailed descriptions of particular
societies at a specific moment in time. There is, however, always the danger
that our definitions become blueprints for explaining the variation in human
movement through time and space. The same definitions that enable us to
speak about the many forms of human mobility ironically also restrict us
and render mobility into a set of timeless, static human interactions with the
natural surroundings. Ironically, the pinnacle of long-term settled existence, the
Near Eastern tell, forces us to describe human-settled habitation as a develop-
ment: the meters of stratigraphical information demonstrate inescapably that
human existence is in constant flux. In contrast, the archaeology of mobility
has too often focused on static categorizations of human reactions to changing
circumstances, a behavioristic approach interpreting the data by resource- or
climate-driven human adaptations.
Most authors in this volume argue against such a static concept of pastoral
nomadism or hunting and gathering. Since the 1970s focus has shifted from
narrowly described modes of resource procurement to the opportunistic use
of several supply strategies, sometimes called multi-resource nomadism (Salzman
1972). The term herder-gatherers, introduced by Rosen (1998, 2002), finds
increasingly widespread use and highlights the frequent occurrence of pastoral-
ists who also embark on hunting and gathering (Betts, Chapter 2). Instead of
becoming fixated on definitions the authors are concerned with the method-
ology of studying archaeological remains in a regional setting (Rosen 1992).
By defining mobility in terms of ‘moment’ (length of time, season), ‘motion’
(mobility pattern charted over time), ‘motivation’ (resources, but also cultural
identity, social or economic circumscription), and ‘segment’ (the parts of
the population defined by gender, age, health or social position), we liberate
ourselves from fixed preconceptions. Rather than considering populations as
either settled or mobile, recent interpretations, of which several can be found in
this volume, emphasize the fluidity of mobility and the role of social organiza-
tion and agency in the process of mobilization or settlement. The scope ranges
from an entire population that is highly mobile to a mostly settled population in
which only small groups move around. Mobility should, therefore, be defined
Table 1.3. Factors and Range of Mobility:
Factor Range
Moment Period of movement Specific day, week, month, season, year
Length of period of movement Total no. of nights
Length of stay in each location Total no. of nights

Repetition of mobility pattern Daily, weekly, monthly, seasonally, yearly, etc.
Stability of mobility Repetition over multiple years, decades or centuries
Variation of mobility over time Inventory of multiple years, decades, centuries
Movement Range of movement Number of km per period of movement
Pattern of movement Fixed, flexible, random (specified with map)
Motivation Purpose of movement Hunting, gathering, herding (specified with

activities)
External factors Climate, war, social pressure, illness, etc.
Decision base for movement Who decides (group or high-status individual), basis
on which decision is made
Assessment of mobility within Identity, cause for pride, neutral, necessity,
the group compassion, scorn
Assessment of mobility by Identity, cause for pride, neutral, necessity,
outsiders compassion, scorn
Mobility integral part of No / Yes (specify how)
identity
Segment Distribution of mobile One group, multiple camps
population
Percentage of population in 0–100%
motion
Part of population in motion Gender, age, social position, profession, cultural
identity
Variation of mobility within Gender, age, social position, profession, cultural
the group identity
Location of burials / cemetery Central location (monumental or not), buried ‘en
route,’ cremation, other means
for each population separately according to very specific questions (Khazanov

1984) and over a long period of time. Table 1.3 gives an outline of research
questions related to a study of mobility and summarizes the factors involved.
It illustrates the complexity of any definition, research program and analysis of
human mobility. Several of these factors have been taken into account in the
important work of Kuznar and Sedlmeyer (Chapter 25).
From Table 1.3 it is apparent that the segments of a society, the mobility
pattern in time and space and, most important, the motivations show an enor-
mous variability in human mobility. Economic or subsistence incentives are
by no means the only factors to be taken into account when seeking to explain
mobility patterns. Most often the motivations that determine the pattern are a
combination of factors, partly dependent on the context and audience, and in

constant development and flux. These motivations are manifold, ranging from
climatic and subsistence to sociocultural, from economic to religious; and from
ideological to a search for personal safety (Cribb, Chapter 24).
Relations Between Mobile and Settled Groups

There has been much debate on the place of sedentism in the evolution of
Homo sapiens (Moore 1983; Bender 1985; Marquardt 1985; Cohen 1985; Keeley
1988; Henry 1991; Liebermann 1993; Rosenberg 1998). The vast majority
of laymen, politicians and scholars, however, consciously or subconsciously
understand settled living as the highest rung on the evolutionary ladder where
permanent housing goes hand in hand with agriculture, landownership, social
stratification and industrialization. The word pharaoh, ruler of one of the earlier
complex societies, literally means ‘great house,’ while in Levantine societies the
concept of ‘the house’ was an all-encompassing social phenomenon (Schloen
2001), so much so that the term is also used for the genealogy by pastoral
nomads in the area (Wendrich, Chapter 23). Similar terminology is used by
government institutions (the House of Representatives) and rich or noble
families (as in Edgar Allan Poe’s ‘The Fall of the House of Usher’). Without
doubt this deterministic point of view is among the reasons that mobile people
are under constant pressure to settle down or leave; in our modern and increas-
ingly complex society even more so than in the past.
The relations between mobile and settled groups are relatively well studied
(Nelson 1973; Aurenche 1984; Khazanov 1984; Khazanov and Wink 2001).
Often this relation is problematic and we should take into account that most
of our historical (written) sources on mobile peoples have been recorded by a
settled population and usually show misconceptions, either vilifying or roman-
ticizing a mobile lifestyle. In the 1st century CE, Strabo already felt the need to
correct this image: “These are nomads and neither many nor warlike, although
they were believed to be so by the ancients because of their frequent raids on
defenceless people” (Strabo, Geography 17.1.53; Eide et al. 1998).
Nomads are often depicted as raiders or freedom loving adventurers, while
the intricacies of social organization, ethnic identity, loyalties and policies are
rarely understood (Burstein, Chapter 11; S. T. Smith, Chapter 16). Part of our
tendency to oppose sedentary and mobile ways of life stems from the represen-
tation of nomadic groups as ‘the other’ or ‘the outsider’ by the settled powers.
Unsettling accounts of people surviving and even thriving in peripheral areas
are, for instance, one way in which the dichotomy between ‘the desert and the
sown’ is maintained (Chang, Chapter 15; Szuchman, Chapter 18; Nelson 1973;
Chang and Koster 1994).
As several of the contributions to this volume point out, there is the

danger of creating a dichotomy between settled and nomadic life. Humans
will adapt their lifestyles to changing circumstances, or proactively decide on
changing their way of life, either as a group, as a specific part of the group, or
as individuals. No firm delineation can be made between settled and mobile
existence. At the same time, we should not imagine the relation between
settled and mobile life as a point or range on a scale between ‘completely
settled’ and ‘completely mobile.’ The decision to move location occasionally,
regularly or frequently is in most cases opportunistic. The idea that mobility
is an adaptation or a response to (often adverse) changing circumstances is an
equally limited view of the different forms of movement and diverse motiva-
tions for mobility.
Considering human society through its movements, rather than through
its settlements, brings to the fore that even in firmly settled societies there is
always a part of the population that is mobile. To mention just a few examples:
in many societies trade is in the hands of either separate mobile groups or
mobile members of the settled group. Trade inherently requires mobility,
varying from farmers traveling to the market town, peddlers and hawkers
moving from village to village, to long-distance traders. Religiously motivated
traffic is another strong reason for mobility of settled populations but is also
a factor in permanently mobile groups. Pilgrimages are known among many
of the world’s populations and are just a step away from religious orders of
mendicants. A third example is the custom of the upper class to move residence
to either cooler or warmer regions during part of the year. In 19th
century
India the British higher ranks moved residence to summer homes in the cooler
mountains. Mobility in our modern Western society includes commuters, vaca-
tioners, homeless people and refugees (Bernbeck, Chapter 3). And yet, these
types of mobility are considered part of the settled mode of living.
Material Culture, Landscape and Fieldwork

Considering the range of potential forms, reasons and objectives for mobility,
we should ask how much of these we can actually retrieve from the archaeo-
logical record. A number of publications have highlighted the potential of
archaeological research of mobility (Monks 1981; Bar-Yosef and Khazanov
1991; Saidel 1998; Rosen 1998, 2002). As Frachetti (Chapter 17) describes
vividly, landscape archaeology is essential for understanding mobility (Chang
and Koster 1986). Within the landscape the features that are of potential signif-
icance for subsistence, resources, ritual and routing are taken into account.
The traditional focus on archaeological ‘sites’ does not answer the ques-
tions involved in the study of mobility. Nevertheless, it is inherent to human
behavior that most material remains will be left, and can be retrieved, at places
where humans stay for at least a short while. In addition, there are traces along
the routes, such as specific signs, that denote ownership of resources or other
messages to those who pass by. Important sources for studying the material
culture of mobile peoples are cached materials: objects that were left behind
on purpose to be used during a future stay in a particular location (Figure 1.2d;
Eerkens, Chapter 14).
The study of pastoral nomads benefits greatly from the methods developed
by prehistorians. Beyond the questions clarifying the nature, scope and scale
of mobility, the research focus diverges when concentrating on the specific
motivations for mobility. The natural circumstances, subsistence, cultural
development as well as the economic and social context of the groups under
study differ, as do our sources of information. Archaeologists working on
understanding mobility patterns put their efforts into establishing not only
the chronology and cultural markers of a region but also the period of stay,
the number of recurrent stays, the seasonality of human presence and the
activities at each location that has a higher density of material remains than the
surrounding landscape. To determine whether a low-density location represents
a one-time period of stay, a short-period production site, or a yearly visited
campsite is extremely difficult and often impossible. The data have to be placed
within the broader context of multisite occupation and landscape analysis
(Bernbeck, Chapter 3; Kuznar and Sedlmeyer, Chapter 25). Conclusions on
the site use and motivation, literally that which causes these specific people
to move or to pause at a particular place for a particular period in a particular
time of year, requires a reconstruction of the entire activity pattern. The study
of mobility is learning about human interaction with the landscape, its limiting
factors, its resources and its meaning.
Different information is gathered from site and regional levels. The most
important techniques when studying low-density assemblages are micro-anal-
ysis of the taphonomy, the faunal and the botanical remains. Archaeozoological
studies potentially provide information on age and size of slaughtered or
hunted animals, level of domestication and seasonality (Monks 1981; Brewer
1989; Van Neer 1993; Bar-Yosef and Rocek 1998; Van Neer et al. 2004).
Subsistence and climatic information, including seasonal use of sites, can be
gained from archaeobotanical research (Shishlina et al., Chapter 10; Holman
and Lovis, Chapter 13), while material culture (pottery, metal, textiles, leather,
basketry) provides information on activity, date, cultural affiliation, gender,
age and organization. Burial sites provide important information on the self-
definition and identity of some mobile groups, while for others the inhumation
seems to be as fleeting and uprooted as their general existence. The study of
sites concentrates on the stationary activities and on the shelters with which
mobile groups equip themselves (Magid, Chapter 20). Storage facilities may
be indications of recurring visits to the same area (Eerkens, Chapter 14;
Akkermans and Duistermaat 1997; Wendrich and Cappers 2005).
On a regional level an inventory of available resources, indications of routing
and other remains of the same chronological period enables the reconstruction
of a mobility pattern. The study of petroglyphs, geoglyphs (such as cleared
areas, stacked stones, border or route indicators) and other visible signifiers
is of great importance (Jacobson-Tepfer, Chapter 9). A caveat should be that
a mobile group can ‘occupy’ an area of hundreds of square km, while a yearly
track can involve a roundtrip in the 1000 km range (Alizadeh, Chapter 4).
Landscape reconstruction, identification of resources and a reconstruction of
the climate, weather and ecology are multidisciplinary tasks (Shishlina et al.,
Chapter 10). Modern archaeological research necessarily involves a team of
specialists from different disciplines and in a way mirrors the explorations of
the mobile groups in the past. A group’s knowledge of resources, such as wells
or raw materials, is passed on to the next generation, but when a group has to
explore virgin territory, its members embark on a process of ‘landscape learning’
(Milne, Chapter 8; Rockman 2003). What archaeologists are doing is learning
the landscape, with the added difficulty that some facets of the landscape have
changed over time. Moreover, particular aspects, such as cultic or symbolic
meaning of landscape features, are difficult to deduce or corroborate.
The lack of stratigraphy in low-density assemblages does not represent a
lack in development of the mobile society and should not cloud our attempts to
understand the complexity of the structure of a society that is highly mobile or
spread out over a large area. Communication or planned (and perhaps ritual-
ized) interaction is very important for societies in which meetings cannot take
place haphazardly, because the space in which the interaction takes place is not
circumscribed or limited (Milne, Chapter 8).
Ethno-archaeology
Many contributors to this volume compare data from archaeological, ethno-
graphic and ethno-archaeological studies. Bernbeck’s statement that the present
state of affairs “cries out for the abandonment of analogical reasoning when
dealing with past mobile groups” (Chapter 3) reflects the ongoing discus-
sion of the pitfalls, values, uses and abuses of ethno-archaeology and analogy
(Wylie 1985; David and Kramer 2001). The direct historical approach, where
present-day populations are considered a continuation or ‘survival’ of ancient
inhabitants of the same area or region, is indeed a dangerous bedfellow for
archaeology because it limits the explanatory power of research and denies
ancient populations the ability to change. Instead, it should be noted that
change is unavoidable and perhaps the only firm given in any archaeological
study. Burstein’s (Chapter 11) and Saidel’s (Chapter 21) contributions to this
volume illustrate this clearly from very different perspectives. Often the direct
historical approach is connected with an evolutionary slant: a linear develop-
ment is implicated from hunter-gatherer (a completed stage) to herding (an
equally completed stage) to sedentism (a stage still in development). The evolu-
tionary approach is often considered one of gradual improvement with modern
(European) humans as the crown of the evolution (Trigger 2006:166–220).
The Great Kalahari Debate (Smith, Chapter 12), for instance, is partly based
on presuppositions about whether the present inhabitants of the Kalahari are
proud hunter-gatherers or pathetic remnants of what once was a number of
well-developed cultures. Equally misleading are ethno-archaeological studies
that concentrate on very limited, mostly ecological, aspects and use these as
a substrate to draw conclusions on a much wider scale. Such a deterministic
approach may provide some valuable insights, but it denies agency to indi-
viduals and groups and rejects the historical development of group agency,
which is often summarized as ‘culture.’
The simplified and limiting uses of analogical reasoning are rightly criti-
cized. But analogical reasoning, when done correctly, is not only extremely
useful but a method we simply cannot do without (Wylie 1985). Most of us
do not have the imagination to formulate the large number of interpretations
that could be given to specific archaeological assemblages. Ethnographies
and ethno-archaeological studies provide an inventory of known occurrences:
a palette of different types of organization, forms of habitation and human
reactions to and interactions with a wide range of physical, social and spiritual
circumstances. A case in point is the examples of groups with a surprisingly
long-distance range of mobility, a yearly round trip that covers more than 1000
km (Alizadeh, Chapter 4). Such an inventory is a source not only of ideas but
also of caveats, such as Alan Roe’s (Chapter 22) example of two groups that
stay in the same area during the same chronological period (late 20th, early 21st
century CE) but in different parts of the year. Roe describes the smaller oases
around Siwa, where pastoral nomads dwell in summer, while settled agricul-
turalists go there for the date harvest in fall. Ethno-archaeological research
is also used to test specific hypotheses, for instance in relation to depositional
and postdepositional processes. Here again the danger looms that we end up
with too limited and too deterministic an approach of ancient society. Barnard’s
(Chapter 19) trials to produce pottery without an existing infrastructure (such
as kilns) and with limited availability of water or fuel exemplify the use of
experimental archaeology to demonstrate that it is feasible to produce high-
quality ceramics in a transient situation. Such experiments do not corroborate
a specific production method or locality but refute unwarranted claims that
mobile people are incapable of producing ceramics because of a lack of infra-

structure (see also Eerkens, Chapter 14).
Related to landscape learning, referred to above, is the notion of the space
that one inhabits. This brings us to the cognitive aspects of the archaeology of
mobility: there is a large difference between inhabiting a range of campsites,
irrespective of the length of stay, and inhabiting the landscape as a whole. Were
groups moving from place to place, or did they consider themselves inhabitants
of one large space, the landscape through which they moved? Ultimately the
question is whether the people we study through the archaeological remains
were looking inward, concentrating on their community, or outward, toward
the world.
Discussion
The reality of a mobile existence is far more complex than the ordering prin-
ciples used to describe them (Figure 1.2). An outline of ‘types of mobility’
is necessarily a simplification that highlights a number of important aspects
but leaves out others. As will be clear from most of the contributions to this
volume, such generalizing categories are helpful to the novice, but they should
probably be abandoned when interpreting the full range of archaeological or
ethnographic data (Table 1.3). This volume gives pointers on how research can
do justice to the complexity and developments of past mobile and partly mobile
societies. Archaeologists studying the remains of mobile groups require specific
methods, which combine a meticulous analysis of ‘ephemeral campsites’ with
a close consideration of the landscape, the availability of resources, as well as
the scale, the layout and, often, the borders of the world through which the
groups moved. Hunter-gatherers may move more than 500 km to follow the
seasonal trek of elk, while settled farmers may temporarily move to a small
reed hut on their fields at a distance of 30 km for the summer nights during
sowing or harvest time. The traces that they leave reflect the fact that they did
not stay in one place permanently but visited the location either one time or
repeatedly for shorter periods. On one level there is remarkably little differ-
ence in the methods employed in the study of the remains of hunter-gatherers,
herder-gatherers, pastoral nomads, or even a settled population that ventured
temporarily outside its usual territory. All of these studies can benefit greatly
from the methods developed by prehistorians to deal with low-density artifact
scatters and landscape archaeology.
Cultures in which mobility plays an integral role are in constant devel-
opment. This is not only a result of changing circumstances but also a
context-dependent effort to improve the status quo. At the same time, humans
are firmly embedded in social relations, as well as ecological, economic and
political circumstances. Central to understanding the archaeology of mobility

is to refrain from defining mobility in too fixed a set of categories, thereby
denying the dynamic and opportunistic development of mobile peoples.
Acknowledgements and Dedication

Thanks are due to many people, among whom are two anonymous reviewers,
Amber Myers and Kandace Pansire (conference assistants), Sam Aroni (Director
of the Special Academic Cooperative Projects, UCLA), William Schniedewind
(Chair of the Department of Near Eastern Languages and Cultures, UCLA),
Jimmy Suo and Dean Abernathy (Visualization Portal, UCLA), Stein Hitland
(University of Bergen, Norway), Sue Rogers (University of Cambridge, UK),
Gary Mattison (The Getty Institute, Los Angeles), Magda Yamamoto, Sheryol
Threewit, Helle Girey, Ernestine Elster and Charles Stanish (Cotsen Institute
of Archaeology, UCLA). Special thanks go to Evangelos Kyriakidis, for sharing
his experience; to Louis van Dompselaar, for making the right software avail-
able; to Krzysztof Pluskota, for his photographs (Chapter 20); to all participants
in Montreal and in Los Angeles, for making the Fourth Cotsen Advanced
Seminar a success; and to Lloyd Cotsen, for making it all possible.
The picture on the cover was taken in 1998 by pastoral nomad Mohamed
Eid, with one of the cameras provided to the Ababda tribe by the Eastern
Desert Antiquities Protection Project (Wendrich, Chapter 23). It shows his
family living in the Eastern Desert, between the Nile and the Red Sea in the
border area between Egypt and Sudan (Burstein, Chapter 11; S. T. Smith,
Chapter 16; Barnard, Chapter 19; Magid, Chapter 20).
This volume is dedicated to the memory of Dr. Margaret B. ‘Peggy’
Holman, Research Associate at the Michigan State University Museum
and Adjunct Faculty in the Department of Anthropology at Michigan State
University (Figure 1.3). Throughout her career and despite her research
focus on North America, Peggy Holman was a dedicated student of mobile
people worldwide, coupling closely their ethnography, ethnohistory and
archaeology. Her work fostered cross-cultural approaches rooted in clear
use of ethnographic analogues to understand the archaeological record.
Peggy contributed substantially to our knowledge of such diverse subject
matter as the origin and role of maple sugaring in the upper Great Lakes
area, the seasonal and technological implications of caching and storage by
mobile hunter-gatherer horticulturalists in the region as well as the seasonal
scheduling of mutually desirable resource areas by adjacent mixed economic
groups. Moreover, she was always available to provide interested students
with her insights and helped to train multiple generations of ‘Great Lakes
archaeologists.’
Figure 1.3. Margaret B. Holman (center) in the field at the Zemaitis site (Michigan) in 1993, sur
rounded by her colleagues (from left to right) Janet Brashler, Elizabeth Garland, William Lovis and
Terrance Martin.
In April 2006, Peggy succumbed to pancreatic cancer, before she could see
the fruit of her work in this volume, a compendium that she greatly anticipated.
Peggy’s many contributions to the literature of mobile people, including her
lead authorship of Chapter 13, testify to her abiding interests in science and
people and will certainly endure. We can only regret that the future literature,
and our small community, will be the poorer for her absence.
References
Akkermans, P. M. and K. Duistermaat
1997 Of Storage and Nomads: The Sealings of the Late Neolithic Sabi Abyad.
Paléorient 22: pp. 17–32.
Aurenche, O.
1984 Nomades et sédentaires: Perspectives anthropologiques. Paris, Éditions Recherche
sur les Civilisations.
Bailey, G. and J. Parkinson (eds.)
1984 The Archaeology of Prehistoric Coastlines. Cambridge, Cambridge University
Press.
Bar-Yosef, O. and A. Khazanov (eds.)

1991 Pastoralism in the Levant: Archaeological Materials in Anthropological Perspectives.
Madison, Prehistory Press.
Bar-Yosef, O. and T. Rocek (eds.)
1998 Seasonality and Sedentism: Archaeological Perspectives from Old and New World
Sites. Cambridge, Harvard University Press.
Bender, B.
1985 Prehistoric Developments in the American Mid-continent and in Britanny,
Northwest France. In T. D. Price and J. A. Brown (eds.), Prehistoric Hunter-
Gatherers: The Emergence of Cultural Complexity. Orlando, Academic Press:
pp. 21–57.
Binford, L. R.
1980 Willow Smoke and Dog’s Tales: Hunter-Gatherer Settlement Systems and
Archaeological Site Formation. American Antiquity 45: pp. 4–20.
Brewer, D. J.
1989 Fishermen, Hunters and Herders: Zooarchaeology in the Fayum, Egypt (ca.
8200–5000 BP). British Archaeological Reports, International Series 478. Oxford,
Archaeopress.
Chang, C. and H. A. Koster
1986 Beyond Bones: Toward an Archaeology of Pastoralism. In M. B. Schiffer (ed.),
Advances in Archaeological Method and Theory, Volume 9. New York, Academic
Press: pp. 97–148.
Chang, C. and H. A. Koster (eds.)
1994 Pastoralists at the Periphery: Herders in a Capitalist World. Tucson, University
of Arizona Press.
Cohen, M. N.
1985 Prehistoric Hunter-Gatherers: The Meaning of Social Complexity. In T. D.
Price and J. A. Brown (eds.), Prehistoric Hunter-Gatherers: The Emergence of
Cultural Complexity. Orlando, Academic Press: pp. 99–119.
Cribb, R.
1991 Nomads in Archaeology. Cambridge, Cambridge University Press.
David, N. and C. Kramer
2001 Ethnoarchaeology in Action. Cambridge, Cambridge University Press.
Eide, T., T. Hägg, R.H. Pierce and László Török
1998 Fontes Historiae Nubiorum. Volume III: From the First to the Sixth Century AD.
Bergen, University of Bergen, Department of Greek, Latin and Egyptology:
828-835.
Finkelstein, I. and A. Perevolotsky
1990 Processes of Sedentarization and Nomadization in the History of Sinai and
the Negev. Bulletin of the American Schools of Oriental Research 279: 67–88.
1991 Foraging, Sedentism, and the Adaptive Vigor in the Natufian: Rethinking
the Linkages. In G. A. Clark (ed.), Perspectives on the Past: Theoretical Biases
in Mediterranean Hunter-Gatherer Research. Philadelphia, University of
Pennsylvania Press: pp. 353–370.
Ingold, T.
1980 Hunters, Pastoralists, and Ranchers. Cambridge, Cambridge University Press.
Irons, W. and N. Dyson-Hudson (eds.)
1972 Perspectives on Nomadism. Leiden, E. J. Brill.
Johnson, D. L.
1969 The Nature of Nomadism: A Comparative Study of Pastoral Migrations in
Southwest Asia and Northern Africa. Department of Geography Research Paper
118. Chicago, University of Chicago.
Keeley, L. H.
1988 Hunter-Gatherer Economic Complexity and “Population Pressure”: A Cross-
Cultural Analysis. Journal of Anthropological Archaeology 7: pp. 373–411.
Kelly, R. L.
1992 Mobility/Sedentism: Concepts, Archaeological Measures, and Effects. Annual
Review of Anthropology 21: pp. 43–66.
1995 The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Washington,
Smithsonian Institution Press.
Khazanov, A. M.
1984 Nomads and the Outside World. Cambridge, Cambridge University Press.
Khazanov, A. M. and A. Wink (eds.)
2001 Nomads in the Sedentary World. Richmond, Curzon.
Liddell, H. G., R. Scott, H. S. Jones and R. McKenzie
1958 A Greek-English Lexicon. Oxford, Clarendon.
Liebermann, D. E.
1993 The Rise and Fall of Seasonal Mobility Among Hunter-Gatherers: The Case
of the Southern Levant. Current Anthropology 34: pp. 599–631.
Marquardt, W. H.
1985 Complexity and Scale in the Study of Fisher-Hunter-Gatherers: An Example
from the Eastern United States. In T. D. Price and J. A. Brown (eds.),
Prehistoric Hunter-Gatherers: The Emergence of Cultural Complexity. Orlando,
Academic Press: pp. 59–98.
Monks, G.
1981 Seasonality Studies. In M. B. Schiffer (ed.), Advances in Archaeological Method
and Theory, Volume 4. New York, Academic Press: pp. 177–240.
Moore, A. M. T.
1983 The First Farmers in the Levant. In T. C. Young, P. E. L. Smith and P.
Mortensen (eds.), Studies in Ancient Oriental Civilization, Volume 36. Chicago,
University of Chicago Press: pp. 91–111.
Nelson, C. (ed.)
1973 The Desert and the Sown: Nomads in the Wider Society. Berkeley, University of
California Press.
Papadopoulos, J. K. and R. M. Leventhal (eds.)
2003 Theory and Practice in Mediterranean Archaeology: Old World and New World
Perspectives. Los Angeles: Cotsen Institute of Archaeology at UCLA.
Rockman, M. A.
2003 Knowledge and Learning in the Archaeology of Colonization. In M. A.
Rockman and J. Steele (eds.), Colonization of Unfamiliar Landscapes: The
Archaeology of Adaptation. New York, Routledge: pp. 3–24.
Rosen, S. A.
1992 Nomads in Archaeology: A Response to Finkelstein and Perevolotsky. Bulletin
of the American Schools of Oriental Research 287: pp.75–85.
1998 The Development of Pastoral Nomadic Systems in the Southern Levantine
Periphery: An Economic Model Based on Archaeological Evidence. In M.
Pearce and M. Tosi (eds.), Papers from the EAA Third Annual Meeting: Pre-
and Protohistory. British Archaeological Reports International Series 717. Oxford,
Archaeopress: pp. 92–97.
2002 The Evolution of Pastoral Nomadic Systems in the Southern Levantine
Periphery. In E. van den Brink and E. Yannai (eds.), In Quest of Ancient
Settlements and Landscapes: Archaeological Studies in Honour of Ram Gophna. Tel
Aviv, Ramot Publishing: pp. 23–44.
Rosenberg, M.
1998 Cheating at Musical Chairs: Territoriality and Sedentism in an Evolutionary
Context. Current Anthropology 39: 5, pp. 653–681.
Rowton, M. B.
1974 Enclosed Nomadism. Journal of the Economic and Social History of the Orient
17: pp. 1–30.
Saidel, B.
1998 Arid Zone Pastoralists in the Early Bronze Age in the Southern Levant.
Harvard University (PhD dissertation).
Salzman, P. C.
1972 Multi-resource Nomadism in Iranian Baluchistan. In W. Irons and N. Dyson-
Hudson (eds.), Perspectives on Nomadism. Leiden, E. J. Brill: pp. 60–68.
Schloen, J. D.
2001 The House of the Father as Fact and Symbol: Patrimonialism in Ugarit and the
Ancient Near East. Winona Lake,Eisenbrauns.
Trigger, B. G.
2006 A History of Archaeological Thought. Cambrigde, Cambridge University Press
(second edition).
Van Neer, W.
1993 Limits of Incremental Growth in Seasonality Studies: The Example of
the Clariid Pectoral Spines from the Byzantino-Islamic Site of Apamea
(Syria, 6th–7th Century A.D.). International Journal of Osteoarchaeology 3: pp.
119–127.
Van Neer, W., A. Ervynck, L. J. Bolle and R. S. Millner
2004 Seasonality Only Works in Certain Parts of the Year: The Reconstruction
of Fishing Seasons Through Otolith Analysis. International Journal of
Osteoarchaeology 14: pp. 457–474.
Veth, P., M. Smith, P. Hiscock (eds.)
2005 Desert Peoples, Archaeological Perspectives. Oxford, Blackwell.
Wendrich, W. Z. and R. T. Cappers
2005 Egypt’s Earliest Granaries: Evidence from the Fayum. Egyptian Archaeology
27: pp. 12–15.
Wylie, A.
1985 The Reaction Against Analogy. In M. B. Schiffer (ed.), Advances in Archaeological
Method and Theory, Volume 8. New York, Academic Press: pp. 63–111.
P a rt I
The Past
at P r e s e n t
Ch a p t e r 2
T hings to Do with
Sheep and G oats
Neolithic Hunter-Forager-
Herders in North Arabia
Alison Betts
F or several decades during the mid 20th century CE, after the spec-
tacular discovery of the Pre-Pottery Neolithic A (PPNA) tower at Jericho
(by Kenyon) had placed the prehistory of the southern Levant firmly in under-
graduate textbooks, the Late or Pottery Neolithic was an orphan period, lost
between the cultural riches of the Pre-Pottery Neolithic B (PPNB) and the
equally colorful evidence for the Chalcolithic period. For some time it was even
suggested that the southern Levant was largely abandoned during the early
part of this period (Kenyon 1957; Perrot 1968; Mellaart 1975:67–68; Moore
1985). Only after the 1980s did this picture begin to change (Table 2.1). While
increasingly intensive study of the Mediterranean climatic zone has produced
evidence for a number of Late Neolithic settlements (Garfinkel 1993; Gopher
1993; Gopher and Gophna 1993; Kafafi 1993; Banning et al. 1994), what has
been much more surprising is that extensive surveys in the steppe and desert
regions have shown that this period is widely represented from the west Jordan
highlands to the Euphrates Valley in Iraq and from the hills around Palmyra to
the fringes of the Nafudh desert (Garrod 1960; Akazawa 1979; Betts 1987; Betts
and Helms 1987, 1993; Garrard et al. 1994a, 1994b, 1996; Stordeur 2000).
The floruit of the PPNB saw the growth of large villages with elaborate
architecture, a rich symbolic corpus, long-range exchange networks and indica-
tions of fairly complex social organization (Kuijt and Goring-Morris 2002). The
transition from the Middle PPNB to the Late PPNB entailed dramatic shifts
in economy and settlement. The occupational focus moved from the core areas
of the Mediterranean climatic zone to the steppic interface zone in the east.
Concurrently, there is increasing evidence for seasonal or intermittent use of the
steppe and desert regions beyond the effective limits of dry farming. By the final
PPNB/PPNC a number of the larger settlements declined or were abandoned.
Some, such as 'Ain Ghazal in central Jordan, continued, although the period is
poorly represented archaeologically. There are also new sites established along
25
26     A l i s o n B e tt s
Table 2.1. Chronology of the Neolithic of the Near East (after Horwitz et al. 1999:64–65):
Period Age BP (uncalibrated) Age BCE (calibrated) Millennium BCE

Pre-Pottery Neolithic A 10,200–9400 9800–8400 10th–9th
(PPNA)
Early Pre-Pottery 9400–9200 8400–8100 Late 9th
Neolithic B (EPPNB)
Mid Pre-Pottery 9200–8500 8100–7500 Early 8th
Neolithic B (MPPNB)
Late Pre-Pottery 8500–8100 7500–7000 Late 8th
Neolithic B (LPPNB)
Pre-Pottery Neolithic 8100–7600 7000–6500 Early 7th
C (PPNC)/Terminal
PPNB
Pottery Neolithic/Late After 7600 BP After 6500 BCE Late 7th onward
Neolithic
the Mediterranean coast (Gopher 1993; Kuijt and Goring-Morris 2002:414). By

the PPNC significant changes in the entire complex of animal resource procure-
ment had occurred in the southern Levant with the introduction of sheep, cattle
and, probably, pig to the domestic economy (Horwitz et al. 1999:70; Kuijt and
Goring-Morris 2002:417). With the Late Neolithic a very new pattern of settle-
ment emerges. What had been a culturally rich and innovative period devolved
into scatters of smaller sites and hamlets in the Mediterranean climatic zone
and along the coast, with an unprecedented spread of short-term campsites in
previously largely unexploited steppe lands.
Within the steppe, too, the changes were dramatic. During the Mid to Late
PPNB the steppe was exploited by hunter-gatherer groups living in seasonal
encampments (Betts et al. 1998; Byrd 1992; Garrard et al. 1994a, 1994b, 1996).
The large herds of gazelle that roamed the steppe were a favored prey (Betts et
al. 1998). Exploitation of the steppe, however, appears to have been restricted to
certain ecologically favored locations. In Jordan this included the oasis of al-Azraq
and the gorge at Wadi Jilat where deep pools offered out-of-season water supplies
(Garrard et al. 1994a, 1994b, 1996), as well as the eastern basalt harra where
deeply incised valleys (wadis) provided seasonal rain pools and the landscape
was suited to the mass exploitation of gazelle (Betts et al. 1998). In Syria, PPNB
sites have been found in the Palmyrene oasis (Akazawa 1979) and in the region
of natural springs at al-Kowm (Stordeur 2000). With the PPNC, and certainly
by the beginning of the Late Neolithic, the steppe sites saw the introduction of
sheep and goats, a development that heralds the start of the spread of short-term
campsites along almost all the wadi systems across the Badiyat ash-Sham. Sites of
this period are still also found clustered around the better water sources so that
they continue at al-Azraq and Wadi Jilat (Garrard et al. 1994a, 1994b, 1996),
Things to Do with Sheep and G o at s      27
while much further east around the seasonal rain pool at Burqu' there is very little
evidence for the PPNB, but the Late Neolithic is strongly represented. In Syria
and Iraq similar sites can be found along most of the deeper wadis.
The weight of evidence suggests that sheep were first domesticated in
southeast Turkey (Peters et al. 1999), spreading out from here and southward
(Horwitz et al. 1999:76). In the southern Levant they occur in significant
numbers at 'Ain Ghazal by the Late PPNB (Horwitz et al. 1999:71; Martin
1999:88). While goats occur in significant numbers earlier than sheep on
southern Levantine sites, there is less consensus on whether they were domes-
ticated there or introduced as a domesticate, although several scholars have
at least postulated some form of ‘proto-herding' (Horwitz et al. 1999:77;
Martin 1999:88; Ducos 1993). In the eastern Jordanian steppe, sheep and
goats form a significant part of faunal assemblages by the PPNC (around 6900
BCE calibrated) and are assumed to appear as introduced domesticates from
the west (Horwitz et al. 1999:75). This constitutes a fairly dramatic change.
Epipaleolithic faunal assemblages are dominated by gazelle with some equid,
whereas PPNB sites have mostly gazelle, hare and fox. The introduction of
sheep and goats signals a marked change in the nature of steppic economies.
The End of the PPNB

The reasons for the decline of the later Pre-Pottery Neolithic have been
debated at some length. Environmental degradation caused by extensive
deforestation and overgrazing has been suggested as a strong contributing
factor (Rollefson and Köhler-Rollefson 1989; Rollefson et al. 1992). The
Pre-Pottery Neolithic saw the rise of the first large population aggregations in
the Near East. They required wood for fuel, construction and manufacture of
the plaster that was a major feature of their house interiors. As the first mixed
farmers in the region, they cleared the woodland for fields and grazed goats on
the open pastures surrounding the settlements. It has been suggested that the
inhabitants gradually began to find that their immediate environment offered
insufficient support to sustain them, slowly bringing about a breakdown of the
larger settlements into smaller and more scattered units. This interpretation is
disputed by some scholars, however, and in any case should not be considered
a monocausal solution to changes that took place over a wide range of envi-
ronmentally diverse regions (Bar-Yosef 2002:53–54).
Social pressure may also have played a role. The PPNB villages represented
the largest sedentary communities ever at that time. Interpersonal stress might
have forced apart communities whose social structures, still rooted in a hunter-
gatherer tradition, may not have been able to cope adequately with conflict
management and large-scale cooperative organization. The complex question
of PPNB social structure has been addressed from a number of differing

perspectives (Kuijt and Goring-Morris 2002), but the nature of the issue makes
it difficult to test against the material evidence. An alternative, or possibly
an additional causal factor, may be climate change. Here too the evidence
is problematic, with considerable regional disparity in terms of the data. As
more information is gathered, there is some consensus in the broader picture
but less in the detail. One interpretation is that the Terminal PPNB/PPNC
coincided with a cooler and more humid climate (8000–7500 BP uncalibrated;
Issar 1998:115), which was followed in turn in the Late Neolithic by increased
warming (7500–7000 BP uncalibrated; Issar 1998:115) and reduced rainfall
(6400–6200 BCE calibrated; Bar-Yosef 2002; Rosen 1998:229). A variant
using much of the same evidence sees a rise in warmer and moister conditions
through the early to mid Holocene interrupted briefly by a colder and dryer
period between 8000 to 7600 BP (uncalibrated) coinciding with the Terminal
PPNB/PPNC (Sanlaville 1997; Wasse 2000). There is also the possibility of a
northerly move of the monsoonal belt during the early Holocene, producing
summer rainfall that would have a marked impact on land-use potential of
marginal steppic regions (al-Moslimany 1994). In general, it is likely that
climate change was a factor in the breakdown of the PPNB agglomerating
settlements and the subsequent developments in the Late Neolithic, but it is
difficult at this stage to define its influence more precisely.
Models of Steppic Economic Str ategies

Martin (1999) has presented detailed evidence for the introduction of sheep and
goats to the east Jordan steppe by the PPNC/Late Neolithic period. She has
also discussed the two main models offered to explain these developments and
presented her own alternative model. Köhler-Rollefson (1988, 1989a, 1989b,
1992), arguing principally in relation to evidence from 'Ain Ghazal, suggests
that nomadic pastoralism developed in eastern Jordan at some time from the
PPNC onward. She proposes that, as a result of environmental pressures on
the potential grazing lands in the immediate vicinity of 'Ain Ghazal, groups
of herders moved out into the steppe in spring and early summer, eventually
becoming specialized pastoralists. She linked this to a shift from meat to milk
production to ensure a food supply without diminishing herd size. The model
is based both on archaeological evidence and analogy with patterns of seasonal
herding practices among modern Bedouin agropastoralists. Byrd (1992) has
presented a very different model involving the integration of domestic resources
into the economy of indigenous hunter-gatherers. He suggests that expansion
and colonization by settled village populations within the fertile Mediterranean
climatic zone in the PPNB pushed indigenous hunter-gatherer groups out into
the steppe where, in order to survive on reduced resources, they responded

by diversifying and exploiting available food sources more intensively. An
example of this might be the development of game traps (the so-called desert
kites) in this period. They then also selectively integrated domesticated crops
and animals into their economic structure to supplement uncertain supplies of
seasonal wild produce with a more predictable yearly return.
Martin (1999), countering the model put forward by Köhler-Rollefson,
argues that the proportions of 20–50% animal bones in the faunal record on
steppe sites does not equate with either the fattening of flocks for slaughter at
village sites or total reliance on ovicaprine pastoralism by steppe populations.
She suggests rather that the proportion of ovicaprine bones is similar to that of
wild animals such as gazelle or hare, which appears to indicate that the steppe
peoples did not specialize in any particular form of resource procurement but
rather were generalist herders, hunters, foragers, trappers and, possibly occa-
sionally, cultivators as well (Martin 1999:97). Cereal remains have been found
at Jilat 13 (Garrard et al. 1996) and chance agriculture is a common practice in
modern times in the North Arabian steppe. Martin also discusses the ways in
which the flocks may have been managed, specifically whether they were used
for milk or solely for meat. Although the evidence is not strong, she believes that
a conservative interpretation suggests culling for meat only (Martin 1999:100).
She goes further to address the question of whether the Late Neolithic steppe
population comprised herders who moved out from the PPNB villages and
adopted additional food procurement strategies to survive effectively or PPNB
hunter-gatherers who adopted herding. Based on continuity in kill patterns of
gazelle and hare across the PPNB/PPNC-LN boundary, she concludes that the
latter is more likely and thus broadly supports Byrd's model.
Evidence from the Harr a and the Hamad

Extensive surveys and a number of excavations by Garrard at Wadi Jilat and
al-Azraq and by Betts in the basalt covered steppe (al-harra) east of al-Azraq
and in the open gravel plains beyond (Figure 2.1; Table 2.2), leading down to
the Euphrates in Iraq (al-hammad), suggest that Martin's conclusions offer
the most reasonable interpretation of the development of herding in the east
Jordanian steppe. In the conclusion to her article she reminds us that, as Ingold
(1980) has pointed out, the keeping of domestic animals involves different
social relationships from hunting, and it also involves different ecological
needs. The changes in behavior necessitated by adoption of a major new
economic strategy are reflected in markedly different patterns of site type and
distribution in the PPNC-LN. Rosen (2002), in a recent overview of the first
herders in the Sinai and Negev regions of the southern Levant, also postulates
Figure 2.1. Map of Jordan, showing the position of sites mentioned in the text.
a similar interpretation, with a hunter-gatherer phase in the PPNB, followed

by a ‘herder-gatherer' phase where domesticates are adopted into peripheral
systems in the Late Neolithic period.
Although there is wide variety in steppic sites of this period, there is also a
broad bipartite division that can be made between sites with a small range of
material traces widely dispersed across the steppe and sites with more complex
remains that tend to cluster in the vicinity of permanent or semipermanent water
sources. Those in the first group are characterized by great frequency, shallow
but sometime extensive spreads of artifacts, few or no associated structures
and, frustratingly, a tool kit constituting an extraordinarily high proportion of
a distinctive chipped-stone artifact, the function of which is unknown. The only
use so far clearly identified for this, the concave truncation burin, is as a core
for the production of spalls that were then used as drill bits. Bead manufacture
Table 2.2. Radiocarbon Determinations for Sites Excavated by Betts in the East of the Jordanian
Steppe:
Site (lab. code) Age BP Age BCE (1)a Age BCE (B)a Millennium
BCE
Dhuweila 1 (OxA-1637) 8350±100 7500–7270 7540–7140 (0.93) Mid–late 8th
7120–7060 (0.07)
Dhuweila 1 (BM-2349) 8190±60 7260–7170 (0.45) 7420–7350 (0.07) Mid–late 8th
7160–7050 (0.55) 7320–7030 (0.93)
Burqu' 35 (OxA-2770) 8270±80 7470–7250 (0.75) 7480–7460 (0.02) Mid–late 8th
7220–7200 (0.06) 7440–7050 (0.98)
7180–7140 (0.12)
7120–7100 (0.07)
Burqu' 35 (OxA-2769) 8180±80 7270–7040 (1.00) 7270–7040 (1.00) Late 8th
6840–6810 (0.01) Early 7th
Burqu' 35 (OxA-2768) 8140±90 7280–7010 (1.00) 7420–6990 (0.85) Late 8th
6970–6770 (0.15) Early 7th
Burqu' 27 (OxA-2766) 7930±80 7030–6965 (0.24) 7060–6640 (0.99) Late 8th
6950–6930 (0.07) 6620–6610 (0.01) Early 7th
6920–6880 (0.14)
6860–6850 (0.01)
6830–6690 (0.54)
Burqu' 27 (OxA-2765) 7350±80 6230–6050 (1.00) 6360–6270 (0.14) Late 7th
6260–6000 (0.86)
Burqu' 27 (OxA-2764) 7270±80 6170–6010 (1.00) 6230–5950 (1.00) Late 7th
Jebel Naja (OxA-375) 7430±100 6380–6170 (1.00) 6430–6040 (1.00) Mid–late 7th
Dhuweila 2 (OxA-1729) 7450±90 6370–6190 (1.00) 6430–6110 (0.95) Mid–late 7th
6090–6050 (0.05)
Dhuweila 2 (OxA-1728) 7140±90 6050–5930 (0.07) 6160–6140 (0.03) Late 7th
5920–5860 (0.30) 6130–5770 (0.97) Early 6th
Dhuweila 2 (OxA-1636) 7030±90 5950–5780 (1.00) 6010–5690 (1.00) Late 7th
Early 6th
Burqu' 03 (OxA-2808) 6900±100 5830–5640 (1.00) 5950–5900 (0.07) Early 6th
5890–5590 (0.93)
a
Dates are calibrated using CALIB3.0.3.c Data set 1 (bidecadal tree ring data set to 9440 BCE
calibrated) and calibration method B (probability method).
was a significant part of the cultural package. The chipped-stone repertoire is

normally very restricted, but the tools tend to be fairly large with generous use
of raw material. These ‘burin Neolithic' sites have been found distributed very
widely across the steppe. They are typically located in sheltered positions on
terraces overlooking the larger wadi systems, which supply year-round grazing
and seasonal water. Depending on rainfall, wadi systems right across the steppe
hold water in pools well into the summer months. These act as conduits through
the steppe, providing corridors linking the Levant with the Tigris-Euphrates
valleys. Sites of this type have been found near el-Kowm (Stordeur 1993) and
the Palmyra basin (Akazawa 1979) in Syria, in the wadis of western Iraq (Garrod
1960), and in northern Saudi Arabia (Garrod 1960; Parr et al. 1978) and are
widely distributed across eastern Jordan from the Iraq border (Betts et al. 1991)
to the edge of the Jordan Valley (Rollefson et al. 1982).
Sites in the second broad category are much more varied. They have stone
structures with hearths and pits, a wide range of chipped-stone and other arti-
facts, imported items and concentrations of faunal and botanical remains. They
are relatively rare and cluster around significant sources of water. The main
examples are the sites adjacent to the gorge in Wadi Jilat, at the oasis of al-Azraq
(Garrard et al. 1994a; b), and Burqu' in eastern Jordan, where an unusually large
rain pool may have provided year-round water supplies in prehistoric times (Betts
et al. 1990, 1991). Another set of sites has been found in the center of the basalt
harra at the south end of Dhuweila (Betts and Helms 1987), where large seasonal
lakes and deep-cut pools may also have provided long-term water supplies. One
more site, which does not fit into any of these categories and can be placed in a
group of its own, is the Late Neolithic occupation at Dhuweila that appears to
be a seasonal hunting camp (Betts et al. 1998). The sites detailed below are only
selected examples of a wider spectrum presented to give a general picture of the
range and diversity of Late Neolithic sites in the North Arabian steppe.
Jebel Naja
An excavated site of the ‘burin Neolithic’ type, Jebel Naja had shallow occu-
pation levels with ephemeral hearths, minimal traces of faunal and botanical
remains, and an area where bead making was carried out. The chipped-stone
tool kit was highly restricted and there were very few other artifacts. The site
was located on the edge of the basalt harra in a sheltered position on the lower
slopes of the massif, overlooking open country with extensive grazing. Water
may have been available in pools in the surrounding wadis but no pools were
immediately adjacent to the site. Based on the minimum number of individuals,
out of a total of eight animals identified from the faunal remains, four were
sheep-goat, two were gazelle, and two were hare.
Wadi Jilat
A number of Neolithic sites are clustered along the rim of the Wadi Jilat gorge
and large open-air Epipaleolithic sites have also been found nearby. The bed
of the gorge contains large, deep pools that retain water for long periods after
flooding, providing a potential year-round water source. The location is in the
limestone steppe on the western edge of the steppe, not far from the modern
limits of the Mediterranean climatic zone. The earliest Neolithic sites date to
the PPNB; they continue on into the Late Neolithic and possibly later. The
sites vary but are generally characterized by subcircular structures made with
one or more rows of upright slabs supported by rubble packing. These may
well have been covered with an organic superstructure. Each site constitutes
a cluster of structures. The chipped-stone tool kits contain a varied array of
artifacts, beads were produced at some sites, ground stone artifacts were found
in moderate amounts, and there were a small number of imported items. The
faunal remains included sheep-goat bones as a significant but not dominant
proportion of the total number of animals and the botanical remains included
cereals, as well as locally available wild plants.
Qasr Burqu'
Qasr Burqu' lies beside a large rain pool that has been augmented in recent
times by a dam to create a lake. The number of prehistoric sites clustering
around the lake shore indicate that it was always an important water source.
The pool is on the eastern edge of the basalt harra at the margins of the
limestone hammad. Apart from a small scatter of Epipaleolithic artifacts, the
sites date mostly from the Late/Terminal PPNB to the Late Neolithic. The
sites consist of small clusters of stone-built subcircular cells constructed using
a combination of dry stone walling and upright slabs supported by rubble
packing. They have some depth of stratigraphy (indicating repeated re-occupa-
tion), internal hearths and pits and a wide range of chipped-stone artifacts. The
chipped-stone tools are moderate in size and there is evidence for increasingly
economic use of raw material in the later periods, with greater emphasis on
use of poor-quality local sources. The sites also have considerable numbers of
ground stone tools and some imported decorative items. The faunal remains
from the sites include significant numbers of sheep-goat remains. At Burqu'
27 (McCartney 1992), for example, sheep-goat bones counted for over 50%
of the total recovered from the site.
Qa' al-Ghirqa
Qa' al-Ghirqa is a mudflat within the basalt harra. Near its southern end,
where a fairly deeply incised wadi has carved a long pool, is a string of sites of
a particularly distinctive type. None have been excavated, but a number have
been extensively surveyed. The sites comprise clusters of circular structures,
in some cases linked by enclosing walls. Some structures have corbelled roofs
and narrow corridor entrances. There are also some examples of upright slab
construction. There are very few ground stone artifacts, but there is evidence
of bead making. The chipped stone tool assemblages are moderately varied
with a significant proportion of concave truncation burins. The chipped stone
assemblages are particularly characterized by the small size of all the artifacts,
yet the sites are only 20 to 30 km from plentiful sources of raw material. No
faunal or botanical remains have been recovered.
Dhuweila
Dhuweila was first occupied in the PPNB by hunter-gathers specializing in
mass killing of gazelle using large stone-built animal traps called ‘desert kites.'
The site is located in the basalt harra on a low promontory overlooking a
large mudflat. There is one main stone shelter built solidly of basalt using a
combination of rough dry stone walling and upright slabs with rubble foot-
ings. The interior was paved with large flat stones, some of which were used
as grinding slabs. The extensive faunal remains consisted of 93% gazelle, with
only 5 identifiable specimens (0.2%) of sheep-goat, and the rest of the remains
comprising birds, fox, equid and hare. After a period of abandonment the site
was re-occupied, but the economy appears to have been similar. Despite the
absence of the large PPNB processing areas with hearths, ash and fire-cracked
stones, the faunal remains consisted of 94% gazelle, with 39 identifiable
specimens (0.5%) of sheep-goat and the rest of the faunal remains as in the
earlier period. There is no positive evidence for domesticated animals. The
botanical remains were all from wild plants that occurred locally and included
edible roots. Evidence suggests that the site was probably occupied seasonally
in winter and early spring. There are limited water sources from rain pools in
the vicinity, but none are large enough to offer supplies into summer. The large
mudflat floods extensively in heavy rain and attracts seasonally migrating birds.
Bones from a Little Bittern were found. This bird requires standing water
for its natural habitat. Finds included a wide range of chipped-stone tools,
numerous ground stone artifacts and some imported decorative items.
Discussion
The North Arabian steppe was used in the later part of the PPNB period by
hunter-gatherer groups whose relationship to the peoples in the settled villages
is unclear. There is extensive evidence for the Epipaleolithic in the steppe,
which is followed by an apparent hiatus in the PPNA. The early to mid PPNB
is only found on the margins of the steppe, such as at Wadi Jilat, while the
deep steppe was only fully re-used in the Late PPNB. This suggests that PPNB
groups had a shared origin with the peoples of the verdant lands to the north
and west. Although they shared a common tradition of chipped stone manu-
facture, there is evidence for culturally distinct practices that suggest that they
established independent populations fully adapted to steppic living. Those in the
basalt harra developed complex hunting techniques using large-scale systems
of traps, while the PPNB hunters at
Table 2.3. Relative Numbers of Sheep-Goat
Dhuweila produced a distinctive style Remains on Jordanian Steppe Sites, Expressed
of rock engraving unique to the steppe. as ‘Minimal Number of Elements’ Counts (after
Martin 1999:Table 3):
The PPNB groups appear to have been
largely confined to the western and Site Sheep/goat Total
northern sectors of the steppe, particu- Late Neolithic
larly in areas of more broken terrain. Burqu’ 27 67 177
PPNB sites have been found in Wadi Jebel Naja 4 8
Jilat and extensively within the basalt Dhuweila 2 38 6355
harra and in the Palmyrene, but very Wadi Jilat 13-3 143 775
little evidence for PPNB occupation is
Wadi Jilat 13-2 164 625
known from the open hammad to the
Wadi Jilat 13-1 387 1782
east. By contrast, most Late Neolithic
Wadi Jilat 25 57 106
sites occur around the edges of the
Azraq 31 281 1217
harra and extensively across the eastern
Late PPNB
hammad. Clearly, there is a complete
shift in land-use patterns that must be Azraq 31 2 56
associated with corresponding changes Dhuweila 1 5 2198
in economic strategies. Equally clearly, Mid/Late PPNB

these must bear some relation to the Wadi Jilat 7-5 0 79
introduction of domestic herds to the Wadi Jilat 32 0 155
traditional hunter-gatherer pattern of Mid PPNB
subsistence (Table 2.3). Wadi Jilat 7-2 0 440
The changes in site location fit Wadi Jilat 7-3 0 172
closely to the new requirements of Wadi Jilat 7-4 0 233
herding, specifically a greater need Wadi Jilat 26 0 12
for water and grazing. The locations
Early PPNB
selected for ‘burin Neolithic’ sites have
Wadi Jilat 7-1 0 245
been almost invariably re-used by more
recent sheep-goat pastoralists, indicating that herding needs strongly governed

the choice of site location. As Martin (1999) points out, however, these people
were not exclusively herders. At most, sheep and goats supplied about half
of their meat needs while they continued to hunt and collect wild plants, as
well as practicing occasional agriculture in wetter areas. Imported items on
the sites indicate contact with lands beyond the steppe and imply some form
of exchange system, something that was already in place in the PPNB. What
is also significant is the marked variety in site types, artifact assemblages and
cultural practices. Given the need for wide-ranging movement associated with
a hunter-herder-forager lifestyle, this variety cannot be explained in terms of
local regional adaptations. What it suggests, rather, is a variety of differing
responses on the part of the former hunter-gatherers to the adoption of a new
resource. With the increased need for water, it may also mean more marked
‘ownership’ of dry-season water sources by particular groups who then each
begin to develop slightly different cultural signatures. Martin (1999:97–98) is
cautious about the faunal evidence for seasonality. The weight of archaeological
evidence, however, leans quite strongly toward the likelihood that the steppe
did support a population of hunter-gatherer-herders year-round. The burin
sites are indicative of wet season use when grazing was plentiful and water pools
were widespread across the steppe. The high frequency of sites and low levels
of artifacts found there suggest short-term occupation and regular movement.
The sites clustered around the more permanent water sources, with more
robust structures, greater depth of stratigraphy and a wider range of artifacts,
suggest that the steppe was certainly used for a substantial portion of the year
when water supplies were limited.
Rightly, much has been made of the problems associated with use of
ethnographic analogies to attempt to understand nomadic groups in the archae-
ological record. However, some of these issues relate to the way in which such
evidence is applied. Some authors place stress on the symbiotic relationship
between pastoralist and farmer, with the implication that this is a fundamental
requirement of successful pastoral practice. Often, too, there is an assumption
that the pastoralist is almost fully dependent on the herds, to the exclusion
of other economic opportunities. While this may be true for some recent
herding groups, it does not necessarily apply in the North Arabian steppe,
as has been shown in a perceptive study of the Rwalla Bedouin by Lancaster
(1981). Although they appear to have been predominantly pastoralists in the
recent past, a closer look at the details of their activities shows a more complex
picture. Lancaster stresses the notion of the ‘multi-resource nomad,’ analyzing
all aspects of social and economic life in terms of ‘assets and options.’ Access
to anything from social contacts, to good grazing, to a job opportunity in the
trucking business is a potential asset. Balancing these options to maximize
economic success in a world of scattered and varied resources is a key skill in

Bedouin life and one that makes the Bedouins remarkably adaptable to change
while the constant, the glue that holds the system in place, is a particular social
system stressing the importance of familial and ‘tribal' relationships. This is not
to say that the modern Bedouin system is in any way directly comparable to
the situation in the Late Neolithic. For a start it is heavily reliant on a complex
social structure that has developed over hundreds, if not thousands, of years.
The model of a multiresource system, however, where adaptability, choice
and variety permit survival in a fairly marginal environment, is a useful one to
consider in examining the nature of the Late Neolithic of North Arabia and
the introduction of herding as one component among a number of different
subsistence strategies.
If Byrd’s (1992) model is correct, then the steppe populations were faced with
constant adaptation and change from the mid-PPNB onward and apparently
managed this condition with a considerable degree of success. In particular,
there is no need to assume economic dependence on settled peoples but rather
possibly only pragmatic contact for exchange purposes. There is no evidence
that the Late Neolithic steppe hunter-herders needed goods that could only be
obtained from agriculturalists. Formerly, hunter-gatherers had survived in the
region and this form of subsistence strategy continued in parallel with herding.
It must also be remembered that even if the climate was as hostile as it is today,
conditions would still have been markedly better. The region today is wholly
overgrazed and almost completely depleted of the rich array of wildlife it once
fostered. In Neolithic times the wadis would have been lined with scatters of
pistachio trees and dense shrub vegetation as tall as a person's head, while the
open steppe would have teemed with vast herds of gazelle, onager, oryx and
flocks of ostrich. Set against this would be the presence of predators: large cats,
hyenas and jackals. This abundance would certainly allow the successful survival
of small hunter-herder-forager groups in the deep steppe, relying wholly on
the wild resources of the land.
This thesis runs counter to the perceived notion of inevitable interde-
pendency between pastoralist and villager that may have developed later as
pastoralism became an increasingly important component of the economic
system of the steppe. It is helpful to contrast developments in North Arabia
with those in the Sinai and Negev region where a similar initial pattern of
hunter-herders has been identified (Rosen 2002). However, while the first
adoption of domestic animals by steppic groups is similar in both areas,
subsequent developments followed different patterns. The Sinai and Negev
regions are rich in mineral resources. Transport, trade and possibly primary
exploitation are all roles in which the desert/steppe population would be
likely to engage and there is evidence to support this. The Sinai Peninsula is
also the gateway to Egypt, offering further trade and transport opportunities,
while the distances to be covered are not great compared with the land mass
of North Arabia. North Arabia is devoid of significant mineral resources, with
the exception of chert and basalt, which were in demand in the later prehis-
toric periods, but these did not offer the same high-market opportunities as
the semiprecious stones and metals to be found in Sinai and the Negev. The
key advantage of the inhabitants of the North Arabian steppe was command
of the shortcut from the Levant to Mesopotamia and to the Middle Euphrates.
In the Late Neolithic, although there is evidence for cultural continuity from
the Jordan Rift to the Tigris-Euphrates basin, this was of limited significance,
but in later times as cultural development in Mesopotamia began to outstrip
that of the Levant, this must have had increasing influence on the economy of
the steppe. Another point of significance is the size of the steppic periphery.
There is evidence for a decline in population in the Negev and Sinai following
the demise of the gateway city of Arad (Rosen 2002) in the Early Bronze
Age, which Rosen suggests is linked to nomadic dependence. According to
Lancaster the options of the North Arabian steppe dwellers were much wider,
encompassing an extensive range of peripheral links with great variations in
economy, social organization and environment. Failure of one key economic
partnership still left many other possibilities to explore.
References
Akazawa, T.
1979 Flint Factory Sites in Palmyra Basin. In K. Hanihara and T. Akazawa (eds.),
Paleolithic Site of Douara Cave and Paleo-geography of Palmyra Basin in Syria II.
Tokyo, University Museum: pp. 177–200.
al-Moslimany, A.
1994 Evidence of Early Holocene Summer Precipitation in the Continental
Middle East. In O. Bar-Yosef and R. Kra (eds.), Late Quaternary Chronology
and Paleoclimates of the Eastern Mediterranean. Tucson, Radiocarbon: pp.
121–145.
Banning, E. B., D. Rahimi and J. Siggers
1994 The Late Neolithic of the Southern Levant: Hiatus, Settlement Shift, or
Observer Bias? The Perspective from Wadi Ziqlab. Paléorient 20: 2, pp.
151–164.
Bar-Yosef, O.
2002 Early Egypt and the Agricultural Dispersals. In H.-G. Gebel, B. D. Hermansen
and C. H. Jensen (eds.), Magic Practices and Ritual in the Near Eastern Neolithic.
Berlin, Ex Oriente: pp. 49–66.
Betts, A. V. G.
1987 Recent Discoveries Relating to the Neolithic Periods in Eastern Jordan. In A.
Hadidi (ed.), Studies in the History and Archaeology of Jordan, Volume 3. London,
Routledge and Kegan Paul: pp. 225–230.
Betts, A. V. G. and S. W. Helms
1987 A Preliminary Survey of Late Neolithic Settlements at el-Ghirqa, Eastern
Jordan. Proceedings of the Prehistoric Society 53: pp. 327–336.
1993 The Neolithic Sequence in the East Jordan Badia: A Preliminary Overview.
Paléorient 19: 1, pp. 43–54.
Betts, A. V. G., S. Colledge, L. Martin, C. McCartney, K. Wright and V. Yagodin
1998 The Harra and the Hamad: Excavations and Explorations in Eastern Jordan, Volume
1. Sheffield, Sheffield Academic Press.
Betts A., S. Helms, W. Lancaster, E. Jones, A. Lupton, L. Martin, and F. Matsaert
1990 The Burqu'/Ruweishid Project: Preliminary Report on the 1988 Field Season.
Levant 22: pp. 1–20.
1991 The Burqu'/Ruweishid Project: Preliminary Report on the 1989 Field Season.
Levant 23: pp. 7–28.
Byrd, B.
1992 The Dispersal of Food Production Across the Levant. In A. Gebauer and T.
Price (eds.), Transitions to Agriculture in Prehistory. Madison, Prehistory Press:
pp. 49–61.
Ducos, P.
1993 Proto-élevage et élevage au Levant sud au VIIe millénaire B.C.: Les données
de la Damascène. Paléorient 19: 1, pp. 153–173.
Garfinkel, Y.
1993 The Yarmukian Culture in Israel. Paléorient 19: 1, pp. 115–134.
Garrard, A. N., D. Baird and B. Byrd
1994a The Chronological Basis and Significance of the Late Paleolithic and Neolithic
Sequence in the Azraq Basin, Jordan. In O. Bar-Yosef and R. S. Kra (eds.), Late
Quaternary Chronology and Paleoclimates of the Eastern Mediterranean. Tucson,
American School of Prehistoric Research: pp. 177–199.
Garrard, A. N., D. Baird, S. Colledge, L. Martin and K. Wright
1994b Prehistoric Environment and Settlement in the Azraq Basin: An Interim
Report on the 1987 and 1988 Excavation Seasons. Levant 26: pp. 73–109.
Garrard, A. N., S. Colledge and L. Martin
1996 The Emergence of Crop Cultivation and Caprine Herding in the ‘Marginal
Zone' of the Southern Levant. In D. Harris (ed.), The Origins and Spread of
Agriculture and Pastoralism in Eurasia. London, University College London:
pp. 204–226.
Garrod, D.
1960 The Flint Implements. In H. Field, North Arabian Desert Archaeological Survey,
1925–50. Cambridge, Peabody Museum: pp. 111–124.
Gopher, A.
1993 Sixth–Fifth Millennia B.C. Settlements in the Coastal Plain, Israel. Paléorient
19: 1, pp. 55–63.
Gopher, A. and R. Gophna
1993 Cultures of the Eighth and Seventh Millennia B.P. in the Southern Levant.
A Review for the 1990's. Journal of World Prehistory 7: 3, pp. 297–363.
Horwitz, L., E. Tchernov, P. Ducos, C. Becker, A. Von den Driesch, L. Martin and A.
Garrard
1999 Animal Domestication in the Southern Levant. Paléorient 25: 2, pp. 63–80.
Ingold, T.
Issar, A.
1998 Climate Change and History During the Holocene in the Eastern Mediter
ranean Region. In A. Issar and N. Brown (eds.), Water, Environment, and Society
in Times of Climatic Change. Dordrecht, Kluwer: pp. 113–128.
Kafafi, Z.
1993 The Yarmoukians in Jordan. Paléorient 19: 1, pp. 101–114.
Kenyon, K.
1957 Digging up Jericho. London, Benn.
Köhler-Rollefson, I.
1988 The Aftermath of the Levantine Neolithic Revolution in the Light of
Ecological and Ethnographic Evidence. Paléorient 14: 1, pp. 87–93.
1989a Changes in Goat Exploitation at 'Ain Ghazal Between the Early and Late
Neolithic: A Metrical Analysis. Paléorient 15: 1, pp. 141–146.
1989b Resolving the Revolution: Late Neolithic Refinements of Economic Strategies
in the Eastern Levant. Archaeozoologia 1: pp. 201–208.
1992 A Model for the Development of Nomadic Pastoralism on the Transjordanian
Plateau. In O. Bar-Yosef and A. Khazanov (eds.), Pastoralism in the Levant:
Archaeological Materials in Anthropological Perspectives. Madison, Prehistory
Press: pp. 11–18.
Kuijt, I. and N. Goring-Morris
2002 Foraging, Farming, and Social Complexity in the Pre-Pottery Neolithic of
the Southern Levant: A Review and Synthesis. Journal of World Prehistory 16:
4, pp. 361–440.
Lancaster, W.
1981 The Rwala Bedouin Today. Cambridge, Cambridge University Press.
Martin, L.
1999 Mammal Remains from the Eastern Jordanian Neolithic and the Nature of
Caprine Herding in the Steppe. Paléorient 25: 2, 87–104.
McCartney, C. J.
1992 Preliminary Report of the 1989 Excavations at Site 27 of the Burqu'/Ruweishid
Project. Levant 24: pp. 33–54.
Mellaart, J.
1975 The Neolithic of the Near East. London, Thames and Hudson.
Moore, A. M. T.
1985 The Development of Neolithic Societies in the Near East. In A. Close and F.
Wendorf (eds.), Advances in World Archaeology, Volume 4. New York, Academic
Press: pp. 1–70.
Parr, P., J. Zarins, M. Ibrahim, J. Waechter, A. Garrard, C. Clarke, M. Bidmead and
al-Badr, H.
1978 Preliminary Report on the Second Phase of the Northern Province Survey
1397/1977. Atlal 2: pp. 29–50.
Perrot, J.
1968 La Préhistoire palestinienne. Paris, Supplément au Dictionnaire de la Bible.
Peters, J., D. Helmer, A. von den Driesch and M. Saña Segumi
1999 Early Family Husbandry in the Northern Levant. Paléorient 25: 2, pp.
27–48.
Rollefson, G. O., Z. Kaechele and J. Kaechele
1982 A Burin Site in the Umm Utheina District, Jabal Amman. Annual of the
Department of Antiquities of Jordan 26: pp. 243–248.
Rollefson, G. O. and I. Köhler-Rollefson
1989 The Collapse of Early Neolithic Settlements in the Southern Levant. In
I. Hershkovitz (ed.), People and Culture in Change: Proceedings of the Second
Symposium on Upper Paleolithic, Mesolithic, and Neolithic Populations of Europe
and the Mediterranean Basin. Oxford, Archaeopress: pp. 73–89.
Rollefson, G. O., A. H. Simmons and Z. Kafafi
1992 Neolithic Cultures at 'Ain Ghazal. Journal of Field Archaeology 19: pp.
443–470.
Rosen, A. M.
1998 Early to Mid-Holocene Environmental Changes and Their Impact on Human
Communities in Southeastern Anatolia. In A. Issar and N. Brown (eds.), Water,
Environment, and Society in Times of Climatic Change. Dordrecht, Kluwer: pp.
215–240.
Rosen, S.
Periphery. In E. van den Brink and E. Yannai (eds.), In Quest of Ancient
Settlements and Landscapes: Archaeological Studies in Honour of Ram Gophna. Tel

Aviv, Ramot Publishing: pp. 23–44.
Sanlaville, P.
1997 Les changements dans l'environment au Moyen-Orient de 20,000 B.P. à 6,000
B.P. Paléorient 23: 2, pp. 249–262.
Stordeur, D.
1993 Sédentaires et nomads du PPNB final dans le desert du Palmyre (Syrie).
Paléorient 19: 11, pp. 187–204.
2000 El Kowm 2: Une île dans le desert. Paris, CNRS.
Wasse, A.
2000 The Development of Goat and Sheep Herding During the Levantine Neolithic.
University College London (PhD dissertation).
Ch a p t e r 3
A n A rchaeology of
Multisited C ommunities
Reinhard Bernbeck1
I n the biblical story describing Jacob’s fate under his father-in-law, Laban,
we read: “And Laban set three days journey betwixt himself and Jacob; and
Jacob fed the rest of Laban’s flocks” (Genesis 30:38). This suggests that Jacob
and Laban lived in a spatially dispersed, partly mobile community. Nowhere,
however, in the archaeological record of the ancient Near East are such
conditions recognized. In this chapter I argue that our sense of communities
as social units with strong spatial limits is inadequate. Instead of considering
all suprahousehold communities as imagined (Anderson 1991:6), I argue that
dispersed, multisited communities are founded on a specific practice, namely
mobility. Such an idea could be of importance for ‘community’ studies in
general and for Near Eastern archaeology and history in particular (Canuto
and Yaeger 2000). It provides a caveat against reconstructions of social, political
or economic processes that are based on a sharp divide between sedentary and
mobile populations.
The idea of dispersed communities is not new. Barth (1973), in a treatise on
mobile lifestyles, maintained that the multifarious lifeways between full mobility
and complete sedentariness can be subdivided into three major categories, based
on subsistence activities: ‘mixed economies’ of herding and agriculture in which
all households perform all subsistence tasks in a nonspecialized manner, with rela-
tively low mobility; ‘integrated communities’ with two segments, one focusing
on herding, the other on agriculture; and ‘fully separate’ herding and agricultural
communities, whereby the sedentary and the mobile communities have formal
exchange relations. Barth’s ‘integrated communities’ are spatially dispersed
groups that are, by necessity, mobile to a certain extent. Anthropology has since
shifted its focus, and we find a starkly different view of mobility exemplified in
writings about transnational migration, diaspora and refugees (Clifford 1994;
Thanks are due to Steve A. Rosen and Henry T. Wright for discussions of an early
1.
version of this chapter. Geoff Emberling, Gabriela Castro-Gessner, Susan Pollock

and Marc Verhoeven gave extensive and essential critiques that improved this chapter
throughout.
43
44    R e i n h a r d B e r n b e c k
Table 3.1. Traditional and Current Views of Mobile Groups:
Traditional view of mobility Current view of mobility

Moves Regular, frequent Infrequent, mostly singular
Reason for moving Environment Political enforcement
Goal Adaptation Refuge, survival
Expectations Continuity of moves Return to origin
Long-term status of migrants Mobile Sedentary
Social structure of migrants Long-term fixed relations Expediently constituted
Typical groups Nomads, hunter-gatherers Diasporic groups, refugees
Lavie and Swedenburg 1996; Daniel 2002). In current anthropology, migra-

tions are described as movements through space, with a result of community
dispersal. Displaced people are assumed to live in a continued state of tension.
Past experience of a specific origin, or imagined origin, results in the expecta-
tion of a return. Cultural anthropology’s shift in mobility studies has gone from
a ‘habitual’ to an ‘enforced’ idea of mobility (Table 3.1). Events of the recent
past, and the concomitant creation of homines sacri (those who are outside of any
social order), fully warrant such a change (Agamben 1998).
Recently, Goldstein (2000) and Preucel and Meskell (2004:221–222) have
suggested that ideas such as diaspora are apposite for an understanding of
mobility in many past societies. This notion may be adequate to some extent,
especially in historical archaeology, but it seems that mobility is too often treated
from a ‘sedentarocentrist’ perspective, which assumes that movement through
space is an undesirable state for any community. I will show that mobility and
the dispersed, multisited constitution of communities can be a long-term way
of life, without any inherent undesirability. I approach the issue of multisited
communities and mobility in four steps. First, since I deal with a case from
the ancient Near East, I will present a number of reasons for the focus among
students of the Near East on either completely mobile or completely sedentary
ways of life, and I will examine the restrictive ideas that underlie terms such as
mobility in this field of research. Next, I discuss methods for empirical analyses
of semisedentary and semimobile lifestyles, as well as their temporal scales.
Third, I will show, by way of an example from the 6th millennium BCE site of
Fıstıklı Höyük in southeast Turkey, how long-term routines or a ‘sequential
organization’ (Walker 2002:164) can be identified archaeologically. Finally, a
few indications from Fıstıklı Höyük and surroundings make it possible to recon-
struct the dynamics of the multisited community of which Fıstıklı Höyük was
one component. My interpretation of the evidence from Fıstıklı Höyük raises
a number of questions on theoretical, methodological and practical levels that
warrant further investigations. I conclude that multisited communities may have
been more widespread than we imagine.
An Archaeology of M u l t i s i t e d C o mm u n i t i e s      45
The Mobile-Sedentary Dichotomy

Near Eastern archaeologists and historians, and to a lesser extent also cultural
anthropologists, have produced an image of past lifestyles as a stark, unrealistic
dichotomy of almost unconnected mobile and sedentary groups. Reasons for
this perception include the practices of fieldwork, analogical reasoning and
ethnoarchaeology in the Near East, ancient historical sources, Orientalist
literature and a simplifying historiography. The following discussion will briefly
address these problems.
The structures that govern archaeological fieldwork have an unacknowl-
edged impact on constructions of communities and the complex interplay
between sedentary and mobile groups. This is especially the case for scholars
based in Western countries. Excavations led by Western archaeologists are
almost invariably restricted to a specific time of the year, mainly early summer
to early fall. Thus, mobility in winter, when villagers move to towns and cities
in search of menial labor, are not experienced. In addition, the physical aspects
of fieldwork play into an overly stable perception of village populations, as
fieldwork depends on support by relatively sedentary workers. Typically, an
employer-laborer relation between researchers and local people develops that
shows signs of hire-and-fire politics, including particularly the selection of
laborers who show up consistently for work. The idea of the stable, steadily
working villager is the result of Western archaeologists’ powerful position,
which enables them largely to determine relations of production in fieldwork.
This working situation is geared toward optimizing the relation between
financial resources and scientific results and inhibits close communication with
people who pick up work one day and drop it when they see no need for further
wages (Bernbeck and Pollock 2004:371–372; Steele 2005:50–52; Pollock,
forth.). This dichotomized perspective is reinforced by chance encounters of
archaeologists with nonsedentary people. Such encounters occur often on a
purely visual basis. On travels, ‘nomads’ may be identified as those who live
in black tents and raise livestock. The construction of a fundamental divide
between sedentary people, who depend on agriculture, and pastoral nomads
is at least in part due to practical matters of fieldwork combined with ad hoc
‘evidential experience’ of nomadism. While such impressions may not be
explicitly inserted in archaeological interpretations, they have a deep influ-
ence on dispositions toward certain kinds of archaeological narratives and
against others.
A second factor that influences our mangled perception of the lifeways of
‘integrated,’ often semisedentary communities in the ancient Near East is
analogical reasoning. In its presentist logic this kind of argument produces
results similar to those of archaeological praxis, again conflating mobility and
pastoral nomadism. There are, however, many ways to construct analogies.

In the case of ancient Near Eastern mobility, archaeologists have employed
them in an unsystematic fashion. Even a brief overview shows that a mixture
of ‘historical’ and ‘formal’ analogies prevails (for terminology see Wylie 1985;
Bernbeck 1997:85–108). Methodologically, a well-developed historical analogy
would constitute a connection between an ‘ethnographic present’ and a past.
Such historical continuity needs to be firmly established in an ‘upstreaming’
method before resorting to analogical reasoning (Stahl 2001:27–31). When
present-day Near Eastern nomadic groups are used as a foil for the past (Hole
1980; Zagarell 1982; Alizadeh 2003), however, the historical continuity is
almost invariably assumed rather than researched, and some features typical
of modern times are omitted (Gilbert 1975). Geographers such as de Planhol
(1968) and Ehlers (1980:154–161) have shown that present Iranian nomadic
structures are the result of the Turko-Tartar invasions of the medieval ages
(around 1000–1500 CE), while earlier nonpermanent ways of life were char-
acterized by short-range semimobility. There is no direct continuity between
Near Eastern prehistoric populations and present-day nomads. A basic condi-
tion for historical analogies is missing. Another kind of analogy, dubbed ‘formal’
by Wylie (1985), is based on an assumption of similarities in environment and
technology between present and past. Where identified, the goal of the analogy
is to search systematically for recurring characteristics among present and past
groups. In such cases one often comes across simplifying assumptions of an
‘insignificant’ change in natural conditions between past and present (as in
Bernbeck 1993). Again, the precondition of some initial similarities between
source and subject sides of an analogy are more assumed than researched. For
present mobile groups the source side, systematic research using the Human
Relations Area Files or the Standard Cross-Cultural Sample (Murdock and
White 1969), is rare.2 One more often encounters anecdotal citations from one
or the other present tribe that lives in the same subregion of the Near East as
the archaeological mobile group. The closer these subregional connections,
the greater the danger is for ecodeterministic arguments, equating similar
environments with similar lifeways (Hole 1980).
Analogies have become what Wobst (1978) termed the “tyranny of the
ethnographic record,” especially in research on past mobility, and this is despite
their unsystematic and highly inconsistent use. The reason for such a state of
affairs is an unwarranted assumption according to which groups that were not
fully sedentary have left only skimpy, unintelligible archaeological remains.
In contrast to sedentary village or city life, mobile groups are thought to have
2.
Both standardized data sets of ethnographic sources have been rightly criticized
for inherent biases (Conkey and Gero 1991:13–14).
always and everywhere possessed minimal material culture. The underlying

idea, again equating mobile life with pastoral nomads as we know them today,
leads to further assumptions, for instance that mobile people lived in tents and
did not stay for long at any single place (but see Hole 1978:149). While ethno-
archaeological research has revealed that even tent dwellers produce significant
material remains (Aurenche 1984; Banning and Köhler-Rollefson 1986; Simms
1988; Cribb 1991), ethno-archaeology has focused too much on the analysis of
taphonomic detail, site-formation processes and postoccupational documenta-
tion of present nomad camps and not enough on the primary task of this field:
the investigation of connections between material culture and social practices
(David and Kramer 2001:6–14).3
Such a state of affairs cries out for the abandonment of analogical reasoning
when dealing with past mobile groups (Khazanov 1994:xli). Outside the Near
East, geographers and anthropologists have documented a continuum between
fully sedentary and fully mobile lifeways. Furthermore, mobility cannot be
isolated from a host of other factors such as subsistence strategies, social rela-
tions, traditional practices, economic structures, the natural environment and
rhythms of life. Mobility and sedentariness need to be inserted into a complex
past reality. Linear correlations with one specific factor, especially the adapta-
tionist idea of an increasing length of stay in one place associated with a growing
reliance on plant resources rather than animals, do not suffice (Cribb 1991:16).
The combining of these two factors, permanence and degree of reliance on
plant resources, with linear sociopolitical evolutionism has further reified these
anthropological constructs (Sadr 1991:9–11). Such schemes are insufficient
because of their restriction of past reality to a few aspects of lifeways and an
assumption of simple correlations. There is no place for groups that move their
settlements every three to four years (Trigger 1968:56), or for groups in which
a few people stay all year-round in a settlement while a substantial part of the
population moves out to pursue some kind of subsistence or other activities in
a distant location, or for combinations of these. I argue that such elements as
partial mobility and spatiotemporal scales of mobility render any categorization
based on a linear relationship between poles of mobile to sedentary, agricultural
to nomadic, egalitarian to hierarchical unrealistically reductive (Hütteroth 1959;
Khazanov 1994:17–25). It follows that a more historical approach to past people’s
3.
Despite this criticism of Near Eastern research on mobility, a lot of highly
commendable work has been done. This is especially true of recent methods for the
identification of seasonality, such as bone isotope analysis (Mashkour 2003) and dental
cementum increment analysis (Lieberman 1998). Much of the sophisticated research
on the mobility issue is, however, restricted to the Paleolithic and the transition to the
Neolithic and focuses almost entirely on aspects of seasonality.
community life is needed and that we have to develop methods that enable us to
investigate mobility patterns empirically in the archaeological record.
For an empirical approach to ancient Near Eastern nomadism we are not
restricted to archaeological sources. There are ample written documents that
mention populations that were not completely sedentary, but written sources
cannot be taken at face value (Burstein, this volume). Painstaking investigations
into political and ideological circumstances of writing, the scribes, vocabulary
meanings and undertones as well as the intent of the documents, need to be
taken into account before they can be integrated into a narrative of ancient
Near Eastern history. One fundamental bias can never be treated adequately:
written sources are the product of sedentary, almost invariably urban, people
who not only take an external standpoint when describing mobile popula-
tions but also tend to be socially and geographically distant from them (Scholz
1995:50). Overall, written sources from Classical antiquity only contribute to
the above-mentioned dichotomy between the ‘Desert and the Sown.’ Ancient
Greek sources depict nonsedentary people in Western Asia as ‘barbarian’ and
‘aggressive’ (Briant 1982:19–25). This perspective was anything but new. Older
Sumerian and Akkadian texts, although devoid of words for sedentary people and
nomads alike, use descriptions such as “those in the steppe” or “the inhabitants
of tents” to describe nonsedentary lifestyles (Klengel 1972:17; Szuchman, this
volume). Literary texts, for instance the hymn ‘Dumuzi and Enkidu,’ juxtapose
shepherd and farmer in a normative way.4 The epic of Gilgamesh opposes a
civilized urban hero to his wild, bad-mannered counterpart, Enkidu, who lives
among animals (Kovacs 1989). Historical inscriptions refer to nonurban groups
by using ethnonyms such as Guti, Yaminites and Hanaeans. The problem is
that they are known to us from an urban, upper-class scribal perspective and
are depicted as wild and basically anarchic. For instance, urban sources on early
second millennium BCE Amorite tribes of the Mesopotamian desert depict them
as “enemies of the gods,” who do not know “ritual or law” and “eat raw meat”
(Buccellati 1966:230–232, this volume; Kirsch and Larsen 1996:151). These
texts have left their marks on numerous Assyriological interpretations that speak
about “waves of invasion” (Winckler 1905), or ‘infiltration’ of the settled lands by
‘reservoirs inépuisables’ of nomads (Kupper 1957), or of “peuplades . . . turbulentes et
pillardes en quête de butin” (Dossin 1939:995). The derogatory tone of the ancient
scribal products has been given an additional Orientalist twist by highly inap-
propriate translations that use terms such as sheikh for the Akkadian word abu,
suggesting an ahistorical social structure from the second millennium BCE to
the ethnographic present (Klengel 1958–59:217; Edzard 1959).
4.
Dumuzid and Enkimdu. ETCSL translation: t.4.08.33 at http://etcsl.orinst.ox.ac
.uk/cgi-bin/etcslmac.ccgi?text=t.4.08.33# (accessed 10 Dec. 2004).
Orientalist literature by travelers, theologians, diplomats and military

personnel from the last few centuries adds another problematic dimension to
the reconstruction of mobile lifestyles in the archaeological record, reinforcing
the impression of a gulf between sedentary and mobile populations. Such
travelogs are readily referenced by European archaeologists and function as
evidence for a ‘timeless’ past. The stock of writers cited covers the 18th-20th
centuries CE (Niebuhr 1776–80; Musil 1908; Morier 1837; Oppenheim 1939).
The implicit logic in citing such works is that before a firm grip of European
colonial empires on the Middle East, local people, and especially nonurban
people, had no history.
While the polemics of ancient Mesopotamian scribes are nowadays treated
in a more sophisticated fashion (Nissen 1980; Buccellati 1990), the general
imparted knowledge of the existence of nonsedentary populations in ancient
Western Asia has led to problems in historical narratives that work on grand
scales, especially those that put a heavy emphasis on long-term demographic
trends. The best known works of this kind are survey interpretations of the
Mesopotamian lowlands and southwest Iran (Adams and Nissen 1972; Gibson
1972; Johnson 1973; Adams 1981; Wright 1981). Interpretations of aggregate
settlement data often resort to nomadism and mobility, as well as the opposite
effect of sedentarization as a convenient deus ex machina whenever a sudden
decrease or increase in sedentary populations cannot be easily explained
otherwise (Luke 1965:22). As need be, it is assumed that either nomads
settled down or that villagers “voted with their feet” (Johnson 1987:126–127;
Grewe 2002:183).
The combination of analogical reasoning, highly biased texts, their uncritical
use, and negative demographic evidence as a way of smoothing over discerned
abrupt transitions produces a highly biased past of nonsedentary groups. I
advocate a more cautious use and interpretation of the mentioned past sources
and present evidence, concomitant with a much greater emphasis than hitherto
presented on historically specific, archaeological evidence of nonsedentary
people. In this way we may obtain a broader view of past ways of living, and
dissolve the unfortunate dichotomy of the ‘Desert and the Sown’. Past life-
ways may have been much more diverse than the limited range documented
in historical and ethnographic sources reveals.
Archaeological Indicators of Mobility

The archaeology of Neolithic and other food-producing societies is, inde-
pendently of theoretical orientation, focused on sedentariness. Processual
archaeology has always assumed a drive toward sedentariness, based on
adaptationist arguments and their underlying instrumentalist reasoning. Such
writings speak less about sedentariness, or degrees of sedentariness, than of

‘sedentarization,’ assuming that the process is inevitable and almost irreversible
(Rafferty 1985:113–127; Garel 2004:75–76). Postprocessual archaeologists have
their own sedentarocentrist blinkers. A focus on the house as a ‘home,’ full of
connotations of personal protection and (assumed) biographic detail (Hodder
1991; Watkins 1990); on landscape, ‘placemaking’ and anchoring identities
(Preucel and Meskell 2004); and on memory as embedded in constructed
landscapes (Pauketat and Alt 2003) is formulated from an implicitly sedenta-
rocentrist perspective. Cultural anthropologists go even as far as suggesting
that a lack of movement may amount to resistance (Clifford 1997). Mobility
is thus treated for what it is, an ephemeral, passing way of life, while quotidian
sedentary life, just as ephemeral, is functionalized and monumentalized.
To investigate mobility, a review of the restricted use of the term in research
on the ancient Near East is in order. Most treatises contain three assumptions.
First, patterns of movement involve all members of a social group, frequently
conceived of as a small community. Second, mobility is seasonal and occurs
two or more times per year, while especially longer-term mobility is neglected
(Bar-Yosef and Rocek 1998:2–3). Third, and connected to the two other points,
evidence for year-round use of a site, derived from faunal and archaeobotanical
data, is equated with sedentariness, although this is only a necessary but not
a sufficient condition. Unless an intricate spatial analysis is performed, such
evidence means no more than that parts of a community were present year-
round (Zeder 1994:110–112; Grigson 2003:229).
The idea of a group’s dependency on annual or more frequent moves alone
does not capture all forms of mobility since it can also work on larger time
scales, as is evident from ethnographies (James 1979: Rivière 1995:197–199)
and research on the central European Neolithic, for example at Bylany or the
Aldenhovener Platte (Soudsky 1969; Lüning 1988). In the latter cases, subsis-
tence-based archaeological indicators of mobility fail because they are geared
solely toward seasonality, not toward a system closer to ‘swidden’ agriculture.
Other archaeological indicators are available, however, and have sometimes
been used. Large-scale excavations of entire settlement systems are at the basis
of the European Neolithic studies mentioned above. In southwestern American
archaeology, site size, artifact density, the ratio of internal to external space
or the placement of midden areas are taken as correlates for relative mobility
(Powell 1983; Rocek 1998:210). The problem with such general measurements
is that they allow an assessment of degrees of mobility only by comparison,
on a very general level that does not lead to the identification of historically
specific practices of mobility.
I propose here that the interpretation of stratigraphies holds much poten-
tial to contribute to investigations of long-term mobility if interpreted in the
spirit of Walker’s (2002) “sequential organization.”5 Archaeological dogma

considers that stratigraphies are particular to sites or even single excavation
units and that stratigraphic events are random (Ford, cited in Walker 2002).
The underlying assumption is that stratigraphies are the result of an entangled
mixture of large-scale to small-scale accidental events, ranging from volcanic
eruptions and wars to more continuous processes, such as decay and the detritus
of quotidian practices of inhabiting a place. The interpretation of stratigraphy,
therefore, serves almost always a purely archaeological, comparative purpose,
namely the establishment of a sitewide or regional chronology. The sole focus
on chronology as well as ‘stratigraphic particularism’ denies the possibility
that stratigraphic sequences could follow a culturally specific diachronic
chain of practices that is based on habitual ‘life cycles’ of inhabiting houses,
compounds or whole sites (but see Pauketat 2000; Tringham 2000; Joyce 2004).
In a burgeoning ‘site abandonment literature’ (Cameron 1991; Cameron and
Tomka 1993; LaMotta and Schiffer 1999:22–25), however, it is becoming clear
that the sudden end of a site or a stratum is not always and necessarily due to
site-external forces, whether natural or human. Apart from well-documented
ethnographic and archaeological instances of ritual house abandonment from
the New World (Walker and Lucero 2000:131–133), similar processes of
deliberate abandonment and house burning are read into the evidence from
Neolithic sites from southeastern Europe and western Asia (Stevanović 1997;
Campbell 2000; Verhoeven 2000). Most of these events are explained as stages
in the life cycle of houses, where the ‘death’ of a house is intentionally marked,
most often by a violent fire.
This is just a beginning in a necessary change in our understanding of
stratigraphy as actively produced by site inhabitants. I would extend such
interpretations in three directions. First, the focus on house abandonment as
a crucial point in the ‘life’ of a structure overemphasizes one kind of a rite de
passage of houses. Indeed, the whole ‘life history’ of buildings should rather
be conceptualized according to Kopytoff’s (1986) notion of “cultural biog-
raphy” (Tringham 2000). House trajectories of founding, first use, remaking
and finally abandonment are in many societies conceptualized as a mirror
of personal biographies or kin-group cycles, and that may well be valid for
the past, too (Carsten and Hugh-Jones 1995:39–42). Second, the notion of
a biography may apply not only to houses but also to whole sites (Bernbeck
5.
Walker’s article on stratigraphy has a terminological flaw: his notion of ‘prac-
tical reason’ is in no way compatible with the general one developed since Kant and
applied by social scientists such as Bourdieu. What Walker means by the term is what
Horkheimer and Adorno (1989), following in part Max Weber, called ‘instrumental
reason.’
et al. 2003). In distinction to Kopytoff’s ‘biographies,’ which are based on a

one-dimensional, temporal sequencing of contexts of objects, biographies of
sites or houses cannot and should not be disentangled from human practices
(Joyce and Hendon 2000:155–157). It is rather the dialectical relation between
spatial structures and practices that is at the core of such ‘biographies.’ With
regard to entire sites, both foundation and abandonment are still almost
invariably envisioned in terms of instrumental reasoning. Sites are founded
by groups splitting off from others because of ‘scalar stress’ (Johnson 1982),
resource depletion or other external factors (Bandy 2004) but not because of an
unquestioned and unquestionable past and its associated practices. Third, the
leap from an analysis of specific processes, such as household cycles (Pfälzner
2001:384–393), to the search for ‘typical’ biographies, from stratigraphies as a
series of singular practices to the inference of practices that constitute recursive
sequences, should be part and parcel of any interpretation of the archaeological
record. I maintain that, if identifiable, Weberian ‘ideal types,’ whether of
actions or of whole stratigraphies, serve an important heuristic function. As
Walker (2002:166) suggests, carefully recorded and evaluated stratigraphies
may be representative (socially, temporally and spatially) of a wider realm. In
those cases where sequential events transcend discrete deposits, they are part
of a past habitus.
Barrett (2000:62–65) has opposed such ‘representational’ arguments, sug
gesting an archaeology that is in all aspects particularistic and concerned with
concrete situations. Such a reduction of human anticipatory faculties to an
immediate future, however, a future that is almost present, reduces past human
beings to animals with a practical consciousness. Such beings are caricatures,
able to act on their immediate spatial and temporal context without reflexive
foresight and oversight (Bernbeck 2003:46). I will try to show by way of a
concrete case how a single-site stratigraphy allows the establishment of a site
‘biography,’ and that an extension beyond it to a whole region allows us to
conceptualize the dialectics of mobility and multisited communities.
The Case Study of Fıstıklı Höyük

The stratigraphy of Fıstıklı Höyük, a small hamlet from the Halaf period in
southeast Turkey, serves as an example for how the biography of a site allows
the construction of mobility practices. The Halaf tradition (commonly called
a ‘culture’) dates to the sixth millennium BCE and has sometimes been inter-
preted as the first hierarchical society in the ancient Near East. Recent work at
several Halaf sites indicates a considerable degree of mobility. From the burned
remains of a very early Halaf site, Sabi Abyad, archaeologists have concluded
that storage buildings were the repositories for the goods of a nonsedentary
component of the population (Akkermans and Duistermaat 1997; Verhoeven

1999:203–232). At later Kazane Höyük and sites in the Balikh Valley in Syria,
there is evidence for shifting, consecutive settlements that result in a large
palimpsest composed of numerous small archaeological sites (Akkermans
1993:163-165; Bernbeck et al. 1999; Akkermans et al. 2006). Investigators
have remarked on the generally short occupation span of many Halaf settle-
ments, and the lack of architecture has often been interpreted as evidence for
a temporary camp (Hole and Johnson 1986-87:184; Akkermans and Schwartz
2003:119-120, 127; Marfoe et al 2003; Cruells 2004:29). These recent findings
cast doubt on earlier assumptions of a hierarchical, sedentary society with a
high degree of craft specialization (LeBlanc and Watson 1973; Watson 1983;
Campbell et al. 1999).
Fıstıklı Höyük is located at the eastern edge of the wide Euphrates river
valley in southeast Turkey, on a small natural elevation (Figure 3.1). It has a
shallow stratigraphy no more than 1.5 m deep. Fieldwork at this half-hectare
site was undertaken in 1999 and 2000 (Pollock et al. 2001; Bernbeck et al. 2002;
Bernbeck et al. 2003). Our initial plans were to investigate seasonal mobility
through a detailed spatial analysis of faunal and archaeobotanical remains.
Figure 3.1. Map showing the sites of the Halaf tradition mentioned in the text.
This part of our project is still underway, but longer-term site development
provides highly relevant insights into mobility beyond such seasonal patterns.
Since Fıstıklı Höyük is both shallow and small, we were able to excavate an
estimated 14% of the whole settlement, with 33% of the central area and 8%
of its periphery (Figure 3.2). Because of this unusually large percentage of
excavated surface, we are confident that the sequence of structures discovered
reflects the general ‘biography’ of the site.
Figure 3.2. Habitation by phase at Fıstıklı Höyük.

Table 3.2. Phases and Their Absolute Dates at Fıstıklı Höyük:
Phase Date (1σ)a Date (2σ)a Maximum range Number of samples

IIIx — — 5520–5320 2
IIIa 5788–5738 5840–5730 5880–5560 9
IIIb 5780–5733 5800–5725 5800–5660 2
IIIc 5980–5840 5990–5800 5990–5800 2
IV — — — 0
Combined dates per phase.

a
Despite the shortness of its occupation, Fıstıklı Höyük has a complex stra-
tigraphy. As the site was covered with trees, stratigraphic sequences of separate
excavation units could not be linked unequivocally by means of profiles. We
therefore used pottery seriation (based on the Robinson coefficient of simi-
larity; Shennan 1988:191–193) to integrate all well-stratified contexts from
different units into a single developmental scheme. We used some 13,000
sherds for this purpose. Phase definitions are based on diachronic similarity
coefficients, where low coefficients serve as phase dividers. Based on the
seriation results, we defined four major phases for the Halaf period at Fıstıklı
Höyük, with IV as the oldest, followed by IIIc, IIIb and IIIa (Table 3.2). These
phases correlate well with available radiocarbon dates, which were combined by
phase for the purpose of comparison with the relative stratigraphy (Bernbeck
et al. 2003:17–23).
There are two additional samples that are approximately 400 years later
(Bernbeck et al. 2003). At present we have no secure archaeological remains
for a Late Halaf phase IIIx of the site and need to investigate in more detail
the contexts from which these samples come. There are parallels to later
revisiting of other Halaf and Halaf-related sites such as Arjoune (Gowlett
2003), Damishliyya and Umm Qseir (Akkermans and Schwartz 2003:120).
The evidence of stratigraphy and of relative and absolute dating provides the
framework for a reconstruction of a ‘sequential organization.’ Initially, in phase
IV, Fıstıklı Höyük was used as a tent camp.6 In this phase the only alteration
of the natural rise on which the site was founded consisted of pits and a small
ditch, which likely served to divert runoff water from the site. In the subsequent
phase IIIc, a small cell-plan building was constructed on the summit of the
small hill, prior to any major habitation structure. Such cellular buildings have a
long tradition as storage buildings in earlier periods (Kirkbride 1982; Bernbeck
Positive evidence for tents consists solely in a few charred wooden sticks driven
6.
into a surface and the spatial arrangement of the pits. These remains are typical for the
earliest levels of many prehistoric Near Eastern sites, such as Matarrah or Tepe Sarab,
and indicate a temporary use of a site that later turned more permanent.
1994:242–243). Associated with this structure was a construction similar to a

dike, which we interpret as protecting the storage building from runoff.
In a next step, in phase IIIb, a second dike was built on the eastern limits of
the site. Prevention of water damage seems to have been a major concern of
the inhabitants of Fıstıklı Höyük, and there was enough foresight to build such
structures before any single house. These large structures were likely erected
by a group of people larger than a single household, and we may therefore
assume that the slow settling in of a community was a ‘project’ in a Husserlian
sense, a practical activity with foresight and planning. Then, two large round
‘tholoi,’ as the habitation structures of the Halaf tradition are called, were
erected. They were located far from each other and, judging by the proximity
of storage buildings, each tholos had its own storage unit. We may therefore
assume that the social groups that were living in each house stored their prod-
ucts separately.7
At the end of this phase a garbage dump was created at the western edge
of the site.8 Since this is the one major space for disposal of garbage that we
found, we can again assume a certain amount of intracommunity agreement
on its location. An oven, which may have been used communally, was set up on
one of the dikes. At Fıstıklı Höyük such ovens likely served to parch grain and
turn it into storable material (Clayton 2004). In this early phase of settling in
at a site, the detailed reconstruction hints at practical activities that were likely
discursively agreed on, such as communal structures, as well as other aspects
that served to reproduce an unquestioned and unquestionable lifeworld, such
as the specific and standardized forms of cellular storage buildings and round
habitation structures.
The next phase, IIIa, is when Fıstıklı Höyük reached is apogee, becoming
what we call a ‘focal site.’ With minimally five tholoi or living structures (Table
3.3), the population was much larger than in the previous phase. The richness
of cultural debris was much greater as well, although this might be due in part
to sampling bias.9 This phase, although short in absolute years, consists of
an intricate sequence of establishments and abandonment of structures and
installations. From the substantial structures that were both newly constructed
and repaired, it is clear that some people at least lived for a multiyear duration
7.
This does not mean that productive or consumption practices were as segregated.
To the contrary, there are good indications that such activities were largely carried out
outside of houses in communally accessible areas.
8.
The development of spatially segregated garbage dumps is sometimes claimed as
an indicator of semi- or full sedentariness (Panja 2003:115–116).
9.
The excavated volume of phase IIIa deposits is many times larger than that of any
other phase.
Table 3.3. Structures and Installations in Use During Each Phase of Occupation at Fıstıklı Höyük:
Phase Number of tholoi Number of ovens Number of storage Ratio of ovens to

structures storage structures
IIIa 5 5 6 1.14
IIIb 2 1 2 0.50
IIIc 0 0 1 0.00
IV 0 0 0 —
at Fıstıklı Höyük. Early in phase IIIa, a small, nonresidential round building

was added in the western part of the site, as well as two large tholoi: one in
the center, another near the garbage dump. The storage unit of one tholos
decreased in size, likely because of less need for such a space because of
decreasing numbers of people in that household. Soon after, two more storage
buildings were added to the northernmost tholoi at the site, and the old storage
building in the center was substantially repaired and extended by at least two
rooms. Another tholos was built near the garbage dump. North of the tholoi, a
large outside working space with an oven was used intensely. Two other ovens
were constructed nearer to the tholoi and the settlement center.
Finally, the site was slowly abandoned, apparently beginning with the entire
southern half. In this abandonment process three tholoi disappeared. Between
the central one and its large storage structure, a painted vessel was smashed
on the final surface. After these events two tholoi were still in use, with only
one storage building and three ovens. If the ovens served to process grain into
a storable material, this might indicate that fields around the site were still
harvested but that the processed food was at least partly stored elsewhere,
likely at another site established by people who had moved away from Fıstıklı
Höyük. During a final phase IIIx that dates several hundred years later and for
which we have so far only radiocarbon dates from problematic contexts, the
site, after having been abandoned, was visited again seasonally by a group that
did not leave any permanent structures.
To summarize this architectural sequence (Bernbeck et al. 2003), the
evidence points toward an early ephemeral use of Fıstıklı Höyük as a location
for a camp, then a more substantial, but still temporary, use during which
goods were stored at the site, likely leading to sojourns that were longer than
before. A relatively permanent establishment of people at the site happened in
phase IIIb, which quickly evolved into a full-blown, densely inhabited settle-
ment in the following phase, IIIa. At the end of this phase slow abandonment
set in, during which some households stayed while others came back from an
unknown new location to process agricultural products that were no longer
stored at the site.
Phytoliths and Stone Tools

The sequence outlined above can be supported by other evidence. I restrict
myself here to the results of phytolith analyses and some tools that were used
in the processing of animals and plants. Among the 40 soil samples from
Fıstıklı Höyük that have been analyzed for phytoliths, 26 are from well-
stratified contexts (Hassan 2002). Admittedly, interpretive conclusions drawn
from such a small sample must remain tentative and acquire their value only
in connection with other results that point in a similar direction. Phytoliths
from wheat and barley husks occur in variable amounts in these samples. A
rough classification into low (including zero), medium and high densities of
such phytoliths produces a clear result. Only the latest phase, IIIa, contains
any samples with a high density of wheat and barley husks, pointing toward
an extensive processing of these plants. All earlier phases, however, also have
some evidence of these plant remains (Table 3.4). This should not necessarily
be taken as an indication that grain was cultivated in the vicinity of Fıstıklı
Höyük from the earliest phase on. Judging by the site’s abandonment process,
when people apparently continued to practice agriculture and process materials
after having moved their houses elsewhere, it is also likely that semiprocessed
products were transported to the site in its early use phases. At least from phase
IIIc on, storing of such material would have been possible. While sample sizes
of different phases may have skewed the outcome in favor of phase IIIa, future
analysis of charred plant remains will be used to assess the extent to which
plants were processed in the early phases, especially during phase IIIb, when
substantial settling-in occurred at Fıstıklı Höyük.
Small finds provide another avenue for investigating the ‘sequential organi-
zation’ indicated by the architectural sequence. Here three items are considered.
These are, according to our present knowledge, related to subsistence activi-
ties. The most frequently occurring of these are ‘jetons,’ an enigmatic find
typical for the Halaf tradition. Jetons are sherds that have been chipped into a
Table 3.4. Phytolith Samples with Wheat/Barley Husks per Phase:
Phase Densities of phytoliths of wheat/barley husks

High Medium Low Total
IIIa 6 1 13 20
IIIb — 1 1 2
IIIc — — 2 2
IV — 1 1 2
Total 6 3 17 26
Figure 3.3. Sherd jetons from Fıstıklı Höyük.
round shape, sometimes with smoothed edges (Figure 3.3). At Fıstıklı Höyük
these items, as well as naturally round, flat river pebbles, have been systemati-
cally analyzed for the first time. Costello (2002) has shown that they can be
categorized into distinct size and material categories. Of importance in this
context is that they are spatially most closely linked to animal bones and are
likely mnemonic devices associated with practices of animal keeping (Costello
2002:212–222).
Sickle blades were used for harvesting plants. The Fıstıklı Höyük specimens
are irregular, and some of them had double uses as notched tools. In the absence
of use-wear analyses of the material, it remains unclear whether the few sickles
found were employed to cut cereal plants or reed. We also found a number of
round ‘calcareous plates’ (Figure 3.4). These were coarsely chipped from large
pieces of locally available limestone. They were shaped into flat, round disks,
some of which have one slightly discolored side, as if they had been exposed
to heat. Cracks in the stone, as well as the location of some of the stones near
hearths, support this idea (Bernbeck et al. 2003:62–63). These plates may have
been used as lids in the preparation of starchy or leafy foods, keeping them at
specific low temperatures or controlling the slow cooling of food.
Figure 3.4. Calcareous plates from Fıstıklı Höyük.
Table 3.5. Occurrences of Jetons, Sickle Blades and Calcareous Plates at Fıstıklı Höyük:
Phase Jetons Sickle blades Calcareous plates

Number
IIIa 116 12 31
IIIb 13 1 4
IIIc 5 — —
IV 2 — —
Percentagea
IIIa 73 7 20
IIIb 72 5 23
IIIc 100 0 0
IV 100 0 0
a
Given the small number of finds, especially in the three early phases, the percentage calculations across
artifact categories are far from ideal and should be seen as rough quantitative indicators only.
Table 3.5 reveals a clear break in the occurrence of these items between
phases IIIc and IIIb. Artifacts that point toward the processing of plant material
appear from phase IIIb onward. Not only were plants harvested near the site,
but the first parching oven is also documented in this phase. With this inclu-
sion of plant processing and settling in of more permanently resident groups,
it became possible to prepare special kinds of food for which heavy cooking
utensils, such as the calcareous plates, were needed. This more settled way of
life was associated with either a different or a more variable cuisine. Finally, in
at least two buildings of phase IIIa, items for processing grain were included
in the foundations. I interpret this as an intentional, symbolic re-use of objects
that had come to the site initially in phase IIIb. A grinding stone was re-used as
a doorsill, and a calcareous plate was turned into a door stone. Both these items
marked a liminal space that is neither completely internal nor external. All of
the above evidence conforms well with the interpretation of the ‘sequential
organization’ that was derived from the architectural sequence. It is important
to note that the results of the phytolith analysis (Table 3.4) seem to contradict
the notion of a transformation between phases IIIc and IIIb. I contend that this
is due to the fact that plant material was consumed in small amounts at the site
from its earliest occupation onward. The tools, and not the plant remains alone,
enable us to unravel past practices of production and processing of animals and
plants, whereas the biological remains (here only the phytoliths) reveal what
was consumed, by animals or by humans.10
How do we make the transition from the stratigraphic occurrence of such
items to an informed guess about past life in and around Fıstıklı Höyük? This
question is essential if narratives about the past aspire to be more than abstract,
bloodless models, and to address it requires a theoretical consideration of
notions of practice. I will show which kinds of practices we can reasonably infer
from the archaeological evidence and what kinds of archaeological fieldwork
are needed to further examine the validity of my interpretation.
Anchoring and Dispositional Pr actices

Ever since Bourdieu’s practice theory was integrated into anthropological
discourse, the emergence and continuity of quotidian activities have become
a mainstay of theoretical debate. In an important paper, Swidler (2001:81)
criticized the prevailing undifferentiated view of practices and asked “whether
we are simply awash in practices, each patterned and habitual, each subject to
revision as it is transposed or replicated, and none more influential than any
other.” In her discussion she argues that some practices are “anchoring whole
larger domains of practice and discourse” (Swidler 2001:87). I follow this
fundamental insight and suggest that at Fıstıklı Höyük those practices that led
to the initial use of the site, as well as its occupation and abandonment, were
anchoring many but not all other practices carried out, for example different
ways of cooking. Thus, much of what I described above in archaeological terms
as ‘sequential organization’ can be imagined as past practices that triggered
In the final analysis of the botanical remains, an effort will be made to distinguish
10.
phase by phase between what Hillman (cited in Moffett 2003:243) has called ‘producer’
and ‘consumer’ sites.
changes in others. Anchoring practices are more likely than others to have
been discursively negotiated. Only the spatiotemporal specificities were nego-
tiable, however, not the anchoring practice itself. For example, the moment
and direction of a household move may have been open to dispute but not the
necessity of the move itself. Thus, I claim that the ‘lifeworld’ (in Habermas’s
sense of Lebenswelt) of the inhabitants of Fıstıklı Höyük did not enable them to
question whether long-term, spatially small-scale mobility should be continued
or abandoned.
It is important to disentangle anchoring, sporadically occurring practices
from other habitual activities. Here it is useful to follow Husserl’s (1976)
distinction between ‘protension’ and ‘project,’ where protension refers to an
immediate future embedded so much in a practice en cours that it is asymptoti-
cally close to the present, and project denotes the future that is physically and
temporally distant and more open to reflexivity and discursiveness. Anchoring
practices may best be conceived of as projectual, whereas temporally smaller-
scale episodes are characterized by protension.11 Quotidian practices, such as
parching grain, cleaning a tholos, leaving the habitation for a nearby field or
repairing a wall are, if one wants to adhere to Bourdieu’s (1990) terms, ‘dispo-
sitional,’ carried out on an ad hoc, situational basis. Such dispositional practices
are not always quotidian. The repair of a damaged house may have happened
once a year, or less often, but was carried out in a similarly embodied fashion
as that of preparing a meal.
Anchoring and dispositional practices are always dialectically related.
Only the expectation of dispositional practices gives the anchoring prac-
tices any meaning. Without the anchoring practices’ distance from daily
routines, however, some aspirations of human life could not be accounted
for. Furthermore, anchoring practices are limited to specific fields. In the
case mentioned here, the setting up of a tent or the construction of a storage
building ground some subsistence and habitation practices but may only
marginally influence other activities. I assume that the specific anchoring prac-
tices, such as deciding to move to a new site, are recursive, not unique events.
They are less particular than many of the daily, improvised activities that were
carried out at Fıstıklı Höyük. The reason for this contention is the interlacing
of initial settling and abandoning practices. It is clear from the final phase at
Fıstıklı Höyük that the majority of the inhabitants had moved on, but some
came back to the site for some time, just as the initial settlers of Fıstıklı Höyük
still made visits to an older site they had abandoned.
Swidler (2001) also uses the phrase ‘constitutive practices,’ which is an unfortunate
11.
choice as it implies an undialectical, hierarchic order of practices.

Making Pr actices Visible

These reflections allow an extension of the archaeological organizational
sequence at Fıstıklı Höyük to other sites that underwent similar cycles from
ephemeral to focal to ephemeral use. Others may have gone through some,
but not all, stages identified at Fıstıklı Höyük (Figure 3.5). Thus, when people
pitched their first tents at Fıstıklı Höyük, they likely came from another, so far
unknown, ‘focal site’ that was structurally similar to Fıstıklı Höyük in its main
phase (IIIa). At the same time, other small groups may have fanned out with
herds, carrying agricultural products to other places in the vicinity of a focal
site. They also established camps, some on older focal sites, others on poten-
tially viable places for future focal sites. A few simple tools were carried along,
including pots and lithics. Camps are likely to have been used only seasonally,
and most of the camps never developed into full-blown focal sites.
The whole gamut of practices that were carried out at focal sites is not yet,
and never will be, fully reconstructible. Apart from agricultural and herding
activities, food processing and crafts are those practices that are most visible
in the archaeological record. The drama of daily social relations seems almost
impossible to infer from material evidence, but likely generational relationships
can be derived from the kind of community that we deal with. It is probable
that the population staying year-round at focal sites included children and the
elderly, with young adults moving out to more ephemeral sites. Thus, any
person would go through an initial stage of life at a focal site, then live a more
mobile life, and in old age again stay in such a focal site.
Shifting a focal site to a new place was probably due to events in the social
life of the community. The death of an elderly person may have triggered
the abandonment of a tholos and the establishment of a new one elsewhere.
Slowly, the whole population moved to the new site. The connection between
events in the life cycle of families and house establishment or abandon-
ment is well documented (Carsten and Hugh-Jones 1995:39–40; Walker and
Lucero 2000:133–134). No close analogy is needed to infer that such times of
shift were filled with tensions and potential for dispute and conflict but also
anticipation.
Future Fieldwork and Its Problems

It seems most appropriate to imagine multisited communities as incorporating
frequent contacts between people at the focal site and its outliers. The shift to a
new focal site therefore did not mean a radical change of the familiar landscape
in which people were living. Distances between elements of a community may
not have been greater than a few kilometers (Abdi 2003:400–402). This is also
Figure 3.5. Dynamics and structures of a Halafian multisited community on the upper Euphrates.
indicated by two excavations in the vicinity of Fıstıklı Höyük, both with major
pre- and post-Halafian components where scatters of Halaf material were
found. Zeytinlibahçe is a mere 250 m south of Fıstıklı Höyük (Frangipane and
Bucak 2001) and Mezraa is 2 km downstream (Karul et al. 2001:168–169). In
neither case do these finds of Halaf artifacts necessarily indicate a campsite
established from the focal site Fıstıklı Höyük, since exact dating cannot be
established from the little material from those scatters. However, many similar
sites can be expected to have been present at the time of Fıstıklı Höyük’s phase
IIIb and IIIa habitation. In this scenario the site of Fıstıklı Höyük was never a
‘community’ in and of itself but always a section of a much larger, dynamic one.
At the core of a multisited community was a focal site that shifted frequently.
From this hub groups left to live temporarily in camps. The archaeological
evidence for these communities can only be assembled from a dense web of
excavations or a small-scale, high-resolution survey.
At present, such archaeological work has not been carried out around Fıstıklı
Höyük. There are several reasons for this. Surveys in the surroundings of
Fıstıklı Höyük were geared toward the salvage of major sites threatened by the
building of dams (Algaze et al. 1994) and therefore were not designed to iden-
tify ephemeral sites. Furthermore, depending on the length of use, such sites
would often contain just a few artifacts, with a great likelihood of a complete
lack of diagnostic, datable pieces. And even if finds are datable to the Halaf
period, it is improbable that they were contemporary with the short occupation
span of Fıstıklı Höyük as a focal site. Furthermore, geomorphologic processes
have probably destroyed or buried most of these ephemeral sites. For example,
the small Halafian site of Tell Amarna, located 25 km downstream from Fıstıklı
Höyük on the right bank of the Euphrates, is blanketed by 2 m of erosional
material (Cruells 2004:26).
Discussion
If understood as an abstract ‘model’ of a settlement system, the results of my
analysis might not be new. They conform relatively well with Barth’s ‘integrated
communities’ mentioned at the beginning of this chapter. Transhumance,
however, is in almost all cases interpreted in terms of cultural ecology, as an
adaptation to particular natural circumstances rather than as a habitual, socially
constructed way of life. My argument here is that mobility is not enforced by
external circumstances, such as resource scarceness or patchiness, and that at
least in some cases it should be seen as ‘community practice.’ The stark contrast
between real, ‘face-to-face’ village communities and larger ‘imagined’ commu-
nities (Anderson 1991:6; Isbell 2000) is overly simple. Communities are always
both imagined and practiced. ‘Practicing community’ is the active establishment
of what Goffman (1963:17) terms ‘co-presence,’ the being of several people in

one and the same space at the same time (Giddens 1984:64–68). Mobility, as a
set of centripetal and centrifugal small-scale moves, is one of the most crucial
practices for a community that imagines itself as multisited. Such small-scale
mobility is fundamental in both establishing and removing co-presence and, as
such, is a deeply dialectical practice. I argue that the establishment of certain
structures at a site may be read as an anchoring practice. Mobility is also an
anchoring practice, in this case for the field of communicative action,12 which
is constitutive of communities.
Multisited communities may be much more widespread in human history
than we think. Recent writings on diaspora, transnational migration and refu-
gees give the overall impression that spatial dispersal of social groups is due to
political or religious oppression and dire economic circumstances. This may
well be true for the capitalist world system and some older, imperial forms of
subjugation. However, neither an ideal of a coherent local (village) commu-
nity with the closely associated assumption of sedentariness, nor the equally
close-knit small group of mobile people is a natural state of being, nor is the
stark, constructed contrast between sedentary and nomadic lifeways. Life in
dispersed communities is a perfectly ‘normal’ way of being, not just in hunting-
gathering societies but also in communities where agriculture or herding are
prevalent. Mobility is an anchor that keeps multisited communities in existence
by enabling the face-to-face, spontaneous, frequent and often aimless chat in
social encounters; a practice in which means and ends, communication and
community, collapse into each other.
References
Abdi, K.
2003 The Early Development of Pastoralism in the Central Zagros Mountains.
Journal of World Prehistory 17: 4, pp. 395–447.
Adams, R. M.
1981 Heartland of Cities. Chicago, University of Chicago Press.
Adams, R. M. and H. J. Nissen
1972 The Uruk Countryside. Chicago, University of Chicago Press.
Agamben, G. (translated by Daniel Heller-Roazen)
1998 Homo Sacer: Sovereign Power and Bare Life. Stanford, Stanford University
Press.
In other contexts practices such as commensality may anchor communicative

12.
actions (Appadurai 1981; Pollock 2003).

Akkermans, P. M. M. G.
1993 Villages in the Steppe. Ann Arbor, International Monographs in Prehistory.
Archaeological Series 5.
Akkermans, P. M. M. G. and K. Duistermaat
1997 Of Storage and Nomads: The Sealings from Late Neolithic Sabi Abyad, Syria.
Paléorient 22: 2, pp. 17–44.
Akkermans, P. M. M. G. and G. M. Schwartz
2003 The Archaeology of Syria: From Complex Hunter-Gatherers to Early Urban Societies
(ca. 16,000–300 BC). Cambridge, Cambridge University Press.
Akkermans, P. M. M. G., R. Cappers, C. Cavallo, O. Neiuwnehuyse, B. Nilhamn and
I. N. Otte
2006 Northern Syria: New Evidence from Tell Sabi Abyad. American Journal of
Archaeology 110: 1, pp. 123–156.
Algaze, G., R. Breuninger and J. Knudstad
1994 The Tigris-Euphrates Archaeological Reconnaissance Project: Final Report
of the Birecik and Carchemish Dam Survey Areas. Anatolica 20: pp. 1–96.
Alizadeh, A.
2003 Some Observations Based on the Nomadic Character of Fars Prehistoric
Cultural Development. In N. Miller and K. Abdi (eds.), Yeki Bud, Yeki Nabud.
Los Angeles, Cotsen Institute of Archaeology at UCLA: pp. 83–97.
Anderson, B.
1991 Imagined Communities. London, Verso (second, revised edition).
Appadurai, A.
1981 Gastro-Politics in Hindu South Asia. American Ethnologist 8: pp. 494–511.
Aurenche, O.
1984 Nomades et sédentaires: Perspectives anthropologiques. Paris, Éditions Recherche
sur les Civilisations.
Bandy, M. S.
2004 Fissioning, Scalar Stress, and Social Evolution in Early Village Societies.
American Anthropologist 106: 2, pp. 322–333.
Banning, E. B. and I. Köhler-Rollefson
1986 Ethnoarchaeological Survey in the Beda-Area, Southern Jordan. Zeitschrift
des Deutschen Palästina-Vereins 102: pp. 152–170.
Bar-Yosef, O. and T. Rocek
1998 Introduction. In T. Rocek and O. Bar-Yosef (eds.), Seasonality and Sedentism:
Archaeological Perspectives from Old and New World Sites. Peabody Museum
Bulletin 6. Cambridge, Harvard University Press: pp. 1–8.
Barrett, J.
2000 A Thesis on Agency. In M.-A. Dobres and J. Robb (eds.), Agency in Archaeology.
London, Routledge: pp. 61–68.
Barth, F.
1973 A General Perspective on Nomad-Sedentary Relations in the Middle East. In
C. Nelson (ed.), The Desert and the Sown. Berkeley, University of California
Press: pp. 11–21.
Bernbeck, R.
1993 Steppe als Kulturlandschaft: Das Agig-Gebiet Ostsyriens vom Neolithikum bis zur
islamischen Zeit. Berlin, Dietrich Reimer.
1994 Die Auflösung der häuslichen Produktionsweise: Das Beispiel Mesopotamiens. Berlin,
Dietrich Reimer.
1997 Theorien in der Archäologie. Tübingen, Francke Verlag.
2003 The Ideologies of Intentionality. Rundbrief der Arbeitsgemeinschaft Theorie in
der Archäologie 2: 2, pp. 44–50.
Bernbeck, R. and S. Pollock
2004 The Political Economy of Archaeological Practice and the Production of
Heritage in the Middle East. In L. and R. Preucel (eds.), A Companion to Social
Archaeology. Oxford, Blackwell: pp. 335–352.
Bernbeck, R., S. Pollock and C. Coursey
1999 The Halaf Settlement at Kazane Höyük: Preliminary Report on the 1996 and
1997 Seasons. Anatolica 25: pp. 109–147.
Bernbeck, R., S. Pollock, S. Allen, A. G. Castro-Gessner, R. Costello, S. K. Costello,
M. Foree, M. Gleba, M. Goodwin, P. Horan, S. Lepinski, C. Nakamura, S.
Niebuhr and R. Shoocongdej
2002 Preliminary Report on the 2000 Season at Fıstıklı Höyük. In N. Tuna and J.
Velibeyoglu (eds.), Salvage Project of the Archaeological Heritage of the Ilısu and
Carchemish Dam Reservoirs: Activities in 2000. Ankara, Middle East Technical
University: pp. 1–42.
Bernbeck, R., S. Pollock, S. Allen, A. G. Castro-Gessner, S. K. Costello, R. Costello,
M. Foree, M.Y. Gleba, M. Goodwin, S. Lepinski, C. Nakamura and S.
Niebuhr
2003 The Biography of an Early Halaf Village: Fıstıklı Höyük, 1999–2000. Istanbuler
Mitteilungen 53: pp. 9–77.
Bourdieu, P.
1990 The Logic of Practice. Stanford, Stanford University Press.
Briant, P.
1982 État et pasteurs au Moyen-Orient ancien. Paris, Editions de la Maison des
sciences de l’homme.
Buccellati, G.
1966 The Amorites of the Ur III Period. Naples, Istituto Orientale di Napoli.
1990 “River Bank,” “High Country,” and “Pasture Land”: The Growth of
Nomadism on the Middle Euphrates and the Khabur. In S. Eichler, M. Wäfler,
and D. Warburton (eds.), Tall al-Hamidiya. Volume 2. Göttingen, Vandenhoeck

and Ruprecht: pp. 87–117.
Cameron, C. M.
1991 Structure Abandonment in Villages. Archaeological Method and Theory. Volume
3: pp. 155–194.
Cameron, C. M. and S. A. Tomka (eds.)
1993 Abandonment of Settlements and Regions: Ethnoarchaeological and Archaeological
Approaches. Cambridge, Cambridge University Press.
Campbell, S.
2000 The Burnt House at Arpachiyah: A Reexamination. Bulletin of the American
School of Oriental Research 318: pp. 1–40.
Campbell, S., E. Carter, E. Healey, S. Anderson, A. Kennedy and S. Whitcher
1999 Emerging Complexity on the Kahramanmaras Plain, Turkey: The Domuztepe
Project, 1995–1997. American Journal of Archaeology 103: pp. 395–418.
Canuto, M. A. and J. Yaeger (eds.)
2000 The Archaeology of Communities: A New World Perspective. London,
Routledge.
Carsten, J. and S. Hugh-Jones
1995 Introduction. In J. Carsten and S. Hugh-Jones (eds.), About the House: Lévi-
Strauss and Beyond. Cambridge, Cambridge University Press: pp. 1–46.
Clayton, L.
2004 The Technology of Food Preparation: The Social Dynamics of Changing Food
Preparation Styles. Binghamton University, Department of Anthropology
(unpublished M.A. thesis).
Clifford, J.
1994 Diasporas. Cultural Anthropology 9: 3, pp. 302–338.
1997 Routes: Travel and Translation in the Late Twentieth Century. Cambridge,
Harvard University Press.
Conkey, M. W. and J. M. Gero
1991 Tensions, Pluralities, and Engendering Archaeology: An Introduction to
Women and Prehistory. In M. W. Conkey and J. M. Gero (eds.), Engendering
Archaeology: Women and Prehistory. Oxford, Blackwell: pp. 3–30.
Costello, S. K.
2002 Tools of Memory: Investigation of the Context of Information Storage in the Halaf
Period. Binghamton, Binghamton University (PhD dissertation).
Cribb, R.
Cruells, W.
2004 Area L: The Soundings. In Ö. Tunca and M. Molist (eds.), Tell Amarna (Syrie)
I: La période de Halaf. Louvain, Peeters: pp. 15–36.
Daniel, V.
2002 The Refugee: A Discourse on Displacement. In J. McClancy (ed.), Exotic No
More. Chicago, University of Chicago Press: pp. 270–286.
De Planhol, X.
1968 Les fondements géographiques de l’histoire de l’Islam. Paris, Flammarion.
Dossin, G.
1939 Benjaminites dans les textes de Mari. In Mélanges Syriens offerts à M. René
Dussaud. Tome II. Paris, Geuthner: pp. 981–996.
Edzard, D. O.
1959 Altbabylonisch nawum. Zeitschrift für Assyriologie 53: pp. 168–173.
Ehlers, E.
1980 Iran: Grundzüge einer geographischen Landeskunde. Darmstadt, Wissenschaftliche
Buchgesellschaft.
Frangipane, M. and E. Bucak
2001 Excavation and Research at Zeytinlibahçe Höyük, 1999. In N. Tuna and J.
Öztürk (eds.), Ilısu ve Karkamıs Baraj Gölleri Altında Kalacak Arkeolojik Kültür
Varlıklarını Kurtarma Projesi 1999 Yılı Çalısmaları, Ankara, ODTÜ Tarihsel
Çevre Arastırma Merkezi: pp. 65–132.
Garel, J.-R.
2004 Néolithisation et évolution cognitive au Proche-Orient. Orient Express 3: pp.
75–79.
Gibson, M.
1972 The City and Area of Kish. Coconut Grove, Field Research Projects.
Giddens, A.
1984 The Constitution of Society. Berkeley, University of California Press.
Gilbert, A. S.
1975 Modern Nomads and Prehistoric Pastoralists: The Limits of Analogy. Journal
of the Ancient Near Eastern Society of Columbia University 7: pp. 53–71.
Goffman, E.
1963 Behaviour in Public Places. New York, Free Press.
Goldstein, P. S.
2000 Communities Without Borders: The Vertical Archipelago and Diaspora
Communities in the Southern Andes. In M. A. Canuto and J. Yaeger (eds.),
The Archaeology of Communities: A New World Perspective. London, Routledge:
pp. 182–209.
Gowlett, J. A. J.
2003 The AMS Radiocarbon Dates: An Analysis and Interpretation. In P. J. Parr
(ed.), Excavations at Arjoune, Syria. British Archaeological Reports, International
Series 1134. Oxford, Archaeopress: pp. 27–30.
Grewe, C.
2002 Die Entstehung regionaler staatlicher Siedlungsstrukturen im Bereich des prähis-
torischen Zagros-Gebirges: Eine Analyse von Siedlungsverteilungen in der Susiana
und im Kur-Flußbecken. Münster, Ugarit Verlag.
Grigson, C.
2003 Animal Husbandry in the Late Neolithic and Chalcolithic at Arjoune: The
Secondary Products Revolution Revisited. In P. J. Parr (ed.), Excavations at
Arjoune, Syria. British Archaeological Reports, International Series 1134. Oxford,
Hassan, A.
2002 Report on Phytolith Samples Taken from the Site of Fıstıklı Höyük, 2000.
Unpublished manuscript.
Hodder, I.
1991 The Domestication of Europe. Oxford, Blackwell.
Hole, F.
1978 Pastoral Nomadism in Western Iran. In R. A. Gould (ed.), Explorations
in Ethnoarchaeology. Albuquerque, University of New Mexico Press: pp.
127–167.
1980 The Prehistory of Herding: Some Suggestions from Ethnography. In M.-T.
Barrelet (ed.), L’Archéologie de l’Iraq: Du début de l’époque néolithique à 333 avant
notre ère: Perspectives et limites de l’interprétation anthropologique des documents.
Paris, Centre National de la Recherche Scientifique: pp. 119–130.
Hole, F. and G. A. Johnson
1986–1987 Umm Qseir on the Khabur: Preliminary Report on the 1986 Excavation.
Annales Archéologiques Arabes Syriennes 36/37: pp. 172–220.
Horkheimer, M. and T. Adorno (translated by John Cummings)
1989 Dialectic of Enlightenment. New York, Continuum.
Husserl, E.
1976 Ideen zu einer reinen Phänomenologie und phänomenologischen Philosophie. Den
Haag, Martinus Nijhoff (first published in 1913).
Hütteroth, W. D.
1959 Bergnomaden und Yaylabauern im mittleren kurdischen Taurus. Marburger
Geographische Schriften 11. Marburg, Selbstverlag des Geographischen
Institutes der Universität Marburg.
Isbell, W. H.
2000 What We Should Be Studying: The “Imagined Community” and the
“Natural Community.” In M. A. Canuto and J. Yaeger (eds.), The Archaeology
of Communities: A New World Perspective. London, Routledge: pp. 243–266.
James, W.
1979 Kwanim Pa: The Making of the Uduk People. Oxford, Clarendon.
Johnson, G. A.
1973 Local Exchange and Early State Development in Southwestern Iran. Anthropological
Papers 51. Ann Arbor, Michigan Museum of Anthropology.
1982 Organizational Structure and Scalar Stress. In C. Renfrew, M. Rowlands
and B. A. Segraves (eds.), Theory and Explanation in Archaeology. New York,
Academic Press: pp. 389–421.
1987 The Changing Organization of the Uruk Administration on the Susiana Plain.
In F. Hole (ed.), The Archaeology of Western Iran: Settlement and Society from
Prehistory to the Islamic Conquest. Washington, Smithsonian Institution Press:
pp. 107–140.
Joyce, R. A.
2004 Unintended Consequences? Monumentality as a Novel Experience in
Formative Mesoamerica. Journal of Archaeological Method and Theory 11: 1, pp
5–29.
Joyce, R. A. and J. A. Hendon
2000 Heterarchy, History, and Material Reality: “Communities” in Late Classic
Honduras. In M. A. Canuto and J. Yaeger (eds.), The Archaeology of Communities:
A New World Perspective. London, Routledge, pp. 143–160.
Karul, N., A. Ayhan and M. Özdogan
2001 1999 Excavations at Mezraa, Teleilat. In N. Tuna and J. Öztürk (eds.), Ilısu ve
Karkamıs Baraj Gölleri Altında Kalacak Arkeolojik Kültür Varlıklarını Kurtarma
Projesi 1999 Yılı Çalısmaları. Ankara, ODTÜ Tarihsel Çevre Arastırma
Merkezi: pp. 162–181.
Khazanov, A. M.
1994 Nomads and the Outside World. Madison, University of Wisconsin Press (second
edition).
Kirkbride, D.
1982 Umm Dabaghiyah. In J. Curtis (ed.), Fifty Years of Mesopotamian Discovery.
Hertford, Stephen Austin and Sons: pp. 11–21.
Kirsch, E. and P. Larsen
1996 Das Verhältnis zwischen Seßhaften und Nichtseßhaften in Mesopotamien am
Ende des 3. und zu Beginn des 2. Jt. v. Chr. In K. Bartl, R. Bernbeck, and M.
Heinz (eds.), Zwischen Euphrat und Indus: Aktuelle Forschungsprobleme in der
Vorderasiatischen Archäologie. Hildesheim, Georg Olms: pp. 148–164.
Klengel, H.
1958–1959 Benjaminiten und Hanäer. Wissenschaftliche Zeitschrift der Humboldt-
Universität zu Berlin: Gesellschafts- und Sprachwissenschaftliche Reihe 8: 2–3, pp.
211–226.
1972 Zwischen Zelt und Palast. Leipzig, Koehler und Amelang.
Kopytoff, I.
1986 The Cultural Biography of Things: Commoditization as Process. In A.
Appadurai (ed.), The Social Life of Things: Commodities in Cultural Perspective.
Cambridge, Cambridge University Press: pp. 64–94.
Kovacs, M.
1989 The Epic of Gilgamesh. Stanford, Stanford University Press.
Kupper, J.-R.
1957 Les nomades en Mésopotamie au temps des rois de Mari. Paris, Société d’Édition
Les Belles Lettres.
LaMotta, V. M. and M. B. Schiffer
1999 Formation Processes of House Floor Assemblages. In P. M. Allison (ed.), The
Archaeology of Household Activities. London, Routledge: pp. 19–29.
Lavie, S. and T. Swedenburg (eds.)
1996 Displacement, Diaspora, and Geographies of Identity. Durham, Duke University
Press.
LeBlanc, S. and P. J. Watson
1973 A Comparative Statistical Analysis of Painted Pottery from Seven Halafian
Sites. Paléorient 1: pp. 117–133.
Lieberman, D. E.
1998 Natufian “Sedentism” and the Importance of Biological Data for Estimating
Reduced Mobility. In T. Rocek and O. Bar-Yosef (eds.), Seasonality and
Sedentism: Archaeological Perspectives from Old and New World Sites. Peabody
Museum Bulletin 6. Cambridge, Harvard University Press: pp. 75–92.
Lüning, J.
1988 Frühe Bauern in Mitteleuropa im 6. und 5. Jahrtausend vor Chr. Jahrbuch des
Römisch-Germanischen Zentralmuseums Mainz 34: pp. 27–93.
Luke, J. T.
1965 Pastoralism and Politics in the Mari Period. University of Michigan, Ann Arbor
(PhD dissertation).
Marfoe, L., P. J. Parr and C. S. Phillips
2003 The Site and Its Excavation. In P. J. Parr (ed.), Excavations at Arjoune, Syria.
British Archaeological Reports, International Series 1134. Oxford, Archaeopress:
pp. 11–22.
Mashkour, M.
2003 Tracing Ancient “Nomads”: Isotopic Research on the Origins of Vertical
“Transhumance” in the Zagros Region. Nomadic Peoples (new series) 7: 2, pp.
36–47.
Moffett, L.
2003 Wild and Cultivated Plants and the Evidence for Crop Processing Activities.
In P. J. Parr (ed.), Excavations at Arjoune, Syria. British Archaeological Reports,
International Series 1134. Oxford, Archaeopress: pp. 241–250.
Morier, J.
1837 Some Account of the I’lyats, or Wandering Tribes of Persia, Obtained in the
Years 1814 and 1815. Journal of the Royal Geographic Society 7: pp. 230–242.
Murdock, G. P. and D. R. White
1969 Standard Cross-Cultural Sample. Ethnology 8: pp. 329–369.
Musil, A.
1908 Arabia Petraea. Band III. Ethnologischer Reisebericht. Wien, A. Hölder.
Niebuhr, C.
1776–1780 Voyages en Arabie et en D’autres Pays Circonvoisins. Amsterdam, Chez
Baalde.
Nissen, H. J.
1980 The Mobility Between Settled and Non-Settled in Early Babylonia: Theory
and Evidence. In M.-T. Barrelet (ed.), L’Archéologie de l’Iraq: Du début de l’époque
néolithique à 333 avant notre ère: Perspectives et limites de l’interprétation anthro-
pologique des documents. Paris, Centre National de la Recherche Scientifique:
pp. 285–290.
Oppenheim, M. Freiherr von
1939 Die Beduinen, Band I. Leipzig, Harrassowitz.
Panja, S.
2003 Mobility Strategies and Site Structure: A Case Study of Inamgaon. Journal of
Anthropological Archaeology 22: pp. 105–125.
Pauketat, T. R.
2000 The Tragedy of the Commoners. In M.-A. Dobres and J. Robb (eds.), Agency
in Archaeology. London, Routledge: pp. 113–129.
Pauketat, T. R. and S. M. Alt
2003 Monds, Memory, and Contested Mississippian History. In R. M. van Dyke
and S. E. Alcock (eds.), Archaeologies of Memory. London, Routledge: pp.
151–179.
Pfälzner, P.
2001 Haus und Haushalt. Damaszener Forschungen 9. Mainz, Philipp von Zabern.
Pollock, S.
2003 Feasts, Funerals, and Fast Food in Early Mesopotamian States. In T. Bray
(ed.), The Archaeology and Politics of Food and Feasting in Early States. New York,
Kluwer: pp. 17–38.
forthcoming Decolonizing Archaeology: Political Economy and Archaeological Practice
in the Middle East. In R. Boytner and L. Swartz-Dodd (eds.), Filtering the
Past, Building the Future: Archaeology, Tradition, and Politics in the Middle East.
Tucson, University of Arizona Press.
Pollock, S., R. Bernbeck, S. Allen, A. G. Castro-Gessner, R. Costello, S. K. Costello,
M. Foree, S. Lepinski and S. Niehbuhr
2001 Excavations at Fıstıklı Höyük 1999. In N. Tuna, J. Öztürk and J. Velibeyoglu

(eds.), The Salvage Project of the Archaeological Heritage of the Ilısu and Carchemish
Dam Reservoirs: Activities 1999. Ankara, Middle East Technical University: pp.
1–64.
Pollock, S. and A. G. Castro-Gessner
2004 Engendering Communities: The Contexts of Production and Consumption in
Early Mesopotamian Villages. Paper presented in “Que(e)rying Archaeology:
The 15th Anniversary Gender Conference” during the 37th Annual Chacmool
Conference, University of Calgary, Alberta.
Powell, S.
1983 Mobility and Adaptation: The Anasazi of Black Mesa, Arizona. Carbondale,
Southern Illinois University Press.
Preucel, R. W. and L. Meskell
2004 Places. In L. Meskell and R. W. Preucel (eds.), A Companion to Social
Archaeology. Oxford, Blackwell, pp. 215–229.
Rafferty, J. E.
1985 The Archaeological Record on Sedentariness: Recognition, Development, and
Implications. Advances in Archaeological Method and Theory 8: pp. 113–156.
Rivière, P.
1995 Houses, Places, and People: Community and Continuity in Guiana. In J.
Carsten and S. Hugh-Jones (eds.), About the House: Lévi-Strauss and Beyond.
Cambridge, Cambridge University Press: pp. 189–205.
Rocek, T. R.
1998 Pithouses and Pueblos on Two Continents: Interpretations of Sedentism and
Mobility in the Southwestern United States. In T. R. Rocek and O. Bar-Yosef
(eds.), Seasonality and Sedentism: Archaeological Perspectives from Old and New
World Sites. Peabody Museum Bulletin 6. Cambridge, Harvard University Press:
pp. 199–216.
Sadr, K.
1991 The Development of Nomadism in Ancient Northeast Africa. Philadelphia,
University of Philadelphia Press.
Scholz, F.
1995 Nomadismus: Theorie und Wandel einer sozio-ökologischen Kulturweise. Stuttgart,
Franz Steiner.
Shennan, S.
1988 Quantifying Archaeology. Edinburgh, Academic Press.
Simms, S. R.
1988 The Archaeological Structure of a Bedouin Camp. Journal of Archaeological
Science 15: pp. 197–211.
Soudsky, B.
1969 Étude de la maison Néolithique. Slovenská Archaeologia 15: pp. 5–96.
Stahl, A. B.
2001 Making History in Banda: Anthropological Visions of Africa’s Past. New York,
Cambridge University Press.
Steele, C.
2005 Who Has Not Eaten Cherries with the Devil? Archaeology and Politically
Problematic Contexts. In S. Pollock and R. Bernbeck (eds.), Archaeologies of
the Middle East: Critical Perspectives. Oxford, Blackwell: pp. 45–65.
Stevanović, M.
1997 The Age of Clay: The Social Dynamics of House Destruction. Journal of
Swidler, A.
2001 What Anchors Cultural Practices. In T. R. Schatzki, K. Knorr Cetina and
E. von Savigny (eds.), The Practice Turn in Contemporary Theory. London,
Routledge: pp. 74–92.
Trigger, B. G.
1968 The Determinants of Settlement Patterns. In K.-C. Chang (ed.), Settlement
Archaeology. Palo Alto, National Press Books: pp. 53–78.
Tringham, R.
2000 The Continuous House: A View from the Deep Past. In R. A. Joyce and
S. Gillespie (eds.), Beyond Kinship: Social and Material Reproduction in House
Societies. Philadelphia, University of Pennsylvania Press: pp. 115–134.
Verhoeven, M.
1999 An Archaeological Ethnography of a Neolithic Community. Leiden, Nederlands
Historisch-Archaeologisch Instituut te Istanbul.
2000 Death, Fire, and Abandonment. Archaeological Dialogues 7: 1, pp. 46–83.
Walker, W. H.
2002 Stratigraphy and Practical Reason. American Anthropologist 104: 1, pp.
159–177.
Walker, W. H. and L. J. Lucero
2000 The Depositional History of Ritual and Power. In M.-A. Dobres and J. Robb
(eds.), Agency in Archaeology. London, Routledge: pp. 130–147.
Watkins, T.
1990 The Origins of House and Home? World Archaeology 21: pp. 336–347.
Watson, P. J.
1983 The Halafian Culture: A Review and Synthesis. In T. C. Young, P. E. L. Smith
and P. Mortensen (eds.), The Hilly Flanks. Studies in Ancient Oriental Civilization
36. Chicago, University of Chicago Press: pp. 231–250.
Winckler, H.
1905 Auszug aus der Vorderasiatischen Geschichte. Leipzig, Hinrichs.
Wobst, M.
1978 The Archaeo-Ethnology of Hunter-Gatherers, or the Tyranny of the
Ethnographic Record in Archaeology. American Antiquity 43: 303–309.
Wright, Henry T.
1981 The Southern Margins of Sumer: Archaeological Survey of the Area of Eridu
and Ur. In R. M. Adams (ed.), Heartland of Cities. Chicago, University of
Chicago Press: pp. 295–345.
Wylie, A.
1985 The Reaction Against Analogy. Advances in Archaeological Method and Theory
8: pp. 63–111.
Zagarell, A.
1982 The Prehistory of the Northeast Bahtiyari Mountains, Iran: The Rise of a Highland
Way of Life. Wiesbaden, Dr. Ludwig Reichert.
Zeder, M. A.
1994 After the Revolution: Post-Neolithic Subsistence in Northern Mesopotamia.
American Anthropologist 96: pp. 97–126.
Ch a p t e r 4
A rchaeology and the

Question of Mobile
Pastoralism in L ate
Prehistory
A bb a s A l i z a d e h 1
M obile pastoralism as a subsistence economy and a way of life is

assumed to have developed parallel to the domestication of sheep and
goats in southwest Asia (particularly in highland Iran),2 and possibly elsewhere.3
But the evidence for interactions between the highlands mobile pastoralist
and lowlands settled farming communities in preliterate times in Iran has
been elusive. There are now a few lines of evidence available that can shed
more light on this problem. The main purpose of this chapter is to present
three sets of archaeological data pertaining to the presence of ancient mobile
pastoralist societies in southwestern Iran. Since the interpretation of the data
is based squarely on inferences from modern-day analogies and social theory,
it is necessary to discuss the ethnographic and historical sources of my infer-
ences. The first part of this chapter therefore deals with some introductory
1.
I would like to express my gratitude to Steven Rosen, Reinhard Bernbeck, Barbara
Helwing and Nicholas Kouchoukos, who read a preliminary draft of this chapter and
made valuable suggestions and criticisms.
2.
I use the term mobile pastoralism instead of nomadism to designate the mobile
herders of the Zagros Mountains, primarily the Bakhtiyāari and Qashqāaii. Unlike
the nomads of the vast steppes or marginal zones (such as the Jazira, Sinai, Negev and
Sahara), the Qashqāaii and Bakhtiyāari occupy primary agricultural regions with high
population density; they are only highly mobile while migrating; they have high social
and economic interactions with the settled farmers; they spend at least a quarter of the
year in regions with high population density; they own villages; they routinely practice
farming in both winter and summer pastures; and they have developed comparatively
complex political organizations and hierarchy.
3.
See for examples, Adams 1974; Gilbert 1975; Oates and Oates 1976; Thomson
Marucheck 1976; Zagarell 1982; Geddes 1983; Levy 1983; Smith 1983; Cribb 1991;
Bernbeck 1992; Köhler-Rollefson 1992 and Akkermans and Duistermaat 1996.
78
Mobile Pastoralism in L at e P r e h i s t o r y      79
remarks on the contemporary mobile pastoralist tribes in southwestern and

southern Iran, as well as some theoretical issues pertinent to the present topic.
The second part is devoted to the discussion of the archaeological evidence,
concluding with remarks derived from the combined ethnographic, historical
and archaeological evidence.
Recent studies suggest that ancient mobile pastoralists may have had
some measure of influence in the development of complex societies in the
ancient Near East, particularly in southwestern Iran (Zagarell 1982; Wright
1987:141-155, 2001). Nevertheless, in the study of the formation of state orga-
nizations in the Near East, the role of mobile pastoralist communities is either
completely ignored or viewed mostly as a contributing factor. Furthermore,
ancient mobile pastoralist communities in southwest Asia are not considered
developing societies that could have reached a level of sociopolitical integra-
tion, which could then lead to the control of their sedentary farming neighbors.
This is understandable, given the difficulties involved in finding pertinent
archaeological clues to the presence of ancient mobile pastoralists.
Elsewhere (Alizadeh 1988a, 1988b, 2003b) I have argued that the small
settlement at Tall-e Bakun A, in highland Fars, south-central Iran (Figures
4.1–4.3), demonstrates a number of features associated with the level of social
complexity attested at some later protohistorical urban centers, but the small
size of the site and the regional settlement patterns during the Bakun A phase
(4400–4100 BCE) as a whole does not conform to the Central Place theory or
tributary economics models, where higher levels of settlements are expected
to exhibit more population and more functions. I have suggested that this
‘anomaly’ may be described, if not explained, if we can demonstrate the under-
lying mobile pastoralist structure of the region.
Tall-e Bakun A is not unique. A number of other sites exhibit most of the
characteristics of the larger regional centers but are nevertheless too small to
have included a large population as a factor. Prominent among these special
sites are Tepe Gawra (Tobler 1950),4 Tell Abada (Jasim 1985), Kheit Qasim
(Forest-Foucault 1980; Margueron 1987) and possibly Tell Madhhur (Roaf
1982, 1987). These special sites constitute a category of settlements that does
not fit our current models of early urban development in which large, circum-
scribed farming populations play a fundamental role in creating socioeconomic
and political complexity. In all descriptive and explanatory models the number
of sites determines the size of a regional population, and the population of each
site is determined by its size. Such estimates obviously account for the settled
farming and urban population of a given region. These models, however, do not
4.
Rothman (1988:461, 599–625) considers Gawra an independent specialized site
with perhaps a nomadic clientele population.
80     A b b a s A l i z a d e h
Figure 4.1. Map of Iran and Mesopotamia.
account for ancient Near Eastern mobile pastoralist communities, though these
communities seem to have coexisted for thousands of years with the settled
communities as part of the socioeconomic continuum of local polities.
Apart from the problem of archaeologically identifying social groups with
residential mobility, it is difficult to attribute to mobile pastoralist communi-
ties the degree of social complexity that evolved in farming societies. This is
primarily because the scope of internal structural and economic variations is
limited in mobile pastoralist societies, where a level of state organization is not
expected to develop internally. Nevertheless, in regions with a high degree of
interaction between mobile pastoralist and settled farming communities, what
Rowton (1981) calls “enclosed nomadism,” chiefly aspirations could be realized
if that control is extended to include sedentary farming communities. In fact,
Rowton (1981:26–27) has shown that in enclosed nomadism it was common
Figure 4.2. Map of southwestern and south-central Iran, showing the Qashqāaii and Bakhtiyāari
territories and the distribution of the Bakun A–style ceramics (dotted area).
for the nomadic tribes to include fully sedentary tribes of a regional population.
The same is remarkably true of the contemporary mobile pastoralist tribes of
the Zagros Mountains.5 Because of the undiversified pastoral economy and its
limitations in accumulating wealth, the desire of mobile pastoralists, particu-
larly the tribal elite, to acquire land-based wealth and power is an important
variable in the dynamic relationship between the settled and mobile pastoralist
communities in the Middle East. The ethnographic literature abounds with
references to acquisition of land by tribal leaders.
To appreciate the comparatively high levels of social and economic interac-
tions between the settled farmers and the mobile pastoralists of southwestern
and south-central Iran, it is important to make the fundamental distinction
between the latter and the nomads of ecologically marginal zones such as
5.
See for examples Barth 1961; Garthwaite 1983 and Beck 1986.
Figure 4.3. Satellite image of northern Fars, showing the intermountain valleys and major sites.
central Asia, the Sahara, the Sinai, the Negev, and the Jazira in northwestern
Mesopotamia. Marginal areas unsuitable for grain agriculture but with excellent
pasture grounds, and thus with low population density, are rare in southwestern
and southern-central Iran. The only winter pastures frequented by the Zagros
mobile pastoralist tribes are the fertile intermountain valleys in Fars and the
southwestern lowlands of Susiana (modern-day Khuzestan), both with a high
density of settled farmers and vast tracks of cultivated lands. This environment
must have created a high level of interaction between the Zagros mobile pastoral-
ists and settled farmers not seen in ecologically marginal areas. Thus, regardless
of when vertical mobile pastoralism developed in Iran as a specialized way of life,
ancient tribesmen must have had comparatively high degrees of interaction with
the settled farmers of their winter haunts from the beginning.
An archaeological survey was conducted in northwestern Fars to test the
validity of the hypothesis that the spatial distribution of the Fars 5th millennium
BCE pottery corresponds to the migration routes of some of the modern-day
mobile pastoralist tribes of the Qashqāaii (Figure 4.2). In the 1995 survey
in the Qashqāaii territory, many permanent and semipermanent Qashqāaii
villages were encountered with strong ties to their pastoralist tribesmen.6
In such a bipolar socioeconomic and political context the entire settled,
semisettled and mobile populations of a tribal territory will have to be taken
into consideration.
Moreover, unlike highly mobile nomadic groups of vast steppes (in central
Asia) and arid zones (in the Negev and the Sinai), vertical mobile pastoralist
tribes of the Zagros Mountains possessed villages with solid architecture
in both their summer and winter pastures, in close proximity to the settled
farmers and urban centers (Figures 4.4-4.5; Alizadeh 1988a; 2003b; 2004). If
this pattern occurred in late prehistory, as argued here, then semipermanent
villages cannot be distinguished from permanent farming villages in surface
surveys. The assumption that in late prehistoric times such villages existed in
the midst of rich agricultural regions is based on ethnographic and historical
data, as well as some archaeological clues that are discussed below. This
assumption is theoretically significant because it addresses the problem of
the economy of scale, which presumably discourages nonpastoral production,
especially pottery, among mobile groups.7
6.
These interactions included ‘endotribal’ marriages between settled and mobile
Qashqāaii, settling of noncriminal disputes through the local Qashqāaii chiefs, and
economic interaction primarily involving hiring of Qashqāaii shepherds to tend flocks
of sheep and goats.
7.
See Eerkens 2003 for a full treatment of this problem and the question of resi-
dential mobility.
Figure 4.4. Semipermanent Qashqāaii village (foreground) and permanent local village (background)
near Firuzabad, central Fars (looking west).
Figure 4.5. Woman and children in a mobile pastoralist semipermanent Bakhtiyāari village near
Behbahan, showing combined tent and solid architecture common in the region.
Once we assume the existence of semipermanent pastoral villages in antiq-

uity, then the difficult problems of attributing industrial activities to the ancient
highland mobile tribes and the spread of certain regional styles of pottery
decoration (Figure 4.6) are not as daunting. The spread of the Bakun style of
pottery from Fars into the Zagros Mountains, lowland Susiana, and even the
Central Plateau may be described as a combined outcome of both segments of
the pastoralist society: the client villages where material goods could be manu-
factured and the mobile population that could carry, use or exchange them. The
spread of a specific class of pottery could also have been augmented through
marriage alliances, where decorated vessels may have been part of the dowry.8
Since interregional marriage alliances occur among the ruling elite of societies,
then the decorated pottery vessels could serve as symbolically significant.
In ecologically marginal regions the degree of political and economic inter-
action between mobile and sedentary communities is comparatively low, and
raids on farming villages and towns are far more frequent as a way of coping
with the highly risky subsistence economy practiced in these regions. Salzman
(1994) defines this situation in Iranian Baluchistan as “ecology of raids,” where
agriculture and animal husbandry are highly risky and demands for grains
often outstrip the local supply. Similar ecology and circumstances fostered
regular raids on the pueblos of the Tewa by the Comanches in the American
Southwest (Ford 1972).
This is in sharp contrast to the territories of the Qashqāaii and Bakhtiyāari,
whose winter pastures are in the most fertile and populated intermoun-
tain valleys of Fars and in lowland Khuzestan. In these regions both tribal
confederations owned not only villages and vast tracks of farmland, but they
also provided protection for the populations in their territories against the
marauding minor tribes of the Mamasani and Boyr Ahmadi, who occupied
the less fertile regions that straddled the Bakhtiyāari and Qashqāaii territories
(Figure 4.2). The Qashqāaii and the Bakhtiyāari also had developed complex
political organizations that simply did not exist among the nomads of Sistan,
Baluchistan, Kerman, and northeastern Iran, even though they all operated
within a state society.9
The interdependence of settled farming and mobile pastoralist communities
created a market from which both societies did benefit. This interaction in turn
created a context within which political and economic hegemony was exercised.
8.
If women were active potters or pot painters in prehistory, and there is no reason
not to consider this alternative, interregional marriages in patrilocal societies certainly
would lead to the spread of specific pottery styles that in the course of time would
become either diluted or would undergo hybridization.
9.
See for examples Bradburd 1994; Irons 1994 and Salzman 1994
Figure 4.6. Samples of ‘dot-motif’ pottery vessels of the Late Susiana 1 phase.
It is this interdependence and close proximity of the two societies in highland

Iran that underlie much of the sociopolitical and economic development in the
Near East in general and in Iran in particular.
Using these insights together with the evidence of surveys and excavations,
I propose that a number of mounded sites in late prehistoric Fars and lowland
Susiana were established as a result of crystallization of mobile pastoralist socio-
economic development (Alizadeh 1988a, 1988b, 2003b, 2004). The Near East
has supported rich and complex societies of mobile pastoralists throughout its
history.10 Near Eastern mobile pastoralist communities in general, and those
of the Zagros Mountains in particular, have had a high degree of economic
and social interaction with the settled farming villages and urban centers. This
interaction has been attributed to ecological and geographic factors that force
mutually dependent, territorially bound and autonomous entities to share
regions that provide the matrix for a web of social, economic and political
interactions. Rowton (1973a, 1973b, 1974, 1981) suggests that economic
interdependence and territorial coexistence of mobile pastoralists and seden-
tary agriculturists are major factors in the development of this high degree of
integration. Moreover, it is argued that uniting both elements (agriculture and
animal husbandry) within the same tribal structure would be advantageous to
all concerned (Rowton 1973b:201–215; Adams 1978:329–335). The mobile
pastoralist way of life can therefore be seen as an environmental, economic
and sometimes political adaptation (Irons 1974).
Until the 1950s the two mobile pastoralist major tribal confederations in
southwestern and south-central Iran were the Qashqāaii and the Bakhtiyāari
(Beck 1986; Garthwaite 1983). Their socioeconomic and political structures
are the best examples of what Salzman (1972) calls “multi-resource nomadism,”
allowing for a high degree of economic and social complexity and diversity.
These characteristics, as noted above, arose from the environmental and
geological features of the Zagros Mountains, which impose specific migration
routes and a choice of winter and summer pastures.
At this point it is important to outline some aspects of mobile pastoralism
in highland Iran, particularly in Fars, that must have contributed to its social
and political complexity in highland Iran. The following is a discussion of the
four basic features of vertical mobile pastoralism that are particularly important
in its socioeconomic development: the complexities of seasonal migration, the
agricultural activities, the externalization of tribal pastoral economy, and the
political and military potentials.
See for examples Mellink 1964; Luke 1965; Roux 1966; Strabo 1969; Herodotus
10.
1972; Bosworth 1973; Lambton 1973; Bottero 1981; Castillo 1981; Digard 1981;
Edzard 1981; Malbran-Labat 1981; Postgate 1981; Rowton 1981; Khazanov 1984 and
Cribb 1991.
Complexities of Seasonal Migr ation

Seasonal migration is of great importance in vertical mobile pastoralism. On
average, a tribal family would strike and repitch a tent many times in the course
of migration. For example, most of the Qashqāaii tribes travel a round trip
of about 1000 km annually. Traveling this long distance requires a great deal
of information to conduct the annual migration as efficiently and peacefully
as possible (Cooper 1925; Barth 1965; Spooner 1972a; Garthwaite 1983:22).
In vertical mobile pastoralism, migration routes are of great importance.
Each of the major tribes in Fars follows its own tribal route, the ‘Il rāah.’
The tribes have a traditional schedule of occupation and departure from a
region. A number of factors determine the departure time of the tribes from
winter and summer pastures. These factors include the location of the clans
in their seasonal pastures; the availability of grass; the possession or lack of
farmland; the type and size of the agricultural fields in the tribal region; tribal
disputes; the weather and a host of other unpredictable factors such as death
and childbirth.11
The complexity involved in dealing with these factors requires careful plan-
ning, scheduling and cooperation to minimize violent conflicts that might arise
if two or more tribes tried simultaneously to follow the same Il rāah (tribal
route) or to occupy the same region while migrating.12 Scheduling and coopera-
tion require large amounts of information processing and decision making by
both the tribal elite and camp leaders (Amanollahi-Baharvand 1981:175–81).13
Barth (1959:9, 1965) argues that the tightly scheduled migration through
the bottlenecks required strong and effective coordinating authorities.14 The
allocation of pastures and scheduling of the movement are affairs decided
collectively by several tribes or confederacies. A lower level of decision making
that involves the location of a camp is equally important in maintaining the
11.
See Varjavand 1967:28 for a detailed discussion of these factors; and Barth
1961:5–7 and Tapper 1979a:95–114, 1979b:84–118 for the relation between grazing
rights and social organizations among the Shāahsevan of northwestern Iran.
12.
See Garthwaite 1983:22 for the same situation and arrangements among the tribes
of the Bakhtiyāari confederation.
13.
See Cribb 1991:13; Johnson 1978, 1987; Wright 1977a:338, 1977b for a detailed
discussion of the processes involved in and the importance of decision making in
antiquity.
14.
See Lefébure 1979:115–126 for an examination of the idea that the structure of
the authority among the Zagros pastoralists is a direct consequence of environmental
constraints. Compare Burnham 1979 and Irons and Dyson-Hudson 1979 for the
importance of mobility in the formation of political centralization among mobile
pastoralists.
structure of smaller segments of the tribes. For more effective cooperation

tribes are divided into smaller segments that usually include wealthy families
with large herds and poor families with few or no animals, the poor providing
labor for the wealthy (Barth 1959:74 and my own observations). Among the
Boyr Ahmadi tribes of southwestern Fars it is the ‘rish sefid’ (literally, white
beard) of the ‘māal’ (a small segment of a tribe) who coordinates the date
of departure with the heads of the families. Thereby the tribesmen reach a
consensus that must be consistent with the general guidelines of the confed-
eration that has already been set forth (Husseini-Kazerooni 1973; Garthwaite
1983:44). This lower level of decision making among the mobile pastoralists of
the region is deemed an important factor in the social cohesion of these tribes
(Barth 1961:25–26, 1965; see also Johnson 1983).
Agricultur al Activities
Wide-range anthropological and historical studies have shown that there has
never been a totally pastoral society, for grain crops have always been an impor-
tant part of the mobile pastoralist diet (Spooner 1972b:245–68; Levy 1983:17;
Teitelbaum 1984). Though farmers supply the bulk of the grains needed by the
mobile pastoralists, the practice of agriculture is also widespread among the
latter in highland Iran. Members of many mobile pastoralist tribes invariably
rely on dry farming and take advantage of arable lands in both summer and
winter pastures.
In the high altitudes of summer pastures, just before leaving the area, some
tribesmen sow crops that are covered by winter snow, sprout in the spring,
and are ready to be harvested by the time the tribe returns. Similarly, tribal
families plant small plots of barley and wheat in December, harvesting them
in April just before they depart for their summer pastures in the mountains.15
At times that the winter crop is not ready, local workers are hired to harvest it
for the tribe. In the Bakhtiyāari Mountains, the tribe of Bamadi leaves for the
mountains in March or April, a month before the crop is ready. The tribesmen
leave some members behind to harvest the crop and to hide it under rocks in
makeshift storage, or they hire some sedentary local farmers to harvest it for
them while they are gone (Varjavand 1967:19).16 Stack (1882:68, 100) reports of
the same practice among the Qashqāaii: “They leave some men behind to reap
their scattered fields which they have ploughed and sown in their Firouzāabāad
qeshlāaq or winter haunts. The grain is buried in pits against the return of the
15.
See, for example, Garrod 1946:33 or Amanollahi-Baharvand 1981:47–48,
86–89.
16.
See Teitelbaum 1984:51–65 for similar practice among the Sudanese mobile
pastoralists.
tribe next year.” Garthwaite (1983:21, 40) also notes the importance of agri-
culture among the Bakhtiyāari and that when the tribe moves to its summer
and winter pastures some men stay behind to harvest and collect the crop.17
In addition, some Bakhtiyāari chiefs showed great interest in even large-scale
agriculture by investing, building and maintaining irrigation systems in western
Iran (Garthwaite 1983:30). This strategically important practice reduces the
risk of total dependence on the farming communities and ensures some security
if the crop fails in other areas.
Exter nal Investment of Tribal

Pastor al Economy
Among the factors that encourage mobile pastoralists to invest in agricultural
land is their awareness of the importance of agriculture as insurance against
the danger of losing the entire flock to epidemics and prolonged spells of dry
weather (Barth 1961:101–104, 1965; Garthwaite 1983:21, 40). This reinvest-
ment does not mean that mobile pastoralists see any advantage in sedentary
life; rather, it is practiced as a measure of security in the event that their live-
stock breeding should fail (Marx 1980:111). Barth (1961:104–106) notes that
sometimes individuals gradually acquire sufficient parcels of land and that once
their economy is determined by such possession, sedentarization seems to be
the natural result.18 While the interest of rank and file mobile pastoralists in
acquiring farmland may be economic and a response to risk, that of the higher
ranking individuals, particularly the chiefs, in acquiring agricultural land can
also be seen as politically motivated, for mobile pastoral economy has a limited
capacity for furthering political ambitions of tribal chiefs.
The processes of sedentarization do not necessarily lead to sedentism, the
outcome of sedentarization. Moreover, sedentism is by no means irreversible
and absolute.19 This is particularly true in times of economic and political uncer-
tainty, when mobile pastoralists keep their options open for shifting from one
way of life to another (Adams 1978; Marx 1980:111). In fact, the processes of
sedentarization, as argued by Barth, do not constitute a threat to the existence of
mobile pastoralism; these processes rather augment pastoralism by maintaining
environmental equilibrium through various mechanisms (Barth 1961:124).
17.
See also the lively description of Freya Stark 1934, who reports the same practice
in parts of Lurestan.
18.
See also Ehmann 1975:113–15, where he reports the same tendency among the
Bakhtiyāri tribes.
19.
For a different view on the processes of sedentarization see Salzman 1980 and
Galaty et al. 1981.
Though part-time farming relieves the mobile tribes from total dependence
on the agriculturists, it does not satisfy their grain requirement, which is
procured either through barter or purchase in market towns. Nevertheless, the
mobile pastoralists’ practice of agriculture and their knowledge of farming have
a strategic significance in that they allow for a greater flexibility in adapting
to various environmental and political calamities (Spooner 1972b:245–268).
The superior knowledge of mobile pastoralists of environmental resources and
geographic features of their vast territories are of strategic importance. Mobile
pastoralists are much more familiar with climatic changes, types of soil and
the location of water sources and other natural resources, so they can easily
shift to a settled way of life. The reverse transition is far more difficult for the
sedentary farmers, particularly if they are not related to the mobile tribes of
their area. In a favorable environment, with many natural resources and ideal
pastures, such as the Zagros Mountains and their piedmonts, the shift from
mobile pastoralism to sedentary farming and vice versa seems to have been the
major adaptive response to environmental and political pressure.20 The most
recent example is the return of part of the Qashqāaii tribes to mobile life after
the Iranian revolution in 1979.
Political and Military Potentials

Before they were forced to settle by Reza Shah in the first half of the 20th
century CE, the number of mobile pastoralists in Iran fluctuated between one
and two million (Barth 1961; Amanollahi-Baharvand 1981; Garthwaite 1983;
Beck 1986; Safinezhad 1989). But these numbers, though large, do not indi-
cate the importance of the role of the highland mobile pastoralists in shaping
Iranian history. They occupy an important place in the society because they
constitute well-organized economic, social, and political units (Sunderland
1968; Ehmann 1975; Briant 1982) that either within a state or in the absence
of centralized state organizations could also pose a military threat to farming
and urban communities.
As moving targets in rugged terrains, mobile pastoralists are difficult to
overwhelm militarily. This mobility, combined with a tribal organization,
enabled highland pastoralists to rule supreme over the settled communities in
the absence of a strong centralized state (in the history of Iran the rule rather
than the exception). In fact, it took the Pahlavi regime several decades of
military operations, aided with fighter jets, gunship helicopters and artillery,
to politically subdue the Qashqāaii and the Bakhtiyāari. One can envisage that
See Adams 1974 for the role of mobile pastoralism in environmental and political
20.
adaptation; see also Adams 1978.

in prehistoric times bands of mobile pastoralists would have been militarily

superior to agriculturists.
It can be argued that the military superiority of mobile pastoralists depended
to a large degree on horses and camels and was thus a late development. This
is certainly true for the vast steppes of central Asia and regions such as the
Sahara. Moreover, in regions with comparatively high population density,
such as lowland Susiana and Fars, the sheer superior numbers of settled
farmers certainly would be a deterrent to any nomadic intrusion and raids
on foot. Nevertheless, in southwestern and south-central Iran, hiding places
were readily available to the tribesmen in the nearby Zagros Mountains but
were comparatively inaccessible and hazardous to the settled farmers. As the
events of the Reza Shah era demonstrate, even a well-organized army with
modern technology cannot easily overwhelm the mountain tribes. The mili-
tary advantage of the vertical mobile pastoralist tribes lies in the geographic
and geological features of their surroundings and in the tribes’ mobility, fluid
subsistence economy and general lack of fixed assets.21
It is equally true that without horses and camels, it is not easy to imagine
how mobile pastoralist tribes could exert their political hegemony on settled
farmers. In the absence of state organizations, or in situations where organized
military response cannot be immediate, fleet-footed mobile tribesmen can
bring a settled regional population to submission by sheer harassment. It is
easy to imagine the vulnerability of farmers during harvest time, when a small
band of mobile pastoralists could easily set fire to the harvest and disappear
without a trace into the mountains. Similarly, flocks of sheep and goats sent
by the farmers to nearby hills could be stolen by the mobile tribesmen, and
stories of such events (real or imagined) abound in the major tribal regions in
Iran. This inherent military superiority of vertical mobile pastoralists should be
considered another factor in their sociopolitical development. As Sáenz (1991)
argues in the case of the Tuaregs of northern Africa, the military advantage of
vertical mobile pastoralist communities alone can lead to extortion, which in
turn may lead to warrior-client interaction and subsequently to stratification
and increased social complexity.
Following Earle (1994), one may postulate that because mobile pastoralist
groups operate regionally over vast areas on a regular basis, the hierarchy
that arises from within can generate overarching levels of social and political
organization not present in any one segment of the society. Such levels of
organization would then result in the integration of economically and politi-
cally segmented groups. The potential military power of the highland mobile
pastoralists can, however, be a double-edged sword. As Earle (1994:956) argues,
military power can be an equalizing force, which not only coerces submission
21.
See, for examples, Irons 1974; 1994.
but also creates resistance to domination. One may assume that the military
aspects of the mobile pastoralist tribes in southwestern Iran in turn became an
important variable in the adaptive reorganization of lowland societies of the
late 5th and early 4th millennia BCE. In this scenario the military capability of
mobile pastoralists could be an important factor in the development of state
organization in the region.
The Zagros Mountains

The above introduction was necessary to provide a context in which mute
archaeological evidence can be interpreted. Apart from the archaeological
evidence from Fars, three lines of evidence provide additional clues to the
presence and activities of ancient mobile pastoralist communities in Iran.
The earliest clues to the presence of socio-economic differentiation among
mobile pastoralists in prehistoric Iran is found in the isolated cemeteries of
Hakalāan and Parchineh in Lurestan, the oldest nomadic cemeteries in Iran
and, in fact, in the entire Near East. Louis Vanden Berghe (1973a, 1973b,
1975, 1987) excavated both cemeteries from 1971 to 1973.22 The cemeteries
are located north of Khuzestan, along the Meimeh River in the Pusht-e Kuh
region of Lurestan, in the southwestern piedmont of the Zagros Mountains
(Figure 4.1). The sites are considered nomadic cemeteries because they are
not associated with any known settlements because they are similar in location
and tomb construction to the later nomadic Bronze and Iron Age tombs in
the same region and also because the region is unsuitable for grain agriculture
and almost devoid of permanent ancient, as well as modern-day, villages with
agriculture as a subsistence base (Alizadeh 2003a, 2004).
In both cemeteries pottery vessels (about 200) were the most abundant
funerary objects. Based on general comparisons to the ceramics of the Early
Middle Chalcolithic in central Zagros (Henrickson 1985), Haerinck and
Overlaet (1996:27) date the cemeteries in Area A at Parchineh to about
4600–4200 BCE. Pottery vessels from both cemeteries show strong affinities
with the pottery of the Ubaid 3 and 4 phases in Mesopotamia and Late Middle
and Late Susiana 1 phases (around 5000–4400 BCE) in lowland Susiana. The
most interesting characteristic of the cemeteries’ artifacts, particularly the
pottery vessels, is the specific regional styles that they exhibit, representing
Mesopotamia, lowland Susiana and highland Iran.
The obvious continuum of the richness of the funerary gifts deposited in the
tombs suggests differential status among those who were buried in the ceme-
teries. At this level of social evolution, and with their assumed inherent military
22.
The final report was superbly published posthumously by E. Haerinck and B.
Overlaet (1996).
superiority over the settled farming communities, the archaeological evidence

suggests that the mobile pastoralist communities of the 5th millennium BCE were
in a position to affect the process of social evolution in southwestern Iran.
Lowland Susiana
Chogha Mish, in lowland Susiana (Figure 4.1), enjoyed a central status in the
entire Susiana plain from the Archaic Susiana (about 6900 BCE) through the
end of the Middle Susiana period (about 5000 BCE), when its monumental
building was destroyed (Alizadeh 1992; Kantor and Delougaz 1996). It is not
known whether hostile forces destroyed the monumental building or that the
fire was accidental. This event, however, coincided with others that, taken
together, suggest a changing organization in the settlement pattern during the
first half of the 5th millennium BCE. The destruction of the monumental ‘Burnt
Building’ coincided not only with the abandonment of Chogha Mish but also
with the desertion of a number of other sites in the eastern part of the plain,
with the appearance of the communal cemeteries of Hakalāan and Parchineh in
the highlands and also with the appearance of a specific class of painted pottery
with a number of pottery shapes and decorative motifs, including the ‘dot
motif’ (Figure 4.6). These motifs are considered specific to the following Late
Susiana 1 phase (the period when Chogha Mish remained unoccupied). The
characteristic vessels decorated with the dot motif are found in highland Fars,
the Central Plateau and the Zagros Mountains (Vanden Berghe 1975:Figure
5:6, Figure 6:7–8, 17, 20 [Late Ubaid style], Figure 5:13–15, 18, Figure 6:9, 13,
16, 18 [Late Susiana 1 style], Figure 6:11 [Central Plateau style], and Figure
5:2, 12 [Fars style]).
With Chogha Mish lying deserted during the Late Susiana 1 phase, it
appears that no single site attained a central status in terms of size and popu-
lation.23 The observed westward movement of the Susiana settlements around
4800 BCE and the appearance of the highland communal burials provide a
relevant context for the observation made by Hole (1987:42, Table 8) that
“...sites were often occupied for only short periods, then abandoned for a
time and reoccupied. About half of the sites changed status from occupied to
unoccupied or vice versa . . . implying that settlements were unstable and that
land was not particularly scarce and therefore not valuable.” This westward
movement continued until the region east of the Shur River became almost
Site KS-04, about 10 km southwest of Chogha Mish is considered by Kouchoukos

23.
(1998) as a large (3–8 ha) population center dating to this phase. Our 2004 excavations
indicated that during the Late Susiana 1 phase the site was not fewer than 4 ha with a
large mud-brick platform at its summit (Alizadeh et al. 2005).
completely deserted before the Late Uruk phase (about 3400 BCE). Even
during the preceding Middle Uruk phase only six sites are reported from this
area (Johnson 1973:141–147).
The presumed correlation between the increased activities in eastern
Susiana of the highland mobile pastoralists and the westward shift of Susiana
settlements at the end of the Middle Susiana period becomes more tenable
when we note that the eastern part of the Susiana plain traditionally has been,
and still is, the locus of the winter pasture for the mobile pastoralists of the
region. If this environmental niche was also used in antiquity, as one might
expect, then the westward shift of the settled community may also be taken as
an indication of an increase in the activities of the mobile pastoralist groups in
the area, and the conflict of interest between the settled and mobile popula-
tions of the region, a dichotomy that remains the leitmotif of Iranian history
throughout the ages.
Until recently, the above observations had not been contextualized in
the field. As part of a joint project by the Iranian Cultural Heritage and
Tourism Organization (ICHTO) and the Oriental Institute and Department
of Anthropology of the University of Chicago, with a grant from the National
Science Foundation (BSC-0120519), geomorphologic surveys in this part of the
region were conducted. To gain additional data on the important Late Susiana
1 phase, excavations were conducted at the site of Dar Khazineh (KS-1626)
(Alizadeh et al. 2004).
KS-1626 is located some 30 kilometers southeast of the provincial town of
Shushtar (Figure 4.2). In this part of the Susiana plain both prehistoric and
historical sites are buried under some 2 m of alluvial deposits, a feature that
Lees and Falcon (1952) had already noticed. Our excavations in a number of
trenches at KS-1626 showed that only certain areas of the site had archaeological
remains; the site is mostly formed by the heavy alluvial deposits cut by torrential
seasonal floods from the nearby mountains. Tony Wilkinson, Nick Kouchoukos
and Andrew Bauer, who conducted the geomorphologic survey, concluded that
the construction of a huge irrigation canal (now the Gargar River) during the
Parthian/Sasanian period was responsible for this phenomenon. The large-scale
Elamite building activities in second millennium BCE Susiana, which required
millions of fired bricks, may have greatly contributed to the deforestation of
the western foothills of the Masjed Suleiman Mountains; this construction and
denuding must have triggered the erosion processes that reached their zenith
during the Parthian and Sasanian periods. As a result, the archaeological sites
in this region are only visible in the wadis (exposed sections of the valleys).
Excavations in the main trench (Figure 4.7, square 379) revealed a peculiar
depositional pattern not recorded before in the region. Clayish and sandy sedi-
ments 5-10 cm thick superimposed thin lenses of cultural deposits. No solid
Figure 4.7. Map of KS-1626, showing the location of the excavated areas.
architecture was found except for extremely badly preserved pisé partition walls
whose faces were usually burnt. Postholes, traces of ash and fireplaces were
also encountered. The ‘floors’ on which these activities took place consisted of
alluvial deposits. Thus, in the main area of excavation, when we factored out
the alluvial levels from the cultural lenses, we were left with just over 30 cm of
deposit for perhaps the entire duration of the 5th millennium BCE. We did not
find any extensive organic horizon that would indicate the presence of animal
pens. This is hardly surprising as most of the site was destroyed, and our 4x4
m exposure may have been too small to reveal such a feature.
In the other excavation areas (Figures 4.7-4.8) we found three simple grave
pits dug into another layer of clayish alluvium. The skeletons were fragmentary
and very badly preserved; legs and hipbones were completely absent. Graves
1 and 2 yielded some rubbing stones and pounders. Grave 3, presumably
belonging to a female, yielded a saddle-shaped stone mill, stone pounders,
and a copper pin that, judging by its position, was used as a hairpin (Figure
4.9). No other archaeological deposit was found below the level of these burial
pits. In square 36 (Figure 4.7), again below the top alluvial deposit, we found a
fragmentary stone pavement embedded with potsherds of Late Susiana 1 date.
Again, we encountered no cultural deposit below this stone pavement.
Dar Khazineh thus seems to be analogous to the modern-day mobile pasto-
ralist campsites (Figures 4.10-4.11). While working at the site, we noticed that
the area was used by some mobile pastoralist tribes as a temporary campsite.
Specifically, we noticed that the tribes used the western bank of the stream in
the wadi as an overnight camp. This gave us an excellent opportunity to make
some ethno-archaeological observations. After a camp of several families left
Figure 4.8. General view of KS-1626, looking south.

Figure 4.9. Objects from Grave 3: (A) saddle shape mill with pounders; (B) copper pin; (C) bitumen
cosmetic vessel? (D) top view of C.
Figure 4.10. Bakhtiyāari spring camp at Chogha Mish (KS-01), looking east.
Figure 4.11. Bakhtiyāari tent in eastern Khuzestan, showing a horseshoe fire pit with low pisé surrounding
walls and various postholes.
at dawn, we found postholes and three shallow fire pits dug some 10 m apart
(Figure 4.12). Twigs and animal droppings were used for fuel. The lumps of
clay that had been dug out to make fireplaces were burnt and blackened by
the overnight fire, but not much else was left behind. This was very similar to
the patterns we excavated in square 379. We also knew that on rainy occasions
mobile pastoralists make stone beddings to protect their belongings against
moisture and rain, a feature similar to the stone bedding in square 36. In
addition, mobile pastoralists use the highest point of natural hills or artificial
mounds to bury their dead, a practice analogous to the graves we found in
square 208, the highest preserved part of the mound.
Subsistence Economy
Ms. Marjan Mashkour of the Centre National de la Recherche Scientifique in Paris
has analyzed the faunal samples. She believes that the faunal assemblage is
characteristic of what one expects from a temporary camp of mobile herders.
Even though we collected every piece of bone from every layer and feature,
only about 400 pieces of bones were found, which is not surprising given the
nature of the site. The main domesticated species identified were sheep and
100     A b b a s A l i z a d e h
Figure 4.12. Dar Khazineh (KS-1626): (A, C and D) remnants of temporary camp fire pits; (B)
Bakhtiyāari encampment.
goats; cattle bones (6–7%) were also present. Wild species included onager
(Equus hemionus onager), fallow deer (Dama dama mesopotamica), medium-size
rodents and some mollusk remains.
Ms. Naomi Miller of the Museum Applied Science Center for Archaeology,
University of Pennsylvania, analyzed the floral samples. Only a small volume
of seeds was recovered, despite the fact that we wet-sieved huge quantities of
dirt, especially from the fire pits. Again, the poor recovery of charred botanical
remains is consistent with the nature of the site, which was exposed to the
elements for much of the year. According to Miller the charred seeds include
two plant families, the grasses (Poaceae) and the legumes (Fabaceae). The only
cultivated plant in the grass family was barley (Hordeum).24
Based on these observations, we concluded that KS-1626 was occupied
seasonally by the prehistoric mobile pastoralists of the region, a pattern that
is still evident in eastern Khuzestan. The analyses of the flora and fauna
samples from the site also corroborate our characterization of the site as a
mobile pastoralist camp. It is important to bear in mind that the primary
Mr. Marco Madella of the University of Cambridge is in the process of analyzing

24.
the phytolith samples collected from the site.

occupation at Dar Khazineh coincides with the Late Susiana 1 phase, a period
we consider to represent the crystallization of a mobile pastoralist mode of
production in Iran.
The Centr al Plateau

Another line of evidence is now available from the Central Plateau, the
primary source of copper in Iran. Remarkable evidence for contact between
Fars, lowland Susiana and the highland Central Plateau comes from a series
of surveys conducted by Mir Abedin Kaboli (2000) of the Tehran ICHTO.
The survey region is located northeast of the city of Qum, some 100 km
south of Tehran (Figure 4.1). The unmistakable characteristic ceramics of the
Late Susiana 1 phase were found on at least six mounds.25 Other contempo-
rary prehistoric mounds in the region yielded only the typical late Cheshme
Ali pottery.
According to Kaboli (2000:133) mobile pastoralism is still practiced by some
families in the region. Sheep and goats are the primary stock, but camels are
also raised. In the hot summer months the pastoralists move to the mountains
near the provincial town of Saveh, northwest of Qum, or to the nearby Marreh
Mountains. While much research is needed to shed light on the dynamics
of the appearance in the Central Plateau of the typical 5th millennium BCE
ceramics of Fars and Susiana, the available evidence suggests that this devel-
opment may be related to the demand for copper in south and southwestern
Iran. More recent archaeological surveys in the region were conducted by
Barbara Helwing, of the German Institute in Tehran, and by Naser Chegini,
of the ICHTO, as part of a joint German-Iranian project. In their survey area,
between Kashan and Natanz, the region with copper mines, Helwing and
Chegini found examples of dot-motif pottery mixed with that of the contem-
porary late Cheshme Ali phase.26
The appearance in the Central Plateau of the typical 5th millennium BCE
southwestern and south-central pottery, decorated with the dot motif, is not
unilateral. Although no genuine pottery of the Sialk II type (Cheshme Ali
phase) has been reported from southwestern Iran, Sialk III type pottery has
been reported from surveys and excavations in the heart of the Bakhtiyāari
mobile pastoralist tribes of central Zagros regions of Khaneh Mirza (Zagarell
1975:146) and from Godin period VI and the mound of She Gabi in the
Kangavar area (Young 1969:Figure 7:1–17; Young and Levine 1974:Figure
25.
See for examples Kaboli 2000:Plates 19:1, 29:1–3, 33:15–16, 36:10, 37:1–5,
39:11.
26.
I am grateful to Dr. Helwing for generously providing this information.
102     A b b a s A l i z a d e h
14:1–20; Levine and Young 1987:Figures 10:50.2–5, 12:10, 17:1–12). This,

together with the evidence of the later third millennium BCE gray ware that is
found both in the Zagros region and in the central plateau, provides evidence
for the continuity of the interactions between southwestern and south-central
Iran with areas to the north and the east. These seem to have started in the
beginning of the 5th millennium BCE.
Discussion
Admittedly, more specific fieldwork is required to justify the basis of our
ethnographic and historical interpretation of the available archaeological
records. Nevertheless, we believe that a mobile pastoralist approach is a viable
alternative to the existing archaeological interpretations of the events and
processes that took place in the 5th and 4th millennia BCE in southwestern
and south-central Iran. Specifically, it is within this interpretive framework
that the development of small highland sites with the characteristics of large
urban centers can be described, if not explained. The unprecedented spread
of a specific decorated pottery in the 5th millennium BCE in southwestern and
south-central Iran and the Central Plateau, as well as the appearance of copper,
lapis lazuli and turquoise in south and southwestern Iran and in northern
Mesopotamia, are consistent with our reconstruction of the development and
crystallization of mobile pastoralism.
The 5th millennium BCE was a time of rapid population increase and of the
development of nascent urban centers in southwest Asia. It is therefore not
difficult or unwarranted to assume that mobile pastoralists’ mode of subsis-
tence economy and way of life put them in a position to become intermediaries
between lowland Susiana, Mesopotamia and highland Iran. Similarly, popu-
lation increase in the settled farming communities of the lowlands and vast
intermountain valleys of Fars would have created a context conducive to the
creation of nomadic surplus production, including meat, dairy by-products,
leather, wool and possibly kilims (carpets). Apart from these tangible prod-
ucts, mobile pastoralist groups could also interact with the settled farmers in
providing services such as labor, military protection, scouting and safe-guarding
commercial routes (Bates 1973; Black-Michaud 1986; Rosen 2003).
Whether items of exchange included pottery is a question that seems to
depend on the degree of residential mobility that imposes restriction on
pottery production (Close 1995; Rice 1999; Skibo and Blinman 1999). Arnold
(1985), while suggesting that less than a third of mobile societies make and use
pottery, argues that a number of practical, logistical and economic (economies
of scale) problems are involved in the production of pottery by groups with
high residential mobility. In a series of articles, however, Eerkens (2003) and
colleagues (Eerkens and Bettinger 2001; Eerkens et al. 2002) discuss a number
of strategies through which such obstacles were overcome by the highly mobile
tribes of Paiute and Shoshone of the southwestern region of the Great Basin
in North America. The pottery manufactured by these Native American tribes
is basically simple, crude and limited in shape and accessories (Eerkens et al.
2002:203–205). The same is true of the Negbite pottery of the Negev that
has been attributed to the nomadic groups of the region (Haiman and Goren
1992). These observations suggest that even when mobile groups do manufac-
ture pottery, their product is technologically and aesthetically inferior to those
produced by sedentary peoples (but see Barnard, this volume).
In the case of the vertical mobile pastoralists of the Zagros Mountains,
this distinction need not be made. First, despite their seasonal mobility, the
Zagros pastoralists spend only a fraction of the year moving from their summer
to winter pastures and vice versa. Whereas in their summer pastures they
occupy regions that are not suitable for grain agriculture and are thus sparsely
populated, in their winter pastures of Fars and lowland Khuzestan they spend
several months in heavily populated and agriculturally rich areas. Some tribes
even own villages with solid architecture or a mixture of tents and mud-brick or
stone houses. If this situation already existed in the 5th millennium BCE, attrib-
uting the manufacture, and thus the spread of the very specific class of the 5th
millennium BCE pottery in southwestern Iran to the mobile pastoralist groups
is theoretically and practically not far-fetched. Interregional marriages, an
important factor in forging interregional alliances through kinship, could also
be considered a contributing factor in the spread of some classes of pottery.
In summary, because of their highly specialized and undifferentiated
economy, mobile pastoralists would be more interested in trade (either
exchanging their own products or serving as intermediaries in long-distance
trade) than sedentary people. But self-sufficient farming villages are, by defi-
nition, not viable markets for the tribesmen. Mobile herders cannot trade
among themselves, however, because of their undiversified economic mode of
production. So we know historically, and expect prehistorically, an association
between the crystallization of the highland mobile pastoralist communities
and the rise of large population centers with diversified economy and a large
population not necessarily engaged in subsistence agriculture.
Once the necessary demographic, economic and political conditions were
present for a pastoral society to engage in the production and distribution of
surplus animal products and material goods, a fixed locus combining produc-
tion, administration and residential quarters would have to be chosen. Tall-e
Bakun A, and similar sites mentioned earlier, may have been the residences of
some of the wealthier and higher-ranking individuals whose economic strength
and social status allowed them to pursue sedentary trade economy. A common
104     A b b a s A l i z a d e h
ethnic background, and perhaps kinship ties between the settled and mobile
communities in Fars and the Zagros Mountains, may have facilitated processes
of economic and sociopolitical development and integration in Fars.
The introduction of the specific Late Susiana 1 pottery in the copper-rich
Central Plateau may be linked with exchange activities of southwestern mobile
pastoralist tribes in procuring copper, turquoise and lapis lazuli, which began
to appear in Fars, lowland Susiana, and Mesopotamia in the 5th millennium
BCE.27 The presence of the typical 5th millennium BCE southwestern pottery in
the Central Plateau can also be explained in terms of a reciprocal social system
involving pottery vessels and their contents as gifts to gain access in ‘foreign’
lands (Hodder 1980; Gregory 1982; Earle 1994). Much work in the region,
however, is required to shed more light on this inference.
The gradual rise of nascent urban centers with industrial and economic
specialization and the rise of a regional elite can be viewed as positive feedback
in mobile pastoralist communities. Specifically, the rising demand for wool
may be considered a contributing factor (Kouchoukos 1998). In an approach
that favors ratios of the number of identified specimens present (NISP) values
among taxa, Richard Redding (1981, 1993) has shown that, with the exception
of Hassunan and Halafian sites in northern Mesopotamia and Syria, prior to
5500 BCE sheep/goat ratios were more or less uniformly low (< 0.5). By 4500
BCE the ratios changed to 1.5–4.5, indicating a changing trend in herding
strategy from a subsistence economy to an economy where animal by-products
became important (Kouchoukos 1998:294–301).
By the late Middle Susiana phase (about 5000 BCE) sheep and goats became
dominant in Susiana, accounting for approximately 65% at Jafarāabāad, with
sheep becoming more dominant in the later phase (Kouchoukos 1998:68).
Similar developments occurred at the contemporary Chogha Mish. If we
consider this development an indication of the increasing importance of wool in
Susiana, as well as in southern Mesopotamia,28 then the concomitant appearance
of the large cemeteries of Hakalāan and Parchineh may not be coincidental.
We can envisage a situation where the initial development of highland mobile
pastoralism in late prehistory was perhaps related to the importance of wool
in the economies of both Susiana and southern Mesopotamia.
While small farming villages could provide the necessary grain for these
newly developed population centers, items of trade not found in the lowlands
were procured through the mobile tribes of southwestern Iran. If, however,
27.
The evidence from Mesopotamia is even earlier, except for that found at Gawra:
Yarim Tepe I, level 9 (Merpert and Munchaev 1987:15, 17); and Arpachyiah, Half levels
(Mallowan and Rose 1935:Plate ivb); and Gawra, level XIII (Tobler 1950:192).
28.
For a detailed study of the importance of wool in the southern Mesopotamian
economy during the Uruk period see Kouchoukos 1998.
as a result of population increase and specialization of crafts, more land was

brought under cultivation in central eastern Khuzestan, to feed that portion
of the population that was not engaged in producing food, one expects to
see a reduction in pasturelands in the same area. This situation could have
created a context where the mobile tribes may have taken measures to reclaim
the lands that they had lost to the farmers. While no direct evidence is avail-
able for intensification of agriculture and the subsequent loss of pasture in
5th millennium BCE Khuzestan, the pattern of competition for the available
land between the contemporary mobile pastoralist tribes and settled farmers
is familiar in Iran.
Growth in both mobile and settled populations can result in an increase
in the amount of farming and pasturelands. This in turn would create closer
proximity for the two populations and intensification of social interaction. In
the context of state organization, or faced with an outside encroaching threat,
mobile tribes may forge confederations that, albeit ephemeral, come close to a
state-level of political organization. In such a context, social complexity would
develop from the constant requirement of the pastoralists for communication
and cooperation to maintain economic and social cohesion. This cohesion was
usually characterized by a loosely structured centralized system culminating
in the single office of chief and welded together by the seemingly dispersed
tribes (Barth 1961:71–73).
The available evidence suggests that as craft specialization developed and
nascent urban centers became more populated, more organized and differen-
tiated socially and economically in the 5th millennium BCE, the demand for
grains, wool, dairy products, animal by-products, timber and exotic goods
(copper, turquoise, lapis lazuli, Persian Gulf shells) increased. In this context
the mobile pastoralist groups were in a strategic position to become the inter-
mediaries between the lowlands and highlands. I have tentatively attributed this
development to the regional conflicts and increasing socio-economic interac-
tion that resulted from the crystallization of mobile pastoralist communities
in the highlands.
References
Adams, R. M.
1974 Anthropological Perspectives on Ancient Trade. Current Anthropology 15: pp.
239–257.
1978 Strategies of Maximization, Stability, and Resilience in Mesopotamian Society,
Settlement, and Agriculture. Proceedings of the American Philosophical Society
122: 5, pp. 329–335.
Akkermans, P. M. and K. Duistermaat
106     A b b a s A l i z a d e h
1997 Of Storage and Nomads: The Sealings from Late Neolithic Sabi Abyad.
Alizadeh, A.
1988a Socio-economic Complexity in Southwestern Iran During the Fifth and
Fourth Millennia BC: The Evidence from Tall-e Bakun A. Iran 26: pp.
17–34.
1988b Mobile Pastoralism and the Development of Complex Societies in Highland Iran:
The Evidence from Tall-e Bakun A. University of Chicago, Department of Near
Eastern Languages and Civilizations (PhD dissertation).
1992 Prehistoric Settlement Patterns and Cultures in Susiana, Southwestern Iran.
University of Michigan Museum of Anthropology Technical Report 24. Ann Arbor,
University of Michigan.
2003a Report on the Joint Archaeological and Geomorphological Research
Expedition in Lowland Susiana, Iran. Oriental Institute News and Notes 177:
pp. 1–7.
2003b Some Observations Based on the Nomadic Character of Fars Prehistoric
Cultural Development. In N. Miller and K. Abdi (eds.), Yeki Bud, Yeki Nabud:
Essays on the Archaeology of Iran in Honor of William M. Sumner. Los Angeles,
Cotsen Institute of Archaeology: pp. 83–97.
Alizadeh, A., N. Kouchoukos, T. Wilkinson, A. Bauer and M. Mashkour
2004 Preliminary Report on the Joint Iranian-American Landscape and
Geoarchaeological Reconnaissance of the Susiana Plain, September–October
2002. Paléorient 29: pp. 69–88.
Alizadeh, A., A. Mahfroozi, L. Niakan, A. Ahrar, M. Karami, A. Zalaghi, K. Aqaii, S.
Ebrahimi and T. Hartnell
2005 Joint ICHTO-Oriental Institute Excavations at KS-04 and KS-108 in Lowland
Susiana, Southwestern Iran. In Oriental Institute Annual Report 2004/2005.
Amanollahi-Baharvand, S.
1981 Pastoral Nomadism in Iran [in Persian]. Tehran, Bongah Tarjomeh va Nashr
Ketab.
Arnold, D. E.
1985 Ceramic Theory and Cultural Process. Cambridge, University of Cambridge
Press.
Barth, F.
1959 The Land Use Pattern of Migratory Tribes of South Persia. Norsk Geografisk
Tidsskrift 17: pp. 1–11.
1961 Nomads of South Persia. London, Allen and Unwin.
1965 Herdsmen of Southwest Asia. In P. B. Hammond (ed.), Cultural and Social
Anthropology. New York, MacMillan: pp. 63–83.
Bates, D. G.
1973 Nomads and Farmers: A Study of the Yörük of Southeastern Turkey. Anthropological
Papers of the Museum of Anthropology 52. Ann Arbor, University of Michigan.
Beck, L.
1986 The Qashqaii. New Haven, Yale University Press.
Bernbeck, R.
1992 Migratory Patterns in Early Nomadism. Paléorient 18: pp. 77–88.
Black-Michaud, J.
1986 Sheep and Land. Cambridge, Cambridge University Press.
Bosworth, C. E.
1973 Barbarian Incursions: The Coming of the Turks into the Islamic World. In
D. S. Richards (ed.), Islamic Civilisation. Oxford, Cassirer: pp. 1–16.
Bottero, J.
1981 Les Habiru, les nomades, et les sédentaires. In J. S. Castillo (ed.), Nomads and
Sedentary Peoples. Mexico City, Colegio de Mexico: pp. 89–108.
Bradburd, D.
1994 Historical Bases of the Political Economy of Kermani Pastoralists: Tribe
and World Markets in the Nineteenth and Early Twentieth Centuries. In C.
Chang and H. A. Koster (eds.), Pastoralists at the Periphery. Tucson, University
of Arizona Press: pp. 42–61.
Briant, P.
1982 Etat et pasteurs au Moyen-Orient ancien. Paris, Éditions de la maison des sciences
de l’homme.
Burnham, P.
1979 Spatial Mobility and Political Centralization in Pastoral Societies. In L’Equipe
écologie et anthropologie des sociétés pastorales (ed.), Pastoral Production and
Society/Production pastorale et société, Actes du colloque international sur le pastoral-
isme nomade, Paris, 1–3 Déc. 1976. Cambridge, Cambridge University Press:
pp. 349–374.
Castillo, J. S.
1981 Tribus pastorales et industrie textile a Mari. In J. S. Castillo (ed.), Nomads and
Sedentary Peoples. Mexico City, Colegio de Mexico: pp.109–122.
Close, A. E.
1995 Few and Far Between: Early Ceramics in North Africa. In W. K. Barnett
and J. W. Hoopes (eds.), The Emergence of Pottery. Washington, Smithsonian
Institution Press: pp. 23–37.
Cooper, M. C.
1925 Grass. New York, G. P. Putnam’s Sons.
Cribb, R.
108     A b b a s A l i z a d e h
Digard, J. P.
1981 A propos des aspects économiques de la symbiose nomades-sédentaires dans
la Mésopotamie ancienne: Le point de vue d’un anthropologue sur le moyen
Orient contemporain. In J. S. Castillo (ed.), Nomads and Sedentary Peoples.
Mexico City, Colegio de Mexico: pp. 13–24.
Earle, T. K.
1994 Positioning Exchange in the Evolution of Human Society. In T. Baugh and
J. Ericson (eds.), Prehistoric Exchange Systems in North America. New York,
Plenum: pp. 419–437.
Edzard, D. O.
1981 Mesopotamian Nomads in the Third Millennium B.C. In J. S. Castillo (ed.),
Nomads and Sedentary Peoples. Mexico City, Colegio de Mexico: pp. 37–46.
Eerkens, J. W.
2003 Residential Mobility and Pottery Use in the Western Great Basin. Current
Anthropology 44: 5, pp. 728–738.
Eerkens, J. W. and R. L. Bettinger
2001 Techniques for Assessing Standardization in Artifact Assemblages: Can We
Scale Material Variability? American Antiquity 66: 3, pp. 493–504.
Eerkens J. W., H. Neff and M. D. Glascock
2002 Ceramic Production Among Small-Scale and Mobile Hunters and Gatherers:
A Case Study from the Southwestern Great Basin. Journal of Anthropological
Archaeology 21: pp. 200–229.
Ehmann, D.
1975 Bahtiyaren-Persische Bergnomaden in Wandel der Zeit. Beihefte zum Tübinger
Atlas des vorderen Orients, Reihe B, No. 15. Wiesbaden, Ludwig Reichert.
Ford, I. R.
1972 Barter, Gift, or Violence: An Analysis of Tewa Intertribal Exchange. In E.
Wilmsen (ed.), Social Exchange and Interaction. Anthropological Papers 46. Ann
Arbor, Museum of Anthropology: pp. 21–45.
Forest-Foucault, C.
1980 Rapport sur les fouilles de Keit Qasim III-Hamrin. Paléorient 6: pp.
221–224.
Galaty, J. G., D. Aronson, P. Salzman and A. Chouinard (eds.)
1981 The Future of Pastoral Peoples. Ottawa, International Development Research
Centre.
Garrod, O.
1946 The Nomadic Tribes of Persia To-day. Journal of the Royal Central Asian Society
33: pp. 32–46.
Garthwaite, G. R.
1983 Khans and Shahs: A Documentary Analysis of the Bakhtiyari in Iran. Cambridge,
Geddes, D. S.
1983 Neolithic Transhumance in the Mediterranean Pyrenees. World Archaeology
15: 1, pp. 51–66
Gilbert, A. S.
1975 Modern Nomads and Prehistoric Pastoralism: The Limits of Analogy. Journal
of Ancient Near Eastern Societies of Columbia University 7: pp. 53–71.
Gregory, C. A.
1982 Gifts and Commodities. New York, Academic Press.
Haerinck, E. and B. Overlaet
1996 The Chalcolithic Parchineh and Hakalāan. Brussels, Royal Museum of Art and
History.
Haiman, M. and Y. Goren
1992 “Negbite” Pottery: New Aspects and Interpretations and the Role of
Pastoralism in Designating Ceramic Technology. In O. Bar-Yosef and A.
M. Khazanov (eds.), Pastoralism in the Levant: Archaeological Materials in
Anthropological Perspectives. Madison, Prehistory Press: pp. 143–152.
Henrickson, E.
1985 An Updated Chronology of the Early and Middle Chalcolithic of the Central
Zagros Highlands, Western Iran. Iran 23: pp. 63–108.
Herodotus
1972 The Histories. London, Penguin.
Hodder, I.
1980 Trade and Exchange: Definition, Identification, and Function. In R. E. Fry
(ed.), Models and Methods in Regional Exchange. Society for American Archaeology
Papers 1. Washington, Society for American Archaeology: pp. 151–156.
Hole, F. (ed.)
1987 The Archaeology of Western Iran: Settlement and Society from Prehistory to the
Islamic Conquest. Washington, Smithsonian Institution Press.
Husseini-Kazerooni, M.
1973 Kuche Mall-e Ka Ebrahim, Il Boyr Ahmadi [in Persian]. In T. Firuzan (ed.),
Ilat va Ashayer. Tehran, University of Tehran: pp. 63–74.
Irons, W.
1974 Nomadism as a Political Adaptation: The Case of the Yomut Turkmen.
American Ethnologist 1: pp. 635–658.
1994 Why Are the Yomut Not More Stratified? In C. Chang and H. A. Koster
(eds.), Pastoralists at the Periphery. Tucson, University of Arizona Press: pp.
175–196.
Irons, W. and N. Dyson-Hudson
1972 Perspectives on Nomadism. Leiden, E. J. Brill.
110     A b b a s A l i z a d e h
Jasim, S. A.
1985 The Ubaid Period in Iraq. BAR International Series 267. Oxford, Archaeo
press.
Johnson, G. A.
1978 Information Sources and the Development of Decision-Making Organizations.
In C. Redman (ed.), Social Archaeology: Beyond Subsistence and Dating. New
York, Academic Press: pp. 87–110.
1983 Decision-Making Organization and Pastoral Nomad Camp Size. Human
Ecology 11: pp. 175–199.
1987 The Changing Organization of Uruk Administration on the Susiana Plain.
In F. Hole (ed.), The Archaeology of Western Iran: Settlement and Society from
pp. 107–140.
Johnson, G. A.
1973 Local Exchange and Early State Development in Southwestern Iran. Anthropological
Papers 51. Ann Arbor, Museum of Anthropology, University of Michigan.
Kaboli, M. A.
2000 Archaeological Survey at Qomrud. Tehran, Iranian Cultural Heritage
Organization.
Kantor, H. J. and P. Delougaz (edited by Abbas Alizadeh)
1996 Chogha Mish. Volume 1: The First Five Seasons of Excavations, 1961–1971. Part
1: Text. Oriental Institute Publications 101. Chicago, Oriental Institute.
Khazanov, A. M.
Köhler-Rollefson, I.
1992 A Model for Development of Nomadic Pastoralism on the Transjordanian
Plateau. In O. Bar-Yosef and A. M. Khazanov (eds.), Pastoralism in the Levant:
Press: pp. 11–19.
Kouchoukos, N.
1998 Landscape and Social Change in Late Prehistoric Mesopotamia. Yale University
(PhD dissertation).
Lambton, A.
1973 Aspects of Saljuq-Ghuzz Settlement in Persia. In D. Richards (ed.), Islamic
Civilisation. Oxford, Cassirer: pp. 25–26.
Lefébure, C.
1979 Acces aux ressources collectives et structure sociale: L’Estivage chez Ayt Atta
(Maroc). In L’Equipe écologie et anthropologie des sociétés pastorales (ed.),
Pastoral Production and Society/Production pastorale et société, Actes du colloque
international sur le pastoralisme nomade, Paris, 1–3 Déc. 1976. Cambridge,
Cambridge University Press: pp. 115–26.
Levine, L. D. and C. T. Young Jr.

1987 A Summary of the Ceramic Assemblages of the Central Western Zagros from
the Middle Neolithic to the Late 3rd Millennium B.C. In J.-L. Huot (ed.),
Préhistoire de la Mésopotamie: La Mésopotamie préhistorique et l’exploration récente
du djebel Hamrin. Paris, CNRS: pp. 15–53.
Levy, T. E.
1983 The Emergence of Specialized Pastoralism in the Southern Levant. World
Luke, J. T.
1965 Pastoralism and Politics in the Mari Period. University of Michigan, Ann Arbor
(PhD dissertation).
Malbran-Labat, F.
1981 Le nomadisme a l’époque néoassyriene. In J. S. Castillo (ed.), Nomads and
Sedentary Peoples. Mexico City, Colegio de Mexico: pp. 57–76.
Mallowan, M. and J. C. Rose
1935 Excavations at Tall Arpachiyah, 1933. Iraq 2: pp. 1–178.
Margueron, J.-C.
1987 Quelques remarques concernant l’architecture monumentale à l’époque
d’Obeid. In J.-L. Huot (ed.), Préhistoire de la Mésopotamie: La Mésopotamie
préhistorique et l’exploration récente du djebel Hamrin. Paris, CNRS: pp.
349–378.
Marx, E.
1980 Wage Labor and Tribal Economy of Bedouin in South Sinai. In P. C. Salzman
(ed.), When Nomads Settle: Processes of Sedentarization as Adaptation and Response.
New York, Praeger: pp. 111–123.
Mellink, M. (ed.)
1964 Dark Ages and Nomads: 1000 BC. Istanbul, Nederlands Historisch-
Archaeologische Institut.
Merpert, N. and R. M. Munchaev
1987 The Earliest Levels at Yarim Tepe I and Yarim Tepe II in Northern Iraq. Iraq
49: pp. 1–36.
Oates, D. and J. Oates
1976 The Rise of Civilization. Oxford, Elsevier-Phaidon.
Postgate, J. N.
1981 Nomads and Sedentaries in the Middle Assyrian Sources. In J. S. Castillo (ed.),
Redding, R.
1981 Decision Making and Subsistence Herding of Sheep and Goats in the Middle East.
University of Michigan, Ann Arbor (PhD dissertation).
1993 Subsistence Security as a Selective Pressure Favoring Increased Cultural
Complexity. Bulletin of Sumerian Agriculture 7: pp. 77–98.
112     A b b a s A l i z a d e h
Rice, P.
1999 On the Origins of Pottery. Journal of Archaeological Methods and Theory 6: pp.
1–54.
Roaf, M.
1982 The Hamrin Sites. In J. Curtis (ed.), Fifty Years of Mesopotamian Discoveries.
London, British School of Archaeology in Iraq: pp. 40–47.
1987 The 'Ubaid Architecture of Tell Madhur. In J.-L. Huot (ed.), Préhistoire de
la Mésopotamie: La Mésopotamie préhistorique et l’exploration récente du djebel
Hamrin. Paris, CNRS: pp. 425–435.
Rosen, S. A.
2003 Early Multi-resource Nomadism: Excavations at the Camel Site in the Central
Negev. Antiquity 77: 298, pp. 750–761.
Rothman, M. S.
1988 Centralization, Administration, and Function at Fourth Millennium B.C. Tepe
Gawra, Northern Iraq. University of Pennsylvania (PhD dissertation).
Roux, G.
1966 Ancient Iraq. London, Penguin.
Rowton, M.
1973a Autonomy and Nomadism in Western Asia. Orientalia 42: pp. 247–258.
1973b Urban Autonomy in a Nomadic Environment. Journal of Near Eastern Studies
32: pp. 201–215.
1974 Enclosed Nomadism. Journal of Economic and Social History of the Orient 17:
pp. 1–30.
1981 Economic and Political Factors in Ancient Nomadism. In J. S. Castillo (ed.),
Sáenz, C.
1991 Lords of the Waste: Predation, Pastoral Production, and the Process of
Stratification Among the Eastern Twaregs. In T. Earle (ed.), Chiefdoms:
Power, Economy, and Ideology. Cambridge, Cambridge University Press: pp.
100–118.
Safinezhad, J.
1989 Nomads of Central Iran. Tehran, Amir Kabir.
Salzman, P. C.
1980 When Nomads Settle: Processes of Sedentarization as Adaptation and Response. New
York, Praeger.
1994 Baluchi Nomads in the Market. In C. Chang and H. A. Koster (eds.), Pastoralists
at the Periphery. Tucson, University of Arizona Press: pp. 165–174.
Skibo, J. M. and E. Blinman

1999 Exploring the Origins of Pottery on the Colorado Plateau. In J. M. Skibo and
G. M. Feinman (eds.), Pottery and People: A Dynamic Interaction. Salt Lake City,
University of Utah Press: pp. 171–183.
Smith, A. B.
1983 Prehistoric Pastoralism in the Southwestern Cape, South Africa. World
Archaeology 15: 1, pp. 79–89.
Spooner, B.
1972a The Status of Nomadism as a Cultural Phenomenon in the Middle East. In
W. Irons and N. Dyson-Hudson (eds.), Perspectives on Nomadism. Leiden, E.
J. Brill: pp. 122–131.
1972b Iranian Desert. In B. Spooner (ed.), Population Growth: Anthropological
Implications. Cambridge, MIT Press: pp. 245–268.
Stack, E.
1882 Six Months in Persia. New York, G. P. Putnam’s Sons.
Stark, F.
1934 The Valleys of the Assassins and Other Persian Travels. London, John Murray.
Strabo (translated by H. L. Jones)
1969 Geography. Cambridge, Cambridge University Press.
Sunderland, E.
1968 Pastoralism, Nomadism, and Social Anthropology of Iran. In W. B. Fisher
(ed.), Cambridge History of Iran, Volume 1. Cambridge, Cambridge University
Press: pp. 611–683.
Tapper, R.
1979a Individuated Grazing Rights and Social Organization Among the Shahsevan
Nomads of Azerbaijan. In L’Equipe écologie et anthropologie des sociétés
pastorales (ed.), Pastoral Production and Society/Production pastorale et société.
Proceedings of the International Meeting on Nomadic Pastoralism, Paris, December
1–3, 1976. Cambridge, Cambridge University Press: pp. 95–114.
1979b Pasture and Politics: Economics, Conflict, and Ritual Among Shahsevan Nomads of
Northwestern Iran. New York, Academic Press.
Teitelbaum, J. M.
1984 The Transhumant Production System and Change Among Hawayzma
Nomads of the Kordofan Region, Western Sudan. Nomadic People 15: pp.
51–65.
Tobler, A. J.
1950 Excavations at Tepe Gawra. Philadelphia, University of Pennsylvania Press.
Vanden Berghe, L.
1973a La nécropole de Hakalan. Archéologia 57: pp. 49-58.
114     A b b a s A l i z a d e h
1973b Le Lorestan avant l’âge du Bronze: Le nécropole de Hakalan. Proceedings of

the 2nd Annual Symposium on Archaeological Research in Iran. Tehran, Iranbastan
Museum: pp. 66–79.
1975 Luristan: La nécropole de Dum-Gar-Parchinah. Archéologie 79: pp. 46–61.
1987 Luristan, Pusht-i-Kuh au chalcolithique moyen (les nécropoles de Parchinah
et Hakalāan). In J.-L. Huot (ed.). Préhistoire de la Mésopotamie: La Mésopotamie
préhistorique et l’exploration récente du Djebel Hamrin. Paris, CNRS: pp.
91–106.
Varjavand, P.
1967 Bamadi, Tayefe-i as Bakhtiari [in Persian]. Tehran, Tehran University.
Wright, H. T.
1977a Recent Research on the Origin of the State. Annual Review of Anthropology 6:
pp. 379–397.
1977b Toward an Explanation of the Origin of the State. In J. N. Hill (ed.), Expla
nation of Prehistoric Change. Albuquerque, University of New Mexico Press:
pp. 215–230.
1987 The Susiana Hinterlands During the Era of Primary State Formation. In
F. Hole (ed.), The Archaeology of Western Iran: Settlement and Society from
pp. 141–155.
2001 Cultural Action in the Uruk World. In M. Rothman (ed.), Uruk Mesopotamia
and Its Neighbors. Santa Fe, School of American Research: pp. 123–148.
Young, C. T., Jr.
1969 Excavations at Godin Tepe: First Progress Report. Toronto, Royal Ontario
Museum.
Young, C. T., Jr. and L. D. Levine
1974 Excavations of the Godin Project: Second Progress Report. Toronto, Royal Ontario
Museum.
Zagarell, A.
1975 An Archaeological Survey in the North-East Baxtiari Mountains. Proceedings
of the IIIrd Annual Symposium on Archaeological Research in Iran. Tehran,
Archaeological Research Center: pp. 23–30.
1982 The Prehistory of the Northeast Bakhtiyari Mountains, Iran: The Rise of a Highland
Way of Life. Beihefte zum Tübinger Atlas des Vorderen Orients. Reihe B, No.
42.Wiesbaden, Ludwig Reichert.
Ch a p t e r 5
Desert Pastoral
Nomadism in the
Longue Durée
A Case Study from the Negev and
the Southern Levantine Deserts
Steven A. Rosen1
A rchaeological studies of pastoral nomadic systems over the longue

durée, in this particular case derived from the Negev and surrounding
areas, reflect a deeper and more complex dynamic of social and ecological
adaptation than is usually acknowledged. Although the inherent adaptive
resilience of pastoral nomadic societies has long been recognized as one of its
primary attributes, this recognition has been limited virtually exclusively to
the short-term ethnographic present. This short-term perspective, on which
many historical interpretations have been based, has resulted in a static under-
standing of the actual history of peripheral nomadic societies. The deep time
perspective, the longue durée, based on archaeological study as exemplified by
work in the Negev and surrounding regions, suggests that some of our basic
assumptions concerning these groups are flawed. Re-evaluation of the develop-
ment of these societies reflects an evolution no less complex than that of their
sedentary cousins, places them in a three-dimensional historical context where
they are not simply a static given in the landscape, and indeed implies that our
very definitions of these societies may be problematic.
With the discovery and excavation of ‘kurgans’ in the Russian steppe in
the early 20th century CE, and the recognition that these monumental burial
mounds represented the mortuary activities of ancient nomadic civilizations,
archaeologists understood early on that, under certain circumstances, the study
of ancient pastoral nomadic societies was an endeavor well worth their while
1.
I am grateful to Emmanuel Marx and Benjamin Saidel who provided impor-
tant comments on an earlier draft of this chapter. I also thank the organizers and
participants of the two meetings from which this volume originated for their input
and discussions.
115
116    S t e v e n A . R o s e n
(P'yankova 1994; Koryakova 1998). That these early studies fall unambiguously
into the culture-historic paradigm of archaeological research, and that analyses
seem to have focused primarily on cultural and ethnic systematics (Belenitsky
1968:45–49; Trigger 1989:148–206; Ochir-Goryaeva 1998; Yablonski 1998),
does not lessen the fact that the kurgan builders were understood to be early
pastoral nomads. The comparisons with the historic and modern nomad
hordes of the Asian steppes were implicit but unambiguous. They were also
generally simplistic.
With the adoption of new research paradigms an archaeology of pastoral
nomadism, similar in conception to the already extant archaeology of hunter-
gatherers, and independent of specific culture-historic frameworks, began to
emerge in the 1970s and 1980s (for instance Cohen and Dever 1978; Hole
1978; Watson 1979; Chang and Koster 1986). This archaeology drew espe-
cially from cultural-ecological approaches to anthropology and was facilitated
by new methods in both fieldwork and laboratory analysis. In the Near East
the discovery that herd animal domestication postdated the development of
farming societies, contrary to the classic 19th century CE hunter-to-herder-to-
farmer evolutionary sequence (Clutton-Brock and Grigson 1984; Davis 1984;
Clutton-Brock 1989; Horwitz and Ducos 1998; Horwitz et al. 1999; Zeder and
Hesse 2000), sharpened focus on the uniqueness of pastoral nomadic societies
and their historical development (Lees and Bates 1974; Khazanov 1984). They
were not merely another step on the inevitable progression to complex society
but required special explanation since they did not, in fact, fall into the linear
evolutionary sequence at all (Gellner 1984). As with hunter-gatherer studies,
however, if the crude social evolution, and implied ethnic parallels, of an earlier
generation were abandoned, the models, analogues, parallels, and hypotheses
generated for understanding pastoral nomadic society and its development
were still derived virtually exclusively from comparisons to modern groups.
That is, the basic paradigms for explaining the origins of pastoral nomadism
and the roles of pastoral nomads in history were still taken exclusively from
modern ethnography and ethnohistory.
The insights derived from these comparisons have been tremendous.
In most ways they are the engine that has driven archaeological analyses,
providing a dynamic set of hypotheses and ideas to be examined against an
ever more detailed archaeological record. These ideas have operated at several
scales, from the relatively low-level spatial analysis of individual sites and
activities (Simms 1988; Cribb 1991), and the understanding of herd manage-
ment and culling (Payne 1973; Cribb 1987; Redding 1984), to larger scale
paradigms, such as sedentary-nomad relations (Lees and Bates 1974; Sumner
1986) and market system mechanics (Wapnish 1981; Wapnish and Hesse
1988; Labianca 1990). Some of these ideas and analyses have proven powerful
D e s e r t P a s t o r a l N o ma d i s m      117
enough to challenge accepted historical concepts, such as the perpetual struggle

between the ‘Desert and the Sown’ (Banning 1986; Avni 1996), and the use of
archaeologically invisible nomads as a deus ex machina for historical explana-
tion (Rosen 1992). These research directions should not be abandoned, but
from another perspective it is important to recognize the ultimate limitations
of ethnography and history, in the sense of texts (both ancient and modern),
for understanding the long-term development of pastoral nomadic societies.
In complement, archaeological analysis offers a unique contribution to under-
standing pastoral nomadic adaptations, most especially over the long term
(Hodder 1986; Bintliff 1991).
Three points are to be considered here, none especially new. First, the prob-
lems inherent in ethnographic analogy have long been debated in archaeology
(Ascher 1961). Hypotheses limited to ideas generated from ethnography may
offer little new to be discovered, except by way of re-ordering the components
of explanation. In the adoption of some version of archaeo-anthropological
uniformitarianism for the explanation of the past, that past social and cultural
processes are all recognizable in the present, we ignore the possibility, indeed
the probability, of the existence of ancient circumstances for which there is no
modern parallel or of modern circumstances with no ancient parallel (Bernbeck,
this volume). Given that the scales of archaeological explanation vary, unique
historical circumstances abound. Furthermore, the assumption of explana-
tory uniqueness is rarely warranted. Similar outcomes may have differing and
multiple causes. In terms of pastoral nomadism, the herding societies of the 19th
century CE, impacted by the armies, technologies, and markets of the devel-
oping industrial world system, may not provide ideal models for earlier times
and relations. Similar points have been made regarding hunter-gatherers.
A second and related point concerns the actual reliability of our historical
and anthropological data. From earliest times, texts dealing with pastoral
nomads, deriving almost exclusively from their sedentary neighbors, have
almost universally depicted them as ‘hostile others’ (for instance Burstein,
this volume; Kupper 1957; Mayerson 1986, 1989). Yet in the documenta-
tion of the complexity of relations between the ‘Desert and the Sown’ both
ethnography and archaeology have demonstrated that such generalized and
simplistic characterizations are essentially wrong. The inescapable biases in
early textual references to nomads render difficult their incorporation into
accounts of pastoral nomadic societies. They reflect the politics and agendas
of the state and its citizens rather than objective attempts to describe another
society. Even recent historical documents, as for example those of the Ottomans
regarding the Bedouin, offer perhaps more in what is not written than in what
is actually said (Zeevi 1996). Critical reading is a necessity (Burstein, this
volume). The lacunae in early accounts are even more difficult to overcome.
118    S t e v e n A . R o s e n
Postmodern critique teaches us that modern anthropology cannot serve as an

unbiased observer either.
A third and final point, obviously related to the first two, concerns construc-
tion and conception of the actual category ‘pastoral nomadism.’ In spite of the
general recognition of a continuous spectrum of behaviors and adaptations,
ranging from fully sedentary full-time agriculturalists (with secondary foddered
animal husbandry) to fully mobile nonagricultural pastoralists (Khazanov
1984), and a universal understanding of the geographic variability inherent in
the societies falling under this rubric (Bacon 1954), most of us persist in viewing
pastoral nomadism as some unique social or economic type whose analytic
isolation may have meaning for understanding some larger questions of the
human career (Ingold 1980, 1987). By nature of our analytic framework, the
conduct of inquiry assumes the existence of threshold values, basic bipolarities
in our perception of the structures of human society. Thus, there are groups
that are pastoral nomadic and groups that are not; or, alternatively, there is
some essential value defining pastoral nomadism, present in varying degrees in
different societies, but allowing the measure of the basic ‘pastoral nomadicness’
of a society. Of course, it is precisely these thresholds and values that allow
comparative study, or else we would risk comparing the proverbial apples and
oranges. Without addressing, however, the issue of the reality or legitimacy
of -etic and -emic constructions of the units and concepts of anthropological
inquiry, their use creates internal analytic dynamics defining research questions
and agendas. Having defined pastoral nomadism as distinct from sedentary
agricultural systems, and clearly separated from hunting-gathering (itself previ-
ously distinguished from farming), further analysis assumes the integrity of the
categories. If the hierarchy of categorization is changed, so might the analysis.
That is, analysis of pastoral nomadic groups emphasizes comparative study of
the groups within the rubric. An alternative approach, à la Leach’s (1954) work
in Burma, might examine shifting political and ecological patterns incorpo-
rating farming, foraging, and herding groups within a region as a single unit
rather than as disparate units. Benhke (1980) has noted that tribal configura-
tions in Cyrenaica incorporate a range of adaptations, and numerous scholars
have noted the constant shift back and forth between sedentism and mobility,
between herding and farming (for instance Barth 1961; Marx 1992).
The ultimate result of this emphasis on ethnography and textual history
for our comprehension of pastoral nomadism seems to be its perception
as somehow fossilized. Once the basic structure of pastoral nomadism was
achieved, at different times in history according to different scholars, there
was little or no further fundamental change. Variability is recognized but only
within certain limited parameters. In a circular logic, if pastoral nomadic behav-
iors are basically static, with little postorigins evolution, then the adoption
of models from ethnography is all the more justified. A similar tendency has
been noted for hunter-gatherers. In somewhat caricatured portrayals, the
ethnographies of the mid 20th century CE, looking for models of early hominid
behavior, have been viewed as demeaning to modern hunter-gatherer groups
in their assumption that somehow these groups represent fossilized Stone
Age behavior. Of course, modern hunter-gatherer societies are the result of
complex evolution and history no less than the societies of their anthropo-
logical analysts, and this is now well recognized by those analysts (Lee and
DeVore 1968; Schrire 1980; Headland and Reid 1989; Wilmsen 1989; Solway
and Lee 1990). Pastoral nomadism as a phenomenon is no different. It has
evolved no less than its village and urban contemporaries. There is a richness
of texture to that history that cannot be reproduced through reference to the
ethnographic present, in all its variability.
Achieving Pastor al Nomadism

There are different ways to define pastoral nomadism, and these different defi-
nitions define differing perceptions of the evolution of the phenomenon. In
particular, four distinct criteria have been cited as defining elements: reliance on
herd animals as a primary means of subsistence in contexts of a general scarcity
of agriculture (Hole 1978), tribal social organization in association with herding
(Tapper 1979), pastoral ideology based on herd ownership (Ingold 1980), and
asymmetric economically dependent relations with settled agricultural soci-
eties (Khazanov 1984). In modern contexts these four criteria often appear as
a package (but see Marx 1992), with major exceptions occurring among some
of the African cattle pastoralists whose independence of sedentary agricultural
society seems clear. Historically, using the southern Levantine deserts as a case
study, this apparently integrated package evolved incrementally over the course
of two to three millennia, more or less in the order indicated above. In other
regions the definitional trajectory need not be the same.
The earliest evidence for the penetration of domesticated herd animals into
the deserts comes from eastern Jordan, where late Pre-Pottery Neolithic B
(PPNB) sites show domestic herd animals (sheep and goats) and evidence for
opportunistic agriculture in what appear to be seasonal camps (Garrard et al.
1996; Martin 1999). Farther south, in southern Jordan, the Negev, and Sinai,
there is no evidence for either herding or agriculture, and the economy was
based on hunting-gathering (Tchernov and Bar-Yosef 1982; Bar-Yosef 1984;
Dayan et al. 1986). Sometime during the late 7th or early 6th millennium BCE
(all dates are calendric absolute dates), the Pottery Neolithic, the first domestic
herd animals were introduced into the hunter-gatherer systems of the Negev,
Sinai, and southern Jordan. The direct evidence for this penetration (around
120    S t e v e n A . R o s e n
6000 BCE) derives from domestic sheep and goat dung layers in Negev
rockshelters (Rosen et al. 2005). Indirect evidence, in the form of changes
in architecture, reduction in arrowhead percentages, the appearance of Near
Eastern sheep and goats in Egypt, and the occurrence of domesticates on
the desert periphery, supports this chronology (Goring-Morris 1993; Rosen
2002). Herd animals ultimately replaced hunted animals in the hunter-gatherer
subsistence economy, in essence initiating a herder-gatherer economy based
on small-band level groups.
By the end of the 6th millennium BCE, formative tribal organization, defined
here as a level of demographic organization able to draw on social groups beyond
band size for various activities,2 can be traced in the archaeological record of
the desert. It is evident in increasing site sizes at one end of the size spectrum,
suggesting seasonal aggregation of multiple bands; in the construction of monu-
mental structures (Avner 1984, 1990; Rosen and Rosen 2003), requiring increased
manpower and labor organization; and in cooperative hunting strategies, as in
the use of ‘desert kites’ (Betts, this volume; Meshel 1974, 1980; Helms and Betts
1987). Specifically, settlement patterns seem to have a wider range of size than
evident in the preceding periods, suggesting greater social interaction and more
complex social patterning (Goring-Morris 1993; Bar-Yosef and Bar-Yosef Mayer
2002; Rosen 2002).3 The construction of solstice shrines, some of which contain
individual blocks of stone weighing up to half a ton, also contrasts with earlier
periods and clearly reflects both the ability to create megaliths and the social need
to do so. The construction of desert kites, apparently also initiated in this period,
reflects a level of planning and cooperative hunting also not evident in earlier
periods. The simultaneous occurrence of increased sophistication in hunting
and the adoption of domestic herd animals need not be seen as contradictory.
Hunting plays a social role among many pastoral groups.
Ingold (1980, 1987) ties the development of pastoral ideologies to fundamental
differences between owning a herd and hunting one, between conservation and
exploitation. The contrast between these activities translates to basic differences
in value systems between hunters and herders. For archaeological purposes it is
useful to divide pastoral ideology into two realms: the belief systems and cult;
2.
The definition of tribe is difficult, basically falling between the so-called egalitarian
band and chiefdom, with incipient stratification and specialized economies (Buccellati,
this volume; Parkinson 2002). My intent is not to plug into a fixed evolutionary frame-
work dictating certain features but to indicate a society beyond the smallest scale, one
that has not yet achieved the elaboration of chiefdoms.
3.
Desert kites are gazelle traps constructed of low walls that funnel the animal into
a pit or corral where it can be killed. They were named by British pilots flying over the
Near Eastern deserts who saw in them a resemblance to kites.
and the changes in lifestyle reflecting different value systems deriving from the
ownership of herds. Although it is difficult to reconstruct the specifics of early
pastoral ideologies, it can be no coincidence that a virtual explosion of desert
cult arose in the middle to late 6th millennium BCE (Avner 1984, 1990), the
late Pottery Neolithic, correlated with what I have referred to as the herder-
gatherer phase of pastoral nomadic evolution. This development is reflected
most especially in the construction of numerous desert shrines (Avner 1984;
1990; Goring-Morris 1993; Rosen 2002), often with solstice alignments (Rosen
and Rosen 2003)4 and usually in association with mortuary ritual. Many of these
seem to have had a megalithic aspect in their size and massiveness. While it is
difficult to see the precise symbolic connection between the new ritual behavior,
as reflected in these shrines and the ‘new’ pastoral way of life, this should not
really be surprising. First, the -emic comprehension of prehistoric symbol systems
is one of archaeology’s least successful endeavors as a discipline. Second, taking
a cue from the preceding agricultural revolution, Cauvin (2000) has emphasized
the symbolic aspects, indeed primacy, of symbolic development, yet most of
the iconography of the agricultural revolution has little or nothing to do with
farming. It is worth noting that attempts to interpret the desert iconography
backward in time, from the known symbolism of historical Mesopotamia or the
biblical period in the Levant (Avner 1984, 1990), are problematic. The chrono-
logical span of thousands of years, and the basic social and cultural contrasts
between early desert tribal pastoralism and later sedentary agricultural, urban,
and state society, must be accounted for if a set of symbols from one is to be
interpretively grafted onto the other. Without detailed consideration of the
mechanics of such a transformation, the value of the scholarly wholesale adop-
tion of myth from one society to another must be questioned.
To return to the original point, if we cannot precisely define the belief and
symbol systems, it nevertheless seems clear that a transformation occurred and
that it can be tied to the rise of herding-gathering. The other half of pastoral
ideology, at least from the archaeological perspective, lifestyles reflecting
pastoral values, can be seen primarily in the organization of space. On the small
scale the architectural changes that accompany the transition to pastoralism in
the desert are one of its clearest correlates. The clustered room architecture of
the PPNB is replaced, at some point in the Pottery Neolithic or Chalcolithic,
by pen-and-attached-room sites (Rosen 2002). This type of structure consists
4.
In the report of the archaeological survey that discovered Ramat Saharonim, this
site was dated between the Pottery Neolithic and Early Bronze Age based on cairn
construction and design. Radiocarbon and optically stimulated luminescence dates have
now established that both the shrine and the cairns are late Pottery Neolithic in date,
about 5000 BCE (calibrated; Porat et al. 2006).
122    S t e v e n A . R o s e n
of a large central enclosure or enclosures (sometimes inappropriately referred

to as courtyards) with the actual habitation structures, or huts, attached to the
periphery. The central enclosure has usually been interpreted as an animal pen
(Kozloff 1981; Haiman 1992; Rosen 2003), but this functional interpretation
is not strictly necessary. The entire organization of the typical occupation has
been re-oriented so that the center is now an open space, instead of a densely
packed cluster. That this change coincides with the development of pastoralism
suggests that the adoption of domestic animals had major effects on the organi-
zation of daily life, a trivial conclusion that nevertheless needs to be emphasized.
Speculating, and assuming that the enclosures really were animal pens, the
centrality of herd animals in these sites dominates their very conception.
On the larger scale, an increase in territoriality is also indicated in the
archaeological record, albeit not strongly. The megalithic aspect of the shrines
mentioned above, and in particular the development of large tomb fields
(nawamis, circular vaulted multiple-burial structures, in south and central Sinai;
tumuli in the Negev), suggests an increased need to demarcate and legitimize
claims to territory (Bar-Yosef et al. 1977, 1986). The development of distinct
zones of settlement, indicated by the clustering of sites, is matched by the
development of two level hierarchies in site size (Haiman 1992), probably both
indicating seasonal movement within a territory and tribal organization. This
increased territoriality, while tied to tribal organization, can also be attributed
to the pressures of herd management (Ingold 1980). Hunter-gatherers are
indeed territorial, especially with regard to specific resources such as favored
water sources or the famous Kalahari mgongo nuts. Although these concerns
feature among desert pastoralists as well, their primary source of territoriality
is the need for pasture, a far more extensive requirement than control over
point source resources (although wells and cisterns may be the focus of dispute).
Territoriality may therefore be more marked. On the other hand, extreme vari-
ability in climate may render territoriality counterproductive, requiring great
flexibility in exploitation of resources. Regardless, evidence for territoriality
increases with the adoption of domestic herds.
It is interesting that, to a degree, these archaeological site-distribution patterns
parallel those of desert Natufian and Harifian hunter-gatherers, some 5000
years earlier. The terminal Pleistocene culture territories, however, restricted
to specific zones within the desert, seem to reflect areas of exploitation rather
than areas of ownership or territorial demarcation. That is, there seem to have
been neither South Sinai Natufians sharing a border with the Negev Natufians
nor any eastern Negev Harifians sharing a border with those of the western
Negev and the Negev Highlands. With the rise of herding societies, territoriality
also becomes an issue of shared borders. Population increase cannot be ruled
out as a factor in the trend toward increasing territoriality, thus explaining the
differences between the hunter-gatherer systems mentioned above and the later
herder-gatherer and pastoral nomadic systems. It seems reasonable to tie such
demographic changes to the transition to food production: herding.
The rise of economic relations between the ‘Desert and the Sown’ is the
final defining characteristic of pastoral nomadism to develop. Unlike the
previous criteria, which to a degree can be seen as autochthonous develop-
ments, the rise of asymmetric economic relations between tribal groups and
sedentary societies is predicated on the evolution of social complexity in the
core regions. In the Levant this does not occur until the Early Bronze Age,
and the earliest evidence for such relations between core and periphery dates
to the end of the 4th and beginning of the third millennium BCE. This period,
the Early Bronze Age I–II, saw increased trade between the desert tribes and
the urbanizing Mediterranean core region, especially focusing on Arad in the
northern Negev, as a gateway town for that trade (Kempinski 1989; Finkelstein
1995:67–86; Amiran et al. 1997; Rosen 2003). Although copper seems to have
been the focus of much of that trade, the desert nomads also produced and
exported milling stones, beads, seashells, hematite, certain chipped-stone
tools, and other trinkets. They seem to have imported grain, based on the
common presence of used milling stones with little evidence for agriculture
(and, given the environment, little opportunity), some pottery, and other
goods. The number of sites that can be attributed to the Early Bronze Age
I–II increases tenfold over preceding periods (Rosen 1987), suggesting a major
desert population increase.5 Even with a somewhat ameliorated climate (Rosen
1995), still less optimal than that of the earlier PPNB or the Chalcolithic, it is
unlikely that these larger populations could be sustained in the desert based on
a herding-gathering subsistence. The Early Bronze Age I–II saw the crossing
of the threshold from herding-gathering to pastoral nomadism as reflected in
a dependence trade relationship. Without the economic supports of the seden-
tary zone, especially the import of grain, the Negev pastoral system was not
viable. To clinch this case, the abandonment of Arad around 2700 BCE, seems
to have entailed the collapse of its pastoral nomadic hinterland, reflected in a
virtual absence of sites in the Negev Highlands for several hundred years.
Technological Developments
If the above review somehow defines the genesis of ‘pastoral nomadism,’
there is nonetheless great developmental variability following the crossing
of the threshold. Technological change plays a defining role in some of this
5.
There are not only more sites, but they are also larger, while the period of occupa-
tion (Early Bronze Age I–II) is no longer than earlier periods.
124    S t e v e n A . R o s e n
development, and indeed some of the postorigins technological changes play

as great a role in the common understanding of pastoral nomadism as do the
definitions. Sherratt (1981, 1983) has stressed the significance, beginning in
the formative period of pastoral nomadism, of secondary product exploitation
in the rise of civilization, and this unquestionably play a major role in the
evolution of desert tribal societies. Russell (1988) concluded that the energy
efficiency of herd exploitation based on dairy products and meat was double
that of carnivorous pastoralism. Adding hair and wool to the equation not
only increased the efficiency of exploitation but, as a cash crop, must have
had significant impact on nomad economics beyond subsistence. Similarly,
the domestication of the donkey in the Pottery Neolithic enabled transport
of bulk items not previously amenable to long-distance trade. Even given the
ability to drag large objects on sledges, as in precontact North America, the
exchange system described above for the Early Bronze Age, including large
quantities of milling stones, ceramics, metals, and grain, was unquestionably
predicated on the earlier domestication of the donkey (Ovadia 1992; Sherratt
1981, 1983). The introduction of domestic horses, especially for riding, into
Near Eastern nomadic society, sometime in the late third millennium BCE
(Wapnish 1997), must also have had a major impact.
The adoption of the domestic camel as a pack animal, however, was a water-
shed event in pastoral nomadic history in the Near East. A strong pack camel can
carry up to 200 to 325 kg of goods and travel for several days without watering,
well beyond the capabilities of any other beast of burden (Gauthier-Pilters and
Dagg 1981:110; Bulliet 1990). Although evidence for camel domestication in
the Persian Gulf area is as early as the late third millennium BCE, the earliest
remains of camels in the Levant date to the Late Bronze Age, the second half
of the second millennium BCE, from Tell Jemmeh near Gaza (Wapnish 1981).
Found at an entrepôt town, these camels probably reflect long-distance, deep-
desert trade with Arabia, virtually impossible prior to the camel. Indeed, the
ability to traverse the deep desert served both to facilitate previously difficult
trade systems and to allow deep retreat into the desert wastes, both characteris-
tics of the classical nomad stereotype. The rise of the Nabateans, the sedentized
North Arabian nomads who established a trading kingdom around Petra and the
spice route through the desert, was predicated on camel caravans. In addition to
its use as a pack animal, the invention and adoption of the North Arabian riding
saddle in the latter half of the first millennium BCE provided control of the camel
and stability to its rider that allowed fighting from camelback (Bulliet 1990).
Thus, desert tribal societies could now constitute a greater military threat, with
capabilities well beyond those of earlier nomads (Eph'al 1984). In this context
of military technologies the introduction of the composite bow must also have
enhanced fighting from camelback, as it did fighting on horseback. Although
well documented on the Asian steppe (Shishlina 1997), the subject has been little
noted in the Near East.
The invention and adoption of large woven tents should also be considered
a technological milestone in the history of pastoral nomadism. Although these
tents are usually difficult to trace archaeologically, the total dominance of
stone-hut bases as an architectural type throughout the southern Levantine
deserts through at least the end of the third millennium BCE suggests a
terminus ante quem for their introduction. Assyrian sources (Eph'al 1984:10)
and biblical references (as in Judges 8:11) to tents as dwelling places suggest
presence in the first half of the first millennium BCE. In the Negev the earliest
unambiguous evidence for tents can be attributed to the Nabateans, in the first
centuries BCE to CE (although it is likely that they were introduced earlier).
This evidence consists of stone lines, cleared areas, hearths, and other features
in the same general configuration, although not in the specific details, as one
can see in more recently abandoned tent camps (Saidel, this volume, 2001;
Simms 1988; Avni 1992; Eldar et al. 1992; Rosen 1993). The implications of
the adoption of the tent are numerous. On the simplest level, the increased
potential for residential mobility enhances the flexibility of the nomadic life-
style. Besides contributing to the relative ease of striking camps, however, tents
allow for great flexibility of spatial arrangements within camps. Indeed, the
ability to tent large spaces allows the construction of communal structures far
more difficult to build of the brush and stone construction materials usually
available in the Levantine deserts. Tents also allow easy internal partition,
virtually impossible in the small huts used by nomads in the Negev prior to
their introduction (Magid, this volume).
Desert agricultural technologies, both the construction of wells and cisterns
and the use of runoff irrigation, have also impacted nomadic societies. Although
most of these systems were built during periods of state incursion into the
desert, the basic infrastructures remained after the withdrawal of the state.
Desert wells and cisterns may serve as major foci for pastoral activities, often
structuring movement, serving as borders between tribes or as bones of conten-
tion between them, and obviously extending grazing territories, especially for
sheep and goats. Even given the disdain often accorded agriculture by nomads,
the ability to farm in the desert, as done opportunistically by modern Bedouin,
affects both the geographic patterning of tribal territories and the nature of
economic relations with sedentary settlements; nomads producing their own
grain need not trade for it. Furthermore, nomadic control of farming areas
probably increases internal tribal inequalities. As Marx (1967) has noted for the
Bedouin tribes of mid 20th century CE Israel, peasant subgroups subservient
to the Bedouin often do the actual farming. In the Negev the earliest cisterns,
cut into soft loess and lined with stone blocks and mud plaster, are dated to
126    S t e v e n A . R o s e n
the Iron Age, in the first half of the first millennium BCE. These are accom-
panied by outposts and small runoff irrigation farms (Bruins 1986). Many of
these cisterns still fill with rainwater seasonally. The earliest rock-cut cisterns
and wells in the Negev date to the Nabatean period (Glueck 1959:Figure 41)
and can be associated with the Nabatean penetration of the Negev. These, too,
still fill up regularly. As with the earlier cisterns, these were accompanied by
relatively small runoff irrigation systems and associated farmsteads. The major
expansion of these systems occurs in the Byzantine and Early Islamic periods,
in the middle of the first millennium CE. These terrace field systems are still
used by Bedouin today, who farm opportunistically in areas receiving as little
rainfall as 100 mm per year, half the usual minimum required for dry farming
of barley and one-third for what is needed for wheat.
The introduction of the gun in the late 18th century CE, and its increasing
use through the 19th and early 20th centuries CE, also had a tremendous impact
on Bedouin society in the Near East (Saidel 2001). It is perhaps difficult to
assess the overall effect of this technology on relations between tribal groups
and sedentary society since virtually all ethnographic observations postdate
the gun. Certainly patterns of warfare must have changed. Saidel (2001) notes
that early matchlock muskets could not be used mounted, and Bedouin raiding
strategy shifted to the ambush. This, of course, is rather small scale, and
whether a real shift in power relations occurred as a result of the introduction
of the gun is difficult to ascertain. Even if one did, whether it should be accom-
modated as part of the ebb and flow of power relations between the ‘Desert
and the Sown,’ or whether it constituted a shift to a new status, is a moot point.
The near extinction of such wild herd animals as gazelle, and the local disap-
pearance of other species, like the roe deer and the ostrich, however, can almost
undoubtedly be attributed to the increased use of guns. These extinctions and
near-extinctions do constitute a major change in ecological balance. In terms
of Bedouin adaptations, the social role of hunting, still important in spite of the
economic dependence on herds, must have been significantly reduced.
The advent of more recent technologies, such as trucks for transporting
animals, wheeled water containers, tractors, and various communication
technologies, have had great impacts as well. In fact, whether the most recent
pastoral nomads should even be compared with their predecessors as models
of nomadic behavior is debatable.
The Impact of the Outside World

Beyond the effects of new technologies, the evolving nature of the adjacent
sedentary societies also had profound effects on pastoral nomadic lifestyles,
relations, and adaptations. Given the general trajectory of increasing complexity
and integration of the external world, with obvious fluctuations, there is also
a trajectory to the effects on interactions with and relations to nomadic
societies. That is, the impact of sedentary societies, beyond technology, on
pastoral nomads evolves over time. These impacts are, on one hand, histori-
cally particular but, on another, can be viewed as cumulative and therefore
directional, given certain general cumulative trends in historical development.
They can be examined from several perspectives, which for our purposes will
be classified as territorial, economic, demographic, and ideological.
Although in the myth of the struggle between the ‘Desert and the Sown’
there is a perception of equality between the forces, in reality, over the long
term, sedentary societies have consistently expanded ever deeper into nomad
and desert territories, albeit with fluctuations and retreats. Such expansions
come at the expense of nomad territories, compressing them and increasing
pressure on resources in the even more marginal lands left to the nomads.
They are made possible by increasing populations within stable agricultural
systems, increasing political integration and military power with the rise of
states and empires, and more effective agricultural technologies, allowing better
utilization of the desert. Even in the retreat of the state from the desert, the
physical infrastructures left behind, as mentioned above in the discussion on
technological changes, change the nature of pastoral adaptation. In the Negev,
sedentary society has advanced and retreated repeatedly since the origins of
pastoral nomadism as a phenomenon. These fluctuations are summarized in
Figure 5.1.
Figure 5.1. Schematic graph of the advance and retreat of the agricultural boundary in the Negev based
on data from the Negev Emergency Survey. Vertical scale is in kilometers.
128    S t e v e n A . R o s e n
The intent here is not to detail each successive phase of settlement but
rather to note that successive incursions by state-level societies from the Early
Bronze Age onward extended successively deeper into the Negev,6 and each
such incursion left an ever greater imprint on the landscape and its inhabit-
ants. An additional and crucial point to consider is that although the periods of
settlement were punctuated by long periods of retreat from the desert, nomadic
presence actually mirrors sedentary presence. During the long periods of
reduced state activity in the Negev, the region reverted to peripheral grazing
grounds, as opposed to tribal residential territory. Each incursion is accompa-
nied by a resurgence of nomadic activity on its fringes, specifically adapted to
the specific circumstances of the incursion.
Obviously partially a function of depth of incursion, the economic impacts
of the outside world on nomadic society can be seen to systematically increase
through time. It is difficult to measure this increase directly, but to measure
trade, for example, petrographic analyses of Early Bronze Age ceramic assem-
blages from Negev nomad sites show high proportions of local pottery
produced by the nomads themselves, in addition to some imports. In contrast,
by classical times (Nabatean, Roman, Byzantine, and Early Islamic) 90% or
more of the pottery found on the nomad sites was imported, produced by
specialists in the towns and sites of the empires of Late Antiquity (but see
Barnard, this volume). By recent and subrecent times virtually all pottery used
by the Bedouin, the famous black Gaza Ware, is imported, along with a vast
array of other manufactured goods. Similarly, one can make a case for the ever-
increasing integration of nomad economic activities into the larger economic
sphere. On one hand, anthropologists have long been aware that pastoralism
is but one in a large set of resources and activities supporting pastoral nomadic
societies. Salzman (1972) coined the phrase “multi-resource nomadism” to
describe the complex economic base of most of these societies.
The long-term history of Negev pastoral nomadism suggests that with time
nomad activities increase in their range, the number of people engaged in them,
and their importance to tribal economy. While the earliest developmental
phase, that of the herder-gatherers, still exhibits relative economic autonomy,
the Early Bronze Age nomads, with a range of cottage industries to supplement
their pastoralism (Rosen 2003), seem already to have expanded beyond the
threshold of herding-gathering carrying capacity. Certainly by Nabatean times,
trade systems played a major role in nomad economies, in some ways providing
the raison d’être for their very presence in the Negev. By later classical times,
6.
While Early Bronze Age society in the southern Levant probably did not achieve
state level political organization as usually defined, it did achieve a level of secondary
urbanism, certainly enough to function as a focus for nomad economies.
in the Byzantine period, local nomads served as guides for the large pilgrimage
system to Sinai, raised camels and donkeys for the Byzantine army, received
subsidies from Rome for keeping the peace, and may well have been wage
laborers in a range of tasks associated with the Byzantine frontier in the Negev.
Raiding should perhaps also be seen in this light as an added economic activity
(Sweet 1965). The difference between Early Antiquity and Late Antiquity is
marked, but difficult to quantify. With modern times the range of nonpastoral
activities and their importance seems to increase, with wage labor and warfare
playing especially greater roles. The integration of pastoral activities, however,
those derived from the animals themselves, also increases with the expansion
of access to markets. In the Negev this is difficult to measure archaeologically,
but in the 20th century CE, with the direct impact of western markets on local
peoples, stock raising emphasized cash returns rather than subsistence.
Although Khazanov (1984) demonstrated some time ago that some form
of dependency characterizes most relations between nomads and their seden-
tary cousins, given the above discussion, economic integration with sedentary
societies seems to increase over time. Indeed, in the Near East of recent times,
perhaps beginning during the Late Ottoman era, the cash economies tying
nomads to sedentary market systems (Black-Michaud 1986) seem qualitatively
different from those of earlier eras.
The increasing number of sites in the Negev through time, punctuated
by periods of archaeological decline, and their generally increasing size have
been well documented in archaeological surveys (Rosen 1987). These trends
undoubtedly reflect long-term increases in the population of the Negev,
again punctuated by periods of demographic decline, the same fluctuations
mentioned above in terms of incursions and pastoral expansions. A large
proportion of this long-term increase was in the nomadic populations and can
be seen archaeologically in increases in the numbers of base camps, camps,
and small campsites, especially in those areas beyond the extent of sedentary
and agricultural settlement. An educated guess given the numbers of sites and
their sizes suggests a general increase from a few hundred nomads at most,
during the herder-gatherer phase of the sequence, to thousands in the Classical
era. By way of comparison, the Bedouin population of the Negev in 1947 was
approximately 90,000, about double that of 1931 (Abu Rabi'a 2001:80–85).
These are changes of orders of magnitude. They must reflect fundamentally
different levels of adaptation and integration clearly tied to the evolution of
the sedentary societies with which the nomads were associated and to the
increasing intensity of the relations between them.
The ideological impact of sedentary societies on their nomadic neighbors can
be seen on two levels: in the adoption of new religious beliefs and their effects
on aspects of nomadic society and in shifts in nomadic ideologies resulting
130    S t e v e n A . R o s e n
from sociological changes due to contact and association with external forces.
In general it is difficult to assess ideological shifts and impacts confidently
because of a paucity of materials (nomads often do not build obvious shrines)
and because of the importance of context and meaning for evaluating symbol
systems. Furthermore, although ‘pastoral ideology’ has been cited as a charac-
teristic element in these societies, beyond Ingold’s (1980) proposed conceptions
of ownership developed primarily to distinguish herding from hunting, there is
no consensus on what the defining elements of pastoral nomadic ideology and
beliefs might be. In terms of the adoption of religious beliefs, although Avner
(1984, 1990) has interpreted the components of many early desert shrines as
representing Near Eastern and Mesopotamian pantheons and rites, and hence
reflecting the influences of sedentary societies, his claims are unconvincing.
Standing stones, in various combinations, are features of societies all over the
world; it is not necessary to invoke Mesopotamian imagery to explain them.
Furthermore, the chronological (3000 years) and sociological (urban state to
desert tribe) distance between the early desert shrines, or even the later ones,
and their supposed symbolic cousins is left without any bridging explanation.
The Iron Age cult site at Kuntillet Ajrud (Meshel 1978; Dever 1984) is
a better case for external influence, but there is nothing to tie it to pastoral
nomadic society. The Nabatean syncretism of the Hellenistic and Roman
pantheon with their own is a clear example of external ideological influence, but
it is not clear that the Nabateans were indeed pastoral nomads in any standard
sense of the word by the time they absorbed Mediterranean religion.
Although there is historical evidence for the adoption of Christianity among
some nomadic tribes in the southern Levantine deserts during the Byzantine
period (Mayerson 1963), we have little evidence indicating how this adoption
affected the tribes socially. The adoption, however, of Islam among desert tribal
peoples changed basic social structures throughout the Near East, including
those of the Negev, Sinai, and Jordan. While there is some historical question
as to the ultimate social and political origins of Islam (Crone and Cook 1977;
Donner 1981; Hawting 1982; Crone 1987), the nomads of the Negev adopted
Islam early in its history, as evidenced clearly in the large number of open-air
mosques found in association with Early Islamic pastoral nomadic camps (Avni
1994; Rosen and Avni 1997) and by hints in some of the texts (Mayerson 1964;
Whitcomb 1995). The adoption of Islam provided cohesiveness to nomadic
tribes over large expanses never before achieved. Politically the adoption
of Islam served as a springboard for major developments in Near Eastern
nomadism. The influence of Islam can be seen on a smaller scale as well. Saidel
(this volume) has noted how the adoption of Islam affected the organization of
space within the tent, deriving especially from the changed roles of women in
nomadic Islamic society. The location of mosques in the Islamic sites, always
somewhat above and removed from the occupation sites (Rosen and Avni 1997),
also differs from that of the preceding pagan shrines, apparently more spatially
integrated within the sites.
Ideological shifts occurring as a consequence of contact, without the direct
adoption of belief systems, can be seen mostly politically. Although egali-
tarianism is the nomadic ideal (Sahlins 1968:21; Marx 1967:181), and perhaps
observed more in the breach than in practice, contact with the outside world
demands leadership, and the greater the demands, and the rewards, from the
state or empire, the larger the scale that leadership needs to be. Zimri Lin, king
of Mari in the second millennium BCE, created nomadic headmen through
his need to govern the tribes under his rule; Rome created the chiefs of large
nomadic tribes; and England and France, after the First World War, the kings
of nomadic states. There is a trajectory here that belies the stereotypical stasis
of pastoral nomadic adaptation.
Discussion
Pastoral nomadism in the southern Levantine deserts is a phenomenon showing
intense variation over time. Although we can trace themes and motifs in this
variation, and indeed perhaps some cyclicity in their appearance, there is no
stability or predictability. Pastoral nomadism seems very much an ad hoc impro-
visation to meet the special circumstances of each historical context. In this sense
the desert tribal societies of the southern Levant, and of course of the Near East
in general, should be tied analytically as much to their sedentary complements
as they are to some general phenomenon of pastoral nomadism. The basic
instability of state adaptations (Marcus 1998) then begins to look structurally
similar to that of pastoral nomadic societies. And in this sense we perhaps ought
to invoke Leach’s (1954) work on the political systems of highland Burma and
see the ‘Desert and the Sown’ as two component parts of an integrated long-
term system, fluctuating back and forth between one another. Adams (1978) has
suggested a similar perspective in his conceptualization of ancient and modern
Near Eastern societies as resilient (meaning adaptable) and shifting between
the urban, the rural, and the desert. Rowton’s (1977) dimorphic society reflects
a similar idea, although it is drawn from ‘enclosed nomadism’ rather than the
geographically peripheral nomadism dealt with here. Of course, in these senses
the distinctions drawn above between external and internal impacts and develop-
ments are purely relative. If pastoral nomadic societies in the Levant, and in the
Near East more generally, should be linked analytically to sedentary societies,
then the developmental history of pastoral nomadic societies cannot be static
but should exhibit trajectories parallel and linked to the increasing complexities
and developments of sedentary societies.
132    S t e v e n A . R o s e n
Thus, the developmental changes sketched so briefly in the preceding review

should come as no surprise. Furthermore, if pastoral nomadic societies are
linked developmentally to their sedentary neighbors, then different pastoral
nomadic societies need not show contemporary or parallel developmental
characteristics. The idea of set stages of social evolution has been passé for a
generation of scholars studying the rise of complex societies. It is reasonable
that the concept be abandoned for nomadic societies as well. This is not to say
that pastoral nomadic societies cannot be studied as a class. The demands of
raising stock and migrating seasonally, without intensively engaging in agri-
culture, must result in similarities among these societies, if only by definition.
But these will constitute only partial explications. In our analytic isolation of
pastoral nomadism as a research subject, we risk losing sight of the essential
genius of the phenomenon, that ability to achieve ever more impressive adapta-
tions to marginal social and physical environments.
References
Abu-Rabi'a, A.
2001 Bedouin Century. New York, Berghahn Books.
Adams, R. M.
1978 Strategies of Maximization, Stability, and Resilience in Mesopotamian Society,
Settlement, and Agriculture. Proceedings of the American Philosophical Society
122: pp. 329–335.
Amiran, R., O. Ilan and M. Sebbane
1997 Canaanite Arad: Gateway City to the Wilderness [in Hebrew]. Jerusalem, Israel
Antiquities Authority.
Ascher, R.
1961 Analogy in Archaeological Interpretation. Southwest Journal of Anthropology
17: pp. 317–325.
Avner, U.
1984 Ancient Cult Sites in the Negev and Sinai Deserts. Tel Aviv 11: pp.
115–131.
1990 Ancient Agricultural Settlement and Religion in the Uvda Valley in Southern
Israel. Biblical Archaeologist 53: pp. 125–141.
Avni, G.
1996 Nomads, Farmers, and Town-Dwellers. Jerusalem, Israel Antiquities Authority.
1992 Survey of Deserted Bedouin Campsites in the Negev Highlands and Its
Implications for Archaeological Research. In O. Bar-Yosef and A. M. Khazanov
(eds.), Pastoralism in the Levant: Archaeological Materials in Anthropological
Perspectives. Madison, Prehistory Press: pp. 241–254.
1994 Early Mosques in the Negev Highlands: New Archaeological Evidence on

Islamic Penetration of Southern Palestine. Bulletin of the American Schools of
Oriental Research 294: pp. 83–100.
Bacon, E.
1954 Types of Pastoral Nomadism in Central and Southwest Asia. Southwestern
Journal of Anthropology 10: pp. 44–68.
Banning, E. B.
1986 Peasants, Pastoralists, and Pax Romana: Mutualism in the Southern Highlands
of Jordan. Bulletin of the American Schools of Oriental Research 261: pp. 25–50.
Barth, F.
1961 Nomads of South Persia. Boston, Little, Brown.
Bar-Yosef, O.
1984 Seasonality Among Neolithic Hunter-Gatherers in Southern Sinai. In J.
Clutton-Brock and C. Grigson (eds.), Animals and Archaeology. Volume 3: Early
Herders and Their Flocks. BAR International Series 202. Oxford, Archaeopress:
pp. 145–160.
Bar-Yosef, O. and D. E. Bar-Yosef Mayer
2002 Early Neolithic Tribes in the Levant. In W. A. Parkinson (ed.), The Archaeology
of Tribal Societies. Ann Arbor, International Monographs in Prehistory 15: pp.
340–371.
Bar-Yosef, O., A. Belfer-Cohen, A. Goren and P. Smith
1977 The Nawamis Near Ein Huderah (Eastern Sinai). Israel Exploration Journal
27: pp. 65–88.
Bar-Yosef, O., A. Belfer-Cohen, A. Goren, I. Hershkovitz, H. Mienis, B. Sass and O.
Ilan
1986 Nawamis and Habitation Sites Near Gebel Gunna, Southern Sinai. Israel
Exploration Journal 36: pp. 121–167.
Belenitsky, A.
1968 Archaeologica Mundi: Central Asia. Geneva, Nagel.
Benhke, R. H.
1980 The Herders of Cyrenaica: Ecology, Economy, and Kinship Among the Bedouin of
Eastern Libya. Bloomington, University of Illinois Press.
Bintliff, J. (ed.)
1991 The Annales School and Archaeology. New York, New York University Press.
Black-Michaud, J.
1986 Sheep and Land. Cambridge, Cambridge University Press.
Bruins, H. J.
1986 Desert Environment and Agriculture in the Central Negev and Kadesh Barnea
During Historical Times. Nijkerk, Midbar Foundation.
Bulliet, R.
1990 The Camel and the Wheel. New York, Columbia University Press.
134    S t e v e n A . R o s e n
Cauvin, J.
2000 The Birth of the Gods and the Origins of Agriculture. Cambridge, Cambridge
University Press.
Chang, C. and H. A. Koster
1986 Beyond Bones: Toward an Archaeology of Pastoralism. Advances in Archaeo
logical Method and Theory 9: pp. 97–148.
Clutton-Brock, J. (ed.)
1989 The Walking Larder: Patterns of Domestication, Pastoralism, and Predation.
London, Unwin Hyman.
Clutton-Brock, J. and C. Grigson (eds.)
1984 Animals and Archaeology. Volume 3: Early Herders and Their Flocks. BAR
International Series 202. Oxford, Archaeopress.
Cohen, R. and W. G. Dever
1978 Preliminary Report of the Pilot Season of the “Central Negev Highlands
Project.” Bulletin of the American Schools of Oriental Research 232: pp. 29–45.
Cribb, R.
1987 The Logic of the Herd: A Computer Simulation of Herd Structure. Journal
of Anthropological Archaeology 6: pp. 376–415.
Crone, P.
1987 Meccan Trade and the Rise of Islam. Princeton, Princeton University Press.
Crone, P. and M. Cook
1977 Hagarism: The Making of the Islamic World. Cambridge, Cambridge University
Press.
Davis, S.
1984 Climatic Change and the Advent of Domestication: The Succession of
Ruminant Artiodactyls in the Late Pleistocene-Holocene in the Israel Region.
Dayan, T., E. Tchernov, O. Bar-Yosef and Y. Yom-Tov
1986 Animal Exploitation in Ujrat el Mehed, a Neolithic Site in Southern Sinai.
Dever, W. G.
1984 Ashera, Consort of Yahweh? New Evidence from Kunitllet Ajrud. Bulletin of
the American Schools of Oriental Research 255: pp. 21–38.
Donner, F. M.
1981 Early Islamic Conquests. Princeton, Princeton University Press.
Eldar, I., Y. Nir and D. Nahlieli
1992 The Bedouin and Their Campsites in the Dimona Region of the Negev:
Comparative Models for the Study of Ancient Desert Settlements. In O.
Bar-Yosef and A. M. Khazanov (eds.), Pastoralism in the Levant: Archaeological
Materials in Anthropological Perspectives. Madison, Prehistory Press: pp.

205–219.
Eph'al, I.
1984 The Ancient Arabs. Jerusalem, Magnes Press.
Finkelstein, I.
1995 Living on the Fringe. Sheffield, Sheffield Academic Press.
Garrard, A., S. Colledge and L. Martin
1996 The Emergence of Crop Cultivation and Caprine Hunting in the “Marginal
Zones” of the Southern Levant. In D. R. Harris (ed.), The Origins and Spread
of Agriculture and Pastoralism in Eurasia. Washington, Smithsonian Institution
Press: pp. 204–226.
Gauthier-Pilters, H. and A. I. Dagg
1981 The Camel, Its Evolution, Ecology, Behavior, and Relationship to Man. Chicago,
University of Chicago Press.
Gellner, E.
1984 Foreword. In A. M. Khazanov (ed.), Nomads and the Outside World. Cambridge,
Cambridge University Press: pp. ix–xxvi.
Glueck, N.
1959 Rivers in the Desert. Philadelphia, Jewish Publication Society.
Goring-Morris, A. N.
1993 From Foraging to Herding in the Negev and Sinai: The Early to Late
Neolithic Transition. Paléorient 19: pp. 65–89.
Haiman, M.
1992 Sedentism and Pastoralism in the Negev Highlands in the Early Bronze Age:
Results of the Western Negev Highlands Emergency Survey. In O. Bar-Yosef
and A. M. Khazanov (eds.), Pastoralism in the Levant: Archaeological Materials
in Anthropological Perspective. Madison, Prehistory Press: pp. 93–105.
Hawting, G. R.
1982 The Origins of the Muslim Sanctuary at Mecca. In G. H. A. Juynboll (ed.),
Studies in the First Century of Islamic Society. Carbondale, Southern Illinois
University Press: pp. 23–47.
Headland, T. and L. Reid
1989 Hunter-Gatherers and Their Neighbors from Prehistory to the Present.
Current Anthropology 30: pp. 43–66.
Helms, S. W. and A. V. G. Betts
1987 The Desert “Kites” of the Badiyat esh-Shaur and North Arabia. Paléorient 13:
pp. 41–67.Hesse, B.
1988 Slaughter Patterns and Domestication: The Beginnings of Pastoralism in
Western Iran. Man 17: pp. 403–417.
Hodder, I. (ed.)
1986 Archaeology as Long-Term History. Cambridge, Cambridge University Press.
136    S t e v e n A . R o s e n
Hole, F.
127–168.
Horwitz, L. K. and P. Ducos
1998 An Investigation into the Origins of Domestic Sheep in the Southern Levant.
In H. Buitenhuis, L. Bartosiewicz and A. M. Choyke (eds.), Archaeozoology of
the Near East. Volume 3. Groningen, ARC Publications: pp. 80–94.
Horwitz, L. K., E. Tchernov, P. Ducos, C. Becker, A. von den Dreisch, L. Martin and
A. Garrard
1999 Animal Domestication in the Southern Levant. Paléorient 25: pp. 63–80.
Ingold, T.
1987 The Appropriation of Nature: Essays on Human Ecology and Social Relations. Iowa
City, University of Iowa Press.
Kempinski, A.
1989 Urbanization and Metallurgy in Southern Canaan. In P. de Miroschedji (ed.),
L’urbanisation de la Palestine a L’age du Bronze Ancien. BAR International Series
527. Oxford, Archaeopress: pp. 163–168.
Khazanov, A. M.
Koryakova, L.
1998 Introduction: Some Problems of Nomadic Studies. In M. Pearce and M. Tosi
(eds.), Papers from the EAA Third Annual Meeting: Pre- and Protohistory. BAR
International Series 717. Oxford, Archaeopress: pp. 82–85.
Kozloff, B.
1981 Pastoral Nomadism in Sinai: An Ethnoarchaeological Study. Production pasto-
rales et société: Bulletin d’ecologie et d’anthropologie des sociétés Pastorales 8: pp.
19–24.
Kupper, J.-R.
1957 Les nomades en Mésopotamie au temps des rois de Mari. Paris, Bibliothèque de la
Faculté de Philosophie et Lettres de L’université de Liège.
LaBianca, O.
1990 Hesban I: Sedentarization and Nomadization: Food System Cycles at Hesban and
Vicinity in Transjordan. Berrien Springs, Andrews University Press.
Leach, E. R.
1954 Political Systems of Highland Burma. Boston, Beacon Press.
Lee, R. B. and I. DeVore (eds.)
1968 Man the Hunter. Chicago, Aldine.
Lees, S. H. and D. G. Bates

1974 The Origins of Specialized Pastoral Nomadism: A Systemic Model. American
Antiquity 39: pp. 187–193.
Marcus, J.
1998 The Peaks and Valleys of Ancient States: An Extension of the Dynamic Model.
In G. M. Feinman and J. Marcus (eds.), Archaic States. Santa Fe, School of
American Research: pp. 59–94.
Martin, L.
1999 Mammal Remains from the Eastern Jordanian Neolithic, and the Nature of
Caprine Herding in the Steppe. Paléorient 25: pp. 87–104.
Marx, E.
1967 Bedouin of the Negev. New York, Praeger.
1992 Are There Pastoral Nomads in the Middle East? In O. Bar-Yosef and A.
M. Khazanov (eds.), Pastoralism in the Levant: Archaeological Materials in
Anthropological Perspectives. Madison, Prehistory Press: pp. 255–260.
Mayerson, P.
1963 The Desert of Southern Palestine According to Byzantine Sources. Proceedings
of the American Philosophical Society 107: pp. 160–172.
1964 The First Muslim Attacks on Southern Palestine (A.D. 633/634). Transactions
and Proceedings of the American Philological Association 95: pp. 155–199.
1986 The Saracens and the Limes. Bulletin of the American Schools of Oriental Research
262: pp. 35–47.
1989 Saracens and Romans: Micro-Macro Relationships. Bulletin of the American
Schools of Oriental Research 274: pp. 71–79.
Meshel, Z.
1974 New Data About the “Desert Kites.” Tel Aviv 1: pp. 129–143.
1978 Kuntillet Ajrud: A Religious Center from the Time of the Judean Monarchy. Museum
Catalogue 175. Jerusalem, Israel Museum.
1980 Desert Kites in Sinai [in Hebrew]. In Z. Meshel and I. Finkelstein (eds.), Sinai
in Antiquity. Jerusalem, Israel Society for the Protection of Nature and the
Israel Exploration Society: pp. 265–288.
Ochir-Goryaeva, M.
1998 Early Iron Age Nomads of Low Volga Region (Model of Cultural
Development). In M. Pearce and M. Tosi (eds.), Papers from the EAA Third
Annual Meeting: Pre- and Protohistory. BAR International Series 717. Oxford,
Ovadia, E.
1992 The Domestication of the Ass and Pack Transport by Animals: A Case of
Technological Change. In O. Bar-Yosef and A. M. Khazanov (eds.), Pastoralism
in the Levant: Archaeological Materials in Anthropological Perspectives. Madison,
Prehistory Press: pp. 19–28.
138    S t e v e n A . R o s e n
Parkinson, W. A. (ed.)
2002 The Archaeology of Tribal Societies. Ann Arbor, International Monographs in
Prehistory 15.
Payne, S.
1973 Kill-Off Patterns in Sheep and Goats: The Mandibles from Asvan Kale.
Anatolian Studies 22: pp. 281–307.
Porat, N., S. A. Rosen, E. Boaretta and Y. Avni
1996 Dating the Ramat Saharonim Late Neolithic Desert Cult Site. Journal of
Archaeological Science 33: pp. 1342–1355.
P’yankova, L.
1994 Central Asia in the Bronze Age: Sedentary and Nomadic Cultures. Antiquity
68: pp. 355–372.
Redding, R. W.
1984 Theoretical Determinants of a Herder’s Decisions: Modeling Variation in the
Sheep-Goat Ratio. In J. Clutton-Brock and C. Grigson (eds.), Animals and
Archaeology. Volume 3: Early Herders and Their Flocks. BAR International Series
202. Oxford, Archaeopress: pp. 223–242.
Rosen, A. M.
1995 The Social Response to Environmental Change in Early Bronze Age Canaan.
Journal of Anthropological Archaeology 14: pp. 26–44.
Rosen, S. A.
1987 Demographic Trends in the Negev Highlands: Preliminary Results from the
Emergency Survey. Bulletin of the American Schools of Oriental Research 266:
pp. 45–58.
of the American Schools of Oriental Research 287: pp. 75–85.
1993 A Roman Period Pastoral Tent Camp in the Negev, Israel. Journal of Field
Periphery. In E. van den Brink and E. Yannai (eds.), Quest of Ancient Settlements
and Landscapes. Tel Aviv, Ramot Publishing: pp. 23–44.
Negev. Antiquity 77: pp. 749–760.
Rosen, S. A., and G. Avni
1997 The 'Oded Sites: Investigations of Two Early Islamic Pastoral Camps South of the
Ramon Crater. Beersheva XI. Studies by the Department of Bible and Ancient Near
East. Beer-Sheva, Ben-Gurion University Press.
Rosen, S. A. and Y. J. Rosen
2003 The Shrines of the Setting Sun. Israel Exploration Journal 53: pp. 1–19.
Rosen, S. A., A. B. Savinetsky, Y. Plakht, N. K. Kisseleva, B. F. Khassanov, A. M.
Pereladov and M. Haiman
2005 Dung in the Desert: Preliminary Results of the Negev Holocene Ecology
Project. Current Anthropology 46: 2, pp. 317–327.
Rowton, M. B.
1977 Dimorphic Structure and the Parasocial Element. Journal of Near Eastern
Studies 36: pp. 181–198.
Russell, K. W.
1988 After Eden: The Behavioral Ecology of Early Food Production in the Near East and
North Africa. BAR International Series 391. Oxford, Archaeopress.
Sahlins, M.
1968 Tribesmen. New York, Prentice-Hall.
Saidel, B. A.
2001 Ethnoarchaeological Investigations of Abandoned Tent Camps in Southern
Jordan. Near Eastern Archaeology 64: 3, pp. 150–157.
Salzman, P. C.
Schrire, C.
1980 An Enquiry into the Evolutionary Status and Apparent Identity of San Hunter-
Gatherers. Human Ecology 8: pp. 9–32.
Sherratt, A.
1981 Plough and Pastoralism: Aspects of the Secondary Products Revolution.
In I. Hodder, G. Isaac and N. Hammond (eds.), Patterns of the Past: Studies
in Honour of David Clarke. Cambridge, Cambridge University Press: pp.
261–315.
1983 The Secondary Exploitation of Animals in the Old World. World Archaeology
15: pp. 90–104.
Shishlina, N. I.
1997 The Bow and Arrow of the Eurasian Steppe Bronze Age Nomads. Journal of
European Archaeology 5: pp. 53–66.
Simms, S. R.
Science 15: pp. 197–211.
Solway, J. S. and R. B. Lee
1990 Foragers, Genuine or Spurious: Situating the Kalahari San in History. Current
Anthropology 31: pp. 109–146.
Sumner, W.
1986 Proto-Elamite Civilization in Fars. In U. Finkbeiner and W. Rollig (eds.),
Gamdat Nasr: Period or Regional Style? Wiesbaden, Springer Verlag: pp.
199–211.
140    S t e v e n A . R o s e n
Sweet, L.
1965 Camel Raiding of the North Arabian Bedouin: A Mechanism of Ecological
Adaptation. American Anthropologist 67: pp. 1132–1150.
Tapper, R.
1979 The Organization of Nomadic Communities in Pastoral Societies of the
Middle East. In L’Equipe écologie et anthropologie des sociétés pastorales
(ed.), Pastoral Production and Society/Production pastorale et société. Cambridge,
Tchernov, E. and O. Bar-Yosef
1982 Animal Exploitation in the Pre-Pottery Neolithic B Period at Wadi Tbeik,
Southern Sinai. Paléorient 8: pp. 17–38.
Trigger, B.
1989 A History of Archaeological Thought. Cambridge, Cambridge University Press.
Wapnish, P.
1981 Camel Caravans and Camel Pastoralists at Tell Jemmeh. Journal of the Ancient
Near Eastern Society of Columbia University 13: pp. 101–121.
1997 Equids. In E. M. Meyers (ed.), The Oxford Encyclopedia of Archaeology in the
Near East. New York, Oxford University Press: pp. 255–256.
Wapnish, P. and B. Hesse
1988 Urbanization and the Organization of Animal Production at Tell Jemmeh in
the Middle Bronze Age Levant. Journal of Near Eastern Studies 47: pp. 81–94.
Watson, P. J.
1979 Archaeological Ethnography in Western Iran. Tucson, University of Arizona
Press.
Whitcomb, D.
1995 Islam and the Socio-cultural Transition of Palestine: Early Islamic Period. In
T. E. Levy (ed.), The Archaeology of Society in the Holy Land. London, Leicester
Press: pp. 488–501.
Wilmsen, E.
1989 A Land Filled with Flies. Chicago, University of Chicago Press.
Yablonski, L. T.
1998 The Nomads of Khwarzem Since the 1st Millennium B.C. to the Middle
of the 1st Millennium A.D. In M. Pearce and M. Tosi (eds.), Papers from the
EAA Third Annual Meeting: Pre- and Protohistory. BAR International Series 717.
Oxford, Archaeopress: pp. 133–141.
Zeder, M. A. and B. Hesse
2000 The Initial Domestication of Goats (Capra Hircus) in the Zagros Mountains
10,000 Years Ago. Science 287 (5461): p. 2254.
Zeevi, D.
1996 An Ottoman Century: The District of Jerusalem in the 1600s. Albany, State
University of New York Press.
Ch a p t e r 6
T he Origin of the T ribe

and of ‘I ndustrial’
Agropastoralism in
Syro ‑M esopotamia
G i o r g i o B u c c e l l at i 1
B y the 19th century BCE, most of Syro-Mesopotamia had come under

the political control of dynasties whose rulers, as well as many of their
subjects, had common onomastics and presumably a common ethnic origin.
The people behind this movement are known to us as ‘Amorites,’ an English
gentilic derived from the Hebrew version of the original name Amurrum.2 The
spread of Amorite presence marked a significant break between the third and
the second millennium BCE and has been explained in the literature as the
result of demographic pressure originating in the Syrian steppe. The Amorites
are universally understood as a population in one state or another of nomadism,
who came as invaders from the steppe into the fertile valleys of the Euphrates
and the Tigris to the north and the east, and of the Orontes to the west. It is
part of the generally accepted scenario that the Amorites were a very distinct
group, linguistically, ethnically and socially, from the populations of the river
valleys. It is further the communis opinio that this difference was rooted in
their historically having hailed from an equally distinct geographical region,
the steppe if not the desert. In other words, they were nomads at the origin,
and they went through a progressive series of developmental stages that led
eventually to their complete sedentarization. In a series of publications I have
1.
In line with the central topic of this volume, I will present my thesis from the
perspective of the origin and development of nomadism in an essay format, without
entering into details or providing documentation and without a proper confrontation
with the literature. For the evidence that underlies my thesis please see my earlier
publications listed in the references (Buccellati 1977b, 1981, 1988, 1990d, 1991, 1992,
1993, 1996a, 1996b, 1997a, 1997b, 1997c, 2004; Buccellati and Kelly-Buccellati 1967,
1986). In this chapter, I will only occasionally refer to these titles.
2.
The Sumerian equivalent is MAR.TU (Wilcke 1969).
141
142    G i o r g i o B u c c e l l at i
presented an alternative interpretation of the data, from the point of view of

history, philology, archaeology and geography. My interpretation revolves
around seven major points.
First, the Amorites were originally not nomads but peasants of the long river
corridor known today as the zôr, the Syrian middle Euphrates from the conflu-
ence of the Balikh to a point south of the confluence of the Khabur, where the
valley floor becomes extremely narrow and where the current border between
Syria and Iraq is situated. This area corresponded to the kingdom of Mari,
the only city to control the entire river corridor. Through the early part of
the third millennium BCE these peasants had sufficient land for farming and
herding within the river valley itself. During that period the steppe remained
largely uninhabited.
Second, as the peasant population increased, the limitation of the zôr became
a major obstacle to any further development. Geography played a significant
role: the valley floor is set in a rather deep and narrow canyon that makes
it impossible for irrigation to reach the plateau above it, while rainfall is so
limited as to render any type of cultivation impossible. As a result of demo-
graphic pressure, the peasants took to the steppe, discovering its suitability as
a vast rangeland for their donkeys, sheep and goats. Such a ‘domestication of
the steppe,’ as I have called it, began in the first half of the third millennium
BCE and reached a level of systemic organization around the middle of the
third millennium BCE.3 In other words, occasional and local exploitation of
the steppe would certainly date to earlier periods, while the systematic and
macroregional scope that characterizes the agropastoralist system, the domesti-
cation of the steppe as a region, would be coterminous with the later expansion
of the urban state. This was in turn influenced by the early experience in the
steppe of an important segment of the social group.
Third, in the process the peasants developed the potential for an almost total
autonomy from the control of the state and its central institutions. At first, this
was primarily economic autonomy: the peasants could remain for indefinite
periods in the steppe, where the brackish water of the shallow water table,
adequate for the animals if not for humans, could easily be tapped through a
3.
The earliest systematic body of evidence pertaining to the Amorites dates to the
Ur III period in the last century of the third millennium BCE (according to the middle
chronology). Apart from significant, if scattered, attestations in the centuries imme-
diately preceding, several considerations point to an earlier presence of the Amorites.
The Ur III evidence comes from southern Mesopotamia, a peripheral area from the
point of view of the Amorites, which they must have reached in the later phase of their
expansion. Their dialect, in my interpretation, is an archaic form of Akkadian and as
such must go back to the period before Akkadian became well established.
Agropastoralism in S y r o - M e s o p o tam i a      143
network of wells that the peasants learned to dig. A few springs provided enough
drinking water for the small human groups of shepherds that accompanied the
animals. By remaining with their flocks in the steppe, these groups could easily
avoid the taxation that the central urban government was intent on imposing.
Fourth, what followed was a growing military and political independence.
Besides taxation, the peasants could avoid conscription as well, all of which is
amply attested in the texts of the royal administration of Mari. As they had been
trained as soldiers, they would retain these skills and use them for their own
needs. This was all the more feasible as the formal urban army of the central
government was quite unprepared for the steppe. It is not attested that it ever
ventured there, all the military events entailing a confrontation between the
regular army and the Amorites taking place in the valley floor.
Fifth, instead of the sedentarization of a nomadic group, we must read these
events as a process of nomadization of peasants. Clearly, this nomadization
took place under special circumstances: the new nomadic lifestyle developed
while its carriers retained strong ties to the ‘homeland,’ the valley floor, where
families and properties were located.4 Just as important, the raison d’être for
the whole process was the expansion of economic opportunities, and these
could only be realized if the interaction with the cities remained in force. As a
result their nomadization could never be complete.
Sixth, eventually the Amorites swelled back toward the river valleys as
invaders from the outside. This is the point where my reconstruction rejoins
the traditional view of things: there was indeed an Amorite conquest of the
entire arc of urban Syro-Mesopotamia, so that by about 1800 BCE the whole
area had come under the control of what are known as ‘Amorite dynasties.’ The
difference is that, in my view, the Amorites are returning to what had always
been their original homeland rather than conquering it as total outsiders.
Seventh, linguistically the Amorites remained clearly distinct from the
other Semitic populations of Syro-Mesopotamia, and this has universally been
considered as reflecting their presumably distinctive geographical origins. In
my view the difference can more easily be explained as reflecting a difference in
socio-economic background: just as the Amorites are in effect Akkadian peas-
ants, so Amorite is the rural version of the same urban language spoken by the
Akkadians. As expected, it retains archaic features that were lost in Akkadian
under the influence of Sumerian, a language with which the Amorites had no
reason to come in close contact.
4.
A qualification of nomadism and mobility is the central theme of this volume, and it
is not my purpose here to contribute to the pertinent theoretical discussion. See Porter
2000 for a lucid exposition of the basic principles, and the discussions in the literature,
as they affect the Amorites.
Against this background I will develop two particular points that I consider of
interest for the topic of this volume because they offer a paradigm that is quite
at variance with other current models. The first point is the origin of the tribe
seen as a political counterpart to the city-state, and the second is the concomitant
development of what I would like to call ‘industrial nomadism.’ First, however, I
will briefly review some salient aspects of urban political institutions against the
backdrop of which the remainder of this chapter can best be understood.
Historical Background:
The Urban Revolution
The ‘urban revolution’ had taken place about a millennium before the
events that I propose to consider here. It was the crystallization of a series of
phenomena that we like to view today in the light of a process of increased
complexity. Long ago I argued that this process is coterminous with the estab-
lishment of the state (Buccellati 1977a). In other words, the city is radically
different from the noncity (the earlier settlements and their communities)
because of its sociopolitical organization. Power was held through vastly
different mechanisms and the perception of power so articulated impacted
deeply on the worldview of the people who supported it. I will summarize my
argument under two headings: functionalization and fragmentation.
First, functionalization: the archaeological evidence that signals great changes
at the turn of the third millennium BCE consists of three classical phenomena:
the vastly expanded size of settlements, the appearance of monumental archi-
tecture and the introduction of writing. But how did the transformation
that we can so clearly read from what we retrieve in the ground impact the
mental template of the people that made it possible? What was the perceptual
response of the people affected? Consider the following, for each of the three
phenomena just mentioned.
Human groups expanded in size beyond the limits that would make face-to-
face association possible, typically placed at 3000 to 5000 individuals. Thus, a
new sense of community developed, one based not on personal acquaintance
but on territorial solidarity. An individual would identify with another because
of assumed shared interests, even if the two individuals had no previous
personal knowledge of each other, and this sharing rested on the physical conti-
guity within the boundaries of the same residential settlement. The tensional
factor that made this linkage possible, and desirable, was the awareness that
the ‘other,’ if unknown at the personal level, was known on the level of func-
tionality: if one needed the services of a carpenter, it mattered little whether
he was already known as a person or not (Buccellati 2005). Functionality was a
new overriding bond, and its efficiency was based on the notion that one would
quickly come to expect the availability of, for example, a carpenter, whoever he
may be. In this regard the perceptual response to increased community density
is the awareness that one can rely on somebody filling any given slot that might
be required to achieve a certain end.
The construction of massive new structures entailed a leadership that could
identify and, if need be, explain what the intended final result was. It also, of
course, entailed the control of financial means that could sustain the operation
over protracted periods of time when the end was not in sight. It required, in
other words, political will and economic resources. The success of the opera-
tion (for instance when a city wall was eventually closed, or a single roof would
eventually cover a sprawling building) helped to propel the person who had
conceived it onto a pedestal that set him apart from those whose communal
effort had in fact made it possible. The single individuals within this anony-
mous mass would, again, recognize themselves as components of a mechanism
that in fact functioned and produced the intended results. Coercion would go
hand in hand with a sense of communal, if impersonal, accomplishment.
Writing would codify and project onto an impersonal carrier, a clay tablet,
the web of relationships that held this human mass together. The written docu-
ment is the outward, extrasomatic embodiment of an abstract image that does
not as such occur in reality. The subdivision of the human mass into crews,
their work assignments for disparate tasks, and the coordination of compensa-
tion over long periods of time are nowhere found together in reality. They are
found together only as a mental template that acquires a physical embodiment
in the written lists. The functionality of the interrelationships is made all the
more apparent by their representation as graphic slots in a matrix.
This progressive functionalization of human relationships goes hand in
hand with the second major point that I want to make here, which I have
called fragmentation. There is, in the process of functionalization, an intel-
lectual dimension that comes to the fore, especially in writing, but that is in
effect present throughout. Things that in the reality of the physical world are
noncontiguous, whether in space or in time, are made contiguous within the
framework of a single overarching vision. The seed is seen as the plant, the mud
as the brick, the wall as the structure, and so on. Single things that, in their
singularity, do not add up to any larger thing are seen instead as potentially
constitutive of a meaningful and efficient whole. They are seen as fragments of
a larger, as yet unrealized, entity. The emphasis here is on ‘seeing’: the potential
for cohesion among the fragments is not imposed; it is discovered. Hence it is
that the fragmentation about which I speak has a twofold dimension. On the
one hand, noncontiguous elements are identified as potentially contiguous; and,
on the other, mechanisms for bringing about their latent contiguity are set in
motion. The perceptual impact of fragmentation is that of widening the hiatus
between humans and nature, which had begun the moment the simplest tool
was created. When homo faber becomes homo civis, direct knowledge of nature
becomes less important as one learns to rely more and more on an ever greater
array of individuals who identify what are, in nature, but fragments and know
how to reconstitute their wholeness for the benefit of the community. But the
more active the control of fragmentation on the part of some individuals, the
tighter their grip on the process that extends their benefit to the other members
of he community. The cost of solidarity is the surrender of individual control.
The Geogr aphical Horizon:

The Steppe as a Perceptual Macroregion
For a proper understanding of the argument that I want to develop with
regard to the tribe and the industrial nature of its economy, I must also
present some considerations about the landscape and, more specifically, about
its perceptual impact as we can presuppose it. My central point is that the
geographical horizon of the Amorites, as they were ‘conquering’ the steppe,
came to encompass the landscape not as a marginal area but as an organically
perceived macroregion. The landscape is highly differentiated within itself,
characterized by prominent mountain ranges, by springs and oases in the
piedmont area, by wadis punctuated by trees, and, as the single human event
that did impact the landscape, by wells. It is, I submit, a major contribution of
the Amorites to have introduced this new element of perceptual geography in
Syro-Mesopotamia, which impacted not only their own historical development
but eventually that of the urban-based states as well.5 For it is at Mari that the
first proper Mesopotamian macroregional urban state came into being. It is
characterized by the vast heterogeneous territory that the state controlled only
vicariously, as it were, through the Amorites in their dual role as, on the one
hand, more or less nominal subjects of the king and, on the other, as those who
had discovered the steppe as a region and de facto controlled it. This is reflected
in the royal title of ‘king of Mari and Khana,’ where the tribe gives its name to
the macroregion. It is the perception of a macroregion that served eventually
as the springboard for the dynamics of political development in Mesopotamia
from the second millennium BCE onward.
‘Perceptual geography’ is a concept that I have briefly discussed elsewhere

5.
(Buccellati 1990b). It refers to the way in which geographical phenomena acquire a

physiognomy of their own, which one might call -emic, in the mind of the individuals
who confront them, potentially different from their objectively measurable, -etic dimen-
sions. Thus a given feature in the landscape may be perceived by some as a ‘mountain’
even if the elevation is objectively minimal.
The notion of a ‘macro-region’ is important and needs some clarification

(Buccellati 1996c). In contrast with it, the landscape of a territorial state estab-
lished around a city, specifically a city-state, is univocal in the sense that it is
homogeneous in terms of geographical features and is dominated by the built
environment of a single structural complex: the city with its ziggurat and city
walls. It is also perceptually present as a whole to all its inhabitants, since each
individual can easily walk across the whole of the urban center and its hinterland
while having the ziggurat visible from most points within the territory. Not so
with the landscape of the steppe. It is too vast to be viewed as a whole from any
given point of view, and it is too differentiated internally to be comprehended
univocally. The concept of contiguity is useful in this respect. The spaces within
a city-state are all perceived as contiguous because they can be so viewed physi-
cally through a minimum of dislocation and because they are uniform. Even a
vast space, such as the river valley dominated by Mari, can fit this description
because its geographical features remain the same from one end to the other of
the long trough. But one cannot perceive of the steppe as contiguous in the same
vein, not only on account of its much greater extension but also on account of
its greater internal differentiation. That it could nevertheless be so perceived
was the accomplishment of the Amorites.6 They unified in their perception a
landscape that the city only sensed as a world beyond. It was through them that
the steppe could become an organic whole because of their intimate knowledge
of its diverse features. The term macroregion refers to just such an organic,
unified perception.
It is significant in this respect that there is in Amorite, as there is in today’s
Syrian Arabic, a proper name for the steppe. In fact, there are two names:
yamina, for the steppe to the right side of the Euphrates; and sam'al for the left
side (Buccellati 1990c; the terms are otherwise understood as referring to south
and north, respectively). The two areas are known today as shamia and jazira.7
6.
This bracketing of noncontiguous perceptions I have defined as “meta-perception”
(Buccellati 2005). In the case presented here it refers to the vast and internally differ-
entiated expanse of the steppe ‘perceived’ as a single entity, even though its constituting
elements cannot be visually grasped at once as a single unit. The same concept applies to
time, as well as space. For instance, the spring season is perceived, or ‘meta-perceived,’ as
a season only because of the juxtaposition that humans can make with the other seasons,
which are not at any given time copresent with spring. The concept, if not the term, is a
central theme of Porter (2000:32–45, 216), who speaks of the “transcendence of space.”
7.
I have argued that the terms “sons of . . . (banū in Amorite, mārū in Akkadian) . . . the
right (yamina)” or “. . . the left (sham’al)” steppe are not to be viewed as tribal names but
as common nouns equivalent to the English nomads or to the Akkadian mārū ugāarim,
meaning peasants or literally “sons of the irrigation district” (Buccellati 1990c).
This split into two steppes is determined by the Euphrates, perceived not only
as a major feature of physical geography, which hinders the movement of the
large herds from one side to the other, but also as the locus where the urban
world resides, with all of its state controls. In contrast, the proper names for
the right and the left steppe are not found in the Sumero-Babylonian south,
where even the common noun şērum has a much more limited use and a more
restricted semantic range than in texts relating to Syria.
The territory is seemingly limitless, yet it can be perceived as a whole
because of its unhindered openness. Movement along given paths, with stops at
fixed and well-known points (especially wells and oases), gives a physiognomy
to the landscape that can be grasped as a single entity in spite of its vastness. It
can safely be assumed that the peasants in the southern agrarian plains would
have hardly any personal knowledge of urban landscapes other than their
own. Most of the peasants we are here considering, instead, would travel long
distances and subsume the wide expanse of the steppe within a single over-
arching perception of its geographical features and resource potential. While
no single individual ever had the whole territory immediately within his or her
field of vision at any given moment (the way this could happen in a southern
city where the ziggurat provided a point of reference visible from most places
within the urban territory), people had a relational perception in that they had
seen most of it with their own eyes. Such a perception of the steppe was crucial
for the impact it had in the political sphere. Control of the steppe’s resources
rested on the unified perception of its organic quality as a macroregion, and
in turn this called for the establishment of new political structures that could
match the macroregional challenge. The answer was the tribe.
The Dynamics of the Sociopolitical Process:

The Origin of the Tribe
It is against the backdrop of the above phenomena that we can appreciate
the broader institutional significance of the Amorite movements. Just as the
‘urban revolution’ coincided with the origin of the state, so the ‘steppe revo-
lution’ coincided, in my view, with the origin of the tribe. And accordingly,
this explains how the tribe emerged as a political institution alternative to
the territorial state. The very brief outline above provides the context within
which the first proposal, concerning the origin of the tribe, makes sense. The
progressive distancing from the urban environment induced an effective state
of autonomy from the coercive institutions represented by the central urban
government. In the steppe the shepherds could avoid primarily taxation and
conscription, yet there were two major ties that remained in effect and kept
them bound to their urban origins.
First, they had personal interests they could not possibly jettison, from
cultivated land to which they held rights to the family network that retained
an essentially agrarian dimension. Second, and more important if less imme-
diately apparent, there was a tensional factor that conditioned their incipient
nomadization, namely their factual dependence on the institutional complex
found in the city. A significant point to be stressed in this connection is that
the villages as they existed in the orbit of the city were by no means ‘a-urban.’
They were instead fully immersed in urban reality as an institutional complex,
however different they may have looked in terms of the built environment in
which they lived. Their real estate was secured by deeds sanctioned by the
central government; the effective use of resources, from availability of traded
goods to control of the river waters by means of canals, was regulated by the
state; conflict resolution was ultimately attributed to royal judgment; access to
a reality beyond the tangible world, in the form of both communal cultic events
and private apotropaic rituals, was in the hands of temple institutions that were
essentially urban; and even such a mundane fact as communicating with distant
correspondents was in the hands of a few specialized scribes. I have defined
this situation as ‘para-urban’ (Buccellati 1996c). Quite unlike prehistoric settle-
ments, the historical village could not exist apart from its urban context.
As these para-urban peasants began to move away from the villages, the
immediacy of the wider urban context was threatened, and the fragility of their
situation emerged. How could these para-urban peasants protect property in
the steppe? How could they co-ordinate access to resources such as specific
pasture grounds or springs and oases? How could they adjudicate conflicts
without courts and judges? How could they relate to the divine world without
access to either the religious technicians who could perform omens or the
organized cultic events unfolding in the temple? How could they send written
messages if no scribe was available? These were institutions they had come
to depend on but would have had to forgo if wholly isolated in the steppe. As
long as their absence from the village was temporary, the problem was all but
nonexistent. As nomadization began to develop as a temptingly permanent
mode of life, however, its consequences were felt at an ever deeper level. To
put it differently, the new situation was undermining the functionalization that
the urban revolution had ushered in. In that context it served as the mechanism
that made it possible for urban society to retain its cohesiveness and efficiency
in spite of its demographic growth. If the pendulum would now swing away
from this context, and encourage the development of a properly nonurban,
rather than para-urban, society, how could the aggregative power of the new
human group be secured? How could group solidarity be maintained beyond
the level of face-to-face association, but without the scaffolding of the territo-
rial contiguity that was at the basis of the urban revolution?
The answer, in my view, was the establishment of the tribe. The city had
shown that some form of functionalization was necessary if large-scale aggrega-
tion was to be achieved and maintained over time. People had come to identify
themselves as components of an organism higher than their personal awareness
of each other might allow. Each recognized another member of the group not
necessarily because of personal acquaintance but because a particular ‘other’
member would fit a particular functional slot, understood to be of service to
the community as a whole. The perceptual link that remained was that of
territorial contiguity: all members of the community lived within a single city,
whose physical integrity was signaled, among other things, by its city walls
or in its immediate hinterland so that each member of the group had a direct
perceptual link with the central city. The territory was a perceptual reality that
helped the individuals to coalesce into a unified whole.
The tribe emerged as an alternative, in some ways mirroring the city-state, or
modeled on it, but in other ways greatly at variance with it.8 It was modeled on
the city-state in that it aimed at serving the same basic needs of political aggrega-
tion beyond the level of face-to-face association. But it replaced the perceptual
basis of aggregation from territorial contiguity to institutionalized interpersonal
relationships. The aspect of mobility, or nomadism, was secondary in this respect
on two grounds. First, it was not pervasive, because many members of the groups
remained de facto sedentary. Second, it was not permanent, because those who
followed the herds in the steppe would retain their roots in the settled areas, to
which they would return on a regular basis. Even if partial and occasional, the
dislocation of a sizable portion of the group in the steppe did propose an alterna-
tive apprehension of what could serve to assure group solidarity. Thus kinship,
rather than territorial contiguity, became the bond: a kinship that was not neces-
sarily biological but could be ascribed through social mechanisms.
The sources document an interesting phenomenon that has been much
discussed in the literature and that, I believe, can best be interpreted in the
light of what I have just said about the origin of the tribe, namely the appear-
ance of people identified as ha-pí-ru in the cuneiform sources. The word has
been read in a variety of ways and has been related to the Hebrew term ‘ibrîm,
from which the very English term Hebrew derives. I have proposed to read
the Akkadian term as hābirū, standing for the Amorite ‘ābirū, meaning “those
who cross over to the clan” (Buccellati 1990b, 1995). I will briefly present
both the linguistic and the historical sides of this argument, because it ties in
with the larger issue of the origin of the tribe. Linguistically, I view the term
as a participle from the same (Amorite) root ‘br, from which ‘ibrum (hibrum in
In this sense, viewed as a state, an Amorite tribe is quite different from the prehis-
8.
toric ‘tribal entities’ (Bar-Yosef and Bar-Yosef Mayer 2002:360–361).

Akkadian) derives; ‘ibrum is the word for ‘clan,’ and it can be understood as a
noun of action (much like ilkum, literally ‘the going,’ comes to mean ‘service’).
It can therefore be understood as ‘the ingathering,’ from which the specialized
lexical meaning ‘the clan’ would have developed. The participle ‘ābirum would
then mean ‘the one who ingathers.’ Given the special lexicalized meaning of
‘ibrum as ‘clan,’ we can then understand ‘ābirum as ‘the one who joins the clan.’
The Hebrew correlate ‘ibrî would retain an alternative adjectival formation:
instead of a participle, it would be (in its original formation) a gentilic derived
from the same word, ‘ibrum and would have the slightly different meaning of
‘the one who belongs to the clan,’ ‘the clansman.’
In my view the term refers, in its Akkadian and Amorite contexts, to those
individuals who are outside the original family nexus of the village but who never-
theless aim to join the tribe, which is perceived as a gathering place alternative to
the territorial state to which otherwise those individuals would belong. They are,
as it were, people in transition between two states, territorial and tribal, taking
advantage of the de facto interstices that exist between the two institutions. To
this extent they are displaced persons in the process of being assimilated into an
alternative, and still rather fluid, sociopolitical institution. This fits quite well
with all the references in the texts to the ha-pí-ru, in the second millennium BCE,
not only in the early part of that period, at the time of the Amorites, but also in
the latter part, the so-called Amarna period. The Hebrew usage fits in as well,
though semantically the term stresses the result, rather than the starting point,
of the process. This is quite consonant with the pertinent historical moment.
In terms of our current interests, the phenomenon of the ‘ābirū, as I under-
stand them, is significant in that it highlights the dynamic pull of the new
institution, the tribe. It is quite frankly an escape, an aspect that is presented in
particularly vivid terms in the so-called Amarna letters but an escape that has
a positive destination. While the motivation would have been in most cases
economical, we may also assume a sociopolitical and even a psychological dimen-
sion. The tribe embodied a certain nomadic ideal, which is even romanticized in
Akkadian literary texts from Gilgamesh to the so-called ‘Theodicy.’ The escape
to the tribe would have appeared to some as a way of eluding not only taxation
and conscription but also the deeper and in some ways dehumanizing effects
of the urban revolution, the urban functionalization of human beings, and the
fragmentation of the relationship with nature. The institutional dimension of this
process lies in its serving as a mechanism for the effective ascription of outsiders
into the new sociopolitical aggregate. It is interesting that no such mechanism
is found for the older state model, that of the city-state, where presumably a
simple transference of residence would have been sufficient to qualify a person
as a member of the new community. In the tribal state, on the other hand, what
was expected was not just immigration but true ‘naturalization.’
The Concomitant Economic Development:

The Nature of ‘Industrial’ Nomadism
The notion of ‘industrialization’ within such a temporally remote context as
that of the third millennium BCE makes sense if we look at it from the point
of view of the fragmentation briefly described above (Buccellati 1990b:98). In
this light, industrialization is the process whereby the moments of a production
sequence are perceived as discrete segments, often at considerable remove one
from the other. The urban users of a bronze object are not necessarily aware,
much less do they individually control, the long chain of events that has led
to the manufacturing of the object. They may know the smith who fashioned
it but hardly ever the tradesman who brought the ingots, the craftsman who
smelted the component metals, or the miners who extracted the ore. It is only
the very few who are aware of the whole sequence who can in fact control it.
The impact on perception was enormous. Think of another even more ancient
case, the introduction of agriculture. To understand the relationship between
the seed and the plant entails precisely such an understanding of, and control
over, a fragmented chain of events. It was as if human perception could collapse
in one single perceptual moment two very distinct segments that are not in and
of themselves perceptually contiguous, temporally (seed and plant) or spatially
(ore and object). There is, as it were, an artificial perception (or ‘metapercep-
tion’), which suspends the rules of natural perception by going beyond them:
humans can now bracket in their abstract vision what is not bracketed in their
concrete apprehension of things.
The economic impact was equally momentous. It was for the first time that
one could give a numeric value to aspects of the human world that previously
did not seem to have one. The seed was worth more than what would normally
be associated with its immediate use. The seed was now worth the plant that
would eventually sprout from it. Thus it was that long-distance trade developed
not only for finished products, or for materials in which one could easily see the
finished product, such as an obsidian core, but for materials such as raw metals,
the potential of which could only be fully realized through a complex series of
intermediate steps. And obviously those who could control these intermediate
steps would stand to reap benefits incommensurate with the value that one
could attribute to each intermediate step by itself.
How does this pertain to our topic? The shepherds who took to the steppe
ended up controlling precisely such a chain of discontinuous segments in
the utilization of their herds. And they could do so without the interference
of third parties, such as merchants, or of the state. The interference of third
parties would have come from individuals who, as in the case of the metal
trade, would have identified points of origin and organized the transshipment
and partial transformation of the desired goods. The Amorites controlled the
entire chain, from the breeding and the caring of the animals to the shearing
and the supply of dairy products. In fact, they were able to extend, in the
process, their mercantile role to the only other critical good found in the
steppe, namely salt.9 The ‘Amorite salt,’ as it is called in the sources, was the
major source of this new commodity so critical for the new urban economy for
its use not so much as a dietary supplement but rather as the main preservative
of perishable foodstuffs.
The interference of the state had typically taken the form of taxation. This
was a way to draw an income from the intermediate steps by assessing a value
on the finished product. Hence it is that the pertinent Mesopotamian texts
speak at great length of the concern of the state to control the time of shearing,
because that was when the finished product, the wool, took concrete shape. But
the state had no control over, had in fact no inkling of, the steps that preceded
the shearing. This involved a knowledge of the steppe territory, the develop-
ment of a network of wells, the care for the animals over long periods in these
areas of otherwise difficult access. The very remoteness of the territory allowed
the shepherds to eschew such interference to the extent that they could avoid
returning to the urbanized area of the river valleys.
This vast and autonomous Amorite trading system is what I call ‘industrial
nomadism.’ It was born in the settled areas, and it irresistibly gravitated toward
them. It could never have entered in the mind of these shepherds to cut off the
ties with their own markets. Quite the contrary, they nurtured the market that
they had in fact created and for which they had all interest in maintaining, since
they were the only ones who could effectively run it. When these urban markets
failed, with the urban demise of the Middle Euphrates (after the collapse of
Mari and Terqa), they still did not cut their ties with the cities. Rather, they
re-oriented their aims to the west, remaining inescapably linked with the urban
dimension. The tribal structure that had developed in the process served as the
institutional backbone that kept the system going for centuries, throughout
the second millennium BCE. Outsiders, the merchants and the state, were
effectively excluded as the tribal organization allowed the Amorites to serve as
their own merchants and their own state. The new functionalization provided
by the tribe consisted precisely in providing the scaffolding for the ever greater
expansion of the economic interests of the group and, in effect, for its ever
stronger affirmation of political autonomy as well.
9.
Truffles are another product indigenous to the steppe, as prized today as in antiq-
uity but obviously not as critical for the urban economy as salt or the produce of the
herds.
The Nature of the Evidence:

The Role of Archaeology
Of central interest to the issue of Syro-Mesopotamian nomadism and mobility
is the question of the archaeological evidence, by which one understands in a
general sense the material culture emerging from the ground. The evidence on
which I have relied so far comes almost entirely from written documents, and
although it is true that these themselves have come from the ground, except for
the biblical references to the ‘ibrîm, they are not typically considered part of the
material culture or the archaeological record.10 We may then ask the question
whether, given that we know the actual situation relatively well from the written
sources, we can use this as a privileged situation to identify correlative elements
in the material culture that can explicitly be linked with the nomadic culture
of the Amorites? There are three areas at which to look. The first is the steppe
itself, where we might seek evidence of nomadic presence. The second is the
reflection of a possible Amorite influence in the material culture of the urban
populations. The third is the impact of the geographical environment and the
ways in which it conditions today the lifeways of analogous human groups.
The first line of research has proven singularly sterile. We still do not have
sites in the steppe that may be called Amorite, not by the numerous wells that dot
the steppe, not in the oases where springs provide fresh water, not in the areas of
the salt playas of Bouara, Palmyra or the Jabbul. It is interesting that there are no
Amorite cities similar to the Aramean cities from the first millennium BCE. One
reason behind the research project that centered on my excavations at Terqa was
precisely to explore the possibility that this might prove to be some sort of an
Amorite border town between Mari and the Amorites. Could there be evidence
of a distinctive population that exhibited different traits from those known to
us from the excavations at Mari? Nothing in the results could have been farther
from this initial hypothesis. In fact, it was especially the evidence from Galvin’s
(1981) analysis of our archaeozoological remains that started me on a different
track, as this suggested that the treatment of the herds was not in keeping with
what one would expect if the herders had been large-scale pastoralists originating
in the steppe (Zeder 1995). The most probable candidate for the status as an
Amorite city is Palmyra, where there may be evidence of an urban settlement
as early as the first part of the second millennium BCE (al-Maqdissi 2000) and
which would indeed appear to have been structurally quite different from the
other urban settlements of Syro-Mesopotamia. An interesting new development
is the discovery of a whole new type of evidence, that from petroglyphs in the
Although I disagree with this understanding of the nature of the archaeological
10.
record, I retain it here as a premise for my considerations about the relevance of the
record for the Amorite question, on account of its being the generally held opinion.
Hemma region (Van Berg et al. 2003). There is no indication that they may be
specifically Amorite, and in fact even the date is uncertain, but it is plausible that
these may be the work of shepherds who were familiar with the general themes
and motifs of the urban world. If so, the petroglyphs would offer a resonance of
the iconography that we otherwise know only from the high achievements of
the urban workshops. On the other hand, one cannot simply link the Amorites
with monuments isolated in the steppe only on account of their location. Thus
the stelas of Jebelet el-Beida are more likely to be associated with a political
program of the Hurrian dynasties of piedmont cities like Urkesh than with the
steppe pastoralists (Dolce 1986).
The second line of research has also limited success, as the interpretations
proposed, however plausible, cannot be argued on archaeological evidence
alone. Apart from references in the texts to artifacts qualified as ‘Amorite,’
we have no artifacts in the archaeological record from the pertinent urban
settlements that can be interpreted as coming from nomadic populations at
home in the steppe, nor is there much in the same record that could be associ-
ated with a nomadic origin were it not for what we otherwise know about the
broader situation from contemporary textual evidence. Let us consider briefly
three specific cases. The diffusion of the iconography of the god Amurru, who
became very popular in the early second millennium BCE and who on the face
of it would appear to be the eponymous tutelary deity of the Amorites, does
not reflect an Amorite religious tradition at all but is rather the projection of
the urban perception of their social reality (Kupper 1961). The construction of
an ‘Amorite wall’ (BÀD MAR.TU; Buccellati 1966:327–328) by King Shulgi
toward the end of the third millennium BCE attests to an awareness of threat
translated into a defensive measure important enough to be the argument of a
year name. If there indeed was a wall running along a defensible border, one
might expect to find a trace of it in the ground, but none has been found so far.
But if it was a string of fortresses, or even a single fortress, strategically placed
to confront the incomers, then any archaeological trace of it would be hardly
distinguishable from that of other small settlements. The most successful effort
at connecting archaeological data from the settled areas to the pastoralists is the
one undertaken by Porter (2000, 2002), who attributes to them the building of
mortuary monuments serving as markers of tribal territory and social identity.
Her insightful interpretation builds on a careful analysis of archaeological data,
though it still remains tied to textual information for the finer interpretive
points. It is interesting, in this respect, how another segment of the written
tradition suggests that the Ur III Sumerian perception of the Amorites was
at odds with the self-perception by the Amorites themselves as interpreted by
Porter, since a Sumerian cliché about the Amorites is that “on the day of their
death they are not buried” (Buccellati 1966:331).
The third line, centering on geography and ethno-archaeology, is produc-

tive but limited in its final import. A careful analysis of the landscape shows
differences that are not immediately noticeable on maps. This is what I have
called “perceptual geography” (Buccellati 1990b), a study of how the perception
of the geographical features, assumed to have been the same in ancient times,
may legitimately be projected in the past and throw light on the reaction of the
ancients to their environment. In our case it is especially the nature of the valley
trough, for which the modern perception has a name, zôr, that is matched, I
have suggested, by the ancient ah Purattim. Analogously, a careful consider-
ation of the lifestyles of the shepherds that take to the steppe, in particular
the different ways in which they adapt to the winter season, is instructive in
showing how the territory can best be exploited by small groups that retain
strong ties to a base in the river valleys (Bernbeck, this volume).
Discussion
In point of fact, the archaeology of the material culture in Syro-Mesopotamia,
in and of itself, tells us very little about the Amorites. Without the informa-
tion derived from the cuneiform texts, we would know nothing today of their
name as a group and much less of their names as individuals, nothing of their
interaction with the urban areas on the valley floors, nothing of their use of the
steppe, nothing, in fact, of even their existence. There is, in my understanding
of events, a very good reason for this. However ‘nomadized’ the Amorites may
have become in the course of time, they had remained essentially rooted in
the rural para-urban areas from which they hailed. The Amorite homeland
remained in the alluvial plains, even though they carved out for themselves
an ecological niche in the steppe that could be considered as an alternative,
or transitional, heartland.11 Their presence in the steppe was thus not only
an extension of village life, but it was also quite ephemeral even by nomadic
standards. The herds were tended by minuscule human groups, whose extended
family remained in the alluvial areas, where they were eager to return and from
whose material culture they amply drew. That they could have given rise to
the tribe as an alternative political institution, as I have proposed, is not in
contrast with this picture. The tribe echoes the central goal of the city as a
territorial state to the extent that it provides effective aggregative means for
As I have argued elsewhere (Buccellati 1990a, in press), it was only in the latter
11.
part of the second millennium BCE that the central position of this heartland was
established as a whole new type of tribal-territorial state, namely the kingdom of
Amurru that I consider to have been centered around Palmyra, resulting in the first
proper ‘steppe kingdom.’
a population that is larger than the sum of persons known to each other on
the basis of face-to-face association. But it does not derive, in our case at least,
from a purely endogenous expansion of a family nucleus that is nomadic in
origin. Rather, it results from the secondary co-alescing of small human groups
that seek effective aggregation outside of the contiguous territorial formula:
an aggregation higher than the village but distinct from the state organization
that controlled the village from an urban perspective. Hence the very fluid
situation that characterizes, in my view, the Amorite movements: they were
peasants turned nomads turning urban.
References
al-Maqdissi, M.
2000 Note sur les sondages réalisés par Robert du Mesnil du Buisson dans la Cour
du Sanctuaire de Bel à Palmyre. Syria 77: pp. 137–158.
Bar-Yosef, O. and D. E. Bar-Yosef Mayer
2002 Early Neolithic Tribes in the Levant. In W. A. Parkinson (ed.), The Archaeology
of Tribal Societies Archaeological Series 15. Ann Arbor, International Monographs
in Prehistory: pp. 340–371.
Buccellati, G.
1966 The Amorites of the Ur III Period. Naples, Istituto Orientale.
1977a The “Urban Revolution” in a Socio-political Perspective. Mesopotamia 12: pp.
19–39.
1977b Fugitives and Refugees of the Amarna Age. Journal of Near Eastern Studies 36:
pp. 145–147.
1981 The Origin of Writing and the Beginning of History. In G. Buccellati and C.
Speroni (eds.), The Shape of the Past: Studies in Honor of Franklin D. Murphy.
Los Angeles, Institute of Archaeology and Office of the Chancellor, University
of California: pp. 3–13.
1988 The Kingdom and Period of Khana. Bulletin of the American Schools of Oriental
Research 270: pp. 43–61.
1990a From Khana to Laqê: The End of Syro-Mesopotamia. In Ö. Tunca (ed.),
De la Babylonie à la Syrie, en passant par Mari (Mélanges Kupper). Liège: pp
229–253.
1990b “River Bank,” “High Country,” and “Pasture Land”: The Growth of
Nomadism on the Middle Euphrates and the Khabur. In S. Eichler, M. Wäfler
and D. Warburton (eds.), Tell al-Hamidiyah 2. Göttingen, Vandenhoek und
Ruprecht: pp. 87–117.
1990c The Rural Landscape of the Ancient Zôr: The Terqa Evidence. In B. Geyer
(ed.), Techniques et pratiques hydro-agricoles traditionnelles en domaine irrigué.
Approche pluridisciplinaire des modes de culture avant la motorisation en Syrie.
Bibliothèque Archéologique et Historique 86. Tome I. Paris, Geuthner: pp.

155–169.
1990d Salt at the Dawn of History: The Case of the Bevelled Rim Bowls. In
P. Matthiae, M. Van Loon and H. Weiss (eds.), Resurrecting the Past: A
Joint Tribute to Adnan Bounni. Leiden, l’Institut Historique-Archéologique
Néerlandais de Stamboul: pp. 17–40.
1991 A Note on the MuŠkênum as a Homesteader. In R. J. Ratner, L. M. Barth,
M. Luijken Gevirtz and B. Zuckerman (eds.), Let Your Colleagues Praise You:
Studies in Memory of Stanley Gevirtz. Maarav 7: pp. 91–100.
1992 Ebla and the Amorites. In Eblaitica. Volume 3. Winona Lake, Eisenbrauns: pp.
85–106.
1993 Gli Amorrei e l’ “Addomesticamento” della Steppa [in Italian]. In O. Rouault
and M. G. Masetti-Rouault (eds.), L’Eufrate e il tempo: Le civiltà del medio
Eufrate e della Gezira siriana. Milan, Electa: pp. 67–69.
1995 Eblaite and Amorite. In E. Eichler, G. Hilty, H. Loeffler, H. Steger
and L. Zgusta (eds.), Namenforschung. Ein internationales Handbuch zur
Onomastik / Name Studies. An International Handbook of Onomastics / Les noms
propres. Manuel International d’Onomastique. Berlin-New York, de Gruyter:
pp. 856–860.
1996a Amorite. In J. Turner (ed.), The Dictionary of Art. Volume I. London-New York,
Macmillan: p. 795.
1996b Review of M. Anbar Les tribus amurrites de Mari. Archiv für Orientforschung
42/43: pp. 233–234.
1996c The Role of Socio-political Factors in the Emergence of “Public” and
“Private” Domains in Early Mesopotamia. In M. Hudson and B. A. Levine
(eds.), Privatization in the Ancient Near East and Classical World. Peabody Museum
Bulletin 5. Cambridge, Peabody Museum of Archaeology and Ethnology:
cover page and pp. 129–151.
1997a Amorites. In The Oxford Encyclopedia of Archaeology in the Near East. Volume I.
New York, Oxford University Press: pp. 107–111.
1997b Syria in the Bronze Age. In The Oxford Encyclopedia of Archaeology in the Near
East. Volume 5. New York, Oxford University Press: pp. 126–131.
1997c Urkesh and the Question of Early Hurrian Urbanism. In M. Hudson and B.
A. Levine (eds.), Urbanization and Land Ownership in the Ancient Near East,
Peabody Museum Bulletin 7. Cambridge, Peabody Museum of Archaeology and
Ethnography: pp. 229–250.
2004 Il secondo millennio a.C. nella memoria epica di Giuda e Israele [in Italian].
La Rivista Teologica di Lugano 9: pp. 521–543.
2005 The Perception of Function and the Prehistory of the State in Syro-
Mesopotamia. In B. D. Dillon and M. A. Boxt (eds.), Archaeology Without
Limits: Papers in Honor of Clement W. Meighan. Lancaster, Labyrinthos: pp.

481–492.
in press Review of Bernard Geyer (ed.), Conquête de la steppe et appropriation des terres
sur les marges arides du Croissant fertile. Archiv für Orientforschung.
Buccellati, G. and M. Kelly-Buccellati
1967 Archaeological Survey of the Palmyrene and the Jebel Bishri. Archaeology 20:
pp. 305–306.
1986 The Glory of Ancient Syria. Terra 24: 4, pp. 6–14; and Terra 24: 5, pp.
22–29.
Dolce, R.
1986 Per una Riconsiderazione delle Opere Figurative da Gebelet el-Beyda [in
Italian]. Contributi e materiali di archeologia orientale (Università degli Studi di
Roma “La Sapienza”), 1: pp. 307–331, Figures 59–60.
Galvin, K. F.
1981 Early State Economic Organization and the Role of Specialized Pastoralism. Terqa
in the Middle Euphrates Region, Syria. University of California, Los Angeles
(PhD dissertation).
Kupper, J.-R.
1961 L’iconographie du Dieu Amurru dans la glyptique de la Ire dynastie babylonienne:
Mémoires de la classe des lettres. 2e Série, Tome 55, Fasc. 1. Bruxelles, J. Duculot,
Imprimeur de l’Academie Royale de Belgique, Gembloux.
Porter, A.
2000 Mortality, Monuments, and Mobility: Ancestor Traditions and the Transcendence of
Space. University of Chicago (PhD dissertation).
2002 The Dynamics of Death: Ancestors, Pastoralism, and the Origins of a Third-
Millennium City in Syria. Bulletin of the American Schools of Oriental Research
325: pp. 1–36.
Van Berg P.-L., N. Cauwe, J.-P. Hénin, S. Lemaitre, V. Picalause, K. Ahmo and M.
Vander Linden
2003 Fieldwork at the Archaeological and Rock Art Sites of the Hemma Plateau
(Hassake, Syria), Season 2002. Adumatu 7: pp. 7–20.
Wilcke, C.
1969 Zur Geschichte der Amurriter in der Ur-III-Zeit. Die Welt des Orients 5: pp.
1–31.
Zeder, M. A.
1995 The Archaeobiology of the Khabur Basin. Bulletin of the Canadian Society for
Mesopotamian Studies 29: pp. 21–32.
Ch a p t e r 7
Pastoral Nomadism in the

Central A ndes
A Historic Retrospective Example
D av i d L . B r o w m a n
O ne of the challenges of this volume is to focus on how pastoral nomadism

can be defined, characterized and recognized in the archaeological record.
The manner in which researchers began considering livestock management in
various regions around the world has influenced the terminology employed
and the way the issue has been approached. This chapter looks at one thread of
the historical definition and recognition of prehistoric animal management in
the Central Andes of western South America. In this region a version of what
has been termed ‘seminomadic pastoralism’ became a principal way of life as
early as seven to eight millennia ago (Wheeler 1984; Wheeler et al. 1976) and
persists into the 21st century CE with some relict groups from central Peru to
northwest Argentina. The essentially treeless grassland expanses of the high-
land region, along with the sharp vertical ecology of the Andes, significant
variations in precipitation and temperature, and the presence of two camelid
species with natural seasonal migration behavioral patterns, made the region
an ideal locale for development of such seminomadic pastoralism.
Roughly a quarter of the land in the Central Andes is composed of semi-arid
and arid grasslands, with the altiplano and puna constituting the bulk of these
areas.1 Precipitation averages nearly 600 mm per year in the northern part of
the Peruvian puna but decreases rapidly as one moves south, to less than 100
mm per year in the southern Bolivian altiplano. This lack of adequate rainfall
is one factor that has led to pastoralism. As elsewhere, grazing is a typical
means to extract subsistence resources from lands too dry for conventional
plant agriculture. But in addition, the Andean altiplano and puna grasslands
range in elevation from 3000 to 4800 m. Some of the grasslands that otherwise
The primarily treeless grasslands of the high Andes are known as the wet puna in
1.
northern Peru, the dry puna in southern Peru, the altiplano in the wide plateau from
Lake Titicaca to central Bolivia and the salt puna in southern Bolivia and northwest
Argentina.
160
P a s t o r a l N o ma d i s m in the C e n t r a l A n d e s      161
might potentially receive adequate moisture for plant agriculture are simply too
high and too cold to be suitable for plant cultivation. Pastoralism once again
becomes the mechanism for extracting resources in these circumstances.
Junin Seminomadic Agropastor alism

The original reconstruction of Andean ways of life, by anthropologists half a
century ago, described camelid pastoralism developing only as a state industry
under the Inca Empire. The great ethnohistorian John Murra and the dean of
Andean archaeology, John Rowe, concurred on this interpretation. Although
I have written extensively about the nature of prehistoric and ethnohistoric
camelid management, the original archaeological field evidence that led to my
research in this area has never been appropriately described. So let me present
that yet unpublished historic evidence here. It provides the ‘hook’ for me to
describe some of the wider archaeological correlates of Andean pastoralism
that are to be observed.
To first follow this example up then, in order to generalize from it later:
when I initially located what appeared to be evidence of seminomadic pasto-
ralism in the Andes, a prehistoric pastoralism that existed at least 3000 years
before the Inca, my reports were greeted with skepticism by my doctoral
dissertation committee. Their view was that the experts had already spoken,
and the evidence that I thought I had for pre-Inca pastoralism ran counter to
accepted wisdom. I had better go back into the field and re-address my find-
ings. This situation had a career-changing impact because it forced me to learn
a good deal about pastoralism in order to convince my committee, as well as
other Andeanists, of the considerable prehistoric time-depth of such pastoral
lifestyles in the Andean puna and altiplano.
At about this same time Thomas Lynch (1971) was analyzing evidence from
his work in northern Peru and shaking up the Peruvianist academy by arguing
for Archaic Period seasonal transhumance, in part based on seasonal exploita-
tion of camelid resources. Although the idea of seasonal transhumance and
seminomadic pastoralism were outside of the normative paradigm then, within
a decade a multitude of Andean scholars had picked up the chase, and today the
idea of camelid pastoralism beginning among the Archaic Period populations in
the high Andean grasslands is recognized as one of the well-known alternative
subsistence exploitation patterns.
One question of semantics that needed to be worked out was what kind of
‘pastoralism’ the Central Andean peoples displayed. Our early reports spoke
of ‘nomadic’ camelid pastoralism. But Old World pastoralism experts, such
as Anatoly Khazanov and others, questioned whether we truly had evidence
of nomadic pastoralism or whether it was really rather more seminomadic
162    D a v i d L . B r o wma n
pastoralism or even a kind of ‘tethered’ nomadism. The significant reliance

on plant agricultural production, at least in later times, made them question
whether this Andean pattern was not better termed ‘agropastoralism.’
What was this early evidence? The project research was in the Jauja-
Huancayo basin of the Mantaro River drainage in Junin Department in the
central highlands of Peru. This basin (about 60 km long and up to 12 km wide)
is relatively flat, being an old Pleistocene lake bed, running from 3125–3350 m
in elevation. In recent periods it has been utilized extensively for grain and root
crop agriculture. Although the majority of the herding activities in prehistoric
and ethnohistoric periods took place along the basin slopes and adjacent high-
land grasslands, at elevations roughly 3500–4000 m, the valley bottom was the
locus of some seasonally permanent habitation sites.
The seminomadic pastoralism here is environmentally driven both by
temperature and precipitation variation, although because of altitude, tempera-
ture is more critical. Rainfall in the basin itself is comparatively abundant, in
terms of grassland requirements, ranging from an average of 635 mm a year at
the north end to an average of 740 mm at the south end. On the higher slopes
the precipitation falls off accordingly with increases in elevation. As is typical
for the Andean puna, the rainfall varies considerably between years, with an
average variation of ±20%. The area can also experience exceptionally dry
years, with precipitation as little as 140 mm, as well as rare abnormally wet
years, with precipitation in some restricted locales up to 2000 mm.
Temperature seems to be the more critical environmental variable, with
respect to the selection of a pastoral production focus. Diurnal ranges are
large, owing to the high elevation, and typically vary more than 20°C in a day.
Summer highs may reach 25°C, and winter lows will drop to -10°C, or lower.
In the central Peruvian highlands average annual temperature drops off at
the standard rate of -0.5°C per every 100 m increase in elevation. For plant
agriculture this translates to about two weeks more time required to reach
maturity for each 100 m of elevation, as well as increased frost and freeze risk,
effectively reducing or eliminating the possibilities of plant agriculture as a
subsistence strategy at higher elevations. Thus frosts and freezes drastically
reduce the growing season at higher elevations in the Jauja-Huancayo zone,
making it difficult to grow crops outside the basin and river valley.
Archaeologically, a radical settlement pattern shift occurred in the Jauja-
Huancayo sector of the Mantaro valley with the development of fully
agricultural sedentary communities at about 600 CE. But in this basin the
development of such fully committed agricultural communities was compara-
tively late for the Andes. Evidence from the project indicated that for at least
the preceding 3000 years (which now of course can be extended much longer),
settlements in this area had been seasonally ephemeral. A large number of the
earlier localities identified were transitory camps on the hillslopes and flanks
and on ridgetops and in the surrounding puna, outside the zone where agricul-
ture was later practiced in the lower valley. In several instances these transitory
camps were clearly associated with corral-like constructions. These prehistoric
sites looked very comparable to the herding residences recorded for ethno-
graphically documented ovicaprid and camelid herders in this region but, when
first found, ran counter to the then-accepted paradigm, as noted above.
Because the original research design had been developed without awareness
of this new subsistence pattern variation, the thrust of fieldwork focused on the
functioning of the later components of this system rather than on searching
for its origins. The basin study region contained most, if not all, of the lower
elevation recurrently utilized wet-season pastoral sites but only a small fraction
of the more ephemeral higher elevation dry-season herding locales. This initial
work did not therefore adequately describe the entire system but rather focused
more on the wet-season sites, of which there was a larger sample.
The wet-season sites were small villages, seasonally occupied, situated near
a spring, stream or other major water resource. This location supplied suffi-
cient water for the frequent artificially enlarged marshy areas, which provided
the softer grasses particularly needed for alpacas, as well as being suitable for
wet-season cultivation. Ceramic assemblages indicated continual habitation
site utilization for decades or, in the case of the largest settlements, centuries.
There were few artifacts related to agriculture at these sites and agricultural
tools in general were infrequent. The typical wet-season villages had up to a
few score semisubterranean or subterranean pithouses, with mean populations
generally estimated in the neighborhood of 100 persons.
There were three to five larger population clusters within the valley basin
during this period, spaced from 10 km to 25 km from one another (Browman
1976). In immediately pre-Inca times there were three major ethnohistoric
subdivisions of the Wanka peoples living in the area: Hatun-Xauxa, Hanan-
Wanka and Lurin-Wanka. But rather than arguing for any direct historical
linkages between these larger early prehistoric valley bottom clusters and
the later three tribal subdivisions, my interpretation was more general: that
population buildup had reached a suitable density that, around 2000–3000
years ago, there were already different political tribal subdivisions among the
agropastoralists in the basin, and that these three to five centers most likely
correlated to such tribal differentiations.
The wet-season sites were the locus of much of the ritual behavior, as is still
the typical pattern among seminomadic camelid herders in the Andes today.
Archaeological evidence of ritual behavior (Browman 1970) was seen in the
production and utilization of ceramic human and animal figurines, whose
examples derived nearly exclusively from the wet-season valley floor sites.
Human figurines were found throughout all phases, from the very earliest
identified sites through the Colonial period, but the animal figurines were
limited to the pastoral-related sites in early phases.
Animal figurines provided some potential insights into herd management
techniques. While among the human figurines male and female variants appeared
in about equal numbers, about 80% of the camelid (llama or perhaps llama and
alpaca) figurines are female. Typically today, herders prefer only a few males to
service a large harem of females; this patterning of figurines suggests a similar
prehistoric sex-selection pattern. The archaeozoological techniques available
when the research was done did not permit documentation of this pattern in the
osteological assemblage, but work in the Central Andes since then by experts such
as Jonathan Kent (1982), George Miller (1979), Katherine Moore (1989), and
Jane Wheeler (1984) has indicated the importance of sex and age profile ratios
in identifying pastoral sites. In addition, all the female camelid figurines depict
the udder or teats, but only a few depict any genital features. Hence, although
Andean camelids provide too little milk to be utilized as a food supplement by
humans, as elsewhere in the world, still the emphasis on desirable female herd
animal traits appears to have been on their infant suckling abilities rather than
merely the sexual components of fertility and reproduction.2
Other livestock management traits similar to modern herding practices were
observed in the figurine collection. ‘Slashes’ comparable to modern ownership
marks were observed on some male figurines. These incisions are interpreted as
ownership marks not only because they closely resemble the devices employed
by modern herders but also because other animal ownership-marking practices
are also known in the central highlands at this time. Ear-notching ownership
marks, identical in location and type to notching procedures still utilized by
today’s herders, have been found on prehistoric llama figurines from that period
in adjacent areas to the north of the Mantaro, making it abundantly clear that
standardized ownership-marking procedures have a history of at least 2500
years in Peru, and probably much longer.
The primitive camelid backpack used for moving goods from field to camp,
as well as for transporting goods on the long-distance llama trade caravans,
also are depicted on the ceramic figurines. And bone toggles, employed for
securing the packing ropes, were recovered during excavation and survey.
These artifacts, from wet-season sites, allowed the reconstruction of a robust
set of pastoral institutions initially estimated to be three to four millennia
duration, now likely much longer.
2.
One of the distinctive, almost definitive, patterns noted at the onset of domestica-
tion of camelids in the Andean region seven millennia ago is the high number of neonate
deaths. It is clear that the survival of newborns has long been of considerable concern
for herders in this region.
During the dry season the populations dispersed to the surrounding

hillslopes, hilltops and puna, where they lived in small hamlets. Dry-season
sites were characterized by as few as 3–15 house platforms, rather randomly
gathered together. The dwellings built on these short-term, seasonal platforms
were constructed of perishable materials. The platforms were most evident on
hillslopes where they were utilized to create level living surfaces. The doorways
or access for habitations built on the hillslopes faced upslope or uphill, which
allowed the resident to enter the structure at ground level rather than having
to climb steps or over embankments. The long axis of these oblong to rectan-
gular platforms was often perpendicular to the hillslope. As it would have been
considerably easier to construct a platform with the axis parallel to the hillslope,
requiring less fill and less revetment, several hypothetical explanations were
investigated to explain this orientation. None has so far proved satisfactory.
These temporary habitation sites were characterized by lithic and bone
tool assemblages related to hunting and herding activities, and they lacked any
agricultural implements. The ceramic assemblages characteristic of this series
of sites dated from the earliest ceramics in the area, perhaps as early as 1500
BCE, to the apparent dramatic termination of this settlement pattern variation
about 500 CE. Some of the hamlets appeared to lack ceramic assemblages, but
it is unknown whether they were preceramic or simply aceramic later sites.
The associated occupation debris indicated that the utilization of these house
platforms was very short-term, possibly only one season, or more likely a few
seasons, and that these grazing-area sites were essentially short-time utiliza-
tion encampments.
These dry-season herding hamlets were usually situated in locations with
good lines of sight. Although the pastoral subsistence pattern was identified
in the basin beginning at least by the late preceramic phases, the majority of
sites in the survey inventory dated to the Early Intermediate Period of the
Peruvian cultural history sequence. The Early Intermediate Period was a time
of general political instability in much of Peru, and occupation sites elsewhere
in Peru during this period often were relocated to higher, defensible locations.
Good sight lines would obviously provide early warning of portending enemy
incursions or rustling, so one explanation is that the dry-season site location
was selected for defensive purposes.
Considering the rather open and vulnerable locations of the wet-season
villages in the valley bottom, as well as the lack of any evidence for substantial
outside contact, the idea of defensive location, other than for rustling protec-
tion, seems unlikely. A more feasible explanation is that these temporary sites
were located at higher elevations to provide the herders with the maximum
view for the monitoring of herds from a distance. It also allowed them to make
early identification of predators, such as pumas and foxes, and provided them
the opportunity to spot, and pursue if they wished, the wild guanaco, deer and
smaller mammals with which they supplemented their diets.
This particular pastoral pattern was abruptly terminated at 600 CE, when
the Wari imperial conquest resulted in the forcible relocation and resettlement
of these valley populations into sedentary communities in maize-growing areas,
a pattern reminiscent of situations described for the Middle East in various
periods. With the collapse of the Wari Empire about three centuries later, the
regional groups then shifted back to a greater pastoral emphasis. The section
of the valley under discussion here seems to have controlled a few million head
of stock by the Inca period. For example, ongoing ethnohistoric research shows
that in the first 15 years of the colonial conquest alone, the Spanish expropri-
ated more than 600,000 llamas and alpacas from two of the three previously
mentioned regional Wanka tribal ‘kingdoms’ that existed here.
Andean Pastor alism Patter ns

The evidence supports the proposition that seasonal pastoralism was involved
in the prehistoric utilization of the Andes, but to what extent can this pasto-
ralism be subsumed under the category of ‘nomadism’? Most current Andean
ethnographies and animal management studies have taken to lumping the
kind of mixed-plant agriculture and animal herding behavior that is found in
the Andean highlands under the general rubric of ‘agropastoralism,’ which
sidesteps the critical issue of the type of movement involved. The fact that
in our example the prehistoric pastoralists had regularly utilized permanent
wet-season villages, to which they returned on a recurring basis and where
they practiced a limited plant agriculture, indicates that these agropastoralist
peoples should better be characterized as seminomadic pastoralists or even
‘tethered’ nomads rather than pure nomadic pastoralists.
Pastoralism frequently involves utilization of land not suitable for plant agri-
culture, land where moisture often is a critical variable so that herders need to
be able to react to local annual precipitation events and vary their exploitation
patterns accordingly. This usually requires herders to move their flocks to graze
in different areas in different years, depending on which area receives adequate
rainfall. To a limited extent this is true of the Andean herders, although, as I have
argued above, temperature is a more important environmental factor here.
Under moisture-driven pastoralism, seasonal migrations are sometimes
seemingly capricious, driven by reports regarding the locations of rains that
season. For the Central Andes, however, the behavioral characteristics of
the llamas and alpacas make it easier to adhere to a much more regular and
constrained migratory pattern. The wild ancestors of the llama and alpaca
migrated seasonally, much like the gazelles formerly did in the Epi-Neolithic
Levant, from winter to summer pastures. Both the wild camelids (guanaco
and vicuña) and the domestic camelids (llama and alpaca) mark their grazing
territory by voiding their bowels in dung heaps on boundary-marking locales.
This activity makes camelid dung collection, for fuel or fertilizer, a relatively
easy process in contrast to, for instance, the dispersed ovicaprid pattern.
These behavioral traits can be viewed as making the camelids ‘pre-adapted’
for pastoral domestication. Early Andean llama and alpaca pastoralism was
integrated into, and generally maintained, the ecosystem in which it was intro-
duced. Innovation here was a conservative process, preserving the former way
of life by simply making the previously hunted wild guanaco and vicuña more
accessible to be sacrificed or harvested when needed, now as herded llama and
alpaca, but otherwise maintaining the extant patterns of camelid migration
and territorial marking.
The territoriality of these species also leads to the development of what can be
viewed as a rather ‘tethered’ sort of nomadism, as the propensity to return to an
established grazing territory has been characterized as tethering the flocks to that
pasturage. Thus, because the natural behavior of the camelids resulted in their
preferring to return to the same areas, the Andean herders could afford to invest
in landscape capital in these pastures, such as enhancing grazing areas through
construction of artificial wet meadows, or bofedales, and similar improvements.
This also made the development of plant agriculture, where possible, a logical
component of the pastoral regime. The prehistoric record appears to support the
following general evolutionary scenario, evolving from this territoriality: first,
hunters following the wild camelids on their annual rounds; second, hunters tran-
sitioning into herders of these same animals, managing them on the same rounds;
and third, herding groups beginning to supplement their subsistence regime with
self-produced domestic plant foodstuffs, thus becoming agropastoralists.
In contrast to Old World pastoral exploitation, Andean camelid herders
never employed milk or blood from live animals as food items. The value of
the animals was in their wool and hides, their meat, and their transport labor.
Over prehistoric time we can observe the principal value of the animals gradu-
ally shifting from their utility as meat producers, for the hunters and early
herders, to their utility as producers of wool and other animal products for the
early agropastoralists, to their more recent prehistoric importance as caravan
animals, as well as producers of pastoral goods.
Agropastoralism in the Andes covers a wide range of possible variations.3
Some of the variants might be subsumed under the “enclosed nomadism” that
I have provided syntheses of the pastoral components in greater detail in other

3.
publications (Browman 1984, 1987, 1990, 1998), so I will not repeat all of that infor-
mation here.
Abbas Alizadeh discusses in this volume for southwest Asia. In some cases of
agropastoralism, where the possibility of plant agriculture is severely limited
by both temperature and moisture, the groups subsumed under this rubric are
mainly pastoral. They maintain a somewhat fictive home community, which
they return to on ceremonial occasions, where they store various equipment
and goods and where they might practice an extremely limited cultivation,
often sowing the seeds and then coming back at harvest time with the hopes of
finding something to reap. But for most of the time, the pastoral community is
split into smaller herding groups, dispersed across the grazing areas.
At the other extreme are the agropastoral communities that are essentially
agricultural but in which the various agricultural families also own livestock,
which are permanently kept in the higher grasslands. In some cases these
families have only a few animals and combine their holdings with neighbors
to make a single viable herd, managed by either hired herding experts or by a
relative of one of the families. In other cases the household livestock holdings
are larger, with the majority of community households having sizable numbers
of animals, herded by relatives on a year-round basis in upper elevation grass-
lands separated from the village and its fields. In most modern ethnographic
examples the location of the flocks is usually within a single day’s travel for
reprovisioning trips, almost never more than two or three days distant.
In other Andean variations of agropastoral groups the community resi-
dents return to the croplands during the wet season, cultivating fields around
a permanent settlement while keeping the animals on fallowed lands from
previous years. At the end of the growing season the community splits into
smaller herding groups, each traveling to their own area of higher elevation
grasslands during the dry season. The location of the dry-season camps may
vary from year to year, depending on the availability of forage, but are usually
situated near some water resource (usually improved as bofedales) to which
the household has traditionally recognized rights. It is this latter pattern, for
example, that is evidenced in the archaeological record of the Upper Mantaro
basin for the millennia prior to 500 CE.
Because of the apparent extensive utilization of long-distance trade caravan
activity, in addition to wet-season fields to secure necessary plant products, the
Andean agropastoralists may have been able to live comfortably with fewer
‘livestock equivalents’ in animals than the theoretical ‘pure’ nomadic pasto-
ralist. Plant resources seem to have provided a significant part of the annual
diet of the recent Andean herding groups, permitting the herders to meet their
subsistence needs with fewer animals.
Ethnographic and ethnohistoric estimates of necessary livestock holdings
vary considerably, but it appears that the basic ‘starter’ herd necessary for the
newly married couple among heavily pastoral-dependent groups was perhaps
no more than roughly 50 camelids. Usually, half of this herd would come
from the bride’s holdings, accumulated through gifts and natural reproduc-
tion growth since her birth, and the other half would come from the groom,
accumulated in a similar manner. The ‘average’ herding family had somewhere
between 100 and 200 animals. Only a few ‘rich’ herders had more than 200
animals, with upper limits on animal holdings usually predicated more on labor
access than on pasture access (Browman 1990).
Maximization, and related least-effort analyses, such as optimal foraging
and linear program modeling, have presupposed a definition of ‘rationality’ as
applied to ‘peasant’ agricultural behavior that involves maximizing of produc-
tion. A logical extension of these analyses is the assumption that ‘rational’
herders, necessarily involved in maximizing animal holdings, will inevitably
overexploit the pasture utilized or held in common. The question of why this
‘tragedy of the commons’ did not occur with Central Andean agropastoral
groups until the late 20th century CE is a puzzling issue under those postulates.
As I have argued in detail elsewhere (Browman 1987, 1994, 1998), the indig-
enous Andean herders dealt with the issues of production uncertainty, both
environmental and social, by risk management or risk minimization. Thus
the minimization of risk rather than the maximization of production was the
operational rationality.
Origins of Centr al Andean

Pastor al Lifestyles
There are a number of competing theories regarding the origin of nomadic
and seminomadic pastoralism. The most frequently cited origin scenarios for
pastoral nomadism include arguments for pastoral nomadism arising from an
adaptation to marginal environments (low temperature or low precipitation);
pastoral nomadism arising as an economic specialization to exploit inter-
regional markets integrated into more complex states, particularly through
control of caravans and transportation; and pastoral nomadism arising as a
defensive strategy by less complex societies against expanding encroaching
state societies, with the opportunity to ‘vote with one’s feet’ as well as highly
mobile resistance allowing one to avoid being unwillingly incorporated into an
expanding hegemony. Animals are the vehicles by which all of these nomadic
trajectories are seen as being operationalized.
In terms of available evidence, Andean seminomadic pastoralists could
fit into the first two of these origin scenarios but not into the third. First,
camelids are herded particularly in areas where plant agriculture is not viable,
because the lands are too high, too cold, or too dry or because the soils are too
poor. This is the pattern observed in Peru since the earliest pastoralists, seven
thousand years ago or earlier. Thus the ‘marginal environment’ origin model
is certainly supported. Camelid pastoralism is the principal, and in many cases
the only, way that these high-elevation lands can be exploited.
But there may be evidence to support a secondary origin, or at least flores-
cence, of pastoralism in the southern altiplano. In areas such as southern Bolivia
and northwest Argentina we have identified archaeological evidence for the
growth of polities associated with pastoralism arising simultaneously with the
development of large interregional markets in the last millennium and a half.
The herding groups owning the transport animals became economic specialists
in moving manufactured goods, as well as staples from one exchange network to
another. This evidence could be marshaled to support the ‘trade and economic
specialization’ model of pastoral nomadism as a secondary development in
response to these conditions.
This secondary florescence in the southern altiplano resulted in a pattern
that Núñez Atencio and Dillehay (1978) have called “gyratory mobility” for
groups trafficking between Chile and Bolivia (probably a rather misleading
appellation). The strategy, however, allowed these pastoralists to thrive during
the Inca Empire and to adapt quickly to the needs of the Spaniards, as the
European conquerors sought to move silver out and goods into the fabled
silver city of Potosi. And even in the late 20th century CE, during periods when
truck fuel was in short supply because of embargoes or too costly because of
runaway inflation, the remnant herding groups experienced brief periods of
resurgence, when their caravan animals once again became a significant means
of moving goods.
Discussion
Identification of the complete suite of archaeological pastoral sites in the Andes
has proceeded slowly. Archaeologists have been able to identify the main-
residence villages, which are the locales of storage and ritual activities, and
often the (at least seasonal) residences of groups that produced part of their own
tuber and grain foodstuffs. But we have had less success in finding temporarily
utilized locales with low archaeological visibility, such as grazing outposts and
hunting camps, and in clearly identifying their functions. We know that until
very recent times, most of the herding groups continued to kill wild camelids,
cervids and rodents on specialized hunting trips, as well as opportunistically
taking the same animals when encountered while herding. Because the faunal
remains and lithic tools are expected to be similar, if not identical, for these
two types of sites, the question arises: how can we distinguish archaeologically
between a hunting encampment and a temporary herding outpost? This issue
becomes even more complicated as we also have to include the temporary
encampments of the trade caravans. Because of the pastoral-derived foodstuffs

brought along on caravan trips, we could make a first-order prediction that
the archaeological record of an overnight caravan stop might not look much
different from that of a hunting camp or a herding station.
Nielsen (2000) has done some pioneering ethno-archaeological work in
trying to help sort this problem out. Traveling with contemporary llama
caravan groups in the Potosi Department, in southern Bolivia, he investigated
the home-based herding locales with main residences, herding posts and
grazing area sites. He also collected information along a set of caravan routes,
looking at the overnight stops, the rest places and the locations where resources
were extracted.4 In the end Nielsen thought he could identify, at least for this
specific contemporary ethnic group, the degree of pastoral specialization,
the goods transported, and their caravan stops as contrasted to their seasonal
herding stations, as well as the ethnic, sociopolitical and geopolitical context
of caravan trade, all based on the specific site context and contents.
More studies of this nature need to be done for other remnant pastoral
groups in the Andes before we can hope to define generalized ‘rules’ for sepa-
rating these site types in all areas. Lessons from these studies need to be applied
to regional surveys. As archaeologists we are currently very good at identifying
farmsteads, villages, and so on, but the present archaeological site-surveying
projects in the Andes are not yet fine-tuned enough to pick up and differentiate
the much more ephemeral herding outpost and caravan stop locations.
The Central Andes is one of the prehistoric locales for independent develop-
ment of pastoral lifestyle. The Andes may have had ‘pure’ pastoral nomadism
at the onset of domestication of the camelids, but very quickly the pattern
evolved into one of seminomadic pastoralism or perhaps tethered nomadism,
and in recent historic periods the area is characterized mainly by seminomadic
agropastoralism.
A pastoral lifestyle still persists in the Central Andes today in marginal zones,
where plant agriculture is severely limited, or impossible, because of climatic
constraints. But more recent prehistoric and historic patterns are also based on
capitalizing on the need for transport of goods between large state-exchange
systems, where the herders control, and have become specialists in, transpor-
tation. Thus the Andean example is not ‘clean;’ it does not fit nicely into an
‘either/or’ category. The agropastoralism is not a ‘pure’ pastoral nomadism,
and its functioning supports more than one of the popular origin models.5
4.
These llameros were involved in a lively trade in rock salt, mined from nearby salt
deposits in dried-up Pleistocene lake beds or salares.
5.
Bonavia Berber (1996) has compiled an 850-page volume, in Spanish, on miscel-
laneous aspects of Andean camelid pastoralism, including an extensive 110-page
bibliography of all sources through the early 1990s.
References
Bonavia Berber, D.
1996 Los camélidos sudamericanos: Una introducción a su estudio [in Spanish]. Lima,
Travaux de l’Institut Français d’Etudes Andines.
Browman, D. L.
1970 Early Peruvian Peasants: The Culture History of a Central Highlands Valley.
Harvard University, Department of Anthropology (PhD dissertation).
1976 Demographic Correlations of the Wari Conquest of Junin. American Antiquity
41: 4, pp. 465–477.
1984 Pastoralism and Development in High Andean Arid Lands. Journal of Arid
Environments 7: 4, pp. 313–328.
1987 Agro-pastoral Risk Management in the Central Andes. Research in Economic
1990 High Altitude Camelid Pastoralism of the Andes. In J. G. Galaty and D.
L. Johnson (eds.), The World of Pastoralism: Herding Systems in Comparative
Perspective. New York, Guilford Publications: pp. 323–352.
1998 Pastoral Risk Perception and Risk Definition for Altiplano Herders. Nomadic
Peoples 38: 1, pp: 22–36.
Kent, J. D.
1982 The Domestication and Exploitation of the South American Camelids: Methods of
Analysis and Their Application to Circum-lacustrine Archaeological Sites in Bolivia
and Peru. Washington University in St. Louis (PhD dissertation).
Lynch, T. F.
1971 Preceramic Transhumance in the Callejón de Huaylas, Peru. American
Antiquity 36: 2, pp. 139–148.
Miller, G. R.
1979 An Introduction to the Ethnoarchaeology of the Andean Camelids. University of
California, Berkeley (PhD dissertation).
Moore, K. M.
1989 Hunting and the Origins of Herding in Peru. University of Michigan, Ann Arbor
(PhD dissertation).
Nielsen, A. E.
2000 Andean Caravans. An Ethnoarchaeology (Bolivia). University of Arizona, Depart
ment of Anthropology (PhD dissertation).
Núñez Atencio, L. and T. D. Dillehay
1978 Movilidad giratoria, armonía social y desarrollo en los Andes meridionales: Patrones
de tráfico e interacción económica [in Spanish]. Antofagasta, Universidad del
Norte.
Wheeler, J. C.
1984 On the Origin and Early Development of Camelid Pastoralism in the Andes.
In J. Clutton-Brock and C. Grigson (eds.), Animals and Archaeology. Volume 3:
Early Herders and Their Flocks. British Archaeological Reports International Series
S202. Oxford, Archaeopress: pp. 395–410.
Wheeler, J. C., E. Pires-Ferreira and P. Kaulicke
1976 Preceramic Animal Utilization in the Central Peruvian Andes. Science 194:
pp. 483–490.
Ch a p t e r 8
C olonization,
Structured L andscapes
and Seasonal Mobility
An Examination of Early Paleo-
Eskimo Land-Use Patterns in the
Eastern Canadian Arctic
S. Brooke Milne1
W hen discussing N ew World archaeology, the word colonization

conjures images of Paleo-Indian sites and Clovis points because these
were the first peoples and lithics to enter the North American continent from
an Asian/Beringian origin, some 12,000 years ago. Several thousand years
later, a second and apparently unrelated pioneering population followed the
1.
Many thanks to Aubrey Cannon and Chris Ellis for reading earlier drafts of this
chapter and for offering many helpful comments and suggestions. I am grateful to
Chris Ellis for his encouragement over the last few years to undertake such comparative
analyses of Paleo-Eskimo and Paleo-Indian material culture. His insights and exper-
tise on Paleo-Indian lithics helped me to formulate several of the ideas presented in
this chapter. Thank you also to Lisa Hodgetts for providing another critical northern
perspective on Paleo-Eskimo culture and to Robert Park for allowing me to repro-
duce the map illustrated in Figure 8.1. This chapter was written during my tenure as
a postdoctoral Research Fellow in the Department of Anthropology at the University
of Western Ontario. Support for this fellowship is generously provided by the Social
Sciences and Humanities Research Council of Canada Post-Doctoral Fellowship
Program. Funding for my doctoral research, on which part of this chapter is based,
was provided by the Social Sciences and Humanities Research Council of Canada
Doctoral Fellowship Program, the Association for Universities for Northern Studies
Studentship Awards Program, the Northern Scientific Training Program and the
McMaster University School of Graduate Studies. It was inspiring to meet, at the
Cotsen Institute of Archaeology at UCLA, where this chapter was first presented, with
colleagues studying such a unique diversity of cultures, from different periods, centering
on issues of human mobility.
174
E a r l y P a l e o -E s k i m o L a n d -U s e P att e r n s      175
Figure 8.1. Map of the Canadian Arctic and Greenland, showing the route of the Paleo-Eskimos
(courtesy of Robert Park).
same entry route into the New World, but instead of moving southward, these
peoples came to occupy the previously uninhabited Arctic regions of Canada
and Greenland (Figure 8.1). These peoples are known archaeologically as the
Paleo-Eskimos.
This movement effectively represents a second discrete colonization of
people into a pristine environment in the New World. Despite some obvious
similarities between the Paleo-Indians and Paleo-Eskimos in terms of culture
origins, small population size, lithic technology and mobility, compara-
tively little consideration has been given to these early Arctic pioneers when
addressing universal archaeological questions of human colonization and
adaptation in unfamiliar environments (Ellis 1998). Given the contributions
that research on the Paleo-Eskimos can make in addressing questions on these
topics, it seems fitting to raise the profile of this culture in the wider field of
archaeology. That said, I am in no way implying we presently know all there is
to know about these peoples. We can state that the Paleo-Eskimos thrived in
the eastern Arctic for roughly 3000 years. We cannot say, though, exactly how
they did survive in what initially would have been unfamiliar surroundings.
Moreover, how did the Paleo-Eskimos stay in contact with one another given
that the average group size is estimated to have been between 10–30 people at
any given time of the year (Maxwell 1985:98)? Some archaeologists have argued
that to maintain basic biological viability in a population, group size needs to be
176    S . B r o o k e M i l n e
larger than 25 (for instance Kelly 2003:51). If it is not, frequent and repeated
contact with outside groups must be established and maintained, no matter
the distance to be traveled (Wilmsen 1973, 1974; Wobst 1974; Mandryk 1993;
MacDonald 1998, 1999; MacDonald and Hewitt 1999). With this in mind,
social factors take on a more significant role than they have been previously
given, especially among small-scale nomadic peoples like the Paleo-Eskimos.
In this chapter I consider how colonization, lithic raw material procurement,
social contact and landscape learning articulated over time to shape Paleo-
Eskimo land-use patterns, particularly those on southern Baffin Island.
To situate this discussion, I begin with an overview of Paleo-Eskimo culture
and current interpretations for what sparked their eastward migration from
Alaska. To highlight some of the flaws with these cultural characterizations,
I draw on Rockman’s (2003:4–7) concepts of environmental knowledge and
landscape learning. I argue that the establishment of habitual land-use patterns
among the Paleo-Eskimos may have been more significantly influenced by
demands for lithic procurement and social contact than by subsistence needs.
The Paleo-Eskimo migration into the eastern Arctic was fairly rapid, and
one of the driving forces structuring this movement may well have been the
motivation to secure an adequate and reliable tool-stone supply. As all Paleo-
Eskimo groups would have shared similar locational priorities on entering
this pristine landscape, a coincidental opportunity for social interaction likely
resulted at these tool-stone sources since there would exist a need to acquire
locational knowledge regarding their distribution in an unfamiliar landscape.
As nomadic peoples, the Paleo-Eskimos did not exist at the mercy of their
surroundings, and interpretations of these peoples living a tenuous existence
must be reconsidered. As Mandryk (2003:xiii) notes, hunger, cold or adven-
ture can no longer be thought of as the principal driving forces behind human
colonization into new lands.
Paleo-Eskimo Culture
Because of poor preservation conditions created by climatic fluctuations, annual
freeze-thaw cycles, and moist acidic soils, organic materials generally do not
survive in Paleo-Eskimo sites. Consequently, much of what we know about the
Paleo-Eskimo culture derives from its lithic remains. The stone tools made
and used by these peoples are very distinctive. Typological and morphological
similarities, particularly among burins, microblades, microcores and endblades,
have led to widespread speculation among Arctic archaeologists that the Paleo-
Eskimos originated somewhere in the Old World, most likely Siberia (Giddings
1967). Perhaps the most striking characteristic of Paleo-Eskimo stone tools is
their small size (Figure 8.2).
Figure 8.2. Examples of Paleo-Eskimo lithic artifacts illustrating their tiny size. This assemblage is
from the Tungatsivvik site (Frobisher Bay, southern Baffin Island) and includes (A) informal tools; (B)
burin spalls; (C) microblade fragments; (D) burins; (E) bifaces and biface fragments; (F) scrapers; and
(G) core fragments.
Among the Inuit, oral histories claim that the makers of these tools must
have been dwarfs precisely because of their small size (Maxwell 1985:40).
Several isolated finds, however, indicate that these tiny lithics were mounted
into larger hafts, handles and shafts. Given the value of these organic imple-
ments in an environment where such materials are often difficult to come by
in large pieces, they were rarely discarded with the lithics, thus leaving the
impression that only small hands could have used the isolated stone parts.
Irving (1957) coined the name Arctic Small Tool tradition (ASTt) and applied
it to all Paleo-Eskimo sites from Alaska to Greenland where these tiny stone
tools were found. He believed, based on this trait, that they all belonged to a
single parent culture.
178    S . B r o o k e M i l n e
When exactly the Paleo-Eskimos entered into the New World is a matter
of debate largely because of problems with radiocarbon dating Arctic sites.2
The point of entry, however, is generally accepted as having been on the north
shore of Alaska. McGhee (1996:73), Maxwell (1985:37, 39) and Schledermann
(1996:39) all speculate that this movement began when several small bands
of people crossed the Bering Strait from the nearby Chukchi Peninsula into
Alaska via the Seward Peninsula some 5000 years ago. Soon after crossing
into Alaska, these people began moving into the eastern regions. Earlier
immigrants into the New World, the Paleo-Indians, had not previously colo-
nized the eastern Arctic because the area remained glaciated until roughly
6500 years ago, making it uninhabitable (Maxwell 1985:37; Schledermann
1996:15). Once the ice retreated, human expansion occurred rapidly, with
full colonization from Alaska to Greenland being estimated at having been
accomplished in 500 years or less (McGhee 1996:73). There remains the
question of why the Paleo-Eskimos moved eastward to colonize the last
uninhabited region in the New World. As Maxwell (1985:45–47) notes, there
is no definitive evidence to explain the impetus for this movement, but there
are several theories.
First, it is thought that with the retreat of the glacial ice sheets there
occurred a biotic explosion in which plants as well as marine and terrestrial
animals thrived in this new environment. Human hunters would have been
attracted to these abundant subsistence resources, which would have been
easy to hunt, having never seen human beings before. Second, population
pressure in Alaska is proposed as a catalyst for colonization. To date, however,
there is no direct evidence indicating that population numbers had exceeded
the carrying capacity of the local Alaskan environment. Third, it is thought
that intergroup tensions or disputes may have resulted in band fissioning, in
which those peoples who were shunned or cast out left in search of a new
‘homeland’ (McGhee 1996:74). Last, the rapid colonization of the eastern
Arctic has been closely tied to musk ox hunting. Steensby (1917) proposed
that the Paleo-Eskimos followed eastward migrating herds of musk ox from
the western subarctic Barrenlands to Greenland. Because of the defensive
2.
The debate in Arctic archaeology surrounding the use of radiocarbon dates deriving
from sea mammal fat and bone has spanned several decades (McGhee and Tuck 1976;
Arundale 1981; Park 1994). Because of problems with the marine reservoir effect, dates
obtained from marine mammal samples are considered less reliable than those obtained
from terrestrial sources. Even though efforts have been made to devise correction factors
for these marine dates, most researchers still consider them suspect, opting instead for
dates obtained from terrestrial species.
tactics of musk oxen, which form a circle or line to protect the herd, all of
the animals could be killed at once. Over time this would invariably put pres-
sure on local populations because of their slow reproductive rates (McGhee
1996:55). Therefore, the Paleo-Eskimos would have had to move further and
further eastward in search of new herds. Evidence from sites in Greenland,
where middens have been found indicating a focused exploitation of musk ox,
seems to support this theory. The ethology of musk ox, and their potential for
population collapse, suggests that a dependence on this animal as a primary
resource would have been extremely risky. Because of this, the musk ox theory
remains questionable.
Whatever the catalyst was for the colonization, we know that it was rapid.
A plausible explanation for the speed of this process is that the environment,
although uninhabited, was not drastically different from the one the Paleo-
Eskimos left in Siberia and Alaska. In other words, the Paleo-Eskimos would
have been pre-adapted to living in the eastern Arctic environment, having
already established sophisticated marine and terrestrial hunting strategies in
their place of origin. Despite the probability that the Paleo-Eskimos were pre-
adapted to this pristine environment, archaeological interpretations of these
peoples’ way of life tend to reinforce an idea that Paleo-Eskimo existence was
frequently tenuous or ‘stressed’ (Maxwell 1997:206). They lived year-round
in skin tents and snow houses (Ramsden and Murray 1995), enduring winter
temperatures of below -40°C. To heat these structures, it is thought that the
Paleo-Eskimos burned animal bones and fat in open fires, contained in box
hearths, rather than using soapstone lamps (Maxwell 1984:361; Schledermann
1996:8–9). Soapstone lamps are far more efficient as sources of heat and light
than open fires are, but few of these items have been found in archaeological
sites, leading to speculation that they were not widely used by these earliest
Arctic inhabitants (Maxwell 1984:361). Existence in these dwellings would not
only be cold but also unpleasantly smoky.
Paleo-Eskimo camps are interpreted as ephemeral given the low density
of material remains that characterizes many of them, suggesting that their
inhabitants were constantly on the move, most likely in pursuit of animal
resources (Andreasen 2000:91; Nagy 2000:144–145). The winter darkness
would complicate efforts to hunt, particularly in the highest Arctic regions.
Knuth (1967) proposed that to mitigate the long winter months of darkness
and isolation, these pioneers may have passed the winter in a kind of semihi-
bernation, during which time little activity of any kind occurred. Presumably,
the Paleo-Eskimos cached food outside their skin tents, gathered as much
fuel, in the form of musk ox bones, as possible, and crawled underneath
heavy skins to sleep the winter darkness away (McGhee 1996:64–65). What
reflects the insecurity of this kind of existence are the population estimates
180    S . B r o o k e M i l n e
for this culture.3 McGhee (1996:65) states that in the High Arctic, there
may have been only 200–300 people scattered over one million km2. If these
same densities apply to the Low Arctic, then the total Paleo-Eskimo popula-
tion may have merely comprised 500–1000 people throughout this massive
region.4 When these people were dispersed in winter camps of 10–15 indi-
viduals, the loss of any one person, particularly a hunter, could spell disaster
for all (Schledermann 1996:101). Moreover, with numbers this small, the
loss of a single regional group could seriously hinder the biological viability
of the entire population (Park 2000:201).
Environmental Knowledge
and Landscape Lear ning
In the Arctic it is easy to explain away certain patterns of human behavior in
the archaeological record, given the sharp seasonal changes experienced in
this environment and their effects on the ecosystem. But the Paleo-Eskimos
would not have survived long if this was their sole reality. To link these peoples’
existence so intrinsically to seasonal shifts in the environment and resource base
undermines their ability to think, act and adapt. It denies them agency as human
beings and makes them appear as automatons that merely wandered about the
landscape, weather permitting, in search of food (Wobst 2000:40). Rockman
(2003) recently proposed that to truly understand the process of human colo-
nization in pristine environments, archaeologists must consider how existing
knowledge (from a place of origin) and landscape learning (the knowledge that
is acquired in a new location) shaped and reinforced the behavior of pioneering
populations in unfamiliar surroundings. This kind of analysis facilitates a
3.
Population estimates for the Paleo-Eskimos are highly speculative because large
parts of the Arctic landscape remain unexplored. This makes it difficult to devise
accurate assessments of Paleo-Eskimo occupation densities. Despite these difficulties,
ethnographic accounts by Boas (1964:18, see table for individual settlement numbers)
for the southern Baffin Inuit indicate that the overall population numbers for the entire
region did not exceed several hundred. Settlement size ranged from 20–82 individuals,
and these numbers were wholly dependent on the seasonal availability of subsistence
resources. It is acknowledged that the Paleo-Eskimo population was smaller than that
of the Inuit (Maxwell 1985). Therefore, it is not unreasonable to speculate that a need
for social interaction would have been of critical importance to these earliest peoples
both on a cultural level and on a purely biological level (Park 2000:201).
4.
The Low Arctic is loosely defined as the region south of the Parry Channel. The
areas north of the channel are considered High Arctic as this is where the conditions
that define the Arctic are at their most extreme (McGhee 1996:44–45).
greater appreciation for how past populations knew and used their environ-
ment, and it works to dispel the image of a perilous existence.
Rockman (2003:4–7) acknowledges that “knowing the environment can
mean many things,” and to simplify discussion, she outlines three basic types
of knowledge. Locational knowledge includes information relating to the spatial
and physical characteristics of particular resources. It also includes the ability to
relocate such resources after their discovery. Locational knowledge is consid-
ered the easiest form of information to acquire. Limitational knowledge refers to
familiarity with the usefulness and reliability of various resources, including the
combination of multiple resources into a working environment. Development
of limitational knowledge depends on the periodicity of the given resource and
its intended use. Social knowledge is the collection of social experiences that
serves as a means of transforming the environment or a collection of natural
resources into a human landscape.
Two important factors that must be considered when discussing the process
of landscape learning include the means by which knowledge is acquired
and the time it takes to acquire that knowledge (Rockman 2003:12). For
the Paleo-Eskimos we know that the process of landscape learning occurred
rapidly. The way this knowledge was acquired, however, has not been explic-
itly considered by Arctic archaeologists. Rockman (2003:13–19) proposes
four approaches through which knowledge and landscape learning can be
incorporated into archaeological investigations. For this chapter I draw on
the Resource Modeling Approach, which focuses on the ability of a colonizing
population to take information acquired in its original environment and
effectively apply it in a new area. The success of transferring this knowledge
depends wholly on “the similarity of necessary resources in terms of location
and distribution, the limitations in terms of carrying capacity, and the social
organization required to access them” (Rockman 2003:18).
In the Arctic the range of subsistence resources is comparatively limited
in terms of terrestrial (such as caribou, musk ox, hare, fox, birds) and marine
mammals (such as seal, polar bear, walrus, whale). The Paleo-Eskimos would
most likely have encountered the same range of species in the eastern Arctic
as they would have in Siberia and Alaska. Accordingly, established hunting
techniques and knowledge of animal ethology would have been transferable
from their old environment to this new one. The social organization required
to hunt and process these resources would have been equally applicable.
Archaeologists studying Paleo-Indian culture (among which Kelly and Todd
1988; Meltzer 1995; Amick 1996; Kelly 1996) have, over the decades, built
on Mason’s (1962:243–246) seminal statement that subsistence systems based
on the exploitation of large fauna are transferable across long distances and
that this transferability would invariably have facilitated the expansion of the
182    S . B r o o k e M i l n e
Paleo-Indians throughout the North American continent. This transference of

hunting strategies would appear to apply in the Arctic for the Paleo-Eskimos as
well. It should be noted that contingent situations, where local faunal resources
experienced periodic crashes, would occur. As the same situations would have
happened previously, the Paleo-Eskimos likely had existing limitational knowl-
edge on how to deal with them. In effect, the Paleo-Eskimos would have been
moving through an environment where the seasonal availability and types of
resources would have been similar.
This model posits that information related to nonorganic resources, such as
lithic raw materials, may be the least transferable from an old environment to
one that is newly colonized (Rockman 2003:19). Lithic raw materials should
be considered an exploitable resource in the same sense as plants and animals
(Rick 1978:4; Ellis 1984:12; Bamforth 1986:40; Daniel 2001:261; Beck et al.
2002:482). Lithic source areas are not mobile, they are easily manipulated when
encountered, and they can be exploited repeatedly once they have been iden-
tified on the landscape (Nelson 1991:77). Source areas, however, are directly
affected by seasonal conditions where ground cover, such as ice and snow,
will reduce their availability (Rolland 1981; Kuhn 1991; Wenzel and Shelley
2001). Lithic raw materials are also not always evenly distributed across the
landscape, and their quality can vary greatly among source areas (Andrefsky
1994a, 1994b). With these factors in mind, Paleo-Eskimo toolmakers would
have needed to quickly locate tool-stone sources and to assess and access them
directly. Consequently, the need to acquire locational and limitational knowl-
edge on lithic source areas in the eastern Arctic would be critical, particularly
since the geological distribution of this material resource would be entirely
unknown to the Paleo-Eskimos in this pristine environment.
In cases where lithic source areas are highly variable in distribution, quality
and seasonal availability, individuals will make every effort to acquire high
quality tool stone even if it means having to travel considerable distances to do
so (Andrefsky 1994a; MacDonald 1998, 1999). To find lithic sources in a new
environment would require exploration, and given the logistical challenges
involved in moving a large group of people across the landscape, it would not
make sense for the entire group to participate on such journeys. We might
expect social organization to change to accommodate long-distance raw mate-
rial procurement (Binford 2001:465–466). Those individuals most capable of
such journeys would include the young and physically fit. These are also the
persons who would be looking for prospective mates in other distant groups.
Mobility, Social Inter action

and Lithic Procurement

For decades archaeologists studying small-scale nomadic hunter-gatherer cultures
have speculated on how these groups survived given their low population densi-
ties and the massive territories they came to occupy (Wobst 1974; Kelly and
Todd 1988; Mandryk 1993). Wilmsen (1973, 1974) proposed that among such
small-scale societies archaeologists should expect to find a reliance on long-
distance mating networks to maintain basic biological viability. MacDonald
(1998:228–230) recently revisited this issue and, building in part on Wilmsen’s
(1974:118–119) ideas, he applied this proposition to Folsom Paleo-Indian sites
located in the northern and central plains. He found, by tracing the distribution
of exotic tool stone throughout these areas, that these Paleo-Indian people not
only used mobility as a critical strategy for finding mates but that, in the process
of doing so, they were also able to maintain strong social ties with otherwise
distant groups, all of whom were living in a recently colonized region. Early land-
use patterns among the Folsom Paleo-Indians appear to have been structured
in large part by the demands for social contact. And the apparent rendez vous
spots were at or near distinct lithic source areas that were, on average, 330 km
away (MacDonald 1999:152). In the process of learning the landscape, the Paleo-
Indians in this region acquired locational and limitational knowledge relating to
tool-stone distribution. Over time this translated into specific social knowledge
in which the local environment, or a collection of resources (such as lithics),
was transformed into a human landscape where socially determined patterns of
activity (such as finding a mate) occurred within and amongst the inhabitants
(Rockman 2003:6). Drawing on this example for comparative purposes, I believe
that Paleo-Eskimo colonization and the establishment of habitual land-use
patterns in the eastern Arctic could have occurred under similar conditions.
In the extreme north, procurement of lithic raw materials is complicated not
only by geological patchiness but also by pronounced seasonal restrictions on
accessibility. On southeastern Baffin Island, Maxwell (1973:10–11) notes that
sources of chert and other stones used in lithic tool production are scarce and that
the available sources consist of small, weathered pebbles on the ocean floor. These
pebbles can only be obtained during the Arctic warm season, when the shore-
fast ice is gone and when the tide recedes to expose the ocean floor. Patchiness
and accessibility would have added to the challenges of finding tool stone in this
new environment. Stone would have been a highly valued resource, and it is
certain that the Paleo-Eskimos were able to find sufficient source locations. Once
identified, a source area was likely used repeatedly simply because it was known,
not necessarily because it was the best (Kelly 2003:51). With this in mind, all
Paleo-Eskimos in the eastern Arctic would have been exploring the landscape in
search of tool stone and, in the process of doing so, may have encountered other
184    S . B r o o k e M i l n e
people exploiting the same source areas. According to MacDonald (1999:152),

Folsom Paleo-Indians traveled 160–500 km in search of tool-stone sources.
Considering that the population density of Paleo-Eskimos in the eastern Arctic
was probably far lower than the estimated figure of 1 person/1000 km2 for the
Folsom Paleo-Indians (MacDonald 1999:154), the former would have had an
even greater need for external social contact than the latter.
It is highly unlikely that it would have taken the Paleo-Eskimos a long
time to learn the locations of tool-stone source areas. This ties back to
Rockman’s (2003) concepts of locational and limitational knowledge. As I have
mentioned, locational knowledge is considered the easiest form of informa-
tion to acquire because it can be gathered rapidly in a matter of days, weeks
or months (Rockman 2003:4). Limitational knowledge relating to lithic raw
materials would not be that different in a new environment because periodicity
would be directly linked to the same seasonal restrictions relating to ground
cover that would have been previously experienced in Siberia and Alaska.
Moreover, the intended use of tool stone, no matter where it was procured,
would be the same, since lithics were such an integral component of the Paleo-
Eskimo technological inventory. Rockman (2003:4) suggests that it would
take roughly a generation to become familiar with such resources in a new
environment in terms of “their fluctuations, their potentials, and their carrying
capacity.” Widely dispersed groups of Paleo-Eskimos would soon locate lithic
source areas throughout the eastern Arctic and, in so doing, would invariably
encounter other groups or at least see evidence of their existence in the region.
This may have provided incentives for these widely dispersed groups to make
contact with one another, which in turn may have provided the basis for them
to establish long-distance social networks where technological needs would be
satisfied, reproductive interests secured, and culture continuity maintained.
Framing the process of Paleo-Eskimo colonization using concepts of
knowledge and learning dispels the idea that life for these peoples was perilous.
Instead, it acknowledges their unique adaptation to a polar environment and
their ability to act on the basis of informed decisions. This approach also looks
at how factors other than those tied to subsistence influenced people’s behavior
and interaction. But how did these patterns change once the Paleo-Eskimos
became established throughout the eastern Arctic? Generally speaking, the next
phase of colonization involves increasing regionalization, where adaptation
takes on a more local or regional focus (Speiss et al. 1998; Mandryk 2003:xiv).
Land-use patterns among Paleo-Eskimo sites on southern Baffin Island indi-
cate, however, that these peoples continued to make long-distance journeys to
acquire tool stone (Figure 8.3), and it appears that social factors were a major
motivation to maintain these long-established patterns.
Figure 8.3. Map of southern Baffin Island, in the Canadian Arctic.
The View from Souther n Baffin Island

On southern Baffin Island, tool stone is geologically and seasonally restricted
(Maxwell 1973, 1985; Odess 1998). In the coastal regions chert occurs in the
form of weathered pebbles that can be found only when the shore-fast ice is
gone and the tides recede, exposing the ocean floor and coastal headlands.
This material is of poor quality and the size of the pebbles restricts the kinds
of tools that can be made from them. Fortunately, the interior of the island has
a rich and reliable supply of stone fit for tool making in the form of secondary
deposits of chert left by retreating glaciers that scoured the inland plains
(Stenton 1991; Milne and Donnelly 2004; Milne 2005a). These secondary
deposits consist of nodules of varying sizes and are highly variable in quality,
color and texture. Amadjuak Lake, one of the large lakes found in the interior,
186    S . B r o o k e M i l n e
is considered an especially important location for acquiring chert. Amadjuak is

an English corruption of the Inuktitut word amaaq or angmalik. Amaaq means
‘chert,’ and ‘amaaq lake’ means, loosely translated, ‘the place chert comes from’
(Stenton and Park 1998:25). This attests to the known presence of chert in the
interior, and recent archaeological investigations in this area further underscore
its abundance and accessibility (Milne 2005a, 2005b).
The most ideal time to acquire stone for tool making is during the warm
season when the ground is not covered by ice and snow. Moreover, seasonal
food resources like nesting waterfowl and caribou are readily available and can
be reliably procured in large quantities with ease (Milne and Donnelly 2004).
This would have permitted toolmakers to focus their attention on renewing
their tool kits and procuring sufficient raw material supplies to take back to
the coast to meet their technological needs for the remainder of the year.
Settlement-pattern analyses on southern Baffin Island indicate that the Paleo-
Eskimos, like Inuit, spent the winter on the sea ice and outer coastal regions,
where they hunted resident populations of ringed seals (Maxwell 1973, 1985;
Stenton 1989; Milne 2003a). The distances between the coast and the interior,
combined with the challenges of winter travel, would have curtailed frequent
trips inland to get tool stone during the winter. Furthermore, procuring
terrestrial subsistence resources at this time of year is difficult because migra-
tory species have long since departed for warmer areas. Boas (1964:22) notes
that in the past some Inuit have wintered in the interior; however, they only
did so for a single season because securing an adequate food supply appeared
tenuous. Based on these seasonal factors, procurement of lithic raw material
undoubtedly was a planned event during the warm season, and inland travels
to get it would have been scheduled accordingly (Milne 2003a, 2005a).
The interior of southern Baffin Island is easily reached from every coastal
location via the coastal uplands (Stenton 1989:112). There are major river
systems draining from each of the interior lakes out to the coast. As Kelly
(2003:48) notes, when foragers are in an unfamiliar area, they rely heavily on
prominent landmarks to navigate their way. Following river systems is perhaps
the simplest way to explore the unknown “because if one goes upstream on the
way out, one simply has to go downstream to return home” (Kelly 2003:48).
Parties of explorers sent out to look for lithic source areas could easily find their
way to the interior of Baffin Island, and the abundance of local chert there, by
following the prominent inland river systems that connect the interior lakes to
the outer coastal regions. If distant Paleo-Eskimos in the Cumberland Sound,
Hudson Strait and Frobisher Bay regions all ventured inland at the same time,
they would most certainly find one another. Ethnographic records document the
importance of the interior of southern Baffin Island as a place where the Inuit
traveled during the Arctic warm season (Bilby 1923; Soper 1928; Boas 1964).
The southern Baffin Inuit occupy several widely separated coastal districts.
During the winter these people experienced a pronounced degree of isolation.
Given the centrality of the interior region, it appears to have served as a focal
point for seasonal interactions on an interregional level (Stenton 1989:335).
Although Inuit populations were larger in the eastern Arctic during the contact
period than those estimated for the Paleo-Eskimos, the Inuit were still widely
distributed over a large geographic region. Therefore, these warm season
interactions were equally critical for maintaining biological viability because
during these meetings prospective mates were given opportunities to interact,
and group alliances were renewed and strengthened (Stenton 1989:119). It is
easy to imagine that the Paleo-Eskimos would have used the interior region
for similar social purposes in addition to lithic acquisition, and five inland
sites on southern Baffin Island provide archaeological evidence to support
this idea. These five sites are variously located along the shores of Nettilling
and Mingo Lakes and the Mingo River, which connects Mingo and Amadjuak
Lake, and all of them have produced radiocarbon dates associated with the
early Paleo-Eskimo period (Milne, in review). These sites are known as Sandy
Point, Mosquito Ridge, LdFa-1, LdFa-12, and LeDx-42 (Stenton 1989; Milne
2003a, 2005a, 2005b, in press, in review; Milne et al., in review).
Sandy Point and Mosquito Ridge are located 10 km from one another on the
western shore of Burwash Bay, which forms the southern margin of Nettilling
Lake (Figure 8.4). Sandy Point is a single-component Paleo-Eskimo site and
was excavated by Stenton in 1985. At the time, the site was being negatively
impacted by forces of mechanical erosion. Consequently, it is not possible to
determine how much of the site was disturbed or how much information may
have been lost as a result. Of those artifacts that were recovered, 1176 are lithic
debitage and 101 are informal and formal tools. Mosquito Ridge is a large
multicomponent site occupied by the Paleo-Eskimos and the later Thule and
Inuit cultures. Partial excavation of the Paleo-Eskimo component at Mosquito
Ridge yielded a lithic assemblage of more than 20,000 artifacts, of which 97%
is lithic debitage (Milne 2003a).
LdFa-1 and LdFa-12 are located 1 km from one another on the northwestern
shore of Mingo Lake, which is approximately 250 km south of Nettilling Lake
(Figure 8.5). LdFa-1 is a large site containing at least 23 well-defined tent-
ring structures, suggesting this location was repeatedly occupied over time.
Few Paleo-Eskimo sites on Baffin Island contain this many intact structural
features, making this site unusual. Because the topography of southern Baffin
Island is characterized by large bedrock outcrops, finding suitable camping
spots near favorable hunting locations can be difficult. As a result, more recent
populations tend to settle on top of existing site remains, thereby obscuring
these earlier occupations. Older tent-ring structures are also occasionally
188    S . B r o o k e M i l n e
Figure 8.4. Map of Burwash Bay, part of the southern littoral of Nettling Lake on Baffin Island, showing
the locations of Sandy Point (LlDv-10) and Mosquito Ridge (MaDV-11) (after Stenton 1989).
dismantled since more recent peoples scavenge perimeter rocks to build new
dwellings (Milne 2003b).
Limited testing at LdFa-1 yielded large quantities of lithic and faunal
material, which is almost exclusively caribou. Caribou are concentrated in the
interior lakes region during the summer and early autumn, and they represent a
rich and reliable food source in this area. The incredibly dense concentration of
lithic debitage at LdFa-1 made for very slow digging. A total of 13.25 m2 were
excavated, yielding a lithic assemblage of 18,743 artifacts, 99.7% of which is
Figure 8.5. Map of the northwest shore of Mingo Lake, showing the locations of LdFa-1 and
LdFa-12.
debitage. LdFa-12 is a small single-component Paleo-Eskimo site and it has no

visible surface features. The site consists entirely of a 115-meter linear scatter
of lithics. LdFa-12 was found serendipitously during a field survey. A well-
worn caribou path runs through the center of the site and numerous artifacts
and flakes were visible on the surface as a result. While ephemeral sites like
LdFa-12 are more difficult to locate given their extremely low archaeological
visibility on the Arctic tundra, postdepositional processes, like caribou tram-
pling and frost heave, churn up the surface vegetation to expose underlying
deposits. In some instances Paleo-Eskimo deposits are found by accident during
the excavation of more recent sites. Despite the ephemeral nature of LdFa-12
and the limited number of units tested (about 5 m2), it also yielded a compara-
tively large lithic assemblage of 3743 artifacts, of which 99.5% is debitage.
Again, caribou dominate the recovered faunal remains, further indicating the
importance of these animals as a subsistence resource in the area.
Last, LeDx-42 is another large Paleo-Eskimo site located on the shores
of the Mingo River, approximately 10 km from LdFa-1 and LdFa-12 (Figure
8.6). Like LdFa-12, numerous caribou paths run through the central portion
of LeDx-42, and scattered along them were diagnostic lithic artifacts indicating
a Paleo-Eskimo cultural affiliation. There is only one possible dwelling struc-
ture visible at the site; however, it has yet to be investigated. Extensive testing
yielded another large lithic and faunal assemblage, containing caribou remains.
The lithics comprise 29,648 flakes, and 87 informal and formal tools.
My analysis of the Sandy Point and Mosquito Ridge lithics was conducted as
part of a larger study to assess whether the Paleo-Eskimos followed a seasonal
round, similar to that recorded for the southern Baffin Inuit, where winters
190    S . B r o o k e M i l n e
Figure 8.6. Map of the Mingo River, where it drains into the southwest shores of Amadjuak Lake,
showing the location of LeDx-42.
were spent in the coastal regions and summers were spent in the interior (Milne
2003a). More than 24,500 lithic debitage and tool artifacts from these two
inland sites and four southern Baffin coastal sites were compared. Patterns of
tool reduction and use isolated among the inland sites indicate that they were
principally used for raw material acquisition. The broad functional interpreta-
tion of the Sandy Point site is that it was used for raw material testing, early
stage core reduction, and the limited production of tool preforms and blanks
(Milne 2005a). These activities would have enabled task groups exploring
the area for tool stone to assess its quality and abundance and, in the process,
acquire locational and limitational knowledge. By comparison, Mosquito
Ridge was used as a lithic acquisition site; however, there is no evidence of
raw material testing, just intense early stage reduction (Milne and Donnelly
2004; Milne 2005a).
Stenton (1989) believes that Mosquito Ridge was a focal place in the inte-
rior region where later Neo-Eskimo populations would meet and interact
for social purposes. I believe that the site had the same significance for the
Paleo-Eskimos, in addition to raw material procurement. Mosquito Ridge
was occupied very early on in the Paleo-Eskimo occupation of the eastern
Arctic given the site’s radiocarbon date of 3800 ± 40 BP (Milne and Donnelly
2004:96–97). This date makes Mosquito Ridge one of the oldest Paleo-
Eskimo occupations on Baffin Island (Milne and Donnelly 2004:97). While
raw-material acquisition may have been the primary incentive leading to the
initial occupation of Mosquito Ridge during the colonization of the island,
its centrality in the interior and its significance as a meeting place would have
further increased its attraction as a place for continuous use. Mosquito Ridge
is located beside the Great Plain of the Koukdjuak, one of the largest nesting
grounds for snow geese in the world. The faunal assemblage recovered from
Mosquito Ridge indicates that the site occupants were intensively hunting snow
geese during the Arctic warm season, most likely during the bird’s annual molt
in early to mid-July. This subsistence resource would have been easy to acquire
with little planning or energy investment, and it would have enabled the Paleo-
Eskimos to focus more of their time on socializing and tool production rather
than hunting (Milne and Donnelly 2004).
Investigations of LdFa-1, LdFa-12, and LeDx-42 were undertaken in 2004.
Analysis of the recovered lithic materials is ongoing, but preliminary results do
complement those from Sandy Point and Mosquito Ridge, further attesting to
the importance of the island’s interior as a place for warm season aggregations
by the Paleo-Eskimos. Lithic raw-material acquisition was also the principal
focus at these three sites. This is most clearly illustrated using flake-to-tool
ratios (Milne and Donnelly 2004; Milne 2005a, in press). Flake-to-tool ratios
are simple measures used to examine the extent to which lost tool utility at
a site was being replaced through tool production activities (Ricklis and Cox
1993:450–451). Ratios are calculated here using all complete and fragmented
burins, microblades, cores, bifaces, scrapers, retouched informal tools and burin
spall tools. Unretouched flake tools, burin spalls, and bifacial edges were not
included since they lack postdetachment modification and, therefore, do not
contribute any by-products to the debitage assemblage.
Ratios for LdFa-1, LdFa-12, and LeDx-42 are extremely high (Table 8.1),
indicating that tool utility was being replaced by intense tool-stone reduction.
Although the ratios for Sandy Point and Mosquito Ridge are low to moderate,
the reduction strategies identified through the formal debitage analysis do
indicate that acquisition was the focus. In other words, the same reduction
and use strategies were employed at all five sites. An obvious reason for these
lower figures at Sandy Point and Mosquito Ridge can be attributed to the
192    S . B r o o k e M i l n e
Table 8.1. Flake-to-Tool Ratios Identified for Five Inland Paleo-Eskimo Sites on Southern Baffin
Island:
Site Flake/tool ratio Number of artifacts Intensity of reduction

MaDv-11 40:1 19,800 flakes | 495 tools Moderate
LlDv-10 14:1 1176 flakes | 84 tools Low
LdFa-12 207:1 3725 flakes | 18 tools High
LdFa-1 406:1 18,691 flakes | 46 tools High
LeDx-42 593:1 29,648 flakes | 50 tools High
fact that LdFa-1 and LeDx-42 are extremely large sites that were intermit-
tently occupied for millennia, resulting in larger deposits of reduction debris.
Moreover, it is impossible to know how much of the assemblage at Sandy
Point was lost as a result of site disturbance. Given the obvious absence of
late stage finishing flakes at all five sites, it appears that formal tools were not
being made in the interior from start to finish. Rather, reduction strategies
isolated in the assemblages indicate that site activities were focused on early
and middle-stage reduction, which is more commonly associated with the
production of preforms and blanks. Because there are so few finished and
intact artifacts in these assemblages, it appears that those implements that were
being roughed out were taken away from the site for completion elsewhere.
These patterns are entirely consistent with those identified in my earlier study
(Milne 2003a), further supporting the inference that the Paleo-Eskimos were
traveling inland to procure raw material during the Arctic warm season and
then transporting back to the coast sufficient supplies to last them throughout
the remainder of the year.
The presence of well-preserved faunal assemblages from LdFa-1, LdFa-12,
and LeDx-42 also provides seasonality data in the form of caribou tooth thin
sections, indicating these sites were also occupied during the warm season.
Last, radiocarbon dates from these three sites further support that seasonal
travels to the interior were established early on in the Paleo-Eskimo occupa-
tion of southern Baffin Island (Table 8.2). Even more interesting is the fact that
additional dates from LdFa-1 and LeDx-42 indicate that these two sites were
continuously occupied throughout the entire Paleo-Eskimo period, spanning
at least 2400 years (Milne, in review). These dates are important because they
coincide with a marked period of climate change experienced in the Arctic
beginning around 2600 BP, which impacted the local environment. Winters
became longer and harsher, and the extent and duration of the sea ice increased.
In response to this the Paleo-Eskimos adopted a more maritime-oriented way
of life, focusing on hunting sea mammals, mostly seals. As a result, it can be
inferred that the importance of the terrestrial ecosystem decreased in signifi-
cance to these peoples’ way of life and that they spent more of the year in the
Table 8.2. Radiocarbon Dates for Five Inland Paleo-Eskimo Sites on Southern Baffin Island:
Site Material Date BP Designation

MaDv-11 Caribou bone 3800±40 Early Paleo-Eskimo
LlDv-10 Peat 2785±115 Early Paleo-Eskimo
LdFa-12 Caribou bone 3670±40 Early Paleo-Eskimo
LdFa-1 Caribou bone 3530±40 Early Paleo-Eskimo
Caribou bone 3490±40 Early Paleo-Eskimo
Caribou bone 3210±40 Early Paleo-Eskimo
Caribou bone 1230±40 Late Paleo-Eskimo
LeDx-42 Caribou bone 3480±40 Early Paleo-Eskimo
Caribou bone 2320±40 Middle Paleo-Eskmo
outer coastal regions. Therefore, finding occupations dating to the middle

and late periods in the inland proper of southern Baffin Island is somewhat
unexpected. Certainly, the Paleo-Eskimos would continue to need tool stone
for their technological needs, but this could be acquired in the coastal regions
given the presence of chert pebbles there. The fact that these people continued
to travel inland and that they reoccupied the same sites for 2400 years strongly
suggests that other factors influenced their seasonal settlement patterns. I
believe that social factors and the need for group interaction were among the
driving forces that maintained these long-established interior land-use patterns.
The Paleo-Eskimos were familiar with this interior landscape and the locations
of these specific sites, which were ideal for raw material acquisition and in close
proximity to ideal caribou and waterfowl hunting grounds.
The coastal areas of Baffin Island do have accessible sources of chert,
although these are in the form of weathered pebbles of inferior quality, and
there is also an abundant supply of food resources in these locations throughout
the Arctic warm season (Jacobs and Stenton 1985; Stenton 1989). Therefore,
the Paleo-Eskimos did not have to keep making these journeys inland purely
for subsistence and material needs, particularly during the later periods given
the arduousness of adapting to an increasingly colder environment. They
would seemingly need to do so for social purposes, however, since the interior
remains the most central place to meet. This strongly suggests that a pattern
of seasonal mobility and land use, which was established very early on in the
Paleo-Eskimo colonization of southern Baffin Island, soon became a habitual
part of this culture’s seasonal round, in spite of local food and material abun-
dances in the coastal regions. The Paleo-Eskimo cultural landscape on southern
194    S . B r o o k e M i l n e
Baffin Island became highly structured by this unique confluence of geese,

caribou, stone and social interaction in the interior. This early adaptation to
an unfamiliar landscape appears to have helped secure the reproductive success
of this population and its cultural continuity over the millennia. Moreover, an
abundance of food and people during the warm season would have created a
very relaxed atmosphere for social interaction (Milne and Donnelly 2004:108).
This stands in sharp contrast to the image of Paleo-Eskimo life being balanced
between a constant search for food and a state of semihibernation.
Discussion
The Paleo-Eskimos were the first peoples to occupy one of the most arduous
environments in the world. Given their knowledge of the Siberian and Alaskan
landscapes, these peoples were essentially pre-adapted for their eastward migra-
tion into the pristine eastern Arctic. In this chapter I have argued that the
Paleo-Eskimos transferred their existing locational and limitational knowledge
from their place of origin to this new environment, which enabled them to
successfully colonize it in a short period of time. Moreover, I suggest that the
need to find suitable tool-stone sources and to establish social contact were
equally if not more important than subsistence in shaping the early exploration
of the eastern Arctic and the establishment of subsequent land-use patterns.
Tool-stone sources are fixed on the landscape and seasonally restricted in
their accessibility. All groups of Paleo-Eskimos would have required this
resource to maintain their tool kits. Traveling distances of up to 500 km to
access good-quality lithic raw material is not unheard of among Paleo-Indian
groups (MacDonald 1999). It is therefore not unreasonable to speculate that
Paleo-Eskimos made similar journeys in search of this resource. Once reli-
able source areas were found, it is highly probable that other distant groups
would have been encountered at these locations. This would provide the basis
to establish long-distance social and reproductive networks in this massive
geographic expanse. Evidence from sites located on southern Baffin Island
further indicates that these early land-use patterns were maintained over time,
and a significant motivation to do so was a need for social interaction. By
shifting my interpretive perspective away from a traditional environmentally
deterministic paradigm focused on climate and subsistence, I have presented
here a very different interpretation of the Paleo-Eskimo lifestyle from that
which pervades the literature. It is not necessary to reduce early hunter-
gatherers to the state of automatons to learn about their archaeological past.
They knew their environments, and they moved across the landscape making
knowledgeable decisions.
References
Amick, D.
1996 Regional Patterns of Folsom Mobility and Land Use in the American
Southwest. World Archaeology 27: 3, pp. 411–426.
Andreasen, C.
2000 Paleo-Eskimos in Northwest and Northeast Greenland. In M. Appelt, J.
Berglund and H. C. Gullov (eds.), Identities and Culture Contacts in the Arctic.
Copenhagen, Danish Polar Centre: pp. 82–96.
Andrefsky, W.
1994a Raw Material Availability and the Organization of Technology. American
Antiquity 59: 1, pp. 21–34.
1994b The Geological Occurrence of Lithic Material and Stone Tool Production
Strategies. Geoarchaeology 9: 5, pp. 375–391.
Arundale, W. H.
1981 Radiocarbon Dating in Eastern Arctic Archaeology: A Flexible Approach.
American Antiquity 46: pp. 244–271.
Bamforth, D. B.
1986 Technological Efficiency and Tool Curation. American Antiquity 51: 1, pp.
38–50.
Beck, C., A. K. Taylor, G. T. Jones, C. M. Fadem, C. R. Cook and S. A. Millward
2002 Rocks Are Heavy: Transport Costs and Paleoarchaic Quarry Behaviour in the
Great Basin. Journal of Anthropological Archaeology 21: pp. 481–507.
Bilby, J. W.
1923 Among Unknown Eskimo. London, J. B. Lippincott.
Binford, L. R.
2001 Constructing Frames of Reference: An Analytical Method for Archaeological Theory
Building Using Hunter-Gatherer and Environmental Data Sets. Los Angeles,
University of California Press.
Boas, F.
1964 The Central Eskimo. Lincoln, University of Nebraska Press (originally published
in 1888).
Daniel, I. R., Jr.
2001 Stone Raw Material Availability and Early Archaic Settlement in the Southern
United States. American Antiquity 66: 2, pp. 237–265.
Ellis, C. J.
1984 Paleo-Indian Lithic Technological Structure and Organization in the Lower Great
Lakes Area: A First Approximation. Simon Fraser University, Department of
Archaeology (PhD dissertation).
1998 The Fluted Point Tradition and the Arctic Small Tool Tradition: What’s the
Connection? Paper presented at the conference “On Being First: Cultural
196    S . B r o o k e M i l n e
Innovations and Environmental Consequences,” 31st Annual Chacmool

Conference, University of Calgary, November 1998.
Giddings, J. L.
1967 Ancient Men of the Arctic. New York, Alfred A. Knopf.
Irving, W. N.
1957 An Archaeological Survey of the Susitna Valley. Anthropological Papers of the
University of Alaska 6: 1, pp. 37–52.
Jacobs, J. and D. Stenton
1985 Environment, Resource, and Prehistoric Settlement in Upper Frobisher Bay,
Baffin Island. Arctic Anthropology 22: 2, pp. 59–76.
Kelly, R.
1996 Ethnographic Analogy and Migration to the Western Hemisphere. In T.
Akazawa and E. J. Szathmary (eds.), Prehistoric Dispersals of Mongoloid Peoples.
Tokyo, Oxford University Press: pp. 228–240.
2003 Colonization of New Land by Hunter-Gatherers: Expectations and Impli
cations Based on Ethnographic Data. In M. Rockman and J. Steele (eds.),
Colonization of Unfamiliar Landscapes: The Archaeology of Adaptation. New York,
Kelly, R. and L. Todd
1988 Coming into the Country: Early Paleo-Indian Hunting and Mobility. American
Antiquity 3: 2, pp. 231–244.
Kuhn, S. L.
1991 “Unpacking Reduction”: Lithic Raw Material Economy in the Mousterian of
West-Central Italy. Journal of Anthropological Archaeology 10: pp. 76–106.
Knuth, E.
1967 The Ruins of Musk-Ox Way. Folk 8–9: pp. 191–219.
MacDonald, D. H.
1998 Subsistence, Sex, and Cultural Transmission in Folsom Culture. Journal of
1999 Modeling Folsom Mobility, Mating Strategies, and Technological Organization
in the Northern Plains. Plains Anthropologist 44: 168, pp. 141–161.
MacDonald, D. H. and B. S. Hewitt
1999 Reproductive Interests and Forager Mobility. Current Anthropology 40: 4, pp.
501–525.
McGhee, R.
1996 Ancient People of the Arctic. Vancouver, UBC Press.
McGhee, R. and J. Tuck
1976 Undating the Arctic. In M. Maxwell (ed.), Eastern Arctic Prehistory: Paleoeskimo
Problems. Memoirs of the Society for American Archaeology Number 31. Salt Lake
City, Society for American Archaeology: pp. 6–14.
Mandryk, C. A. S.
1993 Hunter-Gatherer Social Costs and the Nonviability of Submarginal Environ
ments. Journal of Anthropological Research 49: pp. 39–71.
2003 Foreword. In M. Rockman and J. Steele (eds.), Colonization of Unfamiliar
Landscapes: The Archaeology of Adaptation. New York, Routledge: pp. xiii–xv.
Mason, R. J.
1962 The Paleo-Indian Tradition in Eastern North America. Current Anthropology
3: pp. 227–278.
Maxwell, M.
1973 Archaeology of the Lake Harbour District, Baffin Island. Archaeological Survey of
Canada, Paper Number 6. Ottawa, National Museums of Canada.
1984 Pre-Dorset and Dorset Prehistory of Canada. In D. Damas (ed.), Handbook of
North American Indians. Volume 5: Arctic. Washington, Smithsonian Institution:
pp. 359–368.
1985 Prehistory of the Eastern Arctic. Orlando, Academic Press.
1997 The Canadian Arctic in Transition: Pre-Dorset to Dorset. In G. Gilberg and
H. C. Gulløv (eds.), Fifty Years of Arctic Research: Anthropological Studies from
Greenland to Siberia. Publications of the National Museum Ethnographical Series.
Volume 18. Copenhagen, National Museum of Denmark, Department of
Ethnography: pp. 205–208.
Meltzer, D.
1995 Clocking the First Americans. Annual Review of Anthropology 24: pp. 21–45.
Milne, S. B.
2003a Peopling the Pre-Dorset Past: A Multi-Scalar Study of Early Arctic Lithic Technology
and Seasonal Land Use Patterns on Southern Baffin Island. McMaster University,
Department of Anthropology (PhD dissertation).
2003b Identifying Pre-Dorset Structural Features on Southern Baffin Island:
Challenges and Considerations for Alternative Sampling Methods. Etudes/
Inuit/Studies 27: 1–2, pp. 67–90.
2005a Palaeo-Eskimo Novice Stone Knapping in the Eastern Canadian Arctic.
Journal of Field Archaeology 30: 3, pp. 329–345.
2005b Archaeological Investigations in the Mingo and Amadjuak Lake Districts of
Southern Baffin Island. Permit Report covering the Work conducted under Nunavut
Archaeologist Permit 04-06A. Manuscript on file with the Department of
Culture, Language, Elders, and Youth. Government of Nunavut. Igloolik,
Nunavut.
in press Landscape Learning and Lithic Technology: Seasonal Mobility, Enculturation,
and Tool Apprenticeship Among the Early Palaeo-Eskimos. In A. Cannon
(ed.), Structured Worlds: The Archaeology of Hunter-Gatherer Thought and Action.
London, Equinox.
198    S . B r o o k e M i l n e
in review Why Go Inland? Palaeo-Eskimo Terrestrial Occupations on Southern Baffin

Island. Current Anthropology (submitted 27 Nov. 2006).
Milne, S. B. and S. M. Donnelly
2004 Going to the Birds: Examining the Importance of Avian Resources to
Pre-Dorset Subsistence Strategies in the Interior of Southern Baffin Island.
Arctic Anthropology 41: 1, pp. 90–112.
Milne, S. B., L. Hodgetts and S. Timmermans
in review Pre-Dorset Foragers? New Insights on Pre-Dorset Subsistence Strategies
from the Interior of Southern Baffin Island. Arctic (submitted 3 Nov. 2006).
Nagy, M.
2000 From Pre-Dorset Foragers to Dorset Collectors: Paleo-Eskimo Culture
Change in Ivujivik, Eastern Canadian Arctic. In M. Appelt, J. Berglund and
H. C. Gullov (eds.), Identities and Culture Contacts in the Arctic. Copenhagen,
Danish Polar Centre: pp. 143–148.
Nelson, M. C.
1991 The Study of Technological Organization. In M. Schiffer (ed.), Archaeological
Method and Theory. Volume 3. Tucson, University of Arizona Press: pp.
57–100.
Odess, D.
1998 The Archaeology of Interaction: Views from Artifact Style and Material
Exchange in Dorset Society. American Antiquity 63: 3, pp. 417–435.
Park, R. W.
1994 Approaches to Dating the Thule Culture in the Eastern Arctic. Canadian
Journal of Archaeology 18: pp. 29–48.
2000 The Dorset-Thule Succession Revisited. In M. Appelt, J. Berglund and H.
Gullov (eds.), Identities and Cultural Contacts in the Arctic. Copenhagen, Danish
Polar Centre: pp. 192–205.
Ramsden, P. and M. Murray
1995 Identifying Seasonality in Pre-Dorset Structures in Back Bay, Prince of Wales
Island, NWT. Arctic Anthropology 32: 2, pp. 106–117.
Rick, J. W.
1978 Heat Altered Cherts of the Lower Illinois Valley: An Experimental Study in
Prehistoric Technology. Prehistoric Records Number 2. Evanston, Northwestern
University Archaeological Program.
Ricklis, R., and K. Cox
1993 Examining Lithic Technological Organization as a Dynamic Cultural
Subsystem: The Advantages of an Explicitly Spatial Approach. American
Antiquity 58: 3, pp. 444–461.
Rockman, M.
2003 Knowledge and Learning in the Archaeology of Colonization. In M. Rockman
and J. Steele (eds.), Colonization of Unfamiliar Landscapes: The Archaeology of
Adaptation. New York, Routledge: pp.3–24.
Rolland, N.
1981 The Interpretation of Middle Paleolithic Variability. Man 16: pp. 15–42.
Schledermann, P.
1996 Voices in Stone: A Personal Journey into the Arctic Past. Komatic Series Number 5.
Calgary, Arctic Institute of North America.
Soper, J. D.
1928 A Faunal Investigation of Southern Baffin Island. Canada Department of Mines
Bulletins Number 53. Ottawa, National Museum of Canada.
Speiss, A., D. Wilson and J. Bradley
1998 Paleo-Indian Occupation in the New England-Maritimes Region: Beyond
Cultural Ecology. Archaeology of Eastern North America 26: pp. 201–264.
Steensby, H. P.
1917 An Anthropogeographical Study of the Origin of the Eskimo Culture.
Meddelelser om Grönland 53: pp. 41–228.
Stenton, D. R.
1989 Terrestrial Adaptations of Neo-Eskimo Coastal-Marine Hunters on Southern Baffin
Island, NWT. University of Alberta, Department of Anthropology (PhD
dissertation).
1991 Caribou Population Dynamics and Thule Culture Adaptations on Southern
Baffin Island, NWT. Arctic Anthropology 28: 2, pp. 15–43.
Stenton, D. R., and R. W. Park
1998 Ancient Stone Tools of Nunavu: An Illustrated Guide. Ottawa, Parks Canada.
Wenzel, K. E. and P. H. Shelley
2001 What Put the Small in the Arctic Small Tool Tradition? Raw Material
Constraints on Lithic Technology at the Mosquito Lake Site, Alaska. In
W. Andrefsky (ed.), Lithic Debitage: Context, Form, Meaning. Salt Lake City,
University of Utah Press: pp. 106–123.
Wilmsen, E.
1973 Interaction, Spacing Behaviour, and the Organization of Hunting Bands.
Journal of Anthropological Research 29: 1, pp. 1–31.
1974 Lindenmeier: A Pleistocene Hunting Society. New York, Harper and Row.
Wobst, M.
1974 Boundary Conditions for Paleolithic Social Systems: A Simulation Approach.
American Antiquity 39: pp. 147–178.
2000 Agency in (spite of) Material Culture. In M.-A. Dobres and J. Robb (eds.),
Agency in Archaeology. New York, Routledge: pp. 40–50.
Ch a p t e r 9
T he Emergence of
Cultures of Mobility
in the A ltai Mountains
of Mongolia
Evidence from the Intersection of
Rock Art and Paleoenvironment
Esther Jacobson-Tepfer
T he appearance of North Asian pastoralism is usually located within the

Early to Middle Bronze Age (mid-second millennium BCE), while the
emergence of seminomadic pastoralism is associated with the transition from
the Bronze to Early Iron Age (early first millennium BCE).1 This generalized
dating is based on evidence from mortuary contexts within the Russian Altai
region, the Sayan Mountains of Tuva and the Minusinsk Basin to the north of
the Sayan region. That evidence includes the regular appearance of bones of
sheep, cattle and horses, apparently part of the funerary feast or intended to
be taken to the afterworld, and the regular appearance of ceramic vessels of a
shape and substance indicating use for the storing of meat and milk products.
What little has remained of worldly goods found in North Asian Bronze Age
burials indicates a concern for personal ornament and small items of household
utility. Otherwise, there are no clear reflections of mobility:2 no horse skel-
etons, no horse trappings associated with riding, and virtually no indication of
1.
See Khazanov 1994:19–21 for a provisional understanding of the term seminomadic
pastoralism within the Eurasian steppe. As he points out repeatedly, the concepts of
pastoralism and seminomadism vary from region to region and from period to period.
This is certainly the case that will be made here.
2.
Within this chapter the concept of mobility will refer in its most basic form to
the regular movement of communities to accompany their flocks. This mobility might
have been limited to movement up and down a single mountain or valley, or it might
have been truly extensive.
200
Mobility in the A l ta i M o u n ta i n s      201
wheeled vehicles as a part of household wealth.3 Within the same broad region
of the Altai-Sayan uplift and the Minusinsk Basin, mortuary evidence from
the Late Bronze and Early Iron Ages is abundant. This material indicates an
emerging mobility related to a dependency on horse riding. The great burial
of Arzhan I, dated to the early first millennium BCE, included the remains
of three hundred horses and enough bronze bridle bits to confirm the use of
horses for riding (Gryaznov 1980). Later burials at Tuekta in the Russian Altai,
dated to the 6th century BCE, had been severely plundered in antiquity, but an
abundance of bridle ornaments recovered from the burials attested not only
to horse dependency but also to the extent to which exigencies of mobility
increasingly influenced cultural aesthetics. This transformation of lifestyle and
aesthetic expression is fully demonstrated in the famous burials associated with
the Pazyryk Culture, including those excavated by Gryaznov, in 1950, and
Rudenko, in 1970, and those frozen burials more recently excavated on the
Ukok Plateau (Polos’mak 2001; Molodin et al. 2004). Some of these burials had
been partially plundered in antiquity and subsequently frozen. Surviving organic
matter indicated the lifestyle of people who were horse-dependent pastoralists
and whose way of life involved regular mobility. These burials are all dated
between the 5th and third centuries BCE (Alekseev et al. 2001; Slusarenko et
al. 2001). In addition to the actual bodies of horses intended to accompany the
dead, collapsible stools, small tables and felt horse blankets and carpets were
found, all indicating a pastoral and mobile way of life. Rich felted carpets and
saddle blankets reflect the central importance of sheep within the economy,
while the combination of furs and skins (Polos’mak and Barkova 2005) reveals
a continuing dependency on hunting integrated with herding. In the burials
of both elite and commoners (Kubarev 1986, 1991) were found wooden and
clay vessels of a rough and utilitarian manufacture. Such objects testify to the
fundamental need of these Early Iron Age people for portable material posses-
sions and for vessels intended to prepare and transport milk products and meat.
Many aspects of clothing and ornament confirm the centrality of horse riding
in the life of the Pazyryk Culture (Polos’mak and Barkova 2005).
We do not know the extent to which the pastoralists of the Early Middle
Bronze Age actually moved or, if they did, how they moved. Nor is there
sufficient evidence to determine the nature and temporality of their habitation
sites. We have a much better understanding of the extent to which the Pazyryk
3.
The one exception to this statement is a bronze yoke-shaped object found in a few
Karasuk burials and on deer stones (Vadetskaya 1986:Plate V, no. 28; Novgorodova
1989:159–160). Several scholars have suggested that this object was directly related
to the driving of chariots (Solov’ev 2001:Figures 50, 51), but others have associated it
with bows (Varenov 1984).
202    E s t h e r J a c o b s o n - T e p f e r
Culture engaged mobility. Despite all the signs of mobility embedded in their
burial ritual, their log and plank burial chambers took the form of wooden
dwellings much like those used by modern Altai dwellers for their winter habi-
tations.4 This would suggest that Early Iron Age herders had stable wintering
sites from where they moved their herds into higher pastures for the summer.
Even now in the Mongolian Altai, herders move their dwellings and animals
approximately four times a year, cycling from lower to higher pastures and
back down again. Similar patterns of movement, perhaps within the same long
valleys, most likely characterized the lives of those earlier mobile cultures. Thus
it is possible to describe the mobility of Late Bronze and Early Iron Age cultures
of the Altai region as horse-dependent and seminomadic or transhumant.
The information so relatively abundant for Bronze and Early Iron Age
cultures across a broad swath of South Siberia is virtually absent from the
Mongolian Altai. In this vast region jutting westward between Russia and China
(Figure 9.1), there have been no excavations until very recently and almost no
archaeological surveys relating to settlement sites.5 Nonetheless, within this
archaeological ‘silence,’ there is abundant evidence of the ancient presence of
Bronze and Iron Age cultures. Valleys and ridges are everywhere marked by
large and small altars, standing stones, burial mounds and the altars and stone
images of ancient Turkic populations (mid-first millennium CE). These surface
monuments indicate that the early cultures of mobility attested on the Russian
side of the Altai-Sayan uplift were, if anything, even more fully represented
on the Mongolian side.6 A long-range inventory of surface monuments in
mountainous Bayan Ölgiy, northwestern Mongolia, however, is revealing the
extensiveness of material remains directly relatable to Bronze and Iron Age
cultures of mobility.7 During the summer of 2006, excavations of three burials
within the high permafrost zone of Bayan Ölgiy demonstrated conclusively that
the seminomadic culture we refer to as Pazyryk was well represented within
the Mongolian Altai.8 These are, however, the only excavations yet undertaken
within this critical juncture of North and Central Asia.
4.
This is vividly indicated in reproductions and drawings by Rudenko (1970),
Polos’mak (2001), and Kubarev (1987, 1991).
5.
But see Derevyanko and Natsagdorj 1990 regarding surveys of Paleolithic sites in
northwestern Mongolia.
6.
The prevailing lack of information regarding that region has much to do with
questions of international politics and economy and need not detain us here.
7.
This is the Mongolian Altai Inventory with E. Jacobson-Tepfer, J. Meacham and
G. Tepfer as principal investigators.
8.
The joint Mongolian/German/Russian project was headed by D. Tseveendorj, H.
Parzinger and V. I. Molodin.
Figure 9.1. Map of the Altai Mountain Region, northern Central Asia (prepared by the Mongolian Altai
Inventory and Mapping Project).
Given our present state of knowledge of the relevant archaeology and

theoretical constructs for considering the emergence of mobility in North
Asia, we are still unable to answer a few simple questions. How did mobility
actually happen? Is the model we developed applicable to all regions within
the mountain-steppe of North Asia, or were there variations depending on the
specific physical setting? Within any inquiry into the emergence of cultures
of mobility there is one substantial resource that has hitherto been neglected
and that may be useful in responding to these questions. This is rock art or
petroglyphs, imagery pecked or engraved into the surfaces of bedrock or
boulders. Although the Russian Altai and Tuva have long been known for
their rock art sites, that material has rarely been used as documentation for
the emergence of pastoralism or horse-dependent seminomadic pastoralism.
Within the Mongolian Altai the identification and documentation of rock art
sites has occurred much more recently, in a more organized manner, and with a
greater concern for the preservation of this cultural resource. As a result, there
now exists a vast body of pictorial imagery that could significantly supplement
material derived from burials. Moreover, because this material, executed by
individual artists perhaps in moments of solitude, disrupts cultural conven-

tions reaffirmed within the mortuary context, it holds the promise of reflecting
nuances within cultures of mobility in the ancient Altai (Jacobson 2002). Its
use, however, depends on some understanding of its dating; and that dating, in
turn, is directly related to the reconstruction of the region’s paleoenvironment
and its cultural implications.
The larger climatological context of the North Asian steppe and mountain
steppe zone is well established. During the late Pleistocene the climate of
North Asia was cold and dry and the landscape was characterized by harsh
xeric vegetation (Guthrie 1982). Around 10,000 BP (calibrated), increasing
humidity and moderately rising temperatures marked the beginning of the
early Holocene. Climate amelioration continued into the middle Holocene, a
period characterized by the domination of mesic vegetation and the extension
of boreal forests. Analyses of radiocarbon-dated pollen and diatom records
from Khoton Nuur, the largest lake within the region we are considering,
indicate that before 9000 BP (calibrated) this part of Mongolia was charac-
terized by dry steppe vegetation and the absence of boreal conifers. Between
9000 and 4000 BP (calibrated), wetter conditions supported the development
of a substantial forest dominated by spruce (Picea), pine (Pinus), larch (Larix)
and birch (Betula) (Tarasov et al. 2000). Pollen data from Achit Nuur, a large
lake on the border between the Bayan Ölgiy and Uvs aimags,9 indicate that the
conditions prevailing around Khoton Nuur and one of the sites to be discussed
here (Aral Tolgoi) extended across the present-day mountain steppe of north-
west Mongolia (Gunin et al. 1999), including two other valleys we will be
considering, the Upper Tsagaan Gol and the Oigor drainage.10 The conditions
prevailing through the middle Holocene appear to have dominated until about
4600 BP (calibrated). From then until around 4000 BP, steppe vegetation domi-
nated by mugwort (Artemisia), weeds (a variety of Chenopodiaceae) and grasses
(mostly Poaceae) rapidly increased (Tarasov et al. 2000). Forests retreated
throughout the region, leading to the present environmental conditions by the
late second millennium BCE (Gunin et al. 1999:56). Within our focus region
forests were reduced to fragmentary stands on north-facing slopes or somewhat
more extensively on slopes north, south and west of Khoton Nuur.
The environmental changes associated with the onset and development
of the late Holocene coincided with two profound cultural transformations.
Pastoralism dependent on cattle, horses, sheep and goats entered North Asia,
probably from western Siberia, at some time beginning in the late third or
Mongolia is divided into 21 aimag (provinces), each subdivided into many ‘som.’
9.
These analyses are corroborated by paleoenvironmental studies of the Russian

10.
Altai in the Ulagan uplands (Blyakharchuk et al. 2004).

early second millennium BCE (Vadetskaya 1986; Anthony 1998; Alekseev

et al. 2001). Over the rest of the second millennium, pastoralism based on
cattle, as well as on small animals, gradually became established within the
Minusinsk Basin and in areas farther south and southwest. Wheeled vehicles
were evidently adopted in the same period, but the impact of that innovation on
mobility remains unclear.11 The second major cultural development, the advent
of horse riding, seems also to have been brought to North Asia during the
Late Bronze Age, although it was known earlier in western Siberia (Anthony
1998:101).12 One may argue that the apparent coincidence of the advent of
horse riding with a greater dependency on large animal husbandry gave rise to
several strategic modifications in the lifestyles of steppe and mountain steppe
cultures. Horse riding would have permitted the control of considerably larger
flocks than would be possible by a herder on foot. By facilitating human move-
ment up to higher pastures, the advent of horse riding would have encouraged
the development of more mixed herds, at the same time economizing the
energy required of a single herder or small social unit. Sheep and goats grazing
their way over slopes could be guarded by a single shepherd on foot. Larger
animals, such as cattle, yak, horses and camels, could be driven out to graze on
their own. Horse riding allowed the herder to ride out and bring in his larger
animals when necessary, even from high pastures, thus extending his animals’
access to good pasture. At the same time, the limited carrying capacity of
grassland, affected by the drier cooler environmental conditions and impacted
by larger herds, must have necessitated an increasingly frequent change of
pasture. This, in turn, was also facilitated by the new horse dependency. In
other words, climate change, innovative practices and technologies relating to
the use of large animals and mobility, and related cultural adaptations to both
negative and positive environmental factors functioned together to support the
emergence of seminomadic pastoralism at the end of the Bronze Age.13
Any discussion of the dating of rock art requires, in advance, a number of
qualifications. First, at this time there is no direct way to date petroglyphic
rock art at the spatial and numerical scale presented by sites in northwestern
11.
The evidence for the appearance of wheeled vehicles is primarily visual. Images of
carts appear on stone slabs associated with the Okunev Culture of the Minusinsk Basin
and in association with images of cattle. Novgorodova (1989:173–201) has argued that
the spread of wheeled vehicles in North Asia must be associated with the late Bronze
Age Karasuk Culture.
12.
The exact date of the emergence of horse riding remains a subject of debate.
13.
Such cultural adaptations would have involved the development of the portable
dwelling (gher), portable furnishings, clothing appropriate for riding, and new forms
of food preparation and preservation.
Mongolia.14 Second, with only a few exceptions, the rock art of the Altai
region reflects the real world rather than a world of spirits; or, to be more
precise, if the spirit realm was the subject of some of the art we will consider,
it was rendered in terms of the real world of knowable animals and activities.15
Moreover, the pictorial traditions that have been documented in all the major
sites of the Mongolian Altai are replicated in sites throughout the Russian
Altai and reflected, also, in Bronze Age art of the Altai-Sayan uplift and in
Iron Age art of eastern Kazakhstan; it is thus unnecessary to resort to argu-
ments of memory to explain similar subjects and styles across a variety of sites.
Shared cultural values and references appear to have supported shared visual
and mythic traditions. Finally, given the pictorial basis of Altai rock art in the
experience of the real world, any consideration of its dating must acknowledge
the constraints imposed by the paleoenvironment.
Dating rock art necessitates a constant tacking back and forth between
subject, style, means of execution, environmental constraints, and the informa-
tion derived from excavation archaeology. Although a thorough discussion of
such dating methodology is beyond the scope of this chapter,16 we can briefly
consider how this interreferencing would work. Images of animals that we
know disappeared by the end of the Pleistocene establish a period certain for
the earliest representations. In the case of the Mongolian Altai, those animals
would be megafauna, such as mammoths and rhinoceros, and birds such as
ostriches; all clear indicators of a dry, harsh environment and of a Paleolithic
pictorial culture. Elk (Cervus elaphus sibiricus) depend on the existence of
ecotones of open grassland and protective forest. Wild horses and cattle
(aurochs, Bos bos) presume the existence of integrated extensive steppe, forested
steppe and riparian zones. The Altai bear (Ursus ursus) requires high mountain
meadows, riparian zones and dense forests, while moose (Alces alces) require
the extensive browse of riparian zones and broken forest. Thus many of the
images we have documented can be inserted into a chronology established by
14.
All reliable dating methodologies for the study of rock art refer to pictographs
rather than petroglyphs. There have been many attempts to develop reliable dating
methods for pecked and engraved art but without results applicable to large-scale
studies (Dorn 2001).
15.
The single exception is offered by a horned, faceless anthropomorphic figure
recorded at a number of sites in the Mongolian Altai (Jacobson et al. 2001:II, Plates
77, 106, 207).
16.
For a more complete discussion of the dating of rock art of the Mongolian Altai
see Jacobson 2005 and Jacobson-Tepfer et al. 2007. For a concordance of paleoenviron-
ment, fauna, cultural changes and rock art styles over a period of approximately 12,000
years see Jacobson et al. 2001: I, Chart 1.
environmental factors. Imagery of mammoths, rhinoceros and ostriches reflect

a late Paleolithic culture; the style in which these animals were rendered and
their outline execution by means of heavy stone implements can be used to
propose a date for other, like imagery of individual or overlaid animals. Those
early images also offer an understanding of the nature of early visual expression
in terms of single, static animals or in terms of overlaid but psychologically
unrelated animals. By contrast, the closer that the imagery of animals more
dependent on forested zones takes us to and through the Bronze Age, the more
significant are indications of psychological interaction. There is a clear corre-
spondence of human representations within apparently narrative settings and
a method of execution that reflects finer tools, either stone or metal. Within
compositions that by subject, execution and style can be dated to the Bronze
Age, we frequently see a considered exploitation of the rock surface in terms
of narrative space.
Representations of weaponry offer abundant clues to dating. Excavation
archaeology indicates that Bronze Age hunters used a variety of longbows,
spears, clubs and bronze daggers and knives (Solov’ev 2001). This weapon set
is confirmed by rock art, which also indicates that they carried a kind of tufted
stick, probably used for hunting small animals (Jacobson et al. 2001: II, Plates
XXVIII, 256).17 This element disappears entirely with the advent of scenes
of riding. The adoption of horse riding carried in its wake a different set of
weapons. Bows necessarily became short; to compensate for that change in size,
their shape was recurved, thus offering greater thrust to the archer (Solov’ev
2001:Figure 27). This new bow, a particular kind of quiver (gorytus) with which
it became associated (Jacobson et al. 2001: II, Plates 99, 224, 369), daggers,
knives and a particular kind of short sword (akinakes) have been amply attested
in burials of the Late Bronze and Early Iron Age. Another kind of object
looking like a hanging ball appears in rock art imagery together with figures
carrying recurved bows and wearing the belted tunic and close-fitting pants of
a rider (Jacobson et al. 2001: II, Plate 333). Closely related to the appearance
of the recurved bow is a particular kind of stylized deer (Jacobson et al. 2001:
II, Plates XI, 141). The combination of these elements (bow, gorytus and deer)
is indicated in the great Mongolian tradition of standing stones known as ‘deer
stones,’ dated to the transition period of the Late Bronze–Early Iron (Savinov
1994; Volkov 2002). When this image appears within Altai rock art sites, we
can be assured that we are looking at imagery no earlier than the Late Bronze
Age and no later than the Early Iron Age (Jacobson-Tepfer 2001).
This object is replicated today in a small tufted rod worn by Altai hunters of
17.
marmots. They claim that by twirling it in front of the animals, they are able to distract
and immobilize them.
In the course of a multi-year project in northwestern Mongolia, the Joint

Mongolian/American/Russian Project recorded three highly significant rock
art complexes.18 Two of these are among the largest in North Asia, and all
three complexes include materials extending over several thousand years. The
documentation of North Asian cultures preserved in these materials represents
a continuous record from the late Paleolithic, earlier than 9000 BCE, through
the Turkic Period (6th–8th centuries CE). Even though all three complexes are
in the same general region in northwest Mongolia, it is impossible to ignore
significant differences in their topography and in their pictorial records. These
differences allow us to consider which visual representations in association with
reconstructed paleoenvironments can clarify the conditions for the transition
from a hunting to a pastoral dependency and from a pastoral to a mounted
seminomadic pastoralism.
Location of the Complexes

The three complexes in question are all found within the Altai Mountains of
Bayan Ölgiy aimag in northwestern Mongolia (Figure 9.1). The southernmost
complex is located at Aral Tolgoi, a small hill at the far western end of Khoton
Nuur, near the border of Mongolia and northern China. Aral Tolgoi lies
within the floodplain of the rivers Tsagaan Uss and Khar Salaagiin Gol before
they empty into Khoton Nuur (Figure 9.2). To the north the narrow valley is
bordered by forested slopes leading up to the rocky ridge of a high moraine,
to the northwest and west by rocky, inhospitable mountains, and to the south-
west by mountains still partially covered by thick forest, the remnant of a far
larger forested region in the early and middle Holocene (Gunin et al. 1999).
The elevation at the base of Aral Tolgoi is approximately 2000 m. The best
grassland in this area, indeed the only extensive grassland at the head of this
valley, is that provided by the floodplain itself, from the lake back to the wall
of mountains on the west. At present the valley is inhabited permanently only
by members of a Mongolian border patrol station, but it is visited during the
summer months by vacationing herders. Even now, neither the forested area
nor the limited pasture would support the herding of large flocks year-round.
Presumably, this constraint would have been considerably greater during the
The Joint Mongolian/American/Russian Project, Altay, functioned between

18.
1994 and 2004. Its purpose was to investigate the ecology of ancient cultures in
the Mongolian Altai. Principal investigators in the project included D. Tseveendorj
(Institute of Archaeology, Mongolian Academy of Sciences), E. Jacobson (University
of Oregon) and V. D. Kubarev (Institute of Archaeology and Ethnography, Siberian
Branch, Russian Academy of Sciences).
Figure 9.2. View of Aral Tolgoi from mountains to the south (photograph by Gary Tepfer).
Middle and Late Bronze Age, when the forests around Khoton Nuur were
more extensive and the climate somewhat wetter.
The second complex is that of the Upper Tsagaan Gol, located approximately
45 km north of Aral Tolgoi (Figure 9.1). The complex measures approximately
30 km from east to west, including the uppermost end of the river Tsagaan Gol
and, to the west, the valleys of its two principal tributaries, the Tsagaan Salaa
and the Khar Salaa. Between these rivers rises the sacred mountain, Shiveet
Khairkhan, oriented approximately west to east. To the north and south the
valley is hemmed in by steep slopes; to the west it is blocked by the glaciated
peaks of Tavan Bogd at the intersection of Mongolia, Russia and China. Farther
east the Tsagaan Gol descends to its confluence with the largest river in this
mountainous region, the Khovd Gol.
Although some ravines within the Upper Tsagaan Gol complex facilitate
movement up to the rich pastures on higher elevations, most of the slopes
defining the valley are precipitous and characterized by regular rock falls.
Indeed, most are passable only by the wild goats that live in this high valley and
by their principal predators: wolves and snow leopards. Moreover, the slopes
hemming the valley along the whole complex offer little or no real shelter from
the cold winds funneled down the Khar Salaa and Tsagaan Salaa. The eleva-
tion of the upper Khar Salaa is approximately 2400 m, the summit of Shiveet
Khairkhan is 3320 m, and the highest summit of Tavan Bogd reaches 4374
m. Larch, willow and a few Siberian pine trees can still be found in scattered
patches in the upper valley, usually on those slopes so isolated or precipitous
that they discourage the wood cutting that has decimated the relic forests in
upper valleys throughout the Mongolian Altai. Nonetheless, the rock-pecked
imagery in the Upper Tsagaan Gol complex, including representations of elk,
bear, wild cattle and even moose, indicates that these forests were more exten-
sive in the past, going back to the Early Middle Bronze Age.
About 30 km north of the Upper Tsagaan Gol is the third complex, Tsagaan
Salaa/Baga Oigor (Figure 9.1). This complex extends along the middle flow of
the Baga Oigor and its tributary, a small river also named Tsagaan Salaa. The
complex stretches approximately 25 km along the valleys, for several hundred
meters up south- and east-facing slopes and into deep ravines on the north side
of the valley. The valley is broad and open, water is abundant, and grassland
is extensive and continuous from the valley floor up over easily negotiated
slopes to broad, high pastures (Figure 9.3). Even today the valley includes
small ponds with abundant fish and waterfowl. In the not-too-distant past
this marshland was much more extensive, spreading in some places to either
Figure 9.3. View looking over the Baga Oigor River up the Tsagaan Salaa River (Ulan Khus som, Bayan
Ölgiy aimag). In the distance are the rocky slopes of Tsagaan Salaa I, II and IV; in the upper center is
Tsagaan Salaa III (photograph by Gary Tepfer).
side of the wide Baga Oigor Valley.19 The valley is now entirely treeless, but a
petroglyphic record rich in the imagery of forest-dependent animals indicates
that at an earlier period tree cover must have been relatively extensive along
the rivers and up the slopes. The elevation of the valley at the confluence of
the Tsagaan Salaa and Baga Oigor is approximately 2300 m. Both rivers flow
from drainages that begin at the ridge crowned, ultimately, by Tavan Bogd.
The most negotiable of the passes over the Tavan Bogd ridge is that of Ulaan
Daba at the head of the Oigor river drainage. In contrast to the valleys of Aral
Tolgoi and the upper Tsagaan Gol, those of the Tsagaan Salaa and Baga Oigor
offer relatively easy routes over the high Altai ridge into the extensive grass-
land and forests of present-day south Russia, northern China and northeastern
Kazakhstan. While the climate is just as harsh as in the valleys of the other two
complexes, the physical character of Tsagaan Salaa/Baga Oigor offers much
more protection from the weather that comes down from the west and north-
west. Deep ravines on both the north and south sides of the large Baga Oigor
valley promise shelter from the cold winds, rain and snow in all seasons.
A clear similarity between the subject matter and style of the rock art of the
two largest Mongolian Altai complexes, the Upper Tsagaan Gol and Tsagaan
Salaa/Baga Oigor, as well as that of recorded sites in the Russian Altai,20
indicates that Bronze Age and Early Iron Age cultures regularly crossed the
mountainous ridge that now forms the international border between Mongolia,
China and Russia. In addition, it may be more than accidental that Aral Tolgoi
is only about 62 km from the site with the oldest imagery in the Russian Altai,
that of Kalgut on the Ukok Plateau (Molodin and Cheremisin 1998). The three
Mongolian Altai complexes and their valleys are all in the same mountainous
region. Their primary differences lie in a number of specific physical charac-
teristics that would have influenced the way the valleys might have been used
by early pastoralists. Ancient migration patterns would have been affected by
the proximity of each valley to the high Altai ridge, by the ease of access to
and over that ridge, and by passes that could be negotiated by both humans
and animals. One must assume that the human- and animal-carrying load of
the valleys would have been affected by both the valley width and the existence
of natural shelter from wind and weather. The availability of water and the
extensiveness of pastureland along the valley floor and on accessible slopes
would have qualified the desirability of any valley as a place of temporary or
19.
This is evident in the original air photographs on which the Landscape Model I
of Tsagaan Salaa/Baga Oigor was based (Jacobson et al. 2001: II, 17).
20.
These include, most importantly, Kalbak-Tash, Cheganka, Yelangash and Irbistu,
all in the Chuya river drainage (Kubarev and Jacobson 1996; Jacobson and Kubarev
2003).
more permanent residence. Finally, one must assume that access to forested
areas, with their promise of ample game and wood, would have rendered some
valleys more desirable than others for the purposes of hunting and obtaining
timber for dwellings and burial structures. In other words, these three valleys,
each with its own pictorial record of early cultures, would have offered different
opportunities to peoples of the pre-Bronze, Bronze and Early Iron Ages.
The Rock Art at Ar al Tolgoi

In terms of physical size and number of rock outcrops, Aral Tolgoi is the
smallest of our three complexes. It is also the southernmost site and the only
one still surrounded by a substantial relic forest, now dominated by pine and
larch. The images at Aral Tolgoi are all pecked on the bedrock exposed on the
spine of this whale-shaped hill. These surfaces are exceedingly damaged as a
result of time, weather and the severity of glacial events in the late Pleistocene.
Of the 27 outcroppings on which we have recorded image making (Tseveendorj
et al. 2005), approximately seven are so covered with lichen, or are so badly
crumbling, that it has been almost impossible to make out the subjects of
the imagery. One of the most damaged areas is at the very top of the hill.
Interestingly, here are found what appear to be some of the oldest images.
These images include large ostrichlike birds (Figure 9.4), massive aurochs (wild
cattle), bear, horses and large horned wild sheep. The most significant image is
that of a wooly rhinoceros (Figure 9.5) preserved on a fragment of crumbling
bedrock. More certainly than any other, this image indicates that the rock art
record at Aral Tolgoi began no later than the late Pleistocene and well before
forest had become dominant over dry steppe.
Most of the other carved surfaces of Aral Tolgoi are occupied by images of
aurochs, horses and elk (Figure 9.6), but there are also a significant number
of argali (Ovis ammon) and wild goats. Birds other than ostriches are almost
entirely absent. Representations of humans are relatively crude, and the
contexts in which they occur refer to hunting with heavy clubs, thus indicating
a pre–Bronze Age date (Tseveendorj et al. 2005:Figures 41, 60). The animals
are almost always treated in a style that might be described as monumental
realism. Their representation in terms of static profiles, the frequent appear-
ance of overlay, and the regular indication of the use of heavy stones for pecking
reinforce a general date for the Aral Tolgoi imagery in the late Pleistocene and
early Holocene. There are only a few images datable to the Bronze Age. One
shows archers with longbows (Tseveendorj et al. 2005:Figure 65). There is
one image of a rudimentary cart (Tseveendorj et al. 2005:Figure 46) and one
pair of stylized deer of the Early Nomadic Period (Tseveendorj et al. 2005:
Figure 66). The single horseman dates from the Turkic Period (Tseveendorj et
Figure 9.4. Aral Tolgoi (Tsengel’ som, Bayan Ölgiy aimag): pecked images of large birds (ostriches?) and
a horse, late Pleistocene (photograph by Gary Tepfer).
Figure 9.5. Aral Tolgoi (Tsengel’ som, Bayan Ölgiy aimag): pecked image of a wooly rhinoceros, late
Pleistocene (photograph by Gary Tepfer).
Figure 9.6. Aral Tolgoi (Tsengel’ som, Bayan Ölgiy aimag): pecked image of an elk (Cervus elaphus), late
Pleistocene or early Holocene (photograph by Gary Tepfer).
al. 2005:Figure 170). On the terrace immediately south of the hill and to the
west, however, are a number of ritual structures and burials indicative of Early
Iron Age and Turkic Period dates.21
Rock Art in the Upper Tsaga an Gol Complex

The rock art in the Upper Tsagaan Gol complex and its state of preservation
differ considerably from that at Aral Tolgoi. We have recorded imagery on
more than 2400 surfaces, and all of these are sufficiently clear that one may
identify either what the imagery represents or what it once represented.22 Only
21.
A number of artifacts indicative of Paleolithic and Neolithic workmanship have
been found in the floodplain around the lower Khar Salaagiin Gol.
22.
In the Upper Tsagaan Gol complex, as well as at Tsagaan Salaa/Baga Oigor, many
images have ‘fallen out’: the surface, crushed in the process of pecking the contour or
silhouette, was weathered and weakened and was subsequently dislodged by freezing
and thawing over time. In such cases there remains what might be called a ‘ghost image,’
clear enough, in most cases, to recognize the subject of the former representation
(Jacobson-Tepfer et al. 2007).
Figure 9.7. Upper Tsagaan Gol complex (Tsengel’ som, Bayan Ölgiy aimag): image of an elk, from the
pre–Bronze Age, overlaid by a smaller Bronze Age elk (photograph by Gary Tepfer).
a few representations, including aurochs, elk and argali, suggest execution in a

period earlier than the Bronze Age. One of these is an image of an elk executed
by rough, deeply pecked blows indicative of the use of a heavy stone instrument
(Figure 9.7). Its neck is overlaid by an animal pecked with a much finer instru-
ment in the Bronze Age, and its antlers have also been ‘finished’ (lengthened)
at a later date. Images even more archaic in style and execution are located
on other outcropping in the Shiveet Khairkhan section of the complex, and
a group of clearly archaic animals can be found in the far eastern end of the
complex. Although the mode of execution of many of these images (contours
pecked with a large, heavy object) and their style (static, in profile and monu-
mental) suggest a date as early as the imagery from Aral Tolgoi, the absence of
clearly indicative images (such as mammoths, rhinoceros or ostrich) makes it
impossible to be certain when pre–Bronze Age imagery first appeared in the
Upper Tsagaan Gol valley.
The vast majority of images in the Upper Tsagaan Gol complex can be dated
confidently to the Late Bronze Age, the Early Iron Age and the Turkic Period.
These representations include, for the earliest period, extensive scenes of hunts
for bear, elk and aurochs (Figure 9.8), compositions showing caravans in which
domesticated yak are heavily loaded with children and household goods, and
Figure 9.8. Upper Tsagaan Gol complex (Tsengel’ som, Bayan Ölgiy aimag): pecked scene of hunters
attacking an aurochs, Bronze Age (photograph by Gary Tepfer).
some ‘household’ scenes (scenes with humans and a variety of domesticated

animals; Jacobson-Tepfer et al. 2007: II, Plates 505–509, 529).
There are many images of wheeled vehicles with horses and drivers, as well
as of vehicles alone (Jacobson-Tepfer et al. 2007: II, Plates 361, 531, 651).
Several of these scenes indicate a hunting context (Jacobson-Tepfer et al. 2007:
II, Plates 556, 557), but none indicate the use of the vehicle for carrying loads
or for combat.23 Most can confidently be dated to the Bronze Age by style and
by the bows and quivers carried by the drivers. The beginning of horse riding
in the Bronze Age is well attested in the Upper Tsagaan Gol complex (Figure
9.9). Scenes of herding cattle and horses occur regularly within the complex
and on some occasions with considerable artistry (Jacobson-Tepfer et al. 2007:
II, Plate 451).
Imagery datable to the transitional Late Bronze–Early Iron Age period
includes hunts on foot and on horseback after elk, deer and ibex; some of
these scenes appear to reflect mythic traditions (Jacobson-Tepfer et al. 2007:
II, Plate 390). The image of a highly stylized elk, datable to the Late Bronze
This is in contrast to the imaginative reconstruction of a war chariot in the North

23.
Asian steppe by Solov’ev (2001:Figure 52).

Figure 9.9. Upper Tsagaan Gol complex (Tsengel’ som, Bayan Ölgiy aimag): detail from a larger scene of
a deer hunt, showing a pecked image of a rider preparing to shoot his bow, Late Bronze Age (photograph
by Gary Tepfer).
Age, appears more frequently here than in Tsagaan Salaa/Baga Oigor; increas-
ingly stylized formulations suggest that it continued until the last centuries
of the first millennium BCE (Jacobson 2000a, 2000b, 2001). The number of
representations of wheeled vehicles that can be reliably dated to the Early Iron
Age is perhaps two (Jacobson-Tepfer et al. 2007: II, Plates 151, 540). There
are no scenes of caravans that can be dated to that period with any certainty.
The Turkic Period is represented by scenes of riders and hunting, usually for
ibex (Jacobson-Tepfer et al. 2007: II, Plates 405–414).
Limited but systematic surveys in the upper Tsagaan Gol valley have allowed
us to identify at least two sources of finished Paleolithic tools and related
scatter. By contrast, we have not yet identified Neolithic stone artifacts. This,
in combination with the recorded rock art, allows us to reach the following
tentative conclusions. Artifacts would indicate that the upper valley saw some
human habitation, probably seasonal, in the late Pleistocene, but the record for
such habitation in the early to middle Holocene is intermittent and uncertain.
Beginning in the Middle Bronze Age and extending into the Early Iron Age,
habitation, either seasonal or year-round, became more regular. The apparent
lacuna in imagery between the late Pazyryk Period (5th to third centuries BCE)
and the Turkic Period may be misleading: the people inhabiting this valley at that
time may have been uninterested in image making or they may be represented by
the many surfaces for which the subject matter and style refer to an indeterminate
period in the Iron Age. The Turkic Period is vigorously represented by images
on a number of outcroppings as well as by a significant number of ritual sites
with carved stone figures, enclosures and other surface structures.
The Rock Art at Tsaga an Sala a /Baga Oigor

Of all three complexes, Tsagaan Salaa/Baga Oigor includes the deepest and
most continuous imagistic record, with essentially no cultural interruptions
from the late Paleolithic through the Turkic Period.
Four images of mammoths (Figure 9.10) ‘pin’ the beginning of repre-
sentational traditions here to a period no later than the late Pleistocene.
Representations of wild cattle (Figure 9.11), herds of elk (Figure 9.12), horses,
and a few of bear can be dated to the early and middle Holocene. They conjure
up an ancient environment in which riparian zones supported thickets of willow
Figure 9.10. Tsagaan Salaa/Baga Oigor complex (Ulan Khus som, Bayan Ölgiy aimag): drawing of the
pecked image of a mammoth, late Pleistocene (BO III; Jacobson et al. 2001: I, Figure 912; drawing by
V. D. Kubarev and D. Tseveendorj).
Figure 9.11. Tsagaan Salaa/Baga Oigor complex (Ulan Khus som, Bayan Ölgiy aimag): drawing of the
pecked image of an aurochs, late Pleistocene (BO III; Jacobson et al. 2001: I, Figure 368; drawing by V.
D. Kubarev and D. Tseveendorj).
Figure 9.12. Tsagaan Salaa/Baga Oigor complex (Ulan Khus som, Bayan Ölgiy aimag): drawing of
overlaid images of elk and other animals, early Holocene and Bronze Age (TS IV; Jacobson et al. 2001:
I, Figure 625; drawing by V. D. Kubarev and D. Tseveendorj).
and other leafy vegetation, while the slopes offered forested shelter. Bronze Age
imagery from this complex includes many scenes of hunting wild yak, aurochs,
elk, moose, and argali. Representations from the Middle and Late Bronze Age
include numerous scenes of family caravans, wheeled vehicles, ‘household
scenes’ and scenes of herding (Jacobson et al. 2001: II, Plates 70, 97, 217, 332).
Representations of combat are unusual. Their occasional juxtaposition with
scenes of hunting (Jacobson et al. 2001: II, Plate 370) suggests that incursions
on clan hunting lands may have been one reason for conflict.
In one superb representation from Baga Oigor IV (Figure 9.13), a large
yak bearing a basket, within which can be seen five diminutive figures, is
surrounded by a number of figures shooting arrows, while on the sidelines
stand two women.24 There are many Bronze Age representations of wheeled
vehicles at Tsagaan Salaa/Baga Oigor. Usually they are shown complete with
basket, two wheels, a driver and two horses (Figure 9.14). Sometimes, however,
they are represented in a shorthand form: the cart alone, the cart with horses,
or the cart and driver but no horses, or even just one or two wheels. In one
example a very small cart with driver is shown being pulled by four racing
horses (Jacobson et al. 2001:II, Plate 36).
The images of wheeled vehicles here and in the Upper Tsagaan Gol
complex raise several perplexing questions. Vehicles could be solid-wheeled
(Figures 9.14 and 9.15) or have spoked wheels, with varying numbers of spokes
(Jacobson et al. 2001: I, Figures 215a and 680). Given the extremely rocky
nature of the terrain in these valleys, it is hard to understand how any kind of
wheeled vehicle, solid wheeled or spoked, could have been driven as quickly as
the representations suggest. Even more peculiar is the representation of a hunt
from a solid-wheeled cart after a fleeing deer (Figure 9.15). This implausibility
is all the more striking in that the surface so carved lies on a steep slope strewn
with boulders, through which no cart would have been able to negotiate a track,
let alone one moving at high speed. Clearly, the scenes with wheeled vehicles
so frequently encountered in this complex and in that of the Upper Tsagaan
Gol refer to something other than strict reality.25 Whatever the meaning, it
is interesting that none of the images of wheeled vehicles from this complex
appear to date to the Early Iron Age.
24.
Although caravan scenes are relatively more frequently encountered at this large
site than in the Upper Tsagaan Gol, scenes of raiding are quite unusual; we have docu-
mented no clear representation of the raiding of herds.
25.
Traditionally, most scholars have tended to see the wheeled vehicle as a sign
of Indo-European or Indo-Iranian mythic traditions (Francfort 1998). It is also
possible that the wheeled vehicle refers to a sign of status and wellbeing (Jacobson
1993:125–140).
Figure 9.13. Tsagaan Salaa/Baga Oigor complex (Ulan Khus som, Bayan Ölgiy aimag): drawing of a
scene of a raid on a family caravan, Bronze Age (BO IV; Jacobson et al. 2001: I, Figure 1123; drawing
by V. D. Kubarev and D. Tseveendorj).
scene with two wheeled vehicles, Bronze Age (BO IV; Jacobson et al. 2001: I, Figure 1124; drawing by
V. D. Kubarev and D. Tseveendorj).
hunt scene, Bronze Age (TS IV; Jacobson et al. 2001: I, Figure 590; drawing by V. D. Kubarev and D.
Tseveendorj).
Many scenes of hunting, caravanning and herding at Tsagaan Salaa/Baga

Oigor refer to the transitional Late Bronze–Early Iron Age. In compositions
of this period herding revolves around large cattle and horses (Jacobson et al.
2001: II, Plates 70, 71, 374) rather than the smaller sheep and goats of an earlier
stage in the Bronze Age (Jacobson et al. 2001: II, Plates 89, 90). Household
scenes datable to this period are of particular interest. These include images
of domesticated animals, wild animals, people, dwellings and wheeled vehicles
scattered over the surface of a boulder as if offering a slice of contemporary
life. Several good examples are found in Tsagaan Salaa IV (Jacobson et al.
2001: II, Plates 187–190), Baga Oigor I (Jacobson et al. 2001: I, Figure 779)
and Baga Oigor III (Jacobson et al. 2001: II, Plates 332–334). Like scenes of
caravanning, these compositions reflect a world understood as immediately
present and vital.
At Tsagaan Salaa/Baga Oigor, as in the Upper Tsagaan Gol, the Late Bronze
Age is also represented by the appearance of a highly stylized deer image
(Jacobson et al. 2001: II, Plates XI, XII), and this image persists through the
early centuries of the Iron Age. The Pazyryk Period is represented primarily by
horsemen, hunting scenes and representations of individual animals (Jacobson
et al. 2001: I, Figures 939, 1053, 1140; II, Plates 166, 224, 266), but the vigor
of Bronze Age art has often been replaced by a studied conventionalization.
After the Early Iron Age, imagery of a certain quality and identifiable style falls
off. We have documented only a few compositions from the Turkic Period and
very few from the Mongol Period, as well as a large number of pecked surfaces
of indeterminate date.
Discussion
Rock art of the Bronze and Early Iron Ages in the Mongolian Altai clearly
documents the emergence of pastoralism and, later, of seminomadic, mounted
pastoralism. These are general patterns that are confirmed by excavation
archaeology in a broader region of North Asia. The discussion of three major
rock art complexes and the physical character of their valleys suggests how
geography and climate qualified those transitions, creating various constraints
on the local development of cultures of mobility. Imagery at Aral Tolgoi reflects
a region bordering the extensive glaciation of the high Altai valleys and func-
tioning as a transition zone between open woodland and grassland. Such an
ecotone supported a hunting economy from the late Pleistocene through the
early Holocene. The later, more humid, environment of the middle Holocene
supported fauna adapted to a more heavily forested environment. During the
early Late Holocene, coinciding with the Early Bronze Age in North Asia,
the area around Aral Tolgoi was still sufficiently forested so that emerging
pastoralism would have been quite limited. Indeed, imagery reflective of a

cultural shift to pastoralism is almost entirely lacking from the Aral Tolgoi
repertoire, as is any imagery indicative of the shift to a mounted seminomadic
lifestyle at the end of the Bronze Age. Surface structures (burial mounds, ritual
altars and stele, Turkic image stones and enclosures) found on the valley floor
around Aral Tolgoi and farther west, however, indicate that the valley at the
west end of Khoton Nuur offered a zone of transition for small numbers of
Iron Age herders moving quite probably toward the passes into present-day
China; but the lack of year-round pasture at the west end of Khoton Nuur
would have discouraged extended habitation.
The valley of the upper Tsagaan Gol and its proximity to forest and high
mountain grasslands offered ideal habitat for aurochs, elk, bear, wild sheep and
ibex in the late Pleistocene and early Holocene. The extreme cold of the high
valley would not, however, have encouraged regular human habitation. Once
a pastoral economy was adopted in the Bronze Age, the precipitous slopes that
flank much of the upper valley would have made it difficult to access the rich
pastures above, significantly limiting the size of herds that could be maintained
year-round in the valley. The lack of sheltered places along the edges of the
valley would also have discouraged winter dwelling. Most of the household
and caravanning scenes in the Upper Tsagaan Gol complex are east of Shiveet
Khairkhan; the relatively small number of such scenes in the upper Khar Salaa
and their absence within the Tsagaan Salaa valley suggests that these high
subvalleys were neither heavily nor continuously populated through the Bronze
and Iron Ages although they were used extensively for hunting and herding.
The depth and extensiveness of the pictorial documentation found at
Tsagaan Salaa/Baga Oigor is no accident. Within the Mongolian Altai this
valley presented an optimal environment for the emergence of cultures of
mobility. The size of the valley and its marshlands would have supported
hunting- and fishing-dependent populations in the late Pleistocene and early
Holocene. The broad valley floor and forested slopes would have provided
ample browsing and grazing for wild animals. By the late-middle Holocene,
as forests retreated, grassland became continuous from the valley floor up to
the broad ridges of surrounding mountains, and valley wetlands continued to
support rich wild bird and fish populations. Folds and draws along the walls of
the valley offer shelter from winds from the north and west. In these respects
the Tsagaan Salaa and Baga Oigor valleys offered an ideal environment for the
cultural transition first to pastoralism and then to a seminomadic and horse-
dependent pastoralism.
Scenes representing the herding of cattle, horses and goats vividly reflect the
emergence of a pastoral economy in the Bronze Age. Style indicates that many
of the finest scenes of caravanning in both the Upper Tsagaan Gol and Tsagaan
Salaa/Baga Oigor were contemporaneous with the herding scenes. They reflect
the movement of families up and down the valleys in search of good pasture,
at least by the Middle to Late Bronze Age. When we try to understand the
appearance of wheeled vehicles, however, many images of which are so clearly
contemporaneous with the herding and caravanning scenes, we are faced with a
puzzle. The vehicles never occur in conjunction with the caravans, nor are they
ever represented as loaded.26 There are, however, a few instances where they do
appear in household scenes. The best example is found on a surface from Baga
Oigor III, on which is represented a small family, its domesticated animals and
the proximity of abundant wildlife (Jacobson et al. 2001: I, Figures 979, 980; II,
Plates XXII, XXIII). Despite such images, however, it remains uncertain how
wheeled vehicles participated in an emerging cultural mobility.
The case for the adoption of horse riding is in some respects clearer but not
wholly so. Burials from the Pazyryk Period indicate that horses were expected
to accompany the dead to the next world and were thus, in a sense, a means of
virtual long-range mobility. Reason and experience argue that the development
of herds of large animals required horse riding: numbers of horses or yak could
hardly be controlled by herders on foot nor, of course, would wheeled vehicles
have been any use when the animals were moving up steep slopes. But while the
petroglyphic record clearly reflects the use of horses for hunting, it nowhere
indicates the use of horses for herding or for carrying burdens.27
The final conundrum goes back to the question posed earlier in this chapter:
how did mobility actually happen? As far as we can tell from the vast petro-
glyphic record of the Mongolian Altai, yak were the primary means by which
people first engaged in a mobile way of life in the Bronze Age. Yak carried
the portable dwellings and all the furnishings, the children and the elderly
(Jacobson-Tepfer et al. 2007: II, Plate 530), while adults walked. That, at least,
is completely clear through the Late Bronze Age, but the record for the Early
Iron Age, or Pazyryk Period, is unclear since there are no caravan scenes that
can be certainly so dated. Moreover, although camels are now used extensively
in the Mongolian Altai as primary beasts of burden, there is no record whatso-
ever of their having been so used in the Late Bronze or Early Iron Ages. Again,
when they do appear, it is as riding animals, even within a hunt (Jacobson et al.
2001: II, Plate 184). We are left with the conclusion that mobility did occur in
26.
The only exception with which I am familiar is from the Russian Altai, from the
site of Kalbak-Tash, where a small wheeled vehicle, without horses or driver, is found
within a fine scene of caravanning (Kubarev and Jacobson 1996:Figure 449).
27.
See, for example, a fine scene of a family preparing to move camp, from Baga
Oigor IV. The animals, including horses, are attached to a lead but only the yak are
loaded (Jacobson et al. 2001: II, Plate 332).
the Bronze Age, thanks in large measure to yak as beasts of burden. The role
that horses played in the emergence of seminomadic herding can be inferred
but not proven; and the actual role of wheeled vehicles within the emergence
of mobility is quite unclear, except that they were evidently valued indicators
of material well-being. They were signs of mobility but not necessarily the
means of large-scale social mobility.
This discussion indicates possible sources for a study of North Asian cultures
of mobility and the way in which specific physical contexts may have varied
actual experience within any particular region. Even with the materials now
available for Bayan Ölgiy, it would be possible to multiply the case studies
offered above. Ultimately such a consideration of particular valleys with the
Altai Mountains will help to flesh out the theoretical constructs underlying the
study of the archaeology of mobility in North Asia.
References
Alekseev, A. Y., N. A. Bokovenko, Y. Boltrik, K. A. Chugunov, G. Cook, V. A. Dergachev,
N. Kovalyukh, G. Possnert, J. van der Plicht, E. M. Scott, A. Sementsov, V.
Skripkin, S. Vasiliev and G. Zaitseva
2001 A Chronology of the Scythian Antiquities of Eurasia Based on New
Archaeological and 14C Data. Radiocarbon 43: 2B, pp. 1085–1107.
Anthony, D. W.
1998 The Opening of the Eurasian Steppe at 2000 BCE. In V. H. Mair (ed.), The
Bronze Age and Early Iron Age Peoples of Eastern Central Asia. Washington,
Institute for the Study of Man: pp. 94–113.
Blyakharchuk, T. A., H. E. Wright, P. S. Borodavko, W. O. van der Knaap and V.
Ammann
2004 Late Glacial and Holocene Vegetational Changes on the Ulagan High-
Mountain Plateau, Altai Mountains, Southern Siberia. Palaeogeography,
Palaeoclimatology, Palaeoecology 209: pp. 259–279.
Derevyanko, A. P. and S. Natsagdorj (eds.)
1990 Arkheologicheskie, etnograficheskie i antropologicheskie issledovaniia v Mongolii:
Sbornik nauchnykh trudov (Collection of Research Papers on the Archaeology,
Ethnography, and Anthropology of Mongolia) [in Russian]. Novosibirsk,
Nauka.
Dorn, R. I.
2001 Chronometric Techniques: Engraving. In D. S. Whitley (ed.), Handbook of
Rock Art Research. Walnut Creek, Altamira Press: pp. 167–189.
Francfort, H.-P.
1998 Central Asian Petroglyphs: Between Indo-Iranian and Shamanistic Inter
pretations. In C. Chippindale and P. S. C. Taçon (eds.), The Archaeology of
Rock-Art. Cambridge, Cambridge University Press: pp. 302–318.
Gryaznov, M. P.
1980 Arzhan: Tsarskii kurgan ranneskifskogo vremeni (Arzhan: A Royal Burial of the
Early Scythian Period) [in Russian]. Leningrad, Nauka.
Gunin, P. D., E. A. Vostokova, N. I. Dorofeyuk, P. E. Tarasov and C. C. Black
1999 Vegetation Dynamics of Mongolia. Dordrecht, Kluwer.
Guthrie, R. D.
1982 Mammoths of the Mammoth Steppe as Paleoenvironmental Indicators. In
D. M. Hopkins (ed.), Paleoecology of Beringia. 81st Wenner-Gren Foundation
for Anthropological Research Symposium (Bert Wartenstein, Austria, 1979). New
York, Academic Press: pp. 307–326.
Jacobson, E.
1993 The Deer Goddess of Ancient Siberia: A Study in the Ecology of Belief. Leiden, E.
J. Brill.
2000a “Emblem” Against “Narrative” in Rock Art of the Mongolian Altay [in Russian
and English]. Bulletin of the Siberian Association of Prehistoric Rock Art Researchers
3: pp. 6–14 .
2000b Petroglyphs and Natural History: Sources for the Reconstruction of the
Ecology of Culture. Archaeology, Ethnology, and Anthropology of Eurasia 1: pp.
57–65.
2001 Cultural Riddles: Stylized Deer and Deer Stones of the Mongolian Altai.
Bulletin of the Asia Institute. New Series 15: pp. 31–56.
2002 Petroglyphs and the Qualification of Bronze Age Mortuary Archaeology.
Archaeology, Ethnology, and Anthropology of Eurasia 3: 11, pp. 32–47.
2005 Aral Tolgoi. In D. Tseveendorj, V. D. Kubarev and E. Yakobson (Jacobson)
(eds.), Aral Tolgoin Khadny Zurag (Petroglyphs of Aral Tolgoi) [in English,
Mongolian, and Russian]. Ulaanbaatar, Institute of Archaeology, Mongolian
Academy of Sciences: pp. 103–132.
Jacobson, E. and G. Kubarev
2003 Rock Art Complex on the Irbystu River, Altay Republic. International Newsletter
on Rock Art 36: pp. 12–16.
Jacobson, E., V. D. Kubarev and D. Tseevendorj (edited by J. Sher and H.-P.
Francfort)
2001 Mongolie du Nord-Ouest: Tsagaan Salaa/Baga Oigor. Répertoire des pétroglyphes
d’Asie centrale, Fascicule 6 (2 volumes). Paris, De Boccard.
Jacobson-Tepfer, E., V. D. Kubarev and D. Tseveendorj (edited by H.-P. Francfort
and Y. Sher)
2007 Mongolie du Nord-Ouest: Haut Tsagaan Gol. Répertoire des pétroglyphes d’Asie Centrale,
Fascicule 7 (2 volumes). Paris, De Boccard.
Khazanov, A. M.
1994 Nomads and the Outside World. Madison, University of Wisconsin Press.
Kubarev, V. D.
1986 Kurgany Ulandryka (Kurgan Burials at Ulandryk) [in Russian]. Novosibirsk,
Nauka.
1991 Kurgany Yustyda (Kurgan Burials at Yustyd) [in Russian]. Novosibirsk,
Nauka.
Kubarev, V. D. and E. Jacobson
1996 Répertoire des Pétroglyphes d’Asie Centrale. Sibérie du Sud 3, Kalbak-Tash I
(République de l’Altay). Mémoires de la Mission Archéologique Française en Asie
Centrale. Tome V.3. Paris, De Boccard.
Molodin, V. I. and D. V. Cheremisin
1998 Drevneishie naskal’nye izobrazheniya ploskogor’ya Ukok (The Most Ancient Rock
Art of the Ukok Plateau) [in Russian]. Novosibirsk, Nauka.
Molodin, V. I., N. V. Polos’mak, A. V. Novikov, E. S. Bogdanov, I. Y. Slusarenko and
D. V. Cheremisin
2004 Arkheologicheskie pamyatniki ploskgor’ya Ukoka (Archaeological Monuments
of the Ukok Plateau) [in Russian]. Novosibirsk, Institute of Archaeology and
Ethnography, Siberian Section, Russian Academy of Sciences.
Polos’mak, N. V.
2001 Vsadniki Ukoka (Horsemen of the Ukok) [in Russian]. Novosibirsk,
INFOLIO.
Polos’mak, N. V. and L. L. Barkova
2005 Kostium i tekstil’ pazyryktsev Altaia (IV–III vv. do n.e.) (Costume and Textiles of
the Altai Pazyryk People of the Fourth–Third Centuries BCE) [in Russian].
Novosibirsk, INFOLIO.
Savinov, D. G.
1994 Olennye kamni v kul’ture kochevnikov Evrazii (Deer Stones in the Culture of the
Eurasian Nomads) [in Russian]. St. Petersburg, University of St. Petersburg
Press.
Slusarenko, I. Y., J. A. Christen, L. A. Orlova, Y. V. Kuzmin and G. S. Burr
2001 14
C Wiggle Matching of the “Floating” Tree-Ring Chronology from the Altai
Mountains, Southern Siberia: The Ulandryk-4 Case Study. Radiocarbon 43:
2A, pp. 425–431.
Solov’ev, A. I.
2001 Oruzhie i dospekhi (Weapons and Armor) [in Russian]. Novosibirsk,
INFOLIO.
Tarasov, P., N. Dorofeyuk and E. Metel’tseva

2000 Holocene Vegetation and Climate Changes in Hoton-Nur Basin, Northwest
Mongolia. Boreas 29: pp. 117–126.
Tseveendorj, D., V. D. Kubarev and E. Yakobson (Jacobson)
2005 Aral Tolgoin Khadny Zurag (Petroglyphs of Aral Tolgoi) [in Russian].
Ulaanbaatar, Institute of Archaeology, Mongolian Academy of Sciences.
Vadetskaya, E. B.
1986 Arkheologicheskie pamyatniki v stepyak srednego Yenisey (Archaeology in the
Steppe Region of the Middle Yenisey) [in Russian]. Leningrad, Nauka.
Varenov, A. V.
1984 O funktsional’nom prednaznchenii “modelei yarma” epokhi In’ i Chou
(Regarding the Function of the “Model Yoke” from the Yin and Zhou
Dynasties). In Novoe v Arkheologii Kitaya (New Directions in Chinese
Archaeology) [in Russian]. Novosibirsk, Nauka: pp. 42–51.
Volkov, V. V.
2002 Olennye kamni Mongolii (Deer Stones of Mongolia) [in Russian]. Moscow,
Nauchnyi Mir.
Ch a p t e r 1 0
Nomadic Sites of the

South Yergueni Hills on
the Eurasian Steppe
Models of Seasonal Occupation
and Production
N a t a l y a I . S h i s h l i n a , E . I . G a k a n d A . V. B o r i s o v 1
T he archaeological record that was used to reconstruct the economy

of the ancient population of the northwestern Caspian Sea steppe consisted
of artifacts found in kurgans (burial grounds) dating back to the Bronze Age.
Although it was argued that seasonal sites were very difficult to identify in
the open steppe (Sinitsyn 1931; Shilov 1975; Koltsov 1985), archaeological
surveys found sites, as well as different types of artifacts (Minayeva 1955;
Koltsov 1985), dating back to the Mesolithic, Neolithic and Eneolithic Periods
(Koltsov 1988). The existence of Bronze Age and Early Iron Age sites is still
debated. During our archaeological survey in the southern part of the Kalmyk
Eurasian steppe (Figure 10.1), we found two sites of the Late Bronze and Early
Iron Age: Gashun-Sala and Manych (Figure 10.2).2 The archaeological exca-
vation of these sites produced new evidence of the seasonal economic cycle
of pastoral nomads living in this region during the end of the second into the
first millennium BCE.
Geomorphologic and Geobotanic Descriptions

The sites at Gashun-Sala and Manych are located on the southern fringe of a
high plateau in the Yergueni Hills. The southern border of the Yergueni Hills is
1.
This project was conducted with support of the National Geographic Society,
project number 7978-01. We acknowledge the assistance provided by G. Klevezal, I.
Kirillova, A. Bobrov, M. Pakhomov, M. Novikova, R. Mimokhod, J. van der Plicht, V.
Chichagov and A. Sokolov.
2.
A third site, Chilyuta, was discovered close to Gashun-Sala. Apart from a 1x1 m
bore hole, which reached a depth of 3 m, no excavations were performed here.
230
N o ma d i c S i t e s of the Y e r g u e n i H i l l s      231
Figure 10.1. Map showing the position of the study area in the north Caspian Steppe.
the local linear uplift of the northwest Choloi-Khamur stretch. An uphill valley
of the southern Yergueni Hills stretches to the north of the Shared ravine.
Another large ravine, called Ulan-Zukha, is located to the north. At present,
all large ravines of the Yergueni Hills are dry and are only sporadically filled
with insignificant amounts of water. However, the considerable width of the
valleys, the meander belts, and the two terraces above the flood-prone areas
indicate that in the past rivers flowed through these valleys.
Throughout the Kalmyk steppe, grazing grounds for sheep and goats are
covered with a specific vegetation in which the following plants predominate:
cornflower (Centaurea apiculata Ledeb.), stickseed (Lappula heteracantha Ledeb.
Guerak.), thistle (Carduus thoermeri), yarrow (Achillea nobilis L.), Camphorosma
monspeliacum L., Xanthium albinum (X. Riparium), different species of
Amaranthus and other species that grow near sites exploited by humans. The
vegetation of the ravines includes belts of mixed grasses in contrast to the more
monotonous vegetation that covers the terrace hills. More than a hundred
plant species have been identified, with a predominance of annual plants
(Atriplex, Chenopodium, Lactuca, Linum, Polygonum), mixed perennial plants
(Carduus thoermeri, Galim chumifusum, Phlomis pungens, Potentilla recta, Salvia
sp.), plants growing near water (Aeluropus littoralis [Gouan] Parl., Petrocimonia,
Salicornia eurapaea L.), reeds (Phragmites communis Trin.), Aster tripolium L. and
numerous species of Artemisia. There are seasonal variations in the vegetation
cover resulting in repeated cycles of vegetation and flowering.
232    N ata l i a S h i s h l i n a , E u g e n y G a k and Alexander Borisov
Figure 10.2. Site location map of (A) Manych and (B, C) Gashun-Sala.
Manych, 45°N 40'47.9" / 044°E 33'28.1", is located in the upstream part of

a ravine, called Shared, which was a valley of a small plain river (Figure 10.2a).
The site is located at the foot of the right bank slope, in the place most suitable
for living: an open area, far from the upstream slope and near a river channel
of what might have been a full-flow river. These conditions were favorable in
several respects. The force of the wind was less than on the steppe. Rain and
the runoff from the melting snow were held back by the vegetation and did
not flood the settlement. If water reached the flat area of the terrace, it lost
its energy, spread out thinly, and receded quickly. No landslides or mudflows
were observed near the site, yet traces of many such events were preserved on
the opposite slope. A dense network of short tributaries of the Shared River
descended from the almost vertical slopes in places where water had taken
shortcut routes into the river, subjecting their beds to intensive linear erosion
and erosive leakage. All these tributaries appeared to be erosive washouts
formed along cattle paths that were exploited for a long time.
Gashun-Sala, 45°N 34'44.6" / 044°E 39'36.3" E, is located near the large
Ulan-Zukha ravine (Figure 10.2b, c). The internal part of a large meander,
located upstream from the site, is at present a low-lying boggy plain. When
the site was occupied, this area may have been an oxbow lake. Judging from
the size of the large meanders, and the intensity of the erosion in the valley, the
river near Gashun-Sala was a full-flow river with a permanent flow. Both sites,
therefore, appear to have been located near valleys of full-flow paleorivers with
a permanent inflow and flow rate. The slopes of the Gashun-Sala valley also
show many cracks, both ancient and more recent, the result of erosion along
ancient and modern cattle paths.
Archaeological Excavations
At Manych an area of 80 m² was subdivided into 2x2 m squares. Stratigraphy
was defined according to natural and anthropogenic layering. Three layers
were identified: the turf, the occupation layer, and the rock bed (Figure 10.3).
The turf layer was 5–7 cm thick and the anthropogenic layer, between the
turf and the rock bed, was as much as 40 cm thick in places. It consisted of
dark gray to brown-gray loam. All objects were found in this layer: 58 remains
of animal bones; 77 fragments of handmade, thick-walled, sand-tempered
pottery; and 29 fragments of wheel-thrown amphorae. Comparison of the
potsherds with the Early Iron Age steppe ceramics from adjacent areas (North
Caucasus, Don River Valley) showed a similarity with Sarmatian vessels. Five
fragments of handmade potsherds of the Late Bronze Age Srubnaya culture
were also unearthed (Figure 10.4/2). Another find appeared to be a splash of
copper (Figure 10.4/10). No remains of human dwellings were seen. In the
rock bed a shallow pit (1.20 x 0.90 m with a depth of 0.38 m) was excavated.
This pit was egg-shaped in transect and cup-shaped in cross-section. It
contained some animal bone. The recovered Iron Age ceramics (Figure 10.4/1,
4–9) were tempered with coarse sand; had thick, smooth walls; and did not
preserve any decorations. The most interesting fragments include four rims,
a base, and a massive handle with a worn stamp (Figure 10.4/9). The wheel-
thrown amphorae are orange or gray; fragments included two rims and four
fluted walls.
Figure 10.3. Manych: (top) the excavation trenches; (bottom left) a shallow pit; (bottom right) a splash
of copper.
Figure 10.4. Finds from Manych: (1, 3) sherds of wheel-made amphorae; (2) a handmade Srubnaya sherd
of the Late Bronze Age; (4–9) Early Iron Age handmade potsherds; (10) a splash of copper.
At Gashun-Sala an excavation area was selected near the bottom of the

ravine where a dark gray layer, 0.2–0.3 m thick, with a low content of humus,
was visible at a depth of 0.5–0.7 m from the outcrops. During the cleaning
of these outcrops, this layer yielded 20 animal bones and teeth, as well as 18
fragments of handmade ceramics, including the rim and base of a large vessel
(Figure 10.5/6). Another concentration of animal bones was uncovered near
the western cliff, where two fragments of smooth-walled ceramics without
decoration were unearthed (Figure 10.6/2–3). Bore holes revealed that this
layer, believed to have been anthropogenic in origin, spread along an area of
600 m² adjacent to the western cliff. It has few artifacts and is covered by a
thick deluvial sediment.
An area of 128 m² was divided into 2x2 m squares. Four stratigraphic
layers were identified according to the natural and anthropogenic layering
and the presence of objects. The upper layer (layer 1) was of deluvial origin
and consisted of light gray loam 0.20–0.70 m thick without finds. Layer 2 was
similar to the layer identified along the cliffs. Its thickness varied from 0.20 to
0.60 m in the southern part of the excavated area. Finds included 271 fragments
of animal bones and teeth, 60 fragments of handmade ceramics and large frag-
ments of a wheel-thrown thick-walled red clay vessel. Most handmade ceramics
were gray with a smooth wall (Figure 10.6). One of the rims was decorated with
slanting notches along the edge (Figure 10.6/12); some walls preserved hori-
zontal line incisions (Figure 10.6/4, 6, 10, 15, 17), large teethlike impressions
(Figure 10.6/11), or finger-made pinches (Figure 10.6/9). Other fragments have
traces of combing. There were also fragments of a handmade smooth-walled
vessel with a flat bottom and a salient body. Its neck, which is not quite obvious,
has an ornament in the form of alternating holes and spikes (Figure 10.5/2).
Fragments of the upper part of a large handmade pot-shaped vessel were found
in the southwestern part of the excavated area (Figure 10.5/1). Similarity of
the ceramic assemblage with Srubnaya (Figure 10.5/1, 2, 6; Figure 10.6/8–9,
11–12) and Sarmatian (Figure 6/2–4, 6–7, 10, 12–13, 15–17) vessels suggests
these sites were used by representatives of both cultures. Other finds include
a fragment of a bone tool (Figure 10.5/4), possibly a skate, typical for the Late
Bronze period, a long bone with cut marks (Figure 10.5/3) and a fragment of
a clay item with a hole (Figure 10.5/5), possibly a spindle whorl.
Layer 3 consisted of gray to dark brownish loam without anthropogenic
insertions; its thickness varied from 0.10 to 0.45 m. Most of the finds were
uncovered in its upper part, which yielded 116 remains of animal bones and 62
sherds of handmade undecorated pots with smoothed walls. Their thick walls
preserved traces of polishing with grass. The fabric was tempered with small
organic inclusions and was poorly fired. A few fragments of decorated ceramics
were also present: a rim fragment with a beak (Figure 10.6/1), the wall of a
Figure 10.5. Finds from Gashun-Sala: (1, 2, 6) potsherds; (3) a long bone with cut marks; (4) fragment
of a bone tool, possibly a skate; (5) a ceramic object, possibly a spindle whorl.
Figure 10.6. Assemblage of handmade potsherds found at Gashun-Sala.

Figure 10.7. Gashun-Sala: (top) the excavation trenches; (bottom left) handmade ceramics in situ;
(bottom right) animal bones in situ.
thick-walled vessel decorated with finger impressions (Figure 10.6/18), and

several sherds ornamented with incised zigzag lines (Figure 10.6/5, 14). This
type of ceramic is very typical for the Late Bronze Age steppe cultures, as well
as for the Early Iron Age period. The finds from layer 3 were not distributed
evenly across the excavation area: there were concentrations of small fragments
of ceramics in the eastern part, where almost no animal bones were present.
No remains of human dwellings were found. Layer 3 overlies layer 4, which is
a rock bed of yellow to grayish loam (Figure 10.7).
Neither hearths nor the remains of permanent residential buildings or
auxiliary houses were found in either of the sites. Most of the finds were animal
bones and ceramics dating to the Late Bronze Age, the Srubnaya culture and
to the Early Iron Age.
Archaeozoologic Identifications
The mammal bones from both Manych and Gashun-Sala were poorly preserved.
The taxonomic composition represents a small number of species, but many
small or heavily damaged fragments could not be identified. The assemblage
contained the remains of domesticated horse, cattle, sheep and goats. Hunted
species were also present but in much smaller numbers. The animal bones
from Manych included bones of horse, ungulates and domesticated bull. At

Gashun-Sala sheep bones predominated, but these were absent among the
bones uncovered at Manych. Bones of gazelle, saiga (Saiga tatarica) and kulan
(Equus hemionus kulan) were rarely found at Gashun-Sala.
Archaeobotanic Deter minations

Samples of paleosoil were taken from different occupation layers. Sample 1, from
the occupation layer in the southern part of the trench at Manych, contained a lot
of poorly preserved pollen. Pollen of goosefoot (Chenopodiaceae) predominated,
but pollen of gramineous plants (Poaceae), asters (Asteraceae) and unidentifiable
mixed grasses were found in smaller quantities. Remarkable finds included the
pollen of lily (Liliaceae), pine (Pinus) and genuine fern (Polypodiaceae). Sample
2, taken just above sample 1, also had an abundance of pollen of goosefoot
(Chenopodiaceae), asters (Asteraceae) and mixed grasses. Wormwood (Artemisia)
was found in smaller quantities; in rare cases legumes (Fabaceae), Gramineae plants
and Ephedra were found. Cattail (Typha) was found sporadically; pine (Pinus)
and alder (Alnus) were found rarely. Data of phytolith analysis corroborated the
results of the pollen analysis. It should be noted that the soil could have been
without vegetation for a long time as a result of overgrazing, pasture digression
or wind erosion. Flotation of samples did not yield any results.
A sample from layer 2 at Gashun-Sala had poorly preserved pollen with
a low content of pollen of goosefoot (Chenopodiaceae), asters (Asteraceae),
Brassicaceae, chicory (Cichoriaceae) and mixed grasses. Another sample, from a
layer that contained large quantities of bones, contained phytoliths of gramin-
eous plants. Flotation of the subsoil yielded the fruits of chicory (Cichorium
intybus L.) and amaranth seeds (Amaranthus sp.) (Figure 10.8). White amaranth
seeds (Amaranthus albus L.) were found in a bore pit near the site in another
layer with large quantities of animal bones (Novikova et al. 2002). The presence
of amaranth, which grows near roads, may indicate that, at the time of the occu-
pation of the site, soils around the site were destroyed. This provides indirect
evidence of the possible destruction of the ancient grass and soil cover around
the site, which could have been caused by human activity and the overgrazing
by domesticated animals. Phytoliths in these samples are predominantly of
gramineous plants, characterized by the presence of long cells with wavy edges.
The absence of short cells suggests that the vegetation was once mesophyte
when the climate was more humid than it is today.
Analysis of the botanical remains indicates that, at the time of the occupation
of the site, the area was covered with grasses, amaranth and goosefoot. Plants like
ferns and cattail did grow near the water, and it is quite possible that alder grew
nearby. The adjacent areas must have been used as permanent grasslands.
Figure 10.8. Seeds of amaranth found at Gashun-Sala.
Analysis of the Paleosoil

Morphological properties of the soil profiles were described and soil horizons
were identified in the soil profile. Air-dried soil samples were passed through
a sieve (2 mm mesh) and chemically analyzed. The soil pH was determined
on a saturation paste using a glass electrode pH-meter. The total organic
carbon content was determined by oxidizing the humus with 0.4 n K2Cr2O7,
prepared in a sulphuric acid and water mixture (1/1, v/v). The gypsum content
was measured by extracting SO42- from the soil with 0.2 n HCl, followed by
precipitation of SO42- by 10% BaCl2. The content of carbonates (like CaCO3) in
the soil was determined by titration. The phosphorus content was measured by
determining the concentration of P2O5 in carbonate soils with 1% ammonium
acetate (the Machigin technique).
In the soil profile at Manych a thin layer (2–3 cm) of yellow-brown loam was
discovered at a depth of 12–14 cm. This layer is an outcrop in the domestic pit,
indicating the ancient soil surface at the site at the time that this pit was made.
The layer above is the gray humus horizon of the modern soil, which has been
formed as a result of deluvial processes. This horizon appeared high in organic
carbon and phosphorus, with a neutral to slightly alkaline pH. The occupa-
tion layer proper is located at a depth of 14 cm and below. The high content
Figure 10.9. Chemical properties of soils from Manych (left) and Gashun-Sala (right).
of mobile phosphorus in the soil should be noted (Figure 10.9). At a depth of

50–60 cm buried soil that formed over alluvial sediments was observed. The
site, obviously, has been established in a semihydromorphic landscape on the
ancient alluvial calcareous saline soil. A rather thick occupation layer with
an insignificant number of artifacts, on the one hand, and the relatively high
content of phosphorus and organic carbon, on the other hand, indicates that,
most likely, an active accumulation of artifacts occurred in the occupation
layer when the site was used during a more humid climate and consequently
more intensive water erosion. It can be concluded that the site was used rather
intensively during a relative short time.
We also analyzed a soil sample from Gashun-Sala. The upper 30 cm layer
was, in fact, the result of sedimentation processes after the site was abandoned.
The occupation layer proper was at a depth of 30–80 cm. This horizon had the
highest concentration of artifacts. The chemical properties of this layer differ

significantly from those of the layers above and below. The organic carbon
content of the layer does not vary between 30 and 80 cm, but in the lower
horizons it is drastically reduced. At the same time, the phosphorus content is
rather low (0.3–0.7 mg/100 g soil). Assuming that phosphorus is an indicator
of human occupation, this low phosphorus content can be considered the result
of little human activity. In the deeper layers the content of phosphorus is not
considerably different (Figure 10.9). The buried soil was deposited below the
occupation layer at a depth of 80 cm and below. We suppose that the intense
grazing of cattle in the adjacent areas caused the degradation of the vegetation
cover and intensified the erosional processes resulting in a rapid growth of the
occupation layer. We also think that Gashun-Sala was occupied for a longer,
but less intensive, period than Manych.
Chronology
The dates of the occupation of the sites were determined by radiocarbon
analyses of the animal bones from anthropogenic layers and correlated with
the recovered artifacts. The results of the 14C analysis are presented in Table
10.1. Analysis of the data indicates that both sites were actively exploited
during the Early Iron Age. As a whole, these dates corroborate the results of
the analysis of the ceramic assemblages. Gashun-Sala is likely to have been
used more frequently, and for a longer period, from the Late Bronze Age,
1800–1600 BCE calibrated (the Srubnaya culture) until the 1st century CE.
The chronology of the Srubnaya culture is based on the 14C analysis of graves
in the same area (Mimokhod and Shishlina 2004). The period of the most
intensive activity was probably between the 7th-6th centuries BCE and the 1st
century BCE. This fits the results of our analysis of the paleosoil, which dated
Manych to the 7th-6th centuries BCE. The site could have been visited by local
Srubnaya pastoral groups as well.
Table 10.1. 14C Data from Manych and Gashun-Sala:
Sample (lab. no.) Material Age BP (uncalibrated)a Calibrated agea

Manych Animal bone from layer 2 2560±40 2746–2559 BP
(GrA-19259) 797–610 BCE
Gashun-Sala Animal bone from layer 3 2495±45 2722–2467 BP
(GrA-19226) 773–518 BCE
Gashun-Sala Animal bone from layer 2 1980±30 37–13 BCE
(GrA-17487) 1–59 CE
Analyses performed by J. van der Plicht at the Groningen 14C Laboratory, the Netherlands.
a
Table 10.2. Seasonality of the Death of Animals Found at Gashun-Sala and Chilyuta:
Site Samplea, b Age at death Season of Preservation

death
Gashun-Sala, layer 2 Ovis aries, left lower jaw >2 years Late summer Satisfactory
with teeth, M2 and M3 of the to fall
same species (dp4)
Gashun-Sala, layer 3 Small ungulate dp4 2 years Summer Satisfactory
Gashun-Sala, layer 3 Lower jaw of a small 1–2 years Summer Poor
ungulate with dp4
Gashun-Sala, bore hole A fragment of the left lower 5 years Spring to late Satisfactory
(1.05 m deep) jaw of a sheep or goat with summer
premolars, M1, and M2
Gashun-Sala, layer 2 Unknown teeth from the 3–7 years Late summer Satisfactory
upper jaw of a sheep or goat to fall
Chilyuta, bore hole M1 or M2 from the lower 2–3 years Spring to late Satisfactory
jaw of a sheep or goat with a summer
broken root
Analysis was performed by G. Klevezal and A. Sokolov.

a
M1 = first molar; M2 = second molar; M3 = third molar; dp4 = premolar.

b
Methodological tools to determine seasonality of archaeological sites has

been discussed in Kirillova et al. (2000). We selected six animal teeth from
the osteological finds from Gashun-Sala and from a bore hole at the adjacent
site, Chilyuta. Teeth from the occupation layer at Manych turned out to be
unsuitable for dentum and cementum analysis. The results of our analysis are
presented in Table 10.2, which demonstrates that Gashun-Sala was used as a
seasonal camp from the spring to early fall. Nearby sites that were suitable
as seasonal camps, like Chilyuta, could also have been exploited during the
warmer periods of the year.
Discussion
The pastoral population of the steppe seemed to have settled in valleys of
paleorivers with a continuous flow. They selected flat grounds or the slopes
next to low-lying flood-prone terraces. Sites tended to be built in expanded
valleys of narrow steppe rivers in areas with large meanders. The sites were
conveniently located in shallow open valleys, close to a river and far from the
slope above, where winds blew with almost the same force and frequency as on
the surrounding plain. The botanical remains indicate that trees such as alders
used to grow in the ravines. The vegetation cover was characterized by substan-
tial variations and an abundance of species. This differed significantly from
the scarce vegetation of the watershed areas in the open steppe. Many plants
identified in the occupation layers were good for fodder; their productivity is
at a maximum between May and August. Our determination of the periods

of occupation of Gashun-Sala, and nearby Chilyuta, indicates that they were
only used during the warmer times of the year. The seasons determined for
the occupation during the Srubnaya period, as well as those for the Early Iron
Age burials from nearby burial grounds, is consistent with this. Seasonality of
the camps fits the results of the analysis of the paleosoil, as well as the climatic
characteristics of that time.
The absence of any traces of permanent constructions, dwellings, foundation
pits or permanent domestic pits at the sites indicates that they served as tempo-
rary campsites. We suppose that both wagons and lightweight constructions of
braided mats (such as the ‘kibitka’ of the historic Kalmyks) were used as dwell-
ings. The study of ancient textiles from the region demonstrates that, from the
Bronze Age onward, mats were produced here from whatever materials were
available. These mats were used in burial pits but could also have been used as
construction material (Shishlina 1999). Soil flotation did not reveal the presence
of any domesticated gramineous plants at the sites. The seeds of amaranth and
chicory found at Gashun-Sala indicate that its residents were probably gathering
plants for their nutritional or medicinal properties. They may also have dried
edible plants for the winter. The small number of finds is most likely the result
of the temporary use of these sites. The fragment of a spindle whorl and the
bone with cut marks found at Gashun-Sala and the splash of copper found at
Manych indicate a limited household production (spinning, casting, processing
of bones). We can suppose that sheep were sheared, skins and furs were tanned,
and carpets were felted in such camps. Indirect proof of this can be inferred from
the composition of the animal bone assemblage and the comparison with ethno-
graphic models of pastoral production (Zhitetsky 1893; Tajzscanova 1994).
Bone remains show that the population was engaged in raising domesticated
cattle, sheep and horses. Natural water sources appear to have been used as
watering places for the animals. Apparently the residents of the sites raised a lot
of livestock, and their paths, which must have been used for a long time, eroded
into tributaries to the Shared River. The grasslands around the sites must have
been used intensively, which led to the depletion of the soils and the appearance
of weeds such as chicory and amaranth. The differences in animal species at the
investigated sites are also of interest. The absence of sheep and goat bones in
the assemblage at Manych points to a certain specialization in raising animals:
only cows and horses must have been raised at Manych. Wild animals were
absent at Manych, in contrast to Gashun-Sala, where residents hunted wild
ungulate in addition to raising cows, sheep and horses. The predominance of
sheep and goat bones at Gashun-Sala also indicates that people stayed here
for short times only, as sheep overgraze their pastures much faster than other
domesticated animals.
The summer campsites of the Gashun-Sala type are without dwellings or

domestic constructions. The population was raising domesticated sheep and
goats. Nearby pastures, with an abundance of mixed plants, in the rather wide,
deep valley, on the slopes, and in the watershed area on the open steppe were
used as grazing grounds. Horses and dogs might have been used to facilitate
such grazing. The broken soils of the pastures point to overgrazing. In addition
to herding, people hunted, gathered wild gramineous and other edible plants,
and engaged in simple household productions. Full-flow rivers were used as
natural water reservoirs.
Judging by the size of Gashun-Sala (roughly 600 m²) and the nearby wide
meanders of the Pra-Gashun River, where ceramics and animal bones were also
found, several families could live at this site and the neighboring area at the
same time, forming the network of sites observed in the river valley. Another
explanation might be that the same group migrated from one pasture to another.
Artifacts, 14C data, and the analysis of the paleosoil indicate that Gashun-Sala
was first put to use during the Late Bronze Age, but its more active exploita-
tion occurred during the Early Iron Age. It is quite possible that, as the site was
subjected to the impact of floods and wind erosion, some artifacts were displaced.
Maybe this is the reason why the dry channel at Gashun-Sala yielded fragments
of animal bones and ceramics similar to those found at the site of Gashun-Sala.
Pottery, similar to the pottery found in the kurgan burials of the Early Iron Age
at the Sharakhalsun burial ground near the Kalaus River in the Stavropol area,
indicate that local pastoral groups migrated south. Sheep, which predominated
in the herds, are very suitable for long-distance migration. Therefore, such sites
as Gashun-Sala and its adjacent area were part of an annual system of seasonal
grasslands, with migration from one pasture to the next.
The seasonal campsites of the Manych type are without permanent dwell-
ings and domestic constructions, which indicates that they were used during
the warmer periods of the year. Paleoclimatic characteristics suggest that
Manych existed in more humid and warm conditions with a higher annual
precipitation rate, a likely snow cover, and the impossibility of using this area
as pasture in the winter. The population engaged in raising cattle and horses
on the terraces, slopes and the nearby watershed areas with grasslands rich
in mixed grasses. Excavations in the Shared River valley revealed only small
amounts of ceramics and animal bones. The presence of Srubnaya pottery
indicates that the campsite was visited as early as the Late Bronze Age. The
abundance of artifacts dating to the Early Iron Age in the occupation layer
suggests that it was only actively used during that period. Taking all informa-
tion into account, Manych could only have supported a small group of people,
one or two families. Pottery stylistically similar to Manych ceramics, fragments
of coarse sand-tempered vessels, were found during surveys along the banks
of the Chograi. It is quite possible that pastoral groups migrated there for the
winter. Their winter camps could also have been located farther south, in the
area of the Stavropolye Hills. Such camps appear to belong to the system of
an annual seasonal pastoral cycle economy.
The two types of campsites, Gashun-Sala and Manych, compare fairly well
with existing ethnographic models (Zhitetsky 1893; Bulatov 2000). Kalmyk
khotons, small family groups of pastoralists, consisted of one or two families
and several wagons. They were spread out over the entire steppe, spending the
winter in permanent sites and the summer in the open steppe. Their number
was low, which is why there was no shortage in good winter campsites. The life-
style of the Kalmyks was predominantly nomadic, using the marginal grasslands
of the steppe to raise cattle, horses, sheep and goats. In the morning the women
brought the livestock to grasslands near the khoton; right after breakfast the
teenagers herded the flocks while the men took the cattle to a watering place.
During the day the women were busy with household productions: weaving,
spinning, sewing and preparing food. In the evening the men again took the
cattle to the watering place. The cattle then spent the night in one place, near a
special type of dwelling: the kibitka. These consisted of a wooden frame covered
by wool bands. When families moved to a different temporary campsite, their
kibitkas were dismantled and packed, together with other belongings, on pack
animals and the entire khoton moved to a new place. In the spring these moves
were short. In the summer, especially in case of droughts, they could last for
several days (Zhitetsky 1893:35–37). In a 19th century CE map of the winter and
summer migration routes of the Kalmyk, the steppe areas in the south of the
Iki-Burul district (north of the area where the campsites that we investigated
are located) were used as winter grasslands with migration routes going to the
south, across the Kuma and Kalaus rivers (Bulatov 2000:138–139).
Excavations at Gashun-Sala and Manych, located in the South Yergueni
Hills, and analysis of the collected data helped us to propose two types of
seasonal campsites on the Eurasian steppes in the Late Bronze Age and the
Early Iron Age. They correlate with existing historical models of nomadic
Kalmyks, who have inhabited this area from the 17th-20th centuries CE. In spite
of differences in climatic, social, economic and political conditions, we can
definitely note similarities in the seasonal and economic cycles. The ecological
niches of the steppe developed from the Eneolithic onward. When a pastoral
economy evolved and spread during the Bronze Age, a specific seasonal system
of grassland use was developed. This is reflected in the characteristics of the
sites. So far we have studied only two camps of the Late Bronze Age and Early
Iron Age. We believe that similar studies of other sites in this region will allow
us to reconstruct an overall picture of the economy and the household life of
the earliest populations of the region.
References
Bulatov, V. E.
2000 The System of Seasonal Migrations of the Kalmyks in the Nineteenth
Century [in Russian]. In N. Shishlina (ed.), Seasonality Studies of the Bronze
Age Northwest Caspian Steppe. Papers of the State Historical Museum. Volume 120.
Moscow, Poltex: pp. 131–142.
Kirillova, I. V., G. A. Klevezal, K. E. Mikhailov, A. A. Golyeva, Y. E. Trunova and N.
I. Shishlina
2000 Complex Methods to Determine the Season of the Bronze Age Graves in
Kalmykia [in Russian]. In N. Shishlina (ed.), Seasonality Studies of the Bronze
Age Northwest Caspian Steppe. Papers of the State Historical Museum. Volume 120.
Moscow, Poltex: pp. 30–41.
Koltsov, P. M.
1985 New Neolith and Eneolith Sites of the Black Land [in Russian]. In K. N.
Maksimov (ed.), Ancient Sites of Kalmykia. Elista, Elistinskoye knizscnoye
izdatelstvo: pp. 34–42.
1988 Djangr Site of the Eneolith Epoch [in Russian]. In I. V. Vasilyev (ed.),
Archaeological Cultures of the North Caspian Region. Kuibyshev, Kuibyshevskoye
knizhnoye izdatelstvo: pp. 44–58.
Minayeva, T. M.
1955 Site with Microlith Objects in the Black Lands [in Russian]. Short Reports of
the Institute of the Material Culture 59: pp. 46–53.
Mimokhod, R. A. and N. I. Shishlina
2004 14
C Data of the Final Catacomb Culture Graves of the Mandjikiny Burial
Ground and Problems of the Middle Bronze and Late Bronze Ages Chronology
[in Russian]. In L. T. Yablonsky (ed.), Ancient Caucasus: Retrospection of Culture.
Moscow, Institute of Archaeology RAS: pp. 124–127.
Novikova, M. A., A. G. Devyatov and I. N. Shishlina
2002 Carpological Investigation of the Bronze Age Sites of Kalmykia [in Russian].
In N. Shishlina and E. Tsytskin (eds.), The Ostrovnoy Burial Mound: Results of
the Complex Investigation of the Archaeological Sites of the North-West Caspian Sea
Steppe. Moscow, Elista Tisso Poligraf: pp. 196–203.
Shilov, V. P.
1975 Essay About the History of the Ancient Tribes of the Low Volga Area [in Russian].
Leningrad, Nauka.
Shishlina, N. I. (ed.)
1999 Textile of the Bronze Age Eurasian Steppe. Papers of the State Historical Museum.
Volume 109 [in Russian]. Moscow, Poltex.
Sinitsyn, I. V.
1931 Flint Objects from the Dune Sites of the Kalmykia Area [in Russian]. Bulletin
of the Low-Volga Institute of Local Historical Studies Named After M. Gorky: pp.
83–91.
Tajzscanova, G. E. (ed.)
1994 The Kazakhs: Historical-Ethnographic Research [in Russian]. Almaty,
Kazakhstan.
Zhitetsky, I. A.
1893 Essays About the Way of Life of the Astrakhan Kalmyks: Ethnographic Observation,
1884–1886 [in Russian]. Moscow, M. Volchaninov Publishing House: pp.
1–70.
Ch a p t e r 1 1
T rogodytes =
Blemmyes = Beja?
The Misuse of Ancient
Ethnography
S ta n l e y M . B u r s t e i n 1
T he correlation of textual and archaeological evidence is one of the

chronic problems of the historiography of the margins of the ancient
world. The eastern deserts of Egypt and Lower Nubia in late antiquity are no
exception. Archaeological remains in the Eastern Desert are typically associated
with the Blemmyes, who are considered typical desert nomads. This chapter
argues that the sources offer little support for this reconstruction; rather, they
indicate that in late antiquity the Blemmyes were primarily a sedentary people
located in the Nile Valley.
In 1982 Eric Wolf published an influential book entitled Europe and the
Peoples Without History. Wolf’s work was a pioneering effort to re-integrate
the experience of peoples on the margins of ‘civilizations’ into the narrative of
modern world history. This volume is part of a similar effort in ancient history.
Whether the topic is ancient or modern history, however, the historiography
of the margins is beset by similar methodological problems. Two in particular
stand out: the correlation of archaeological cultures with specific peoples
named in written sources and the validity of the ethnographic present.
The reasons for the persistence of these two problems are clear. The historical
record, with all of its problems and deficiencies, seems to provide the basis for a
rich and detailed diachronic account of developments within a region, replete
with references to individuals and ethnic groups and their conflicts and interac-
tions. By contrast, in the view of historians, the archaeological record cuts a poor
figure, with its anonymous architectural, funerary, tool and ceramic assemblages
often spread over wide swathes of territory and rarely capable of registering
I would like to thank Markus Wiener Publisher, Inc., for permitting me to reprint
1.
my translation of Agatharchides’ description of the Trogodytes, and Dr. R. H. Pierce,

University of Bergen, Department of Greek, Latin and Egyptology, for his comments
on earlier versions of this chapter.
250
T r o g o d y t e s = B l e mm y e s = B e ja ?      251
change in any mode except the longue durée. The temptation to unite the two
kinds of evidence has proved irresistible. One need only think of the endless
debate over the archaeological correlates of the Indo-Europeans (Mallory
1989:164–185) or the Celts and Germans (Wells 2001; Burns 2003:18–26);
indeed, the latter correlation seems to have ended in the paradoxical conclusion
that the distinction between the two peoples is illusory.
Figure 11.1. Map of southeast Egypt and northeast Sudan, showing the location of the places mentioned
in the text (drawn by H. Barnard).
252    S ta n l e y M . B u r s t e i n
The problem of the ethnographic present is equally vexing. Ideally, histo-

rians should have at their disposal multiple and comprehensive ethnographies
of the same ethnic group compiled at regular intervals and using the same
theoretical frameworks. Comparison would then allow historians to identify
and trace over time changes in socio-economic and cultural structures. The
reality, of course, is different. For many peoples, and especially those without
written native historical traditions, historians are fortunate if they have at their
disposal one ethnography and a handful of travelers’ accounts, supplemented
by accounts in the writings of foreign historians, geographers, and government
officials. As a result, the most authoritative ethnography of a people is treated
as universally valid for its whole history, creating an impression of changeless-
ness as characterizing the history of peoples living in nonstate societies. The
problem was first recognized in North America, where ethnographies of Native
Americans, compiled by anthropologists on 19th and 20th century CE reserva-
tions, were long erroneously treated as reliable descriptions of the cultures of
their precolumbian ancestors (Berkhofer 1978:62–69). Both problems bedevil
efforts to reconstruct the history of the peoples of the Eastern Deserts of Egypt
and the Sudan.
The Trogodytes 2
The Eastern Desert occupies a vast area bounded by the Nile in the west and
the Red Sea in the east and extending from Egypt in the north to Ethiopia and
Eritrea in the south. The historical record of this area is long and complex,
beginning in the 4th millennium BCE and extending to the present. This record
includes texts in numerous languages including Egyptian, Coptic, Greek,
Latin, Ge’ez, Arabic and a variety of European languages. The variety of the
textual record is as varied as the languages in which it is written, including royal
inscriptions, letters, ethnographies, travelers’ accounts, histories, geographies,
poetry and graffiti in several languages, to mention only the most prominent
categories. The one common denominator among these texts is that none of
them were written by the people who inhabit the Eastern Desert. All reflect
perspectives of the inhabitants of the Nile Valley, to whom the Eastern Desert
was a strange and hostile place but whose resources they sought and whose
inhabitants they tolerated or feared, depending on whether they needed their
services or feared their attacks. Only the meager surviving material record
was created by the local populations. Not surprisingly, correlating the mate-
rial and textual records is a central theme in the historiography of the Eastern
Desert. Obvious examples are the identification of the Medjay with the Pan
2.
For the spelling of Trogodyte see Burstein (1989:109, note 1).
Grave culture (Lacovara 2001:3, 20); the ascription of virtually all late ancient
non-Greco-Roman archaeological material in the Eastern Desert, including
Eastern Desert Ware, to the Blemmyes (Pierce 2001:157; Barnard 2002:53–57;
Sidebotham et al. 2004:16; but see Barnard 2005); and the closely related
controversy over whether the creators of X-Group Culture and the occupants
of the elite tombs at Ballana and Qustul, in the Lower Nubian Nile Valley,
were Nobatai or Blemmyes (Kirwan 1937:47–62; Adams 1977:419–422; Török
1987:219–221; Welsby 2002:20–23).
The validity and usefulness of such correlations is called into question by the
complex ethnography of the Eastern Desert region. Today the Eastern Desert
is primarily inhabited by the Beja, a loosely related conglomeration of tribal
groups united by geography, a lifestyle based mainly but not exclusively on
herding (some Beja are sedentary agriculturalists), and the fact that they speak
languages belonging to the Tu-Bedawie language family (Magid, this volume;
Paul 1954:12–19). The multiplicity of names for groups supposedly living in
the Eastern Desert in antiquity (Medjay, Bega, Bougaeitoi, blhm, Blemmyes,
Trogodytes and Megabari, to mention only the most prominent) suggests a
similar complexity in antiquity.
Not surprisingly, scholars have long attempted to simplify the ethnography
of the region by arguing that, like today, the Beja were also in antiquity the
dominant population in the Eastern Desert and that the various ethnic terms
mentioned above designated various Beja tribes (Krall 1900:6; Updegraff
1978:156–159). This hypothesis is made more attractive by the fact that the
various terms are language specific. Thus, Medjay occurs only in Middle and
New Kingdom Egyptian texts; blhm in Demotic texts; Blemmye in Greek, Latin
and late Egyptian/Coptic texts; Bega and Bougaeitoi in Aksumite Greek texts;
Trogodytes in Greek and Latin texts; Megabari in Greek, Latin and Egyptian
texts; and Beja in Arabic texts. The idea that the variety of ethnic terms is the
result of individual peoples giving different names to the same ethnic group is
an obvious solution to the problem and seemingly supported in the sources.
So the linguistic similarities of the terms Beja, Medjay and Bega, the explicit
equation of the Bega and Blemmyes by the 6th century CE Christian traveler
and geographer Cosmas Indicopleustes (Wolska-Conus 1968, volume 1:63),
and the equation of Beja and Blemmye in Arabic and Coptic texts (Plumley and
Adams 1974:278) appear to confirm this hypothesis. Finally, since the Bega,
Blemmyes and Trogodytes are all attested as possessing livestock (Eide et al.
1998:1094–1100, 1147–1153, 1158–1165), it has seemed safe to assume that,
like the modern Beja, the ancient Beja were also pastoral nomads. The principal
evidence for this thesis is the description of the Trogodytes preserved among
the fragments of On the Erythraean Sea of the 2nd century BCE Greek historian
Agatharchides of Cnidus (Burstein 1989:62–64, 1998:47–50):
Now the Trogodytes are called ‘Nomads’ by the Greeks and live a wandering
life supported by their herds in groups ruled by tyrants. Together with their
children they have their women in common except for the one belonging to the
tyrant. Against a person who has sexual relations with her the chief levies as a
fine a specified number of sheep.
This is their way of life. When it is winter in their country—this is at the time
of the Etesian winds—and the god inundates their land with heavy rains, they
draw their sustenance from blood and milk, which they mix together and stir in
jars, which have been slightly heated. When summer comes, however, they live
in the marshlands, fighting among themselves over the pasture. They eat those of
their animals that are old and sick after they have been slaughtered by butchers,
whom they call ‘unclean.’
They do not assign to any human being the appellation ‘parent’ but to a bull
and a cow, calling the former ‘father’ and the latter ‘mother’ and likewise to a ram
and a ewe because they do not obtain their daily sustenance from their parents
but from these beasts. For drink the mass of the people employs a preparation
made from the Christ’s thorn plant, but the tyrants drink one that is prepared
from a particular flower and is like poor quality sweet wine. They leave the rest
of their body naked but cover their loins with a girdle of skins. It is customary for
the other Trogodytes to circumcise their genital organs, just as do the Egyptians,
but the tribes the Greeks call ‘Colobi’ have the custom of cutting off with razors
during infancy the whole portion that others circumcise. From this practice they
gained for themselves the appellation just mentioned.
For armament the tribe of Trogodytes called Megabari have circular shields
made of raw oxhide and clubs tipped with iron knobs, but the others have bows
and spears.
He says that the Trogodytes deal as follows with the dead. They bind the
neck to the legs with withies made from the Christ’s thorn plant. Then, after
they place the body on a mound, they pelt it with stones large enough to be held
in a hand while jeering and laughing until they have hidden the corpse. They
then place on top (of the cairn) the horn of a goat and depart free of sadness
and completely cheerful. In conducting such a funeral, he says, they act sensibly
since to not cause themselves grief on account of those who are free from pain
is a sign of intelligence.
They do not fight with each other, as the Greeks do, over land or some other
pretext but over the pasturage as it sprouts up at various times. In their feuds, they
first pelt each other with stones until some are wounded. Then for the remainder
of the battle they resort to a contest of bows and arrows. In short time many die
as they shoot accurately because of their practice in this pursuit and their aiming
at a target bare of defensive weapons. The older women, however, put an end to
the battle by rushing in between them and meeting with respect. For it is their
custom not to strike these women on any account so that immediately upon their
appearance the men cease shooting.
They do not, he says, sleep as do other men. They possess a large number of
animals which accompany them, and they hand cowbells from the horns of all the
males in order that their sound might drive off wild beasts. At nightfall, they collect
their herds into byres and cover these with hurdles made from palm branches.
Their women and children mount up on top of these. The men, however, light
fires in a circle and tell traditional tales and thus ward off sleep, since in many
situations discipline imposed by necessity is able to conquer nature.
Those individuals who are unable to follow the herds because of age wind the
tail of a cow around their necks and willingly free themselves from life. But should
one seek to postpone death, anyone has the right to fasten the noose as though
from kindness and with a rebuke to deprive him of his life. It is likewise their
custom to remove from life those who have been crippled or are suffering from
incurable diseases. For they consider the greatest of evils to be for a person to
desire to live when unable to do anything that makes life worth living. Wherefore,
one can see that all Trogodytes are sound of body and still in the prime of life
since none is over sixty years of age.3
The Evidence Reconsidered

Despite its brevity, Agatharchides’ description of the Trogodytes is unique for
its high quality among Greek and Roman accounts of pastoral nomads. It is
not, of course, without flaws. It reveals clear evidence of one of the main weak-
nesses of classical ethnography: the tendency of Greek and Latin writers to
moralize about their own societies by holding up supposedly ‘simpler’ peoples
as models of natural behavior (Anderson 1938:ix-xix; Dihle 1962:207–232;
Hartog 1988:310–381). It also shows signs of a priori reasoning derived from
Agatharchides’ use of Peripatetic theories of sociocultural evolution to orga-
nize his accounts of Nubian peoples in his work (Burstein 1989:26–28). For
example, his insistence on the limited character of Trogodyte marriage, the
restriction of property to their livestock, and the fact that warfare among them
is directly connected to the protection of their herds’ access to pasture are
diagnostic features of the pastoral stage that forms the second stage of cultural
evolution in the Peripatetic scheme of cultural evolution. The pastoral stage is
marked by the breakup of the undifferentiated gatherer society of the ‘Golden
Age,’ which is represented in Agatharchides’ work by the Fish-eaters of the Red
3.
Reprinted with permission from the publisher. This translation is a synthesis of
the three parallel and overlapping surviving versions of Agatharchides’ account of the
Trogodytes.
Sea coasts of Egypt and Sudan. More important are two other characteristics
of Agatharchides’ ethnography of the Trogodytes.
First, Agatharchides’ Trogodyte ethnography is almost unique among
ancient and medieval accounts of nomads in that it lacks any trace of the two
defining characteristics of what the Roman historian Shaw (1982:6) calls “the
ideology of the pastoral nomad,” namely, “the complete separation of nomads
and sedentarists into two polarized and isolated taxonomic compartments” and
the characterization of the nomad “as the ultimate barbaric human type who
is directly opposed to the ‘civilized’ sedentary agriculturalist.”
Second, and equally important, numerous aspects of his description, such as
blood eating, the Trogodyte’s quasi-kinship relationship to their animals, the
existence of ‘pariah’ groups such as the butchers mentioned by Agatharchides,
and their funerary customs, are all paralleled in later ethnographies of east
African pastoralists (Burstein 1989:109–115). Clearly, Agatharchides’ claim
that his account was based on Ptolemaic explorers’ reports is justified, and it is
a sound basis for reconstructing the culture of the Trogodytes. But can it also
be used to reconstruct the life of the Blemmyes in late antiquity?
The identification of the Blemmyes as a Beja tribe has led scholars to assume
it can (Paul 1954:35; Updegraff 1978:192), but there is reason to believe that
this is not true. Thus, while it is probable that the Blemmyes were ethnically
Beja,4 not all modern Beja are pastoral nomads, and there is no reason to
assume that the situation was different in antiquity.5 Equally important, no
ancient source identifies the Trogodytes with the Blemmyes. Moreover, even
if one did, Greek and Roman writers were as susceptible to the pitfalls of
the ethnographic present as modern scholars. Once a standard ethnography
appeared, it tended to be copied century after century, leaving the described
peoples frozen in time and space. For example, India in Greek literature was
always north India in the 3rd century BCE (Dihle 1962, 1963), and Gaul was
always southern France in the late 2nd and early 1st century BCE (Tierney
1960:189–275; Nash 1976:111–126).
The case is similar for the Trogodytes. Five centuries and hundreds of
miles separate the Trogodytes described by Agatharchides from the Blemmyes
4.
For evidence indicating that the Blemmyes spoke a language of the Tu-Bedawi
language family see Zhylarz 1940–1941:1–2 and Browne 2004:243–244.
5.
A good example of the misuse of the ethnographic present with regard to the
Blemmyes is Pierce’s comment (2001:159) that “the primary reason for believing that
the Blemmyes were nomadic in life-style and had a clan-based (segmentary) social
structure is the evidence that links them to the Beja, who undoubtedly fit this descrip-
tion,” despite the hundreds of years that separate them from the earliest medieval
accounts of the Beja.
encountered by the Romans in Lower Nubia and Egypt in late antiquity. The
conditions in which the peoples of the Eastern Desert lived changed signifi-
cantly in those centuries, but Agatharchides’ account continued to stand as the
standard ethnography of the Trogodytes until the end of antiquity (Burstein
1989:33–36). Agatharchides described the Trogodytes on the basis of the
reports of Ptolemaic explorers, who encountered them in the hill country
adjacent to the Red Sea coasts of the central Sudan during the early phases
of Ptolemaic elephant hunting in the mid-3rd century BCE. Later authors
repeated his account for the rest of antiquity. The Ptolemaic footprint in the
Eastern Desert at the time that Agatharchides’ sources were compiled was light
and mostly limited to the road linking Berenike to Koptos, a few forts, the
gold mines in Wadi Allaqi and some ports and hunting stations on the Red Sea
coast. In contrast, during the Roman period trade caravans and supply convoys
regularly crossed the region using one of several roads linking the Nile Valley
with the Red Sea coast and with the watering stations, forts, mines, quarries
and settlements of various types that were scattered throughout the region.
These circumstances make it unlikely that Agatharchides’ description of the
Trogodytes could also accurately characterize the life of the Blemmyes four
or five centuries later.
For almost half a century a number of scholars, including Herzog
(1967:57–58), Christides (1980:134–136), and, most recently, the Sudanese
scholar Samia Dafa'alla (1987:34–39), have protested against the tendency
to reconstruct the culture of the Blemmyes on the basis of their presumed
identification with the Trogodytes. The most comprehensive of these critiques
is that of Dafa'alla, who argued in an important but little noticed article that
there were two groups of Blemmyes, one nomadic and one, more important,
sedentary group living near the Nile in Lower Nubia and southern Egypt.
Dafa'alla suggested that it was this latter group to which most classical texts
referred. Analysis of the evidence for the political organization and location
of the Blemmyes in late antiquity confirms the validity of much, but not all,
of this reconstruction.
The textual evidence concerning the Blemmyes’ political organization
in late antiquity is remarkably varied, including texts in Greek, Latin and
Coptic, in a wide variety of categories. Included among these texts are tax
records, royal decrees, letters, inscriptions, graffiti and allusions in historical
and hagiographical works. The picture that emerges from them, however, is
clear and consistent.
While it is true that the Greek geographer Strabo (Eide et al. 1996:828–835)
included the Blemmyes among the nomadic peoples of Nubia in the late 1st
century BCE, it is equally true that the focus of Blemmye life in the 5th century
CE had shifted from the desert to a group of towns in lower Nubia. Around
390 CE the theologian Epiphanius (Eide et al. 1998:1115–1121) claimed that

the Blemmyes controlled the emerald mines of Mons Smaragdus from their
capital at Talmis (Kalabsha), while a quarter of a century later, around 420 CE,
the ambassador and historian Olympiodorus (Eide et al. 1998:1126–1128),
who visited the Blemmyes, described them as barbarians occupying Talmis
and holding also the towns of Primis (Qasr Ibrim), Phoinikon, Khiris and
Thapis (Taphis). Later in the century the Nobatian king, Silko (Eide et
al. 1998:1147–1153), boasts of seizing Talmis and Taphis from them and
conquering their towns from Primis to the first cataract; while the Blemmyan
king, Phonen (Eide et al. 1998:1158–1165), demands that Silko’s successor,
Abourni, return Talmis, and the rest of their territory, to the Blemmyes. There
is textual evidence not only for town officials (Eide et al. 1998:1134–1138) but
also for various royal officials, who seem to be arranged in a hierarchy including
kings (basiliskoi), tribal chiefs (phylarchoi), subchiefs (hupotyrannoi), court offi-
cials (domestici) and scribes (Updegraff 1978:144, 180–181). The kings had the
power to levy taxes and grant exemptions, as well as authority over territory
(Eide et al. 1998:1207–1210). As Herzog (1967:58) pointed out almost half
a century ago, these features of Blemmye political organization indicate that
in late antiquity the Blemmyes were no longer stateless nomads of the type
described in Agatharchides’ account of the Trogodytes but a people with “a
state organization with nobles ruling islands” and “a royal court on the model
of a sedentary population with specialized subordinate officials.”
The situation is similar with regard to the evidence for the territory of
the Blemmyes in the 4th–5th centuries CE. In his important monograph Late
Antique Nubia Török (1987:28) claimed that the Blemmyes lived in “the area
of the Red Sea hills” and that “late classical sources locate their tribes in the
Eastern Desert.”6 The actual situation, however, is exactly the reverse of this
statement. With one exception (the 6th century CE historian Procopius) sources
consistently place the Blemmyes in the Nile Valley. Five sources are of relevance
here: the 4th century CE theologian Epiphanius, the 5th century CE historian
Olympiodorus, the late 4th century CE historian Ammianus Marcellinus, the
contemporary Latin poet Claudian, and the biography of the Egyptian monk
Shenute compiled by Besa. Epiphanius’s and Olympiodorus’s references to the
Blemmyes occupying cities in the Nile Valley have already been discussed. Even
more revealing are two passages of Ammianus Marcellinus. In paragraph 14.4.3
Ammianus describes the inhabitants of the Eastern Desert of Egypt, whom he
calls Saracens. He treats this term as a synonym for the term Scenitae, ‘tent-
dwellers’ (Ammianus Marcellinus 22.15.2, Rolfe 1935-1939):
His claim that “some accounts place them falsely ‘near the source’ of the River
6.
Nile” is arbitrary (Török 1987:28).

Among those tribes whose original abode extends from the Assyrians to the cata-
racts of the Nile and the frontiers of the Blemmyes [ad Nili cataractas porrigitur,
et confinia Blemmyarum] all alike are warriors of equal rank, half-nude, clad in
dyed cloaks as far as the loins, ranging widely with the help of swift horses and
slender camels in times of peace or war. No man ever grasps a plough-handle or
cultivates a tree, none seeks a living by tilling the soil, but they rove continually
over wide and extensive tracts without a home, without fixed abodes or laws.
Two facts emerge from this text. First, Ammianus explicitly excluded the
Blemmyes from the category of tent-dwelling nomads whom he calls Saracens.7
Second, his linking of the Blemmyes’ territory to the cataracts clearly places it
in the Nile Valley. Confirmation is provided by a passage in paragraph 22.15.24.
Commenting on the disappearance of hippopotami from Egypt as a result of
overhunting, Ammianus observes that “as the inhabitants of these regions
(Egypt) conjecture, they [the hippopotami] were forced from weariness of the
multitude that hunted them to take refuge in the land of the Blemmyes.” Given
the hippopotami’s need for water, Ammianus clearly implies that the Blemmyes
live near the Nile and not in the desert. The evidence of Claudian and Besa is
similar. The former (Eide et al. 1998:1125–1126), in describing the course of
the Nile, says that the river “winds through Meroe and the fierce Blemmyes and
black Syene,” while the latter (Eide et al 1998: 1107–1109) describes a Blemmye
raid on the White Monastery as coming from ‘the south,’ not the east.
The evidence of the majority of the late ancient sources is clear: the terri-
tory of the Blemmyes in the 4th–5th centuries CE lay in the Nile Valley, not in
the Eastern Desert. As mentioned, the only exception is Procopius, but the
exception is only apparent. In describing the distribution of peoples south
of Egypt in his own time Procopius (Eide et al. 1998:1188–1193) says that
“the Blemmyes inhabit the interior of this country (Lower Nubia), while the
Nobatai possess the lands on either side of the River Nile.” He points out later
in the same passage, however, that the Nobatai originally lived in the oasis of
Kharga but settled in the Nile Valley at the invitation of Emperor Diocletian,
who wished them “to drive off the Blemmyes.” The historicity of Procopius’s
account of how the Nobatai gained control of the Lower Nubia is doubtful
(Updegraff 1978:52–54; Eide et al. 1998:1191–1193), but it is clear that he
believed that the Blemmyes originally lived near the Nile and that they took
refuge in the Eastern Desert only after their defeat by the Nobatai, specifically
by Silko and his successors.
7.
Török’s (1987:32) identification of the Saracens with the Nabateans, the inhabit-
ants of northwestern Arabia, ignores Ammianus Marcellinus’s explicit statement that
their territory bounded Egypt on the east (Ammianus Marcellinus 22.15.1, Rolfe
1935-1939).
Discussion
The evidence reviewed in this chapter suggests a different course of Blemmye
history than that generally found in standard accounts. The emphasis in the
sources on Blemmye raids into Egypt has led to the simplistic characteriza-
tion of them as typical desert nomads, obscuring a more complex historical
trajectory. The Blemmyes probably did originate as nomads living east of the
Nile in the central Sudan under Kushite suzerainty. That is at least how the
geographer Eratosthenes (Eide et al. 1996:557–561) described them in the
3rd century BCE. Aksumite sources indicate that their relatives (the Bega/
Bougaeitoi) continued to live that way.
The later history of the Blemmyes, however, differed from that of the
Bega. While Strabo (Eide et al. 1996:828–835) in the early 1st century CE
and Ptolemy, in Geography 4.7 (Berggren and Jones 2000), in the 2nd century
CE still placed the Blemmyes in the central Sudan, by late antiquity they had
relocated into the Nile Valley in Lower Nubia, possibly as part of the Kushite
colonization of the region. There they underwent a process of sedentarization.
As Roman power in the area weakened, the Blemmyes occupied the principal
towns of the region, establishing a state based on a mixed agricultural and
herding economy that was strong enough by the end of the 4th century CE to
seize control of the ‘emerald’ mines in the Eastern Desert.
This review of the textual evidence concerning the political organization and
territory of the Blemmyes supports the identification of a riparian Blemmye
population in lower Nubia but not the existence of a separate and distinct popu-
lation of nomadic Blemmyes in the region as suggested by Dafa'alla. Instead,
it suggests that the expansion of Blemmye influence into the Eastern Desert
occurred after they had moved into Lower Nubia. This represented their
opportunistic adaptation to the opportunities afforded by Roman development
in the region, specifically the growth of mining and caravan traffic between
the Nile and the Red Sea. It was only after their defeat by the Nobatai in the
second half of the 5th century CE that they were forced to abandon the Nile
Valley and withdraw into the Eastern Desert. Procopius located them there
in the 6th century CE, and other papyri and Arabic sources indicate that they
survived as a distinct population until at least the 8th century CE before finally
merging with the populations of the region.
If this reconstruction of Blemmye history is correct, it has important impli-
cations for the study of archaeological material in the Eastern Desert, including
Eastern Desert Ware (Barnard 2002). It is clear that Agatharchides’ account
of the Trogodytes in east-central Sudan during the 3rd century BCE cannot be
used as a basis for reconstructing the life of the Blemmyes living in late antique
Lower Nubia. This significantly reduces the ethnographic evidence relevant to
Blemmye society. If Eastern Desert Ware, and other Eastern Desert artifacts,
was created by the Blemmyes, it was not created by people who were primarily
nomads. Alternately, if Eastern Desert Ware was created by nomads, then it
cannot be used as diagnostic for the presence of the Blemmyes.
This conclusion should not be surprising. The assumption that the Blemmyes
were the sole inhabitants of the Eastern Desert was always a gross oversim-
plification that obscured more than it explained. As mentioned, Ammianus
described the Eastern Desert as occupied by a variety of nomadic tribes that
he called ‘Saracens.’ His account of the Saracens is, unfortunately, marred by
‘pastoral ideology,’ but the general validity of Ammianus’s picture of conditions
in the Eastern Desert is not in doubt. Even after they left the Nile Valley, the
Blemmyes were only one of the many peoples occupying Lower Nubia and
its desert hinterlands. As elsewhere, in the Eastern Desert the correlation of
archaeological evidence with peoples mentioned in textual sources is never
simple and straightforward but always problematic and uncertain.
References
Adams, W. Y.
1977 Nubia: Corridor to Africa. Princeton, Princeton University Press.
Anderson, J. G. C. (ed.)
1938 Cornelii Taciti De origine et situ Germanorum. Oxford, Clarendon.
Barnard, H.
2002 Eastern Desert Ware: A First Introduction. Sudan & Nubia 6: pp. 53–57.
2005 Sire, il n’y a pas de Blemmyes: A Re-evaluation of Historical and Archaeological
Data. In J. C. M. Starkey (ed.), People of the Red Sea: Proceedings of the
Red Sea Project II, Held in the British Museum, October 2004. Society for
Arabian Studies Monographs number 3. BAR International Series 1395.
Oxford, Archaeopress: pp. 23–40.
Berkhofer, R. F., Jr.
1978 The White Man’s Indian: Images of the American Indian, from Columbus to
the Present. New York, Random House.
Berggren, J. L. and A. Jones
2000 Ptolemy’s Geography: An Annotated Translation of the Theoretical Chapters.
Princeton, Princeton University Press.
Browne, G. M.
2004 Blemmyica. Zeitschrift für Papyrologie und Epigraphik 148: pp. 243–244.
Burns, T. S.
2003 Rome and the Barbarians, 100 B.C.–A.D. 400. Baltimore, Johns Hopkins
University Press.
Burstein, S. M. (ed.)
1989 Agatharchides of Cnidus on the Erythraean Sea. London, Hakluyt Society.
1998 Ancient African Civilizations: Kush and Axum. Princeton, Markus Wiener.
Christides, V.
1980 Ethnic Movements in Southern Egypt and Northern Sudan: Blemmyes-Beja
in Late Antique and Early Arab Egypt Until 707 A.D. Listy Filologické 103:
pp. 120–143.
Dafa'alla, S.
1987 The Historical Role of the Blemmyes in Late Meroitic and Early X-Group
Periods. Beiträge zur Sudanforschung 2: pp. 34–40.
Dihle, A.
1962 Zur Hellenistische Ethnographie. In H. Schwabl, H. Diller, O. Reverdin, W.
Peremans, H. C. Baldry and A. Dihle (eds.), Grecs et Barbares: Entretiens
sur l’antiquité classique. Geneva, Vandoeuvres-Genève: pp. 205–232.
1963 The Conception of India in Hellenistic and Roman Literature. In Proceedings
of the Cambridge Philological Society 190: new series 10, pp. 15–23.
Eide T., T. Hägg, R. H. Pierce and L. Török
1996 Fontes Historiae Nubiorum, Volume 2. University of Bergen, John Grieg.
1998 Fontes Historiae Nubiorum, Volume 3. University of Bergen, John Grieg.
Hartog, F.
1988 The Mirror of Herodotus: The Representation of the Other in the Writing
of History. Berkeley, University of California Press.
Herzog, R.
1967 Zur Frage der Kulturhohe und der Wirtschaftsform der Frühen Bedja.
Paideuma 13: pp. 54–59.
Kirwan, L.
1937 Studies in the Later History of Nubia. Liverpool Annals of Archaeology and
Krall, J.
1900 Beiträge zur Geschichte der Blemyer und Nubier. In Denkschriften der
Kaiserlichen Akademie der Wissenschaften: Philosophisch-Historische Klasse.
Volume 47. Vienna, Kaiserlichen Akademie der Wissenschaften: pp. 1–26.
Lacovara, P.
2001 Pan-Grave People. In D. B. Redford (ed.), The Oxford Encyclopedia of
Ancient Egypt. Volume 3. Oxford, Oxford University Press, pp. 20–22.
Mallory, J. P.
1989 In Search of the Indo-Europeans: Language, Archaeology, and Myth. London,
Thames and Hudson.
Nash, D.
1976 Reconstructing Poseidonios’ Celtic Ethnography: Some Considerations.
Britannia 7: pp. 111–126.
Paul, A.
1954 A History of the Beja Tribes of the Sudan. Cambridge, Cambridge University
Press.
Pierce, R. H.
2001 Past and Present in the Eastern Desert. In K. Krzywinski and R. H. Pierce,
Deserting the Desert: A Threatened Cultural Landscape Between the Nile
and the Sea. Larvik, Alvheim, and Eide Akademisk Forlag: pp. 143–169.
Plumley, J. M. and W. Y. Adams
1974 Qasr Ibrim, 1972. Journal of Egyptian Archaeology 60: pp. 212–278.
Rolfe, J. C.
1935–1939 Ammianus Marcellinus (3 volumes). Cambridge, Harvard University
Press.
Shaw, B.
1982 Eaters of Flesh, Drinkers of Milk: The Ancient Mediterranean Ideology of
the Pastoral Nomad. Ancient Society 13: pp. 5–31.
Sidebotham, S. E., H. M. Nouwens, A. M. Hense and J. A. Harrell
2004 Preliminary Report on Archaeological Fieldwork at Sikait (Eastern Desert of
Egypt) and Environs: 2002–2003. Sahara 15: pp. 7–30.
Tierney, J.
1960 The Celtic Ethnography of Posidonius. In Proceedings of the Royal Irish
Academy (section C) 60: pp. 189–275.
Török, L.
1987 Late Antique Nubia: Antaeus. Communicationes ex Instituto Archaeologico
Academiae Scientiarum Hungaricae 16. Budapest, Archaeological Institute of
the Hungarian Academy of Sciences.
Updegraff, R. T.
1978 A Study of the Blemmyes. Brandeis University (PhD dissertation).
Wells, P. S.
2001 Beyond Celts, Germans, and Scythians. London, Duckworth.
Welsby, D. A.
2002 The Medieval Kingdoms of Nubia: Pagans, Christians, and Muslims Along
the Middle Nile. London, British Museum Press.
Wolf, E.
1982 Europe and the Peoples Without History. Berkeley, University of California
Press.
Wolska-Conus, W. (ed.)
1968 Cosmas Indicopleustès: Topographie chrétienne. Tome 1. Paris, Les Éditions
du Cerf.
Zhylarz, E.
1940–1941 Die Sprache der Blemmyer. Zeitschrift für Eingeborenen-Sprachen 31:
pp. 1–21.
Ch a p t e r 1 2
Is A bsence of Evidence ,
Evidence of A bsence?
Problems in the Archaeology
of Early Herding Societies of
Southern Africa
Andrew B. Smith
T he Khoekhoen (known to the world at large as ‘Hottentots’) became

one of the earliest known ‘primitive’ people that existed beyond medieval
Christendom. While the African “monstrous races” of Pliny (Friedman 1981)
seem to have been based on real people, their fanciful descriptions made them
more myth than reality. The Khoekhoen, by contrast, although probably within
Pliny’s category of ‘anthropophages’(man-eaters), were people whose ‘savage’
cultural attributes could be (and were) described and repeated (plagiarized) ad
nauseam in the travel literature (Raven-Hart 1967), from the early 15th century
CE onward (Smith 1993a). They were initially seen to be closest to what human
beings would have been like before the Christian ‘original sin’ (living in the
Garden of Eden), as can be seen in the comparison between Albrecht Dürer’s
Adam and Eve and Hans Burgkmair’s people living at the end of Africa (Figures
12.1 and 12.2). Perhaps this is not surprising, since Dürer and Burgkmair were
members of the same artist’s club in Augsburg, Germany (Smith and Pasche
1997). Later, like Rousseau, Denis Diderot questioned one of the more enlight-
ened travelers among the Khoekhoen, Col. Robert Gordon, in The Hague in
1774 (Cullinan 1989) and used the information to support his ideas on the ‘noble
savages’ in the Pacific. Cullinan (1989:150) shows that there is good evidence that
Diderot, contrasting the colonial Dutch lifestyle with that of the Khoekhoen,
said the Khoe dwell in “the happiness, innocence, and tranquility of a patriarchal
life.” The Khoekhoe press, unfortunately, was not able to sustain this mythical
aura, and they became archetypical savages up to the 20th century, as the lowest
level of humans in the ‘Great Chain of Being’ (Lovejoy 1942). Even the eminent
anthropologist Alfred Kroeber, writing in the 1920s, used them as an example of
the argument why no geniuses have shown up in ‘primitive’ society, whose cultural
environment is ‘atrophied and sterile,’ noting that the extreme southwest of Africa
“is in possession of the backward Bushmen and Hottentots” (Kroeber 1923:96).
264
Herding Societies of S o u t h e r n A f r i c a      265
Figure 12.1. Adam and Eve (engraving by Albrecht Dürer, 1504).
The early travel literature is full of descriptions of the large herds of animals
owned by the Khoekhoen, and although they were initially eager to trade their
stock for pieces of iron, by the beginning of the 17th century the market was so
glutted that the Khoe began demanding more valuable metals, such as copper
and brass. Sometimes they even refused to trade with the sailors, as seen in the
complaint of John Jourdain in 1617:
In the tyme of our beeinge in this roade wee could not get any refreshinge after
the first daie, although the people came downe with great store of cattle and
sheepe . . . The next daie after our arrival they brought downe above 5000 head
of cattle . . . butt of these cattle they would not sell any unlesse wee would goe
with our people to Cories [a Khoe who could speak English] house . . . when we
were come to the toppe of the hill some four miles from the tents, we saw in the
valley aboute 10,000 head of cattle (Raven-Hart 1967:87–88).
266     A n d r e w B . S m i t h
Even if the numbers were exaggerated, there were no doubt large herds in
Khoe hands.
By contrast, the Khoekhoen, so well known in the historic literature, are
virtually invisible in the archaeological record. No sites from the period after
around 800 BP are recognizable as specifically ‘herder’ and, with the excep-
tion of 17th century CE Dutch colonial structures such as the castle at Cape
Town and the redoubt at Oudepost (Schrire 1988), no site has produced large
numbers of cattle bones. We can assume that many of the bones found associ-
ated with these fortified structures were from animals traded from the Khoe.
Small rockshelters inhabited by aboriginal people on the edge of the coastal
plain were occupied during the colonial period. One such site, Voelvlei (Smith
et al. 1991), which also yielded colonial artifacts, was probably not occupied by
herders, since the cultural material would suggest they were hunters, although
Figure 12.2. Khoekhoe family at Algoa Bay (woodcut by Hans Burgkmair for Balthasar Springer’s Journal
Die Merfart und Erfarung nuwer Schjffung und Wege, published in 1508 in Munich).
one attribute of the herders, large ostrich eggshell beads, was found. This
suggests separation, but contact, between hunters and herders even up to the
colonial period (Smith 1998).
Extensive surveys for Khoe sites have been carried out, even in areas where
they were recorded in the historic literature (Hart 1987), but none have been
found. We have to accept that their archaeological ‘invisibility’ must be due to
their mobility and transience across the landscape, as well as the fact that the
coastal forelands are also where modern agriculture takes place. Deep plowing
of the soil may have obscured signs of transient camps. This appears to have
been less so when sheep were the economic base, although the sites are only
found on the coast where marine shell makes them more visible (Smith 1987,
1993b; Sealy et al. 2004). A cattle-based economy and the needs of bulk grazers,
particularly in a low-nutrient area like the Western Cape, must have meant
the herds were probably constantly on the move, and little cultural residue
was being left behind.
Material Culture and

Archaeological Visibility
Even though the Khoekhoen were massively disrupted by colonial intrusion in
the 17th and 18th centuries CE, a picture of their material culture can be gleaned
from historic writings. The center of a woman’s world was her hut, over which
she had absolute control (Smith and Webley 2000). These huts were made
from bent saplings, dug into the ground to form a structure across which were
placed reed mats to form a dome shape. An encampment in the past would have
consisted of a number of these huts laid out to form a ring, leaving an open
space in the middle to keep the stock at night (Kolb 1719:Tab XIV, Figure 1,
opposite p. 470; Smith 1992:Figure 8.6). Like most modern African pastoral-
ists, the Khoekhoen were very concerned about their personal appearance. As
can be seen from the Khoekhoen depicted in Figure 12.3, probably drawn at
the end of the 17th century CE, women would have spent a great deal of their
time making personal decorative items, which is why copper and brass became
so popular as trade items in the 17th century CE.
Robertshaw (1978) was first to recognize that modern Nama (a group of
Khoekhoen) camps along the Orange River (the northern border of South Africa
with Namibia) were so ephemeral that they quickly leave no trace once they are
abandoned. He describes one such encampment at Sendeling’s Drift:
Though the site was occupied within the last twenty years, the position of the
dwelling hut or huts could only be estimated from the small amount of wood or
matting that remained. As this will undoubtedly perish within a few years no trace
268     A n d r e w B . S m i t h
of any huts will survive for future archaeologists to excavate. The hut(s) also leave
no post-holes or other features in the ground. A similar situation exists for the
kraal in which livestock were kept. Once the perimeter fence has gone and the
dung within it decayed nothing will remain (Robertshaw 1978:29).
Figure 12.3. Khoekhoe women interacting with colonists at the Cape. Unknown artist, probably drawn
at the end of the 17th century CE (Smith and Pheiffer 1993).
More recent work along the river (Webley 1997; Smith et al. 2001) showed that
when stone tools were still being used, herder sites could be identified. In part
this was due to massive amounts of silt covering the sites, before flood-control
dams were built along the river, and subsequent erosion of the terraces. Once
stone cutting tools were replaced by metal ones, the preponderance of wood for
making bowls and herding equipment (as well as leather for bags, dress, etc.) all
compounded the problem of survival of material culture in the archaeological
record. This is exemplified by the Khoekhoe exhibit in the Ethnographic
Gallery of the South African Museum, Cape Town, which has only a handful
of items surviving from the long history of these people. By the middle of the
18th century CE the Khoekhoen as independent herders were becoming very
sparse, and only a few material culture objects of known Khoe manufacture
from then have survived in European museums (Rudner and Rudner 1957).
From an archaeological perspective we can assume ceramics would play
an important role in identifying Khoe sites. Indeed, when sites of herders
can be found during the sheep period, such as at Kasteelberg, there are large
numbers of sherds. Away from these coastal areas, however, the distribution
of sherds is meager, possibly a result of the more intense use of coastal sites,
with the attraction of marine resources (Smith 1993b), but also because of the
constant movement across a landscape that today is the main agricultural area
of the Cape.
The Earliest Herders in Souther n Africa

Having noted the case for the problem of visibility of pastoralists at the Cape,
we can turn our attention to a similar problem, that of identifying the earliest
food producers in southern Africa. It used to be assumed that the first domestic
animals were introduced by incoming Iron Age farmers from the north (Smith
1990a, 1990b). It is now clear, however, that sheep bones found in association
with cultural material that indicates aboriginal hunters of the Western Cape
predate the arrival of the Iron Age farmers by possibly several centuries. Sites
such as Spoegrivier Cave (Webley 2001), Blombos Cave (Henshilwood 1996),
Die Kelders Cave (Schweitzer 1979), and Kasteelberg G, an open site (Sadr et
al. 2003), all have dates between 2000 and 1900 BP for sheep bones and pottery.
The question is: “Were these mostly small cave sites occupied by herders?” Sadr
(2003), taking the most parsimonious line with the available archaeological data
of sheep bones associated with cultural material that would identify aboriginal
hunters (Smith et al. 1991), would say that these were hunters independently
becoming shepherds and were thus the earliest food producers at the Cape. But
what if the scenario posed above with the later historical Khoekhoen pertained
at this early period as well, and there were herders living on open sites who
moved across the landscape leaving little archaeological residue, but in turn
were the donors of sheep to the hunters? Sadr does not address how the sheep
in the caves came to the Cape. He argues that diffusion of stock could have
occurred and interprets the advent of small stock in rockshelters in Botswana
to support his case (Sadr and Plug 2001). Diffusion among the hunters could
have occurred via internal exchange mechanisms, such as hxaro (Wiessner 1994)
seen among the Ju/’hoansi Bushmen today.
Prior to the advent of food production in southern Africa the entire subcon-
tinent was occupied by hunter-gatherers. These Later Stone Age hunters were
part of a hunting lifestyle trajectory that began more than 20,000 years ago.
It is into this population that food producers first introduced domestic sheep,
and some of these hunter-gatherers at least made the transition to become food
producers themselves and became the Khoekhoen. In spite of the fact that all
the aboriginal people of southern Africa appear to form a unique biological
population, there are three distinct Khoisan languages. All are ‘click-based,’ but
are mutually unintelligible, which suggests a long period of separation. Today
these are found among the best known Ju/’hoansi (or !Kung speakers) in the
Kalahari and southern Angola, among the few remnants of Cape Bushmen
descendants who still remember the language (the !Ui-Taa), and among Khoe
speakers in Botswana (where the Khoekhoen existed historically).
Linguistic data suggest that the words for ‘ram,’ ‘milk ewe,’ and ‘young
ram’ are loan words only into Khoe languages. According to Ehret (1998)
270     A n d r e w B . S m i t h
these came via a putative donor East Sahelian language from the north (Ehret
1998), who might be best represented archaeologically by Bambata, a ceramic
cultural entity identified in southern Zambia and Zimbabwe. It is unclear
exactly what this tradition is, or how it fits with the Early Iron Age, although
there appears to be little doubt of its affinity (Huffman 1994). Our ‘invisibility’
problem may also be at play here. The best evidence for pastoral activity in East
Africa comes from Ngamuriak, in Kenya, close to the border with Tanzania
(Robertshaw 1978). This site has minimally decorated ceramics with spouts.
Pastoral people are often less concerned with decoration on their vessels than
agriculturalists, who imbue their pots with symbolic significance (David et al.
1988; Hall 1998). Hodder (1991) has shown how calabash decoration among
East African pastoral groups is often more for personal identification than for
group symbolism, although certain statements might be made when decoration
is used, such as young men and women opposing elder hegemony among the
Il Chamus, or possibly resisting male control among the Khoekhoen (Smith
and Webley 2000).
The artificial colonial boundary that is the modern state frontier between
Kenya and Tanzania would not have inhibited pastoral movement in the past,
but we know much less about the archaeology of pastoral societies in Tanzania
and Zambia, so any possible southward expansion may have been masked by
research into the more dominant, ceramic-rich, Iron Age farming occupa-
tions of these countries. One might postulate that small groups of herders
colonizing new areas would have been even less likely to leave significant
archaeological residues; thus our ‘invisibility’ scenario may mean this is why
we do not see the archaeological contacts between pastoral groups in eastern
and southern Africa.
Alternatively, if Bambata is indeed an Iron Age entity, perhaps it is to
agrarian societies arriving in southern Zambia and Zimbabwe that we have to
look for the donor society. Regardless of the donor, the existence of a tsetse-free
corridor from East Africa (Figure 12.4) would have facilitated the southward
movement of domestic stock.
Morris (1992) has analyzed skeletal material from both hunting-gathering
and farming contexts and has found no Khoisan individuals north of the
Zambezi. This may be partially because of the limited number of skeletons
available, but it could well mean that the domestic animals had to come to the
Khoisan along the Zambezi and into the Caprivi with pastoral groups from the
north. Thus, the initial entry would have been from outside groups.
There is no archaeological information to tell us what happened then.
The closest we can come is from the distribution of ripple-rim ceramics,
which occur across a large area of southern Africa, from northern Namibia to
the Limpopo Province of South Africa (Figure 12.4). These are from small,
Figure 12.4. Estimated route of movement of domestic animals into southern Africa.
well‑made vessels but consist of only a handful of sherds (Smith and Jacobson
1995), and we are unsure how they fit with the Bambata ceramics. We are
also not sure if their distribution was the result of spread by Khoe-speakers
who took on the package of sheep and pottery and who subsequently spread
through the river systems of the Vaal/Orange drainages, ultimately arriving in
the Western Cape. Who were these first herders? Is it possible that we have a
‘bow-wave’ of small numbers of Khoe-speaking herders bringing their stock
to the Cape and being the donors of the animals found in the caves?
As an alternative, Sadr (2003) would like to see the local hunters becoming
‘Neolithic’ food producers independent of Khoe speakers. He uses the open
site of Kasteelberg G, which is occupied before and after the appearance of
sheep and ceramics around 2000 BP, to support his thesis (Sadr et al. 2003).
This pattern is repeated at Witklip Cave, where there is no change in the
272     A n d r e w B . S m i t h
material culture across the food-production divide (Smith et al. 1991). The
beads from the cave sites continue to be small (>5 mm) before and after the
appearance of sheep bones (Smith 1998). The small number of sherds (10/
m3), and the fact that none of the rim-sherds can refit, suggests they might be
coming in from other groups.
Archaeologically recognizable herding groups with substantial numbers of
sheep are visible at Kasteelberg A site by at least 1630 BP (Sealy and Yates 1994)
and possibly as early as 1860 BP (Smith 1987). The ostrich eggshell beads from
these herder sites are large (<5 mm), and, since they appear abruptly, we might
infer that they came via immigrant groups.
Theories of Colonization
The spread of food production and Neolithic farming techniques has been the
subject of great debate. Previous assumptions were that farming groups spread
out of the Near East and swamped Mesolithic communities in Europe (Piggot
1965). This was seen as an incremental intrusion and colonization by migrants.
More recent work has suggested that local Mesolithic hunters adopted farming
themselves and transformed the social landscape (Gronenborn 2003). We thus
have several models of diffusion or independent adoption of food produc-
tion. The latter gives credit to indigenes who are capable of adopting new
economic ideas and changing their lifestyle. We also have to accept that ideas
and commodities do not travel in a vacuum, however, so the people involved
may also be spreading and interacting with indigenes. The degree to which
incorporation, or not, takes place may depend on competition for resources.
Alexander (1984) has suggested that colonization takes place in two steps:
a movable frontier where small groups of immigrants, initially acceptable to local
people, move into new territory (in the case of expansion of early food produc-
tion this would mean entering areas already occupied by hunter-gatherers);
a static frontier where the immigrants consolidate their position in the landscape
and start to dominate the indigenes (by this time all the best land is being used,
and boundaries inhibiting further expansion have been reached; more marginal
lands will be progressively occupied, particularly by people being pushed to
the periphery of the more dominant society).
This model seemed to conveniently fit the expansion of Europeans into the
American West and the subjugation of the Native Americans by a combination
of firepower and merchant capital (Turner 1920), although Mikesell (1960)
showed how contentious these ideas became and discussed frontiers of assimila-
tion and exclusion. How relevant might the model be for southern Africa? The
model of exclusion might fit the expansion of Europeans out of the original
settlement at the Cape in the 17th and 18th centuries CE, where, like America,
racial attitudes on the part of Europeans kept the cultural groups apart (except
for the initial window when there were few European women either on the
American or South African frontiers and men would take local or slave wives).
In South Africa the power of the church quickly closed ranks against interracial
marriages, and a society developed along both class and race lines.
By contrast, polygynous Iron Age farmers have always been quite content to
take wives from hunter societies, so there was a far greater potential for inte-
gration and assimilation between colonist and indigene, although hypergyny
(one-way gene flow) was almost always the result (Morris 1992). Few hunters
could have amassed the number of cattle needed for bride-wealth payments
to secure farmer wives. There seems to be little doubt that Iron Age farmers
had arrived in southern Africa by 1800 years ago, although their initial impact
was not very significant along the eastern seaboard of Mozambique and South
Africa. Once the central highland was occupied, 1500 years ago, this appears
to have been much more intrusive.
Resistance against Bantu-speaking farmers by indigenous hunters certainly
took place once the farmers were established (Guenther 1997), but it would
seem that most hunters may have been relatively easily subsumed into farming
societies elsewhere. This does not explain how domestic sheep and pottery
arrived as far as the Cape before Iron Age farmers were in South Africa. There
may have been a domino effect, with East Sahelian pastoralists being pushed
ahead of expanding farmers out of eastern Africa and local Khoe-speaking
hunters being the first Khoisan groups to meet these herders in and around the
Caprivi Strip. If they took on sheepherding themselves, this raises the question
of how and why this took place. It is in this region that Khoe-speaking groups
seem to have established their dominance of the river systems (Cashdan 1986),
and such a well-watered, tsetse-free environment would have been conducive to
pastoral expansion. The East Sahelian herders could perhaps be seen as a ‘bow-
wave’ of expanding food-producing populations out of East Africa, which was
increasingly becoming very diverse in the mixture of different economies. The
archaeological problem of identifying the first pastoralists would come from
the small size of these initial groups, combined with mobility and the necessary
light, portable material culture made from organic materials. Pottery might
be the only easily identifiable indicator, and, as suggested above, may not have
been particularly heavily decorated and thus easily ignored in archaeological
surveys (but see Barnard, this volume). The ceramics that seem to have the
strongest association with the earliest pastoralists are usually small, thin-walled
vessels (10 cm in height), possibly used for sheep milking, which break up into
tiny pieces, making them even more difficult to find on small open sites.
274     A n d r e w B . S m i t h
The Tr ansition to Herding

The transition from hunters becoming independent herders, as opposed to
hunters becoming encapsulated and subsumed into food producing societies,
may depend on a number of factors. Hunters, generally, have a problem
competing as equals with food producers (Smith 1998). This does not mean
they cannot live separately and maintain their hunting/foraging lifestyle, but
this can depend on the hunters having a ‘bolt hole’: a place they can retreat to
out of the way of food producers. The arguments around how and why this
might have occurred among Kalahari foragers became very heated in the 1990s
and was generally known as ‘The Great Kalahari Debate.’
Ranged on one side were researchers who had studied the Kalahari Bushmen
(Ju/’hoansi) before massive impact from the outside world (Lee and DeVore
1976; Marshall 1976; Howell 1979; Lee 1979). On the other side were those
who, following Wolf (1982), believed that the Ju/’hoansi were a proletariat,
whose present condition was part of worldwide colonial expansion and impov-
erishment of traditional peoples, and that they had been under the hegemony
of wider polities since Iron Age times (Wilmsen 1989).
There is no question that many Bushmen were indeed encapsulated by Iron
Age people, and once European traders and hunters were involved in the ivory
and skin trade, firearms changed the political world forever. The Ju/’hoansi of
Nyae Nyae (Marshall 1976) appear only to have been minimally affected by all
of these events. The oral traditions suggest black people arrived in Nyae Nyae
for the first time as refugees as a result of German genocide against the Herero
and as herders in the 1940s (Lee 2002). European traders avoided Nyae Nyae,
as it was easier to move around the area (Lee and Guenther 1993). Any contact
with the outside by Ju/’hoansi seems to have been on their terms (Lee 2002), and
archaeological work in this part of the Kalahari indicates a minimal amount of
material culture penetrating from outside (Smith and Lee 1997; Smith 2001).
The archaeological indicators of encapsulation of Bushmen by Iron Age
people are clearly stated by Sadr (2002:44), who says that paucity of traditional
and prevalence of foreign remains are the clues. These materials include low
numbers of stone tools and high numbers of ceramics coming from the more
dominant society, as well as shifts in the frequencies of ostrich eggshell and
glass beads and metal. Sadr also states that there are clear distinctions between
ex-hunters living on the edge of farming communities and those still living
in the ‘bush.’
Although the Kalahari debate highlighted diverse opinions on the status of
modern hunters in southern Africa, it would appear from these studies that it is
possible to separate out independent hunters from food producers, the former
becoming subsumed into food-producing societies in the archaeological record.
Each of the economic groups had different settlement patterns, which make
archaeological visibility variable. Farmers, whose attachment to the land allows
accumulation of material goods, particularly ceramics, may be quite dominant
in the landscape. Herders, by virtue of their mobility, are really only visible
when they repeatedly re-occupy a site, and this is often for reasons other than
their pastoral activities, such as fishing or exploitation of marine resources.
Hunters are also mobile, but rockshelters are attractive to small groups, and
this increases their archaeological visibility.
Discussion
The relations between aboriginal hunters and incoming food producers
from the north would appear to have been variable. Some hunters retained
their foraging lifestyle, while others took on sheep herding. Still others were
subsumed into farming societies.
The archaeological problems of the first herders to southern Africa would
appear to stem from a combination of high mobility and small populations,
either of people or sheep, entering the region. Since we can make linguistic
connections between potential donors of domestic animals outside southern
Africa and Khoe speakers, they make the most sense as immigrants to the
Cape, as they are ultimately recognizable as the historic Khoekhoen. The
questions are: “When did they arrive? Were they the initial donors of stock
and ceramics to the cave occupants 2000 years ago, did they arrive when the
first herding sites with large numbers of sheep can be seen (around 1800–1600
BP), or did they only appear when large numbers of cattle were introduced to
the Cape (sometime after 900 BP)?” To be really contentious, since there is no
archaeological evidence for the historic Khoekhoen, Sadr has even suggested
that perhaps they only came to the fore once there was a market for the animals
when Europeans arrived in increasing numbers after the 16th century CE.
References
Alexander, J. A.
1984 The End of the Moving Frontier in the Neolithic of North-Eastern Africa. In
L. Krzyzaniak and M. Kobusiewicz (eds.), Origin and Early Development of Food-
Producing Cultures in North-Eastern Africa. Poznan, Poznan Archaeological
Cashdan, E. A.
1986 Competition Between Foragers and Food-Producers on the Botletli River,
Botswana. Africa 56: 3, pp. 299–318.
276     A n d r e w B . S m i t h
Cullinan, P.
1989 Robert Jacob Gordon and Denis Diderot: The Hague, 1774. Quarterly Bulletin
of the South African Library 43: 4, p. 146.
David, N., J. Sterner and K. Gavua
1988 Why Pots Are Decorated. Current Anthropology 29: 3, pp. 365–389.
Ehret, C.
1998 An African Classical Age: Eastern and Southern Africa in World History, 1000 BC
to AD 400. Oxford, James Currey.
Friedman, J. B.
1981 The Monstrous Races in Medieval Art and Thought. Cambridge, Harvard
University Press.
Gronenborn, D.
2003 Migration, Acculturation, and Culture Change in Western Temperate Eurasia,
6500–5000 cal BC. In M. Budja (ed.), 10th Neolithic Studies: Documenta
Praehistorica 30. Univerza v Ljubljani, Oddelek za Arheologia: pp. 79–91.
Guenther, M.
1997 Lords of the Desert Land: Politics and Resistance of the Ghanzi Basarwa of
the Nineteenth Century. Botswana Notes and Records 29: pp. 121–141.
Hall, S. L.
1998 A Consideration of Gender Relations in the Late Iron Age “Sotho” Sequence
of the Western Highveld, South Africa. In S. Kent (ed.), Gender in African
Prehistory. Walnut Creek, AltaMira Press: pp. 235–258.
Hart, T.
1987 Portervill Survey. In J. Parkington and M. Hall (eds.), Papers in the Prehistory
of the Western Cape, South Africa. BAR International Series 332. Oxford,
Henshilwood, C.
1996 A Revised Chronology for Pastoralism in Southernmost Africa: New Evidence
of Sheep at c. 2000 b.p. from Blombos Cave, South Africa. Antiquity 70: pp.
945–949.
Hodder, I.
1991 The Decoration of Containers: An Ethnographic and Historical Study. In W.
A. Longacre (ed.), Ceramic Archaeology. Tucson, University of Arizona Press:
pp. 71–94.
Howell, N.
1979 Demography of the Dobe !Kung. New York, Academic Press.
Huffman, T. N.
1994 Toteng Pottery and the Origins of Bambata. Southern African Field Archaeology
3: pp. 3–9.
Kolb, P.
1719 Caput bonae spei hodiernum. Nürnburg, Peter Conrad Monath.
Kroeber, A. L.
1923 Anthropology. London, Harrap.
Lee, R. B.
1979 The !Kung San. Cambridge, Cambridge University Press.
2002 Solitude or Servitude? Ju/’hoansi Images of the Colonial Encounter. In S. Kent
(ed.), Ethnicity, Hunter-Gatherers, and the “Other.” Washington, Smithsonian
Lee, R. B. and I. DeVore
1976 Kalahari Hunter-Gatherers. Cambridge, Harvard University Press.
Lee, R. B. and M. Guenther
1993 Problems in Kalahari Historical Ethnography and the Tolerance of Error.
History in Africa 20: pp. 185–235.
Lovejoy, A.
1942 The Great Chain of Being. Cambridge, Harvard University Press.
Marshall, L.
1976 The !Kung of Nyae Nyae. Cambridge, Harvard University Press.
Mikesell, M.
1960 Comparative Studies in Frontier History. Annals of the Association of American
Geographers 50: pp. 62–74.
Morris, A. G.
1992 The Skeletons of Contact. Johannesburg, Witwatersrand University Press.
Piggot, S.
1965 Ancient Europe. Chicago, Aldine.
Raven-Hart, R.
1967 Before van Riebeeck. Cape Town, Struik.
Robertshaw, P. T.
1978 The Archaeology of an Abandoned Pastoralist Camp-Site. South African
Journal of Science 74: pp. 29–31.
Rudner, I. and J. Rudner
1957 A. Sparrman’s Ethnographic Collection from South Africa: Stockholm: Statens
Etnografiska Museum.’ Smärre Meddelanden 25: pp. 5–28.
Sadr, K.
2002 Encapsulated Bushmen in the Archaeology of Thamaga. In S. Kent (ed.),
Ethnicity, Hunter-Gatherers, and the “Other.” Washington, Smithsonian
2003 The Neolithic of Southern Africa. Journal of African History 44: pp.
195–209.
Sadr, K. and I. Plug
2001 Faunal Remains in the Transition from Hunting to Herding in Southeastern
Botswana. South African Archaeological Bulletin 56: pp. 76–82.
278     A n d r e w B . S m i t h
Sadr, K., A. B. Smith and I. Plug

2003 Herders and Foragers on Kasteelberg. South African Archaeological Bulletin 58:
pp. 27–32.
Schrire, C.
1988. The Historical Archaeology of the Impact of Colonialism in 17th-Century
South Africa. Antiquity 62: pp. 214–225.
Schweitzer, F. R.
1979 Excavations at Die Kelders, Cape Province, South Africa: The Holocene
Deposits. Annals of the South African Museum 78: pp. 101–233.
Sealy, J., T. Maggs, A. Jerardino and J. Kaplan
2004 Excavations at Melkbosstrand: Variability Among Herder Sites on Table Bay,
South Africa. South African Archaeological Bulletin 59: pp. 17–28.
Sealy, J. and R. Yates
1994 The Chronology of the Introduction of Pastoralism to the Cape, South Africa.
Smith, A. B.
1987 Seasonal Exploitation of Resources on the Vredenburg Peninsula After 2000
BP. In J. Parkington and M. Hall (eds.), Papers in the Prehistory of the Western
Cape, South Africa. BAR International Series 332. Oxford, Archaeopress: pp.
393–402.
1990a On Becoming Herders: Khoikhoi and San Ethnicity in Southern Africa.
African Studies 49: 2, pp. 51–73.
1990b Pastoralism in Africa. London, Hurst.
1993a Different Facets of the Crystal: Early European Images of the Khoikhoi at the
Cape, South Africa. In M. Hall and A. Markell (eds.), Historical Archaeology in
the Western Cape. Goodwin Series 7. Cape Town, South African Archaeological
Society: pp. 8–20.
1993b Exploitation of Marine Mammals by Prehistoric Cape Herders. South African
Journal of Science 89: pp. 162–165.
1998 Keeping People on the Periphery: The Ideology of Social Hierarchies
Between Hunters and Herders. Journal of Anthropological Archaeology 17: 2,
pp. 201–215.
2001 Ethnohistory and Archaeology of the Ju/’hoansi Bushmen. African Study
Monographs: Supplement 26, pp. 15–25.
Smith, A. B., D. Halkett, T. Hart and B Mütti
2001 Spatial Patterning, Cultural Identity, and Site Integrity on Open Sites:
Evidence from Bloeddrift 23, a Pre-Colonial Herder Camp in the Richtersveld,
Northern Cape Province, South Africa. South African Archaeological Bulletin
56: pp. 23–33.
Smith, A. B. and L. Jacobson

1995 Excavations at Geduld and the Appearance of Early Domestic Stock in
Namibia. South African Archaeological Bulletin 50: pp. 3–14.
Smith, A. B. and R. B. Lee
1997 Cho/ana: Archaeological and Ethnohistorical Evidence for Recent Hunter-
Gatherer/Agro-Pastoral Contact in Northern Bushmanland. South African
Archaeological Bulletin 52: pp. 52–58.
Smith, A. B. and W. E. Pasche
1997 Balthasar Springer at the Cape (1506). Quarterly Bulletin of the South African
Library 51: 3, pp. 93–98.
Smith, A. B. and R. H. Pheiffer
1993 The Khoikhoi at the Cape of Good Hope: Seventeenth-Century Drawings in the
South African Library. Cape Town, South African Library.
Smith, A. B., K. Sadr, J. Gribble and R. Yates
1991 Excavations in the South-Western Cape, South Africa, and the Archaeological
Identity of Prehistoric Hunter-Gatherers Within the Last 2000 Years. South
African Archaeological Bulletin 46: pp. 71–91.
Smith, A. B. and L. Webley
2000 Men and Women of the Khoekhoen of Southern Africa. In D. L. Hodgson
(ed.), Rethinking Pastoralism in Africa. London, James Currey: pp. 72–96.
Turner, F. J.
1920 The Frontier in American History. New York, Henry Holt.
Webley, L.
1997 Jakkalsberg A and B: The Cultural Material from Two Pastoralist Sites in
the Richtersveld, Northern Cape. Southern African Field Archaeology 6: 1, pp.
3–19.
2001 The Re-excavation of Spoegrivier Cave on the West Coast of Southern Africa.
Annals of Eastern Cape Museums 2: pp. 19–49.
Wiessner, P.
1994 The Pathways of the Past: !Kung San Hxaro Exchange and History. In M.
Bollig and F. Kees (eds.), Überlebens-strategien in Afrika. Köln, Heinrich Barth
Institut: pp. 101–124.
Wilmsen, E. N.
1989 Land Filled with Flies. Chicago, University of Chicago Press.
Wolf, E.
1982 Europe and the People Without History. Berkeley, University of California
Press.
Ch a p t e r 1 3
T he Social and
Environmental
C onstraints on Mobility
in the L ate P rehistoric
Upper Great L akes R egion
Margaret B. Holman and William A. Lovis1
T o better understand the archaeology of late prehistoric mobility in the

Great Lakes region of the United States, theoretical concepts of mobility in
small-scale societies are used to frame analogies derived from the ethnohistoric
record. Despite disruptions associated with European contact, ethnohistoric
observations remain relevant for examining late prehistoric mobility. This
relevance is a consequence of relative stability in the structure of the environ-
ment and the use of seasonal resources across the contact period boundary,
which in turn allows the development of analogue models. Local constraints on
mobility were highly variable: some territories were essentially filled to capacity
with hunter-gatherers, some societies had regular and sustained contact with
settled horticulturalists, and all were compelled to respond to the fur trade and
consequent interactions with Euro-American settlers. We demonstrate concor-
dance between the late prehistoric archaeological record and ethnohistorically
documented flexibility in mobility strategies and integrating social mechanisms
that acted to buffer environmental variability.
All of the ethnographically documented native people resident in Michigan
during the early years of direct and consistent European contact, around
1650 CE, were mobile to some degree, even after the late adoption of horti-
culture. Groups regularly moved to obtain seasonally available foods and
1.
Dr. Margaret B. Holman fell victim to pancreatic cancer before the publication of
this volume, a synthesis about which she was tremendously excited from its inception as
a Cotsen Advanced Seminar. I regret that she is not able to appreciate the publication
of our chapter, for which she is largely responsible. I know she would be proud and
humbled by the dedication of this volume to her memory, and as her friend, colleague
and co-author I am deeply grateful for this kindness.
280
Constraints on M o b i l i t y      281
other resources that were unevenly distributed across the landscape. Some
incorporated crops into their subsistence cycle, whereas others grew very little
food, if any at all. Mobility patterns depended on the distribution of resources
sought and the social constraints on regularly exploited territories. Variations
included seasonal aggregation and dispersal of related households, and some
degree of sedentation, combined with task-specific logisttic mobility. The Late
Woodland peoples (650–1650 CE), immediately preceding European contact,
were likewise seasonally mobile and used strategies similar to those seen among
historically known groups.
To understand changes in mobility patterns around the Straits of Mackinac
and the northern lower peninsula of Michigan (Figure 13.1), we use the estab-
lished triad of ethnographic analogy, documentary evidence and archaeology.
The observed changes that occurred over a millennium range from responses
to the use of all available territories by hunter-gatherers, to contact with settled
horticulturalists, to disruptions related to the European fur trade, to interac-
tions with American settlers. Thus, we have an opportunity to continually
refine our understanding of how mobility in this area was a way of life used to
solve problems similar to those faced by nomadic, or mobile, people in other
parts of the world.
Figure 13.1. Location of the study area, in the Great Lakes region on the border between Canada and
the United States.
282     M a r g a r e t H o l ma n and W i l l i am L o v i s
Theoretical Fr amework
We understand mobility as a means by which relatively egalitarian societies
secure subsistence when food and other resources are spatially and seasonally
separated. Such mobility may be characteristic of people who live entirely
on wild foods (hunters, gatherers, fishers) or people who combine hunting,
gathering or fishing with horticulture. Movement as part of these kinds of
subsistence-settlement systems is through a territory that provides resources
throughout an annual cycle. Territories are occupied and ‘controlled’ by
groups whose members usually are extended families related on the basis of
primary kinship ties. Territories supplying the subsistence needs of one group
may overlap those of other groups (Williams 1968:129; Cleland 1992a:101),
suggesting fluid boundary maintenance, and there may be neutral zones avail-
able for exploitation by more than one group (Pilling 1968:155).
Mobility may be residential or logistic, or, recognizing that these are
extremes of a continuum, a combination of the two (Binford 1980). Residential
mobility involves movement of entire households, or groups, to locations where
a variety of resources can be obtained nearby and brought back to a base camp
for processing. Residential mobility pools the labor of different genders and
ages at a single location, where they procure and process food, manufacture
tools, construct shelters, make clothing and undertake other tasks. Thus, several
needs may be met at the same time and place. Logistic mobility operates in
situations where important resources are separated from one another spatially
or temporally (Binford 1980:5–10). In such situations one or a few members
of a group will move to secure a specific resource to be brought back to the
residential base. Task groups, such as hunting parties, may travel considerable
distances and be away for extended periods of time before returning with meat
and hides that can be processed at the residential location. Binford (1980:14)
emphasizes that the structure of the environment—that is, how resources are
distributed within it—is critical as to which type of mobility is most likely to
characterize the patterning of a group’s movements. We take the position that,
along with environmental structure, social constraints on the use of territory
affect mobility strategies. Emphasis on residential or logistic mobility varied
over time within our study area, even though the distribution of wild resources
remained constant, as did the potential for growing crops in certain loca-
tions. There is evidence for fluctuations in climate, but the structure of these
resources within the environment did not change markedly. As the potential for
food failure is ever present, mobile people may buffer risk by interacting with
their neighbors (Spielmann 1986:280-281; O'Shea and Halstead 1989:124).
Co-operative social interactions are important to hunter-gatherers when their
normally exploited territory does not supply sufficient food in a given year.
The option of residential movement into adjacent territories, occupied or

controlled by other people, is facilitated by good relationships, as is the sharing
of overlapping territories. Social constraints and interactions also condition
mobility when critical resources are produced or controlled by disparate groups
(Spielmann 1986:288–290). Mobility under these circumstances tends to be
logistic for purposes of exchange or conflict.
Ethnogr aphic Analogy and Late

Prehistoric Mobility
The predominant residents of Michigan at the time of contact were three
related Algonquian-speaking groups: the Potawatomi in the south, the
Chippewa in the north, and the Ottawa, who lived between the other two
(Figure 13.2). Each occupied an environment that differed in its configuration
from the others, and each employed mobility as part of economic and social
strategies that were adaptive to the local environment. Since use of the “direct
historical approach” (Strong 1935, 1936; Wedel 1938) is rarely possible, it is
these adaptive patterns that form the basis for our inquiries into the mobile
lifestyle of prehistoric Late Woodland peoples in the same region (Fitting and
Cleland 1969). Our intent here is not to provide a comprehensive overview
of the uses of ethnographic analogy in archaeology but rather to historically
contextualize its use as a vehicle for explaining late prehistoric mobility within
our study region of the upper Great Lakes, particularly Michigan.
As noted, use of a direct historical approach has not been a productive
vehicle for such regional explanation, as we cannot normally associate artifacts,
or artifact styles, found at a site occupied by a historically documented group
with similar artifacts used by their precontact ancestors. We cannot even be
confident that these groups, whose initial contact was with the French, were
the same Late Woodland peoples who lived in Michigan just prior to contact,
early in the 17th century CE (Kinietz 1965:v; Clifton 1984:1). By their own
oral traditions the three groups arrived in the region near the end of the Late
Woodland period, around 1500 CE. The gap between historically known
groups and Late Woodland peoples is attributable to several factors. Continuity
and change are apparent in Late Woodland archaeology, but beginning with the
arrival of Europeans, the different cultures cease to be archaeologically ‘visible’
(Cleland 1992b:29). Native Americans adopted European manufactured goods
quite quickly. For example, stylistically sensitive pottery was largely discarded
in favor of metal kettles (McClurken 1988:8). There was also such widespread
disruption and displacement of peoples that pottery known to have been made
by a particular group might be found at a site reported to have been occupied
by a different group entirely (Fitting 1975).
Figure 13.2. Locations of Native American people in the Great Lakes region during the Historic Period
(note overlapping and shared territories).
The major disturbances creating such confusion in the Great Lakes archaeo-
logical record were one consequence of the warfare between different Native
American groups who were competitors in the fur trade (Tanner 1987:1–4).
They were variously allied with the French or the British, who themselves
were vying with one another for control of the trade. At times, whole areas
of the region, including western Ontario and the lower Michigan peninsula,
were depopulated, while other areas, such as northeastern Wisconsin, were
filled with refugees of differing tribal or ethnic affiliations. The situation is
further complicated by the fact that, as a result of this disruption, some groups
combined with others in temporary communities, and some formed new multi-
ethnic corporate groups. Some populations were so decimated by European
diseases, and other consequences of contact, that their sites cannot be identified
(Mason 1988:66–67).
The fashion in which Michigan archaeologists have used ethnohistoric
accounts to address late prehistoric mobility, particularly on a seasonal basis,
owes itself in large part to the now forty-year-old work of G. I. Quimby. In a
major chapter of his book Indian Culture and European Trade Goods, Quimby
(1966) employed the 18th century CE narratives of Alexander Henry, specifi-
cally with citations to Henry, to Bain’s 1901 editing of Henry (reprinted in
1969), and to Quaife’s (1921) annotations of the same narrative. Henry traveled
with the Wawatam family of Chippewa for a year, in 1763–1764, and Quimby
used the accounts of their travels to both interpret and provide insights into the
archaeological signatures that one might expect, including notions of seasonal

mobility, technology and resource use. Coupled with other postcontact narra-
tives, such as that of John Tanner (1956) and James Smith (1907), archaeologists
refined and developed models that closely reflected their origins in Quimby’s
original work, situated them in an ecological framework, and applied them to
the precontact, late prehistoric past (Fitting and Cleland 1969; Fitting 1975;
Holman and Krist 2001, among others ).
Deviations from the fundamental model were minor, relatively rare, and
incorporated similar elements; there was almost always winter season fission
into smaller social units that exploited the interior of Michigan’s two penin-
sulas. In the north, where horticulture was less common and fishing more
prominent, seasonal aggregation was primarily during periods of peak fish
abundance, spring and late fall; in the south, where horticulture was dominant,
general warm season aggregation took primacy. In both cases group territories
were viewed as spanning both coastal and interior resource zones, with no
independent social or ethnic groups resident in either one or the other and
with seasonal movement between zones. In truth, adherence to such deriva-
tive models actually explained much, but not all, of what was being observed
in late prehistoric context.
Changes in the archaeological view of these more traditional ethnographic
analogs have come in the form of altered perspectives on the scale of archaeo-
logical inquiry, as well as the role of symbiotic ritual and exchange relationships
in fostering more formalized intergroup interactions (Milner and O'Shea 1998;
O'Shea and Milner 2002; Howey and O'Shea 2006). While such approaches
have not necessarily diminished the ecological perspective that has domi-
nated late-period research, they have nonetheless called into question strict
adherence to the primacy of ethnohistoric documents, particularly given the
temporal disjunction of more than a century between the arrival of Europeans
and the documents on which the models are based (O’Shea 2004). This has
given rise in lower Michigan to a competing model of symbiotically interac-
tive coastal horticulturalists and interior foragers, each with relatively separate
social territories (O’Shea 2003). These new perspectives, too, have explanatory
merit and provide an additional dimension to the approach we take here.
Despite the difficulties caused by confusing archaeological remains and
written documents of varying utility, Michigan archaeologists have effectively
combined both sources of information to address questions of mobility and
changes in mobility strategies over time. These investigations have been
productive, in part, because Late Woodland peoples were associated with
the same broad environmental zones as the historically documented groups,
coped with similar environmental uncertainties, and had similar decentralized
social organizations (Quimby 1966:179; Cleland 1992a:97; Holman and Krist
Figure 13.3. Regional environmental zones in the study area.
2001:20). Thus, the patterning of the adaptive strategies employed by the

Potawatomi, the Ottawa and the Chippewa in their respective environmental
zones have served as a framework guiding archaeological research focused
on Late Woodland settlement (Fitting and Cleland 1969:292). But only the
patterns of adaptation exhibited by these groups have been traced back into
prehistory, not the groups themselves.
The Environment of the Study Area

The Straits of Mackinac and the northern lower peninsula of Michigan are
particularly useful for understanding mobility in the Great Lakes region
(Figure 13.3). This area occurs at a juncture of environmental zones, where
two historically observed adaptive patterns, those of the Chippewa and the
Ottawa, were present (Fitting and Cleland 1969). Not only are the subsistence
and settlement systems of these groups the best known, but the adaptive use of
the broad environmental configuration has also been the focus of surveys and
excavations used to test the Late Woodland occurrence of the ethnohistorically
derived adaptive patterns (Cleland 1967, 1982; Lovis 1976; Holman 1978;
Martin 1989). The environmental zones in the study area are the Canadian
Biotic Province and a transition zone between the Canadian Biotic Province
and the Carolinian Biotic Province to the south (Albert et al. 1986). Late
Woodland and Historic mobility included movement both within and across
these environmental zones. Sharing of overlapping territory, as well as move-

ment into environments regularly occupied by other groups, was a significant
part of the various adaptations (McClurken 1988). At times, food and other
items produced by residents of these zones were exchanged among groups
(Kinietz 1965; McClurken 1988; Smith 1996).
People in the Canadian Biotic Province occupy an environment that is
highly variable in the distribution of food across the landscape and through
the year (Yarnell 1964:5–7; Cleland 1966:9–11). Important differences in food
potential can be seen between the Upper Great Lakes and the forests of the
interior. There are additional significant differences in habitats between two
major forest types. Systematic and scheduled mobility solves the problem of
subsistence in the Canadian Biotic Province, where wild resources are unevenly
distributed both spatially and seasonally. Great Lakes coastal waters, and the
mouths of streams emptying into them, are sources of fish that are particularly
abundant and reliable during spawning periods in late spring to early summer
and again in the late fall (Cleland 1982, 1989:766–767; Martin 1989:602).
For people dependent on wild foods, catching spawning lake sturgeons and
suckers can be critical in the spring, when other resources are not yet available.
Similarly, fall spawning fish, like lake trout and whitefish, are key to a nutritious
supply of food that can be preserved and transported for winter use. Fish found
in interior lakes are a source of food throughout the year (Cleland 1966:10;
Martin 1999:224). Most soils in the Canadian Biotic Province are too poor,
and the growing season too short, to support a reliable horticulture. While
some shoreline locations along Lake Michigan and Lake Huron are within the
120 frost-free-day limit for successful aboriginal maize agriculture, they are
subject to late frosts in spring and early frost in fall (Yarnell 1964:128–129, 133;
Cleland 1966:9–11). When crops are grown in these areas, a successful harvest
is so doubtful that crops can only be used as supplements to the wild food
supply. Thus, even in favorable locales, lakeshore settings are more significant
for catching fish than for planting gardens.
Productivity of the forests in the Canadian Biotic Province varies between
the climax vegetation of northern hardwoods, such as beech and maple, and
several coniferous subclimax forests, including pines and hemlocks on higher
sandier soils and spruce, cedar, fir and birch in swampy soils (Yarnell 1964:5–7).
The differences between these various forest compositions are significant
because the hardwoods provide foods such as maple sap and beechnuts, as well
as forage for a variety of animals. In contrast, vast stands of pine and extensive
coniferous bogs and swamps produce little in the way of food for humans or
the animals they hunt. Economically important plants do not grow together in
the same habitats, nor are they productive at the same times (Yarnell 1964:77).
Thus, in order to collect a variety of plants, either concurrently or successively,
it is necessary to position oneself in proximity to several habitats. Mobility must

also be employed in the Canadian Biotic Province to hunt northern animals
such as moose, woodland caribou, bear and beaver (Cleland 1966:10). Most of
these animals are essentially solitary, and none are abundant in any one loca-
tion. Whitetail deer are important game, although they are more numerous
in the oak- and hickory-dominated forests and forest edges of the Carolinian
Biotic Province (Cleland 1966:65). Unlike southern deer populations, northern
deer aggregate, or ‘yard’ in winter in coniferous forests near cedar swamps,
where they are sheltered from the wind and can browse on evergreen vegeta-
tion (Doepker and Ozoga 1991). Thus, habitat where deer might be found in
the harsh Canadian Biotic Province winter is fairly predictable, and the animals
are not necessarily solitary.
The transition zone between the Canadian and the Carolinian biotic prov-
inces is significant for human subsistence because it offers an unusually diverse
array of plant and animal foods (Cleland 1966:6–7; Fitting 1966; Hambacher
1992:36). Additionally, the average of between 160 and 140 frost-free days is
well above the limit of 120 frost-free days for prehistoric horticulture (Yarnell
1964:127–128). Both biodiversity and the potential for horticulture are most
marked along the eastern shores of Lake Michigan and Lake Huron, where
climate is ameliorated by westerly winds blowing across the lakes. The west-
erlies blow lake-warmed air over the land in winter and lake-cooled air in
summer. The transition zone is abrupt along the western shore of Lake Huron,
where the climate modifying winds have less effect because of the intervening
land mass. The transition zone is populated by plants and animals found in the
beech, maple and coniferous forests of the Canadian Biotic Province and by
plants and animals associated with the oak and hickory forests of the Carolinian
Biotic Province (Cleland 1966:224–246; Hambacher 1992:36). Thus, the food
supply for mobile hunter-gatherers is characterized by a diversity of species
that can be exploited successively through the year. For example, the transition
zone yields an abundance of northerly distributed berries along with several
species of nuts that are more abundant to the south (Yarnell 1964:78–79). The
berries can be obtained as they ripen during the late summer, whereas the
nuts can be procured as they become available through the fall. Both foods
can be preserved for winter use. Game present in northern portions of the
transition zone included more northerly moose and woodland caribou until
the 19th century CE (Cleland 1966:10). Whitetail deer and elk were available
in lesser numbers than in the Carolinian Biotic Province to the south. Because
the various resources of the transition zone occur in patches, movement
involves travel to locations where several different micro-environments are in
proximity to one another (Yarnell 1964:77–79; Hambacher 1992:36). These
micro-environments offer different food plants and, consequently, the animal
communities they support. The numerous lakes and streams in the transition
zone are sources of multiple fish species, and the rivers in the lower peninsula
are transportation routes between the coast and the interior (Hambacher
1992:35). Large portions of the transition zone are south of the spawning
grounds of the whitefish, making coastal locations unattractive for fall fishing
(Hambacher 1992:35–36).
The Chippewa and the Mackinac Phase People

The Chippewa were highly mobile people who depended on hunting, gath-
ering and fishing in the Canadian Biotic Province. Chippewa mobility consisted
of residential moves by extended family households who moved seasonally to
locations where foods were available. The numbers of households residing at
a location depended largely on the amount of food available. Always, however,
households acted as the basic economic decision-making group that moved
together and stayed together throughout the year (Kinietz 1965:321; Cleland
1992b:46–47). Although they planted maize and other crops, the Chippewa
could not rely on a harvest sufficient for winter use. Instead, they acquired
maize from the sedentary Huron as part of an exchange process mediated by
their Ottawa neighbors (McClurken 1988:14). In turn, the Chippewa provided
the Huron with meat and hides.
During the warm season the Chippewa hunted, fished, gathered plants, and,
in some places, planted gardens (Kinietz 1965:321; Cleland 1992b:46–47). As
food was readily available at such times, extended families would come together
with related households in settlements often located along the shores of the
Great Lakes. Larger aggregations of people occurred at key fall fishing loca-
tions along the Great Lakes shorelines and at the rapids of the Saint Mary’s
River at Sault Ste. Marie. When food was less abundant and more scattered
during the winter, households would disperse across the interior landscape.
Movement toward winter hunting grounds was often along river systems,
where people caught deer and beaver along the way. In spring, households
came together again to make maple sugar before returning to their summer
settlements. Storage of food procured in each season was an important buffer
against times of scarcity (Densmore 1979:40, 119–131). Lined storage or cache
pits were dug and filled at settlements, camps and other locations while people
were on the move. Scaffolds were used when the ground was frozen in the
winter (Smith 1907:35; Tanner 1956:46; Bain 1969:130).
The adaptive strategy proven archaeologically for the earliest phase of the
Late Woodland in northern Michigan is comparable to the Chippewa pattern
of residential mobility (Holman 1978). During the period between 800 and
1000 CE, known as the Mackinac phase, people were hunter-gatherers who
Figure 13.4. Mackinac phase and other early Late Woodland sites mentioned in the text.
systematically moved to obtain the same seasonal resources in the same natural
settings as the Chippewa (Figure 13.4). Like communal practices among the
Chippewa, social organization in the Mackinac phase was probably charac-
terized by extended family economic units and by seasonal aggregation and
dispersal of such families. Unlike the Chippewa, there is no evidence for a
formal system of exchange whereby Mackinac phase people provided meat
and hides to settled horticulturalists and received maize in return. Although
maize was present in the Eastern Woodlands, it was not until about 1000 CE
that some Upper Great Lakes peoples became dependent on maize, beans and
squash (Murphy and Ferris 1990:261–263, 273; Parker 1996:314; Brashler et al.
2000:570), so there may not have been settled horticulturalists with whom to
trade. There were neighbors, however, with whom good relations were main-
tained in order to buffer risk in an unpredictable environment (Holman and
Kingsley 1996). Mackinac phase peoples in the study area were mainly residents
of the Canadian Biotic Province, but their sites along the Lake Michigan coast
south from the Straits of Mackinac to the north end of Grand Traverse Bay are
in the Carolinian-Canadian transition zone. Mobility involved seasonal travel
between the coast and interior.
Like the Chippewa, Mackinac phase groups normally spent the warm season
at coastal sites, situated near streams populated by spring-spawning fish, where
food of all kinds could be found within a relatively short distance (Holman
1978; Martin 1989). In these favorable environmental settings, faunal and
floral analyses indicate warm season habitations (Lovis 1973; Holman 1978).
As expected from repetitive extended family residential occupations, coastal
sites are large, exhibit thick organic midden accumulations, show relatively
evenly distributed features (including dwellings), and yield artifacts repre-
senting a wide variety of gender- and age-based activities. These activities
include hunting, gathering, fishing, cooking, hide scraping, woodworking,
manufacturing pottery (including toys) and producing stone tools. This broad
range of tasks, and the partitioned spatial structure evident at two sites, indi-
cates that labor was apportioned along age and gender lines similar to that
of the Chippewa and other hunter-gatherer groups. At both sites near the
source of Norwood chert, stone tools of this material were manufactured in
contexts separate from areas where hearths, pottery and other artifacts indi-
cate domestic activities (Lovis 1973; Holman 1978). Mackinac phase domestic
areas were arranged along the back edges of lakeshore bluffs in a linear pattern
similar to that seen in Chippewa summer camps, where each extended family
established its own space (Densmore 1979:122). The absence of more formal
partitioning of space suggests that Mackinac phase extended families, like those
of the Chippewa, were relatively independent. Flexibility of movement in the
season of plentiful food included the option of visiting friends and family at
other coastal sites or avoiding social contact altogether by gathering food in
the interior (Holman 1978). Mackinac phase coastal sites in the transition zone
contain ceramics made by Mackinac phase potters along with pottery made
by their neighbors immediately to the south (Lovis 1973). These neighbors
shared the use of some transition zone locations with Mackinac phase people,
including the Pi-wan-go-ning quarry, where both groups obtained Norwood
chert (Lovis 1973; Holman 1978; Hambacher 1992).
The fall component of Mackinac phase settlement centered on the Straits
of Mackinac, where whitefish and lake trout came into shallow waters to spawn
in late October to early December (Cleland 1966:172). As was the case with
the Chippewa, these fish served as a kind of ‘survival insurance’ for the winter,
when food was neither abundant nor reliable. The best known example of a
late-fall fishing site is the Juntunen site on Bois Blanc Island, in the Straits of
Mackinac (McPherron 1967). Juntunen exhibits intensive occupation along
with abundant remains of fish, including sturgeon, whitefish and lake trout
(Cleland 1966:157–210). Although plant and animal remains indicate warm-
season occupation occurred at Juntunen, this was a particularly desirable site in
late fall. Its central location and the abundance of fall-spawning fish created the
opportunity for large numbers of people to meet and interact. The intensity of
Mackinac phase occupation at Juntunen suggests that it represents an aggrega-
tion of local bands comparable to that seen at the Chippewa fall fishery in the
rapids at Sault Ste. Marie and fall fishing in the Straits of Mackinac. These
gatherings reinforced social ties and were chances to acquire information about
winter hunting prospects.
Mackinac phase households occupied winter sites near inland lakes a short
distance south of the Straits of Mackinac (Holman 1978:45–48, 149). These
sites were situated in sheltered locations where they were surrounded by a
wide variety of habitat types and associated game. Such settings increased the
probability that winter hunts would be successful. By comparison with coastal
sites, excavated Mackinac phase interior sites are characterized by an uneven
distribution of materials, fewer features and no developed organic middens.
Hearths and postmolds, coupled with abundant pottery and few stone tools,
suggest a relatively narrow range of activities focused on domestic concerns.2
Among the stone tools are projectile points and hide scrapers, both of which are
related to capturing and processing game. Plant remains are limited in number
and variety, as would be expected from cold-season sites where there was no
substantial plant storage as people arrived late in the year with only plant foods
they carried with them (Densmore 1979:120). There is no ceramic evidence
that interior locations were occupied by anyone other than Mackinac phase
people during the early Late Woodland. As expected at winter sites, people
lived there in relative isolation.
It is probable that Mackinac phase people, like the Chippewa, made maple
sugar at sites temporally and geographically located between winter and
summer locales (Holman 1978:154). The sap run provided food in the early
spring season of scarcity at a location where related households aggregated
after a winter of separation. Evidence for the existence of Mackinac phase sites
used for sugaring includes settings with maples distributed so the length of the
sugaring season could be prolonged and a site with features such as abundant
fire-cracked rock and artifacts probably related to sugaring (Lovis 1978:44;
Holman 1984:79).
The early Late Woodland was marked by a pattern of seasonal subsistence
that, in most respects, mirrors that of the Historic period Chippewa in the same
region. Mackinac phase groups faced the same environmental uncertainties as
the Chippewa, but exchange with settled horticulturalists was not a possible
solution to potential scarcity of food within their regularly exploited territories.
Simply moving into unoccupied land was also impossible because Mackinac
phase peoples were surrounded by neighbors who, like them, relied on wild
food and faced potential food scarcities. Even though adjacent territories
were occupied, people needed to be able to move outside their own primarily
exploited territories on an emergency basis. Mackinac phase people and other
2.
Postmolds are organic soil stains remaining after the decomposition of a post, in
this case indicative of the remains of temporary dwellings (cf. Chang, this volume).
early Late Woodland peoples of the lower peninsula solved the problem of
possible resource shortfalls by engaging in a system of cooperative buffering,
whereby they were able to use territories normally within the domain of other
groups (Holman and Kingsley 1996:360–361). Cooperative buffering in the
lower peninsula is evidenced by joint use of overlapping territory in the transi-
tion zone along the northwest coast of Lake Michigan (Holman and Kingsley
1996:361). Some warm-season sites in the transition zone were occupied by
Mackinac phase groups and their southern neighbors, while the Norwood
chert from the Pi-wan-go-ning quarry in the same area was freely used by both
groups (Cleland 1967; Lovis 1973, 1990).
Maintaining good relations through visiting and other interactions is indi-
cated by small amounts of Mackinac phase pottery at the Skegemog Point site
near Grand Traverse Bay, which was predominantly occupied during the warm
season by a neighboring group (Hambacher 1992). Similarly, the Fletcher site,
a warm-season site near Saginaw Bay, belonged to a different set of neighbors
and has a few examples of Mackinac phase pottery, suggesting that visiting took
place (Brashler 1973; Holman and Kingsley 1996:360–361). The importance of
fostering cooperation is seen in a Mackinac phase winter hunting component
at the Butterfield site, which is on the northeastern side of the Saginaw Valley,
not far from the Fletcher site (Wobst 1968:246). This site is evidence for the
option to move into another group’s territory in the most problematic season
of the year (Holman and Kingsley 1996:363).
The Ottawa and the Juntunen Phase People

The Ottawa homeland, at the northern end of Ontario’s Bruce Peninsula
and islands in Georgian Bay (Lake Huron), was at the southern end of the
Canadian Biotic Province and, in Michigan and Wisconsin, in the transition
zone between the Carolinian and Canadian biotic provinces (Cleland 1966:73;
McClurken 1988:12). Mobility related to Ottawa subsistence, including
exchange, was both logistic on the part of specific task groups and residential
involving entire households. Mobility and social organization among the
Ottawa reflected their flexible subsistence strategy centered on the transition
zone (McClurken 1988:13–14). The Ottawa raised corn, beans, squash and
sunflowers in the fields around their semipermanent villages. In the summer,
meat was provided by hunting parties who sought game within 120–160 km
from the village (Kinietz 1965:237). When the harvest was good, the Ottawa
stored their surplus maize and remained in their villages throughout the winter.
Important additional sources of storable foods were the whitefish and lake
trout caught in the Straits of Mackinac. Winter stores of maize and fish were
augmented by logistic hunting carried out by small parties, usually of men,
who would travel long distances in search of game to bring back to the village.
Ottawa winter hunts often took place along the river systems in the transition
zone of the lower peninsula and in the Saginaw Valley.
Harvests were not always sufficient to provide adequate winter stores as the
growing season in the transition zone was subject to late frosts in the spring
and early frost at the end of summer (Yarnell 1964:128–129, 133). Additionally,
Ottawa crops probably failed during the serious droughts that occurred on
average once a decade (McClurken 1988:14). In the winters of years when the
harvest was poor, Ottawa subsistence was based on hunting and gathering. At
such times, extended family households left their villages and dispersed across
the landscape to employ a residential mobility system similar to that of the
Chippewa. In addition to seasonal hunting, logistic mobility was employed
by the Ottawa in the course of their occupation as middlemen in a system
of exchange between neighboring groups (Kinietz 1965:245; McClurken
1988:14). This activity involved the exchange of items that were important
for the subsistence of neighbors and for the Ottawa themselves. The Ottawa
brokered an exchange of meat and furs, obtained by the northern Chippewa,
for maize and other agricultural products grown by the Huron, who cultivated
crops in fields around their sizable and relatively permanent villages (Trigger
1969:9–11). Because the territory of the Huron was subject to depletion of
game by overhunting, the Huron needed the meat and hides supplied by their
trading partners (Smith 1996:285). By virtue of exchange, the Chippewa were
able to obtain supplies of maize to augment their food supply, while the Ottawa
were able to make their subsistence more secure. Logistic mobility by Ottawa
men was a key factor in the exchange network that provided the Chippewa
and Huron with food and other items they could not independently produce
in sufficient quantities (Kinietz 1965:245; McClurken 1988:23). These men
traveled the Great Lakes region in birch-bark canoes along family-controlled
routes, where exchange relationships were carefully nurtured by reciprocal gift
giving and by the arrangement of marriages with trading partners. Maintaining
strong partnerships was important to all parties because the network redistrib-
uted key subsistence items throughout the region (Trigger 1969:39).
After the Huron were permanently driven from their homeland by the
Iroquois, between 1648 and 1651 CE (Kinietz 1965:2), the old subsistence
system depending on trade with the Huron ceased (Smith 1996:105–106).
When the threat from the Iroquois had passed, and it was safe to return to
the Straits of Mackinac region, many Ottawa established horticultural villages
in their old hunting grounds in the transition zone of the northern lower
peninsula, where they resumed their pattern of subsistence (McClurken
1988:29–30, 33, 35–36). This included participation in the fur trade, once
again as middlemen. The familiar role of broker was employed not between
the Huron and the Chippewa but between the French in the east and groups
situated to the west of the Ottawa. Once again, logistic mobility was employed
to provide furs to the French and European goods to the western bands and
tribes. Later, the Ottawa were providers of maize, fish and equipment to French
traders passing through the Straits of Mackinac on their way to the interior.
Beginning with the Juntunen phase, about 1200 CE, the region was home
to people who practiced a settlement system comparable to the Ottawa pattern
(Figure 13.5). Evidence for the Ottawa pattern during the Juntunen phase
includes semipermanent horticultural villages, intensive fall fishing locales,
long-distance mobility for hunting and exchange and a backup system of
residential moves for winter hunting. Whether semipermanent horticultural
villages were common during the Juntunen phase and early Historic periods
(ca. 1200–1650 CE) is uncertain. There are sites that were intensively occupied
during the warm season and through late fall or year-round. The coastal loca-
tions of these sites, however, are surrounded by minimal arable land, suggesting
that fishing and accessibility, not gardening, were prime considerations in
choosing site situations (Cleland 1966:194; Smith 1996:246–252). In all cases
maize is found in relatively low densities, whereas the remains of sucker, lake
sturgeon, whitefish and lake trout are abundant. Again, the locations of these
sites at prime all-season fishing spots, coupled with the faunal evidence, show
that the most important foods obtained were fish rather than maize (Cleland
1966; Martin 1989; Smith 1996). The remains of large animals at coastal sites
Figure 13.5. Juntunen phase and other later Late Woodland sites mentioned in the text.
reveal that woodland caribou, deer, beaver and bear were desired foods (Cleland
1966:194; Smith 1996:240–241). These animals were not as readily available as
fish because they were not present in sufficient numbers to sustain local animal
populations in the face of continued hunting pressure (Smith 1996:252). Thus,
although some game may have been captured nearby, much must have been
obtained by logistic hunting parties or exchange with northern neighbors.
The abundant evidence of fall-season fishing at coastal sites, most likely
with the assistance of gill nets, attests to the importance of this activity in
the Juntunen phase (Cleland 1966; Smith 1996, 2004). Fall fishing, as part of
the Ottawa adaptive pattern, is characterized by the aggregation of smaller,
seasonally mobile social units at key fishing locales (Kinietz 1965:239–240;
McClurken 1988:13–14). Evidence for aggregation, and consequent displays of
more overarching corporate group identity, may be seen in the ossuary burials
found at Juntunen (McPherron 1967:229–232; O’Shea 1988). O’Shea (1988:78)
notes, “Here the emphasis on territoriality by marking claim to an important
seasonal resource location is retained in the mortuary symbolism, although
now it is within the context of a large, collective ossuary. Furthermore, within
the Juntunen ossuary, the remains of individual family or lineage groups were
kept distinct and segregated [...]. This suggests that the deceased from several
distinct groups were processed and curated for ultimate interment within the
collective ossuary.”
While admittedly at a smaller scale, these burials resemble the histori-
cally known ‘Feast of the Dead’ mortuary ritual practiced by Algonquian and
western Iroquoian (particularly Huron) groups and documented at the nearby
17th century CE Lasanen site, at the Straits of Mackinac (Cleland 1971). On
these occasions all deceased members of a band or tribe were collectively
interred in a single pit, at intervals of a few to as many as ten years. Such collec-
tive, integrating, rituals re-inforced the social identity of both individuals and
smaller social units with the larger regional group by bringing together people
whose villages might be located across the broader region. This link to aggre-
gation and associated social ties is underscored by Brown’s (2003:219) recent
observations that “collective burial is present when communal solidarity is the
backdrop for burial rites in communities of a certain size and stability” and
“population size has to be of a sufficient size to warrant and to make scheduled
or periodic rites possible.” In this vein Cleland (1992a:101), consistent with
O’Shea (1988), notes that sites like Juntunen, which may be situated at the
point where band territories overlap, are natural points of aggregation.
The year-round security of food supplies at semipermanent sites was
problematic given the unreliability of harvests and the low density of animal
populations nearby. This unreliability implies that subsistence security had
to be insured through reciprocal exchange between groups comparable to
that seen in the Ottawa adaptive pattern (Milner and O’Shea 1998:199–201;
Smith 1996:286). Ceramics and other evidence supports this implication. Later
Juntunen phase ceramics have some Huron-like stylistic attributes, as might
be expected if the Huron were their trading partners (McPherron 1967:116;
Milner and O’Shea 1998:194). Further evidence of reciprocal exchange is
presented by Smith (1996:270–273), who suggests that the ritual killing and
burial of dogs at upper Great Lakes sites may represent the remains of feasts
offered by the hosts of exchange events. Widespread ritual feasting is indica-
tive of the cross-cultural ideology underlying exchange. Dog burials are found
at intensively occupied year-round sites such as the Providence Bay site on
Manitoulin Island in northern Lake Huron. Providence Bay is a shoreline
site easily accessible to traders from the southeast via Lake Huron, from the
northwest via the Saint Mary’s River, and from the west through the Straits
of Mackinac.
Evidence for reciprocal exchange can also be seen at Late Woodland earth-
work sites in the northern lower peninsula that date from around 900 to 1500
CE (Milner and O’Shea 1998:181). These sites have earthen berms surmounted
by wooden palisades and surrounded by ditches. Many earthworks have gaps
in the perimeter facilitating entry for trading parties along with clear areas in
the center where exchange could take place (Milner and O’Shea 1998:184,
187). One of the few earthworks to be excavated, Mikado (Carruthers 1969),
is characterized by relatively few artifacts, hearths and pit clusters, suggesting
short-term usage (Milner and O’Shea 1998:188). The fact that Mikado yielded
a large sample of maize, including stems and husk fragments, as well as cobs,
points to this short-term use as focused on scheduled exchange. The argument
that earthworks such as Mikado functioned as points of exchange is supported
by the fact that they lie at ecological and social boundaries (Milner and O’Shea
1998:199). These sites are situated in the headwaters of primary streams near
the juncture of interior uplands and coastal lowlands and near the transition
between the Carolinian and Canadian biotic provinces. Thus, the products
of diverse environments could be brought to these accessible locations for
exchange (O’Shea 2003:13).
Ethnographic analogues reveal that exchange between groups often occurred
at social boundaries, which constituted safe places that facilitated exchange. A
social boundary at Mikado is indicated by the patterning of ceramic attributes
on Juntunen ware found there (Milner and O’Shea 1998:199). These attri-
butes indicate distinct stylistic differences with ceramics made by horticultural
neighbors to the south (Younge tradition), but there is also a mix of attributes
suggesting interaction with these same neighbors. It is noteworthy that Mikado
is in a boundary location rather than a central place such as Providence Bay.
Milner and O’Shea (1998:200–201) note that such locations provide a place
of safety for socially disparate peoples to trade and, like Smith (1996), suggest
that feasting and dancing might have taken place in the earthwork ‘plazas’ to
symbolically cement the exchange relationship.
An Ottawa-pattern strategy of residential dispersal into extended family
groups to hunt in the winter, as needed, was employed during the Juntunen
phase. The same interior sites occupied in winter by Mackinac phase extended
family–sized groups were used by Juntunen phase peoples (Holman 1978:45).
Again, these sites were situated in relatively isolated and sheltered locations with
ecological settings high in potential yield for winter resources. Features and
artifacts at interior sites are indicative of domestic activities and hunting, with
ceramics limited to Juntunen ware, as would be expected for sites occupied by
small groups wintering in isolation. One site, O’Neil, is particularly interesting
because it shows that residential mobility on the part of Juntunen phase house-
holds was not restricted to winter, nor was it employed only in times of necessity.
Thus, households had the ability to make independent decisions with regard to
mobility. Lovis (1990:198) has isolated three Juntunen phase occupations at the
stratified multicomponent O’Neil site, including an early one, 1200–1300 CE,
one about 1440 CE, and a protohistoric occupation about 1700 CE.
O’Neil, located on the Lake Michigan coast south of Lake Charlevoix, was
clearly a warm-season habitation (Lovis 1973, 1990). During each Juntunen
phase occupation, O’Neil was a spatially partitioned residential site with
domestic activities taking place around hearths, separate areas for discarding
fish refuse, and still others set aside for the manufacture of stone tools from
Norwood chert obtained at the nearby Pi-wan-go-ning quarry (Cleland 1973).
Most fish were caught in the spring, and, clearly, the proximity of the quarry
was an important factor in deciding to occupy O’Neil during the Juntunen
phase. The diverse range of activities represented and the remains of an oval
house from the 15th century CE are indicative of residential moves to the
O’Neil site during the Juntunen phase. Evidence for logistic occupation at
O’Neil by contemporary ‘traverse phase’ residents of the transition zone shows
mutual use of the site, but whether this was scheduled for different times or
was concurrent is unknown (Lovis 1990:207). It is noteworthy that O’Neil is
the only Late Woodland coastal site south of the Straits of Mackinac that has
substantial Juntunen phase residential occupations. In contrast, other sites
along the coast were occupied repeatedly in the early Late Woodland Mackinac
phase (Holman 1978). Juntunen phase peoples made little use of the Lake
Michigan coast south of their fall fishing grounds. Rather, their occupations
were concentrated to take advantage of fall fishing and to act as middlemen in
the exchange of foodstuffs. Nonetheless, extended families made residential
moves in seasons of plenty as well as in times of scarcity and clearly functioned
as basic economic units in a decentralized society.
Discussion
Scheduled mobility along the coast and through the interior of the Canadian
Biotic Province and the transition zone was consistently used to procure
the wild resources of these environmental zones. Historically documented
mobility patterns, after about 1650 CE, are also evident in the Late Woodland
archaeological record. As predicted from the documents, sites are found in
coastal and interior settings where seasonal resources are likely to be found.
Furthermore, site structure, features and artifacts reflect site functions and
group compositions comparable to those of the historically known patterns,
while floral and faunal remains indicate the same seasons of occupation and
exploitation of the same resources. Variations in mobility include the Chippewa
and early Late Woodland Mackinac phase emphasis on residential mobility
in the Canadian Biotic Province and the Ottawa and Juntunen phase use of
logistic mobility, coupled with seasonal residential mobility in the transition
zone. This variability cannot be explained entirely by environmental differences
in the abundance and variety of foods. Like Mackinac phase peoples in the
Canadian Biotic Province, early Late Woodland populations in the transition
zone probably employed mobility to obtain a range of wild foods (Hambacher
1992). In cooperation with their Mackinac phase neighbors they buffered risk
from environmental uncertainties of the transition zone by sharing overlap-
ping territory and by using other groups’ territories when necessary (Holman
and Kingsley 1996). Thus, the structure of wild resources in the two environ-
mental zones fostered comparable residential mobility strategies in the early
Late Woodland, while cooperation between occupants of adjacent territories
provided options for movement in times of environmental stress. The differ-
ences between the Chippewa and Ottawa patterns are attributable to the fact
that coastal areas adjacent to and in the transition zone are more suitable for
horticulture than is the Canadian Biotic Province.
The Ottawa pattern appeared in the Juntunen phase, when maize became
significant in the Great Lakes region, and is predicated on growing crops and
maintaining an active Great Lakes fishery. Both horticulture and fishing were
conducted by extended family households at semipermanent coastal sites.
These activities acted as tethers for most of the labor force, as did production
of trade items such as reed mats (McClurken 1988:24). The Ottawa pattern
was a commitment to a semisedentary way of life supported by crops and fish
to secure sufficient food in a variable environment. Nonetheless, mobility was
critical because harvests were problematic, and fish was not the only source
of meat. Subsistence was made more secure by the use of logistic mobility for
both exchange and hunting. Such moves, for specific purposes, supplied addi-
tional maize and game from more distant sources when neither was plentiful
nor available around semipermanent residential sites. Logistic mobility was

particularly important after about 1500 CE, when the onset of the ‘Little
Ice Age’ made subsistence even more problematic (Smith 1996; Milner and
O'Shea 1998).
Logistic mobility, while critical for subsistence, was also subject to social
constraints. These strictures were most apparent in the exchange with settled
horticulturalists, like the Huron. The exchange which was a significant benefit
to both groups nonetheless required fostering of relationships. This trade was
reciprocal, with expectations of an immediate return in foodstuffs and other
items of equal value. Actions such as giving of gifts and marrying members of
other groups were necessary to ensure the exchange would continue. Without
constant effort this trade would fail (McClurken 1988:14–15). Social relations
also facilitated logistic mobility for hunting by providing access to territory that
was either shared with or controlled by another group. Unlike the case with
the Huron, these were relationships with groups such as the Chippewa, who
were regarded as close kin and carried the obligations of kinship. Thus, the
generosity with which kin treated one another fostered long-term cooperative
buffering via the mechanism of mobility. The Ottawa pattern, with its use of
residential mobility as a subsistence backup, reflects the continuing impor-
tance of maintaining the option to disperse across the landscape. Thus, people
were not entirely dependent on reciprocal exchange to make up subsistence
shortfalls. The use of residential mobility in turn implies that households had
considerable autonomy in economic decision making.
Both the Chippewa and Ottawa patterns of mobility persisted through
most of the 19th century CE (McClurken 1988; Cleland 1992b). Until their
northern lands were desired for American settlements and industries, such as
lumbering and mining, mobility was employed in the same systematic ways
to obtain the same seasonal resources in the same natural settings (Quimby
1966:179; Cleland 1983). Additionally, the same social devices were used to
buffer risk, especially including kinship ties, that provided access to overlap-
ping territories and reciprocal exchange with Europeans and the first American
farmers. Mobility was no longer an adaptive subsistence strategy when both
groups ceded most of their land to the United States and, thus, lost ready access
to dispersed resources.
Multiple lines of evidence demonstrate concordance between the late
prehistoric archaeological record and ethnohistorically documented flex-
ibility in mobility strategies and integrating social mechanisms that acted to
buffer environmental variability. This evidence includes site locations, site
structures, artifact assemblages and floral and faunal remains. After more
than 35 years of archaeological research in northern Michigan and the Straits
of Mackinac region, there is no evidence contradicting models based on the
historically documented Chippewa and Ottawa patterns. It is critical, however,

to carefully, continuously and independently test ethnohistorically derived
analogues against regional prehistoric evidence. For example, the prehistoric
evidence from southwestern lower Michigan does not support expectations
of the ‘Potawatomi pattern’ in that region. This pattern predicts large and
permanent agricultural villages like those inhabited by the Potawatomi, and
such large villages have yet to be found (O’Gorman 2003, 2007; O’Gorman
and Lovis 2006).
References
Albert, D. A., S. R. Denton and B. V. Barnes
1986 Regional Landscape Ecosystems of Michigan. Ann Arbor, University of
Michigan.
Bain, J. (ed.)
1969 Travels and Adventures in Canada and the Indian Territories Between the Years
1760 and 1776. By Alexander Henry, Fur Trader. Rutland, Charles E. Tuttle
(originally published in 1809).
Binford, L. R.
1980 Willow Smoke and Dogs’ Tails: Hunter-Gatherer Settlement Systems and
Archaeological Site Formation. American Antiquity 45: 1, pp. 1–17.
Brashler, J. G.
1973 A Formal Analysis of Prehistoric Ceramics from the Fletcher Site. Michigan State
University, Department of Anthropology (Master’s thesis).
Brashler, J. G., E. B. Garland, M. B. Holman, W. A. Lovis and S. R. Martin
2000 Adaptive Strategies and Socioeconomic Systems in Northern Great Lakes
Riverine Environments: The Late Woodland of Michigan. In T. E. Emerson,
D. L. McElrath and A. C. Fortier (eds.), Late Woodland Societies: Tradition and
Transformation Across the Midcontinent. Lincoln, University of Nebraska Press:
pp. 543–582.
Brown, J. A.
2003 Collective Burial Practices Across the Agricultural Transition in the Eastern
Woodlands. In G. Burenhult and S. Westergaard (eds.), Stones and Bones:
Formal Disposal of the Dead in Atlantic Europe During the Mesolithic Neolithic
Interface, 6000–3000 BC. BAR international Series 1201. Oxford, Archaeopress:
pp. 207–223.
Carruthers, P. J.
1969 The Mikado Earthwork: 20Aa5. University of Calgary, Department of
Archaeology (Master’s thesis).
Cleland, C. E.
1966 The Prehistoric Animal Ecology and Ethnozoology of the Upper Great Lakes Region.
Anthropological Papers Number 29. Ann Arbor, Museum of Anthropology,
1967 Progress Report on National Science Foundation Grants Nos. GS-1026 and
GS-1669: The Environmental Adaptations of the Prehistoric Cultures of the
Grand Traverse Bay Area of Michigan. Report on file at the Consortium for
Archaeological Research, Michigan State University, East Lansing.
1971 The Lasanen Site: An Historical Burial Locality in Mackinac County, Michigan.
Publications of the Museum of Michigan State University. Volume 1, Number 1.
East Lansing, Michigan State University.
1973 The Pi-wan-go-ning Prehistoric District at Norwood, Michigan. In H.
M. Martin (ed.), Geology and the Environment: Man, Earth, and Nature in
Northwestern Lower Michigan. Ann Arbor, Michigan Basin Geological Society:
pp. 85–87.
1982 The Inland Shore Fishery of the Northern Great Lakes: Its Development and
Importance in Prehistory. American Antiquity 47: 4, pp. 761–784.
1983 Indians in a Changing Environment. In S. Flader (ed.), The Great Lakes Forest:
An Environmental and Social History. Minneapolis, University of Minnesota
Press: pp. 83–95.
1989 Comments on “A Reconsideration of Aboriginal Fishing Strategies in the
Northern Great Lakes Region,” by Susan R. Martin. American Antiquity 54:
3, pp. 605–609.
1992a From Ethnohistory to Archaeology: Ottawa and Ojibwa Band Territories of
the Northern Great Lakes. In B. J. Little (ed.), Text-Aided Archaeology. Boca
Raton, CRC Press: pp.97–102.
1992b Rites of Conquest: The History and Culture of Michigan’s Native Americans. Ann
Arbor, University of Michigan Press.
Clifton, J. A.
1984 The Pokagons, 1683–1983: Catholic Potawatomi Indians of the St. Joseph River
Valley. Lanham, University Press of America.
Densmore, F.
1979 Chippewa Customs. St. Paul, Minnesota Historical Society Press. (originally
published in 1929 as Bulletin 86 of the Bureau of American Ethnology,
Smithsonian Institution, Washington).
Doepker, R. V. and J. J. Ozoga
1991 Wildlife Values of Northern White Cedar. In Workshop Proceedings on Northern
White Cedar in Michigan: Agricultural Experiment Research Report 512. East
Lansing, Michigan State University: pp. 15–34.
Fitting, J. E.
1966 Archaeological Investigations of the Carolinian-Canadian Edge in Central
Michigan. In J. E. Fitting (ed.), Edge Area Archaeology (special issue), Michigan
Archaeologist 12: 4, pp. 143–150.
1975 The Archaeology of Michigan: A Guide to the Prehistory of the Great Lakes Region.
Bulletin Number 56. Bloomfield Hills, Cranbrook Institute of Science.
Fitting, J. E. and C. E. Cleland
1969 Late Prehistoric Settlement Patterns in the Upper Great Lakes. Ethnohistory
16: pp. 289–302.
Hambacher, M. J.
1992 The Skegemog Point Site: Continuing Studies of Cultural Dynamics in the
Carolinian-Canadian Transition Zone. East Lansing. Michigan State University
(PhD dissertation).
Holman, M. B.
1978 The Settlement System of the Mackinac Phase. East Lansing. Michigan State
University (PhD dissertation).
1984 The Identification of Late Woodland Maple Sugaring Sites in the Upper
Great Lakes. Midcontinental Journal of Archaeology 9: 1, pp. 63–90.
Holman, M. B. and R. G. Kingsley
1996 Territoriality and Societal Interaction During the Early Late Woodland in
Southern Michigan. In M. B. Holman, J. G. Brashler and K. E. Parker (eds.),
Investigating the Archaeological Record of the Great Lakes State: Essays in Honor
of Elizabeth Baldwin Garland. Kalamazoo, New Issues Press: pp. 341–382.
Holman, M. B. and F. Krist Jr.
2001 Late Woodland Storage and Mobility in Western Lower Michigan. In T. C.
Pleger, R. A. Birmingham and C. I. Mason (eds.), Papers in Honor of Carol I.
Mason. (special issue), Wisconsin Archeologist 82: 1–2, pp. 7–32.
Howey, M. C. L. and J. M. O’Shea
2006 Bear’s Journey and the Study of Ritual in Archaeology. American Antiquity 71:
pp. 261–282.
Kinietz, W. V.
1965 The Indians of the Western Great Lakes, 1615–1760. Ann Arbor, University of
Michigan Press.
Lovis, W. A.
1973 Late Woodland Cultural Dynamics in the Northern Lower Peninsula of Michigan.
East Lansing. Michigan State University (PhD dissertation).
1976 Quarter Sections and Forests: An Example of Probability Sampling in the
Northeastern Woodlands. American Antiquity 41: pp. 364–372.
1978 A Numerical Taxonomic Analysis of Changing Woodland Site Location
Strategies on an Interior Lake Chain. Michigan Academician 1: 1, pp 39–48.
1990 Site Formation Processes and the Organization of Space at the Stratified
Late Woodland O’Neill Site. In G. E. Gibbon (ed.), The Woodland Tradition
in the Western Great Lakes: Papers Presented to Elden Johnson. Publications in
Anthropology Number 4. Minneapolis, University of Minnesota: pp. 195–211.
Martin, S. R.
1989 A Reconsideration of Aboriginal Fishing Strategies in the Northern Great
Lakes Region. American Antiquity 54: 3, pp. 594–604.
1999 A Site for all Seasons: Some Aspects of Life in the Upper Peninsula During
Late Woodland Times. In J. R. Halsey and M. D. Stafford (eds.), Retrieving
Michigan’s Buried Past: The Archaeology of the Great Lakes State Bulletin 64.
Bloomfield Hills, Cranbrook Institute of Science: pp. 221–227.
Mason, C. I.
1988 Introduction to Wisconsin Indians. Salem, Sheffield Publishing.
McClurken, J. M.
1988 We Wish to Be Civilized: Ottawa-American Political Contests on the Michigan
Frontier. East Lansing. Michigan State University (PhD dissertation).
McPherron, A.
1967 The Juntunen Site and the Late Woodland Prehistory of the Upper Great Lakes
Area. Anthropological Papers Number 30. Ann Arbor, Museum of Anthropology,
Milner, C. M. and J. M. O’Shea
1998 The Socioeconomic Role of Late Woodland Enclosures in Northern Lower
Michigan. In R. C. Mainfort and L. P. Sullivan (eds.), Ancient Earthen Enclosures
of the Eastern Woodlands. The Ripley P. Bullen Series. Gainesville, University
Press of Florida: pp. 181–201.
Murphy, C. and N. Ferris
1990 The Late Woodland Western Basin Tradition of Southwestern Ontario. In C.
J. Ellis and N. Ferris (eds.), The Archaeology of Southern Ontario to A.D. 1650.
Occasional Publications of the London Chapter, Publication No 5. London, Ontario
Archaeological Society: pp. 189–278.
O’Gorman, J. A.
2003 The “Agricultural Village” Myth and Recent Research at Moccasin Bluff.
Michigan Archaeologist 49: 1–2, pp. 91–100.
2007 The Myth of Moccasin Bluff—Rethinking the Potawatomi Pattern.
Ethnohistory 54: in press.
O’Gorman, J. A. and W. A. Lovis
2006 Before Removal: An Archaeological Perspective on the Southern Lake
Michigan Basin. Midcontinental Journal of Archaeology 31: pp. 21–56.
O’Shea, J. M.
1988 Social Organization and Mortuary Behavior in the Late Woodland Period in
Michigan. In R. Yerkes (ed.), Interpretations of Culture Change in the Eastern
Woodlands During the Late Woodland Period. Occasional Papers in Anthropology

Number 3. Columbus, Ohio State University: pp. 68–85.
2003 Inland Foragers and the Adoption of Maize Agriculture in the Upper Great
Lakes Region of North America. Before Farming: The Archaeology of Old-World
Hunter-Gatherers 2: pp. 1–21.
O’Shea, J. M. and P. Halstead
1989 Conclusion: Bad Year Economics. In P. Halstead and J. O’Shea (eds.), Bad
Year Economics. Cambridge, Cambridge University Press: 123–126.
O’Shea, J. M. and C. M. Milner
2002 Material Indicators of Territory, Identity, and Interaction in a Prehistoric
Tribal System. In W. Parkinson (ed.), The Archaeology of Tribal Societies. Ann
Arbor, International Monographs in Prehistory.
Parker, K. E.
1996 Three Corn Kernels and a Hill of Beans: The Evidence for Prehistoric
Horticulture in Michigan. In M. B. Holman, J. G. Brashler and K. E. Parker
(eds.), Investigating the Archaeological Record of the Great Lakes State: Essays
in Honor of Elizabeth Baldwin Garland. Kalamazoo, New Issues Press: pp.
307–339.
Pilling, A.
1968 Discussions: Part III. In R. B. Lee and I. DeVore (eds.), Man the Hunter.
Chicago, Aldine: pp. 155–156.
Quaife, M. M.
1921 Alexander Henry’s Travels and Adventures in the Years 1760–1776. Chicago,
Lakeside Classics.
Quimby, G. I.
1966 Indian Culture and European Trade Goods: The Archaeology of the Historic Period
in the Western Great Lakes Region. Madison, University of Wisconsin Press.
Smith, B. A.
1996 Systems of Subsistence and Networks of Exchange in the Terminal Woodland and
Early Historic Periods in the Upper Great Lakes. East Lansing. Michigan State
University (PhD dissertation).
2004 The Gill Net’s “Native Country”: The Inland Shore Fishery in the Northern
Lake Michigan Basin. In W. A. Lovis (ed.), An Upper Great Lakes Archaeological
Odyssey: Essays in Honor of Charles E. Cleland. Bloomfield Hills, Cranbrook
Institute of Science: pp. 64–84.
Smith, J.
1907 An Account of the Remarkable Occurrences in the Life and Travels of Col. James
Smith During His Captivity with the Indians in the Years 1755, ’56, ’57, ’58,
’59 (with an appendix of illustrative notes by Wm. M. Darlington of Pittsburgh).
Cincinnati: Robert Clarke (originally published as Ohio Valley Historical
Series Number 5).
Spielmann, K. A.
1986 Interdependence Among Egalitarian Societies. Journal of Anthropological
Strong, W. D.
1935 An Introduction to Nebraska Archeology. Miscellaneous Collections Number 93.
Washington, Smithsonian Institution.
1936 Anthropological Theory and Archaeological Fact. In R. H. Lowie (ed.), Essays
in Anthropology presented to A. L. Kroeber. Berkeley, University of California
Press: pp. 359–370.
Tanner, H. H.
1987 Atlas of Great Lakes Indian History. Norman, University of Oklahoma Press.
Tanner, J.
1956 A Narrative of the Captivity and Adventures of John Tanner During Thirty Years
Among the Indians in the Interior of North America. Minneapolis, Ross and
Haines.
Trigger, B. G.
1969 The Huron Farmers of the North. New York, Holt, Rinehart and Winston.
Wedel, W. R.
1938 The Direct-Historical Approach in Pawnee Archeology. Miscellaneous Collections
Number 97. Washington, Smithsonian Institution.
Williams, B. J.
1968 The Birhor of India and Some Comments on Band Organization. In R. B.
Lee and I. DeVore (eds.), Man the Hunter. Chicago, Aldine: pp. 126–131.
Wobst, H. M.
1968 The Butterfield Site, 20 BY 29, Bay County, Michigan. In J. E. Fitting, J.
R. Halsey and H. M. Wobst (eds.), Contributions to Michigan Archaeology.
Anthropological Papers Number 32. Ann Arbor, Museum of Anthropology,
University of Michigan: pp. 173–275.
Yarnell, R. A.
1964 Aboriginal Relationships Between Culture and Plant Life in the Upper Great
Lakes Region. Anthropological Papers Number 23. Ann Arbor, Museum of
Anthropology, University of Michigan.
Ch a p t e r 1 4
Nomadic Potters
Relationships Between Ceramic
Technologies and Mobility
Strategies
J e l m e r W. E e r k e n s
T here is always a give-and-take when people, and societies collectively,

incorporate the production and use of new material technologies. People
usually alter the technology to suit their needs, but they are also often funda-
mentally changed by the technology as well. Such changes can come quickly
and consciously, as deliberate adjustments are made, or they may happen at a
slower and imperceptible rate. This chapter examines the interplay between
material technology and one aspect of prehistoric lifestyle: settlement strate-
gies, especially residential mobility.1 Lithic analysis has been heavily engaged
in this type of research because stone tools are a common find at archaeological
sites spanning the transition between different mobility strategies (Rafferty
1985; Parry and Kelly 1987; Kelly 1988; Basgall 1989; Henry 1989; Lurie
1989; Rosen 1989; Andrefsky 1991; Bamforth 1991; Odell 1996; Thacker
1996). This chapter examines a second, in this context much less studied, tech-
nology: pottery. As a case study I draw from the Mojave Desert and the North
American Great Basin, or more properly the Basin and Range, where I have
been focusing my research for the last decade (Figure 14.1). I will refer to this
area simply as the Western Great Basin, although I recognize that it includes
parts of the Mojave Desert and the Western Sierra Nevada Mountains, which
are technically not part of the Basin and Range geographic province.
The late prehistoric (around 700–100 BP) archaeological record in the
Western Great Basin provides an interesting case for examining these relation-
ships for several reasons. First, the aboriginal populations of this area were
1.
In this chapter I focus on residentially mobile hunting-and-gathering populations.
However, as discussed during the workshop at the Cotsen Institute of Archaeology at
UCLA, June 2004, there ought to be analogous predictions for the material technolo-
gies of nomadic pastoralists as well. Some of the discussion in this chapter may apply in
such settings, but others, such as the heaviness of technologies, may be less restrictive
elements.
307
308     J e l m e r W. E e r k e n s
Figure 14.1. Map of the border region between California and Nevada, showing the study areas (in
gray).
composed of hunting-and-gathering groups, controlling for a subsistence

mode in the study. Second, ethnographic and archaeological data indicate
several settlement patterns were found within the region, ranging from
nearly sedentary in the Owens Valley to highly mobile in the Mojave Desert.
Third, although ethnographic data on pottery making are scarce, pottery is a
common component of late prehistoric sites across the entire region and has
been the subject of archaeological investigation. Although the coarseness of
the archaeological record does not allow us to determine exactly how pottery
affected people and vice versa, it does allow us to examine general patterns
between pottery making, pottery use, and mobility strategies.
N o ma d i c P o tt e r s      309
Mobility and Pottery Technology

There has been a common wisdom or stereotype among many archaeologists
that sedentism, agriculture and pottery technologies are necessarily and positively
correlated. Indeed, some archaeologists use the presence or absence of pottery in
the archaeological record as an independent measure of residential mobility. The
presence of potsherds at a site would indicate sedentism, and a lack of pottery
would suggest some degree of seasonal transhumance. Although examples of
pottery in sites occupied by mobile hunter-gatherers are known (Arnold 1985), as
Sassaman (1993:2–3) notes, in such cases the pottery is usually defined as ‘crude’
and ‘technologically unimpressive’ thereby relegating it to a status of little or
no significance and reinforcing the stereotype (but see Barnard, this volume).
There are, of course, good reasons to believe that mobile hunter-gatherers should
not make pots. I present five conflicts or problems that hinder the use of pots
among mobile societies, both hunter-gatherers and pastoralists. To make pottery
a worthwhile technology, mobile societies must resolve these issues.
First, pots are heavy relative to other containers and are taxing to carry
around during seasonal movements. This is particularly true for societies
that do not have pack animals and societies in which goods must be carried
by people on their backs. And even among societies that have pack animals,
including dogs, the extra energy required by the animals to move heavy objects
requires additional nourishment and grazing time. When lighter alternatives,
such as baskets, gourds, animal-skin pouches or other containers are available,
such technologies should be more attractive from an energy standpoint than
heavy fired clay ones.
Second, because they are relatively fragile, pots are exposed to high rates of
breakage when carried around during residential movements. Although they
can be insulated from impact shock by being packed in softer materials, each
packing up and unpacking of pots increases the chance of breakage. Accidents,
such as dropping during transport, increase the risk of breakage. These risks,
however small, make ceramic pots inferior relative to other container technolo-
gies for mobile groups. Again, baskets and hides are more resistant to impact
shock and should be preferable to pots.
Third, mobile peoples may not stay in one place long enough to see the
pottery production cycle through. From the collection of clay, to forming a pot,
to drying and firing it, the making of a clay pot can take from several days to
several weeks (Arnold 1985). In particular, the crucial step of drying an unfired
pot may take several days and may require significant oversight to ensure an
even and thorough drying. Mobile people, particularly during certain times
of the year, may have to move before they can complete the steps necessary
to produce pots.
310     J e l m e r W. E e r k e n s
Fourth, and related to the third issue, is that the most opportune time to
produce pots, the dry season, is also the time when many seeds, nuts, berries
and greens ripen. Time conflicts between gathering food that is only available
during narrow temporal windows and the production of ceramic pots may
make the latter too expensive. This is particularly true if a significant quantity
of plant products must be harvested and stored for later consumption. One
way to solve such a time conflict is to divide these tasks by sex or some other
societal division. Yet plant gathering and pottery production are both typically
performed by women, even among mobile hunter-gatherers.
Fifth, the small population sizes typically encountered among mobile
hunter-gatherers tends to limit the demand for pots. As discussed by Brown
(1989), one of the significant advantages of pots over other containers is that
there is an economy of scale in the production of the former. As multiple pots
can be fired at once, this step can be performed almost as easily for one pot as
for a dozen. This is not true of other common containers, such as baskets or
sewn skins, where each item must be made individually, and making a dozen
takes twelve times as long (and twelve times as much raw material) as making
one. In arid landscapes, such as the Great Basin, limited fuel resources make
this factor of particular importance (Bettinger et al. 1994). Fuel needed for
cooking and warmth may be in such high demand that the fuel needed to fire
only a small number of ceramic vessels may simply be unavailable.
In sum, ceramic technologies do not lend themselves well to a mobile
lifestyle. Yet we know of many archaeological and ethnographic examples
of pottery making in mobile societies (Arnold 1985; Barnard, this volume).
Indeed, the origins of ceramic technologies often occur within such settings
(Ikawa-Smith 1976; Reid 1984; Aikens 1995; Close 1995: Hoopes and Barnett
1995; Rice 1999). So why do these mobile groups engage in pottery produc-
tion? And when they do, how does their mobility affect the way they organize
the production and use of pots? And finally, how do they resolve the conflicts
and problems listed above? These issues have not been extensively investi-
gated by archaeologists. Simms et al. (1997; see also Bright and Ugan 1999)
have explored some of these questions in the Eastern Great Basin, and Braun
(1983) has done so in the eastern United States, but there are few published
accounts on this topic. This chapter builds on their work but takes it in a
different direction by examining pottery and mobility practices within the
Western Great Basin.
Steward (1938) believed residential mobility in the Great Basin to be
inversely correlated with population density, which itself was correlated
to precipitation and bioproductivity. Residential mobility was necessary to
exploit spatially variable and low-density food resources. People followed the
distribution and availability of ripening plant foods, especially piñon nuts and
Table 14.1. Population Levels, Average Precipitation and Estimated Degree of Residential Stability for
Six Regions of the Western Great Basin:
Region Population Annual Mobility Sherds/acre Sherds/projectile Pottery

densitya rainfalla rank point rank
Western Sierra 0.5 57.7 1 N/A 11.9 3
Northern Owens 2.1 16.0 2 0.02 5.2 6
Southern Owens 2.1 14.5 3 0.14 33.9 1
Deep Springs 10.7 15.4 4 0.06 7.9 4
Death Valley 30.0 5.3 5 0.09 17.7 2
Northern Mojave >30 11.9 6 0.05 7.7 5
a
The population density is given in persons/square mile (Delacorte 1990; Eerkens 2003a), the annual
rainfall in cm/year.
small seeds. Table 14.1 gives the estimated population density (persons/square
mile), precipitation (cm/year), and level and density of pottery for six regions,
representing a range of different populations within the Western Great Basin,
though all are hunter-gatherers practicing some degree of residential mobility.
The table ranks residential mobility and the degree of engagement in pottery.
It is evident that the degree of residential mobility did not have a predictable or
consistent effect on the amount or density of pottery in a region. This suggests
that mobility did not affect the degree of reliance on pots in the material culture
(Eerkens 2003a).
Although there is variability, vessels from the Western Great Basin are
generally medium-sized (18–22 cm high and 18–25 cm wide at the mouth) and
undecorated. In most areas, less than 10% of the rim sherds have fingernail-
impressed decoration around the rim (Eerkens 2001). Painting, slipping and
burnishing are not evident. Straight-sided and direct-rimmed conical boiling
pots are the most common vessel type, though spherical bowls with recurved
rims are also present, particularly in the eastern part of the Western Great
Basin (Bettinger 1986; Pippin 1986; Prince 1986; Lyneis 1988; Touhy 1990).
Vessels were constructed mainly by stacking coils of clay onto a circular disk
base, welding the coils together by scraping with the fingers, a bundle of sticks
or a small object. Sand or crushed rock temper is usually present and was prob-
ably part of the local clay matrix. Organic temper is occasionally present in the
form of grass blades and other vegetable matter. Vessels were fired at relatively
low temperatures (around 600°C) and appear brown-red in color, giving rise
to the general category of ‘brown wares.’ Figure 14.2 shows a nearly complete
pot found in Owens Valley, typical for the size and shape of the region. Holes
near the rim of the pot were drilled on either side of a crack. Cordage would
have been looped through the holes in an attempt to hold the pot together
and extend its use life.
312     J e l m e r W. E e r k e n s
Figure 14.2. Whole cooking pot from Owens Valley, California, with repair holes near the rim.
Pottery making is clearly a late technology in the Western Great Basin.

Dating of ceramics has not been actively pursued in most regions, but it is clear
that they are consistently associated with other artifacts that date to the latest
period in prehistory, after 700 BP (Pippin 1986; Rhode 1994; Feathers and
Rhode 1998). In Owens Valley, people seem to have been experimenting with
ceramic technologies around 1200 years ago (Eerkens et al. 1999), yet the craft
does not become commonplace until 500–700 years ago (Delacorte 1999). Pots
seem to have been used mainly for boiling seeds and other vegetable products
and only rarely were used to process meat (Eerkens 2005). Pots were rarely
carried more than 80 km outside their region of manufacture, and production
was organized at a small-scale or family level (Eerkens et al. 2002).
Resolving Conflicts
The presence of pottery among the mobile hunter-gatherers of the Western
Great Basin is particularly interesting in light of the problems and conflicts
mentioned above. Basketry technologies were highly developed in this region,
such that they could perform virtually all the tasks that pots could. Baskets were
woven so tightly that they could hold water, were durable and strong enough
to boil foods, and were long-lasting enough to store and serve foods. All these
activities could be performed at a fraction of the weight of ceramic pots and
with much greater resistance to impact stress. Baskets, then, would seem ideally
suited to a mobile lifestyle. So why did these people ever get into the business
of making pots? Similarly, how did they modify pottery technologies to suit
their lifestyle, and how did the use of pottery modify their lifestyle? I address
these issues by examining how the Paiute, Shoshone, and the Mojave Desert
people of the study area resolved the five problems listed above.
One solution to the heaviness and fragility of pots is simply not to move
them at all. Caching pots may have been a way to avoid carrying them during
the seasonal round (Eerkens 2003a). In the Western Great Basin two pieces
of information suggest that caching was an important strategy used to deal
with these problems. First, though uncommon, cached pots from rockshelters
and caves have occasionally been recorded and described by archaeologists
(Campbell 1931; Wallace 1965; King 1976; Bayman et al. 1996). The discovery,
Figure 14.3. Distribution of pottery in Owens Valley (data from Bettinger 1975).
314     J e l m e r W. E e r k e n s
Figure 14.4. The Owens River, near Big Pine.
in March 2002, of a nearly complete pot stowed away in a narrow rock crevice
near Little Lake, just south of Owens Valley, is a clear example of a cached pot
(Eerkens n.d.). All these caches are in lowland locations. Second, the distribu-
tion of potsherds across the landscape is clearly uneven, heavily skewed toward
valley-bottom and wetland locations, as shown in Figure 14.3 (Eerkens 2003a).
Figures 14.4 through 14.7 depict some of these different environments in and
to the east of Owens Valley, California, all within several kilometers of one
another. Note the dramatic changes in both the density and makeup of vegeta-
tion communities in these different environments.
Alfred Kroeber (1922) recognized this pattern but did not attach any partic-
ular meaning to it. Subsequent archaeological investigations have supported
his impressions. Surveys in many valleys in the Central and Western Great
Basin demonstrate that the frequency of pottery is significantly higher in
riverine and lakeside locations on the valley bottom (Hunt 1960; Thomas 1971,
1983; Bettinger 1975; Wallace 1986; Weaver 1986; Delacorte 1990; Plew and
Bennick 1990; Gilreath and Hildebrandt 1997; Hildebrandt and Ruby 1999).
Table 14.2 shows the results from five regions with comparable survey strate-
gies and coverage. Between 63–100% of the potsherds are located near the
valley bottom, even when adjusted for the total number of artifacts found or
the area surveyed (Thomas 1971; Bettinger 1975; Delacorte 1990; Gilreath and
Hildebrandt 1997; Hildebrandt and Ruby 1999). Despite the fact that archaeo-
logical sites from the ceramic period are present in all parts of the landscape,
Figure 14.5. The Owens Valley desert shrub landscape.
Figure 14.6. The Piñon-Juniper landscape of the White Mountains (courtesy of the Far Western
Anthropological Research Group).
316     J e l m e r W. E e r k e n s
Figure 14.7. The high altitude alpine desert landscape of the White Mountains.
from valley bottom to alpine zones, pottery is differentially distributed within

the former. This suggests that locations across the valley bottom were where
people broke, and presumably used, the majority of their pots and that pots
were not often carried to other parts of the landscape. Besides the fact that it
has all the resources necessary to make pots (clay, sand, water and firewood), a
major advantage of the valley bottom is that this is a predictable source of water.
As a result, the food resources in these locations are spatially and temporally
predictable, particularly when compared with other Great Basin resources such
as piñon nuts and dryland seeds (Thomas 1972). Thus, caching only works in
areas that have relatively stable and predictable food resources.
Although caching pots would have solved the heaviness and fragility prob-
lems, it would have had major repercussions for the lifestyles of people doing
so. In particular, since they could not have cached a pot at every spot on the
landscape, caching would have tethered people to particular points on the
landscape, where they had left their pots. This would result in higher rates of
site re-occupation, also referred to as ‘occupational redundancy’ or ‘persistent
places’ in the literature (Eerkens 2003a). Such tethering may have promoted
reliance on foods associated with these locations and may have encouraged the
notion of landownership and territoriality. It is also likely that caching behavior
would have prompted people to modify pottery technologies to make their
products more suitable for storage during the off-season. Thicker and stronger
pots may have been the by-product of this behavior.
Table 14.2. Distribution of Pottery by Environmental Zone (adjusted by area surveyed):
Valley bottom Piñon/juniper Above piñon Reference

Northern Owens 86% 12% 4% Bettinger 1975
Deep Springs 64% 36% 0% Delacorte 1990
Gilreath and Hildebrandt
Northern Mojave 63% 37% N/A 1997; Hildebrandt and
Ruby 1999
Reese River 100% 0% 0% Thomas 1971
Monitor Valley 76% 15% 9% Thomas 1983
The third conflict, the need to be in one place long enough to see the pro
duction cycle through, may have been solved by remaining in certain spots
for longer periods or to occupy them more frequently. Combined with the
reasoning above, spending more time at these locations may have led to a
positive feedback cycle with the tethering and caching behavior. In particular,
spending longer periods of time in places where pots were made and cached
may have promoted an increased reliance on the resources available in those
areas. If pots had to be constructed for use, and were not already available as
cached items, people may have needed to arrive several days or weeks ahead
of the availability of such resources to prepare for such activities. Time spent
making pots while seed resources were ripe would have subtracted from time
that could be devoted to gathering. It is possible that women traveled to valley
bottom seed patches several months prior to, or after, the availability of seeds,
constructed pots there and cached them for future use. This would have also
solved the fourth conflict, namely time conflicts in the dry season between seed
gathering and pot production.
All these factors would have forced individuals to alter the way that they
made pottery by limiting the amount of time devoted to production. A
minimum of time investment is consistent with the ceramic technology seen
in the Western Great Basin. The minor amount of decoration, the lack of
extensive surface finishing (burnishing, polishing, slipping) and the minimum
attention given to symmetry and evenness (rims are often undulating and walls
often have a variable thickness) all indicate that pots were hastily made. The
use of sand and crushed rock as temper, most likely native to the matrix from
which the clay was collected (Schaefer 2003), also indicates little investment in
production activities. Moreover, if pots were constructed in the rainy season,
there would have been a need for quick-drying pots, to minimize the chances
of their getting wet again while drying.
Several methods exist to reduce the time required to dry a pot prior to firing,
including addition of fiber temper, roughening the exterior or thinning the pot
(Skibo et al. 1989; Schiffer et al. 1994). All of these factors are evident in areas
318     J e l m e r W. E e r k e n s
where people were more mobile. Although fiber temper is not dominant in pots
from the Western Great Basin, it is present in most sherds in small amounts,
again supporting the notion that these pots were constructed to minimize time
investment. These findings are consistent with those of Simms et al. (1997), who
suggest that mobile foragers invested less in their ceramic pots than sedentary
agriculturalists in the Eastern Great Basin. At the same time, there are indica-
tions that within the mobile hunter-gatherer groups, greater residential mobility
may have fostered greater investment in ceramic technologies, likely because of
the greater demands that residential mobility places on the material technology
(Eerkens 2003a, 2003b). Pots in regions where people were more mobile are
smaller in diameter, thinner, and more often roughened on their exterior surface;
they also contain finer temper and are less diverse in size and shape. Although
other factors can contribute to these attributes, such as the intended vessel func-
tion, the nature and availability of clays, and different learning traditions, there
are good reasons to believe that mobility would have an effect on these attributes
(Simms et al. 1997). Finally, as Brown (1989) has argued, one of the main advan-
tages of pottery technologies is the economy of scale in production. Pottery is
a particularly advantageous technology when large numbers of containers are
needed. Unfortunately, mobile societies usually maintain low population densi-
ties that do not allow them to take advantage of this attribute of pottery.
All indications from ethnographic and archaeological data confirm that
the Western Great Basin was home to low population densities. Although
population density certainly varied across the region (Steward 1938), it is
clear that areas with higher population did not necessarily produce more pots.
Some of the highest densities of pottery in the study area occur in areas where
people were quite mobile, and some of the lowest densities occur in areas with
semisedentism (Eerkens 2003a). Nor does it appear that pottery production was
organized at a higher regionwide level by a few specialists who provided pots
to a large area to take advantage of the economy of scale. Pots in the Western
Great Basin seem to have been produced on a small scale at a local family or
village level (Eerkens et al. 2002).
Low-temperature firing, to conserve fuel, and extensive efforts to increase
the use life of pots, for example by repairing cracks (Figure 14.2), may have
been responses to the expenses involved in pottery production. These strate-
gies may have been employed to make the costs of production versus artifact
use life for pots more equal to that of baskets. Once people were investing in
the production of pots, it is likely that pots were put to an increasing range of
uses. Analysis of the potsherds from the region suggests an increase in shape
and size diversity over time. As people became familiar with the technology,
they altered the design to increase heating efficiency and minimize the amount
of raw materials needed by making pots thinner (Eerkens 2003b). Once the
technology was incorporated into the lives of prehistoric people in the Western
Great Basin, it gradually encouraged other changes in day-to-day activities.
Discussion
There are many hurdles for mobile peoples to clear before they can incorporate
pottery technologies within their material culture. Overall, these hurdles may
account for the general relationship between pottery use and mobility strate-
gies seen worldwide. More specifically, most fully sedentary and semisedentary
societies use pots (91% and 75% respectively), but less than a third of mobile
nomadic people engage in this activity (Arnold 1985). Yet in some instances
mobile groups are able to resolve these issues to make the technology work
for them. The Western Great Basin was one of these areas.
Rather than disregard ceramic technologies altogether, late prehistoric
societies of the Western Great Basin were extremely inventive and designed
ways around the problems that typically beset mobile societies. They actively
manipulated aspects of the form, function, use and production of their pots to
fit this technology within their mobile lifestyles, that is, to suit their specific
needs. It also appears that the use of pots had effects on the lifestyles of these
people. It probably tethered people to particular tracts of land, promoted an
increased reliance on the resources available in these areas, especially seeds, and
may have required much travel and foresight to produce and cache pots ahead
of time in patches where high seed yields were anticipated. While tethering
may promote decreased mobility or sedentism in the long run (Kelly 1990;
1995), mobile peoples can also be tethered to certain locations by making
consistent and repeated use of them (occupational redundancy). Caching and
occupational redundancy allowed such groups to take advantage of technolo-
gies that are normally reserved for more sedentary groups, including the use
of heavy or fragile tools such as ceramic pots. In this respect caching may be
an important strategy for mobile pottery-using hunting-and-gathering groups.
The success of caching is highly dependent, however, on the spatial predict-
ability of the resources for which the tools are needed.
There is little patterning in the degree of residential mobility versus the
amount of pottery (Table 14.1). This suggests that once people resolved the
conflicts associated with pottery production, they were free to engage in vessel
production to whatever degree was necessary. In other words, once people had
figured out how to incorporate pot production and use into their lifestyles (by
redesigning shape, temper and texture, as well as by caching), the degree of resi-
dential mobility did not have an influence over how many vessels they made. It
did, however, affect how they made pots. A design favoring rapid drying, increased
postfiring strength, overall lightness and durability was clearly favored.
320     J e l m e r W. E e r k e n s
The above may explain how it was that mobile people were able to make
and use pots. A remaining question, of course, is the matter of why they did.
Although this question lies beyond the scope of this chapter, I have argued else-
where that the main reason for this change relates to demands on the time and
labor of women (Eerkens 2001; see also Crown and Wills 1995). Prior to the
adoption of pottery, stone boiling in baskets was the main method for boiling
foods, especially small seeds. Figure 14.8 shows the density of small seeds
recovered from flotation studies from house floors in Owens Valley. Clearly,
low numbers of seeds were eaten from the 3rd century CE onward (Eerkens
2004). These were probably boiled in baskets; however, around 1350 CE the
density of small seeds greatly increased, and seed boiling must have become
a major activity. Stone boiling in baskets is an inefficient method because it
demands constant attention from women to replace cooled stones with heated
ones and to avoid burning holes in the bottom of the basket. Pots may have
provided a more efficient boiling container than baskets because they can be
set over the fire with little further attention. As a result, greater numbers of
seeds could be processed at once.
In conclusion, the restrictions on technology imposed by a residentially
mobile lifestyle may force mobile groups to modify technologies in predictable
ways. For example, we may expect to see more standardization in certain attri-
butes, especially size, shape and weight. A mobile lifestyle may not allow for a
range of shapes to be made and used, and experimentation with new designs
Figure 14.8. Density over time of the concentration of small seeds found on house floors in Owens
Valley (after Eerkens 2004).
may not be possible, particularly in more marginal environments where the

costs of failure are high. Similarly, we may expect to see a low amount of time
invested in these technologies (Simms et al. 1997; Bright and Ugan 1999). Only
after people become more sedentary, and the craft becomes established, will we
see elaboration in shapes, sizes and styles as the technology is applied to other
purposes (Hoopes and Barnett 1995; Simms et al. 1997:783). For items that are
cached, we may not see much in the way of decoration or other modifications.
While potters may add decoration for their own artistic enjoyment, such effort
may not be worth the time if the goal is to transmit social information (such as
status or faction membership) when it would be out of view most of the year. It
is important to stress that these are expectations only and not blind rules to be
applied to the archaeological record. As with all archaeological interpretation,
the design, standardization and distribution of material artifacts should help
our reconstructions of mobility, but, if at all possible, they should represent
just one window on this aspect of prehistoric ways of life.
References
Aikens, C. M.
1995 First in the World: The Jomon Pottery of Early Japan. In W. K. Barnett
Andrefsky, W.
1991 Inferring Trends in Prehistoric Settlement Behavior from Lithic Production
Technology in the Southern Plains. North American Archaeologist 12: pp.
129–144.
Arnold, D. E.
1985 Ceramic Theory and Cultural Process. Cambridge, Cambridge University
Press.
Bamforth, D. B.
1991 Technological Organization and Hunter-Gatherer Land Use. American
Basgall, M. E.
1989 Obsidian Acquisition and Use in Prehistoric Central Eastern California: A
Preliminary Assessment. In R. E. Hughes (ed.), Current Directions in California
Obsidian Studies. Contributions of the University of California Archaeological
Research Facility 48. Berkeley, University of California Press: pp. 111–126.
Bayman, J. M., R. H. Hevly, B. Johnson, K. J. Reinhard and R. Ryan
1996 Analytical Perspectives on a Protohistoric Cache of Ceramic Jars from the
Lower Colorado Desert. Journal of California and Great Basin Anthropology 18:
pp. 131–154.
322     J e l m e r W. E e r k e n s
Bettinger, R. L.
1975 The Surface Archaeology of Owens Valley, Eastern California: Prehistoric Man-Land
Relationships in the Great Basin. University of California, Riverside (PhD
dissertation).
1986 Intersite Comparison of Great Basin Brown Ware Assemblages. In S.
Griset (ed.), Pottery of the Great Basin and Adjacent Areas. University of Utah
Anthropological Papers 111. Salt Lake City, University of Utah Press: pp.
97–106.
Bettinger, R. L., D. B. Madsen and R. G. Elston
1994 Prehistoric Settlement Categories and Settlement Systems in the Alashan
Desert of Inner Mongolia, PRC. Journal of Anthropological Archaeology 13: pp.
74–101.
Braun, D.
1983 Pots as Tools. In A. Keene and J. Moore (eds.), Archaeological Hammers and
Theories. New York, Academic Press: pp. 107–134.
Bright, J. R. and A. Ugan
1999 Ceramics and Mobility: Assessing the Role of Foraging Behavior and Its
Implications for Culture-History. Utah Archaeology 12: pp. 17–29.
Brown, J. A.
1989 The Beginnings of Pottery as an Economic Process. In S. E. van der Leeuw
and R. Torrence (eds.), What’s New? A Closer Look at the Process of Innovation.
London, Unwin Hyman: pp. 203–224.
Campbell, E. W.
1931 An Archeological Survey of the Twenty Nine Palms Region. Southwest Museum
Paper Number 7. Los Angeles, Southwest Museum.
Close, A. E.
1995 Few and Far Between: Early Ceramics in North Africa. In W. K. Barnett
Crown, P. L. and W. H. Wills
1995 The Origins of Southwestern Ceramic Containers: Women’s Time Allocation
and Economic Intensification. Journal of Anthropological Research 51: pp.
173–186.
Delacorte, M. G.
1990 The Prehistory of Deep Springs Valley, Eastern California: Adaptive Variation in
the Western Great Basin. University of California, Davis (PhD dissertation).
1999 The Changing Role of Riverine Environments in the Prehistory of the Central-
Western Great Basin: Data Recovery Excavations at Six Prehistoric Sites in Owens
Valley, California. Report submitted by Far Western Anthropological Research
Group to Caltrans District 9, Bishop, California.
Eerkens, J. W.
2001 The Origins of Pottery Among Late Prehistoric Hunter-Gatherers in California
and the Western Great Basin. University of California, Santa Barbara (PhD
dissertation).
2003a Residential and Pottery Use in the Western Great Basin. Current Anthropology
44: pp. 728–738.
2003b Towards a Chronology of Brownware Pottery in the Western Great Basin: A
Case Study from Owens Valley. North American Archaeologist 24: pp. 1–27.
2004 Privatization, Small-Seed Intensification, and the Origins of Pottery in the
Western Great Basin. American Antiquity 69: pp. 653–670.
2005 GC-MS Analysis and Fatty Acid Ratios of Archaeological Potsherds from the
Western Great Basin of North America. Archaeometry 47: pp. 83–102.
n.d. Notes on a Cached Brownware Pot near Little Lake, Eastern California.
Manuscript in possession of the author.
Eerkens, J. W., H. Neff and M. Glascock
1999 Early Pottery from Sunga’va and Implications for the Development of
Ceramic Technology in Owens Valley, California. Journal of California and
Great Basin Anthropology 20: pp. 275–285.
2002 Ceramic Production Among Small-Scale and Mobile Hunters and Gatherers:
A Case Study from the Southwestern Great Basin. Journal of Anthropological
Feathers, J. K. and D. Rhode
1998 Luminescence Dating of Protohistoric Pottery from the Great Basin.
Geoarchaeology 13: pp 287–308.
Gilreath, A. J. and W. R. Hildebrandt
1997 Prehistoric Use of the Coso Volcanic Field. Contributions of the University of
California Archaeological Research Facility Number 56. Berkeley, University of
California Press.
Henry, D. O.
1989 Correlations Between Reduction Strategies and Settlement Patterns. In D.
O. Henry and D. H. Odell (eds.), Alternative Approaches to Lithic Analysis.
Washington, American Anthropological Association: pp. 139–155.
Hildebrandt, W. R. and A. Ruby
1999 Archaeological Survey of the Coso Target Range: Evidence for Prehistoric and Early
Historic Use of the Pinyon Zone at Naval Air Weapons Station, China Lake. Report
on file with the Naval Air Weapons Station, China Lake, California.
Hoopes, J. W. and W. K. Barnett
1995 The Shape of Early Pottery Studies. In W. K. Barnett and J. W. Hoopes
(eds.), The Emergence of Pottery. Washington, Smithsonian Institution Press:
pp. 1–7.
324     J e l m e r W. E e r k e n s
Hunt, A.
1960 Archaeology of the Death Valley Salt Pan, California. Anthropological Papers
Number 47. Salt Lake City, University of Utah Press.
Ikawa-Smith, F.
1976 On Ceramic Technology in East Asia. Current Anthropology 17: pp. 513–515.
Kelly, R. L.
1988 The Three Sides of a Biface. American Antiquity 53: pp. 717–734.
1990 Marshes and Mobility in the Western Great Basin. In J. C. Janetski and D. B.
Madsen (eds.), Wetland Adaptations in the Great Basin. Occasional Papers Number
1. Provo, Museum of Peoples and Cultures: pp. 233–258.
1995 The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Washington,
King, T. J., Jr.
1976 A Cache of Vessels from Cottonwood Spring (Riv-937). Journal of California
Kroeber, A. L.
1922 Elements of Culture in Native California. University of California Publications
in American Archaeology and Ethnology 13: 8, pp. 257–328.
Lurie, R.
1989 Lithic Technology and Mobility Strategies: The Kosher Site Middle Archaic.
In R. Torrence (ed.), Time, Energy, and Stone Tools. Cambridge, Cambridge
Lyneis, M. M.
1988 Tizon Brown Ware and the Problems Raised by Paddle-and-Anvil Pottery in
the Mojave Desert. Journal of California and Great Basin Anthropology 10: pp.
146–155.
Odell, G.
1996 Economizing Behavior and the Concept of “Curation.” In G. Odell (ed.),
Theory and Behavior from Stone Tools. New York, Plenum: pp. 53–81.
Parry, W. and R. L. Kelly
1987 Expedient Core Technology and Sedentism. In J. K Johnson and C. A.
Morrow (eds.), The Organization of Core Technology. Boulder, Westview Press:
pp. 285–304.
Pippin, L. C.
1986 Intermountain Brown Wares: An Assessment. In S. Griset (ed.), Pottery of the
Great Basin and Adjacent Areas. Anthropological Papers Number 111. Salt Lake
City, University of Utah Press: pp. 9–21.
Plew, M. G. and M. Bennick
1990 Prehistoric Pottery of Southwestern Idaho: A Report on the Southwest Idaho
Ceramics Project. In J. M. Mack (ed.), Hunter-Gatherer Pottery from the Far
West. Nevada State Museum Anthropological Papers 23. Carson City, Nevada
State Museum: pp. 107–122.
Prince, E. R.
1986 Shoshonean Pottery of the Western Great Basin. In S. Griset (ed.), Pottery of
the Great Basin and Adjacent Areas. Salt Lake City, University of Utah Press:
pp. 3–8.
Rafferty, J.
1985 The Archaeological Record on Sedentariness: Recognition, Development,
and Implications. In M. B. Schiffer (ed.), Advances in Archaeological Method
and Theory. New York, Academic Press: pp. 113–156.
Reid, K. C.
1984 Fire and Ice: New Evidence for the Production and Preservation of Late
Archaic Fiber-Tempered Pottery in the Middle Latitude Lowlands. American
Rhode, D.
1994 Direct Dating of Brown Ware Ceramics Using Thermoluminescence and Its
Relation to the Numic Spread. In D. B. Madsen and D. Rhode (eds.), Across
the West: Human Population Movement and the Expansion of the Numa. Salt Lake
City, University of Utah Press: pp. 124–130.
Rice, P. M.
1999 On the Origins of Pottery. Journal of Archaeological Method and Theory 6: pp.
1–54.
Rosen, S. A.
1989 The Origins of Craft Specialization: Lithic Perspectives. In I. Hershkovitz
(ed.), People and Culture in Change: Proceedings of the Second Symposium of
Upper Palaeolithic, Mesopolithic, and Neolithic Populations of Europe and the
Mediterranean Basin. Oxford, British Archaeological Reports: pp. 107–114.
Sassaman, K. E.
1993 Early Pottery in the Southeast: Tradition and Innovation in Cooking Technology.
Tuscaloosa, University of Alabama Press.
Schaefer, J.
2003 Petrographic and INAA Analysis of Southern Owens Valley Clays and Brown
Ware Ceramics from the Olancha/Cartago Four-Lane Project. In B. F. Byrd
and M. Hale (eds.), Lacustrine Lifestyles Along Owens Lake: NRHP Evaluation
of 15 Prehistoric Sites for the Olancha/Cartago Four-Lane Project, U.S. Route 395,
Inyo County, California. Encinitas, ASM Affiliates: pp. 567–605.
Schiffer, M. B., J. M. Skibo, T. C. Boelke, M. A. Neupert and M. Aronson
1994 New Perspectives on Experimental Archaeology: Surface Treatments and
Thermal Response of the Clay Cooking Pot. American Antiquity 59: pp.
197–217.
326     J e l m e r W. E e r k e n s
Simms, S. R., J. R. Bright and A. Ugan

1997 Plain-Ware Ceramics and Residential Mobility: A Case Study from the Great
Basin. Journal of Archaeological Science 24: pp. 779–792.
Skibo, J. M., M. B. Schiffer and K. C. Reid
1989 Organic Tempered Pottery: An Experimental Study. American Antiquity 54:
pp. 122–146.
Steward, J. H.
1938 Basin-Plateau Aboriginal Sociopolitical Groups. Washington, Smithsonian
Institution Press.
Thacker, P. T.
1996 Hunter-Gatherer Lithic Economy and Settlement Systems: Understanding
Regional Assemblage Variability in the Upper Paleolithic of Portuguese
Estremadura. In G. Odel (ed.), Theory and Behavior from Stone Tools. New York,
Plenum: pp. 101–124.
Thomas, D. H.
1971 Prehistoric Subsistence-Settlement Patterns of the Reese River Valley, Central
Nevada. University of California, Davis (PhD dissertation).
1972 Western Shoshone Ecology: Settlement Patterns and Beyond. In D. Fowler
(ed.), Great Basin Cultural Ecology: A Symposium. Reno, University of Nevada:
pp. 135–153.
1983 The Archaeology of Monitor Valley 2: Gatecliff Shelter. Anthropological Papers
of the American Museum of Natural History 59: 1, pp. 1–52.
Touhy, D. R.
1990 Second Thoughts on Shoshoni Pots from Nevada and Elsewhere. In J. Mack
(ed.), Hunter-Gatherer Pottery from the Far West. Carson City, Nevada State
Wallace, W. J.
1965 A Cache of Unfired Clay Objects from Death Valley, California. American
1986 The Pottery of Mesquite Flat, Death Valley, California. In S. Griset (ed.),
Pottery of the Great Basin and Adjacent Areas. Salt Lake City, University of Utah
Press: pp. 71–74.
Weaver, R. A.
1986 Notes on the Production, Use, and Distribution of Pottery in East-Central
California. In S. Griset (ed.), Pottery of the Great Basin and Adjacent Areas. Salt
Lake City, University of Utah Press: pp. 75–81.
P a rt II
The Present
and the Future
Ch a p t e r 1 5
Mobility and Sedentism

of the I ron Age
Agropastoralists of
Southeast K azakhstan
Claudia Chang1
E vidence from the Iron Age settlements in the Talgar region of southeast
Kazakhstan indicates that the ‘fierce horse riding warrior’ confederacies
such as the Saka (Scythian), Wusun and Yuezhi practiced both farming and
herding along the well-watered steppes north of the Tian Shan Mountains
(Chang et al. 2003). The splendid kurgan tradition dating from the 8th century
BCE through the 5th century CE of the reputed Iron Age nomads also suggests
that such pastoral confederacies were hierarchical and stratified (Frachetti,
this volume; Shishlina, this volume). To what degree were the members of
these confederacies mobile or settled? And how did mobility, or the ability to
move with flocks, personnel, military cavalries or caravans, contribute to the
social evolution of these chiefdom or incipient state-level steppe societies? I
draw on our previously published interpretations of the excavations at three
Iron Age settlements in the Talgar region, and the surface surveys conducted
in the highland and lowland zones of the mountain-steppe region along the
1.
The following granting agencies supported the archaeological investigations in
the Talgar area: the National Science Foundation, Grant Number BCS-9603661, for
fieldwork from 1997 to 2002; the Wenner-Gren Foundation, for fieldwork from 1994
to 1996; and the National Geographic Society, for fieldwork from 1994 to 1996 and in
2003 and 2004. I wish to express my gratitude to the archaeologists of Kazakhstan who
contributed to this work: Dr. Karl M. Baipakov, Director of the Institute of Archaeology
(Almaty, Republic of Kazakhstan); Fedor P. Grigoriev; Murat M. Nurpeisov; Boris A.
Zheleznyakov and Yuri M. Peshkov. Perry A. Tourtellotte helped in all stages of the
fieldwork. Over the years I have found Philip Salzman’s work an inspiration and his
spare prose a model for ethnographic writing and analysis. I share his enthusiasm for
Fredrik Barth’s (1961) Nomads of South Persia, the first and most influential ethnography
I know on pastoral nomads.
329
330    C l a u d i a C h a n g
Tian Shan Mountains in southeast Kazakhstan to put forth a set of questions

pertaining to the nature of mobility among these Iron Age agropastoralists
(Chang et al. 2002). Did the Iron Age inhabitants of the Talgar region vacillate
between ‘settled’ and ‘mobile’ lifestyles, and how would mobility or sedentism
be apparent in the archaeological record? A set of ethnographic premises,
drawn from contemporary pastoral societies in the Middle East, provides a
basis for comparison between contemporary and ancient pastoral nomadism.
The interface between these two sets of variables is explored: sedentarization
versus mobility and hierarchy versus equality.
The Iron Age farmers and herders of the Talgar region appear to have been
integrated into a larger ramifying chiefdom or even an incipient state-level
society. What role did military mobilization, caravan trade, and predatory
raiding play in determining their degree of mobility? Today in southeast
Kazakhstan, pastoral nomads practice short-distance pastoral transhumance.
Many herders move their flocks of sheep and goats and their herds of cattle
and horses between the highland alpine meadows for summer pasture and
the lowland steppe areas for fall through spring pasture. Ancient Iron Age
pastoralists could have followed similar patterns of transhumance in which
at least a portion of the Iron Age population in the Talgar region practiced
seasonal mobility. During the Iron Age the Talgar region, just to the north
of the foothills of the Tian Shan Mountains, was a node along an east-west
corridor used for trade, predatory raiding and military activities. The ancient
people occupying this corridor during the first millennium BCE to the first
half of the first millennium CE have been labeled the Saka, the Wusun and
the Yuezhi nomadic confederacies (Pulleyblank 1970, 1995; Moshkova 1992;
Di Cosmo 1994).
In the past the study of nomadic cultures such as the Saka (Scythians) has
resulted in the romanticized notions of Iron Age nomads as the perpetrators of
social and cultural change across the Eurasian steppe. It has been the assumption
that the ‘horse riding nomads’ governed the steppe regions through their use
of mounted archery and their mobility (Pulleyblank 1995). Such romanticized
notions of the Eurasian Iron Age nomadic cultures should be evaluated by two
empirical lines of investigation: a more informed theoretical view of the varia-
tion extant in contemporary pastoral societies, thus the use of better pastoral
models, and the application of more rigorous archaeological methods to test
such pastoral models. The organization of this chapter follows this program-
matic order of investigation. First, the ethnographic models are presented.
Second, the current state of archaeological data from Talgar is assessed in light
of these models. Finally, future research directions are outlined.
Agropastoralists of S o u t h e a s t K a z a k h s ta n      331
The Study Area

Since 1994 the Kazakh-American Talgar Archaeological Project has conducted
surface surveys and excavations in the Talgar region of southeast Kazakhstan.
We have conducted surface surveys in three areas: the Talgar alluvial fan, the
highland plateau of Orman, and the valleys of Turgen and Asi (Figure 15.1).
The turbulent Talgar River originates in the Tian Shan Mountains and flows
northward emptying into the Ili River. This region is called the Semirechy’e
or Seven Rivers Area, which extends from the east of the Chinese-Kazakhstan
border to the Chu River and is characterized by seven rivers that flow north-
ward from the mountains and empty into the Ili River basin (more than 600 km
in length). The Talgar alluvial fan is a cone-shaped landform, and its apex is at
the base of the foothills where the Talgar River cuts into the valley floor. The
fan is situated just north of the highest peak of the Zailiisky Alatau Mountains
(in the northern Tian Shan range), Peak Talgar. The steep gradient from the
snowcapped Peak Talgar, 5000 m above the semi-arid steppe regions of the
Ili River basin, starts at 80–100 km distance. Both the historic and prehistoric
herdsmen, who maintained flocks of sheep and goats, as well as herds of cattle,
horses and even some camels, probably practiced vertical transhumance moving
between highland summer pastures, at 1800–2400 m, and winter pastures at
550–1100 m, just north of the mountain range. Less than one percent of this
alluvial fan has been surveyed archaeologically. This relatively small area has
yielded a total of more than 500 finds, ranging from settlements and burial
mounds to isolated artifacts. Two-thirds of these finds have been dated to
the Iron Age. Three Iron Age settlements have been excavated (Chang et al.
2003). Archaeozoological and archaeobotanical analysis (both flotation samples
and phytoliths) indicate that the prehistoric inhabitants of these settlements
practiced cereal farming (millet, wheat, barley and possibly rice) and animal
herding (cattle, sheep, goats, horses and possibly camels).
Models of Nomadic Tribes and Confeder acies

The ethnographic model posited by Salzman (2004) outlines the processes
by which contemporary and historic tribes and nomadic confederacies in the
Middle East have been encapsulated into regional polities and the nation-state.
I have used this model to re-examine the archaeological data on Iron Age
settlement and burials from the Talgar region. According to Salzman’s (2004)
model of encapsulation, any pastoral tribe or confederacy that is incorporated
into regional or national markets or has been subjected to the rule of the State
will be less capable of maintaining a semblance of egalitarianism and individual
autonomy among all the members of the group. The examples used to test these
Figure 15.1. Map showing the location of the study area in southeast Kazakhstan.
assumptions include the Yomut Turkmen, the Basseri and the Sarhadi Baluch,
as well as other Middle Eastern nomadic tribes that maintain flocks of sheep
and goats, and in some cases keep camels as well. The Yomut Turkmen, the
Basseri and the Sarhadi Baluch are tribal confederacies that are headed by khans
(chiefs). Yet these particular groups of nomadic pastoralists also pose a striking
contrast to hierarchically ranked tribes or confederacies: each camp group is
defined as a group of households that share common pasture resources, govern

themselves through consensus rather than coercion, and express egalitarian
values, at least in political if not economic arenas. Salzman (2004) then asks
why some pastoral tribes remain nomadic while others become settled and
why some pastoral societies remain independent and egalitarian while others
become hierarchically ordered?
Salzman (2004:34) argues cogently that mobility and sedentism are but
opposite ends of a continuum of strategies used by nomadic peoples in response
to changing environmental, social or political circumstances. Usually, pastoral
mobility is defined as an adaptive strategy allowing herdsmen to maximize the
use of dispersed environmental resources such as natural pasture, forage or
water. Pastoral peoples traveling with their herds of cattle, sheep and goats,
horses or camels across diverse landscapes engage in pastoral productive activi-
ties (raising animals for meat, milk, wool or hair), either for subsistence or
market consumption. Why do pastoralists ‘settle’ or restrict their mobility to
a fixed location? The usual factors are that they leave herding completely and
become farmers; engage in multiresource strategies, including activities that
require residence in a fixed location, like agriculture; become wage laborers
attached to settled communities; or that they are forced to settle in permanent
places as a result of coercive state or government policies. In pre-industrial
societies nomadic pastoralism or the periodic movement of humans and their
herds, may also have a social and political dimension. Herdsmen may also move
frequently in order to escape oppressive regimes, as a military strategy, or to
expand social and economic networks.
For Iranian groups, such as the Baluch and the Yomut Turkmen in the 1960s,
mobility or sedentism can be seen as adaptive responses to the outside forces
of the state (Salzman 2004). The Yomut Turkmen used mobility as a political
strategy to avoid Persian control over their wealth and independence (Irons
1974). The Sarhadi Baluch, on the other hand, reduced their mobility in the
1960s and 1970s as they obtained jobs in settlements and began to engage in
agriculture (Salzman 2004:35). In both cases the choice to remain mobile or
to become settled was an economic and political response to the increasing
encapsulation of these nomadic tribes by the larger nation-state.
The image of the fiercely independent and egalitarian nomad is an ethno-
graphically pleasing stereotype. Pastoral nomads, continually confronted
with decisions about how to best maintain their individually owned herds on
communal pasturelands, are characterized as free agents expressing the ultimate
democratic values of self-reliance, independence and autonomy. The opposing
ideal, that of the nomadic herdsman or tribal member oppressed by a chief
(khan) or autocratic leader, suggests that the confederacy or tribal organization
is subject to the outside influences of the state or empire (Khazanov 1994).
Yet to what degree does the chief or khan derive his power from the larger
sociopolitical hierarchies of the state or centralized bureaucracy? Salzman
(2004) challenges Barth’s (1961) description of the khan as an autocratic and
oppressive ruler who controls the otherwise autonomous and self-sufficient
Basseri nomads. He argues that the khan, as the autocratic leader of the
Khamseh Confederacy of the Basseri, maintains his coercive and autocratic
authority over the Basseri nomads because the herdsmen collude with the
leader in order to project the chief’s image as an autocratic leader (Salzman
2004:90–91). This act of ‘collective impression management’ is done to boost
the confederacy’s position in the multi-ethnic arenas of urban Persian society.
Salzman (2004:91) argues that the herdsmen ultimately do maintain their
autonomy and self-determination since they exercise ‘choice’ in selecting their
leaders. In fact, the camp group may choose to switch its allegiance to those
chiefs who will best ‘serve their interests and protect them.’ This ‘collective’
image of the khan as a despot, capable of oppressive and autocratic rule, is a
useful fiction promoted by the tribesmen, as a political strategy for maintaining
the Khamseh Confederacy’s overall clout in the larger multi-ethnic political
arena of the Persian state. As Salzman (2004:85) points out, tribesmen could
resist coercive and autocratic power but choose not to do so. The khan, with
his splendid residency in Shiraz and his extensive networks of political ties to
other elites, must serve his subjects by attending to their needs.
Mobility and Sedentism in the

Archaeological Record
In the Talgar region mixed farming and herding was practiced on the alluvial
fan and coupled with short-distance vertical transhumance from the 8th century
BCE through the 1st century CE (Chang et al. 2003). Surface surveys of the
Talgar alluvial fan demonstrate that the majority of Iron Age places visible on
the landscape are burial mounds, ‘kurgans’ (183 or more), and settlement sites
(at least 59). The burial mounds are prominent features of the steppe landscape
(Figure 15.2). The high number of kurgans and the low number of settle-
ments is due to geomorphologic processes by which the Iron Age settlements
are buried under 0.5–1.0 m of deposits. Those settlement sites visible on the
surface have been exposed by agricultural plowing or are in erosion cuts or
streambed profiles. It is noteworthy that even during ancient times the kurgans
must have been dominant features in the landscape while, in comparison,
village hamlets were modest in appearance.
Our archaeological excavations at the three Iron Age settlements of Tuzusai,
Tseganka 8 and Taldy Bulak 2, from 1994 to 2004, provide another window into
the problem of Iron Age social organization (Figure 15.3). At these settlements
Figure 15.2. A Kazakh herdsman and his flock in front of an Iron Age kurgan (burial mound).
the archaeozoological and archaeobotanical remains indicate the presence of

agriculture based on millet, wheat and barley and the herding of cattle, sheep
and goats, horses and maybe camels. The architecture at Tseganka 8 shows a
range of domestic structures, such as at least seven semisubterranean round and
rectangular pithouses. At Tuzusai at least 30 circular and rectangular storage
pits, a possible pithouse structure and plastered floor fragments associated with
stone-lined postholes were identified. At Taldy Bulak 2 several rectangular
stone alignments, probably the foundations of houses, postmolds associated
with ephemeral dwellings, and overlapping activity areas were found.2
Additional features found at all three sites include storage pits, fire pits,
hearths, and outdoor activity areas. All three sites have artifact inventories that
include grinding stones, small iron or bronze fragments, and red and buff ware
utility pottery (bowls, jars, cooking and storage vessels). At Tuzusai there were
at least four different occupation periods, including a post–Iron Age occupa-
tion, and at Tseganka 8 and Taldy Bulak there were eight to ten occupational
levels. These sequences suggest that the village hamlets were semisedentary or
year-round sedentary occupations, where a core group of household members
stayed to plant, care for, and harvest cereal crops, while others moved with the
flocks and herds of cattle, sheep, goats and horses between highland summer
pastures and lowland winter grazing. There is insufficient evidence to prove
the existence of seasonal occupation, yet the distinct occupation levels at the
settlements do seem to indicate intermittent periods of activity at the sites. The
2.
Postmolds are soil stains remaining after the decomposition of a post, in this case
indicative of the remains of temporary dwellings (cf. Holman and Lovis, this volume).
Figure 15.3. Close-up of a topographic map of the Talgar alluvial fan, showing the northern section of
the fan where the excavated sites of Tuzusai, Tseganka 8, and Taldy Bulak 2 are located.
artifact inventories, and the relatively unsophisticated nature of the architecture

at these settlements, suggest that the inhabitants were common folk, engaged in
farming and herding activities. We have not been able to locate the cemeteries
or graves of these common folk.
How mobile or sedentary were the Iron Age inhabitants who occupied the
small village hamlets of Talgar? An intriguing comparison with the Yomut
Turkmen might provide some insight here. The Yomut Turkmen included
both nomadic agricultural and nomadic pastoral households that could shift
between these occupations (Salzman 2004:33). The ability to shift from a
greater emphasis on farming to herding was always a viable strategy for the
Iron Age inhabitants of Talgar. At the site of Tuzusai the cultivation of rice
increased between 415 BCE and 75 CE, while at Tseganka 8 more emphasis
was placed on the cultivation of millet from 775–40 BCE. At Taldy Bulak 2 a
paucity of cereal remains in the phytolith samples (Rosen 2003) indicates that
animal husbandry might have been favored over cereal cultivation. Households
and village communities underwent shifts in their overall subsistence strategies,
some favoring cultivation over herding, others herding over farming. This
fluctuation between the dependence on crop cultivation or livestock herding

at various hamlets in the Talgar region may also signal the shift in and out of
nomadic and sedentary lifestyles.
Currently, it is not possible to determine whether these three sites were
year-round or seasonal occupations or what kind of mobility strategies were
present. Yet some intriguing possibilities exist. The Iron Age village hamlets
were used for a wide range of subsistence activities, such as herding, dry
farming, irrigation farming and foraging. The local Talgar population may have
rotated in and out of these productive strategies based on individual household
decisions. Residential mobility, within a single household, might have consisted
of several members engaged in vertical transhumant herding while other
members engaged in settled subsistence farming. Families or households may
have rotated or shifted their places of residence, depending on their reliance
on various productive activities at a given point in time. This may have led to
preferential migration to Tuzusai, where irrigation agriculture was practiced,
or to Tseganka 8, where millet production was favored, or to Taldy Bulak 2,
where metal finishing took place.
Equality and Hier archy in the

Archaeological Record
The historical data based on texts and inscriptions suggest that the distant Iron
Age nomadic confederacies, such as the Saka of the Ili River (Semirechy’e),
inhabited territories far from the irrigated desert oases of Chorasmia, Sogdiana,
Bactria and Margiana (Central Asian polities) but could have been encapsu-
lated or controlled by such polities throughout the first millennium BCE
(Dandamaev 1994). The proposed social structure for the Saka was that of
clan and tribal organization. Akishev and Kushaev (1963) divided the 25 burial
mounds of Bes Shatyr, a famous megalithic kurgan sanctuary of the Ili River
Basin, into three categories: the clan leaders, the aristocratic elite and glorified
warriors, and the enlisted warriors. These categories were determined on the
basis of burial-mound size and limited information on the burial inventories
from each category. Even if we dismiss the criteria on which this three-tiered
social stratification was constructed, there still remains sufficient evidence
that points to the vast status differentiation between the chiefly leaders and
the commoners. For example, the Golden Warrior tomb of Issyk, only 20
kilometers east of the Talgar alluvial fan, was discovered in the side chamber
of a tsar kurgan dating to the 3rd or 2nd century BCE. It is the tomb of a youth,
robed in a cloth cloak decorated with golden plaques, a magnificent headdress
with bird and ram figurines, an inscribed silver bowl, ceramic vessels and other
items. At settlement sites in the Talgar fan, with occupation levels from the
same time period, there is a lack of any luxury items. When metal objects are
found, they usually are utilitarian, such as iron knife fragments, trident arrow
points, bronze cauldron fragments, hinges, etc.
Our surveys and excavations on the Talgar alluvial fan have discovered
these types of archaeological places: semisedentary village hamlets with simple
pottery, metal and farming technology interspersed among the burial kurgans;
and a burial cult, as apparent from the large quantity of burial kurgans of
different size ranges (heights and diameters) and of different time periods,
spanning the first millennium BCE. Is this sufficient evidence to prove the
existence of a ramifying clan structure of a chiefdom or incipient state-level
society? The artifact inventories at the excavated small hamlets at Tuzusai,
Tseganka 8 and Taldy Bulak 2 indicate the vast social differences between the
ordinary common folk and the aristocratic elite buried in the burial kurgans.
One can infer that, during the Iron Age, there must have been a ranked
hierarchy of a chiefly elite group with high-status burial goods buried in the
kurgans and the common folk buried in simple graves or cemeteries yet to be
located by our surface survey.
In early agricultural states the bureaucracy and elite at the top stratum of
society were able to extract valuable economic, political and social resources
from the lower classes in the form of tribute or taxation. What kind of resources
did the Eurasian nomadic clan chiefs or khans extract from the bottom stratum
of society? Furthermore, how did the clan leaders convert simple subsistence
resources, animals or grain, into precious luxury items such as gold and silver
that in turn became status markers in ‘death or burial’ cults? Such luxury goods,
by their very disposal in burial kurgans, were then effectively removed from
everyday circulation. And more important, how did the chiefs or the khans orga-
nize the tribesmen, who were the backbone of the internal economic structure
of society, into a viable military force? The ‘nomadic elite’ of the hinterland may
have garnered specific localized control through important economic, political
and ideological links to the powerful desert-oases kingdoms. Perhaps the king-
doms relied on the far-flung nomadic confederacies as military retainers who
controlled distant territories and valuable trade and caravan routes. Therefore,
the precious luxury items of gold, silver and bronze found in local burial kurgans
were part of a supralocal circulation of elite status items used to bolster the
position of the local leaders. Such objects might represent the local leaders’
abilities to obtain luxury items through a set of sociopolitical and economic
linkages to the more powerful states and kingdoms. This may also explain why
the Saka (eastern Scythian) zoomorphic imagery found on precious metals and
textiles throughout the first millennium BCE in Eurasia was part of a stylized
iconography shared among many disparate local and regional nomadic groups
rather than one that represented a single, shared cultural tradition.
The burial mounds, especially those with splendid burial inventories, may
have been constructed by the ancient inhabitants as part of a larger ‘image-
building’ strategy to make the leaders of the common herdsmen and farmers
appear to be powerful, wealthy and militarily capable, especially in the face of
challenges, military and otherwise, from outside nomadic groups. The burial
mounds, the largest measure 80–100 m in diameter and are 17 m high, repre-
sent the most visible markers of Iron Age occupation of the Talgar alluvial fan.
These mounds often form linear patterns along dry streambeds or along the
banks of the Talgar River. The human effort invested in building each burial
mound, even the smaller ones, is considerable. Many burial mounds have
been robbed either in antiquity or in recent history. Certainly these mounds
represent a higher investment of labor than the simple domestic architecture of
pithouses and simple rectangular room blocks found at the village hamlets.
The archaeological evidence from the three excavated Iron Age settlements
in the Talgar region suggests that there was little to no status differentiation
among the common folk. The artifact inventories rarely include any luxury
or status items, aside from bronze fragments or a jadeite bead or fragments of
imported ceramic wares. Within less than a kilometer, however, most of the
settlement sites appear to be situated in close proximity to the lines of burial
kurgans from contemporaneous chronological periods. Relative egalitarianism
existed among the common folk of herders and farmers who inhabited the
small village hamlets, while ancestor worship in the form of a mortuary cult
emphasized inequality and the presence of hierarchy.
Mobility and Hier archy in the Talgar Region

This following scenario is put forth as a possible interpretation of the Iron
Age archaeology of Talgar. How did mobility as a strategy and the nature of
hierarchy within the agropastoral population both function on a larger regional
level? Did the ‘nomadic mobility strategies’ and apparent hierarchy within the
confederacy contribute directly to processes of state formation? Iron Age social
organization might have been egalitarian at the local level and hierarchical at
the regional level. The common farmers and herders functioned day to day
through a system of equality and relative autonomy. A herdsman might become
a farmer, or vice versa, and an entire household or members within the house-
hold might engage in seasonal transhumance or practice year-round sedentism.
Individuals, households or residential groups operated in a fluid manner alter-
nating between the opposing poles of mobility and sedentism based on rational
economic decision making. The elite, emerging out of ranked kin-based
units, maintained and cultivated the image of ‘nomadism.’ Their mobility or
nomadism was necessary for the coalescence of larger tribal units, segments or
confederacies and for activities such as predatory raiding or military battles that
took place on a regional level. Local elites maintained their ability to coerce
the common folk into providing labor, tribute or personnel for these activities
through ‘collective impression management.’ An aristocratic leader maintained
his charismatic position through the manipulation of obvious status markers,
including his ability to participate in an ancestral death cult, which was marked
by the linear arrangement of burial mounds along the stream channels of the
most productive agricultural and natural pasture areas. The ancestral cult of
burying the dead with luxury items way beyond the reach of most commoners
reinforced the social and ideological distance between the elite rulers and the
rest of society. But was this ideological distance real or fictive? And was some
of the rulers’ wealth also a display of the leader’s allegiance to larger central
polities? How were the two worlds, that of the commoner and that of the
ranked elite, fused together to form a cohesive whole?
Perhaps the leaders’ positions were augmented through another set of
mobility strategies unrelated to nomadic pastoralism but instead tied to the
group’s ability to conduct predatory raiding or to act as a cohesive military
force. The hierarchical ranking as apparent from the burial cult preserved
the image-building strategy of local elites. They, in turn, relied on the larger
nomadic states or empires for their supralocal and regional power. The
commoners contributed labor to building this ancestral cult of burial mounds
because it served their collective interests to do so. In this way they were guar-
anteed both access to productive lands and protection from outside marauders.
Mighty clan leaders and the aristocratic elite were supported because they could
defend and protect the very productive cultivated land and the natural pastures
from outside encroachment. Like the Basseri khan, the nomadic elite also best
represented the concerns of the tribe or confederacy within larger multi-ethnic,
sociopolitical arenas of kingdoms, states and empires. No doubt the herders
of Talgar did engage in both long and short distance transhumance, but it
was the hierarchical authoritarian organization of the clan or tribe that really
defined the nomadic nature of steppe social organization. This hierarchical
organization was most necessary at a regional and supraregional level in order
to rapidly deploy traders, raiders or warriors. The burial cult was indeed a form
of ancestor worship, inscribed on the alluvial fan to commemorate the ances-
tors of the aristocratic elite and to remind the commoners of their duties and
responsibilities toward the larger, regionally organized, tribe or confederacy.
Discussion
This chapter poses far more questions than it answers. Our archaeological
research on the Talgar alluvial fan is a first step toward outlining the nature
of economic strategies along the northern edge of the Tian Shan Mountains.
Archaeological research that links the artifact inventories from the burial
mounds to the artifact inventories of the settlement sites needs to be conducted.
This will help establish the temporal relationship between the burial cult and
the settlements and may also indicate whether or not the burial inventories
represent hierarchical ranking and the settlement inventories represent a more
egalitarian society. More settlements need to be excavated in the Talgar region
to determine the range of variability present in artifact inventories, as well as
in faunal and plant remains. As the parameters for the range and variation
within settlement sites are more firmly established, such as the occupational
histories of such sites and the variation in domestic architecture and artifact
inventories, it may be possible to develop a clear method for evaluating whether
a given site was used by mobile or settled groups. Surface surveys have been
conducted on the Talgar alluvial fan, a nearby highland plateau and a valley at
2400 m above sea level. The survey data from Talgar should be incorporated
into large-scale regional studies that will provide a perspective on the Iron Age
social organization from several discrete regions of Semirechy’e. Furthermore,
the spatial patterning of sites across a large region of diverse environmental
zones is necessary for a complete understanding of the range of subsistence
strategies practiced during the Iron Age. Finally, we need to redefine Iron
Age mobility on the Eurasian steppe, to include strategies for mobility that go
beyond subsistence and trading activities such as those tied to militarism and
predatory raiding. The interface between the range of mobility strategies and
the nature of social hierarchy during this period necessitates an archaeology
more closely informed by the fine-grained ethnographic analysis of nomadic
pastoral cultures from the contemporary world.
References
Akishev, K. A. and G. A. Kushaev
1963 Ancient Saka and Wusun Culture of the Ili River Valley [in Russian]. Alma-Ata,
Nauka.
Barth, F.
1961 Nomads of South Persia. Oslo, Oslo University Press.
Chang, C., N. Benecke, F. P. Grigoriev, A. M. Rosen and P. A. Tourtellotte
2003 Iron Age Society and Chronology in South-East Kazakhstan. Antiquity 73:
296, pp. 298–312.
Chang, C., P. A. Tourtellotte, K. M. Baipakov and F. P. Grigoriev
2002 The Evolution of Steppe Communities from the Bronze Age Through Medieval
Periods in Southeastern Kazakhstan (Zhetysu). Almaty, Sweet Briar College
and the A. K. Margulan Institute of Archaeology, Ministry of Education and

Science of the Republic of Kazakhstan.
Dandamaev, M. A.
1994 Media and Achaemenid Iran. In J. Harmatta, B. N. Puri and G. F. Etemadi
(eds.), History of Civilizations of Central Asia. Volume 2: The Development of
Sedentary and Nomadic Civilizations, 700 B.C. to A.D. 250. Paris, UNESCO
Press: pp. 35–65.
Di Cosmo, N.
1994 Ancient Inner Asian Nomads: Their Economic Basis and Its Significance in
Chinese History. Journal of Asian Studies 53: 4, pp. 1092–1126.
Irons, W.
Khazanov, A. M.
edition).
Moshkova, M. G.
1992 Steppe Region of the Asiatic Part of the USSR in the Scythe-Sarmatski Time [in
Russian]. Moscow, Nauka Izdatel’stvo.
Pulleyblank, E. G.
1970 The Wu-Sun and the Sakas and the Yueh-Chih Migration. Bulletin of the School
of Oriental and African Studies 33 (Special Issue in Honour of Sir Harold Bailey):
pp. 154–160.
1995 Why Tocharians? Journal of Indo-European Studies 23: 3–4, pp. 415–430.
Rosen, A. M.
2003 Preliminary Report on Phytolith Samples from Taldy Bulak 2. Unpublished report
on file at Sweet Briar College, Department of Anthropology and Sociology
(Sweet Briar, Virginia).
Salzman, P. C.
2004 Pastoralists: Equality, Hierarchy, and the State. Boulder, Westview Press.
Ch a p t e r 1 6
Crossing Boundaries
Nomadic Groups and
Ethnic Identities
S t u a r t T. S m i t h
E thnicity has proven elusive in the archaeological record, causing some

archaeologists to despair of ever identifying ethnic groups, especially
when dealing with nomads, who often leave only ephemeral traces in the
archaeological record. Symbiotic relations with other groups often characterize
nomadic peoples, especially pastoralists, producing an assemblage of mixed
materials that defies characterization as a distinctive tradition, bounded in
space and time. Ethnicity is usually defined this way, with the expectation of
a distinct, bounded people with a shared culture and primordial attachments.
Ethnic identities are identified through real or perceived commonalities of
culture, history and language, but others can also ascribe ethnicity, particu-
larly in contexts like the ones discussed in this chapter. Archaeologists have
generally assumed that ethnicity should appear as a distinctive material assem-
blage, reflecting ethnicity’s primordial attachments. This attitude reflects a
fundamental misunderstanding of ethnic dynamics. Ethnicity is a powerful
phenomenon, affecting us today in fundamental ways that color our perceptions
and expectations by creating a focus on ethnic groups. There is, however, an
emerging consensus among anthropologists and sociologists that we should
not expect to find absolute and bounded ethnic groups either archaeologically
or ethnographically. Instead, ethnic identities are situational and overlapping,
constructed and negotiated by individuals in specific social contexts. As a result,
archaeologists must re-orient their perspective away from the search for neatly
bounded groups in favor of an agent-centered approach that focuses instead
on ethnic dynamics.
This chapter begins with an exploration of the dynamics of ethnicity, looking
at different theoretical models and archaeological examples. The following
section examines some different ways in which archaeological evidence can be
used to examine ethnic dynamics, using two colonial communities on ancient
Egypt’s southern frontier, Askut and Tombos, in Nubia, as examples. It will
finish with some thoughts on the problems and potentials of studying the
343
344    S t u a r t T . S m i t h
ethnicity of nomadic groups with a brief consideration of the potential for

documenting ethnicity at Berenike (a Greco-Roman harbor on the Egyptian
Red Sea coast).
Ethnicity, Boundaries and the Nation-State

Ethnicity is often regarded as a recent phenomenon, linked closely to the
dynamics of European colonialism and the nationalist movements of the 19th
century CE (Kohn 1944; Handler 1988). Some argue that before this time, an
elite made up of a handful of administrators and aristocratic landlords ruled
over an agrarian society made up of a larger and more heterogeneous popu-
lation of peasants. There was no notion that these disparate groups shared
a common identity. Elites in many cases shared more in common across
state boundaries than with the people they ruled (Banks 1996:123–131). In
contrast, Smith (1986) sees evidence for ethnic solidarity, if not nationalism,
stretching back into antiquity as ethnie, or a combination of a collective name,
common myth of descent, shared history, culture, specific territory and sense
of solidarity. These ethnie provided the basis and necessary prerequisite for
the modern emergence of nationalism. This view reflects the essentialist
notion of ethnicity that does not take into consideration the possibility that
these features may be constructed post hoc to match the needs of nationalist
movements. Nevertheless, Smith’s ethnie are a good characterization of the
primordial way in which ethnic identities are almost always framed. Smith
asserts that ethnie can be traced into the distant past, specifically the ancient
Near East, but since the ethnie reside ultimately in people’s minds, Banks
(1996:129–130) is skeptical that these features can really be identified and
traced so far back in time.
The rich historical, artistic and material records of ancient Egypt and
the Classical world, however, do provide evidence for the construction of
ethnicity. For example, the ancient historian Herodotus defined the Greeks
as “the kinship of all Greeks in blood and speech, and the shrines of the gods
and the sacrifices that we have in common, and the likeness of our way of life”
(Rawlinson 1964: VI, 44). Tonkin et al. (1989) and Hall (1997) note that the
ancient Greeks used the term genos to refer to themselves, employing ethnos
primarily for the ‘barbaric’ other, including animals. Renfrew (1996) points
out that Herodotus’s focus on genetic, linguistic and cultural foundations for
group identity corresponds to modern definitions of ethnicity that emphasize
a common territory, descent (or a myth of origins), language, culture and
beliefs (especially religion). In this essentialist construction ethnic identity
is monolithic and bounded, immutable and self-defined. Tonkin et al. (1989)
and Hall (1997) also argue that the ‘barbaric’ ethnos played a key role in its
N o ma d i c G r o u p s and E t h n i c I d e n t i t i e s      345
opposition to the civilized Greek genos. Such ‘self/other’ oppositions are a

fundamental component of ethnic dynamics, so we can see the ‘genos-ethnos’
terminology as an early expression of ethnicity in spite of some inconsistencies
in the etymology of the word ethnic.
Ancient Egyptian texts and representations reflect a strong sense of ethnicity
(Smith 2003). For example, Pharaoh Akhenaton (1353–1336 BCE) provides
a surprisingly modern definition of ethnic groups in the great Hymn to the
Aton (Lichtheim 1976:131–132):1
You set every man in his place . . .
Their tongues differ in speech,
Their characters likewise;
Their skins are distinct,
For you distinguished the peoples.
Such state ideologies demonstrate a strong sense of national identity that is

at least theoretically applied to all Egyptians, regardless of social position,
and is placed in opposition to ethnic others. The state ideology explicitly
linked ethnic groups with territory (Loprieno 1988; Liverani 1990; Smith
2003), creating a sense of Egyptian-ness akin to Anderson’s (1983) ‘imagined
communities’ that provide a foundation for the modern nation-state. In spite
of extensive conquests and the complete cultural and economic incorpora-
tion of Lower Nubia into the Egyptian state during the New Kingdom, the
traditional national border of Egypt remained at Aswan in the south and the
Mediterranean in the north, coinciding precisely with the Egyptian ethnie, the
state ideology’s imagined community (Smith 2003).
This evidence contradicts the notion that ethnicity, national borders and
national identities are a product of 18th century CE absolutism and the emer-
gence of the modern state (Ratzel 1897; Prescott 1987:1; Wendl and Rösler
1999:7; Parker 2002); in fact, it suggests that all three elements already played
a key role in one of the earliest primary states. With this background in mind
I will first focus on the identification of ethnicity in the archaeological record,
including an examination of ethnic dynamics at two sites (Askut, an ancient
Egyptian fortress occupied around 1850–1050 BCE; and Tombos, a cemetery
used around 1400–1050 BCE), and wrap up with a consideration of ethnicity
among nomadic groups, like the Blemmyes at Berenike.
The dates given of the ancient Egyptian dynasties are those suggested by Baines
1.
and Malek (2000:36–37).

346    S t u a r t T . S m i t h
Ethnicity and Archaeology

The ancient view of ethnic groups as distinctive cultures tied to specific terri-
tories matches popular perceptions of ethnicity, and indeed ethnic identities
are usually constructed in similar terms. Not surprisingly, archaeologists have
continued to apply this framework, tending to equate archaeological cultures
and ethnic groups in spite of skepticism within both processual and postpro-
cessual approaches (Díaz-Andreau 1996; Hides 1996). This view relies on the
essentialist assumption that ethnic groups are bounded and uniform with a
set of shared beliefs handed down in a continuous tradition (Jones 1996). Yet,
on closer inspection, the seemingly immutable characteristics of ethnicity are
surprisingly mutable and socially contingent (Glazer and Moynihan 1963; Royce
1982). The expectation that material culture should fall into consistent, neatly
bounded assemblages corresponding to ethnic groups, or indeed international
borders, should therefore be abandoned in favor of a more complex model.
Recent studies of ethnicity reject essentialist approaches, instead empha-
sizing the dynamic nature of ethnic identity. Ethnic groups are subjectively
constructed, derived by actors who determine their own ethnicity, regardless
of the objective ‘reality’ of their cultural similarities or differences (Graves-
Brown 1996). Barth (1969) observed that the nomadic Pathans founded their
ethnic identity on a narrow selection of social elements, not broadly shared
cultural features. Instrumentalists argue that actors can shift or create ethnic
allegiances for social and economic gain. Royce (1982) documents how the
Zapotec of Juchitán manipulated ethnic allegiances to maintain political and
economic dominance, shifting between Zapotec and Mexican identities as the
situation demanded. In a similar but negative vein the supposedly primordial
ethnic polarization that characterized the most recent conflict in the Balkans in
reality consisted primarily of post hoc rationalizations (Graves-Brown 1996).
An overly instrumental approach runs the risk of reducing ethnicity to
no more than an economic or political strategy, and even instrumentalists
point out that ethnic identities are not made up of arbitrary features in order
to meet the instrumental needs of a particular situation. Jones (1997) argues
that ethnicity is grounded in Bourdieu’s (1977) notion of the habitus. Jones
avoids an extreme instrumental position by arguing that ethnic identity takes
on particular forms as a product of the habitus and the instrumental social
conditions of a particular context. This approach has the advantage of both
allowing for individual agency and acknowledging the important influence of
the particular social milieu of those actors.
Most important, ethnic identities are constructed through a consciousness
of difference with reference to the specific cultural practices of ethnic ‘others.’
As a result, competition and conflict sharpen ethnic polarization (Spicer 1962;
Figure 16.1. Ramesses III (1187–1156 BCE) leads ethnic prisoners before the god Amun-Re in this
relief at his mortuary temple in Medinet Habu (near modern Luxor, Egypt).
Isajew 1974; Hodder 1979; Royce 1982; Comaroff and Comaroff 1992; Jones
1997). For example, Herodotus reflects a Greek consciousness of both commu-
nity and superiority (Daugé 1981; Díaz-Andreau 1996). Ancient Egyptian and
Near Eastern ideology created and manipulated a positive ethic self juxtaposed
with negative ethnic others to legitimize the power and authority of their kings
(Loprieno 1988; Liverani 1990; Smith 2003). In each case the king pacified the
‘barbaric’ foreigners, protecting the inner order and civilization represented
by the ethnic self (Figure 16.1). Classical civilization created similar self/other
oppositions between the civilized and barbarian.
Kurzban et al. (2001) suggest that this emphasis on difference may have an
evolutionary explanation as a by-product of coalition building by our early ances-
tors. Racial and ethnic stereotypes were easily undermined when the researchers
replaced racial categories with an arbitrary yet visible cue correlated with group
membership, like shirt color. This is consistent with the notion that racial and
ethnic categories are socially constructed rather than having some fundamental
biological basis, as some continue to suggest. Constructions of ‘us-them’ quickly
trump physical differences that previously provided the basis for group identity.
Kurzban, Tooby and Cosmides later stated on a web-site that “in our experi-
ments, people quickly came to use an arbitrary visual marker to predict coalitional
alliances. This happened—with tragic consequences—as soon as it was suspected
348    S t u a r t T . S m i t h
that Middle Eastern terrorists were responsible for the WTC attacks: hate
crimes were perpetrated against individuals (e.g., Sikhs) who had coalitional
cues (turbans) reminiscent of headgear worn in the Middle East.”2 On a more
positive note, an ethnic group like Latino or Arab can incorporate considerable
physical and biological diversity. In a similar way, both the ancient Egyptian and
the Classical civilizations avoided the kind of pseudobiological color prejudice
characteristic of racism while practicing cultural chauvinism (Smith 2003).
When tied to power relations, the features selected to define the ethnic
other are often negative and subordinating, like recent derogatory character-
izations about Arab dress and culture in the wake of the World Trade Center
attacks or, in the case of ancient Egypt, the assertion that Kush must always
be ‘wretched’ (Smith 2003). In a military context (Figure 16.2) Nubians were
cowards, instantly defeated by the king if they even fought at all (Liverani
1990). Not only are foreigners wretched cowards, but they are hardly people
at all (Loprieno 1988). At the end of the New Kingdom text The Instruction of
Ani the student complains that no one could possibly learn everything that Ani
presents to him. Ani replies that animals can be trained, and that
One teaches the Nubian to speak Egyptian,
The Syrian and other strangers too.
Say “I shall do like all the beasts,”
Listen and learn what they do. (Lichtheim 1976:144)
In a similar way Afrikaner settlers called the Tswana of South Africa skepsels
(creatures), but the Tswana retaliated by calling the Afrikaners makgoa (white
bush lice) (Comaroff 1978; Crapanzano 1985). The Romans also created
negative stereotypes of the groups who resisted them. For example, to Pliny
the Blemmyes were headless people with eyes and ears in their chests, and
Solinus described them as barbaric savages who simply leaped on their prey
like animals when hunting.
Ethnicity in the Archaeological Record

Given the subjectivity and mutability of ethnicity, how can we get at it archaeo-
logically? The first thing to do is to abandon the search for the chimera of
neatly bounded ethnic groups corresponding to a particular material culture
assemblage. Instead, we need to focus on those parts of the material culture
that reflect social contexts where ethnicity is salient. Ethnicity is only one
of multiple identities that people assume in different social contexts. At any
http://www.psych.ucsb.edu/research/cep/erasingrace.htm (accessed 6 August
2.
2007)
Figure 16.2. Nubian warriors fall beneath the chariot of Pharaoh Tutankhamun (1332–1322 BCE) in
this detail of a painted box found in his tomb (now in the Egyptian Museum, Cairo).
particular time, other forms of identity, like gender or social position, might
play a more important role (Meskell 1994). Social arenas where ethnicity might
be, but is not inevitably, expressed include foodways, religious practice (espe-
cially relating to ancestors) and funerary practice (Santley et al. 1987).
Foodways vary both between and within cultures as a marker of status
(Goody 1982). Several archaeologists emphasize the importance of cuisine in the
construction of social identities, including ethnicity (Santley et al. 1987; Stanish
1989; Lightfoot and Martinez 1995; Burmeister 2000; Bunimovitz and Faust
2001). Funerals and burial practice demonstrate primordial ties, a key element
in the construction of ethnicity (Santley et al. 1987). It is no co-incidence that
cemeteries in Bosnia and Kosovo were targeted in campaigns of ethnic cleansing
(Chapman 1994). Funerals provide opportunities for the dead to make a final
assertion of identity but also, since the dead cannot bury themselves, they
allow for an active re-assertion or renegotiation of the social position and iden-
tity of their offspring and relatives (Hodder 1982; Morris 1987; Metcalf and
Huntington 1991; Meskell 1994). Two case studies reflecting Egyptian-Nubian
interactions during the second millennium BCE (Askut for foodways and reli-
gion, Tombos for funerary and religious practice) will be used to illustrated how
ethnic dynamics can be recognized in specific social contexts (Smith 2003).
350    S t u a r t T . S m i t h
Askut
At first glance Egyptian colonists in Nubia apparently forged a society identical
to the Egyptian core, unlike the northern empire in Syro-Palestine (Trigger
1976; Adams 1977, 1984; Kemp 1978; Higginbotham 2000). The fortress at
Askut is no exception. Built around 1850 BCE, this small fortified settlement
was occupied continuously through the initial period of Egyptian colonization
in the Middle Kingdom, through Nubian control in the Second Intermediate
Period (1630–1520 BCE) until the end of the New Kingdom in 1075 BCE
(Smith 1995). As is the case elsewhere in Lower Nubia, Askut’s whitewashed
walls would have symbolized Egyptian dominance and probably Egyptian
ethnicity as well. The architecture of the houses and the small chapel also
signaled an Egyptian identity (Figure 16.3). Egyptian-style artifacts dominate
the material assemblage, as is the case with the contemporary forts in the
region. As a result, Egyptologists have emphasized the emulative character
of these settlements, either characterizing them as a transplant of Egyptian
Figure 16.3. Plan of the fortress at Askut (Sudanese Nubia).

culture through colonization or as an indication of the complete assimilative

acculturation of native groups.
Archaeological studies of frontier communities caution against equating
the overall percentage of native versus colonial artifacts with cultural and
ethnic groups, calling instead for a nuanced analysis that focuses on different
components of the archaeological assemblage (Stanish 1989; Aldenderfer and
Stanish 1993; Lightfoot and Martinez 1995; Stein 1999). Farnsworth (1992)
breaks the material culture of California Native Americans incorporated into
the Spanish Mission system into a series of subassemblages, examining each for
continuity, replacement, and innovation. Similarly, Rogers (1990) breaks down
Arikara culture into three contexts: domestic, ritual and funerary. For Askut I
will focus on two areas, the domestic ceramic and ritual assemblages.
The ceramics from Askut were divided into three subassemblages: service,
storage and cooking pots (Figure 16.4). Nubian and Egyptian pottery differ
dramatically in manufacture and decoration, so it is easy to separate even body
sherds along cultural lines. As is the case at the other forts, Nubian pottery
Figure 16.4. Distribution of the inferred function of the pottery identified as Nubian found at Askut in
contexts dated to the Middle Kingdom (1975–1640 BCE), the Second Intermediate Period (1630–1520
BCE) and the New Kingdom (1539–1075 BCE).
352    S t u a r t T . S m i t h
appears consistently in Askut, in the small numbers that meet the expectations
of the Egyptological colonization and acculturation model. When broken down
into subassemblages, however, an interesting pattern emerges (Smith 2003).
The percentage of Nubian pottery in the service subassemblage fluctuates,
starting very low in the Middle Kingdom, increasing substantially in the Second
Intermediate Period, when Nubians controlled the area, and declining sharply
in the New Kingdom, correlating with the new colonial policy of assimilation.
Storage vessels show a similar distribution, but the pattern of cooking pottery
is very different. Nubian cooking pots are drastically overrepresented at Askut,
starting at nearly half of the cooking subassemblage, growing to two-thirds in
the Second Intermediate Period, and dominating during the New Kingdom.
Preliminary results of residue analysis using gas chromatography followed by
mass spectrometry imply that different Egyptian and Nubian cuisines also
existed at Askut. If we suppose, as Egyptian historical sources indicate, that
women did most of the cooking, then Askut reflects a kind of counteraccultura-
tion, with Nubian women transforming colonial foodways.
The spike in Nubian fine wares during the Second Intermediate Period may
reflect an instrumental assertion of ethnic ties demonstrating links between the
community and their new Kerma overlords through display during feasting,
perhaps driven by the men who helped manage the lucrative trade in luxuries.
This might correspond to Goody’s (1982:151–152) notion that the elements
of foodways that connect to larger political systems tend to change in order to
meet political contingencies. Goody also observes that those culinary practices
without external entanglements tend to be conservative, and the prominence of
Nubian cooking pots and cuisine at Askut may reflect the less overt influence
of Nubian women on the community’s foodways. The fact that the proportion
of Nubian cooking pots, and presumably Nubian cuisine, increases steadily
over time, however, implies that this is more than just a passive retention of a
Nubian habitus, that it is instead an active assertion of Nubian ethnic identity
that eventually came to dominate this particular social context. Although rooted
in the habitus, foodways play an important role in engendering and negotiating
social identities (Wood 1995). In a similar colonial context native women in
Spanish Saint Augustine used local pottery and maintained native foodways
(Deagan 1983). Lightfoot and Martinez (1995) argue that native women in
California’s Russian colony at Fort Ross used cuisine, and the organization of
domestic space, to assert their native identity. Foodways and burial practice
allowed slaves on southern plantations to maintain a separate ethnic identity,
and cuisine continues to play a key role in African American ethnicity (McKee
1999:235). Nubian women within Egyptian colonial communities like Askut
may have used foodways to provide an ethnic counterpoint to Egyptian political
and cultural hegemony.
Figure 16.5. Two figurines from Askut (Sudanese Nubia).
Nubian influence also appears in the presence of Nubian-style fertility and

cattle figurines (Figure 16.5). The increasing proportion of the latter may reflect
the Nubian religious emphasis on cattle, although the shapes are simple and
also appear in Egypt. Less ambiguous is a statuette of a seated pregnant woman
in classic Nubian style. It was found adjacent to a typically Egyptian household
shrine. Like cooking, fertility magic and household religion were under the
354    S t u a r t T . S m i t h
purview of women (Pinch 1993, 1994; Robins 1993). This pattern may reflect
a transculturation (Ortiz 1940; Deagan 1998) or blending of Nubian elements
into an Egyptian ritual setting.
Tombos
Located at the headwaters of the third cataract of the Nile, Tombos lay on an
important border, marked and explicitly indicated by a number of stelae carved
to commemorate the defeat of Kush by Pharaoh Thutmose I in 1502 BCE. It lies
only 10 km from Kerma, the former capital of the kingdom of Kush. Preliminary
evidence from two seasons of excavation indicates that the cemeteries at Tombos
were used from around 1400–600 BCE, and perhaps later, although whether
there is continuity between the New Kingdom and the later use of the cemetery
is still being investigated (Smith 2003). The location of the Egyptian colonial
settlement has also proven elusive. Apart from small amounts of Nubian pottery,
funerary architecture and grave goods reflect Egyptian burial practice. An elite
area contained perhaps 10 large pyramid tombs of a type popular with high-level
bureaucrats during the New Kingdom (Figure 16.6). Like the massive fortresses
of Upper Nubia, these impressive structures would provide a symbol of Egyptian
control, marking this important internal border within Egypt’s southern empire.
One of these tombs even had funerary cones dedicated to Siamun and his wife,
Weren, a type of decoration rarely found outside Thebes (Ryan 1988). As
‘Overseer of Foreign Lands,’ Siamun would have played a prominent role in the
colonial administration, regulating traffic across the internal border and perhaps
assembling the annual tribute in gold, cattle, slaves, ivory and other precious
goods from the conquered kingdom of Kush.
In a nearby middle-class cemetery, remains of decorated and inscribed
coffins, evidence for mummification, and specialized items (like Ushabti
figurines) reflect an Egyptian belief system (Smith 1992). The use of vaulted
subterranean chambers as family crypts also reflects Egyptian practices. Objects
of daily life, including personal items (like cosmetic equipment), furniture
(including the remains of a folding seat), a boomerang for hunting birds, and
an almost entirely Egyptian ceramic assemblage demonstrate its overwhelm-
ingly Egyptian cultural orientation. Luxuries like a rare Mycenean juglet attest
to the prosperity of the community (Smith 2003).
During the second season of excavation, however, we found burials of four
women in Nubian style (Figure 16.7), flexed and oriented with their heads to
the east while the bodies above and around them were buried mummified,
extended on their back and with their heads to the west in typical Egyptian
fashion. Looters had shoved one of these Nubian burials into an unusual posi-
tion in order to steal valuable jewelry but missed a set of Egyptian style amulets
Figure 16.6. Plan of the Pyramid of Siamun and Weren at Tombos in Sudanese Nubia.
dedicated to the dwarf god Bes, who protects the household from physical and
spiritual dangers. A Kerma-style cup was placed at the head of another set of
two Nubian burials, but apart from the cup and a couple of shell beads, the
grave goods associated with all four burials were Egyptian. Since Egyptian, and
presumably also Nubian, funerals were public events, the burial of women in
Nubian style at Tombos would make a very overt assertion of ethnic identity
against the very Egyptian monumentality of the elite cemetery’s pyramids.
356    S t u a r t T . S m i t h
Figure 16.7. Nubian-style burial of a woman, found on an Egyptian-style tomb at Tombos in Sudanese
Nubia.
Even though, at first glance, both Nubian sites seem to represent the kind
of colonial cultural conservatism discussed by Kopytoff (1987), we can see
the appearance of both Egyptian and Nubian cultural features as an active
social practice, an assertion of ethnic identity by different actors in different
social contexts. In settings like Egypt’s Nubian frontier, acculturation models
oversimplify cultural interactions by emphasizing the transmission of core
cultures to peripheral societies with little consideration for individual agency
(Lightfoot and Martinez 1995). The evidence of Egyptian and Nubian influ-
ence at Askut and Tombos re-inforces the notion that cultural interaction on
colonial frontiers is not an either-or proposition of assimilation versus innova-
tion. Individual men and women played a role in cultural and ethnic dynamics
that transformed colonial society into a dynamic social field for negotiating
cultural differences, not just producing a cultural hybrid. Even in a situation of
apparent ethnic polarization, coinciding with political borders and ideological
boundaries, we can see the permeability of putative ethnic boundaries and the
socially contingent nature of both ethnic identities and cultural interaction.
Nomads and Ethnicity

Pastoral nomads play a special role in their interactions with settled groups,
crossing, and in some cases undermining, state and ethnic boundaries. For
example, Welsby (1996:204) credits the collapse of the Meroitic state to the
infiltration of nomadic Nubians. Ethnic groups like the Nubians often have
complex subdivisions that seem to contradict the notion of a consistent, single
ethnic group, especially when they include a mobile component that ranges
over a wide territory (Barth 1969). Historical sources mention a variety of
ethnic groups and subgroups that might be classified as Nubian during the
Classical era, including the Nubae, Nobades, Annoubades, Noba, Nuba, Red
Noba and the Blemmyes, who may be related to the Pharaonic Medjai. It is not
clear to what extent these names refer to the same people or different groups
entirely (Burstein, this volume). Even within a single group there are differ-
ences. For example, there are indications that factions within the Blemmyes
allied with Rome, while others fought against Roman control. Evidence from
historical sources and sites like Deir al-Ballas in Egypt point to similar divisions
within the Nubian Kerman state.
Given the inconsistencies in terminology and the fact that the terms reflect
the construction of ‘other’ ethnic categories by Greeks, Romans, Byzantines
and Arabs, can we even start to investigate the ethnic dynamics of groups like
the Blemmyes? Some archaeologists would refute this proposition, but this
position represents a fundamental misunderstanding of ethnic dynamics. Perry
and Paynter (1999:306–307) note that archaeologists often uncritically adopt
an essentialist view of ethnicity, assuming that ethnic groups should be easily
visible through bounded sets of material culture. This reflects a highly norma-
tive view of culture. Attempts to correlate ethnic groups with archaeological
cultures are misguided when they are based on the assumption that material
correlates should match a primordial ethnicity. As ethnicity is situational and
contingent, we can expect a considerable degree of archaeological variability in
the expression of ethnic identity and overlapping rather than mutually exclusive
material culture distributions (Hodder 1982; Jones 1996; Hall 1997), especially
given the complex interactions that characterize a colonial frontier.
Ethnicity is not constant but dynamic, constructed and negotiated by
individual actors in specific social contexts. This requires an agent-centered
approach making it necessary to identify individuals in the archaeological
record, a difficult task even under ideal conditions, because individuals, and
not groups, make decisions about group identities, which can furthermore shift
depending on individual interests and desires. This is especially the case for
nomadic groups. Barth (1969) describes a considerable degree of heterogeneity
among the Pathans, who nevertheless come together as a unified ethnic group,
358    S t u a r t T . S m i t h
especially in the context of interaction with outsiders. In a similar way we might

expect a considerable degree of heterogeneity among groups ranging in the
Eastern Desert, between the Nile and the Red Sea, who might nonetheless
have exhibited a degree of ethnic solidarity, especially in their interactions with
the great powers that controlled Egypt.
Berenike
Several lines of evidence point to the presence of nomadic Blemmyes at Berenike
(Burstein, this volume; Sidebotham and Wendrich 1996). The following remarks
represent some preliminary suggestions for the exploration of ethnic dynamics at
this important entrepôt. Pottery now identified as Eastern Desert Ware appears
in domestic contexts, increasing in quantity in the last phase of the excavation
(Barnard 2002). This change correlates to a shift in the faunal assemblage to large
quantities of sheep and goats from the 4th century CE onward. Nubian-style
tumulus burials, overlooking the wadis (valleys) outside of the city, also appear in
this period. But does the presence of distinctive pottery and Nubian burial prac-
tice in an Eastern Desert context like Berenike represent an active assertion of
ethnicity or just a passive expression of habitus within a cosmopolitan setting?
Pottery, of course, does not necessarily equal the presence of people, but
foodways play a central role in enculturation. Meals and dining etiquette are
also tied to ethnicity; just think of the importance of ‘ethnic’ food in the defini-
tion of ethnic groups. Ethnic cuisine is often deployed as an active assertion of
ethnic identity, rather than reflecting a shared habitus. When displayed during
feasts, ceramics can also promote ethnic solidarity both within the group and
to outsiders. The correlation of Eastern Desert Ware with changes in the
faunal assemblage suggests a shift in cuisine as well as the ceramic assemblage,
like that seen at Askut. The possible presence of Nubian religious practices at
Berenike may also play a role in the ethnic dynamics of the city. In a lecture,
Rob Hughes suggested that a localized version of the Nubian Sol Invectus was
worshiped in Berenike’s North Shrine. Offering basins from both the North
and West Shrines date to the 4th century CE and may also reflect a Nubian
influence transmitted by nomads from the Eastern Desert.
Cemeteries like the 4th–6th century CE Nubian tumuli at Berenike provide
a clear connection between ethnicity and territory through an emphasis on
primordial ties. Against this, Shanna Kennedy argued in a lecture that although
Berenike’s tumuli superficially resemble tombs at Deraheib and Wadi Qitna
attributed to the Blemmyes,3 the differences in detail, and the widespread use
http://www.archbase.com/berenike/UCstudentLA6.html (accessed 6 August
3.
2007).
of similar Nubian practices throughout the Eastern Desert, make it implausible

to attribute all of them to a single ethnic group like the Blemmyes. This kind
of ethnic diversity, however, is by no means inconsistent with a unified ethnic
identity, especially when dealing with a nomadic people that range across a
large area and operate within different social and ecological environments (Nile
Valley, desert, Red Sea coast). Barth (1969) argued that not every diversifica-
tion within the group represents a first step in the direction of subdivision.
Diversity within an ethnic group, however, can also be approached from a
multiscale perspective, with seemingly disparate groups assembling together,
especially in the face of a common threat or competitor, only to dissolve into
different groups in different social and political settings. Deleuze and Guattari
(1987) observe that nomadic clan lineages are essentially segments in action;
they meld and divide, and vary according to the ancestor considered, the tasks
and the circumstances. So it is in fact quite plausible to associate the tumuli at
Berenike with the Blemmyes. Regardless, the tumuli’s strategic location along
desert ridges outside of Berenike would serve as an active assertion of ethnic
territoriality in the context of interaction with Rome.
Discussion
Perry and Paynter (1999:306–307) attribute the tendency for archaeologists and
historians to think of ethnic groups as distinctly bounded and exclusive to the
very nature of ethnic identities. These are inevitably constructed in essentialist
terms that obscure the actual complexity and contextual, gendered nature of
ethnicity in frontier communities like St. Augustine, Fort Ross, Askut, Tombos
and perhaps Berenike. The key attribute to understand about ethnicity is that
although people construct ethnic identities as an essential quality ascribed
at birth and immutable, in fact ethnicity is socially contingent and can shift
depending on the social and economic interests of individual actors. This has
profound implications for the archaeological search for ethnicity, which is
often an ultimately futile search for neatly bounded units that can be linked
to ethnic groups. We should abandon an overly normative view of ethnicity
and look carefully at different social contexts and expect inconsistencies, since
ethnic identities can shift in different situations. This means that we need to
think as much, if not more, in terms of individual actors than group dynamics
in order to get at ethnicity.
Nomads represent a similar dynamic of cultural contact to that between
Egypt and Nubia. Nomads generally interact with settled peoples, so we would
expect a mixed material assemblage but can still look for key social areas for
ethnic expression. For example, Eastern Desert Ware and the persistence of
tumulus burials stand out as two areas that might reflect assertions of ethnic
360    S t u a r t T . S m i t h
identity. The evidence from Berenike suggests a constellation of features,

domestic, religious and funerary that points to a possible ethnic influence or
perhaps an enclave within a multi-ethnic Berenike toward the end of the settle-
ment. The fact that some features, like the tumuli, are not as distinctive as one
might expect, or that Eastern Desert Ware constitutes only one component
of the ceramic assemblage, need not bother us. At Askut and Tombos, where
one would expect complete assimilation, we see a more complex pattern, with
Nubian artifacts as a whole accounting for a small percentage of the overall
assemblage but nevertheless with a Nubian ethnic identity emphasized in
specific social contexts. The Eastern Desert Ware found at Berenike may
operate in a similar way, reflecting the maintenance of ethnic foodways within
the settlement at the household level. Other pottery, and even elements of
foodways with Egyptian or classical connections, could be more salient to
social status. The cemeteries outside of Berenike might represent a primordial
claim to territory in line with ethnic dynamics. Like the Pathan, we should
expect heterogeneity among the Blemmyes but ethnic expression at bound-
aries like Berenike, where they came in contact with Romans, Egyptians and
others. Regardless, a more detailed examination of the archaeological record
at Berenike along these lines will shed light on the ethnic dynamics of this
important frontier community and its interactions with desert nomads like
the Blemmyes.
References
Adams, W. Y.
1977 Nubia: Corridor to Africa. London, Penguin.
1984 The First Colonial Empire: Egypt in Nubia, 3200–1200 B.C. Comparative
Studies in Sociology and History 26: pp. 36–71.
Aldenderfer, M. S. and C. Stanish
1993 Domestic Architecture, Household Archaeology, and the Past in the South-
Central Andes. In M. S. Aldenderfer (ed.), Domestic Architecture, Ethnicity,
and Complementarity in the South-Central Andes. Iowa City, University of Iowa
Press: pp. 1–12.
Anderson, B.
1983 Imagined Communities: Reflections on the Origins and Spread of Nationalism.
London, Verso.
Baines, J. and J. Malek
2000 Cultural Atlas of Ancient Egypt. Abingdon, Andromeda.
Banks, M.
1996 Ethnicity: Anthropological Constructions. London, Routledge.
Barnard, H.
Barth, F. (ed.)
1969 Ethnic Groups and Boundaries. Boston, Little, Brown.
Bourdieu, P.
1977 Outline of a Theory of Practice. Cambridge, Cambridge University Press.
Bunimovitz, S. and A. Faust
2001 Chronological Separation, Geographical Segregation, or Ethnic Demarcation?
Ethnography and the Iron Age Low Chronology. Bulletin of the American
Schools of Oriental Research 322: pp. 1–10.
Burmeister, S.
2000 Archaeology and Migration. Current Anthropology 41: pp. 539–567.
Chapman, J.
1994 Destruction of a Common Heritage: The Archaeology of War in Croatia,
Bosnia, and Hercegovina. Antiquity 68: pp. 120–126.
Comaroff, J. L.
1978 Rules and Rulers: Political Processes in a Tswana Chiefdom. Man 13: pp.
1–20.
Comaroff, J. and J. Comaroff
1992 Ethnography and the Historical Imagination. Boulder, Westview Press.
Crapanzano, V.
1985 Waiting: The Whites of South Africa. New York, Random House.
Daugé, Y. A.
1981 Le Barbare: Recherches sur la conception romaine de la barbarie et de la civilisation.
Brussels, Révue d’Études Latines.
Deagan, K.
1983 Spanish St. Augustine: The Archaeology of a Colonial Creole Community. New
York, Academic Press.
1998 Transculturation and Spanish American Ethnogenesis: The Archaeological
Legacy of the Quincentenniary. In J. Cusick (ed.), Studies in Culture Contact:
Interaction, Culture Change, and Archaeology. Carbondale, Southern Illinois
Deleuze, G. and F. Guattari
1987 A Thousand Plateaus. Minneapolis, University of Minnesota Press.
Díaz-Andreau, M.
1996 Constructing Identities Through Culture: The Past in the Forging of Europe.
In P. Graves-Brown, S. Jones and C. Gamble (eds.), Cultural Identity and
Archaeology. London, Routledge: pp. 48–61.
Farnsworth, P.
1992 Missions, Indians, and Cultural Continuity. Historical Archaeology 26: pp.
22–36.
362    S t u a r t T . S m i t h
Glazer, N. and D. P. Moynihan (eds.)

1963 Ethnicity: Theory and Experience. Cambridge, Harvard University Press.
Goody, J.
1982 Cooking, Cuisine, and Class: A Study in Comparative Sociology. Cambridge,
Graves-Brown, P.
1996 All Things Bright and Beautiful? Species, Ethnicity, and Cultural Dynamics.
In P. Graves-Brown, S. Jones and C. Gamble (eds.), Cultural Identity and
Archaeology. London, Routledge: pp. 81–95.
Hall, J. M.
1997 Ethnic Identity in Greek Antiquity. Cambridge, Cambridge University Press.
Handler, R.
1988 Nationalism and the Politics of Culture in Quebec. Madison, University of
Wisconsin Press.
Hides, S.
1996 The Genealogy of Material Culture and Cultural Identity. In P. Graves-
Brown, S. Jones and C. Gamble (eds.), Cultural Identity and Archaeology.
London, Routledge: pp. 25–47.
Higginbotham, C. R.
2000 Egyptianization and Elite Emulation in Ramesside Palestine. Leiden, E. J. Brill.
Hodder, I.
1979 Economic and Social Stress and Material Culture. American Antiquity 44: pp.
446–454.
1982 The Present Past: An Introduction to Anthropology for Archaeologists. London,
Batsford.
Isajew, W.
1974 Definitions of Ethnicity. Ethnicity 1: pp. 111–124.
Jones, S.
1996 Discourses of Identity in the Interpretation of the Past. In P. Graves-Brown,
S. Jones and C. Gamble (eds.), Cultural Identity and Archaeology. London,
1997 The Archaeology of Ethnicity: Constructing Identities in the Past and Present.
London, Routledge.
Kemp, B. J.
1978 Imperialism in New Kingdom Egypt (c. 1575–1087 B.C.). In P. D. A. Garnsey
and C. R. Whittaker (eds.), Imperialism in the Ancient World. Cambridge,
Cambridge University Press: pp. 7–57, 283–297.
Kohn, H.
1944 The Idea of Nationalism: A Study of Its Origins and Backgrounds. New York,
Macmillan.
Kopytoff, I.
1987 The African Frontier: The Reproduction of Traditional African Societies.
Bloomington, Indiana University Press.
Kurzban, R., J. Tooby and L. Cosmides
2001 Can Race be Erased? Coalitional Computation and Social Categorization. In
Proceedings of the National Academy of Sciences 98: 26, pp. 15387–15392.
Lichtheim, M.
1976 Ancient Egyptian Literature: A Book of Readings. Volume 2: The New Kingdom.
Berkeley, University of California Press.
Lightfoot, K. G. and A. Martinez
1995 Frontiers and Boundaries in Archaeological Perspective. Annual Review of
Liverani, M.
1990 Prestige and Interest: International Relations in the Near East, ca. 1600–1100 B.C.
Padova, Sargon.
Loprieno, A.
1988 Topos und Mimesis. Wiesbaden, Harrassowitz.
McKee, L.
1999 Food Supply and Plantation Social Order: An Archaeological Perspective. In
T. A. Singleton (ed.), “I, Too, Am America”: Archaeological Studies of African-
American Life. Charlottesville, University Press of Virginia: pp. 218–239.
Meskell, L. M.
1994 Dying Young: The Experience of Death at Deir el-Medineh. Archaeological
Review from Cambridge 13: 2, pp. 35–45.
Metcalf, P. and R. Huntington
1991 Celebrations of Death: The Anthropology of Mortuary Ritual. Cambridge,
Morris, I.
1987 Burial and Ancient Society: The Rise of the Greek City-State. Cambridge,
Ortiz, F.
1940 Contrapunteo cubano del tabaco y el Azúcar [in Spanish]. Havana, Montero.
Parker, B. J.
2002 At the Edge of Empire: Conceptualizing Assyria’s Anatolian Frontier, ca. 700
BC. Journal of Anthropological Archaeology 21: pp. 371–395.
Perry, W. and R. Paynter
1999 Artifacts, Ethnicity, and Archaeology of African Americans. In T. A. Singleton
(ed.), “I, Too, Am America’: Archaeological Studies of African-American Life.
Charlottesville, University Press of Virginia: pp. 299–310.
Pinch, G.
1993 Votive Offerings to Hathor. Oxford, Griffith Institute.
364    S t u a r t T . S m i t h
1994 Magic in Ancient Egypt. Austin, University of Texas Press.

Prescott, J. R. V.
1987 Political Frontiers and Boundaries. London, Unwin Hyman.
Ratzel, F.
1897 Politische Geographie. München-Berlin, Oldenbourg.
Rawlinson, G.
1964 The Histories of Herodotus. London, Dent.
Renfrew, C.
1996 Prehistory and the Identity of Europe, or, Don’t Let’s Be Beastly to the
Hungarians. In P. Graves-Brown, S. Jones and C. Gamble (eds.), Cultural
Identity and Archaeology. London, Routledge: pp. 125–137.
Robins, G.
1993 Women in Ancient Egypt. Cambridge, Harvard University Press.
Rogers, J. D.
1990 Objects of Change: The Archaeology and History of Arikara Contact with Europeans.
Washington, Smithsonian Institution Press.
Royce, A. P.
1982 Ethnic Identity: Strategies of Diversity. Bloomington, Indiana University
Press.
Ryan, D. P.
1988 The Archaeological Analysis of Inscribed Egyptian Funerary Cones. Varia
Egyptiaca 4: pp. 165–170.
Santley, R., C. Yarborough and B. Hall
1987 Enclaves, Ethnicity, and the Archaeological Record at Matacapan. In R. Auger,
M. F. Glass, S. MacEachern and P. H. MacCartney (eds.), Ethnicity and Culture.
Calgary, Archaeological Association of the University of Calgary.
Sidebotham, S. E. and W. Z. Wendrich
1996 Berenike 1995: Preliminary Report of the 1995 Excavations at Berenike (Egyptian
Red Sea Coast) and the Survey of the Eastern Desert. Leiden, Research School
CNWS.
Smith, A. D.
1986 The Ethnic Origins of Nations. Oxford, Blackwell.
Smith, S. T.
1992 Intact Theban Tombs and the New Kingdom Burial Assemblage. Mitteilungen
des Deutschen Archäologischen Instituts Kairo 48: pp. 193–231.
1995 Askut in Nubia: The Economics and Ideology of Egyptian Imperialism in the Second
Millennium BC. London, Kegan Paul.
2003 Wretched Kush: Ethnic Identities and Boundaries in Egypt’s Nubian Empire.
London, Routledge.
Spicer, E.
1962 Cycles of Conquest. Tucson, University of Arizona Press.
Stanish, C.
1989 Household Archaeology: Testing Models of Zonal Complementarity in the
South Central Andes. American Anthropologist 91: pp. 7–24.
Stein, G. J.
1999 Rethinking World Systems: Diasporas, Colonies, and Interaction in Uruk
Mesopotamia. Tucson, University of Arizona Press.
Tonkin, E., M. McDonald and M. Chapman
1989 Introduction. In E. Tonkin, M. McDonald and M. Chapman (eds.), History
and Ethnicity. London, Routledge: pp. 1–21.
Trigger, B. G.
1976 Nubia Under the Pharaohs. London, Thames and Hudson.
Welsby, D. A.
1996 The Kingdom of Kush. London, British Museum Press.
Wendl, T. and M. Rösler
1999 Frontiers and Borderlands: The Rise and Relevance of an Anthropological
Research Genre. In M. Rösler and T. Wendl (eds.), Frontiers and Borderlands:
Anthropological Perspectives. New York, Peter Lang: pp. 1–27.
Wood, R. C.
1995 The Sociology of the Meal. Edinburgh, Edinburgh University Press.
Ch a p t e r 1 7
Variability and
Dynamic L andscapes
of Mobile Pastoralism
in Ethnography and
Prehistory
Michael D. Frachetti1
P rehistoric nomadic pastoralism presents a unique analytical and

theoretical problem for archaeologists, in that often we are trying to
explain the prototypical forms of a social and economic way of life that regu-
larly defies a ‘typical’ classification, even in a given context (Dyson-Hudson and
Dyson-Hudson 1980). From the wealth of ethnographic studies concerning
contemporary nomadic pastoralists, we may only be able to generalize two
rudimentary facts: nomadic pastoralism reflects an intensive engagement in
herding of domesticated animals as a primary economic and social way of life;
and the strategies and practices (movement, animal management, settlement,
trade, warfare, etc.) of nomadic pastoralists are adapted in response to the
geographic and temporal dynamics of their environment; their socio-ideolog-
ical, political and economic relationships; and their individual or group health
and well-being. An additional caveat to these observations is that the frequency
and amplitude of change across such factors is both irregular and codependent
on the nature of the strategies employed.
These conditions can be confounding for the archaeologist, because in the
first case the prevalence of domesticated animal remains in archaeological
contexts is not sufficient to argue for a nomadic way of life in prehistory. So,
although they can be a significant indicator, such data can be easily overvali-
dated as a requisite aspect of ‘nomadic sites.’ The second condition presents
a frustrating feedback loop, in that the complex layering of environmental,
political, and socio-economic considerations that ultimately affects the choices
The research on which this chapter is based was funded by grants from the National
1.
Science Foundation and the George F. Dales Foundation. Radiocarbon dates were
analyzed by the Arizona AMS laboratory.
366
Mobile Pastoralism in E t h n o g r a p h y      367
and practices of nomadic pastoral societies is impacted and even shaped by

the spatial and temporal patterning of those very strategies, thereby indexing
a highly dynamic way of life that sometimes appears categorically ‘nomadic,’
sometimes looks more ‘sedentary’ and is regularly recast in different places
and at different times in different forms. Thus, a paradox lies in the fact that
perhaps the only ‘regular’ aspect of nomadic pastoral lifestyles is the condi-
tion of variability. This observation constitutes the first point of this chapter
and is illustrated below through some well-developed ethnographic studies of
nomadic societies of Western and Central Asia.
Within archaeology the recovery of variation in the layout of domestic
contexts, economic strategies, ritual constructions and material culture often
leads to typological classifications in the attempt to order distinct social or
cultural groupings: A’s with A’s, B’s with B’s, and C’s with C’s. This is especially
the case when the relative chronology of sites is in question. Correlating typo-
logical distinctions with particular social or economic forms, however, may be
the wrong approach in the archaeology of mobile pastoralism, as categorical
classifications can mask the potential plurality of strategies employed by a
given society. A more useful approach may be to recognize that A’s, B’s and
C’s can reflect the variation of strategies, such as different settlement or camp
configurations, that enables pastoralists to maintain social cohesiveness and
adaptive success within the geographic and temporal fluctuations of their
experienced landscape.
The second aim of this chapter is to propose an analytical approach to
the archaeology of prehistoric mobile pastoralists that takes the focus away
from the identification and categorization of ‘nomadic’ or ‘non-nomadic’ data
classes and turns it toward the issue of geographic and temporal variability, as
reflected archaeologically across pastoral landscapes. The argument here is that
prehistoric mobile pastoral societies can be better understood by relating the
archaeological variation within various data classes to the optional practices
and adaptations relevant to different environmental and social contexts charted
across geographic and temporal planes. In other words, this approach advocates
modeling how changes in strategy and choice are mapped onto recoverable
archaeological landscapes and how the range of options co-varies with other
dynamic factors (environment, technology, etc.) over time. Logically, chrono-
logical contemporaneity within a range of data is key to the argument, as
variation can essentially reflect two scenarios: change over time or variation
within a range set. In practice these scenarios combine to produce considerably
complex social, economic and political landscapes.
In the final part of this chapter the proposed approach is applied to a case
study of Bronze Age societies in eastern Kazakhstan, illustrating that these
pastoral groups may have employed a variety of strategies that range across the
368     M i c h a e l D . F r a c h e tt i
ideal categories of ‘nomadic’ or ‘sedentary’ herders and thereby contributed to

the formation of extensive networks of social and economic interaction during
the second millennium BCE.
Ethnogr aphy of Nomadism as

a Study of Variation
Nomadic pastoralism is most commonly understood as a way of life based
predominantly in the social and economic strategies associated with a routine
migratory management of domesticated herd animals (Lattimore 1940:54; Barth
1964:4; Khazanov 1994:17). Etymologically, the words nomadism and pasto-
ralism both imply pasturing or the raising of herds (Spooner 1973:3; Salzman
2002:245). A number of scholars, such as Barfield (1993:4), however, note that
the term nomadism is also sometimes used in association with other mobility
strategies, such as hunting and gathering. Thus, Barfield distinguishes nomadic as
a referent to movement, or mobility, and pastoralism as a referent to a productive
strategy, raising livestock on natural pastures (see also Salzman 2002:245). A
number of scholars have long recognized that nomadic-pastoral strategies reflect
a considerable degree of variation that makes normative categories, generated on
the basis of ideal economic or social types, inadequate as explanatory paradigms.
Contemporary ethnographers have noted that a broad definition of nomadic
pastoralism rather inadequately describes the wide range of socio-economic
strategies recorded among societies who rely on herding (Salzman 1972:67;
Spooner 1973:4) and does not in itself describe the variability in social and
political practices that are documented within these societies (Dyson-Hudson
and Dyson-Hudson 1980). Yet they commonly agree that the broad range of
pastoral strategies (mobility, multiresource exploitation, etc.) are adaptive in
view of specific environmental variations (Bacon 1954:54), human and animal
ecology (Barth 1964; Spooner 1973; Koster 1977), and socio-ideological and
political contexts (Irons 1974; Tapper 1979). In fact, it is difficult to emphasize
one of these contributing factors over the other in forming typological defini-
tions, as ethnographic examples illustrate differing emphasis on each of these
factors. In some cases these factors may even fluctuate in their importance to
the organization and practices of a particular pastoral group. Dyson-Hudson
and Dyson-Hudson (1980:18) summarize this point nicely: “since a unique
constellation of ecological, political, economic, and affective factors determines
the patterns of movement of each pastoral group, and the specific movements of
each independent herd owner within every pastoral society, it is not surprising
that there is enormous variation in patterns of mobility.”
The ethnographic record exhibits a diversity of adaptations and particular
pastoral strategies that confound categorization of ideal types. Migration
patterns, for example, are often altered through iterated engagements in

productive practices such as agriculture or market trade (Bradburd 1990:34–39),
changes in political organization (Shahrani 1979:171–172) or participation
in complex and changing routines of social and economic interaction with
neighboring populations (Barth 1964:109; Beck 1991). Salzman (2002:256)
reiterates this point directly: “shifting between strategies of adaptation [on the
part of nomadic societies] in response to changes in conditions has been very
common throughout the Middle East and North Africa. We must also keep
in mind that ‘settled’ and ‘nomadic,’ rather than being two types, are better
thought of as opposite ends of a continuum with many gradations of stability
and mobility.” Ethnographic studies show that nomadic societies often consist
of groups who exhibit variously specialized economic practices as part of one
sociopolitical structure, thereby defying rigid dichotomies between peasant
and nomadic ways of life.
Irons’s ethnography (1974:636–637) of the Yomut Turkmen provides a clear
example of such a nomadic pastoral system in the Gurgan Plain of Iran. The
Yomut Turkmen maintain two occupationally different factions within their
tribal organization. These subgroups are called the chomur and the charwa, the
former being primarily agriculturalists, the latter engaged more exclusively in
pastoralism. The economic relationship between these groups is supportive,
and socially they adhere to a common tribal organization. Both groups employ
a degree of mobility in their exploitation of the limited resources of their envi-
ronment, though the charwa rely more heavily on seasonal migration than do
the chomur, even though both groups could feasibly lead far more sedentary
lifestyles given their economic demands. Both groups distinguish themselves
politically and ideologically from non-Turkmen groups of the same region and
use their flexibility in residence as a strategy for resisting political control. Irons
(1974:654) contends that the Yomut are strategically able to negotiate multiple
political contexts more effectively because their fluctuating patterns of mobility
and symbiosis in agricultural and pastoral production enable them to evade
taxation and state control, while maintaining viable economic productivity.
Irons’s example of the Yomut shows that ‘nomadic pastoralism’ can encompass
variations in seasonal migration, settlement, agricultural emphasis and social
interaction, making it a highly adaptive strategy.
Salzman (1972:66–67) has proposed that the Yarahmadzai and Gamshadai
pastoral tribes of Baluchistan engage in ‘multi-resource nomadism,’ varying
their movement patterns to accommodate the demands of pastoral production
and to take advantage of productive date cultivation and the sale of labor in
regional markets. Salzman cites these alternative strategies as evidence that
pastoralists often maximize their economic and social success in marginal
environments by engaging in practices that are not typically associated
with societies classified as ‘nomads.’ Variation in mobile pastoral systems is

commonly linked to both the ecology of herding and sociopolitical negotiations
(Tapper 1979:111). These factors can contribute to significant changes in the
way pastoralists manage territory and lay claim on locations in their landscape
(pastures and campgrounds).
Barfield’s study (1981:44–46) of the Central Asian Arabs of Afghanistan
describes how some nomadic Arabs claim exclusive rights to particular pasture
zones, based on rights established through complex political dealings with
regional and national political bodies in the early part of the 20th century
CE. In light of the environmental variability in pasture quality from year
to year, ownership and control of particular locations and resources such as
summer and winter pastures (ailoq and qhishloq) and seasonal cisterns (yekhdon)
engendered various forms of social interactions, such as trading of resources,
political alliances, and land rental, to meet the needs of domesticated herds.
Barfield’s example describes how the environmental variability of mountainous
environments conditioned social practices of greater investment in demarcated
locales, contributing to an ecologically ordered but socially negotiable pattern
of mobility and pastoral land use.
Pastner (1971:175–180) describes an interesting case where environmental,
political and ideological systems of the Makran Baluch of western Pakistan
result in various patterns of interactions related to territorial and social affili-
ations at local and regional scales. Pastner emphasizes how localized patterns
of mobility, or the ‘micro-pastoral orbit’ used by the Makrani nomads to
accommodate the demands of herd animals and social groups in a marginal
environment, are also extended for purposes of resource exploitation and socio-
economic strategies not specific to herd needs. Alternative aims, such as trade,
raiding, itinerant agriculture or the sale of labor, introduce unique mobility
patterns and bring nomads into close interactions with sedentary villagers while
settling in peripheral residence camps near agricultural villages, often during
the time of haman (harvest). The nature of camp formation and territorial
use in the micropastoral orbit during this time is also affected by ideological
concerns that arise from the social pressures associated with the dynamics of
interaction around village groups. Islamic concern with ‘purdah,’ the protec-
tion of the honor and purity of women, is a factor that affects economic and
practical decisions, as life around imposing sedentary villagers is seen as a risk
for the women. Thus, there is an ideological justification for the formation of
group camps, where women may be better shielded from sexual predation on
the part of outsiders. For the Makrani Baluch, the year-to-year variability of the
mobility strategies in their nomadic search for pasture also brings about inter-
action and overlap between various contiguous micropastoral orbits, forming
what Pastner calls a ‘macro-pastoral orbit’ or territory. This macropastoral
orbit generates for the pastoralists a wide range of regional alliances and
social affiliations between groups. Such social affiliations become significant in
negotiating economic and political relationships when disparate groups come
together on the outskirts of sedentary village contexts during haman, once
again serving the purpose of protecting their ideological concerns. Pastner
(1971:182) notes that “it is at this point that social parameters of the macro-
pastoral orbit pay off . . . Co-resident encampments of nomads are composed
of people united in the web of consanguinity, affinality, and friendship of the
macro-orbit . . . These co-resident members of the macro-orbit provide the
means of alleviating the apprehensions of men about their women’s sexual
safety, particularly vulnerable, as it were, during haman.” Pastner’s example
illustrates the overlapping forces of environmental adaptation and political
interaction that contribute to the variable scale and pattern of micro- and
macropastoral orbits. It also demonstrates how the patterned and variable
mobility of the Makrani Baluch results in the formation of social alliances
and cohesive social units at camps through the organization of territory along
political, economic and ideological lines.
Shahrani’s (1979:112–116) study of the Kirghiz of the Wakhan Corridor and
Pamir Mountains of Afghanistan describes a case of nomadic pastoralism where
pastoral mobility patterns and associated social interactions with neighboring
populations were drastically affected by changes in the political geography
of northern Afghanistan, China, and the USSR in the early 20th century CE.
Shahrani provides a detailed discussion of the ecological impact of the harsh
high-altitude environment of the Pamir Mountains on pastoral strategies and
illustrates that the ethnic Kirghiz practice an ‘intensive’ pattern of pastoral
mobility, which varies considerably within a confined territory according to
the seasonal alpine climate and pasture dynamics. Shahrani (1979:116) notes
that “the Kirghiz are intensive pasturage users and the distance covered in
their pendular migrations is relatively short but not uniform. The farthest
distance between camps occupied by the same herding unit during a year may
range from fifteen to thirty-five kilometers. However the distance covered
during a change of encampment (e.g. winter to spring) . . . may be less than
that suggested above.” He further shows that this pattern of mobility was not
always typical: prior to 1949 the Kirghiz employed more extensive migrations,
moving over 150 km into lowland valleys in present-day China and Tajikistan,
interacting with the Wakhi, who are settled agriculturalists of the Wakhan
corridor (Shahrani 1979:171). Nevertheless, Shahrani finds historical conti-
nuity in the social identity of the Kirghiz in spite of these changes (Shahrani
1979:170). Looking at this case, we observe two scales in which the Kirghiz
exhibit variation in their migratory pastoral adaptation. The first is the local
and contemporary scale, where their intensive adaptation to the mountain
ecology alters their mobility patterns from year to year. The second is the
regional and historical scale, whereby their mobility pattern has paralleled
considerable change in both environmental exploitation and social interaction
over the past 55 years.
The key observation from these ethnographies, among many others, is that
mobile pastoral systems often reflect a highly changeable strategy for managing
social and ecological demands within a variety of environmentally, politically
and ideologically dynamic contexts. Therefore, archaeologists may benefit
from the observation that societies engaged in mobile forms of pastoralism
commonly construct a social landscape that on the one hand is ordered by their
patterns of herd management in response to fluctuating ecological contexts,
while at the same time produces variations in social contexts according to the
negotiation of social, economic, ritual or political conditions. On this basis
typological categorizations of nomadic pastoralism in current ethnography
have been superseded by more focused attention on the historical and practical
particulars of ‘mobile pastoral’ ways of life (Humphrey and Sneath 1999), which
may lead one to agree with Kavoori’s (1999:14) optimistic remark that “we are
well past the earlier sterile typological concerns that sought to classify pasto-
ralists as nomads, semi-nomads, transhumants, and so on.” Yet oddly, it is still
common in archaeological studies to rely on basic categories of economic and
social modes of nomadic pastoralism. This is especially the case in studies of
the Eurasian steppe (Kosarev 1984; Khazanov 1994), which is the focus of the
case study below. Perhaps this is due to the fact that archaeologists often have
less refined evidence than ethnographers to describe the complex pressures that
contributed to dynamic prehistoric pastoral systems. Yet this complication does
not justify a categorically simple description of mobile forms of pastoralism
in prehistory. Archaeologists can productively investigate the archaeological
signatures of variation in pastoral contexts and benefit from the ethnographic
recognition that choice and strategic variability are key aspects to the success
and evolution of pastoral societies over time.
The Archaeology of Mobile Pastor alism

The archaeology of prehistoric mobile pastoralism has suffered from the
lack of an approach to target the condition of variability within pastoral
systems. Archaeological data such as site layouts, faunal remains and artifact
assemblages, which are presented as analogous indicators of socio-economic
strategies known from ethnographically recorded ‘nomadic’ societies, have
typically served as the basis for identifying prehistoric nomadic pastoralism in
the Near East and Central Asia (Cribb 1991; Bar-Yosef and Khazanov 1992;
Kohler-Rollefson 1992). Although these data are compelling evidence for
prehistoric pastoral adaptations, they may not present a complete picture if

categorically separated from other ‘less analogous’ archaeological contexts.
Even though variations exist in various classes of data across these archaeo-
logical landscapes, the approach has more often been to place, for example,
tent camps and permanent architecture in categorical distinction, rather than to
conceive of these features as part of a range of settlement options reflecting less
distinction between pastoralists and others in the fabric of regional prehistoric
societies (but see Rosen 2003). Perhaps one rationale for separating nomadic
and agricultural populations in Near Eastern archaeological settings lies in
the greater formal difference between the archaeological remains of campsites
and large urban settlements, and the apparent distinctions in the political
economies of protostates and contemporary tribal groups (but see Lamberg-
Karlovsky 2003). Even so, archaeologists working in this part of the world are
quick to recognize that these groups were likely linked in economic, if not
social, symbiosis (Danti 2000), but nomads are still relegated to the periphery
in terms of their social and political agency in such contexts.
The relationship between large-scale agricultural settlements and the devel-
opment of specialized pastoralism is not well documented to date in Central
Asia and the Eurasian steppes (Lamberg-Karlovsky 2003). Eneolithic ‘culture
groups’ such as the Atbasar and the Botai of the central steppes reveal little
evidence for a developed agricultural economy in the third millennium BCE
(Kislenko and Tatarintseva 1999). Like the Atbasar, third millennium BCE
societies in Inner Asia, such as the Afanas’ev in the northeastern forest steppes,
were primarily hunter-fishers with only limited herding of cattle (Khlobystina
1973; Shilov 1975; Vadetskaya 1986). Faunal evidence in this region indicates
that pastoral exploitation of horses, cattle and sheep only became predominant
by the end of the third millennium BCE (Tsalkin 1964), such that the model of
emerging pastoralism in the Eurasian steppe becomes increasingly dissimilar to
that proposed for the Near East. In the western Eurasian steppes, in regions of
Southern Russia, North Caucuses, and north of the Caspian Sea, the possibility
that specialized pastoralism emerged from mixed agropastoral subsistence
strategy is better documented. Settled agricultural practices of societies such
as the Srubnaya and Tripolye are well documented for the late third millen-
nium BCE and aggregate sites, such as Sintashta and Arkaim in the southwest
Ural region, illustrate that a mixed economy of agriculture and pastoralism
was developing by the beginning of the second millennium BCE (Chernykh
1997; Jones-Bley and Zdanovich 2002). Thus, evidence for both agricultural
and pastoral economies is known, albeit scantily, in this part of Eurasia.
Archaeobotanical studies in the western steppe region and the Samara
Valley, however, have yet to recover any evidence of domesticated plants, even
where comprehensive flotation strategies were employed (David Anthony, pers.
comm.). Shishlina (this volume) also notes this trend at Bronze Age sites in the
north Caucasus, where her archaeobotanical studies have not revealed domes-
ticated plants. From these results we might propose that ‘pastoral’ systems
in the western steppe reflect a degree of specialization, where some groups
were engaged in agricultural production, while other groups, throughout the
broader region, were not. Unfortunately, the relationship between these two
Bronze Age strategies, either as socially specialized economies or as part of a
common adaptive strategy, are still underdocumented. These archaeological
debates are hindered by the lack of an approach to situate archaeological data
within a framework highlighting the variable social and economic strategies
of mobile pastoralists in prehistory. In part, the investigation of variability
in prehistoric pastoralism is limited by the desire to match archaeological
evidence to the paradigm of set economic modes of production. This approach
leaves our understanding of the emergence of prehistoric mobile pastoralism
foggy at best. When overly generalized categories are used paradigmatically
to explain prehistoric processes, archaeologists are snared somewhat unwit-
tingly within an analytical tautology. Namely, typical categories are used to
model prototypical scenarios, which in turn are used to justify the generation
of the category itself. To be sprung from this loop, an approach to prehistoric
mobile pastoralism is needed that documents the variability of mobile pastoral
systems in the past and that accounts for the possibility of unique combinations
of ecological, social, political and ideological practices. Landscape archaeology
is a useful springboard for developing such an approach.
Dynamic Pastor al Landscapes

Landscape-oriented studies have been part of archaeology long enough that the
term alone does not convey a singular approach (Ashmore and Knapp 1999;
Stoddart 2000). Anschuetz et al. (2001:158) remark that the imprecise definition
of landscape is a problem that plagues archaeology, as well as other disciplines
such as geography, as all struggle to understand the “fundamental nature of the
relationship between people and the spaces they employ.” Beyond this observa-
tion, most agree that landscape archaeology situates past populations in both an
environmental and social milieu, where they create and negotiate the ecological,
political, ideological and ritual boundaries of their way of life (McGlade 1995;
Knapp and Ashmore 1999; Anschuetz et al. 2001). Ingold (1993:152) construc-
tively points out that ‘landscapes’ reflect the impact of agents situated in time
and space, a vantage point specifically useful for studying mobile pastoralists,
whose pattern of life is often synchronous with environmental cycles and whose
economic and political activities can be both patterned and flexible (Barth
1969; Beck 1991). From this perspective mobile pastoralism can be studied as
the mobile activation of various geographic, economic, ideological, social and

political landscapes united into one mode of life.
The landscape approach promoted here assumes that various contexts of
pastoral praxis distributed over a given territory contribute to discernible anthro-
pogenic ‘footprints’ that correspond to specific adaptive practices employed over
time, while changing the natural and social environment according to strategic
choices (McGlade 1995; Erickson 2000). What is perhaps most appealing about
this definition is the allowance for variability in human strategies within periodi-
cally different snapshots of the environmental and social context. The creation
of landscapes by societies, over lifetimes and longer durations of time, will be
reflected by the adaptive stability of certain ways of exploiting the environment
and by variations in the social employment of both natural and anthropogenic
locales. Ultimately, ecological and archaeological documentation of periods
of stability and change in the constructed landscape provides an entrée to
discussions of more slippery topics, such as how social, political, economic and
ideological frameworks impacted regional populations over time.
As stated by McGlade (1995:114), “we need to understand the conception of
nature and the location of humans within its ambit—not simply as a dynamical
system, but as part of a social historical process.” He proposes that in order
to bridge the dialectic between nature and culture, archaeologists should be
concerned with ‘human eco-dynamics,’ which he defines as “the dynamics of
human modified landscapes set within a long-term perspective, and viewed as
a non-linear dynamical system” (McGlade 1995:126). This use of a nonlinear
model of causation loosens the relationship between human strategies and
historical outcomes, while taking into account the fact that human actions do
result in recoverable and distinctive structures over time. McGlade’s paradigm
is powerful in that it situates the agent in the foreground of landscape concep-
tion yet recognizes that the practice of building social relationships is indeed
conditioned by the historically extant structure of the landscape. Thus, the
spatial and temporal constraints of the natural environment are conditioned
and negotiated through patterns of land use and the variability of human
interaction within both the ecological and social affordances of the landscape.
Human ecodynamics is a useful concept for tracing mobile pastoralism in that
many pastoral activities are economically tied to the potential of the environ-
ment, yet strategies are altered to accommodate social, political and ideological
pressures applied across those very same territories. Thus, the pastoral ‘land-
scape’ represents the amalgamation of these factors into a recoverable, and
conceptually real, spatial and temporal entity.
Nevertheless, to deny that the environment has a life of its own is to
ignore the visible ecological balance that often defines the natural context
of pastoral societies. Many times typical mobility orbits are strategically
changed by pastoralists in reaction to short-term fluctuations in the natural

environment, such as extremely wet or cold summers in alpine meadows.
In such a case upland meadows would not be grazed as usual both because
of the inclement conditions at high altitude and the greater abundance of
adequate pasture at lower elevations. The effects of this altered plan are
then passed back to the environment, as midland pastures become overused
and alpine meadows become overgrown. Thus, for each series of reciprocal
reactions, there is an anthropogenic ripple effect that lasts longer than the
immediate condition. From an archaeological perspective these elemental
changes are difficult to document. Thus our ‘graphic’ for human ecody-
namics is necessarily smoothed, and we are constrained by the average case
scenario, regardless of our knowledge that human groups are challenged to
deal with variability in the actual environment. To better understand the
reality of pastoral strategies, and the potential social implications that stem
from various modes of interaction, a landscape approach that emphasizes
temporal and spatial currents within the human-ecological sphere offers a
useful synergy between the fixed archaeological record and the patterns of
land use by populations whose resource catchments were variable in terms
of distance and accessibility over time.
Human ecodynamics are folded into the pastoral landscape, which encom-
passes the exploitation and living strategies employed by societies over time and
space, within the limits and opportunities of specific environments. Modeling
the landscape entails comparing its periodic productivity and identifying
potentially successful strategies for its exploitation. This is not to propose a
deterministic relationship between environmental productivity and human
exploitation. Societies frequently exercise their choice not to utilize certain
resources for cultural reasons or to modify their environment to suit their needs
given the available technology (Salzman 2002). Pastoral landscapes reflect many
practices that are less dependent on the environment, such as the creation of
ritual spaces, or patterns of mobility that define routes and boundaries within
the landscape not tied to environmental considerations. These spaces may
have an equally reflexive impact as the natural setting on the development of
patterned occupation of the landscape. We now turn to a concrete archaeo-
logical case study, to better understand how variation in economic and social
strategies of Bronze Age pastoralists living in the Dzhungar Mountains in
Kazakhstan contributed to the formation of a distinct cultural landscape and
set into motion wider interregional networks of interaction.
The Pastor al Archaeology of

Easter n K azakhstan
The Dzhungar Mountains Archaeology Project (DMAP) was initiated in
1999 to address the nature of Bronze Age pastoralism in one region of the
Eurasian steppe, the Semirech’ye and the Dzhungar Mountains (Figure 17.1).
The DMAP presents a comprehensive program for scientific archaeological
research concerning the economy, social organization, and structure of inter-
regional interaction of Bronze Age societies in the eastern Eurasian steppe
zone (Frachetti 2004a). The analytical approach of the DMAP draws from
landscape archaeology, which provides a conceptual framework for addressing
the distribution of archaeological data from different analytical scales while also
considering its spatial variation across a given territory (Frachetti 2006). This
approach roots archaeological interpretations in intensive studies of particular
locations within the wider distribution of sites across the landscape and justifies
them in relation to a number of concomitant factors, such as environmental
resources, topography and site-to-site correlations. Spatial analysis of the
relative location of settlements, burials and other related locales (such as rock-
art sanctuaries), along with detailed scientific analysis of the material culture
and archaeobotanical and archaeozoological remains, and the chronology of
Figure 17.1. Location of the study area in the Eurasian steppe on the border of Kazakhstan and
China.
various sites, provides a rich fabric of data at regional, local and site-specific
levels. Applying this multiscalar approach, the archaeological methodology
included archaeological survey, archaeological excavation, paleo-environmental
reconstruction and computer-assisted spatial modeling using a geographic
information system. This phase was carried out between 2002 and 2004 and
resulted in one of the first contemporary studies of Bronze Age pastoralism in
the region (Frachetti 2004b).
Our archaeological survey resulted in the discovery of more than 380 new
archaeological sites distributed throughout the study area in the Koksu River
Valley in Eastern Kazakhstan (Frachetti 2004b). The Koksu Valley was selected
for a number of reasons. First, the environment of southeast Kazakhstan varies
drastically from sandy deserts to grassy steppe-lands and alpine meadows
within a geographic extent of fewer than 100 km. This geographic variation
enabled a concise investigation of different environmental contexts within
a logistically reasonable territory and allowed for the correlation between
archaeological contexts and their corresponding environmental niches. Second,
earlier archaeological research showed that the Koksu Valley had been host to
Bronze Age societies engaged in, roughly defined, pastoralism (Mar’yashev and
Goryachev 1993; Goryachev and Mar’yashev 1998), while the ‘Dzhungarian
Gates,’ the historical name for the mountain passes through Semirech’ye, are
documented trade and travel routes in the region (Bartol’d 1943).
The survey phase of the DMAP documented a variety of site types from
different periods within the Koksu Valley, the most common being settlements,
burials and rock art, though sporadic finds and unique features were also docu-
mented. Within this data set the archaeology of the Koksu Valley dates from
the earliest find of a Neolithic flint blade core to the most recent settlements of
the past 100 years. Over 80% of the sites can be attributed to the Bronze Age
(second and early first millennia BCE). This rough chronology was assigned on
the basis of comparable formal characteristics in the construction style of sites,
as well as on the basis of datable ceramics and other archaeological materials
collected in shovel tests. The chronology was then better justified through
archaeological excavations, which produced a more accurate range of absolute
dates for these materials on the basis of radiocarbon dating.
Small-scale excavations were carried out at the Bronze Age site of Begash,
which contributed material and analytical data concerning the chronology,
domestic economy, environment, patterns of land use and burial rituals of
Bronze Age pastoral groups (Frachetti 2004b, 2006; Mar’yashev and Frachetti,
forthcoming). The excavations included a Bronze Age settlement (Begash), as
well as three burials from the nearby cemetery (Begash-2), located 350 m from
the settlement (Figure 17.2). Excavations at the settlement provided a sequence
of radiocarbon dates that illustrate the site’s long-term use throughout the
Figure 17.2. Map of the Koksu River Valley, showing the distribution of Bronze Age sites as recorded
by the Dzhungar Mountains Archaeology Project.
Bronze Age (2500–1000 BCE calibrated). They also provided archaeobotanical

and archaeozoological data, helping to formulate a preliminary picture of the
Bronze Age domestic economy in the Koksu Valley. In addition, geological
samples were analyzed to establish a local paleoclimatic sequence, which aided
in estimating the environmental carrying capacity of pasture resources during
the Bronze Age.
The excavations at the associated Bronze Age cemetery provided anthro-
pological data concerning the diet, health, behavior, and rituals of individuals
and groups in this region, as well as unique finds of bronze and gold jewelry
that, along with ceramics from the settlement excavation, provided a diverse
assemblage of artifacts, suggesting an interregional range of interaction for
this Bronze Age pastoral population. The most likely economic strategy for
societies living in the Koksu Valley during the mid to late second millennium
BCE was a vertically transhumant form of mobile pastoralism, predominantly
based in herding sheep and cattle between upland pastures in the summer
and lowland regions in the winter. This conclusion is based on four lines of
evidence (Frachetti 2004b), which will not be discussed in detail here. They
include the overwhelming percentage of domestic fauna compared with the
limited amount of wild animal remains; the vertical zonality and restrictive
nature of the mountain steppe environment in the Koksu Valley, which would
promote a strategy of vertical transhumance to support intensive sheep and
cattle herding; the provisional ‘lack of evidence’ from Bronze Age archaeo-
logical contexts for alternative subsistence strategies, such as the cultivation of
domesticated plants; and ethnohistorical evidence that the traditional economy
of the region is characterized by vertically transhumant mobile pastoralism
since the 3rd century BCE. Also significant, the archaeological stratigraphy
at the settlement of Begash presents evidence for seasonal or stochastic use
of winter settlements by mobile pastoralists as opposed to settled herders. At
Begash, this interpretation is supported by the iterated infilling shown in the
stratigraphy and formation of the site.
Such a general observation about the seasonal pastoral economy is only the
starting point for a more detailed understanding of the potential variation in
the land-use patterns and social strategies that such a transhumant lifestyle can
entail; thus one must examine more closely the variation within archaeological
and environmental data to understand the broader impact that such a way of life
can have on the cultural geography of the region. In the case of the Dzhungar
Mountains, variability in the pastoral strategy of Bronze Age populations was
tied to both environmental conditions and sociopolitical choices made on
the part of various groups or individuals. These choices and adaptations are
reflected in the diversity of archaeological contexts distributed throughout the
valley in different environmental contexts.
Variation in the Environment

It should be noted that the Bronze Age environmental reconstructions are
derived from contemporary satellite imagery, adjusted according to data
concerning the climatic and environmental changes in the study area.
Paleoclimatic studies conducted in the Dzhungar region suggest that the
climate and vegetation of the second millennium BCE was broadly comparable
with that documented today (Rhodes et al. 1996). This is a topic of debate;
paleoclimatologists working in the western steppes and northern Kazakhstan
have argued that the climate, not to be confused with the vegetation, of the
second millennium BCE was different from today’s (Kremenetski 2002).
Archaeobotanical research within the scope of the DMAP suggests that the
steppe vegetation during the second millennium BCE, at Begash, was compa-
rable to contemporary vegetation (Aubekerov et al. 2003). This conclusion is
also supported by archaeobotanical studies in the steppe zone, which argue
that in spite of climatic oscillations the general geographic distribution of
grassland vegetation in the region has remained unchanged for the past 4000
years (Khotinskiy 1984). Thus, the modern environment may serve as an

approximation of the vegetation and environmental geography experienced
by pastoralists during the Bronze Age.
Although water sources are abundant throughout the Koksu Valley, soils
are poor and generally unproductive for cultivation (Sobolev 1960). More
than 80% of the natural vegetation is classified as natural pasture; thus the
region has been effectively exploited by mobile pastoralists for millennia.
The size and productivity of pastures in the study zone is directly correlated
with variations in seasonal climatic conditions and altitude. As a general rule,
high-altitude pastures (more than 1400 m above sea level) are three to six
times more productive than pastures below 800 m above sea level during the
months of June, July and August (Frachetti 2004b). This is because of the
aridity in the lowlands during the summer, as well as high summer tempera-
tures in the valley basin. The lowland areas, however, do provide dry fodder in
the winter months and are not covered by snow, as are the highland pastures.
Therefore, the geography of pasture resources in the study zone can be recti-
fied according to known botanical horizons at different altitudes, and pasture
areas can be assigned ‘quality’ based on their ability to support herd animals
during different times of the year. According to range-productivity calcula-
tions (Frachetti 2004a), upland pastures are prospective locations for herding
during the summer, while lowland areas are more suitable for winter habitation.
To summarize the seasonal variations and vertical geography of the valley’s
resources, as they pertain to the ecodynamics of pastoral populations, we can
qualify the environment through a map of seasonal fitness from an economic
point of view (Figure 17.3). Naturally, these trends fluctuate both temporally
and spatially, meaning that some years are wetter, colder, drier or warmer, and
the locations suitable for pasturing or settlement may also change from time
to time. Thus, the variation in environmental productivity must be seen as
a dynamic factor that contributes to various choices on the part of pastoral-
ists. The impact of periodic environmental variation must also be considered
alongside social and ritual concerns; thus, we now turn to some examples of
variation in these aspects of the pastoral landscape.
Burial and Settlement Geogr aphy and For ms

One of the aims of this chapter is to examine the ways in which societies manip-
ulated and changed the local boundaries of their experienced landscape by
recasting economic, ritual, political and social experiences within the temporal
and geographic routines of their settlement and migration and through their
investments in social contexts such as burials and rock art. As Giddens (1984)
argues, structures are never static and, even though the economic, ritual and
Figure 17.3. Seasonal environmental conditions, related to pastoral activities and settlement, in the
Koksu River Valley and surrounding upland meadows.
sociopolitical landscapes of Bronze Age societies do reflect some qualities of

regularity, the inherent variation in strategies to negotiate the environmental
and social components of the landscape provided scenarios for the boundaries
and coherence of that very landscape to constantly be renegotiated. Here I will
focus on two archaeological data classes: Bronze Age cemeteries, and settlement
typology and geography across the study area.
The burial ground of Begash-2 is located 1.3 km to the northwest of the
modern village of Begash on the piedmont terrace of the Chibandy Mountains
and roughly 350 m to the northeast of the prehistoric settlement ‘Begash.’ The
cemetery is situated on a flat terrace where a small stream emerges from a steep
gorge and consists of 33 stone formations, with multiple burials, and six kurgan
mounds, likely from later periods (Rosen, this volume; Shishlina, this volume).
The Bronze Age burials themselves appear as rectangular, oval or circular stone
formations, with 1–5 stone boxlike burial cists with flat capstones inside the
stone formation (Figure 17.4). Stone cist burials represent the most common
form of burial for the Bronze Age in the region and are common to the other
known burial grounds in the Koksu Valley: Talapty, Kuigan and Begash-1
(Goryachev and Mar’yashev 2004). Like Begash-2, Talapty and Kuigan repre-
sent a large number of burials constructed at the opening of gorges and are
closely associated with substantial settlements and extensive rock art in the
Figure 17.4. Excavated Bronze Age stone cist burial at Begash-2 (Koksu River Valley, eastern
Kazakhstan).
nearby cliffs. By contrast, the site of Begash-1 is a cluster of Bronze Age stone
cist burials (Karabaspakova 1987). In their construction and material forms
the stone arrangements and cist burials at Begash-1 are similar to the other
cemeteries in the valley.
Begash-1 significantly reflects a deviation in both its overall scale and
geographic context. First, when compared to the other Bronze Age burial
groups, Begash-1 is comparatively small. The other burial grounds in the
Koksu Valley are large, ranging from 17 burials (more than 35 cists) at Talapty,
35 burials (more than 70 cists) at Begash-2 and at least 40 burials (more than 80
cists) at Kuigan. Begash-1 has roughly 10 stone arrangements, each with two
to three cists; thus it is less than half the size of its contemporary cemeteries.
A second difference is the geographic context of Begash-1. Unlike the other
burials, Begash-1 is located in an open area rather than near a ravine or gorge
and, more significant, is not associated with a nearby rock-art site or Bronze
Age settlement. From a social perspective it would seem that the group that
used Begash-1 as its burial grounds was somehow disarticulated from groups
that identified with the larger, more established burials and settlements, such
as at Begash-2. One might argue that Begash-1 and Begash-2 are not contem-
porary and simply represent two distinct periods in the mortuary history of
the valley, but from the palimpsest of Iron Age and Bronze Age burials at
Begash-2 it is clear that Begash-2 retained its ritual significance for many
centuries after the Bronze Age. The burial construction at the two cemeteries
is also comparable, which further suggests that they were contemporaneous.
Therefore, Begash-1 more likely represents the splintering off, or new arrival,
of a social group that decided to establish its own burial ground. The excava-
tions and materials from Begash-1 are not extensively published, so it is difficult
to make more detailed statements concerning the root of the variation that
may be represented by Begash-1. Regardless of the reasons for the foundation
of Begash-1, its disarticulation from other domestic and ritual contexts and
its small size represent a geographical and scalar departure from the more
common sites of Bronze Age burial grounds in the valley. It may stand as an
example of the alternative choices of different groups in the creation of the
social and ritual landscape.
In addition to variation in burial contexts there are three different settlement
types recovered in the Koksu Valley. Although all seem to have been in use during
the Bronze Age, they may reflect various aspects of the economic, social, and
political choices of Bronze Age pastoralists. These settlement forms include semi-
subterranean houses, small camps and ephemeral settlements. Of the more than
20 Bronze Age settlements recovered in the survey, 50% are classified as semi-
subterranean houses, 35% as small camps and 15% as ephemeral settlements.
Semisubterranean houses have substantial stone foundations and are multiroom
structures. These structures are generally rectangular and the exterior founda-
tion forms a large (as large as 20x20 m) polygon. In addition, individual housing
units frequently are arranged in rowlike groups of 5–10 rooms. These house
groups are located most often on the flat shoulders and river terraces of small
tributary valleys and canyons, such as in the case of the settlement at Begash.
The construction of semisubterranean houses includes a stone foundation and
wall, typically dug into the earth at a depth ranging from 0.5–1.5 m, with stone
coursing mortared with dirt and clay (Figure 17.5). By analogy with similar types
of settlements from other steppe contexts, the superstructure was likely made
of wood and grass, although there is no evidence for this from excavated settle-
ments in the Koksu Valley. Semisubterranean settlement construction is well
known in Semirech’ye from other excavated late Bronze Age settlements such as
Talapty (Goryachev and Mar’yashev 1998), as well as from excavated settlements
in central Kazakhstan (Margulan et al. 1966). A distinguishing aspect of these
substantial settlement contexts is their proximity to large rock-art sites and large
Bronze Age cemeteries. Material from excavations at Begash, and shovel tests at a
number of these settlements, includes a wide variety of domestic ceramics, burnt
and discarded animal bones, grinding stones, stone pestles and metal implements.
The detailed excavations at Begash revealed that these structures were likely dug
out and re-used on a seasonal basis (Frachetti 2004b).
‘Small camps’ are similar to semisubterranean houses in some respects, as
Figure 17.5. Reconstruction of a Bronze Age semisubterranean house in Buguly (central Kazakhstan)
comparable to those excavated in Begash (after Margulan et al. 1966).
they often show similar construction techniques, but they are different in terms
of size, general shape and geographic location. Small camps are characterized
by small stone foundations, most often circular and ranging from 4–5 m in
diameter. These smaller structures are found in groups of three to four, often
located in small ravines with steep slopes, and are frequently built on small level
terrace platforms with little surrounding area. Given their limited spatial extent,
these settlements likely serviced smaller groups or were used as short-lived
stopping camps for mobile groups. Shallow foundations (often fewer than 0.3
m) and observed thinner cultural strata detected in shovel tests, compared to
semisubterranean houses, both support this interpretation. Small camps revealed
a mixed assortment of material similar to that of the semisubterranean houses,
including handmade ceramics and animal bones. The ceramic material recov-
ered from these settlement types is typical of Bronze Age handmade pottery,
which is the primary evidence used to chronologically relate small camps to
semisubterranean houses. Although not formally excavated, based on the struc-
tural qualities and geography of the small camps, they might be interpreted as
satellite habitations or seasonal retreats for smaller groups of the resident popu-
lations of the larger settlements. An exemplary group of small camps is located
in a steep tributary canyon to the south of the Koksu River. Shovel tests within
these structures revealed ceramic fragments clearly associated with known late
Bronze Age forms. The settlement area is wedged into a highly inaccessible
ravine, although there is a year-round water source. It is likely that this settle-
ment context was not regularly used but rather was either a fail-safe location
when more regularly used settlement areas, such as Begash, were undesirable.
Alternatively, this instance of small camps might have been occupied by some
disarticulated subgroup of the society that was forced, or chose, to make its own
settlement outside the more common lowland contexts.
Another example of variation in the settlement geography of the Koksu
Valley is demonstrated by a group of small camps located in the ravines on the

western slopes of Mount Alabasy, overlooking the Mukri River. These settle-
ments are located far into the arid lowlands and are nestled into a small ravine
that cuts into the terrace plain. The small camp within this settlement group
consists of fewer than 10 structures. Unlike the other small camps discussed
above, a small group of stone-arrangement burials were constructed near this
location, but no rock art was detected. This may suggest that the population that
used the area was beginning to invest in it as a more regularly visited location,
although a major settlement of the semisubterranean type was not warranted
or affordable. Small groups may have used this location in years when settle-
ment zones were inundated by unusually deep snow or when other conditions
caused them to extend their domestic and ritual world farther into the lowlands.
These small camps are significant in that they reflect the maximum extents
to the geographic patterns that characterized the Bronze Age settlement and
suggest that either social or environmental factors, or both, could lead groups
to introduce variation into the creation of their environment.
Ephemeral settlements are the most difficult to interpret. They do not
exhibit any permanent foundations and are known only by the chemical
residue they reflect through vegetation. Ephemeral features are typically oval
or circular, about 3–5 m in diameter, and are located on the grassy shoulders
and open terraces of midland and upland elevations. Often, a nearby rectilinear
auxiliary ‘footprint,’ most likely representing an animal corral, accompanies the
oval marking (Figure 17.6). The organization of seasonal summer settlements,
in the form of yurt camps, is known from my own ethnographic documenta-
tion in the valley (Figure 17.7), as well as from broader ethnographic studies
of pastoral settlements (Andrews 1999). At this time, however, we cannot
comment scientifically about the chronological antiquity of this settlement
type in the Koksu Valley. We can at best suggest that these settlement locations
share the consistency over time that is more concretely exhibited in the other
types of Bronze Age settlement locales. The distribution of ephemeral settle-
ments illustrates that upland areas have been settled in the past, minimally on
a seasonal or transient basis, and that the kinds of structures in this ecological
niche likely consisted of nonpermanent foundations.
Different settlement types illustrate a varied pattern of settlement choices
in light of practical factors, which suggests that the ecology played a role in
the selection and revisitation of particular locales. Groups of semisubterranean
houses are located in wide ravines or on nearby wide, flat terraces; small camps
are tucked into steep gorges and occupy small tributary terraces; and ephemeral
settlements are located in upland zones on flat plains or nearby grassy pastures.
In addition, the geographic location and scale of these settlements influenced
the way the landscape was experienced in space and time, as larger or smaller
Figure 17.6. Remains of an ephemeral settlement in the Koksu River Valley, evident only by differences
in the vegetation.
Figure 17.7. Contemporary Kazakh settlement in the Koksu River Valley, showing the summer yurt
and the corral.
groups would come together to various extents at each location.

Semisubterranean houses reflect the largest and most elaborate settlements
and are located both in environmentally attractive niches and in socially
elaborated spaces, in the proximity of rock art and burials. Small camps, on
the other hand, reflect smaller-scale settlements located in areas that cannot
support large groups for extended periods of time without frequent relocation.
Finally, ephemeral settlements, taken as proxies for the kind of summer high-
land settlements that may have been constructed by Bronze Age pastoralists,
are the most transient or unfixed settlements in the landscape, likely reflecting
yearly, or monthly, choices for short-term settlement during summer migra-
tions to higher elevations. Accordingly, the dynamics of social interaction may
have been ordered in relation to the social or political identities of Bronze Age
groups as derived from the status and scale of domestic spaces. The variation
in settlement evidence illustrates that Bronze Age groups employed a number
of different habitation types, some of which were geographically permanent,
some short-lived, and some of which accommodated frequent movements
under changing environmental conditions. These various settlement contexts
also suggest dynamic social and political conditions among the valley’s popula-
tions, though more detailed investigations are necessary to illuminate those
factors more clearly. Given the geographic distribution of archaeological
sites, however, such as burials and settlements, and the location of pasture
resources and ecumenical environmental niches described above, we can now
more accurately reconstruct the patterns and extent of mobility of Bronze Age
pastoralists. This is a first step toward a better understanding of Bronze Age
sociopolitical and economic interactions.
Modeling the Landscape Dynamics

To model the dynamic engagement of Bronze Age pastoralists, with their
constructed and natural landscapes in the Koksu Valley, we must draw from
those data sources that we know are relevant to pastoral choices, keeping in
mind the likelihood that other, less recoverable, factors also affected the range
of archaeological variation presented above. Mobility, for example, is a primary
correlate to the archaeological variation of pastoral contexts in the Koksu
Valley. The data used to model pastoral mobility patterns include the locales
that we can safely assume were visited and exploited by groups and individuals,
and where we can link the location and distribution of settlements to other
archaeologically recorded social venues (burials, rock-art sanctuaries, etc.)
according to the seasonally variable productivity and geographic distribution
of natural pastures. On the basis of archaeozoological data, we know that the
primary herd animals for Bronze Age groups in the Koksu Valley were sheep
and cattle. It is fair to suggest that Bronze Age groups recognized the value in
exploiting highland pastures in the summer and the environmental protection
of lowland areas in the winter. The archaeology from our survey supports this
assumption. Pastoral mobility was at least one practice contributing to the
variation in the activation and deactivation of particular loci in the landscape
by the Bronze Age population.
Patterns of land use and mobility can be modeled using a geographic
information system. This allows for different properties to be emphasized, or
diminished, to study their role in mobility patterns while being realistically
constrained to the actual data. In this way modeling the dynamics of the Bronze
Age pastoral landscape is not left to hypothetical or arbitrary simulation.
Patterns of mobility and land use, the examples explored here, can be rooted
in the relevant variables that are recovered archaeologically or geographically.
The simulated aspect of the model lies in the way ‘value’ is assigned to various
factors, while significant correlation is achieved when independent variables
are shown to be mutually significant. In the case of the Koksu Valley, modeling
the economic considerations of Bronze Age pastoralists entails understanding
the productive capacity of the region’s pastures and their geographic proximity
to social and domestic contexts. The technical methods used to calculate these
figures are discussed elsewhere (Frachetti 2004b; Frachetti 2006).
The capacity of the region’s pastures in the height of the growing season is
high, such that the pastures located on average 20 km from midland settlements
could easily support flocks of more than 20,000 sheep over a given growing
season. Thus, the economic pressures on pastoral migration orbits would not
demand extensive mobility. In fact, if we return to the site of Begash, there
are extremely productive highland pastures within 15 km that could easily
provide ample fodder for large herds or flocks (Figure 17.8). Sites located
farther into the lowlands may not have been selected for their proximity to
summer pasture, however, but for their proximity to ritual sites. The settlement
at Talapty is located in an area more than 30 km from rich upland pastures,
but the winter settlement area is adjacent to an extensive stone ridgeline in
the Eshkiolmes foothills, which contain more than 10,000 rock-art images
commonly attributed to the Bronze Age, as well as large cemeteries. It would
seem that sometimes the choice to travel farther, and to ignore the economic
cost, was justified in terms of the social, ritual or political capital attributed
to various territories throughout the broader landscape (Figure 17.9). The
construction of a socialized and ritual landscape, demarcated by such features
as rock-art sanctuaries, cemeteries, and other socialized spaces, played a role
equally significant to that of the environment in affecting the dynamic exploi-
tation of the Koksu River Valley by Bronze Age populations. As noted above,
nearly every large settlement in the region of Begash is associated with a group
Figure 17.8. Variable land-use pathways between rich highland pasture (dark gray) and lowland
settlements, shown as points with 2-km ‘activity zones’ around them. Routes are calculated based on
the quality and productivity of the vegetation. Distance is iterated according to the size and nutritional
demand of the herd, ranging between 1000 and 40,0000 animal units to be supported annually. Annual
travel, adequate to support even the largest herds, rarely exceeds 35 km.
Figure 17.9. Variable land-use pathways between rich highland pasture (dark gray) and lowland
settlements, shown as points with 2-km ‘activity zones’ around them. Routes are calculated to account
for pasture quality, terrain (slope) and the location of significant social locales (burials, rock art, etc.).
Routes can be more than 50 km to satisfy both herd needs and socioritual practices.
of rock art and a nearby cemetery. The spatial relationship between rock-art,
burials, and settlements, coupled with the seasonal economic patterns proposed
above, allows for some initial interpretations concerning the land-use schedule
and sociopolitical interactions that resulted from group investment in specific
locations in the landscape.
The location of settlements provides information concerning the seasonal
use of the landscape, which can be tied to social and political strategies. The
Bronze Age settlements at Talapty and Kuigan are located in dry lowland
areas and likely represent winter or fall settlements, considering the lack of
productive pasture in the area during the summer and the favorable winter
conditions of the terrace on the south side of the Eshkiolmes range. Begash is
located at a slightly higher elevation (around 950 m above sea level) and, given
its situation in a protected canyon, probably represents a winter settlement.
According to the seasonal economic reconstruction above, these settlements
could have been inhabited for 3–7 months, during the late fall, winter and early
spring, and would likely have been unoccupied for about 3–6 months during
the late spring and summer, while groups migrated to highland pastures. The
existence of the complex of burials and rock art near these winter settlements
may indicate that investments in the landscape, in part, served to communicate
ownership or control over domestic locations while the population was away
at highland pastures in the summer. There is little rock art in the upland areas,
even though usable rock faces exist there as well. Since the upland pastures are
treacherously cold and uninhabited during the winter, it is unlikely that there
would be significant human traffic there except during the summer. Logically,
there would be no need to protect or mark settlement areas in the highlands. As
the nature of the summer pasture resources is much more variable from year to
year than winter conditions, marking areas of settlement might actually serve to
limit the possibilities of claiming prime locations from year to year. Most likely,
the boundaries of summer pastures and settlement zones were negotiated at the
time of migration, when pastoralists naturally came together while providing
for their herds at the limited territories of highland pastures. Although it may
have been unnecessary to mark summer settlement areas, lowland winter settle-
ment areas were accessible to any group passing through the area, so more
overt displays of control, power and status, such as symbolic invocations of
shamanism, folklore and ancestry (Mar’yashev and Goryachev 2002; Frachetti
2004b), were encoded into the landscape near choice settlements. By socializing
these areas with ritual and ideological signs, specific groups could signify their
definition of territorial boundaries and could communicate their engagement
in specific locations, even though their overall patterns of movement led to
periods when these sites were not physically occupied.
Discussion
I have argued that the archaeology of mobile pastoralism is best approached as
a patterned yet variable socio-economic strategy. The recovery of this strategy,
or strategies, can be facilitated by conceiving of the wider geographic and
temporal layouts of pastoral experience and practices within dynamic land-
scapes, as opposed to categorically rigid paradigms of economic or political
systems. The select archaeological examples from the Bronze Age of the Koksu
River Valley illustrate that pastoralists of this region were constantly renego-
tiating the character of their social and economic geography according to the
variability of their environmental and social contexts. Ultimately, archaeolo-
gists may better understand the wider impact of nomadic pastoral groups in
prehistory by documenting the variation within pastoral systems over time
rather than by viewing such societies within strict social, economic or political
frameworks.
References
Andrews, P. A.
1999 Felt Tents and Pavilions: The Nomadic Tradition and Its Interaction with Princely
Tentage. London, Kölner Ethnologische Mitteilungen Melisende.
Anschuetz, K. F., R. H. Williams and C. L. Scheick
2001 An Archaeology of Landscapes: Perspectives and Directions. Journal of
Archaeological Research 9: 2, pp. 157–211.
Ashmore, W. and A. B. Knapp
1999 Archaeologies of Landscape. Malden, Blackwell.
Aubekerov, B. Z., S. A. Nigmatova and M. D. Frachetti
2003 Geomorphological Particulars in the Region of the Archaeological Monument
Begash, Northern Dzhungar Alatau [in Russian]. In Aktual’nye Problemy
Geostistem Aridnikh Territorii. Almaty, Kazakh National University: pp.
287–289.
Bacon, E.
1954 Types of Pastoral Nomadism in Central and Southwest Asia. Southwestern
Journal of Anthropology 10: pp. 44–68.
Barfield, T. J.
1981 The Central Asian Arabs of Afghanistan: Pastoral Nomadism in Transition. Austin,
University of Texas Press.
1993 The Nomadic Alternative. Englewood Cliffs, New Jersey, Prentice Hall.
Barth, F.
1964 Nomads of South Persia. New York, Humanities Press.
1969 Ethnic Groups and Boundaries. Boston, Little, Brown.
Bartol’d, V. V.
1943 Ocherk Istorii Semirech’ia [in Russian]. Frunze, Kirgizgosizdat.
Bar-Yosef, O. and A. M. Khazanov
1992 Pastoralism in the Levant: Archaeological Materials in Anthropological Perspectives.
Madison, Prehistory Press.
Beck, L.
1991 Nomad: A Year in the Life of a Qashga’i Tribesman in Iran. Berkeley, University
of California Press.
Bradburd, D.
1990 Ambiguous Relations: Kin, Class, and Conflict Among Komachi Pastoralists.
Chernykh, E. N.
1997 Kargaly: Zabytyi Mir [in Russian]. Moskva, Institut Arkheologii (Rossiiskaia
akademiia nauk).
Cribb, R.
Danti, M. D.
2000 Early Bronze Age Settlement and Land Use in the Tell Es-sweyhat Region, Syria.
University of Pennsylvania (PhD dissertation).
Dyson-Hudson, R. and N. Dyson-Hudson
1980 Nomadic Pastoralism. Annual Review of Anthropology 9: pp. 15–61.
Erickson, C.
2000 The Lake Titicaca Basin: A Precolumbian Built Landscape. In D. Lentz (ed.),
Imperfect Balance: Landscape Transformations in the Precolumbian Andes. New
York, Columbia University Press: pp. 311–356.
Frachetti, M. D.
2004a Archaeological Explorations of Bronze Age Pastoral Societies in the Mountains
of Eastern Eurasia. Silk Road 2: 1, pp. 3–8.
2004b Bronze Age Pastoral Landscapes of Eurasia and the Nature of Social Interaction in
the Mountain Steppe Zone of Eastern Kazakhstan. University of Pennsylvania
(PhD dissertation).
2006 Digital Archaeology and the Scalar Structure of Pastoral Landscapes: Modeling
Mobile Societies of Prehistoric Central Asia. In T. Evans and P. Daly (eds.),
Digital Archaeology. London, Routledge: pp. 128–147.
Giddens, A.
1984 The Constitution of Society: Outline of the Theory of Structuration. Berkeley,
University of California Press.
Goryachev, A. A. and A. N. Mar’yashev
1998 Nouveaux Sites du Bronze Recent au Semirech’e (Kazakhstan). Paléorient 24:
1, pp. 71–80.
2004 The Bronze Age Archaeological Memorials in Semirechie. In K. M. Linduff

(ed.), Metallurgy in Ancient Eastern Eurasia from the Urals to the Yellow River.
Lewiston, Edwin Mellen Press: pp. 109–138.
Humphrey, C. and D. Sneath
1999 The End of Nomadism? Society, State, and the Environment in Inner Asia. Durham,
Duke University Press.
Ingold, T.
1993 The Temporality of the Landscape. World Archaeology 25: pp.152–174.
Irons, W.
Jones-Bley, K. and D. G. Zdanovich (eds.)
2002 Complex Societies of Central Eurasia from the 3rd to the 1st Millennium BC:
Regional Specifics in Light of Global Models. Journal of Indo-European Studies.
Monograph 46. Washington, Institute for the Study of Man.
Karabaspakova, K. M.
1987 K Voprosu o Kul’turnoi Prinalezhnosti Pamyanikov Epkhi Pozdnei Bronzy
Severo-Vostochnogo Semirech’ya i ikh Svyaz’ s Pamyatnikami Tsentral’nogo
Kazakhstana [in Russian]. In K. G. Omarova (ed.), Voprosy Periodizatsii
Arkheologicheskikh Pamyatnikov Tsentral’nogog i Severnogog Kazakhstana.
Karaganda, Karaganda University Press: pp. 90–101.
Kavoori, P. S.
1999 Pastoralism in Expansion: The Transhuming Herders of Western Rajasthan. Studies
in Social Ecology and Environmental History. New Delhi, Oxford University
Press.
Khazanov, A. M.
Khlobystina, M. D.
1973 Origins et Developpment de la Civilization de Premiere Age du Bronze dans
la Siberie de Sud’ [in Russian]. Sovetskaya Arkheologiya 1: pp. 24–38.
Khotinskiy, N. A.
1984 Holocene Vegetation History. In A. A. Velichko, H. E. Wright and C. W.
Barnosky (eds.), Late Quaternary Environments of the Soviet Union. Minneapolis,
University of Minnesota Press: pp. 179–200.
Kislenko, A. and N. Tatarintseva
1999 The Eastern Ural Steppe at the End of the Stone Age. In M. Levine, Y.
Rassamakin, A. Kislenko and T. N. Kislenko (eds.), Late Prehistoric Exploitation
of the Eurasian Steppe. Cambridge, McDonald Institute for Archaeological
Research: pp. 183–216.
Knapp, A. B. and W. Ashmore

1999 Archaeological Landscapes: Constructed, Conceptualized, and Ideational. In
W. Ashmore and A. B. Knapp (eds.), Archaeologies of Landscape: Contemporary
Perspectives. Oxford, Blackwell: pp. 1–30.
Kohler-Rollefson, I.
1992 A Model for the Development of Nomadic Pastoralism on the Transjordanian
Plateau. In O. Bar-Yosef and A. Khazanov (eds.), Pastoralism in the Levant:
Press: pp. 11–18.
Kosarev, M. F.
1984 Zapadnaia Sibir’ v Drevnosti [in Russian]. Moskva, Nauka.
Koster, H. A.
1977 The Ecology of Pastoralism in Relation to Changing Patterns of Land Use in the
Northeast Peloponnese. University of Pennsylvania (PhD dissertation).
Kremenetski, C. V.
2002 Steppe and Forest-Steppe Belt of Eurasia: Holocene Environmental History.
In M. Levine, C. Renfrew and K. Boyle (eds.), Prehistoric Steppe Adaptation
and the Horse. Cambridge, McDonald Institute for Archaeological Research:
pp. 11–28.
Lamberg-Karlovsky, C. C.
2003 Civilization, State, or Tribe? Bactria and Margiana in the Bronze Age. Review
of Archaeology 24: 1, pp. 11–19.
Lattimore, O.
1940 Inner Asian Frontiers of China. Boston, Beacon Press.
Mar’yashev, A. N. and M. Frachetti
forthcoming The Bronze Age Burials of Begash [in Russian]. Vestnik, Journal of the
Institute of Archaeology, Almaty, Kazakhstan.
Mar’yashev, A. N. and A. A. Goryachev
1993 Typological and Chronological Questions of Bronze Age Sites of Semirechye.
Rossiyaskaya Arkheologiya 1: pp. 5–20.
2002 Rock-Art of Semirech’ya. Volume 2: Updated and Completed [in Russian]. Almaty,
Fond XXI c. Press.
Margulan, A. X., K. A. Akishev, M. K. Kadirbaev and A. M. Orazbaev
1966 Ancient Cultures of Central Kazakhstan [in Russian]. Almaty, Nauka Kazakhskoi
SSSR.
McGlade, J.
1995 Archaeology and the Ecodynamics of Human Modified Landscapes. Antiquity
69: pp. 113–132.
Pastner, S.
1971 Ideological Aspects of Nomad-Sedentary Contact: A Case Study from
Southern Baluchistan. Anthropological Quarterly 44: 3, pp. 173–184.
Rhodes, T. E., F. Gasse, L. Ruifen, J.-C. Fontes, W. Keqin, P. Bertrand, E. Gilbert, F.

Melieres, P. Tucholka, W. Zhixiang and C. Zhi-Yuan
1996 A Late Pleistocene-Holocene Lacustrine Record from Lake Manas, Zunggar
(Northern Xinjiang, Western China). Palaeogeography, Palaeoclimatology,
Palaeoecology 120: pp. 105–121.
Rosen, S. A.
Negev. Antiquity 77: 298, pp. 749–760.
Salzman, P. C.
2002 Pastoral Nomads: Some General Observations Based on Research in Iran.
Journal of Anthropological Research 58: 2, pp. 245–264.
Shahrani, M. N. M.
1979 The Kirghiz and Wakhi of Afghanistan: Adaptation to Closed Frontiers. Seattle,
University of Washington Press.
Shilov, V. P.
1975 Models of Pastoral Economies in the Steppe Regions of Eurasia in the
Eneolithic and Early Bronze Ages [in Russian]. Sovetskaya Arkeologiya 1: pp.
5–16.
Sobolev, L. N.
1960 Fodder Resources of Kazakhstan [in Russian]. Moskva, Akademiya Nauk.
Spooner, B.
1973 The Cultural Ecology of Pastoral Nomads. An Addison-Wesley Module in
Anthropology Number 45. Reading, Addison-Wesley.
Stoddart, S.
2000 Landscapes from Antiquity. Cambridge, Antiquity Publications.
Tapper, R.
1979 Pasture and Politics: Economics, Conflict, and Ritual Among Shahsevan Nomads of
Northwest Iran. New York, Academic Press.
Tsalkin, V. I.
1964 Nekotorye Itogi Izucheniia Kostnykh Ostatkov Zhivotnykh iz Rskopok
Arkheologicheskikh Pamiatnikov Pozdnego Bronzovogo Veka [in Russian].
Kratkie Soobshcheniia Instituta Arkheologii 101: pp. 24–34.
Vadetskaya, E. B.
1986 Arkeologicheskie Pamyatniki v Stepyakh Srednevo Yeniseya [in Russian]. Leningrad,
Akademiya Nauk CCCP.
Ch a p t e r 1 8
Mobility and
Sedentarization in L ate
Bronze Age Syria
Jeffrey J. Szuchman
A rchaeologists who work in Mesopotamia are well aware of the chal-

lenges in the archaeology of pastoral nomadism in that region. Wilkinson
(2003:50) has pointed to the disappointing record of pastoral nomadic archae-
ology in the Near East: “Despite the large number of enthusiastic references to
the importance of the nomadic element, they continue to be underrecognized
in the survey record.” One nomadic group that has received special attention
despite its absence from the archaeological record is the Aramaean tribes of
late second millennium BCE Syro-Anatolia. These tribes were traditionally
depicted as the ambassadors of doom for the settled populations of Syro-
Anatolia at the end of the second millennium BCE. Roux (1969/1992:275), for
example, described the “barbaric Aramaeans” as “originally uncouth bedouins,”
who “contributed nothing to the civilizations of the Near East.” A few years
later, Hawkins (1982:375) explained the development of Aramaean kingdoms
as the result of “a new and intrusive population group” whose “penetration of
Syria . . . must have exerted pressure on the already settled Anatolian peoples.”
More recently, analyses of textual references to Aramaeans and archaeo-
logical surveys in northern Syria have resulted in a reconceptualization of the
Aramaean presence in Syro-Anatolia (Schwartz 1989; McClellan 1992; Sader
1992, 2000; Lipinski 2000; Akkermans and Schwartz 2003:367). This new
image of the Aramaeans differs from the old view in two important ways.
First, Aramaeans are now understood not as intrusive elements but rather
as a local Syrian pastoral people who had been active in the Syrian social and
economic sphere for some time before they turned to sedentism. Second,
the sedentarization of Aramaeans is no longer assumed to be the result of an
organized military campaign against the urban population of Syria and Turkey;
rather, Aramaeans became sedentary agriculturists during the Dark Age that
followed the demise of the Hittite and Assyrian kingdoms in the final centu-
ries of the second millennium BCE. In doing so, Aramaean tribes filled and
exploited a power vacuum that already existed in Syro-Anatolia.
397
398     J e f f r e y J . S z u c h ma n
The main evidence for the reconceptualization of Aramaean sedentariza-

tion comes from the clear continuity of material culture in areas of Aramaean
occupation and the sharp increase in Iron Age settlements in these same areas
during the transition from the Late Bronze to Early Iron Age (McClellan 1992;
Sader 1992, 2000). The cultural continuity certainly suggests that Aramaeans
had been part of the Syrian social landscape for some time before they became
sedentary and thus are unlikely to have infiltrated suddenly into northern
Syria from the desert (Figure 18.1). However, the notion that Aramaean
pastoral tribes sedentarized in Syria during a period of administrative chaos
and in the absence of any alternative centralized regime in Syro-Anatolia is
less compelling.
Ethnographies of pastoral nomads in the modern period suggest that
sedentarization primarily occurs during periods of peace and under relatively
stable political conditions rather than in periods of political upheaval (Barth
1961; Cribb 1991:61–64). A famous example of the mechanics of sedentariza-
tion shows that an increase in agricultural production will force farmers to
expand their cultivated fields into pastoral lands. The pressure now exerted
Figure 18.1. Map of border region between Anatolia and Mesopotamia (in modern Turkey, Syria and
Iraq), showing the place names mentioned in the text.
L at e B r o n z e A g e S y r i a      399
on pastoral nomads encourages sedentarization as an alternative to what has

become a risky pastoralist venture (Khazanov 1994:212–221). Such a scenario
is generally made possible only by a strong central administration that fosters
agricultural expansion at the expense of pastoralism. Based on the ethnographic
record, Aramaean sedentarization probably began either before or after the
upheavals of the late second millennium BCE, when the political conditions
of Syro-Anatolia were less volatile. Aramaeans appear to have already estab-
lished strong sedentary political dynasties in Syria and Anatolia by the time
that the Dark Age came to an end at the turn of the first millennium BCE, in
the early years of the Neo-Assyrian period (Bunnens 1999). Thus, if Aramaean
sedentarization did occur against a backdrop of political stability, it must have
begun before the ‘Crisis Years’ (Ward and Joukowsky 1989) at the close of the
Late Bronze Age. It was during this period that the Middle Assyrian kings
restored stability to Syro-Anatolia after the fall of the Mitanni kingdom in the
14th century BCE.
Were Ar amaeans Nomadic?

The problem with discussions of Aramaean sedentarization in general lies in the
available evidence for Aramaean nomadism. The sources for early Aramaean
nomadism are primarily textual and unfortunately scarce. They begin with the
description in the annals of King Tiglath-Pileser I of his campaign against the
Aramaeans in 1111 BCE:
With the support of the god Aššur, my lord, I took my chariots and warriors
(and) set off for the desert. I marched against the Ahlamu-Aramaeans, enemies
of the god Aššur, my lord. I plundered from the edge of the land Suhu to the
city Carchemish of the land Hatti in a single day. I massacred them (and) carried
back their booty, possessions, and goods without number. The rest of their
troops, who had fled from the weapons of the god Aššur, my lord, crossed the
Euphrates. I crossed the Euphrates after them on rafts (made of inflated) goat-
skin. I conquered six of their cities at the foot of Mount Bišri, burnt, razed, and
destroyed [them, and] brought their booty, possessions, and goods to my city
Aššur (Grayson 1991:23).
From this and similar statements in the annals of Tiglath-Pileser I, scholars

have assumed that, as late as the 11th century BCE, Aramaeans were primarily
mobile. Jebel Bišri lies outside the agricultural centers along the Syrian
Euphrates, and Tiglath-Pileser I does locate the target of his campaigns in the
‘desert.’ But the term he uses, mudbaru (in line A.0.87.1 45), is also used in
Neo-Assyrian texts to refer to the steppe in contrast to the cultivated country.
Whether it refers specifically to the desert or to the steppe, the term clearly
implies that Aramaeans occupied an environmental zone outside the limits of

sedentary irrigation agriculture. The use of the term itself does not, however,
rule out the possibility that Aramaeans practiced small-scale cultivation in
addition to pastoral activities.
More interesting in regard to the nomadic background of the Aramaeans is
their association with the Ahlamu of North Syria. Throughout the inscriptions
of Tiglath-Pileser I in the 12th century BCE, the term Aramaean is preceded
by the term Ahlamu. This designation appears sporadically in texts from the
second millennium BCE, and it seems to refer to a mobile population in Syria
that may have had possession of some lands (Sader 2000; Schachner 2003).
As early as the 14th century BCE, King Adad-Nirari I claimed that his father,
King Arik-Den-Ili (1319–1308 BCE), had conquered the lands of the Ahlamu
(Grayson 1987:132). In the 13th century BCE, during a campaign against
a coalition of Mitanni, Hittites and Ahlamu, King Shalmaneser I “became
ruler over their lands” and “set fire to the remainder of their cities” (Grayson
1987:184). Ahlamu also appear in second millennium BCE Babylonian sources,
which suggests that they had far-reaching contacts with urban kingdoms
throughout Mesopotamia (Brinkman 1968:278; Heimpel 2003:28). By the end
of the second millennium the term Ahlamu probably referred to one or several
mobile tribes that may have been attached to a particular territory. Although
there is little evidence to suggest that the Ahlamu were the forerunners of the
Aramaeans or that a genetic link between the two groups existed (Brinkman
1968:277–278), the association of Aramaeans with Ahlamu by Tiglath-Pileser I
may suggest that, if the two groups were not one and the same, the Aramaeans
were probably organized either politically, socially or territorially in the same
way as the second millennium BCE Ahlamu, so that the two were indistinguish-
able to the Assyrian rulers.
The association of Aramaeans with Ahlamu and the reference to the
mudbaru in the Assyrian texts suggests on the one hand that Aramaean tribes
were highly mobile. On the other hand, Tiglath-Pileser I boasts of conquering
and plundering Aramaean cities and of taking their valuable possessions as
booty to his capital Aššur. In these passages he seems to describe a sedentary
population, or at least a semisedentary group with assets rich enough to take
back to the capital, rather than a primarily mobile people. Furthermore, the
language he uses is standard in texts that describe the conquest of cities within
an established kingdom. For example, in his account of the conquest of the
lands of Katmuhhu and Nairi, to the northwest of Assyria, Tiglath-Pileser
I says, “I brought out their booty, property [and possessions]. Their cities,
I burnt, razed, [and] destroyed” (Grayson 1991:14). In a later text, Tiglath-
Pileser I emphasizes that he crossed the Euphrates 28 times in pursuit of the
Aramaeans, which also might suggest a high degree of mobility (Sader 1992,
2000). But it may also reflect the difficulty that Tiglath-Pileser I encountered
when he tried to extend his control to such a distance from the Assyrian heart-
land, where the city of Carchemish constrained Assyrian expansion. That the
Aramaeans exhibited such tenacity also indicates their fundamental attachment
to their territory, an attachment that may have led to, or been generated by,
the establishment of permanent settlements.
The first attestation of Aramaeans in the Assyrian annals, which has tradi-
tionally been interpreted as evidence of Aramaean mobility in the 11th century
BCE, suggests, in fact, just the opposite: that Aramaean sedentarization had
been underway for quite some time before the reign of Tiglath-Pileser I. It may
be best, therefore, to think of Late Bronze Age Aramaeans as tribal societies
with mobile components rather than as nomads, a term that is too imprecise in
this context. This terminology implies that Aramaean tribes also had sedentary
components, although the extent and nature of sedentary Aramaean commu-
nities in the Late Bronze Age have not yet been addressed. Indeed Middle
Assyrian kings had introduced favorable conditions for sedentarization into
Syria for over a century before Tiglath-Pileser I ascended the throne.
Ar amaeans and Assyrians in

the Late Bronze Age
Assyria’s interest in northern Syria and southeastern Turkey, the land they
called Hanigalbat, began when King Ashur-Uballit I (1365–1330 BCE) threw
off the yoke of the Mitanni kingdom around 1350 BCE. The newly indepen-
dent Assyrian king now called himself ‘Great King,’ although the territory
of Assyria remained confined to its traditional heartland and barely extended
to the west of the Tigris (Harrak 1987:57). Assyrian occupation of Syria may
have begun in the reign of King Adad-Nirari I (1307–1275 BCE). In his first
campaigns against Hanigalbat, at the beginning of the 13th century BCE,
Adad-Nirari I made the weak Mitanni king an Assyrian vassal. Later, following
a failed Mitanni revolt, Adad-Nirari I marched to Waššukanni, the Mitanni
capital, and deported the royal family to Ashur. A new Assyrian capital was built
at Taidu,1 securing his control over north Syria and southeastern Turkey from
the Tigris to the Euphrates, including the entire Khabur basin.
After the death of Adad-Nirari I, his son Shalmaneser I (1274–1245 BCE)
faced another revolt in Hanigalbat that was promptly put down. Following
the campaigns of Shalmaneser I, clear evidence of Assyrian settlement in
Taidu was most likely at, or in the vicinity of, Tell Hamidiye on the Jaghjagh River
1.
(Meijer 1986; Harrak 1987), or at Üçtepe, along the Tigris River in Turkey (Kessler
1980).
Hanigalbat appears. At Tell Brak (Oates et al. 1997:13–15), for example, the
Mitanni Palace was probably destroyed first by Adad-Nirari and later by
Shalmaneser I. Shortly after the second palace destruction, Middle Assyrian
occupation began. The earliest texts from Tell Sheikh Hamad date to the time
of Shalmaneser I, which suggests that Assyrian occupation of that site also
began during his reign (Kühne 2000). In fact, Tell al-Rimah, on the border of
the Assyrian heartland, is the only site in Hanigalbat with evidence of Middle
Assyrian occupation predating Shalmaneser I (Postgate et al. 1997:56).
Part of what made Assyrian occupation of this territory possible was the
administrative structure that Shalmaneser I and his successor, Tukulti-Ninurta
I (1244–1208 BCE), instituted. Shalmaneser I removed the king of Hanigalbat
to Assyria, along with a large number of other Hanigalbateans. At the same
time, people from the north, Assyrians, and others from within Hanigalbat
were settled in the now vacant cities of Syro-Anatolia (Machinist 1982:18–19).
The entire territory was subdivided into pahutu (provinces), each administered
by an Assyrian bel pahete (official) who reported directly to the Assyrian sukallu
rabu or šar Hanigalbat (king of Hanigalbat), who himself answered to the king
of Assyria. Under this system the Assyrian king had direct control over all of
the territories west of the Tigris, and Shalmaneser I was able to extend his
kingdom as far west as the Balikh River. This system of authority probably
remained in place throughout the Middle Assyrian period and probably as late
as the 10th century BCE. It is a testament to the success of this system that the
Neo-Assyrian kings of the first millennium BCE ruled the entire Near East via
an administrative system that was essentially the same as that implemented by
Shalmaneser I in the 13th century BCE (Machinist 1982:34).
The administration of the Assyrian provinces and the redistribution of the
Syrian population would have had a profound effect on the pastoral nomads
of Syro-Anatolia. Since the decline of the Mitanni kingdom in the 14th century
BCE, pastoral nomadic groups had not been subject to a strong central
administration. The drastic administrative and demographic restructuring
of Hanigalbat under the Middle Assyrian kings would have stripped mobile
peoples of their relative autonomy and fundamentally impacted their way of
life. New urban centers may have altered pastoralist migration routes so that
they could benefit from the wealth of the urban elite; a strong military pres-
ence would have reduced the ability of nomadic groups to raid villages and
towns; conscription may have limited their mobility; taxation and political
stability may have expanded agricultural lands into what was previously
Aramaean pasture. Irrigation agriculture and a new extensive canal irrigation
system (Kühne 1990) would certainly have expanded cultivable areas into lands
previously exploited exclusively by pastoral nomads. Any one or combination
of these factors might have encouraged the mobile peoples to settle, so that
by the time Tiglath-Pileser I encountered them, the sedentarization of the

Ahlamu-Aramaeans would have been well under way.
Models of Sedentarization
That the Ahlamu-Aramaeans became sedentary is not in itself surprising. Since
Barth’s (1961) groundbreaking work on the Basseri tribe of southwest Iran,
ethnographies of Near Eastern pastoral nomads have continued to demonstrate
not only the close relationship between nomadic and sedentary society but
also the diverse nature of nomadism as a subsistence economy (Salzman 1972;
Bates 1973; Marx 1977). In most cases Near Eastern pastoral nomads were
neither purely nomadic nor solely reliant on herding sheep and goats as their
main economic activity. Salzman (1972) labeled nomads with diverse economic
bases ‘multi-resource nomads.’ Among the Yarahmadzai tribe in Baluchistan,
for example, nomads supplement pastoral activities with date palm cultivation,
hunting and gathering, and small-scale grain cultivation, as well as raiding
(Salzman 1972; 1980a, 1980b). In southwest Iran some Basseri work as hired
laborers in village fields (Barth 1961), and a number of Qashqa’i tribe members
practice agriculture in addition to pastoralism (Beck 1986, 1991). The diverse
nature of their economies enables these multiresource nomads to settle with
little effort when pastoralism ceases to be a viable subsistence strategy.
The diversity of nomadic economies and lifestyles corresponds to a similar
diversity in the mechanics of sedentarization. Salzman (1980a, 1980b) stresses
the voluntary and temporary nature of sedentarization. Nomads move back
and forth between degrees of mobility “in response to changing pressures,
constraints, and opportunities both internal and external to the society”
(Salzman 1980a:14). For archaeologists such flexibility means that historical
events alone cannot account for periods of sedentarization or nomadization in
antiquity. At best, external factors can only suggest that sedentarization was at
least a viable option for those members of a mobile society who saw the shift
to agriculture as a necessary alternative to pastoralism.
A number of studies of modern Near Eastern nomads illustrate the variety
of sedentarization processes. Despite the diverse impulses toward sedentariza-
tion, the ethnographic record makes clear that, in several cases, sedentarization
depends on political factors. Barth (1961:118) was explicit about the fact that,
although personal wealth is the primary determinant in a Basseri family’s deci-
sion to settle, sedentarization was more likely to take place in peaceful periods
when the central administration was strong. In periods of weak administration
a switch to nomadism became much more common. Similarly, the economic
effects of land policies instituted by a strong Turkish government led to the
sedentarization of Yörük nomads (Bates 1973, 1980). As the government
encouraged large-scale cotton, wheat and rice farming, irrigation and cultiva-
tion expanded in rural Turkey. This policy reduced the pasturage available to
nomads and forced them into a sedentary agricultural lifestyle. Although the
presence of a strong central administration does not necessarily lead to seden-
tarization, the process is certainly more likely to take place against a backdrop
of political and economic stability.
A second important archaeological implication of ethnographies of sedenta-
rization is that in most cases, recently settled nomads retain a distinct identity
from their sedentary neighbors. This distinction remains, in part, because in
order to allow for a return to nomadism if conditions change, the cultural iden-
tity of nomads must, to some degree, remain unique. Indeed, for the Basseri,
it appeared important to retain some tribal connections as sedentism leads to
increased birthrates, and shares of inheritance are not always enough to allow
a landowner’s children to remain sedentary (Barth 1961). In cases when Basseri
settled as a group, tribesmen continued to identify as Basseri and maintained
an identity distinct from the villagers (Barth 1961:117). Maintenance of a tribal
identity in the case of Yörük sedentarization in Turkey was directly expressed
in their material culture (Bates 1973). Permanent dwellings were arranged
according to a pattern adapted from the distribution of tents within a campsite
and organized according to two branches of one patrilineage. In some cases
Yörük families remained mobile but pitched tents as home bases in Yörük
village quarters when they grazed their flocks on Yörük fields in the summer.
Bates (1973:222) found that, on the whole, although some social changes took
place after Yörük settlement, the new adaptations were economic rather than
a “massive change in formal institutions or social rules.”
Settled Qashqa’i nomads serve as another example of the retention of tribal
identity despite a change in residence and subsistence patterns. Sedentary
Qashqa’i are as integral a part of the tribe as are nomadic Qashqa’i (Beck
1986, 1991). Sedentarization does not change any of the political affiliations or
social relations within the tribe (Beck 1986:185). In some cases tribal cultural
affiliation is both tangibly experienced and visually expressed by constructing
permanent dwellings over earlier tent sites (Beck 2003:293). Beck (2003:293)
describes one striking example of sedentary Qashqa’i tribal identity: a black
tent pitched against a permanent dwelling, where a son and his new bride lived
and guests were entertained.
The anthropology of sedentarization in the 20th century CE has substantial
implications for the archaeology of early Aramaeans. The fact that sedenta-
rization of modern nomads most often occurs when a central administration is
politically and economically stable suggests that the same may have been the
case for mobile Aramaeans. Mass sedentarization of Aramaean tribes is there-
fore unlikely to have occurred during the Dark Age that followed the reign of
King Tiglath-Pileser I. Rather, the conditions for successful agriculture and

sedentarization had been established much earlier, as Middle Assyrian kings
brought political stability and economic control into Hanigalbat. Accordingly,
it is in the 14th–13th centuries BCE that mobile pastoralists would have been
most inclined to become sedentary. Moreover, because tribal identity is often
maintained after sedentarization, and may be expressed in the material culture
of sedentarized nomads, Aramaean settlements in Late Bronze Age Syria and
Turkey are likely to exhibit a material culture similar to that of their mobile
kinsmen but unlike that of their sedentary neighbors.
Toward an Archaeology of Sedentarization

Although the literature on nomadism in ancient Mesopotamia recognizes the
continuum of mobility and the fluid boundary between nomadic and sedentary
lifestyles (Adams 1974; Rowton 1974; Nissen 1980; Zagarell 1989; Finkelstein
and Perevolotsky 1990; Cribb 1991; Khazanov 1994), archaeologists have
tended to concentrate on one extreme of this continuum and often draw a bold
line between sedentism and nomadism in practice (Hole 1980; Cribb 1991;
Alizadeh 2003). The result is that, despite a clear awareness of the variations
and nuances of pastoral nomadic systems and sedentary-nomadic interaction,
the archaeology of nomadism in Mesopotamia is limited to identifying the
ancient ‘campsite’ or ‘tentsite.’ It is no wonder, then, that the best representa-
tive of excavated nomadic sites in Mesopotamia remains Tepe Tula’i (Hole
1974), although its interpretation as a nomadic campsite is disputed (Wheeler
Pires-Ferreira 1975). Such controversy over the interpretation of possible
nomadic sites is not unique to Tepe Tula’i. The faunal remains of Early Bronze
Age Sos Höyük and Büyüktepe Höyük (Howell-Meurs 2001a, 2001b) suggest
that the inhabitants of those sites practiced sedentary, rather than nomadic,
pastoralism as the excavators originally concluded based on architectural
context (Sagona et al. 1996:37).
By searching for only nomadic or sedentary settlements, an opportunity
is missed to widen the archaeological perspective to encompass transitional
sites or to focus archaeological investigations of ancient nomadism toward the
center of the spectrum, to sites of sedentarizing nomads. As they sedentarize,
nomads begin to use durable construction materials and to accumulate preserv-
able debris (Cribb 1991; Wendrich, this volume), but they also retain strong
cultural connections to their tribes. Thus, the remains of sedentarizing nomads
will be more visible than, but not necessarily distinct from, that of their mobile
counterparts. Sites of sedentarizing nomads can potentially offer a great deal
of information about how their occupants interacted with agricultural village
or urban communities. An archaeology of sedentarization can begin to address
Wilkinson’s (2003:50) lament over the lack of progress in the archaeology of

pastoral nomadism in Mesopotamia. In fact, Wilkinson and Tucker’s own
survey (1995) of the North Jazira, in Iraq, provides an illustration of how such
a methodology might work to determine the extent of Aramaean sedentariza-
tion in the Late Bronze Age.
In four seasons, from 1986 to 1990, 497 square km in northwest Iraq were
surveyed by Wilkinson and Tucker. The survey region consisted of a shallow
basin in the North Jazira, which is drained by small wadis (seasonal rivers) in
the north and south that empty into the Wadi al-Mur, which itself flows south-
east to empty into the Tigris. The basin is flanked by rolling hills to the west,
northeast and southeast. In most years dry farming is possible on the North
Jazira plain, though the flat wadi basin has a higher agricultural yield than the
surrounding hills (Wilkinson and Tucker 1995:7). Wilkinson and Tucker divide
the second millennium BCE in the North Jazira into three historical periods:
the Khabur period (2000–1500 BCE), identified by a range of Khabur wares;
the Mitanni period (1500–1300 BCE), identified by Nuzi ware; and the Middle
Assyrian period (1300–1000 BCE), based on characteristic Middle Assyrian
pottery (Wilkinson and Tucker 1995:iv, 59). To address the difficulties iden-
tifying Nuzi ware and distinguish between Middle and Late Assyrian wares,
Wilkinson and Tucker take Khabur wares as generally representative of the first
half of the second millennium BCE and Middle Assyrian wares as generally
representative of the second half of the second millennium BCE (Wilkinson
and Tucker 1995:59). Thus, they effectively collapse the tripartite historical
division of the second millennium BCE into two broad archaeological periods
based on ceramic indicators: a ‘Khabur period’ and a ‘Middle Assyrian period.’
This may well be the best solution, as Khabur ware has since been found in
later Mitanni levels at Rimah (Postgate et al. 1997:54). Moreover, Pfälzner
(1995), in his analysis of ceramics from Tell Sheikh Hamad, has confirmed that
Mitanni wares give way to Middle Assyrian types beginning in the 13th century
BCE. One problem with this scheme is that maps of the second half of the Late
Bronze Age may conflate the effects on settlement patterns of two significant
political developments: the decline of the Mitanni kingdom, beginning in the
14th century BCE, and the subsequent growth of the Middle Assyrian kingdom.
Thus, the overall thinning of the number of settlements in this period prob-
ably reflects the ruralization, perhaps accompanied by an increase in pastoral
nomadism, that followed the collapse of the Mitanni kingdom and not the
expansion of Middle Assyrian control.
Further complicating the settlement picture in this period is the fact that
any increase in settlement that may have occurred during the Middle Assyrian
period is dwarfed in comparison to the dramatic settlement growth that took
place in the first millennium BCE (32 new settlements). Wilkinson and Tucker
(1995:62) conclude that at least some of these new Late Assyrian settlements
may have been occupied by sedentarized tribesmen, who they suggest turned
to agriculture in this period. A close look at the changes in settlement between
the Khabur and Middle Assyrian periods, however, shows that sedentarization
may have begun as early as the Middle Assyrian period. In the Khabur period
in the North Jazira, several new settlements appear along wadis or as satellites
of large urban centers. These locations are appropriate for an agricultural
economic base, and the small sites that emerge associated with large centers
may indicate a general growth in urbanization in the first half of the second
millennium BCE. In the Middle Assyrian period, by contrast, fewer new settle-
ments are established, compared with the preceding period. Whereas in the
Khabur period 25 new sites appear, in the Middle Assyrian period there are
only seven new sites. Like the new sites of the Khabur period, some of these
new Middle Assyrian sites occur along wadis, which is consistent with primarily
agricultural activities. Some of these new sites, however, are established at some
distance from the wadi beds and urban centers of the region, a trend that is new
in the Middle Assyrian period. Two Middle Assyrian period sites (numbers 69
and 157) are located on the western and eastern slopes of the large wadi basin,
on land that is not as productive as the flat terrain on the floor of the basin. A
third site (number 105) is located in an area that was entirely devoid of settle-
ment in the preceding period. Whereas in the Khabur period settlement was
concentrated in the northern and southern ends of the survey region, these
three new Middle Assyrian settlements fill the space between these two poles.
Furthermore, all three of these sites are quite small, site 157 being 1.2 ha and
sites 69 and 105 only 0.8 ha. Might these new Middle Assyrian period sites
be home to sedentarizing pastoral nomads, transitioning from temporary to
permanent settlements?
Discussion
As the Middle Assyrian regime introduced the political stability that made
sedentarization feasible, along with the economic stability that intensive irri-
gation agriculture demanded, the pastoral nomads of the region would have
pursued agriculture more intensively. As the pastoral activities that had required
seasonal campsites gave way to year-round cultivation, these temporary settle-
ments may have taken on more permanent fixtures and, by the end of the Late
Bronze Age, may already have become productive agricultural communities
integrated economically and politically into the provincial landscape of the
Middle Assyrian kingdom. In the Late Assyrian period, settlement continued
in this central area, and the whole region became more densely occupied.
Sedentarization of nomads certainly may have played some role in the Iron
Age settlement increase (Wilkinson and Tucker 1995:62), but this should not
be understood as a sudden and explosive event. The cursory re-evaluation
presented here of the North Jazira Survey shows that Aramaean sedentariza-
tion began in the Middle Assyrian period at the end of the Late Bronze Age.
A much more detailed study of the abundant surveys of Late Bronze Age
Syro-Anatolia will be required to confirm these suggestions, but the aim of
this chapter has been to advocate for an archaeology of sedentarization as a
way to approach the survey data. The disappointing record of the archaeology
of pastoral nomadism in Mesopotamia stems in part from the fact that the
implications of the anthropology of sedentarization have largely been ignored
by archaeologists in practice, if not in theory. In the case of early Aramaean
nomads, Middle Assyrian occupation of Syro-Anatolia created the political
and economic conditions that tend to encourage sedentarization. Once we
move away from a false nomadic-sedentary dichotomy in terms of settlement
types by investigating sites of sedentarizing nomads, the survey record appears
to correspond with those anthropological implications. The archaeology of
sedentarization may be one way to confront the challenges of pastoral nomadic
archaeology in Mesopotamia.
References
Adams, R. M.
1974 The Mesopotamian Social Landscape: A View from the Frontier. In C. B.
Moore (ed.), Reconstructing Complex Societies: Supplement to Bulletin of the
American Schools of Oriental Research 20: pp. 1–13.
Akkermans, P. M. M. G. and G. Schwartz
2003 The Archaeology of Syria. Cambridge, Cambridge University Press.
Alizadeh, A.
2003 Some Observations Based on the Nomadic Character of Fars Prehistoric
Cultural Development. In N. F. Miller and K. Abdi (eds.), Yeki Bud, Yeki
Nabud: Essays on the Archaeology of Iran in Honor of William M. Sumner. Los
Angeles, Cotsen Institute of Archaeology: pp. 83–97.
Barth, F.
1961 Nomads of South Persia. Boston, Little, Brown.
Bates, D.
1973 Nomads and Farmers: A Study of the Yörük of Southeastern Turkey. Anthropological
papers of the Museum of Anthropology 52. Ann Arbor, University of Michigan.
1980 Yoruk Settlement in Southeast Turkey. In P. C. Salzman (ed.), When Nomads
Settle: Processes of Sedentarization as Adaptation and Response. New York, Praeger:
pp. 124–139.
Beck, L.
1986 The Qashqa’i of Iran. New Haven, Yale University Press.
1991 Nomad: A Year in the Life of a Qashqa’i Tribesman in Iran. Berkeley, University
2003 Qashqa’i Nomadic Pastoralists and Their Use of Land. In N. F. Miller and
K. Abdi (eds.), Yeki Bud, Yeki Nabud: Essays on the Archaeology of Iran in Honor
of William M. Sumner. Los Angeles, Cotsen Institute of Archaeology: pp.
289–304.
Brinkman, J. A.
1968 A Political History of Post-Kassite Babylonia (1158–722 B.C.). Rome, Pontificium
Institutum Biblicum.
Bunnens, G.
1999 Aramaeans, Hittites, and Assyrians in the Upper Euphrates Valley. In G. del
Olmo Lete and J.-L. Montero Fenollós (eds.), Archaeology of the Upper Syrian
Euphrates, the Tishrin Dam Area. Barcelona, Institut del Pròxim Orient Antic:
605–624.
Cribb, R.
Finkelstein, I. and A. Perevolotsky
1990 Processes of Sedentarization and Nomadization in the History of Sinai and the
Negev. Bulletin of the American Schools of Oriental Research 279: pp. 67–88.
Grayson, A. K.
1987 Assyrian Rulers of the Third and Second Millennia BC. Volume 1 (to 1115 BC).
Toronto, University of Toronto Press.
1991 Assyrian Rulers of the Early First Millennium BC. Volume 2 (1114–859 BC).
Toronto, University of Toronto Press.
Harrak, A.
1987 Assyria and Hanigalbat: A Historical Reconstruction of Bilateral Relations from the
Middle of the Fourteenth to the End of the Twelfth Centuries B.C. Hildesheim,
Georg Olms.
Hawkins, J. D.
1982 The Neo-Hittite States in Syria and Anatolia. In J. Boardman (ed.), Cambridge
Ancient History. Volume 3, Part 2. Cambridge, Cambridge University Press:
pp. 372–441.
Heimpel, W.
2003 Letters to the King of Mari: A New Translation with Historical Introduction, Notes,
and Commentary. Winona Lake, Eisenbrauns.
Hole, F.
1974 Tepe Tula’i: An Early Campsite in Khuzistan. Paléorient 2: pp. 219–242.
1980 The Prehistory of Herding: Some Suggestions for Ethnography. In M. T.
Barrelet. L’Archéologie de l’Iraq: Du début de l’époque néolithique à 333 avant notre
ère: Perspectives et limites de l’interprétation anthropologique des documents. Paris,

Editions du Centre National de la Recherche Scientifique: pp. 119–128.
Howell-Meurs, S.
2001a Archaeozoological Evidence for Pastoral Systems and Herd Mobility: The
Remains from Sos Höyük and Büyüktepe Höyük. International Journal of
Osteoarchaeology 11: pp. 321–328.
2001b Early Bronze and Iron Age Animal Exploitation in Northeastern Anatolia: The
Faunal Remains from Sos Höyük and Büyüktepe Höyük. BAR International Series
945. Oxford, Archaeopress.
Kessler, K.
1980 Untersuchungen zur historischen Topographie Nordmesopotamiens: Nach keilschrift-
lichen Quellen des 1. Jahrtausends v. Chr. Beihefte zum Tübinger Atlas des Vorderen
Orients B 26. Weisbaden, Ludwig Reichert.
Khazanov, A. M. (translated by Julia Crookenden)
1994 Nomads and the Outside World. University of Wisconsin Press.
Kühne, H.
1990 The Effects of Irrigation Agriculture: Bronze and Iron Age Habitation Along
the Khabur, Eastern Syria. In S. Bottema, G. Entjes-Nieborg and W. van
Zeist (eds.), Man’s Role in the Shaping of the Eastern Mediterranean Landscape.
Rotterdam, A. A. Balkema: pp. 15–30.
2000 Dur-katlimmu and the Middle Assyrian Empire. In O. Rouault and M. Wäfler
(eds.), La Djéziré et l’Euphrate syriens de la protohistoire à la fin du IIe millénaire
av. J.-C. Subartu 7. Turnhout, Brepols: pp. 271–279.
Lipinski, E.
2000 The Aramaeans: Their Ancient History, Culture, Religion. Leuven, Peeters.
Machinist, P.
1982 Provincial Governance in Middle Assyria and Some New Texts from Yale.
Assur 3: pp. 1–37.
Marx, E.
1977 The Tribe as a Unit of Subsistence: Nomadic Pastoralism in the Middle East.
American Anthropologist 79: pp. 343–363.
McClellan, T. C.
1992 The 12th Century BC in Syria: Comments on H. Sader’s Paper. In W. A.
Ward and M. S. Joukowsky (eds.), The Crisis Years: The 12th Century B.C.:
From Beyond the Danube to the Tigris. Dubuque, Kendall-Hunt: pp. 164–173.
Meijer, D.
1986 A Survey in Northeastern Syria. Istanbul, Nederlands Historisch-Archaeologisch
Instituut te Istanbul.
Nissen, H.
1980 The Mobility Between Settled and Non-settled in Early Babylonia: Theory
and Evidence. In M. T. Barrelet. L’Archéologie de l’Iraq: Du Début de l’époque
Néolithique à 333 avant Notre Ère. Paris, Editions du Centre National de la

Recherche Scientifique: pp. 286–290.
Oates, D., J. Oates and H. McDonald
1997 Excavations at Tell Brak. Volume 1: The Mitanni and Old Babylonian Periods.
Cambridge, McDonald Institute for Archaeological Research.
Pfälzner, P.
1995 Mittanische und mittelassyrische Keramik: Eine chronologische, funktionale, und
produktionsökonomische Analyse. Berichte der Ausgrabunki Tall Šeh Hamad /
Dur-Katlimmu. Band 3. Berlin, Dietrich Reimer Verlag.
Postgate, C., D. Oates and J. Oates
1997 The Excavations at Tell al-Rimah: The Pottery. Wiltshire, British School of
Archaeology in Iraq.
Roux, G.
1969/1992 Ancient Iraq. London, Penguin.
Rowton, M. B.
1974 Enclosed Nomadism. Journal of the Economic and Social History of the Orient
17: pp. 1–30.
Sader, H.
1992 The 12th Century BC in Syria: The Problem of the Rise of the Aramaeans.
In W. A. Ward and M. S. Joukowsky (eds.). The Crisis Years: The 12th Century
B.C.: From Beyond the Danube to the Tigris. Dubuque, Kendall-Hunt: pp.
157–163.
2000 Aramaean Kingdoms of Syria: Origin and Formation Processes. In G. Bunnens
(ed.), Essays on Syria in the Iron Age. Ancient Near Eastern Studies Supplement 7.
Louvain, Peeters Press: pp. 61–76.
Sagona, A., M. Erkmen, C. Sagona and I. Thomas
1996 Excavations at Sos Höyük, 1995: Second Preliminary Report. Anatolian Studies
46: pp. 27–52.
Salzman, P. C.
1980a Introduction: Processes of Sedentarization as Adaptation and Response. In P.
C. Salzman (ed.), When Nomads Settle: Processes of Sedentarization as Adaptation
and Response. New York, Praeger: pp. 1–19.
1980b Processes of Sedentarization Among the Nomads of Baluchistan. In P. C.
Salzman (ed.), When Nomads Settle: Processes of Sedentarization as Adaptation
and Response. New York, Praeger: pp. 95–110.
Schachner, A.
2003 From the Bronze to the Iron Age: Identifying Changes in the Upper Tigris
Region. The Case of Giricano. In B. Fischer, H. Genz, É. Jean and K. Köroğlu
(eds.), Identifying Changes: Proceedings of the International Workshop “The
Transition from Bronze to Iron Age in Anatolia and Its Neighbouring Regions.”
Istanbul, 8–9 November 2002. Istanbul, Türk Eskiçag Bilimleri Enstitüsü.
Schwartz, G. M.
1989 The Origins of the Aramaeans in Syria and Northern Mesopotamia: Research
Problems and Potential Strategies. In O. M. C. Haex, H. H. Curvers and P.
M. M. G. Akkermans (eds.), To the Euphrates and Beyond: Archaeological Studies
in Honour of Maurits N. van Loon. Rotterdam, A. A. Balkema.
Ward, W. and M. S. Joukowsky (eds.)
1989 The Crisis Years: The 12th Century B.C.: From Beyond the Danube to the Tigris.
Dubuque, Kendall-Hunt.
Wheeler Pires-Ferreira, J.
1975 Tepe Tûlâ’i: Faunal Remains from an Early Campsite in Khuzistan, Iran.
Wilkinson, T. J.
2003 Archaeological Survey and Long-Term Population Trends in Upper Meso
potamia and Iran. In N. F. Miller and K. Abdi (eds.), Yeki Bud, Yeki Nabud:
Essays on the Archaeology of Iran in Honor of William M. Sumner. Los Angeles,
Cotsen Institute of Archaeology: pp. 39–51.
Wilkinson, T. J. and D. J. Tucker
1995 Settlement Development in the North Jezira, Iraq. Iraq Archaeological Reports 3.
Wiltshire, British School of Archaeology in Iraq.
Zagarell, A.
1989 Pastoralism and the Early State in Greater Mesopotamia. In C. C. Lamberg-
Karlovsky (ed.), Archaeological Thought in America. Cambridge, Cambridge
University Press.
Ch a p t e r 1 9
Suggestions for a Chaîne

Opératoire of Nomadic
Pottery Sherds
Hans Barnard1
A mong the more enigmatic, and sometimes controversial, aspects of

a nomadic, mobile lifestyle are the desire and the ability of people that
have not (yet) settled to manufacture, or even use, ceramic vessels. Current,
well-balanced opinions on the subject of pottery production by mobile people
are reflected in several other contributions to this volume.
There has been a common wisdom or stereotype among many archaeologists
that sedentism, agriculture and pottery technologies are necessarily positively
correlated. Indeed, some archaeologists use the presence or absence of pottery in
the archaeological record as an independent measure of residential mobility. The
presence of potsherds at a site would indicate sedentism and a lack of pottery some
degree of seasonal transhumance. Although examples of pottery in sites occupied by
mobile hunter-gatherers are known . . . in such cases the pottery is usually defined
as ‘crude’ and ‘technologically unimpressive’ thereby relegating it to a lesser or
unimportant status and reinforcing the stereotype (Eerkens, this volume).
1.
I would like to thank David Verity for sharing some of his vast experience with
me; Steve Sidebotham, Roberta Tomber, Pamela Rose, Anwar Abdel-Magid, Richard
Pierce, Knut Krzywinski, Eugen Strouhal, Jana Součková, Jitka Barochová, Manfred
Bietak, Elfriede Reiser-Haslauer and Roswitha Egner for making the necessary material
available; John Bintliff and Steve Rosen for their stimulating remarks; Jelmer Eerkers
and Paul Nicholson for their comments on earlier versions of this chapter; Anna
Barnard-van der Nat for her financial and logistical support; and Willeke Wendrich for
her unending encouragement. The license to excavate the site of Tabot was issued by
the Department of Antiquities and National Museums, Khartoum, Sudan, to Dr. Anwar
Abdel-Magid. Test excavations carried out by the license holder in 1994 and 1995 were
sponsored by the Committee for Development Research and Education (NUFU) of
the Norwegian Council of Universities within the framework of their Sudan Program
(Phase II: Archaeology Project). The license holder authorized Hans Barnard to study
and publish the pottery from Tabot.
413
414    H a n s B a r n a r d
There seems to be consensus, among the authors contributing to this volume,

that a fraction of mobile people will produce their own pottery and that this
pottery will usually be of poor quality.
These observations suggest that even when mobile groups do manufacture pottery,
their product is technologically and aesthetically inferior to those produced by
sedentary peoples. Arnold (1985), while suggesting that less than a third of mobile
societies make and use pottery, argues that a number of practical, logistical and
economic (economies of scale) problems are involved in the production of pottery
by groups with high residential mobility (Alizadeh, this volume).
The relative low quality and inferior aesthetics of the vessels produced by
mobile people usually extends to a lack of decoration of the vessels. This adds
to the difficulty of identification, in an archaeological context, if remains of
such vessels are found at all.
Pastoral people are often less concerned with decoration on their vessels than
are agriculturalists, who imbue their pots with symbolic significance . . . The
ceramics that seem to have the strongest association with the earliest pastoralists
are usually small, thin-walled vessels (10 cm in height), possibly used for sheep
milking, which break up into tiny pieces, making them even more difficult to find
on small open sites (A. B. Smith, this volume).
Over time, the character of residential mobility may have changed, slowly
losing some of the properties of hunting-gathering or herding-gathering,
including the production of pottery, and growing more dependent on the
surrounding settled communities.
It is difficult to measure this directly, but to measure trade, for example, petro-
graphic analyses of Early Bronze Age ceramic assemblages from Negev nomad
sites shows high proportions of local pottery produced by the nomads themselves,
in addition to some imports. In contrast, by classical times (Nabatean, Roman,
Byzantine and Early Islamic), 90% or more of the pottery found on the nomad
sites was imported, produced by specialists in the towns and sites of the empires
of Late Antiquity. By recent and subrecent times, virtually all pottery used by
the Bedouin, the famous black Gaza Ware, is imported, along with a vast array
of other manufactured goods (Rosen, this volume).
Despite these intricacies, the pottery used and left behind by mobile people
can, in specific circumstances, be recognized and used to show, and even date,
their presence.
Sherd scatters of coarse gray pottery identified as ‘Gaza Ware’ have been found at
many abandoned Bedouin tent camps in the Negev and Sinai. Such sherd scatters
Chaîne Opératoire of N o ma d i c P o tt e r y      415
at archaeological sites in the Negev Highlands constitute the only evidence that
these settlements were re-used by Bedouin . . . Dating Gaza Ware is problematic
because it has not been the subject of systematic typological and chronological
research. Current studies indicate that the production of Gaza Ware may have
begun as early as the second half of the 17th century CE, or the beginning of
the 18th century CE, and that it continued to be manufactured until the end of
the 20th century CE . . . This suggests that the abandoned Bedouin tent camps
situated in the Negev Highlands should be dated between the 17th‑20th centuries
CE (Saidel, this volume).
In this chapter I will deal with the production aspect of a corpus of ceramic
vessels, Eastern Desert Ware, that does not fully fit the above characteristics.
Although these vessels are believed to have been handmade by pastoral nomads
in the arid landscape between the Nile and the Red Sea, during the 4th-6th
centuries CE, most vessels are not only rather well finished but also distinc-
tively decorated (Figure 19.1). This makes it possible to identify even small
sherds of Eastern Desert Ware among many sherds of ‘imported’ wheel‑thrown
Figure 19.1. Examples of Eastern Desert Ware, believed to have been made by pastoral nomads, during
the 4th-6th centuries CE, in southeastern Egypt and northeastern Sudan (Figure 19.2). EDW 17 is from
Berenike (on the Red Sea coast); EDW 234 is from Wadi Sikait (in the Mons Smaragdus area); KHM
76918 is from Sayala (in the Nile Valley, courtesy of the Kunsthistorisches Museum in Vienna); and P 840
is from Wadi Qitna (just west of the Nile Valley, courtesy of the Náprstek Museum in Prague).
416    H a n s B a r n a r d
vessels. This special case may indicate that the production of pottery less
recognizable than Eastern Desert Ware may have been more common among
mobile groups than is usually suggested.
Given the dearth of additional archaeological finds and the ambiguous
historical sources (Burstein, this volume; Barnard 2005) Eastern Desert Ware
may remain our only source of information about its producers, the dwellers
of the Eastern Desert (Figure 19.2). Next to a careful study of the ancient
artifacts, currently undertaken from the macroscopic to the molecular levels
(Barnard and Strouhal 2004; Barnard et al. 2006), additional insights may be
acquired through experimental and ethno-archaeological studies (Shepard
1976; Rye 1981; P. J. Arnold 1991; Arnold et al. 1991; Stark 1991; Gosselain
1992; Longacre and Stark 1994; Schiffer et al. 1994; Rice 1996; Kramer 1997;
Deal 1998; Arthur 2002).2 At present the inhabitants of the area do not produce
any pottery, rendering an ethno-archaeological study pointless. Therefore, this
chapter reports on my experimental work trying to reproduce vessels similar
to Eastern Desert Ware, in a setting as close as possible to that of the ancient
pastoral nomads. Based on these experiments, I put forward a chaîne opératoire
(operational sequence) with an archaeological rather than an ethnographic
perspective (Table 19.1). Archaeologists typically deal with sherds rather than
vessels and usually do not attribute separate meanings to sherds or whole
vessels. The chaîne opératoire that I suggest, therefore, aims to explore the
processes resulting in the sherds of handmade burnished and decorated vessels
recently found in the Egyptian and Sudanese Eastern Deserts. In no way is it
meant as proof that Eastern Desert Ware vessels were indeed made by pastoral
nomads rather than by settled desert dwellers or inhabitants of the Nile Valley;
it serves only to show that it would have been feasible for them to do so.
Easter n Desert Ware

During the first survey and excavation season at the Greco-Roman harbor
Berenike, on the Egyptian Red Sea coast, in 1994, a number of remarkable
potsherds were found (Figure 19.1). These were of handmade cups and bowls
with burnished surfaces and incised decorations (Rose 1995). The closest
parallels for this pottery are described at sites in the Nile Valley, most notably
in Kalabsha, Wadi Qitna and Sayala (Kromer 1967; Bedawi 1976; Strouhal
1984:157–177; Barnard and Strouhal 2004; Barnard and Magid 2006; Barnard
et al. 2006; Barnard and Rose, in press), a considerable distance across an arid
landscape to the west. Since then, similar sherds have been recognized at sites
An excellent overview of the recent literature on these related, and sometimes
2.
confused, subjects can be found in Stark 2003.

in southeastern Egypt and northeastern Sudan (Luft et al. 2004; Nordström

2004; Sidebotham et al. 2005; Barnard and Magid 2006; Barnard, in press;
Barnard and Rose, in press), always in small quantities among many sherds of
late-Roman (Byzantine) Egyptian or Meroitic (X-group) vessels (Figure 19.2).
These, combined with a few other datable finds (like coins) and radiocarbon
Figure 19.2. Map of the border area between Egypt and Sudan, showing the places where Eastern
Desert Ware has been described.
418    H a n s B a r n a r d
Table 19.1. A Possible Chaîne Opératoire for Archaeologically Recovered Sherds of Eastern Desert
Ware:
Phase Tools Skilla Timeb

Vessel as concept
obtaining raw materials (clay, temper, water, fuel) receptacles (shovel, axe) + ++
preparation of the paste (sieving, mixing, levigating?, receptacles, sieve? ++ ++
drying)
shaping the vessel (coiling, pinching) none + +
surface treatment (wiping) and decoration (impressing) cloth, pointed tool + +
drying until leather hard none - +
Vessel as creation
surface treatment (smoothing) and decoration (incising) abrasive, blade ++ +
drying until bone dry none - ++
decoration (slipping) and surface treatment (burnishing) brush, slip, pebble, oil? ++ +
heating, prefiring none ++ +
firing, refiring saggar? ++ +
Vessel as object
first use, seasoning sealant + +
intended use none - +++
re-use (for instance as grave gift) none - +++
Vessel as tool
breaking of the vessel none - +
repair and re-use or utilization of (some of) the sherds drill, thread, adhesive? +/++ +
discarding the remains of the vessel none - +++
a
Skill levels: - = no skills required; + = limited skills required; ++ = expert skills required.
b
Time estimates: + = phase may take 0–6 hours; ++ = phase may take 0.5–7 days; +++ = phase may last
years.
dates (Strouhal 1984:265; Sadr et al. 1995:227; Magid 2004:157–159), allow

the conclusion that these vessels must have been produced between at least the
end of the third and the beginning of the 8th century CE. Given its distribution
the corpus is now identified as Eastern Desert Ware (Barnard 2002).
The majority of Eastern Desert Ware vessels are made of an orange to
rusty-red fabric with few organic but abundant poorly sorted mineral inclu-
sions (identified in petrographic thin sections as angular quartz and feldspars).
Macroscopic and microscopic inspection of this fabric places it outside the
‘Vienna System’ that classifies the common clay sources used for pottery
in Ancient Egypt (Arnold and Bourriau 1993). The technology, shape and
decoration of the vessels also make it unlikely that they were produced by the
permanent inhabitants of the Nile Valley as the pottery is very different from
that usually encountered. Preliminary interpretation of the elemental composi-
tion of more than 140 Eastern Desert Ware sherds, obtained by laser ablation
inductively coupled plasma mass spectrometry (LA-ICP-MS),3 indicate that

these vessels were made in several geologically different areas, all most likely
outside the Nile Valley. A comprehensive discussion of the origin of Eastern
Desert Ware falls outside the scope of this chapter, but based on the technical
research summarized above, the distribution of the finds (Figure 19.2), and
the fact that the Eastern Desert at the time was inhabited by pastoral nomads
(Burstein, this volume; Magid, this volume; Barnard 2005), as it is today
(Magid, this volume; Wendrich, this volume; Murray 1935; Paul 1954), it is
now assumed that Eastern Desert Ware was made and used by the pastoral
nomads roaming the area in the 4th-6th century CE (Rose 1995; Barnard 2002,
2005). When used, most likely as serving vessels as suggested by their shape
and size, the vessels, so different from those used by the settled people in the
region, must have also acted as cultural, and possibly even ethnic, markers (S.
T. Smith, this volume). The identification of these ancient nomads is elusive
because of the lack of material remains other than the pottery and because
historical sources on the area are both scarce and ambiguous for this period
(Burstein, this volume; Barnard 2005). Several explanations can be proposed for
the occurrence of such small numbers of remarkable potsherds, always mixed
with large numbers of sherds from imported vessels, in such a large area. These
explanations range from a demand-driven production, by settled or traveling
professionals, to a household production taking place where and when the need
arose or the opportunity presented itself. The validity of the latter explana-
tion depends on whether it is feasible for mobile people to produce pottery
as well made and finished as Eastern Desert Ware. The aim of this chapter
is to investigate this feasibility for which I tried to reproduce Eastern Desert
Ware, giving particular attention to the problems and possibilities likely to be
encountered by mobile people.
Experimental Pottery Production

The pastoral nomads in the region today, the Ababda and the Bisharyyin,4
do not produce any pottery but instead use imported ceramic cups and
coffeemakers (djabana) alongside metal and plastic containers (Wendrich, this
3.
This research was done on the GBC Optimass Orthogonal Time-of-Flight
ICP-MS, with attached New Wave LUV Laser Ablation System, owned by the Institute
for Integrated Research in Materials, Environments, and Society (IIRMES) at California
State University, Long Beach, and sponsored by Dr. Hector Neff (IIRMES) and the
Cotsen Institute of Archaeology at UCLA.
4.
Both groups claim to be among the many tribes of the Beja (Paul 1954), along with
the Hadendowa (Magid, this volume) and the Beni Amer.
420    H a n s B a r n a r d
volume). Modern pottery most like Eastern Desert Ware, in technology and
appearance, is at present regularly manufactured in the southwestern United
States and northwest Mexico (LeFree 1975; Bell 1994; Wisner 1999).5 For a
while I therefore joined longtime amateur potter David Verity in his endeavor
to master the ceramic techniques that are most famously practiced in Mata
Ortiz (Chihuahua, Mexico). Using this experience, and data obtained in some
more experimental settings, I suggest the following archaeologically focused
chaîne opératoire for the sherds of Eastern Desert Ware (Table 19.1),6 assuming
the presence of the necessary tools, skills, and time (Shepard 1976; Rye 1981;
Bourriau et al. 2000). Not all sherds or archaeological vessels, which are the
end products of this sequence, necessarily see all phases or go through the
phases in the exact order as given, as will be explained below. Both in Arabic,
the language currently spoken in the area where Eastern Desert Ware is found,
and in English, clay means different things to geologists, potters and archaeolo-
gists; this is even more so for temper (Rice 1987; Hertz and Garrison 1998). A
prospective nomadic potter in a familiar landscape, however, will need to see
how the raw materials present themselves only once or twice to be able to find
and recognize them. In the desert (Figure 18.3), clay and materials that can be
utilized as inorganic temper are usually found in the same area, and they are
frequently naturally mixed in adequate proportions. As these areas are often
devoid of water and fuel, it will be necessary to either bring those or to carry
out clay and temper. Settled potters typically do the latter (P. J. Arnold 1991;
Gosselain 1992; Deal 1998; Wisner 1999), while mobile people may do the
former as they will usually be carrying water and fuel for other purposes.
Although clay and potential temper are relatively easy to recognize, their
behavior when shaped and fired is impossible to predict. It is therefore very
likely that mobile people, like their settled counterparts, returned to sources
that had proven to yield good raw materials, or at least raw materials with
known properties (P. J. Arnold 1991; Arnold et al. 1991; Gosselain 1992; Deal
1998; Wisner 1999). They may have included such valued sources in their
routes, as they will almost certainly have done with sources of special supplies,
such as temper rich in mica or clay suitable for slips. After collecting clay, water
and temper, these are combined into a paste that can be shaped and fired.
One way of doing this is to break up the raw clay, which will be a mixture of
5.
The closest local, but ancient, parallels are the vessels of the C-Horizon, produced
by Nubian groups that inhabited the Nile Valley from the first to the third cataract
between 2300–1500 BCE.
6.
A chaîne opératoire is typically used as a way to describe the production process of
a class of objects. I use it here as one of many tools to study the sherds or complete
vessels of Eastern Desert Ware found in archaeological context.
Figure 19.3. Maps of southern California and northern Egypt, indicating the places where clay and
(inorganic) temper were collected for the experiments described in this chapter.
422    H a n s B a r n a r d
clay, silt and other inclusions, and suspend the actual clay particles in water.
As these particles are very small, by definition,7 they will remain suspended
for hours while larger particles will quickly sink to the bottom. After a while
the suspension is carefully decanted into a second receptacle through a sieve,
if available, or cloth and allowed to settle. As this technique requires much
time and water, naturally levigated clay sources (at the bottom of dry lakes or
ponds) would have been preferred and may have been used exclusively. Pure
clay, however, is not suitable for the production of pottery because it will shrink
dramatically while drying, causing cracks and breaks. Therefore, not all inclu-
sions should be removed from the raw clay, or some ‘filler’ should be added to
the paste. These nonplastic materials will be cemented by the clay particles,
forming a network that will sufficiently reduce, but not completely eliminate,
shrinkage. Many materials can be used as filler, including silt,8 dung, volcanic
ash, chopped straw, crushed shells or pottery (grog), each with its own effect
on both the technological process and the appearance of the vessel (Shepard
1976; Rye 1981; Arnold et al. 1991; Schiffer et al. 1994; Bourriau et al. 2000).
I chose to add about one part of silt (Figure 19.4), by volume, to four parts
of clay and allowed them to settle together. It is unclear whether some of the
inclusions in Eastern Desert Ware were added in a similar way, as necessary in
Santa Clara, New Mexico (LeFree 1975), or if they were naturally present, as
in Mata Ortiz, Mexico (Wisner 1999). Experience would have shown which
sources naturally produced clay and temper in a favorable ratio. These would
have attracted mobile people to return, especially as such raw material required
much less preparation.
After clay and silt have settled, which may take several days, the water is
removed and the paste dried to a workable plasticity. This is best done on a
slab of plaster of paris (Figure 19.4), or by wrapping the paste in cloth, to slow
down the drying process, but can also be done in the sand and probably even
on the move. Once the paste is fit to be shaped, which can again take several
days, it can be modeled into the desired form by connecting two, or more,
rings to a base (coiling) or by pinching a ball into the right shape (Figure 19.5).
The surfaces can be smoothed with a wet finger, or a damp wad of cloth, and
decorations can be impressed into the wet surface. Impression can be made
with, among many other things, a fingernail, a blade, a shell, an animal bone
or a potter’s ‘comb’ (Shepard 1976; Rye 1981). Experiments have shown that
many of the decorations on Eastern Desert Ware were made with thorns of a
7.
Depending on the scientific context, clay is defined as platy particles weathered to
smaller than a 2–4 µm (0.002–0.004 mm) diameter.
8.
Again depending on the context, silt is defined as particles weathered to a diameter
between 2–4 µm (0.002–0.004 mm) and 0.05 mm.
Figure 19.4. Preparation of potter’s clay (by David Verity).

424    H a n s B a r n a r d
date palm (Phoenix dactilifera), which are the underdeveloped leaves at the base
of a palm frond (Barnard, in press). In the Nile Valley, palm fronds are often
used for fuel, and their thorns would have been readily available to potters. In
the desert, where there are no palm trees, they may have been imported. The
shaping and decorating of a vessel takes 30–60 min, after which the vessel is
allowed to dry until it is ‘leather hard:’ when the paste has lost its plasticity
but still holds 20–30% free water (Shepard 1976; Rye 1981; Bourriau et
al. 2000).
Once a vessel is leather hard, which may take several hours, its shape can no
longer be changed, but small repairs and additional decorations can be made
(LeFree 1975; Shepard 1976; Rye 1981). It is also possible to further smooth
the surface with a piece of damp cloth or leather (Bell 1994:53). It is likely that
the potters who made Eastern Desert Ware did either of these, but I worked
on my vessels only in the plastic and in the ‘bone dry’ stages. This latter stage
is reached after all the free water has evaporated from the fabric. This can
take several days but may be accelerated by carefully warming the vessel, for
instance by placing it in the sun and regularly turning it around (Shepard 1976;
Gosselain 1992; Kramer 1997; Deal 1998; Wisner 1999; Bourriau et al. 2000).
Between the leather-hard and the bone-dry stages the vessel is susceptible to
damage from handling and spontaneous cracking. Until firing, the production
process can be completely reversed by adding enough water. Minor repairs can
be made, or the paste can be completely recycled. It also means that the drying
vessel must be shielded from water and sweat. When bone dry, the vessel can be
smoothed, with sandpaper or another abrasive, slipped and burnished (Figure
19.5). If they did not do so when the vessel was still leather hard, the potters
working on Eastern Desert Ware may have smoothed their bone-dry vessels
with sand or an abrasive stone, like pumice or vesicular basalt, a technique remi-
niscent of burnishing. Slipping is the application of a thin suspension of clay
with a distinctive color (naturally or because of an added pigment) by pouring
or brushing this on the desired areas (Shepard 1976:67–69). Clays that make a
good slip, bonding securely while delivering a bright color, are rare and would
have been collected when encountered, or even warranted a detour, and carried
around until needed. Burnishing is the polishing of a vessel by rubbing it with a
hard object, like a pebble or the back of a spoon, after wetting the surface with
slip, water or oil. The frequent combination of the slipping and burnishing of
Eastern Desert Ware makes it likely that these were joined actions. The high
luster of many Eastern Desert Ware vessels indicates that these were fired at
relatively low temperatures (below 750–800°C or 1400–1500°F) as such luster
tends to fade on exposure to higher temperatures.
The next stage of pottery production, the firing of the clay vessel, is
the shortest and most dramatic. When the paste reaches sufficiently high
Figure 19.5. The shaping and smoothing of vessels.

426    H a n s B a r n a r d
temperatures, it will mature: the clay minerals irreversibly lose their ability
to turn back into a plastic paste. Another important effect, depending on
the conditions during the firing, is the burning off (oxidation) or deposition
(reduction) of carbon, which greatly influences the color of the vessel (Rye
1981:114–118). At higher temperatures the iron oxides in the clay can also be
reduced or oxidized, changing between black and red, respectively (Bourriau et
al. 2000). If Eastern Desert Ware was indeed produced by nomadic potters, it
was most likely fired in an open fire, unless space and time were negotiated in
kilns belonging to settled potters in the Nile Valley. As is apparent from their
results, several undesirable effects of firing vessels in an open fire were evidently
circumvented by the nomadic potters. An important difference between a kiln
and an open fire is that the temperature in a kiln can be better controlled and
reach a higher maximum. As the water bonded to the clay minerals needs to
be driven out gently, to prevent blistering, cracking or even exploding of the
vessel, an open fire needs to be carefully monitored. The choice of fuel can
facilitate this: dry wood will burn swift and hot, while animal dung or charcoal
will take longer to heat up. At present, the pastoral nomads in the area use
camel dung for cooking when wood is scarce and burn charcoal where wood is
plentiful. The former is possible since the spread of the camel in Egypt during
the last centuries BCE, and I saw the latter still done in the late 1990s. The
dung of donkeys or sheep and goats may also have been used. A method to
prevent the vessels from being destroyed by the firing is to force most of the
water out first, by heating the vessel to a moderate temperature for a prolonged
period. One or two summer days in the desert sun may be hot enough to do so
(Shepard 1976; Wisner 1999). Alternatively, the vessels may have been buried
in heated sand, as is currently the way that the nomads in the area bake their
unleavened bread (Wendrich, this volume).
Another crucial difference between an open fire and a kiln is the contact
between the vessel and the fire (flames and ashes). In a kiln the fire is sepa-
rated from the vessels, which are heated by the hot gasses released by the fire.
Allowing more or less air into the kiln generates an oxidizing or a reducing
environment, respectively. Placing a clay vessel directly into an open fire allows
the flames to create color differences on the fired surfaces, leaving so-called ‘fire
clouds,’ while the collapsing fire will create a reducing environment. This will
induce the paste to take up carbon, released by burning organics outside and
inside the fabric, turning the vessel black. As most Eastern Desert Ware vessels
do not show fire clouds or reduction, the ancient nomadic potters must have
found ways to prevent them from appearing. The simplest way to reverse some
of the fire clouds and most of the reduction is to take the hot vessel out of the
fire, before it collapses, and allow it to cool in the open air (Rye 1981). A better
method is to protect the vessel with a ‘saggar’ or quemador (Wisner 1999). A
Figure 19.6. Saggars used in California (above, photograph by W.Z. Wendrich) and in Egypt (below,
photograph by the author).
428    H a n s B a r n a r d
saggar can be interpreted as a very simple kiln. It consists of a metal or ceramic

container, holding the clay vessel, which is placed in or on top of the fire. I have
successfully used an upturned terracotta flowerpot, an old paint drum (with
a few holes to secure an oxidizing environment) and a perforated cookie tin
(Figure 19.6). Ancient potters could have separated their vessels from the fire
with larger vessels, such as cooking vessels, or may have constructed ad hoc
saggars with the sherds of broken vessels or slabs of stone (LeFree 1975). Like
the receptacles used for the preparation of the paste, such items would have
been relatively easy to clean and to re-employed for their original function.
For this study I made about a dozen vessels and fired them using a variety
of techniques. Put directly into an open fire, vessels did not survive, and vessels
buried below an open fire did not mature (they failed to lose their ability to
suspend in water). Clearly the temperature must be raised slowly, to gently
drive out the water and reach a maximum above 360°C or 680°F (Figure 19.7).
The vessels fired in a saggar in a slowly started open fire did mature without
cracking or reducing but lost some of their luster. The same was true for vessels
fired in the controlled environment of an electric kiln set to switch off at 866°C
or 1591°F (using pyrometric cone 012). The optimal temperature for Eastern
Figure 19.7. Temperature curves of three experimental firings, the first representing the theoretical
temperatures inside an electric kiln controlled by a kiln-sitter (with pyrometric cone 012), on 27 June
2003; the second showing the measured temperatures below a slow-started bonfire failing to mature three
buried vessels, on 11 August 2003; the third showing the measured temperatures in the center of a fast-
started bonfire destroying not previously heated vessels, on 4 June 2005. Measurements were performed
with an Omega Type K thermocouple (kindly made available by Dr. Brian Damiata, University of
California, Riverside), which has a Chromel (nickel-chromium) positive and an Alumel (nickel-aluminum)
negative lead, attached to a RadioShack digital multimeter. A plastic thermos flask with melting ice (0ºC
= 32ºF) was used as external reference.
Desert Ware must therefore have been 410–810°C (770–1490°F) but probably
closer to the latter (Shepard 1976; Rye 1981; Gosselain 1992; Schiffer 1994;
Bourriau et al. 2000).
A vessel placed on top of a small fire of 2 kg (about 4.5 lb) charcoal, kindled
with kerosene (paraffin) after the vessel was in place, also matured without
cracking (Figure 19.8). This setup appeared to allow enough oxygen to reach
the vessel to prevent absorption of reduced carbon released by the fuel. In
places where olive oil was applied to the vessel, the surplus of carbon could
not be oxidized, leaving a black surface (Figure 19.9). This carbon was later
removed, turning the surface reddish brown, by firing the vessel again in a fully
oxidizing environment. Some fire clouds remained, however, especially on
the bottom of the vessel where it had been in direct contact with the glowing
embers.
The resulting unglazed earthenware, quite similar to Eastern Desert Ware,
will be more or less porous, especially when new or if used relatively little. To
reduce this property, such vessels are often ‘seasoned’ by heating, for instance,
milk, oil, butter or honey (with beeswax?) in a new vessel. This saturates its
walls and diminishes the permeability of the fabric.9 Resin or bitumen may
also have been used to this effect or to repair broken vessels (Eerkens 2002).
Little is known about the intended use of the vessels, but given their size and
shape, they most likely functioned as serving vessels. At the same time, they
probably acted as cultural, or ethnic, markers as they are markedly different, in
technology and appearance, from the vessels of the settled population in and
around the region (Sidebotham et al. 2002; Luft et al. 2004).
Many Eastern Desert Ware vessels were recovered from graves, where
they had been placed as grave goods for the deceased, to whom they probably
belonged during life (Habachi 1967; Strouhal 1984). A very similar custom is
still practiced by the pastoral nomads now living in the area, despite centu-
ries of Christian and Islamic discouragement (Barnard 1998; Wendrich, this
volume). Sherds of many other vessels were found in private and public build-
ings in settlements in the Nile Valley and in the Eastern Desert. It takes little
skill and time to break a vessel, but this should not always be interpreted as an
accident. The intentional breaking of vessels can be part of a rite de passage or
be an expression of joy or mourning, for instance during a Jewish wedding, a
Greek dance, or in the context of the ancient Egyptian ritual ‘breaking the red
pots,’ where the breaking was usually followed, and sometimes replaced, by
the burial of the vessels (Ritner 1993:144–153). There are no indications that
9.
This technique is still widely practiced in modern Egypt and should be taken into
account during the analysis and interpretation of ancient organic residues (Bourriau et
al. 2000:128; Barnard et al. 2007).
430    H a n s B a r n a r d
Figure 19.8. Firing vessels in California (above, photograph by W.Z. Wendrich) and in Egypt (below,
photograph by the author).
Figure 19.9. Examples of the vessels produced in California (above) and in Egypt (below).
432    H a n s B a r n a r d
Eastern Desert Ware was ever deliberately broken, but given the importance
evidently attached to the vessels, it certainly cannot be ruled out.
There are indications that attempts were made to repair broken Eastern
Desert Ware vessels, not by using an adhesive (such as a resin or bitumen) but
rather by ‘stitching’ the sherds by threading holes drilled along the breaks,
a technique commonly used (Eerkens, this volume). Eight of the 290 sherds
that I studied in detail preserved a small hole, some of which may have been
intended to suspend the vessel; others may be repair holes. No traces of adhe-
sives, wire or string were ever seen. As the vessels must have been highly valued,
it is hardly surprising that attempts were made to extend their functional life
as long as possible, albeit possibly for a different task. After a broken vessel
was judged beyond repair, its sherds may have been used as cover, scraper, toy,
gaming piece or as a surface to receive writing (ostrakon). Finally, the remains
of the vessel may have been crushed to serve as temper for a vessel still to be
made (grog), or they may simply have been discarded to be studied by archae-
ologists centuries later.
Discussion
There is sufficient archaeological and experimental evidence to conclude that
the production of pottery, including vessels as nice as Eastern Desert Ware, by
mobile people, and certainly by pastoral nomads, is eminently possible. Apart
from a substantial investment in time and resources, however, ceramic vessels
are relatively heavy, fragile and uneconomical to produce in small numbers
(Alizadeh, this volume; Eerkens, this volume). The discussion of whether
mobile people could produce pottery, therefore, is replaced by the question of
what would make them decide to do so. The value of the specific properties of
ceramics, for the completion of certain tasks, is clearly appreciated by mobile
people worldwide, as is apparent from their propensity to use vessels acquired
from settled outsiders. Among the few things that the nomads currently
roaming the area where Eastern Desert Ware is found will always carry with
them are the essentials for their ‘coffee ceremony,’ including a terracotta
coffeemaker (made by settled potters in Sudan) and several small porcelain
cups (industrially made in China). These fragile, but not very heavy, ceramics
are stored in custom-made containers of basketry, leather or wood to prevent
breakage (Wendrich, this volume).
One reason for mobile groups to produce their own pottery may be a limited
availability of imported vessels. Settled potters, or other possible sources of
pottery, may be far away or difficult to reach. These will likely be associated
with other cultural, religious, or ethnic groups, which may obstruct the neces-
sary contacts, and the available material can be of limited practical use or too
expensive, in terms of currency, barter or otherwise. Any of these factors may

prompt members of the mobile group to produce some of their own pottery.
Modern developments have decreased the need to produce pottery, replacing
it with metal or plastic containers and providing access to ceramics produced in
places as distant as China. During the 4th-6th centuries CE there was certainly
no shortage of imported ceramic vessels in the Eastern Desert. The number of
vessels that passed through as containers for trade items can only be guessed,
but an abundance of vessels is found at the ancient harbors, road stations, mines
and quarries throughout the region (Sidebotham et al. 2001, 2002). Some may
have been intended for long-distance trade but remained behind because they
were damaged or appropriated; others would have contained supplies for those
temporarily working and living in the desert. It seems unlikely that the pastoral
nomads could not somehow have obtained a sufficient number of vessels from
this copious source, and they must have had other, and obviously important,
reasons to produce their own.
Mobile potters can try to emulate vessels produced by settled potters, espe-
cially when they are already using their imported products, or develop their
own type of vessels. The technology will be similar, but nomadic potters will
have to develop their own techniques to adjust to their specific environment
and needs. Such adaptations are likely to differ from place to place, according
to the local situation and the availability of the necessary materials. Features
that are not essential, like micaceous temper or red slip, can be omitted, and
decorations preferably made with the thorn of a date palm can be made with the
thorn of an acacia instead. Several acacia species, such as Acacia nilotica and A.
raddiana, occur in the region where Eastern Desert Ware is found, and several
Eastern Desert Ware vessels appear to have been decorated with a round tool.
This has left less distinctive marks than the triangular thorn of the date palm
but may well have been another thorn, for instance one of the Acacia species
mentioned above.
The apparent use of date palm thorns on many Eastern Desert Ware vessels is
remarkable. As palm fronds are often used for fuel by potters in the Nile Valley,
their triangular thorns are readily available to them.10 In the desert there are very
few palm trees, and the use of palm thorns seems therefore indicative of produc-
tion in the Nile Valley. However, this is not concurrent with the origin of the
Fronds of the doam palm (Hyphaene thebaica), rather than those of the date palm
10.
(Phoenix dactylifera), can also be used for fuel. Doam palms were more abundant in the
Nile Valley in ancient times; they have been slowly replaced by date palms because of
climatic changes and human intervention. Doam palm fronds carry real thorns on their
stem, reminiscent of a rose or bramble branch, with an oval rather than a triangular
cross-section.
434    H a n s B a r n a r d
clay matrix, outside the Nile Valley, as suggested by petrographic and chemical
analysis of the fabric. Furthermore, the use of palm thorns to apply incised or
impressed decorations on pottery is rarely attested for vessels originating in the
Nile Valley after the C-Horizon (2300–1500 BCE). Like Eastern Desert Ware,
vessels of the C-Horizon are handmade, partly burnished and decorated with
incised decorations, but there is no evidence to suggest continuous production
during the 1800 years that separate them, nor of a revival of the C-Horizon
culture after Lower Nubia had been under more or less long-lasting Egyptian,
Napatan, Meroitic and Roman influence. We must therefore assume that either
the clay for Eastern Desert Ware was brought into the Nile Valley, where the
vessels were subsequently made and fired, or that palm thorns were taken from
the Nile Valley into the desert, to be used for the decoration of Eastern Desert
Ware and probably primarily other chores. Many of the current inhabitants of
the Eastern Desert live, at least part of the time, in dwellings made of rugs and
mats over a dome-shaped wooden frame (Magid, this volume; Wendrich, this
volume). These mats are made of palm leafs (from Phoenix dactylifera or Hyphaene
thebaica), to which inhabitants obviously have access, held together by wooden
pegs not unlike date palm thorns. Such dwellings appear to be mentioned in
Egyptian Middle Kingdom and Late Kingdom texts, while Strabo reports, in
the 1st century CE, that the nomads in the desert live in dwellings made of
interwoven split pieces of palm leaves (Magid, this volume).
In the period during which Eastern Desert Ware was produced, the 4th–6th
century CE, there was a substantial influx of outsiders into the arid landscape
between the Nile and the Red Sea. A network of trade routes connected the
Mediterranean Basin and the Nile Valley with the Red Sea coast, Arabia,
sub-Saharan Africa and India (Sidebotham and Wendrich 1996; Wendrich
et al. 2006). Next to these transient traders, the Eastern Desert was more
permanently inhabited by numerous quarrymen, miners and early Christian
hermits (Sidebotham et al. 2001, 2002, 2005). The resulting infrastructure
of settlements, tracks and supplies, not equaled until the development of the
Red Sea coast for tourism in the 1990s, allowed the pastoral nomads to settle
temporarily when they accepted employment as laborers, guards, guides or
prostitutes. Even more fleeting contacts, including those with a hostile nature,
must have introduced the indigenous inhabitants of the Eastern Desert to the
pottery of the more recent immigrants. Both groups would have benefited
from the large volume (attested by the quantity of recovered sherds) of pottery
imported into the region. That the mobile inhabitants of the Eastern Desert
apparently chose this period of relative plenty to produce their own pottery
may be attributed to the following three points.
Being in the same place for a longer period than they probably would have
been previously may have enabled the pastoral nomads to see the pottery
production process, as reflected in Table 19.1, through for the first time
(Eerkens, this volume). The infrastructure that allowed this would also have
provided them with the necessary surplus of water and fuel.
Some of the immigrants (traders, miners and quarrymen) may have given
more or less detailed instructions, suggestions or inspiration to the nomadic
potters. Despite the possibility that they were educated by outsiders, the
nomadic potters decided to create their own corpus rather than imitate
imported vessels. This decision seems to have been based on their desire to
separate themselves from the more recently arrived inhabitants of the Eastern
Desert. The growing number of immigrants and their increasing influence,
partly fueled by the changing politics of the Roman Empire toward ethnic and
cultural minorities, would have increased this need, which sometimes gave rise
to violent confrontations (Eide et al. 1998; Barnard 2005).
Finally, sherds of Eastern Desert Ware will necessarily be concentrated near
the settlements of its producers and users. The more significant settlements,
like those associated with mines, quarries and harbors, will attract the attention
of archaeologists prior to the ephemeral campsites of pastoral nomads. Many
isolated sherds of Eastern Desert Ware vessels may lay scattered unobserved
over the vast stretches of arid landscape between the places where it has so far
been found (Figure 19.2; A. B. Smith, this volume). These could date from
periods well before and after the time of contact between the indigenous
nomads and the immigrant traders, miners and quarrymen. More research,
including a more systematic survey of the area, will be necessary to understand
the relation between Eastern Desert Ware and the pastoral nomads of the
Eastern Desert between the 4th‑6th centuries CE.
References
Arnold, D. and J. Bourriau (eds.)
1993 An Introduction to Ancient Egyptian Pottery. Deutsches Archäologisches Institut
Abteilung Kairo Sonderschrift 17. Mainz am Rhein, Verlag Philip von Zabern.
Arnold, D. E.
1985 Ceramic Theory and Cultural Process. Cambridge, Cambridge University
Press.
Arnold, D. E., H. A. Neff and R. L. Bishop
1991 Compositional Analysis and “Sources” of Pottery: An Ethnoarchaeological
Approach. Journal of Archaeological Method and Theory 7: pp. 333–375.
Arnold, P. J.
1991 Domestic Ceramic Production and Spatial Organization: A Mexican Case Study in
Ethnoarchaeology. Cambridge, Cambridge University Press.
436    H a n s B a r n a r d
Arthur, J. W.
2002 Pottery Use-Alteration as an Indicator of Socioeconomic Status: An
Ethnoarchaeological Study of the Gamo of Ethiopia. Journal of Archaeological
Method and Theory 9: pp. 331–355.
Barnard, H.
1998 Human Bones and Burials. In S. E. Sidebotham and W. Z. Wendrich (eds.),
Berenike 1996: Report of the 1996 Excavations at Berenike (Egyptian Red Sea
Coast) and the Survey of the Eastern Desert. Leiden, Research School CNWS:
pp. 389–401.
Data. In J. C. M. Starkey (ed.), People of the Red Sea: Proceedings of Red Sea
Project II. British Archaeological Reports International Series 1395. Oxford,
Archaeopress: pp.23–40.
in press Eastern Desert Ware from Marsa Nakari and Wadi Sikait. Journal of the
American Research Center in Egypt.
Barnard, H., S. H. Ambrose, D. E. Beehr, M. D. Forster, R. E. Lanehart, M. E.
Malainey, R. E. Parr, M. Rider, C. Solazzo and R. M. Yohe II
2007 Mixed Results of Seven Methods for Organic Residue Analysis Applied to
One Vessel with the Residue of a Known Foodstuff. Journal of Archaeological
Science 34: pp. 28–37.
Barnard, H., A. N. Dooley and K. F. Faull
2006 New Data on the Eastern Desert Ware from Sayala (Lower Nubia) in the
Kunsthistorisches Museum, Vienna. Ägypten und Levante 15: pp. 49–64.
Barnard, H. and A. A. Magid
2006 Eastern Desert Ware from Tabot (Sudan): More Links to the North. Archéologie
du Nil Moyen 10.: pp. 15-34.
Barnard, H. and P. J. Rose
in press Eastern Desert Ware from Berenike and Kab Marfu’a. In S. E. Sidebotham
and W. Z. Wendrich (eds.), Berenike, 1999–2000: Report of the 1999 and 2000
Excavations in Berenike, Siket, and Wadi Kalalat and the Survey of the Egyptian
Eastern Desert, Including the Beryl Mines in Wadi Sikait. Los Angeles, Cotsen
Institute of Archaeology.
Barnard, H. and E. Strouhal
2004 Wadi Qitna Revisited. Annals of the Náprstek Museum Prague 25: pp. 29–55.
Bedawi, F.A.
1976 Die Römische Gräberfelder von Sayala-Nubien. Berichte des Österreichischen
Nationalkomitees der UNESCO-Aktion für die Rettung der Nubischen Altertümer
VI. Philosophisch-Historische Klasse Denkschriften 126. Vienna, Verlag der
Österreichischen Akademie der Wissenschaften.
Bell, J.
1994 Making Pottery at Mata Ortiz. Kiva: Journal of Southwestern Anthropology and
History 60: 1, pp. 33–70.
Bourriau, J. D., P. T. Nicholson and P. J. Rose
2000 Pottery. In P. T. Nicholson and I. N. Shaw (eds.), Ancient Egyptian Materials
and Technology. Cambridge, Cambridge University Press: pp. 121–147.
Deal, M.
1998 Pottery Ethnoarchaeology in the Central Maya Highlands. Salt Lake City,
University of Utah Press.
Eerkens, J. W.
2002 The Preservation and Identification of Piñon Resins by GC-MS in Pottery
from the Eastern Great Basin. Archaeometry 44: 1, pp. 95–105.
Eide, T., T. Hägg, R. H. Pierce and L. Török
1998 Fontes Historiae Nubiorum: Textual Sources for the History of the Middle Nile
Region Between the Eighth Century BC and the Sixth Century AD. Volume 3:
From the First to the Sixth Century AD. University of Bergen, Department of
Greek, Latin, and Egyptology.
Gosselain, O. P.
1992 Technology and Style: Potters and Pottery Among Bafia of Cameroon. Man
(New Series) 27: pp. 559–586.
Habachi, L.
1967 Tongefässe. In H. Ricke (ed.), Ausgrabungen von Khor-Dehmit bis Bet El-Wali.
The University of Chicago Oriental Institute Nubian Expedition Volume 2. Chicago,
University of Chicago Press: pp. 46–68.
Hertz, N. and E. G. Garrison
1998 Geological Methods for Archaeology. New York, Oxford University Press.
Kramer, C.
1997 Pottery in Rajasthan: Ethnoarchaeology in Two Indian Cities. Washington,
Kromer, K.
1967 Römische Weinstuben in Sayala (Unternubien). Berichte des Österreichischen
Nationalkomitees der UNESCO-Aktion für die Rettung der Nubischen Altertümer
IV. Philosophisch-Historische Klasse Denkschriften 95. Vienna, Verlag der
Österreichischen Akademie der Wissenschaften.
LeFree, B.
1975 Santa Clara Pottery Today. School of American Research Monograph Series Number
29. Albuquerque, University of New Mexico Press.
Longacre, W. A. and M. T. Stark
1994 Kalinga Ethnoarchaeology: Expanding Archaeological Method and Theory.
438    H a n s B a r n a r d
Luft, U., A. Almásy, M. A. Farkas, I. Furka, Z. Horváth and G. Lassányi

2004 Preliminary Report on the Fieldwork at Bir Minih, Arabian Desert.
Mitteilungen des Deutschen Archäologischen Instituts Abteilung Kairo 58: pp.
373–390.
Magid, A. A.
2004 The Site of Tabot: An Old Waystation in the Southern Red Sea Hills, Sudan.
In S. Wenig (ed.), Neueste Feldforschungen im Sudan und in Eritrea: Akten des
Symposiums vom 13 bis 14 Oktober 1999 in Berlin. Wiesbaden, Harrassowitz
Verlag.
Murray, G. M.
1935 Sons of Ishmael: A Study of the Egyptian Bedouin. London, George Routledge.
Nordström, H.-Å.
2004 Pottery Production. In D. A. Welsby and J. R. Anderson (eds.), Sudan, Ancient
Treasures: An Exhibition of Recent Discoveries from the Sudan. London, National
Museum: cat. no. 255 (p. 269).
Paul, A.
1954 A History of the Beja Tribes of the Sudan. Cambridge, Cambridge University
Press.
Rice, P. M.
1987 Pottery Analysis: A Sourcebook. Chicago, University of Chicago Press.
1996 Recent Ceramic Analysis. Journal of Archaeological Research 4: pp. 133–202.
Ritner, R. K.
1993 The Mechanics of Ancient Egyptian Magical Practice. Studies in Ancient Oriental
Civilization 54. Chicago, Oriental Institute of the University of Chicago.
Rose, P. J.
1995 Report on the Handmade Sherds. In S. E. Sidebotham and W. Z. Wendrich
(eds.), Berenike 1994: Preliminary Report of the 1994 Excavations at Berenike
(Egyptian Red Sea Coast) and the Survey of the Eastern Desert. Leiden, Research
School CNWS: pp. 41–43.
Rye, O. S.
1981 Pottery Technology: Principles and Reconstruction. Manuals on Archaeology 4.
Washington, Taraxacum.
Sadr, K., Alf. Castiglioni and Ang. Castiglioni
1995 Nubian Desert Archaeology: A Preliminary View. Archéologie du Nil Moyen 7:
pp. 203– 235 .
Schiffer, M. B., J. M. Skibo, T. C. Boelke, M. A. Neupert and M. Arenson
1994 New Perspectives on Experimental Archaeology: Surface Treatments and
Thermal Response of the Clay Cooking Pot. American Antiquity 59: pp.
197–217.
Shepard, A. O.
1976 Ceramics for the Archaeologist. Washington, Carnegie Institution (fifth printing
of the 1954 manuscript).
Sidebotham, S. E., H. Barnard, J. A. Harrell and R. S. Tomber
2001 The Roman Quarry and Installations in Wadi Umm Wikala and Wadi Semna.
Journal of Egyptian Archaeology 87: pp. 135–170.
Sidebotham, S. E., H. Barnard, L. A. Pintozzi and R. S. Tomber
2005 The Enigma of Kab Marfu’a: Precious Gems in Egypt’s Eastern Desert.
Minerva 16: 1, pp. 24–26.
Sidebotham, S. E., H. Barnard and G. Pyke
2002 Five Enigmatic Late Roman Settlements in the Eastern Desert. Journal of
Egyptian Archaeology 88: pp.187–225.
1996 Interpretative Summary and Conclusion. In S. E. Sidebotham and W. Z.
Wendrich (eds.), Berenike 1995: Preliminary Report of the 1995 Excavations at
Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert. Leiden,
Research School CNWS: Figures 1–2, 1–3 (pp. 3–4), pp. 441–452.
Stark, M. T.
1991 Ceramic Production and Community Specialization: A Ceramic Ethno
archaeological Study. World Archaeology 23: pp. 64–78.
2003 Current Issues in Ceramic Ethnoarchaeology. Journal of Archaeological Research
11: 3, pp. 193–242.
Strouhal, E.
1984 Wadi Qitna and Kalabsha-South: Late Roman-Early Byzantine Tumuli Cemeteries
in Egyptian Nubia. Volume 1: Archaeology. Prague, Charles University.
Wendrich, W. Z., R. S. Bagnall, R.T. J. Cappers, J. A. Harrell, S. E. Sidebotham and
R. S. Tomber
2006 Berenike Crossroads: The Integration of Information. In N. Yoffee and B. L.
Crowell (eds.), Excavating Asian History: Interdisciplinary Studies in Archaeology
and History. Tucson, University of Arizona Press: pp. 15–66 (an earlier version
appeared in the Journal of the Economic and Social History of the Orient 46, 1;
2003: pp. 46–87).
Wisner, M.
1999 The Ceramic Technology of Mata Ortiz. In S. Lowell (ed.), The Many Faces
of Mata Ortiz. Tucson, Rio Nuevo Publishers: pp. 187–197.
Ch a p t e r 2 0
History of the Nomadic

A rchitecture of the
H adendowa in Northeast
Sudan
A n wa r A b d e l - M a g i d 1
T he Hadendowa is the largest and politically dominant group among

the Beja tribal confederation. The other main groups are the Bisharyyin,
the Amarar (Paul 1971:18-19, 21; Vaagenes 1990:35; Egemi 1994:2; Manger
1996:19; Nautrup 2004:11), and the Atman (Vaagenes 1990:35; Hjort af
Ornäs and Dahl 1991:1). All of these tribal groups are believed to be indig-
enous inhabitants of the area between the Nile River and the Red Sea Hills
in the Sudan (Figure 20.1). They speak an unwritten Cushitic language called
Tu-Bedawie (Walker 1987; Vaagenes 1990:35, 1998:1; Hjort af Ornäs and Dahl
1991:2; Egemi 1994:2).
There is a fifth tribal group, called the Beni Amer, that is believed to be
among the indigenous inhabitants of the Southern Red Sea Hills in Sudan. As
they speak a different language, Tigre, their connection to the Beja remains
a controversial issue. Some researchers, like Egemi (1994:2), regard them as
a separate group, whereas others, like Manger (1996:19), Paul (1971:21) and
Vaagenes (1990:35), consider them one of the Beja tribal groups. The Beja,
especially the Hadendowa, do not consider the Beni Amer as one of their tribal
groups. Nautrup (2004:11) assumes that the Beni Amer are part of the Beja
tribal confederation but emphasizes that their language difference creates a
‘symbolic difference’ between them and other Beja groups.
More Beja tribal and subtribal groups live in the Red Sea Hills area and
further south in Kassala State and along the River Atbara. Detailed accounts
on these, including their origins and their history, are presented in several
I am grateful for the help that I received during the process of data collection and
1.
the writing of this chapter from Professor Leif Ole Manger and Professor Richard
Holton Pierce of the University of Bergen (Norway), from Hans Barnard of the
University of California, Los Angeles (USA) and from Krzysztof Pluskota.
441
442     A n wa r A - M a g i d
Figure 20.1. Beja main tribal groups and the location of the study area (map prepared by H. Barnard
after Manger et al. 1996:204).
publications (Seligmann 1913:593–705; Owen 1937:183; Murdock 1959:315;

Arkell 1961:170; Paul 1971:12–37, 137–139; Holt and Daly 1979:1). Nomadic
pastoralism based on the rearing of livestock has always been the basic element
of the economy of the Beja (Egemi 1994:84). They are probably one of the
earliest pastoral groups, not only in Northeast Africa (Grigg 1974:119) but
in Africa at large (Murdock 1959:314). Egemi (1994:86) believes them to be
“the first pastoralists to inhabit Sudan.” Fattovich (1993:445) suggests that
in 2000–1500 BCE the pastoral Gash Delta Group people “spread to the
east and north, as far as the Red Sea coast, occupying the whole northern
Ethiopian-Sudanese borderland.” Fattovich did not, however, engage in
discussing the relationship between the Gash group and the Beja. In his book
Surat al-ard, written in 988 CE, the Arab geographer Ibn Haugal described
N o ma d i c A r c h i t e c t u r e of the H a d e n d o wa      443
the Beja as pastoral nomads wandering about with their animals in search of
good pasture and water in the area extending from the Nile River to the Red
Sea and from Egypt to Eritrea (Ibn Haugal 1979). Historically, the emergence
of the Hadendowa (as well as the Bisharyyin and Amarar) can be traced to the
15th–16th centuries CE (Paul 1971:25).2
The Beja and the Hadendowa can hardly be studied without taking into
account the role of the camel and its impact on their lifestyle. There is no doubt
that the introduction of the camel and the spread of its use in the Beja terri-
tory was a landmark that redefined pastoral economic strategies, demographic
mobility, social structure and political power. Most of these issues are beyond
the scope of this chapter, which will focus on the effects of the introduction of
the camel on the nomadic architecture. The camel has been used as a pack and
riding animal in the Eastern Deserts of Egypt and Sudan since the 3rd century
BCE (Prussin 1995:14). As the oldest pastoral groups roaming the area, the
ancient Beja must have been owners and breeders of camel since the 3rd century
BCE, if not earlier. The earliest direct evidence for the camel’s being used as a
pack animal for the nomadic dwellings of the Beja dates to the 14th century CE
(Burckhardt 1822:182; Vantini 1975:621). This, however, cannot be considered
the earliest date for the use of the camel in the Beja territory.
The introduction of the camel greatly facilitated the transport of the dwell-
ings and belongings of the Beja across their otherwise impenetrable territory.
It enabled them to reach far pastures, to keep larger herds of livestock, and to
obtain new building and furnishing materials from the wool and hides of those
herds. Ethnographic and anthropological research among the Hadendowa has
shown that the camel is indispensable to transport their nomadic dwellings,
furnishings and baggage whenever they travel “distances in excess of a dozen
kilometres” (Ausenda 1987:192). When traveling shorter distances than that,
donkeys are also used, especially in the southern part of the Hadendowa area
(in Kassala State). It takes four to six trips with a donkey to move all parts and
furnishings of a dwelling. It has been reported that, in emergency situations
like floods or fire, a man can move his dwelling over a maximum of one kilo-
meter taking at least fifteen trips to bring all components of the tents to safety
(Ausenda 1987:192). This clearly illustrates that it is labor intensive and time
consuming to move nomadic dwellings, even over limited distances, without
the help of a camel.
Until the late 19th and early 20th centuries CE the Hadendowa had large
herds of cattle and smaller herds of camels in their territory, which extends
from the Sinkat District in the Southern Red Sea Hills to the Gash Delta in
Egemi (1994:86) implies that the Hadendowa have been known since the 10th
2.
century CE.
444     A n wa r A - M a g i d
Kassala State (Seligmann 1913; Sadr 1991). Paul (1971:16) also reports that
large herds of cattle were kept in these areas. Repeated droughts and increased
aridity, especially in the northern part of their territory, forced the Hadendowa,
like many other pastoral groups in the Sahel zone, to abandon cattle herding
and to rear camels, sheep and goats instead. They also cultivate Sorghum
vulgare (durra in Arabic) after the infrequent rainfall. The subtribal groups
of the Hadendowa who inhabit Kassala State still keep small herds of cattle.
As they are more settled, they practice flush-irrigated cultivation of sorghum
on a regular basis. Nevertheless, the characteristic feature of Hadendowa
life is mobility, in search for water and pasture. The type of dwellings they
use reflects their economic and environmental adaptation to the landscape,
as well as their customs. It is in the area of the northern Hadendowa (in the
Sinkat District) that I carried out archaeological, ethnographic and ethno-
archaeological research. This included a study of the Hadendowa past and
present nomadic architecture.
North African Nomadic Tent Dwellings

Theoretically and academically, tent has been defined as “an eminently prac-
tical, completely portable house demonstrating all the principal elements
of permanent architecture” (Cribb 1991:85). Architecturally, a tent may be
defined as “a prefabricated structure consisting of a flexible covering and struc-
tural supports temporarily brought together to form an integrated architectural
unit” (Cribb 1991:85). Both definitions encompass a variety of architectural
units ranging from extremely temporary and disposable to more permanent,
and all combine a tent form with a house function. The Hadendowa nomadic
dwelling fulfills all the elements and variables of these definitions. The data
collected during my fieldwork show that the Hadendowa people know and
distinguish between a tent form and a house function, yet they consider and call
their nomadic dwelling ‘a house’ (beit in Arabic). One is tempted to suggest that
such use and understanding are meant to emphasize a functional rather than an
architectural purpose. My Hadendowa informants, however, offered a cultural
explanation, asserting the distance and distinction between the Hadendowa and
other nomadic groups residing in the area and neighboring territories (like the
Rashaida Arabs). We do not know whether similar tendencies existed among
past Beja populations. The name of the nomadic dwelling of the contemporary
Hadendowa is Badaigaw, meaning ‘mat-house’ (beit al-birsh in Arabic). This is
self-defining and explicitly proclaims such a dwelling to be a house made from
mats. These ‘houses’ are contrasted to ‘mud houses,’ ‘sun-baked brick houses’
and ‘burnt-brick houses’ emphasizing that houses have a range of which the
Hadendowa dwelling is one form. The Hadendowa conception and naming
of their dwelling as houses is consistent with the definition of a house as “a

building intended for human habitation, especially one used as residence of a
family” (Merriam-Webster’s Dictionary) or “a building which people, usually a
family, live in” (Cambridge Dictionary). Owen (1937:204–205) seems to have
been one of the first authors to describe the Hadendowa dwellings as ‘houses’
rather than as ‘tents’ indicating his understanding of their culture. Murray
(1978:80) reports that the Bedouin of the Sinai call their tent a ‘hair house’
(beit al-shaàr in Arabic), while they use ‘tent (kheima in Arabic) to denote a
European-style canvas tent.
In the literature on the nomadic groups and their dwellings in North
Africa the reader often encounters confusion in the use and meaning of the
terms house, tent, hut, dwelling, shelter, and so forth. For instance, Nicolaisen
(1963:327, 332–349), in his ethnographic study of the Tuareg of the Sahara
Desert, uses the term hut to denote three types of permanently built dwellings.
Lewis (1969:67) uses the same term to describe both the portable skin- and
mat-covered dwellings of the Somali nomads in East Africa and the permanent
dwellings of farmers in the same region (Lewis 1969:85). In other instances,
different terms are used to describe a single type of portable nomadic dwelling.
Murray (1978:81) called the portable nomadic dwellings of the Beja ‘huts’ while
Prussin (1995:5) calls the same dwellings ‘tents.’ Another confusing use of the
terms is evident in the work of Oliver (1971:25–35) who used the terms shelter
and tent to describe the dwellings of the nomadic Kababish nomadic group
in western Sudan. I propose the term tent-dwelling (maskan in Arabic) when
referring to the ancient and present nomadic architecture of the Beja and the
Hadendowa, particularly those dwellings in which mat coverings are used. In
this chapter the term tent-dwelling denotes ‘a temporary and spacious structure
meant for human habitation, architecturally designed to meet a particular need
to suit a mobile pattern of exploitation of natural resources and, at the same
time, to symbolize an expression of cultural personality.’
The oldest and simplest form of human-made dwellings is probably a
frame of one, or several, poles, or rods, covered with bark, skin, grass, reeds or
mats. Prehistoric hunter-gatherers were perhaps the first to invent this type
of dwelling, which continues to exist among contemporary hunter-gatherer
groups (Rosen, this volume). Skins are among the rewards of hunting; bark,
reeds, grass or leaves were used where available. All these materials are flex-
ible and can be folded or rolled and neatly packed and moved. Similar to early
hunter-gatherer practice, early pastoral nomadic groups most likely took the
covers and left the frame (the heavy and nonflexible poles or rods) behind when
they moved as “it is easier to cut new poles than to carry the old ones along”
(Faegre 1979:60–61). Obviously, if people moved in areas where wood is scarce,
the frame must have been taken along as well, their poles or rods lightened as
446     A n wa r A - M a g i d
much as possible. As the tent-dwelling is set up and dismantled many times,

systems making the disassembling, transportation and subsequent assemblage
easier undergo continual improvement and modification to adapt to changing
environments, building materials and resources, as well as to meet cultural and
individual preferences (Faegre 1979:60–61).
Prussin (1995:54) divides the tent-dwellings of the African pastoral groups
into two basic types: tensile structures and armatures. A tensile structure consists
of a central pole (or system of poles) kept in position by stretching a sheet (or
sheets) of membranous covering or coating tightly over the pole. One form of
tensile structure, commonly called a tent, is architecturally based on the struc-
tural interdependence between the pole(s) and the covering(s). The armature
has a system of poles that, unlike those of the tensile, can stand up independently
without any tight covering. The type or form of covering, or the tension it puts
on the pole(s), has no structural function (Faegre 1979; Prussin 1995).
Some have suggested a simple evolutionary development from an armature
to a tensile structure without providing cultural, economic or environmental
reasons for such development. Others, like Prussin (1995:55), note that the
architecture of the dwellings of African nomadic groups does not always follow
this evolutionary sequence. Sometimes the transitions are retrograde, and
in other instances both types, tensile and armature, co-exist. There are also
forms of African nomadic architectural structures that appear to be composites
between armature and tensile structures. The contemporary Hadendowa dwell-
ings, portable structures consisting of a frame of wooden poles covered with
mats, can be regarded as such a composite type because they combine the struc-
tural interdependence between the wooden frames (of the tensile structure) and
the independent coverings that characterize the armature structure.
The Hadendowa Tent-Dwelling

The emphasis of this chapter is on the history of the Hadendowa nomadic
architecture, its development and its economic, social, symbolic and environ-
mental implications. A few problems were encountered during the process
of data collection and analysis. First, previous research on the history of the
Hadendowa nomadic dwellings was scarce and fragmentary. Furthermore, the
Hadendowa nomadic dwellings were, and still are, made of perishable building
materials of organic origin. These materials (wood, leather, wool, etc.) decay
and perish over time. It is therefore difficult to find the evidence of ancient
dwellings or settlements. The settlements and campsites of the Hadendowa
are often occupied for a short time, which does not allow for the accumulation,
and recovery, of indicative cultural debris. Because of their continuous move-
ment, pastoral groups tend to have few possessions or tools, most of which are
multifunctional and lightweight. When a nomadic dwelling is moved, what is

left behind is a hearth, a few shallow postholes, some stones and perhaps frag-
ments of broken pottery or grinding stones. These will get buried, blown or
washed away by wind, rain or the water running off the mountains in a gully
or stream (khor or wadi in Arabic). The Hadendowa area is predominantly
mountainous, which adds to the other constraints on finding archaeological
evidence of nomadic dwellings. The mobile settlement pattern of the Beja, the
perishable nature of the building materials of their dwellings, and the rocky
nature of the Red Sea Hills are the three main reasons for the lack of archaeo-
logical evidence for this people (but see Saidel, this volume).
The research questions of this study of the Hadendowa tent-dwellings
include the following:
— Did the Hadendowa once have types of dwellings other than the contem-
porary mat house (the Badaigaw), and, if so, what was the nature and
architectural structure of these dwellings?
— How far back can we trace the history of the contemporary Badaigaw and
the other types of dwellings, if any existed?
— What are the forms and designs of the earlier types of dwellings, and how
did they develop over time?
— Were these forms of dwellings subject to change as a result of external
influence after contacts with other regions?
— How is the development associated with the changes in environmental
conditions and the Hadendowa economy?
In the course of discussing these questions, I will also deal with issues related
to the history of the ancient nomadic architecture, such as the building tech-
nology, the transportation technology and the symbolism associated with these
structures. There is textual and archaeological evidence indicating that the
history of the nomadic dwellings in the Beja territory is much older than the
emergence of the contemporary Beja tribal groups, including the Hadendowa.
It should also be emphasized that the contemporary Beja nomadic dwellings
as used by the various tribal groups seem to be closely related, exhibiting simi-
larities in frame structure, form, coverings and building material. Therefore,
the present study deals with the ancient history of the nomadic dwellings in
the Beja territory as a whole.3 Mention is made where there is any concrete
evidence of specific names or areas.
The study of the ancient nomadic dwellings in the territory of the contem-
porary Beja would not be complete without considering the history of similar
3.
Whether past population(s) that lived in the territory of the contemporary Beja are
related to or different from the present Beja population is a question beyond the scope
of this chapter (see Burstein, this volume).
448     A n wa r A - M a g i d
and related ancient dwellings within the wider context of North Africa. The first
written evidence for tent-dwellings in Egypt is found in royal correspondence
from Dynasty VI (2325–2175 BCE), in which no information on the shape of
these dwellings is provided (Prussin 1995:7).4 Textual evidence from around
1500 BCE indicates that the tent became part of Egyptian military equipment
because all references are in a military context (Drew 1979:5). Reliefs in Luxor
and Abu Simbel clearly show Egyptian royal and military tents, but no informa-
tion is available on the materials from which these were built, how they were
built, or if any tents existed that were used by ordinary people (Prussin 1995:7).
Other textual sources indicate that the ancestors of the contemporary Beja tribal
groups first came to the attention of the Egyptians around 1900 BCE. The
first depiction of a ‘Bejawie’ (probably related to the modern Beja) appears in
a Dynasty XII (1991–1786 BCE) tomb-chapel in Upper Egypt, but there is no
reference to any type of Beja dwellings or to the name of the owner of the tomb
(Prussin 1995:6). It is possible that the ancient Bejawie resided in tent-dwellings
as mentioned in the Egyptian texts and even that the Egyptian tent was adopted
from the tent-dwellings used by the Bejawie people.
Later archaeological and textual evidence provides us with more, although
still incomplete, information on the history of the Beja tent-dwelling. The
available data indicate that the area has had a pastoral nomadic population since
times immemorial (Burckhardt 1822; Seligmann 1913; el-Maqrizi 1922; Owen
1937; Paul 1971; Hasan 1975; Vantini 1975; Updegraff 1978; Egemi 1994).
Their dwellings were characterized by their simple design, their light weight
and the ease with which they could be dismantled, moved and reconstructed.
The contemporary Beja mat tent-dwellings illustrate these characteristics.
Several sources mention skin, hair and mat tent-dwellings (Burckhardt 1822;
el-Maqrizi 1922; Paul 1971; Vantini 1975; Prussin 1995). The mat tent-
dwelling is the only type that has survived and has continued to be commonly
used by the modern Beja. A detailed documentation on the contemporary
Hadendowa mat tent-dwelling will be the subject of an in-progress ethno-
graphic study on the social and spatial symbolism of the contemporary nomadic
architecture of the Hadendowa tribal group in northeast Sudan. Therefore, the
following focuses on the ancient skin, hair and mat tent-dwellings.
Skin Tent-Dwellings
One of the earliest mentions of leather tent-dwellings in North Africa, used
mainly by soldiers, dates to the Egyptian Middle Kingdom (1975–1640 BCE).
The given dates of the ancient Egyptian dynasties are those suggested by Baines
4.
and Malek (2000:36–37).

Of interest here is a remark in the story of Sinuhe, around 1875 BCE, in which
it is mentioned that Sinuhe stayed with the Bedouin who live in tents in the
desert between Egypt and Palestine (Helck and Otto 1983:1372). It is implied
that these tents were made of leather. Other textual evidence indicates that
these Bedouin were living in present-day northeast Sudan and southeast Egypt
(Sadr 1991). Leather tents may also have been used by the Libyans who came
to Egypt around 1194–1188 BCE, but it is not clear whether they brought
these tents or used tents that they found in Egypt (Roe, this volume; Helck and
Otto 1986:1372). Egyptian accounts on military campaigns against the Syrians
also mention tents. In 1274 BCE the Egyptians captured the son of the Syrian
king of Qadesh and took the tent in which he was found. Assyrian reliefs on
this event include drawings that clearly show what this tent looked like. An
intriguing remark by Prussin (1995:7–8) states that the mat tent-dwellings of
the contemporary Hadendowa “bear close resemblance to the structure of the
ancestral Assyrian tents.” In addition, Prussin states that the two Egyptian
terms used in Dynasty XIX texts (1292–1190 BCE) imply that the Nubians,
against whom the Egyptians were fighting, used mats and leather as coverings
for their tents (Prussin 1995:7). On the basis of this evidence Prussin argues
that the Egyptian army, finding the southern Nubian tent suitable to their
needs, may have adopted and used it.
The earliest archaeological evidence of leather tent-dwellings in North
Africa dates to Dynasty XXI (1075–945 BCE) and consists of a large leather
tent found among royal mummies in what was originally a Dynasty XI tomb
complex (Brugsch 1891). The hieroglyphic texts on the leather indicate that
the tent belonged to Isimkheb, a princess and high priestess, and that it was
a funerary symbol of a high-ranking office (Prussin 1995:7–8). This tent was
made of the skins of hundreds of goats and gazelles (Prussin 1995:7). An
intriguing question has been raised regarding the possibility of its being related
to the Beja (Prussin 1995:8). As no information is available on its supports and
its shape when erected, the nature of its relation to the Beja, as suggested by
Prussin, remains unclear.
The first and second centuries CE were the period of contact between the
Romans and the indigenous pastoral tribes in North Africa. These contacts
involved exchange and adaptation of the type of dwellings these pastoral groups
used at the time. It has been reported that during their military campaigns
in Africa, the Roman soldiers slept in skin tents. These tents consisted of a
velum made of a number of calf or goat skins, stitched together to form rect-
angular coverings. These were fixed to a pole and held taut with ropes (Prussin
1995:7–8). A 310–330 CE mosaic showing hunting scenes and tents was found
at Hippo Reguis on the Mediterranean coast. Tents are also shown on a mosaic
from around 300 CE, found at Oudna in Tunisia, depicting a rural scene.
450     A n wa r A - M a g i d
When the Arabs arrived in North Africa, during the 7th century CE, a
large network of trade routes was developed linking the West African coast
to the Red Sea and further east across the Red Sea into modern Saudi Arabia.
These trade routes provided a stimulus for the spread, exchange and adapta-
tion of nomadic architectural knowledge and technology. By the 11th century
CE nomadic dwellings were in use all over this vast area (Prussin 1995:7–8).
At present, the northern Tuareg of the Sahara are renowned for their skin
tent-dwellings (Faegre 1979:29; Prussin 1995:6). Skin tents are also used
by other contemporary nomadic groups roaming the western Sahara, for
instance the Teda. Skin and mat tent-dwellings are also reported to be in use
by pastoral groups living in the deserts of Egypt, Sudan, Ethiopia and Somalia
(Faegre 1979:64).
During my archaeological survey and excavations in the southern Red Sea
Hills, mainly in the Hadendowa area, I neither saw skin tent-dwellings nor
was I told about the presence of this type of dwelling anywhere in the area.
The archaeological survey and excavations conducted in the southern Red Sea
Hills area did not yield any evidence of exclusively nomadic camp or settle-
ment sites. Thus, our only sources are the fragmentary accounts of historians,
early travelers and geographers who traveled in the area. Among these is the
medieval Arab historian el-Maqrizi (around 1300 CE) who described the Beja
as a nomadic people living in tent-dwellings made of skin. He also stated that
they kept large numbers of ‘well-bred, long horned’ cattle, as well as camels,
sheep and goats, and that they roamed the area searching for good pasture for
their animals (el-Maqrizi 1922). On the basis of this account one could infer
that, contrary to the present time, the climatic conditions were less arid and
more favorable for large herds of cattle, camel, sheep and goats. Even though
el-Maqrizi tells us that the tent of the Beja magician has a dome shape (Vantini
1975:631), he does not specify what type of hides was used, if the hides used
were treated (tanned), or how many hides were needed to build this or other
types of tent-dwellings.
The archaeological evidence for the presence of cattle, sheep and goats in
the Beja territory predates the account of el-Maqrizi by thousands of years,
suggesting a population of cattle pastoralists in what was then a savannah envi-
ronment (Sadr 1991). A text dating about 2575 BCE states that the Egyptian
king Snofru invaded the land of the Beja and Nubia and returned to Egypt with
200,000 head of cattle (Sadr 1991:93). Excavations at the site of Tabot in the
southern Red Sea Hills, west of Sinkat, revealed bones of cattle, sheep and goats
with a radiocarbon date of 225 CE ±60 (Magid 1998). A large rock drawing
found at the site of Samadi, also situated in the southern Red Sea Hills, west
of Sinkat, portrays a large number of herds of long-horned cattle. The date of
this drawing is unknown, but burial tombs found in its vicinity have been dated
to the 3rd century CE (Magid et al. 1995). This is consistent with the fact that
the Beja tribes owned large herds of cattle and camels until a hundred years
ago (Seligmann 1913; Paul 1971; Sadr 1991). Historical sources (Paul 1971:16),
oral history (Dirar 1992:731), and field observations (Nautrup 2004:268) also
report that the Hadendowa bred cattle and that it was the women who did the
tanning and built the tent-dwellings.
Complementary evidence includes the undated rock paintings in the moun-
tain massif of Tassili and Ajjer in the central Sahara. These show accurate
“renditions of the nomadic ethnographic present” (Prussin 1995:4). One of
the paintings shows a woman, surrounded by herds of cattle, holding a bent
wooden frame used for an armature tent-dwelling in a form similar to that
of the contemporary Rendille nomads, in Somalia, and the contemporary
Hadendowa. Another painting portrays an accurate elevation of a tent armature
(Prussin 1995:4). It has been stated that the dome frame, made of bent poles,
is not only the most commonly used frame for the African nomadic tent-
dwellings but also that it is “undoubtedly the oldest” (Faegre 1979:62) in both
West and East Africa. The women of the contemporary nomadic Tuareg and
Teda in the Sahara are known for their outstanding skills as leather workers.
They produce both the skin tent-dwellings and the leather farthings of these
dwellings (Faegre 1979:65). The most commonly used skin for the cover of
the skin tent-dwelling is that of goats, but sheepskin is also used. Cattle skin
is used where available (western Sudan). Treatment and preparation of the
skin includes the removal of the hair and tanning with fruits and bark of the
Acacia arabica. It is smeared with butter or tar to make it water resistant, soft
and foldable. Ethnographic studies among contemporary pastoral nomads in
Africa show that cattle hide is less desirable because it shrinks and hardens when
it dries after being exposed to (rain)water (Faegre 1979:70). The coverings of
very large Tuareg tent-dwellings are made of as many as 150 goatskins, while
an average size tent-dwelling is made of 30–40. The tent-dwelling of a typical
contemporary nomadic Danakil in Ethiopia has a diameter of around 3.3 m
and is about 1.3 m high at the center (Faegre 1979:75). Such a dwelling can be
made with only 10–15 goatskins.
From the archaeological evidence of nomadic dwellings from the central
Sahara, and the ethnographic parallels from both the Sahara and East Africa,
it can be inferred that the ancient skin tent-dwellings of the Beja were dome
shaped with a wooden frame and that women were responsible for both
processing the building materials and building the actual dwelling. The
tradition of such skin tent-dwelling probably dates back thousands of years.
Classic sources and accounts of travelers and geographers on the Beja attest
to the presence of skin tent-dwellings in the Beja territory at least six to seven
centuries after the arrival of the Arabs. There are strong indications that skin
452     A n wa r A - M a g i d
tent-dwellings used by the ‘ancient’ Beja were similar to those used by the
ancestors of the contemporary Tuareg and by other nomadic groups in West
and East Africa. Accordingly, there are reasons to assume that the African skin
tent-dwelling (including that of the ‘ancient’ Beja) was one of the products of
an indigenous African nomadic architecture.
Neither the written sources, the archaeological evidence, nor the oral tradi-
tion indicate when and why the use of the skin tent-dwellings was abandoned.
A steady decrease of the rainfall may have been the main factor that led to a
marked reduction in both the size and composition of the herds leading to the
abandonment of skin tent-dwellings for which substantial numbers of skins
are required.
Hair Tent-Dwellings
Information on the ancient hair tent-dwellings of the Beja is mostly obtained
from the accounts of early travelers and geographers writing about the region.
The earliest of these are by Arab writers like al-Istakhari (932–950 CE), Ibn
Haugal (around 977 CE) and al-Hamadhani (Vantini 1975:112, 151, 633; Ibn
Haugal 1979). They report that the Beja live in hair tent-dwellings but do
not name the animal(s) providing the hair or the shape of the tent, although
al-Hamadhani describes the hair of the tent as being ‘woven.’ Two centuries
later, el-Maqrizi wrote about the Beja territory, its environment, wildlife, vege-
tation, its people and their food, economic strategies, spiritual life and some of
their customs and habits (el-Maqrizi 1922:280). Of interest here is their diet,
which consisted of only milk and meat. This indicates the importance of the
animal herds that they owned. El-Maqrizi also made a general remark on the
dwellings of the Beja, describing them as tents made of hair and skin.
Reporting on his visit to a Beja, and most likely Hadendowa, encampment
in the early 19th century CE, Burckhardt tells us that he found tents of woven
hair supported by bent poles (Ausenda 1987). Although Burckhardt did not
provide detailed information on the structural architecture or form of these
dwellings, his general description of the frame as having bent poles suggests
that the hair tent-dwellings have a dome shape. If this inference is correct,
then there is reason to believe that, similar to the skin and the contemporary
mat tent-dwellings, the frame structure of the ancient Beja hair tent-dwelling
was a product of an indigenous nomadic architecture. During my survey and
archaeological excavations in the southern Red Sea Hills, I did not see hair
tent-dwellings, nor did I find any archaeological evidence or traces indi-
cating the presence of these in the past. Knowledgeable informants from the
Hadendowa area could not recall having ever seen or having ever been told
about Beja dwellings made of animal hair.
Unlike the Beja, some of the pastoral nomadic groups in North Africa
replaced their traditional dome-shaped skin and mat tent-dwellings with a
modified form of the Arabian rectangular ‘black tent’ (Saidel, this volume;
Faegre 1979:62). Examples of these groups in Sudan are the al-Rashaida in East
Sudan,5 the Kababish in West Sudan and other pastoral groups in North Africa,
such as the Berber of the Djebel Nefousa in Tunisia. They have introduced
some modifications that gave the tent an African “characteristic profile of a high
curved ridge line instead of the Arabian flat roof” (Faegre 1979:29). This means
that the history of the ‘black tent’ in North Africa is not older than the arrival
of the Arabs in Africa, in the 7th century CE. The black tent is a tensile struc-
ture that requires few vertical wooden poles and few tension bands. Goat hair
is predominantly used for making the black tent, which explains its name, but
sheep wool and camel hair are often added. Women design and make the black
tent. It casts adequate shade, while its loose weaves allow good ventilation. The
hair provides warmth during cold winter nights (Faegre 1979:7–12).
Accounts on the Beja territory in the 10th-19th centuries CE describe plants
and animals similar to those existing today in the savannah zone, at least 700 km
further south. This shows that the climatic conditions were once less arid than
today. Because a hair tent-dwelling will leak after prolonged rain and become
too heavy for its frame structure, it is likely that such dwellings were not desir-
able in such a wet climate (Faegre 1979:12). The earliest date for the presence
of hair tent-dwellings in the Beja area is from the 10th century CE, almost three
centuries after the Arabs arrived in North Africa. The dome-shaped type of
the hair tent-dwelling continued to be used in the Beja area until, or shortly
after, the late 19th century CE.
Available sources do not show who was responsible for building the hair
tent-dwelling in the Beja territory. But the building of other ancient and
contemporary Beja types of tent-dwellings is assigned to women. Ethnographic
studies among North African pastoral groups show that the collection, prepa-
ration and processing of the building material, as well as the building and
maintenance of the hair tent-dwelling, are performed by the women. It is
therefore most likely that building and maintaining the hair tent-dwelling of
the Beja was also performed by women. The history of the hair tent-dwelling
in the Beja area is short compared to that of the skin and mat dwellings. The
construction resembles the older tent-dwellings in form and frame structure
(a dome shape on a bent wooden frame). Goat hair was probably the preferred
fiber to make the tent covering, but sheep wool and camel hair may also have
been used. It is not possible to infer the size of the dwellings or the amount of
hair needed to build each type of dwelling.
5.
The Rashaida are an Arab pastoral group who migrated from Saudi Arabia to the
Sudan about 150 years ago (Egemi 1994).
454     A n wa r A - M a g i d
The presence of a dome-shaped hair tent-dwelling in the Beja area can be

explained in two ways. The Beja may have borrowed the idea and technology of
making a woven-hair covering for their traditional dome-shaped frame structure.
Unlike other North African pastoral groups who adopted the Arabian rectangular
hair tent, the Beja were apparently reluctant to embrace this dwelling, despite
their gradual Islamization and intermarriage with the Arab invaders. This resis-
tance may, on the one hand, have been based on sociocultural or political reasons.
On the other hand, the Arabs who settled in the Beja territory and intermarried
with the Beja may have adopted the traditional dome-shaped frame structure,
while retaining the hair covering of their Arabian black tents. In other words, the
Arabs who settled in the Beja area may have adapted the hair covering of their
rectangular nomadic dwellings to the dome-shaped frames used by the Beja.
Mat Tent-Dwellings
The contemporary mat tent-dwelling is the most common, if not the only, form
of housing used by the present nomadic Hadendowa and other Beja nomadic
groups (Figure 20.2). The archaeological survey and test excavation carried
out in the southern Red Sea Hills yielded no evidence of any form of mat
tent-dwellings. Textual sources and ethnographic sources, however, provide
some information about the history of the mat tent-dwelling of both the Beja
and the Hadendowa.
Figure 20.2. Typical Hadendowa Badaigaw (mat tent-dwelling).

The earliest mention of mat tent-dwellings occurs in a text dated to Dynasty

XIX (1292–1190 BCE), which suggests that the Nubians used both mat and
skin tent-dwellings (Prussin 1995:7). Although there is no reference in the text
to dwellings of the population inhabiting the territory of the contemporary
Beja, it is likely that they used similar dwellings. This assumption is based on
the suggestion that the Egyptian skin tent-dwelling of Dynasty XXI (1075–945
BCE) was a modified version of the Nubian or Beja skin tent-dwelling and
that mat tent-dwellings were known in Egypt since the Old Kingdom (Helck
and Otto 1986:1372; Prussin 1995:8). In the 1st century CE Strabo reports
that the Beja nomadic housing was “built by interweaving split pieces of palm
leaves” (Prussin 1995:6). But Strabo stopped short of pointing out who was
responsible for building the mat dwellings or of describing their form and
frame structure.
Evidence, both archaeological and ethnographic, from elsewhere in North
Africa provides pointers that may help fill in these gaps in our knowledge. This
evidence includes ethnographic studies that show that the Beja mat tent-dwelling
is closely related to that of the contemporary pastoral Tuareg and Teda in the
Sahara Desert (Faegre 1979:75). The archaeological evidence from Tassili and
Ajjer, in the central Sahara, indicates that the contemporary Tuareg and Teda
mat tent-dwelling is a product of a building practice that existed at the time when
these rock paintings were made. It has been stated that the form of the present
Hadendowa mat tent-dwelling bears close resemblance to that of an Assyrian
tent depicted on an Assyrian relief (Prussin 1995:8). Furthermore, the oral tradi-
tion among the Hadendowa indicates that the contemporary building practice,
as well as the form and frame, of the mat tent-dwellings has been known in the
area since time immemorial. Accounts of early Arab travelers and compilers, as
for instance el-Maqrizi, include description of different species of trees that may
have provided the raw materials for a mat tent-dwelling. These include the doam
palm (Hyphaene thebaica) and Ziziphus spina-christi. The accounts also contain
detailed descriptions of species of plants concurrent with conditions similar to a
savannah type of environment (el-Maqrizi 1922:270; Vantini 1975:151).
Unlike the skin and hair tent-dwellings, the mat tent-dwellings continued
to be widely used in the Beja territory. Among the reasons for its survival are
the availability of the necessary building materials in the local environment;
the simplicity of its construction, dismantling and transport; and its durability
in all types of weather. Mat tent-dwellings were probably used in the region of
the contemporary Beja since the first millennium BCE, and a direct associa-
tion of these dwellings with the Beja was first made in the 1st century CE. The
ancient mat tent-dwellings had a dome shape similar to those used at present.
They were built of the same materials, mats made from leaves of the doam
palm and bent wood.
456     A n wa r A - M a g i d
Reed Huts, Caves and Rock Shelters

Historians often refer to the description by the 5th century BCE Greek histo-
rian Herodotus of the portable dwellings, made on reeds and rushes, in the
region of the ancient city Meroe (Prussin 1995:5). Indirect association of these
reed dwellings with the ancient Beja was first indicated in the inscription of
the Aksumite king Ezana in recording his military campaign against Meroe,
and his destruction of their royal capital, in 350 CE. The inscription mentions
the ‘Noba’ people’s dwelling in reed huts and the fact that the army arrived at
the frontier of the ‘Red Noba.’ Trimingham (1949:45) suggested that the ‘Red
Noba’ might be the Blemmyes, who probably owned and used the same type of
reed dwellings. The earliest direct evidence of reed dwellings in the Beja terri-
tory dates to around 1267 CE, when the Arab traveler el-Maqrizi wrote that
the majority of the houses in the Red Sea Aydhab were made of reeds (Vantini
1975:148). No description of similar dwellings is known from anywhere else
in the Beja territory. It should be emphasized that el-Maqrizi identified these
dwellings as ‘houses’ and not ‘tents’ indicating that the Aydhab dwellings were
probably permanent rather than pastoral ones. As Aydhab was at the peak of
its activities as a sea port for pilgrims and trade caravans crossing the Red Sea
during the 12th–13th centuries CE, it seems likely that some, if not all, of these
houses were owned by Beja to rent out to outsiders passing through the port
(Trimingham 1949:13).
Based on archaeological and ethnographic data from the southern Red
Sea Hills, caves and rock shelters were, and still are, used in the Beja area
as temporary shelters (Magid et al. 1995:163–190). Two caves and one rock
shelter were located during our survey. The size of the rock shelter and the
two caves appeared too small to accommodate more than one person comfort-
ably. Rock paintings of circles, some with a cross inside, were found on the
walls of the caves. Several fireplaces in and just outside one of these caves were
excavated, revealing a few potsherds in three different layers, each 10 cm thick.
These cultural layers were thin and badly eroded. The quantity and type of
these remains indicate that the cave was occupied for short and discontinuous
periods. Radiocarbon analysis of charcoal samples from the oldest, bottom
layer of the cave (about 30 cm below the surface) showed them to date to the
13th century CE. The second cave was surveyed, and one recent fireplace,
but no other cultural remains, was found. Our informants believe that such
types of shelters have been used since time immemorial during cold nights or
periods of rain.
The scarcity of archaeological material in the caves and the infrequent
mention of reed huts in the accounts of classic writers and early Arab travelers
indicate that both types of dwellings seem to have been used for purposes other
than those connected with livestock and herding. The reed huts were used
for accommodation of outsiders, pilgrims and traders, while the rock shelters
were used by Beja travelers and passersby for temporary protection. One may
therefore cautiously conclude that reed huts and caves are not common forms
of Beja nomadic dwellings, nor were they regularly used by the ancient popula-
tion of the Beja territory.
Euphorbia Stem-Dwellings
These dwellings (bokar in Hadendowa) were first reported, in the Hadendowa
area, at the beginning of the 20th century CE by Seligmann (1913). They are
made by driving the dead stems of Euphorbia abyssinica into the ground to form
circular or oval structures slanted in such a way that they produce walls and
domes (Figure 20.3). Mat tent-dwellings usually are dispensed with when a
Euphorbia stem-dwelling is built. As these dwellings are more permanent, their
nomadic owners often use them as base dwellings for most of the year while
using mat tent-dwellings during seasonal movements. The Euphorbia stem-
dwellings are still used by contemporary Hadendowa but only in a limited
area where Euphorbia is relatively abundant (in the area of Erkweit, east of
Figure 20.3. A Euphorbia stem-dwelling in the north of the study area; note the stone supporting wall
(photograph by Krzysztof Pluskota).
458     A n wa r A - M a g i d
Sinkat, in the eastern part of the Hadendowa territory). This area receives
summer and winter rain and is characterized by its high elevation (about 1000
m above sea level).
Our informants confirmed that they use Euphorbia stems because they
survive the humid climate better than mats or other types of building mate-
rial. The ancient dome-shaped design, which is traditionally associated with
the ancient skin, hair and mat tent-dwellings, is retained in the Euphorbia
stem-dwellings. We observed that many Hadendowa living in Euphorbia stem-
dwellings have settled and have completely abandoned a pastoral nomadic
lifestyle. There are dwellings in the area of Erkweit, Sinkat, and farther west
that are similar in architectural design but with frames made of curved wooden
beams and covered with mats of doam-palm leaves. These are also called bokar
in the Hadendowa language.
Litters and Palanquins

Palanquins and litters (howdaj and shibria in Arabic) are inseparable parts
of the traditional nomadic dwelling (Prussin 1995:17). Both consist of a
superstructure fastened on top of a packsaddle and set on a camel’s back to
transport people, mainly women and children, and baggage (Figure 20.4).
They are also used to shelter infants when the nomads take a break during
the day or camp overnight during their journey. Thus, litters and palanquins
can be considered dwellings in transit (Prussin 1995:46–47, 107). The frame
of the superstructure of the litter and palanquin is commonly made of wood,
bent in the shape of a bow, which is assembled over a pad so that they form
a dome. Mats of doam-palm leaves or of wool and sheets of cloth cover the
frame. Slight differences are observed in the details of the frame design or the
ornamentation of litters and palanquins among the different North African
nomadic groups. With the covers put over the frame, the litter and the palan-
quin give the impression of an armature (Prussin 1995:52). The materials
for the litters and palanquins are collected and processed by women, who
are also responsible for the dismantling, re-assembling and mounting of the
litter (Prussin 1995:42).
Litters and palanquins are commonly used among many North African
nomadic groups like, for instance, the Tuareg and the Teda in the Sahara; the
Kababish, the Hassanyia and the Beja nomads in the Sudan; and the Somali
in East Africa. Our informants state that the palanquin has been used in the
Hadendowa area since time immemorial and that the technological knowledge
and experience of its manufacture is passed on from mothers to daughters. The
available sources provide no information on the history of litters and palanquins
in the Beja territory. The depiction of several camels carrying palanquins on
Figure 20.4. Detail of a 16th century CE bronze map of the world, showing camels carrying palanquins
in Nubia Sarracenorum (currently kept in the Vatican Museum, Rome; photograph by Krzysztof
Pluskota).
a 16th century CE bronze map of the world, currently in the Vatican Museum
(Rome), proves that palanquins were used in Nubia, and probably by the Beja
as well, by that time and probably earlier.
The Packsaddle
In the history of the camel as a means of transport of the Beja people, their
dwellings, furnishings and other possessions, the packsaddle has played an
important role. The difference in loading a camel with a person or a dwelling,
combined with the anatomy of the camel, dictated the development of two
different types of saddles, the riding saddle (makhlufa in Arabic) and the
packsaddle (haweya in Arabic). It is noteworthy that the packsaddle is multi-
functional (it can also be used for transporting individuals), while the riding
saddle is function-specific (it cannot be used for moving loads). As nomads
require a packsaddle more than a riding saddle, it seems safe to assume that
the packsaddle was invented first and that the riding saddle was a later develop-
ment. The earliest archaeological evidence in Sudan of a camel packsaddle is
a figurine that was found in Meroe and has been dated to the 1st century BCE
(Epstein 1971:566; Murray 1978:104; Bulliet 1990:117). One of the earliest,
and often cited, literary references to the camel in North Africa dates to 46
460     A n wa r A - M a g i d
BCE and tells us that when the Romans defeated King Juba of Numidia, at the
battle of Thapsus, they captured twenty camels. These may have been pack
animals normally used for agriculture (Gauthier-Pilters and Dagg 1981:117).
Evidence from North Africa for the distinction between the packsaddle and
the riding saddle is found on a frieze in a Roman house at al-Djem, in Tunisia,
dated to 140–160 CE. The frieze shows “a camel being ridden on what appears
to be a saddle with two side panels” (Prussin 1995:14). Other evidence consists
of Assyrian reliefs in Egypt and terracotta figurines of loaded camels dated to
the Greco-Roman period in Egypt (332 BCE–395 CE).
Similar to the structure of several African camel saddles, that of the contem-
porary Beja and Hadendowa is characterized by a framework put together as
a permanently assembled unit. The assembly of the packsaddle, as well as that
of the riding saddle, requires adequate knowledge of the camel anatomy and
thorough experience with knots and the rawhide used to fasten the wooden
components of the saddle (Prussin 1995:51). All literary and archaeological
evidence indicates a continuity of the nomadic building practices, as well as
the materials and production techniques of packsaddles. It is therefore likely
that the procedures of loading and unloading, used by the contemporary
Hadendowa, is also a continuation of ancient practices.
Discussion
Limited research has been done on the Beja, and Hadendowa, ancient nomadic
architecture, its development and its socio-economic and environmental
context. Our archaeological and ethnographic research aimed at studying these
largely neglected issues, but the results remain inadequate to draw compre-
hensive conclusions. Other sources of information are therefore included
to construct a historical account on the nomadic architecture of the Beja.
These sources consist of textual evidence, accounts of classic writers and early
Arab travelers and geographers, historical and ethnographic studies and oral
history. Combining the data from these sources suggests that there has been
a continuity of forms and building practices of the nomadic dwellings in the
Beja territory for at least the last 2000 years. The dome shape and the building
materials, mainly plant material and to a lesser degree leather and hair, of the
dwellings has continued to be the same. There are reasons to assume that the
building practices and the form of the dwellings are products of indigenous
knowledge and experience. Skin, mats and hair were used to cover the frame
structure. The use of skin and hair coverings, evident in rock drawings and
historical accounts, attests to the presence of large herds of livestock in the
recent past. Combined with the accounts of the ancient flora and fauna in the
Beja territory, these large herds indicate that the climate was once less arid. The
disappearance of the skin and hair as coverings for the Beja nomadic dwellings
may be attributed, at least partly, to the loss of the herds as a result of decreasing
rainfall. The mat tent-dwelling continued to be used by the Beja because of
the availability of its building material. Similar to ancient nomadic groups in
North Africa, the Beja women were the collectors, processors and producers
of building materials, as well as the architects, builders and maintainers of the
nomadic dwellings. There are reasons to suggest, therefore, that the dome
shape of the dwellings was the reflection of gender skills and a marker of both
the production process and the final product.
Although little is known about the origins and history of the palanquin in
the Hadendowa and the Beja area, it bears witness to the creative process of
nomadic architecture in general. It is built of the same materials, retains the
same form, and serves the same functions as the tent-dwelling: protection and
privacy. It illustrates the relationship of the nomadic dwellings during move-
ments and reflects the proficiency and practicality of the nomadic technology
of transport. The combination of the camel and the packsaddle provided an
efficient nomadic vehicle, enabling the Beja and the Hadendowa to break
physical and geological barriers and allowing them to reach new and better
natural resources. It also enabled them to keep larger herds of cattle, sheep and
goats, which in turn supplied them with wool, hair and leather for covering
and furnishing their dwellings.
References
Arkell, A. J.
1961 A History of the Sudan from the Earliest Times to 1821. London, Athlon Press.
Ausenda, G.
1987 Leisurely Nomads: The Hadendowa (Beja) of the Gash Delta and Their Transition
to Sedentary Village Life. University of Columbia (PhD dissertation).
Baines, J. and J. Malek
2000 Cultural Atlas of Ancient Egypt. Abingdon, Andromeda.
Brugsch, H. K.
1891 A History of Egypt Under the Pharaohs. London, John Murray.
Bulliet, R. W.
1990 The Camel and the Wheel. New York, Columbia University Press.
Burckhardt, L.
1822 Travels in Nubia. London, John Murray.
Cribb, R.
462     A n wa r A - M a g i d
Dirar, M. S.
1992 Tareikh Sharaq al-Sudan: Mamalik al-Beja, Qabaelaha wa Tareikhaha: Al-Juzu
al-Awal wa al-Tnai [in Arabic]. Cairo, Maktab Abkass.
Drew, P.
1979 Tensile Architecture. London, Granada.
Egemi, O. A. M.
1994 The Political Ecology of Subsistence Crisis in the Red Sea Hills, Sudan. University
of Bergen, Department of Geography (PhD dissertation).
el-Maqrizi (edited by M. Gaston)
1922 El-Mawaiz wa`l-I`tibar fi Dhikr El-Khitat wa`l-Athar [in Arabic]. Cairo,
Imprimerie de L’Institut Français.
Epstein, H. (with I. L. Mason)
1971 The Origin of Domestic Animals of Africa. New York, Africana Publications.
Faegre, T.
1979 Tents: Architecture of the Nomads. London, John Murray.
Fattovich, R.
1993 The Gash Group of the Eastern Sudan. An Outline. In L. Krzyzaniak, M.
Kobusiewicz and J. Alexander (eds.), Environmental Change and Human Culture
in the Nile Basin and Northern Africa Until the Second Millennium B.C. Poznan,
Poznan Archaeological Museum: pp. 439–448.
Gauthier-Pilters, H. and A. I. Dagg
1981 The Camel: Its Evolution, Ecology, Behaviour, and Relationship to Man. Chicago,
University of Chicago Press.
Grigg, D. B.
1974 The Agricultural Systems of the World: An Evolutionary Approach. Cambridge,
Hasan, Y. F.
1975 Dirasat fi taryikh Al-Suodan: Aljuzu Al-Awal [in Arabic]. Khartoum, Khartoum
University Press.
Helck, W. and E. Otto
1986 Lexikon der Ägyptologie. Volume 6. Wiesbaden, Otto Harrassowitz.
Hjort af Ornäs, A. and G. Dahl
1991 Responsible Man: The Atmaan Beja of North-Eastern Sudan. Stockholm-Uppsala,
Stockholm Studies in Social Anthropology in cooperation with Nordiska
Afrikainstitutet.
Holt, P. M. and M. Daly
1979 The History of the Sudan from the Coming of Islam to the Present Day. London,
Weidenfeld and Nicolson.
Ibn Haugal, A. G.
1979 Surat al-Ard [in Arabic]. Hairout, Al-Haiat Publishing House.
Lewis, I.
1969 People of the Horn of Africa, Somali, Afar, and Saho. London, International
African Institute.
Magid, A.
1998 Ancient Waystations in the Southern Red Sea Hills, Sudan: A New Discovery.
Sudan Notes and Records 15: pp. 5–15.
Magid, A., R. H. Pierce and K. Krzywinski
1995 Test Excavations in the Southern Red Sea Hills (Sudan): Cultural Linkages
to the North. Archéologie du Nil Moyen 7: pp. 163–190.
Manger, L.
1996 General Introduction. In L. Manger, H. Abd el-Ati, S. Harir, K. Krzywinski,
and O. R. Vetaas (eds.), Survival on Meagre Resources: Hadendowa Pastoralism
in the Red Sea Hills. Stockholm, Gotab: pp. 18–36.
Murdock, G.
1959 Africa: Its Peoples and Their Culture History. New York, McGraw-Hill.
Murray, G. W.
1978 Sons of Ishmael: A Study of the Egyptian Bedouin. New York, AMS Press (reprint
of the 1935 manuscript).
Nautrup, B. L.
2004 Gender, Society, and Religion: Changing Everyday Life Among the Hadendowa-
Beja of the Gash Delta, Eastern Sudan. University of Copenhagen, Institute of
Anthropology (PhD dissertation).
Nicolaisen, J.
1963 Ecology and Culture of the Pastoral Tuareg: With Special Reference to the Tuareg of
Ahaggar and Ayr. Copenhagen, National Museum.
Oliver, P.
1971 Shelter in Africa. London, Barrie and Jenkins.
Owen, T. R.
1937 The Hadendowa. Sudan Notes and Records 20: 2, pp. 183–208.
Paul, A.
1971 A History of the Beja Tribes of the Sudan.Cambridge, Cambridge University
Press.
Prussin, L.
1995 African Nomadic Architecture: Space, Place, and Gender. Washington, Smithsonian
Institution Press and the National Museum of African Arts.
Sadr, K.
1991 Development of Nomadism in Ancient Northeast Africa. Philadelphia, University
of Pennsylvania Press.
464     A n wa r A - M a g i d
Seligmann, C. G.
1913 Some Aspects of the Hamitic Problem in the Anglo-Egyptian Sudan. Journal
of the Royal Anthropological Institute of Great Britain and Ireland 63: pp. 593,
705.
Trimingham, J. S.
1949 Islam in the Sudan. London, Oxford University Press.
Updegraff, R. T.
1978 A Study of the Blemmyes. Brandeis University (PhD dissertation).
Vaagenes, V.
1990 Women Going Public: Social Change and Gender Roles in the Red Sea Hills, Sudan.
University of Bergen, Department of Geography (Candidacy thesis).
1998 Women of the Interior, Men of the Exterior: The Gender Order of Hadendowa
Nomads, Red Sea Hills, Sudan. University of Bergen, Department of Geography
(PhD dissertation).
Vantini, G.
1975 Oriental Sources Concerning Nubia. Heidelberg-Warsaw, Society for Nubian
Studies, the Polish Academy of Science and the Heidelberger Akademie der
Wissenschaffen.
Walker, P.
1987 Food for Recovery: Food Monitoring and Targeting in the Red Sea Province, Sudan,
1985–1987. Oxford, Oxfam.
Ch a p t e r 2 1
T he Bedouin T ent
An Ethno-archaeological
Portal to Antiquity or a
Modern Construct?
Benjamin A. Saidel1
O ne objective of ethno-archaeological research is to record human

behavior in traditional communities in order to shed light on ancient
societies. Kramer (1979:1) states that “observations of contemporary behavior
can facilitate the development and refinement of insights into past behav-
iors, particularly when strong similarities can be shown to exist between the
environments and technologies of the past and contemporary sociocultural
systems being compared.”2 In the Old World, ethno-archaeological investi-
gations of Bedouin campsites have been conducted in the southern Levant,
northern Arabia and on the Sinai peninsula (for instance Zarins et al. 1980:23,
Plate 11; Bar-Yosef 1984:155; Banning and Köhler-Rollefson 1986, 1992;
Simms 1988; Avni 1992; Eldar et al. 1992; Zarins 1992; Goren-Inbar 1993).
These projects demonstrated that the campsites of pastoral nomads are
archaeologically visible and that the archaeological signature of tent camps is
influenced by the nature of the surrounding terrain (such as Rosen 1992:76,
1993:443; Saidel 2001). Ethno-archaeological fieldwork carried out on the
tent camps of modern pastoral nomads was performed on the assumption that
1.
An early version of this chapter was presented as part of Symposium 82, Nomads in
Archaeology, during the 69th Annual Meeting of the Society for American Archaeology
(Montreal, 2 April 2004). I thank Anna Belfer-Cohen, Laura Mazow and Steven Rosen
for their incisive comments on the earlier drafts of this chapter. I also thank Edna
Sachar for her meticulous copy editing. This paper was written during my tenure as
the Ernest S. Ferichs fellow at the W. F. Albright Institute of Archaeological Research
in Jerusalem in 2004/2005, and I thank the Albright Fellowship Committee for its
support.
2.
See David and Kramer (2001:12, Table 1.1) for additional definitions of ethno-
archaeology.
465
466    B e n jam i n A . S a i d e l
this research would provide insights into ancient forms of mobile pastoralism
(for example Hole 1979:196, 1980:120–121). In particular, it was assumed
that these structures are comparable in form and function to similarly shaped
tensile structures in antiquity. This perspective is represented, for example, in
a preliminary report of a survey carried out in the central and southwestern
provinces of Saudi Arabia: “This best modern parallel to the post-Neolithic
sites comes from a recently abandoned (last ten years) Bedu tent encamp-
ment in Qatar (al-’Uqda) where multiple stone cobbles were used to clearly
anchor the entire tent walls . . . leaving a clear rectangular tent outline”
(Zarins et al. 1980:23). In this example the basis for drawing the comparison
between the Bedouin tent and post-Neolithic tents is that both types of
structures are rectangular (Zarins et al. 1980:23, Plates 11, 10A, 31A). The
shape of the Bedouin tent, combined with recent ethnographies describing
the sedentarization of the Bedouin, has been used to identify the sedentariza-
tion of nomads in the archaeological record of the early Iron Age I in the
southern Levant (for example Finkelstein 1988:237–258, 1995:46–49). The
methodological approaches outlined above are problematic because no one
has yet considered the age of the Bedouin tent and the customs carried out
inside of this structure. Implicit in much of the research is the premise that
the Bedouin tent is a relic from a primordial past.
In this chapter I use various lines of evidence to propose two possible dates
for the age of the Bedouin ‘black tent.’ The first is based on the organization of
space within this structure; the second is based on some of the material culture
found inside it. The latter is significant as a means to establish the age of some
of the customs and material culture found inside the black tent as described by
Western and non-Western travelers and ethnographers. By dating the mate-
rial culture, we can also provide relative dates for some of the social customs
carried out inside this tent. This study does not investigate the origins of tensile
architecture in the southern Levant or in the Old World. Rather, its purpose is
to ascertain the age of a specific type of tent, namely the Bedouin ‘black tent’
(for example Magid, this volume; Musil 1928:61–76; Dickson 1951:66–80;
Simms 1988), and some of the activities carried on inside it. The geographical
focus of this study is primarily the southern Levant, northern Arabia and the
Sinai, with data also drawn from other parts of the Near East. This research
relates to Bedouin black-tent dwellers who are, or have been defined by others
as, Bedouin (Burckhardt [1831] 1967:3; Musil 1928:xiii, 44; Kirk 1941:60, zone
2; el-Aref [1944] 1974; Simms 1988).
T h e B e d o u i n T e n t      467
Table 21.1. Selected Examples of the Size of Bedouin Tents:
Location or tribe Tent dimensions (m) Reference

or size (m2)
Petra, Jordan 15x4 m Simms 1988:202
Petra, Jordan 51.5 m2 (main tent) Bienkowski and Chlebik 1991:Table 3
Petra, Jordan 18.4 m2 (‘second tent’) Bienkowski and Chlebik 1991:Table 3
Wadi Fatima, Saudi Arabia 10x4 m Katakura 1977:73
Rwala tent without main pole 4–6x2.5–3 m Musil 1928:72
(harbus)
Rwala tent with one main pole 12x3.5 m Musil 1928:72
(katba)
Rwala tent for children or the 4x2 m Musil 1928:72
‘poorest inhabitants of the camp’
(tuzz)
The Bedouin Black Tent

The Bedouin black tent, the beit shaàr or ‘house of hair,’ is defined as a rectan-
gular structure that varies in length and width (Table 21.1) (Weir 1976:1).3 This
type of tent was found throughout the Levant, northern Arabia, and the Sinai,
as well as in other parts of the Near East and North Africa (Andrews 1990).
The Bedouin black tent is primarily used during the winter months because
it is water-repellant and, when it rains, the fabric expands, making the tent
waterproof (Burckhardt [1831] 1967:37, 67; Musil 1928:61; Dickson 1951:66;
Weir 1976:1; Kay 1978:11). During the summer months, those Bedouin who
could afford to purchase a second tent, one made of lighter materials such as
cotton, did so, but ‘poverty’ compelled others to continue to live in the black
goat-hair tent (Jabbur 1995:249). The Bedouin had various means to acquire
the panels of goat hair that were sown together to make a tent. Some Bedouin
women wove their own tent panels; however, this was not possible for the
mobile pastoralists who lived in deserts because goats could not survive in
such arid conditions (Weir 1976:46). Goat-hair panels, or entire tents made of
goat hair, were purchased by the Bedouin from towns or itinerant merchants
(Musil 1928:61; Dickson 1951:73; Weir 1976:49). Many towns in the Levant
manufactured Bedouin tents. According to Weir (1976:49), during the British
Mandate in Palestine, tents were manufactured in Safad, Majd al-Kurum,
Samakh, Beisan, Anabta, Tulkarm, Nablus and Hebron. Tents were also
3.
It is acknowledged that some Bedouin, such as the Jebaliya, use tents that are not
rectilinear and that during the summer they do not use the beit shaàr at all because it is
too hot (Bar-Yosef 1984:155; Goren-Inbar 1993:418). Aside from tents, the Bedouin use
a variety of other types of spaces for shelter, such as caves and rock shelters (Magid, this
volume; Bienkowski 1985; LaBianca 1990:83, 87, 222, 232; Simms and Russell 1997).
made in the village of Shhim, in Lebanon (Weir 1976:49; Jabbur 1995:250).

According to Jabbur (1995:250) a major tent-producing center was located in
the Syrian town of Yabrūd: “In this small town there were about 2000 people
who earned their living making tents. There are about 300 looms there, and
these supply the Bedouins of Syria and export to Jordan, Iraq, Najd, Kuwayt,
and the oil companies.”
Bedouin tents are often physically divided, by using a curtain or rug, into two
sections; one for men (al-shigg) and one for women (al-mahram) (Burckhardt
[1831] 1967:40; Musil 1928:64; Dickson 1951:71–73; Cole 1975:64–65; Weir
1976:17). The men’s, al-shigg, side of the tent is used primarily for entertaining
guests and other tribesmen (Musil 1928:64, 97). Items found in this area include
a hearth, an assortment of rugs and cushions and camel saddles for sitting and
reclining. One of the more important activities carried out on this side of the tent
is the preparation of coffee for consumption by guests and friends (Musil 1928:
100–102; Dickson 1951:195–201; Weir 1976:7–10). In general, the men’s section
of the black tent contains fewer material culture items than the women’s side.
Most household chores like cooking, however, are carried out in the women’s,
al-mahram, side of the structure, where domestic items and baggage are stored
(Table 21.2) (Doughty [1888] 1979:267–268; Musil 1928:100; Weir 1976:17).
Lancaster (1997:61) observes that the division of the Bedouin tent into two parts
is a reflection of the duality of Bedouin society: the men’s part of the tent is asso-
ciated with the public aspects of society, whereas the women’s side is restricted
to ‘private’ aspects of Bedouin life. The gendered division of space inside the
Bedouin tent is attributed to Islamic beliefs and customs that determine how
space is used and where activities are conducted (Insoll 1999:62, 72, 90).
Table 21.2. Objects Found in the Bedouin Tent as Recorded by Burckhardt, Musil and Dickson
Reference Found in al-shigg (men’s section)a Found in al-mahram (women’s

section)a
Burckhardt (1967) Camel bags and packsaddles (for Butter, cooking utensils, water skins
comfort), Persian carpet, wheat sacks
(cushion)
Musil (1928:64) Camel saddle, pillows Barley, butter, carpets, copper
utensils, coffee, coverlets, dates,
litters, loom, rice, sâğ (iron sheet for
baking bread), salt, spindle, sugar,
wheat, wooden utensils
Dickson (1951:76–77) Carpet, case for coffee cups, Blankets, camel saddles, cardamom,
coffeepots, incense burner, mattresses, coffee, cooking utensils, dates,
men’s camel saddle, mortar and pestle flour, hand loom, litters, rice, rifle
for coffee, peg for stirring coffee ammunition, spare rifle, salt silks,
beans, pillows, rifle spindles, sugar
Only major material culture items are presented.

a
T h e B e d o u i n T e n t      469
Concerning this division of space within, Insoll (1999:62) writes: “The

primary and over-riding concern is with privacy and the protection and seclu-
sion of women (in certain cultural contexts referred to as purdah) and the sanctity
of the family. Both wife and domestic space are to be protected, and domestic
life is linked to ideas of purity.” In this context there is a well-defined etiquette
that describes how a stranger should approach a Bedouin tent, as recognized by
Bedouin and Western travelers alike (for example Doughty [1888] 1979:39–40;
el-Aref [1944] 1974:108; Diqs 1969:18). El-Aref ([1944] 1974:136–137), for
example, describes how a guest should behave on entering a Bedouin tent:
There are strict rules of social conduct to which the guest must conform if he
would not be considered ill mannered and undeserving of kindness. Facing the
entrance to a Bedawi tent, the hareem is always on the right. On the left of the
tent is the Shigg where the guest is entertained. It is very distasteful and offensive
to a Bedawi to have another man even pass closely to his harem, therefore it is
incumbent on a visitor to approach the tent from the rear and to keep the tent
on his left as he comes to the front. That is rule (1).
Rule (2) is that he shall not take undue notice or peer at a woman moving about
in the tent. He should appear to ignore their presence.4
Given the gendered division of space within the Bedouin tent, Musil (1928:64)
recorded an anecdote that demonstrates how the Rwala manipulated these
rules to disarm their guests when they entered the men’s section of the tent:
“The master of the tent and his guests remove their shoes or boots, which
they place behind them against the dividing wall, while they hang their rifles
on the main pole in the women’s compartment.” Although the interpreta-
tion is speculative, this behavior appears to be one mechanism for defusing
potential tensions between host and guests: while their shoes are kept in the
men’s section, their firearms are stored in the women’s part of the tent, which
is off-limits. If a violent disagreement arose between a host and his guests their
rifles would, at least theoretically, be inaccessible. I suggest that the Bedouin
black tent as we know it today emerged some time after the mid 7th century
CE, that is to say, following the Islamic conquest of Palestine. This terminus
post quem is based on the division of space and gender roles assigned to men
and women in Islamic societies.
4.
Aref el-Aref was born in Jerusalem and educated in Istanbul. During the period
of the British Mandate in Palestine he held various administrative posts, including
governor of Beersheba (Abu-Rabià 2001:86–87). As a result, he became well acquainted
with the Bedouin who lived in the Negev and wrote a number of books on this popula-
tion (such as el-Aref 1938, [1944] 1974; see also Abu-Rabià 2001:87).
Coffee, Tobacco and Pottery

Another means to determine the age of the Bedouin tent is to establish relative
dates for some of the objects found inside these structures. In the following
section I propose relative dates for the accoutrements associated with coffee
consumption and tobacco smoking among the Bedouin. Both activities are
well attested in Western travelogues and in the ethnographic literature (such
as Conder 1879:282). The presence of pottery known as Gaza Ware is another
line of evidence that can be used to date Bedouin tents.5 According to historic
and ethnographic accounts, coffee is an important part of Bedouin society.
Coffee is the only foodstuff prepared in the men’s part of the Bedouin tent,
and the equipment used to prepare it is usually stored there. The preparation
of fresh coffee is a hallmark of hospitality among the Bedouin (Doughty [1888]
1979:289, 290; Musil 1928:469; Dickson 1951:195; Weir 1976:10). Both sexes
in Bedouin society consume coffee (Musil 1928:100; el-Aref [1944] 1974:36;
Dickson 1951:76–77, 84, 198–199). The social value of coffee is illustrated
by the accoutrements used for coffee preparation and consumption (Musil
1928:100), as well as by the many poems extolling the virtues of this beverage
(Musil 1928:102–114). The grinding of coffee beans is also considered an
art and a means of personal expression: “The noise made by the mortar and
pestle, hess al-mihbas, is heard everywhere in the vicinity, and the people give
their opinion as to whether these sounds are regular and artistic or not. The
pounding of coffee is an art, and musical ability is judged according to the way
in which it is done” (Musil 1928:101).
Information contained in travelogues and ethnographies indicates that both
Bedouin of humble means and tribal leaders spent considerable portions of their
income on coffee consumption. One travel account from the early 19th century
CE indicates that one Bedouin spent as much as 29% of his yearly income on
luxury goods that included coffee (Burckhardt [1831] 1967:70).6 The Rwala
leader, An-Nûri eben Ša`lân, spent an appreciable portion of his income on
the purchase of coffee and sugar (Table 21.3). Dickson (1951:78) notes that the
Bedouin in Kuwait hide their coffee with other valuables: “Treasures such as
money, coffee beans, cardamom, sugar, salt, silks and special holiday attire are
kept by the housewife locked up in a small tin or wooden box. The key of this she
always keeps on her person, and tied to a portion of her head veil or milfa.”
5.
Firearms are problematic chronological indicators because some Bedouin continued
to use obsolete weapons for economic or social reasons, while others had the means to
procure modern rifles (Saidel 2000:197–213). They are therefore excluded from this
discussion.
6.
Other luxuries purchased included a “dried apricot jelly” and “a sweet jelly made
of grapes” (Burckhardt [1831] 1967:70).
T h e B e d o u i n T e n t      471
Table 21.3. Prices of Foodstuffs by ‘Load.’
Foodstuffa Cost per ‘load’ (in $) Amount spent (in $)

Butter (no number of loads given) 247.50 247.00
Burrul (husked wheat) 1.60 11.25
Camel (for consumption) 45.00 225.00
Coffee 60.00 90.00
Flour 36.00 1440.00
Rice 6.00 18.00
Sugar 12.50 25.00
Sheep (for consumption) 2.70 270.00
Wheat 0.68 13.50
a
Livestock purchased for human consumption is included for comparison; foodstuffs for animals are
excluded (data drawn from Musil 1928:59).
The introduction of coffee to the Middle East provides a relative date for the
introduction of this beverage to Bedouin society. The use of coffee as a beverage
did not become widespread in the Ottoman Empire until the 17th century CE
(Arendonk 1978:450–451; Baram 1996:120, 1999:140). According to Baram
(1999:142), by 1610 CE, coffee and coffeehouses were present throughout
the Ottoman Empire. Because coffee is a stimulant, there were times during
the 17th century CE when consumption was banned on religious grounds
(Arendonk 1978:451; Baram 1999:141; Insoll 1999:106). Archaeologists have
interpreted the finds of small porcelain coffee cups at archaeological sites in
the southern Levant as evidence of the import and consumption of coffee in
this region (for example Edelstein and Avissar 1997:133; Baram 2000:147, 154;
Boas 2000:553–554; Ward 2000:189–190; Kletter 2004:198–200). Therefore,
the dating of the objects necessary to prepare coffee, and its social role within
Bedouin society, cannot be earlier than the 17th century CE. This only applies
to the use of coffee, however, and does not mean that other (hot) beverages
were not offered as expressions of hospitality
Tobacco, another stimulant used by both men and women in Bedouin
society, was exported from the New World, also during the 17th century CE
(Von Gernet 1995:79; Baram 2000:149). The adoption and use of tobacco
among the Bedouin is well documented for the 19th and early 20th centuries CE
(Doughty [1888] 1979:355–356; Musil 1928:127; el-Aref [1944] 1974:36). They
acquired tobacco from various sources, by means of purchasing it in markets
in towns or from itinerant tobacconists, casual exchange and personal cultiva-
tion (Doughty [1888] 1979:355, 487, 525; el-Aref [1944] 1974:36). Like coffee,
tobacco was often a subject for Rwalan poetry (Musil 1928:128–131).
In the Ottoman period the Bedouin acquired tobacco pipes both through
purchase and by making their own from local stone. According to information
contained in travel accounts and provided by pipe collectors, the Bedouin in
northern Arabia and Sinai crafted their own pipes (Doughty [1888] 1979:288;
Baram 1996:192, 193, figure 9). Citing Musil, Baram (1996:192–194) points
out that male and female smokers among the Rwala Bedouin use different
types of tobacco pipes. Tobacco pipes, dating from the 17th–20th centuries CE,
have been found, in archaeological surveys, along the coast of northern Sinai
(Cytryn-Silverman 1996:147–149) and in the Negev Highlands. Examples of
the latter, however, are either poorly illustrated or not illustrated at all, making
it impossible to date them at this time. Given the origin of tobacco, and the
archaeological evidence of tobacco smoking, the Bedouin could not have
adopted it prior to the 17th century CE.
Sherd scatters of coarse gray pottery identified as ‘Gaza Ware’ have been
found at many abandoned Bedouin tent camps in the Negev and Sinai. Such
sherd scatters at archaeological sites in the Negev Highlands constitute the
only evidence that these settlements were re-used by Bedouin (Rosen 1981,
1994:21–22). Dating Gaza Ware is problematic because it has not been the
subject of systematic typological and chronological research. Current studies
indicate that the production of Gaza Ware may have begun as early as the
second half of the 17th century CE or the beginning of the 18th century CE
and that it continued to be manufactured until the end of the 20th century CE
(Rosen and Goodfriend 1993; Ziadeh 1995:211, 220; Bulle and Marmiroli
2000:93). This suggests that the abandoned Bedouin tent camps situated in the
Negev Highlands should be dated to the 17th–20th centuries CE.7
7.
Legal definitions of what constitutes an archaeological site vary, and the implica-
tions are significant for the preservation of abandoned tent campsites. In Israel, remains
of human activities that date to before 1700 CE are considered archaeological sites and
are accordingly given legal protection (Rosen and Goodfriend 1993:147). Publication
of archaeological materials dating after 1700 CE is dependent on each excavator. For
example, the excavation of archaeological deposits in Jaffa from the “late nineteenth to
early twentieth centuries” was not published because the sediments and artifacts are not
“considered as antiquities by Israeli law” (Kletter 2004:205–206). In contrast, Ustinova
and Nahshoni (1994:176) published the results of their excavations of farmsteads located
in Beersheva, which they dated from the 1930s to the late 1940s CE. In contrast to the
antiquities laws in Israel, the Archaeological Resources Protection Act of 1979 in the
United States of America defines an archaeological site as any traces of human activities
that are at least 100 years old.
T h e B e d o u i n T e n t      473
The Bedouin Tent in Archaeological Context

Ethno-archaeological investigations in the southern Levant have identified the
archaeological remains of recently abandoned tent camps (for example Simms
1988; Avni 1992:242–247; Saidel 2001). At present there are no published exam-
ples of excavated Bedouin tent camps in the southern Levant that predate the
20th century CE. The best published example of a Bedouin tent and tent camp
in an archaeological context is Ras Abaruk Site 5 in Qatar (Garlake 1978). This
site contains the outlines of at least six tents. The pottery from surface collec-
tion and test excavation indicates that occupation at this settlement spanned the
17th–19th centuries CE (Garlake 1978:166). Although surveys in the Negev have
documented the presence of Bedouin tents in this region, no Bedouin tents dated
to the 6th–18th centuries CE have as yet been excavated. However, Gaza Ware
pottery at these tent sites, as well as in locations where there are no architectural
remains of tents, is visible evidence of Bedouin occupation from the 17th–20th
centuries CE (Rosen 1981, 1994:21–22). Archaeological and historical data
have been used to posit a Bedouin presence in the northwestern Negev during
the Mamluk period (Schaefer 1989:54–56). Sherd scatters of Mamluk pottery,
hearths and a lack of architectural remains are interpreted as evidence that these
‘camps’ were occupied by Bedouin (Gazit 1996:18). The data are problematic,
however, as many of the Mamluk campsites have evidence of multiple-period
occupations (Gazit 1996:40–41, 50). Caution needs to be exercised in the iden-
tification of these sites as ‘Bedouin camps,’ because the site formation processes
may be misleading. Citing this problem in relation to Chalcolithic sites in the
northwestern Negev, Gilead (1992:31) writes, “The dried mudbricks were made
of local silt, identical to the natural soil. When they decay, it is sometimes impos-
sible to separate them from the natural soil. This may give the wrong impression
that such sites featured no architecture, and were, therefore, nomadic camps.”
Therefore, the ‘Bedouin campsites’ identified by Schaefer (1989) need to be
investigated further to determine whether they are indeed the remains of tent
camps or rather deflated farmsteads. Stone outlines of ancient and modern recti-
linear tents were documented at the sites of Har 'Oded and Nahal 'Oded, located
to the south of the Ramon Crater (Rosen and Avni 1997:22–23, Figure 4.2, 44,
54, 55, Figure 5.12, Figure 5.13, 59, 70). At both of these sites pottery attributed
to the 6th–8th centuries CE was found adjacent to the stone outlines of rectilinear
tents. This is tantalizing evidence for the emergence of the Bedouin black tent.
The best example of rectilinear tents in an archaeological context in the
southern Levant is represented at the site of Givòt Reved in the Ramon Crater.
Rosen’s (1993:448) research at this site identified a series of rectangular tents,
arranged in a row, dated to the second and third centuries CE. While the stone
foundations of these structures do provide evidence of the existence of rectilinear
tents in pre-Islamic periods, their presence in these archaeological contexts does

not constitute evidence for the Bedouin tent per se as it remains to be determined
if the use of space inside both types of structures is similar or dissimilar to one
another (Doughty [1888] 1979:267–268; Musil 1928:64–68, 97, 100; Insoll
1999:62, 72, 90).8 The results of such research, however, indicate a relatively
shallow time-depth for the Bedouin tent and its related material culture and has
immediate methodological implications for archaeological and ethno-archaeo-
logical research in the southern Levant, as well as in the wider Middle East.
From an archaeological perspective our knowledge of the antiquity of the
Bedouin tent is hindered by several issues. Scholars working in the southern
Levant have traditionally not been interested in conducting archaeological
investigations at sites from the Ottoman period (1515–1917 CE), although this
no longer is the case (for instance Ziadeh 1995; Schick 1997–1998; Baram and
Carroll 2000). The paucity of archaeological research on historic and ancient
forms of pastoralism can also be attributed to the commoditization of archaeo-
logical sites and data as sources of prestige for scholars and their institutions.
In practical terms this is illustrated in at least four forms that are neither mutu-
ally exclusive nor the sole purview of any subdiscipline of archaeology. First,
the excavation of large urban sites, by virtue of their size, is more prestigious
than fieldwork carried out at smaller settlements, be they farmsteads, forts or
pastoral encampments. Second, importance is often attached to sites, regardless
of their size, that are mentioned or described in documents and literature from
antiquity. Third, prestige is bestowed on some archaeological sites because they
were excavated in the early 20th century CE by a prominent archaeologist and
therefore merit further investigation. Fourth, inherent importance is attached
to research that makes a contribution, or is perceived to make a contribution, to
subjects that are deemed ‘sexy’ in Western civilization, such as the biblical flood
story (for example Wilford 1999, 2001; Leary 2000).9 These forms of prestige
are projected in scholarly and popular media for academic and lay audiences.
As Homan (2002:4) notes, the habitations of pastoral nomads, be they from
prehistoric or historic periods, are generally not of interest to archaeologists:
“even when practicable, excavating tent fragments and elliptical settlement
patterns is less romantic and fruitful in both publications and fundraising than
8.
Archaeological research conducted in the Negev Highlands demonstrated that
most tents from the Chalcolithic through the Early Arab periods were circular (Rosen
1984:119, 1994:18-20; Haiman 1989; see also Magid, this volume).
9.
This phenomenon is not unique to Old World archaeology or sub-branches of
Near Eastern archaeology like biblical archaeology. For an example of the commodifi-
cation of archaeological research in the New World, consider the title of Adovasio and
Page’s (2003) book: The First Americans: In Pursuit of Archaeology’s Greatest Mystery.
T h e B e d o u i n T e n t      475
digging a massive urban center.”10 In other words, it appears that archaeological

research on mobile pastoralists is perceived to generate less academic currency
than fieldwork carried out at ancient urban centers.
The Ethno-archaeological Value

of the Bedouin Tent
In archaeological studies of the southern Levant some scholars have argued
that the pillared houses of the Iron Age I, also known as the ‘Israelite
four-room house’ (Figure 21.1), are copies in stone of the Bedouin tent
(Figure 21.2), with the pillars mirroring the position of the wooden tent poles
(Herzog 1984:76–77; Finkelstein 1988:257). This analogy is based on three
factors. First, the inhabitants of these sites were Israelite pastoral nomads
in the process of sedentarization (Herzog 1984:76–77, 82–83; Finkelstein
1988:74–80, 238). Second, ethnographic information is cited indicating that
when the Bedouin construct houses made of stone or cement, the layout of
these homes mirrors the design of their tents: “It is widely accepted that
nomads in the process of sedentarization retained traditions from their
pastoral existence, at least initially. An obvious example is the transference of
their tradition of dwelling in portable tents made of perishable materials to
their permanent architecture” (Finkelstein 1988:245). Third, it is proposed
by Finkelstein (1988:248) that the rectilinear shape of the four-room house
and its division of space are based on the Bedouin tent: “From our hypothesis
that the elliptical site originated in the nomadic encampment, it follows that
the individual unit of construction—a broadroom or ‘casemate’—reflected the
individual desert tent. In this connection, it should be stressed that the tradi-
tion of the tent shape was apparently even stronger than stone construction,
for it was deeply rooted in centuries of an unchanging lifestyle and consistent
geographical setting . . . It is therefore unlikely that the shape of the desert
tent in our region was altered over the course of time.”
The parallel drawn between the four-room house and the Bedouin tent is
methodologically flawed. There are no similarities between the layout and
use of space inside the Bedouin tent and in the pillared houses of the Iron
Age I period in the southern Levant other than that both structures are rect-
angular. Furthermore, the division of space inside Bedouin tents, and in their
10.
It should also be mentioned, however, that the publication of fieldwork carried out
at pastoral encampments is relatively easy, given the limited amount of material culture
and small size of these settlements. Moreover, in the arid zones of the southern Levant
it is also relatively easy to draw the architectural plan of an entire mobile-pastoralist site,
which cannot be said of tells and other sites containing complex stratigraphy.
Figure 21.1. An early Iron Age I four-room house in Fields A and B at Tall al-'Umayri in Jordan
(redrawn and modified after Clark 2003:36). The function of the interior spaces follows Clark’s (2003:37)
reconstruction of this building as a two-story structure. The activities conducted on the ground floor
(GF) are indicated in regular font; those carried out on the upper floor (UF) are in italics.
homes made of durable materials, such as mud brick (Figure 21.3), are totally
different from the layout of the four-room house. The analogy proposed by
scholars such as Finkelstein also does not take into consideration that the
division of space within the Bedouin tent and the manner in which it is used
are directly impacted by gender relations and social distances predicated on
Islamic sensitivities.
The implications of this for the ethno-archaeological research on Bedouin
tents are profound. Based on information contained in historic and ethnohis-
toric sources, it is apparent that tent camps recorded by ethno-archaeologists
over the past 20 years have few similarities in size and layout to Bedouin
encampments from the 19th–20th centuries CE. At the beginning of the 19th
T h e B e d o u i n T e n t      477
Figure 21.2. Plan of a Bedouin tent in Wadi Fatima, Saudi Arabia (redrawn and modified after Katakura
1977:75). The interior contents include (1) mats; (2) a manual sewing machine; (3) a closet box; (4) a
loom; (5) lamps; (6) an incense burner; (7) a coffeepot; (8) kitchen utensils; (9) a pen made of thorny
branches for sheep and goats; (10) a twig shelter to cover leather water bags.
century CE the campsites of the Aeneze, of which the Rwala were a subtribe,
varied in size from 10–800 tents (Burckhardt [1831] 1967:32). By the turn
of the 20th century CE, the Rwala distinguished three sizes of tent camps:
those comprising fewer than ten structures (ferîz), those containing 10–29
tents (neğe), and those exceeding thirty tents (nezel) (Musil 1928:77). Ethno-
archaeological research conducted in Israel, Jordan and Egypt has seldom
documented tent camps comprising more than two contemporaneous tents
(Banning and Köhler-Rollefson 1992; Eldar et al. 1992).11
Travel accounts and ethnographies from the Ottoman period demonstrate
that Bedouin tent camps also changed in layout. By the beginning of the 19th
century CE, Aeneze tent camps were arranged in both circular and linear
Studies of larger tent camps have been conducted by Cribb (1991) in Turkey and
11.
by Evans (1983) in Iran.

Figure 21.3. Plan of a Bedouin ‘clay house’ in Wadi Fatima, Saudi Arabia (redrawn and modified after
Katakura 1977:75). The following domestic items are located in the open courtyard and in the rooms:
(1) mats; (2) a manual sewing machine; (3) a closet box; (4) kitchen utensils; (5) lamps; (6) water vessels;
(7) a chicken coop; (8) a pen made of thorny branches for sheep and goats.
alignments. In some instances tents were arranged in multiple rows, as many

as three or four tents deep. Historic and ethnographic sources indicate that
the layout of the tents in Bedouin encampments was influenced by two factors:
animal husbandry and security. Bedouin who raised sheep and goats tended to
arrange their tents in the shape of an ellipse, whereas those who herded camels
pitched their tents in parallel lines (Musil 1928:180). Musil (1928) observed, at
the turn of the 20th century CE, that the layout of Rwala tent camps was influ-
enced by security conditions. In areas with no threat of physical attack tents
were scattered or arranged in parallel rows. When the Rwala were expecting
an attack, tents were grouped together to form larger concentrations, although
Musil does not specify shape.
None of this information is hinted at in the current ethno-archaeological
literature as this describes a relict population of pastoralists whose spatial
T h e B e d o u i n T e n t      479
organization of tent camps has changed significantly over the past 110 years
and bears little resemblance in plan to the tent camps from the Ottoman and
Mandate periods. A major problem with ethno-archaeological research on
the Bedouin is the presumption that the Bedouin black tent and tent camps
are unchanging relicts from a chronologically undefined past. What is often
overlooked, however, is that the lifestyle, economy and settlement patterns
of the Bedouin have been impacted by political events and governmental
programs aimed at controlling and settling them since the end of the 19th
century CE and continuing to the present day (for example Marx 1967:6,
Map 1, 36–58; Lewis 1987:124–147; Perevolotsky et al. 1989; Bienkowski
and Chlebik 1991:162–177; Kressel and Ben-David 1995:119, 138; Rogan
1999:180–182, 249–250).
Discussion
The information presented above indicates a relatively shallow time-depth
for the Bedouin tent. At this stage of research it appears that the division
of space within this structure should be attributed to gender relations in
Islam, suggesting that the use of space inside the Bedouin black tent most
likely developed sometime after the 7th century CE. Based on the fact that
the adoption of coffee, tobacco and pottery by the Bedouin in the southern
Levant, and the activities and social customs associated with them, cannot be
older than the 17th–18th centuries CE, the parallels drawn by Finkelstein, and
others, between the four-room house of Iron Age I Palestine and Transjordan
and the Bedouin tent are baseless because the main evidence for this analogy
is that both structures are rectilinear. The Bedouin of the past 200 years are
a powerful icon for ethno-archaeologists working in the southern Levant and
wider Near East, regardless of their academic training and discipline (Bernbeck,
this volume). In a number of publications on the Bedouin, the first chapter
is devoted to a description of the Bedouin tent (for instance Weir 1976:1–6;
Kay 1978:11–32). A common observation on the Bedouin made by 19th–20th
century CE travelers and archaeologists was that, except for their use of coffee,
tobacco and firearms, the Bedouin were essentially living fossils (for example
Conder 1879:282; Glueck 1970:11). Ironically, ethno-archaeological research
on these populations has fallen into the same seductive trap. In reality, dispari-
ties in the size and shape of contemporary Bedouin tent camps, as recorded by
ethno-archaeologists, with those from the19th–20th centuries CE demonstrate
that Bedouin encampments in the southern Levant and northern Arabia have
changed significantly over the past 100 years.
References
Abu-Rabià, A.
2001 A Bedouin Century: Education and Development Among the Negev Tribes in the
Twentieth Century. New York, Berghahn Books.
Adovasio, J. M. and J. Page
2003 The First Americans: In Pursuit of Archaeology’s Greatest Mystery. New York,
Modern Library.
Andrews, P. A.
1990 Vorderer Orient Nomadenzeltformen. Wiesbaden, Dr. Ludwig Reichert
Verlag.
Arendonk, C. van
1978 Kahwa. In E. Donzel, B. Lewis and C. Pellat. The Encyclopedia of Islam (New
Edition. Volume 4: IRAN-KHA. Leiden, E. J. Brill: pp. 449–453.
Avni, G.
Implications for Archaeological Research. In O. Bar-Yosef and A. M. Khazanov
Banning, E. B. and I. Köhler-Rollefson
1986 Ethno-archaeological Survey in the Beda Area, Southern Jordan. Zeitschrift
des Deutschen Paläestina-Vereins 102: pp. 152–170.
1992 Ethnographic Lessons for the Pastoral Past: Camp Locations and Material
Remains Near Beidha, Southern Jordan. In O. Bar-Yosef and A. M. Khazanov
Baram, U.
1996 Material Culture, Commodities, and Consumption in Palestine, 1500–1900.
University of Massachusetts, Amherst (PhD dissertation).
1999 Clay Tobacco Pipes and Coffee Cup Sherds in the Archaeology of the Middle
East: Artifacts of Social Tension from the Ottoman Past. International Journal
of Historical Archaeology 3: pp. 137–151.
2000 Entangled Objects from the Palestinian Past: Archaeological Perspectives
for the Ottoman Period, 1500–1900. In U. Baram and L. A. Carroll (eds.), A
Historical Archaeology of the Ottoman Empire: Breaking New Ground. New York,
Kluwer: pp. 137–159.
Baram, U. and L. Carroll (eds.)
2000 A Historical Archaeology of the Ottoman Empire: Breaking New Ground. New
York, Kluwer.
T h e B e d o u i n T e n t      481
Bar-Yosef, O.
1984 Seasonality Among Neolithic Hunter-Gatherers in Southern Sinai. In J.
Clutton-Brock and C. Grigson (eds.), Animals and Archaeology. Volume 3:, Early
Herders and Their Flocks. BAR International Series 202. Oxford, Archaeopress:
pp 145–160.
Bienkowski, P.
1985 New Caves for Old: Beduin Architecture in Petra. World Archaeology 17: pp.
149–160.
Bienkowski, P. and B. Chlebik
1991 Changing Places: Architecture and Spatial Organization of the Bedul in Petra.
Levant 23: pp. 147–180.
Boas, A. J.
2000 Pottery and Small Finds from the Late Ottoman Village and the Early Zionist
Settlement. In Y. Hirschfeld (ed.), Ramat Hanadiv Excavations: Final Report of
the 1984–1998 Seasons. Jerusalem, Israel Exploration Society: pp. 547–580.
Bulle, S. and B. Marmiroli
2000 Gaza, une ville en Devenir. In J.-B. Humbert (ed.), Gaza Méditerranéenne:
Histoire et Archéologie en Palestine. Paris, Ministère des Affaires Étrangères:
pp. 87–100.
Burckhardt, J. L.
1967 Notes on the Bedouins and Wahabys: Collected During His Travels in the East
(Reprinted Edition). New York, Johnson Reprint (originally published in
1831).
Clark, D. R
2003 Bricks, Sweat, and Tears: The Human Investment in Constructing a “Four
Room” House. Near Eastern Archaeology 66: pp. 34–43.
Cole, D. P.
1975 Nomads of the Nomads: The Al Murrah Bedouin of the Empty Quarter. Chicago,
Aldine.
Conder, C. R.
1879 Tent Work in Palestine: A Record of Discovery and Adventure. London, Palestine
Exploration Fund.
Cribb, R.
Cytryn-Silverman, K.
1996 The Islamic Period in North Sinai: The Pottery Evidence. Hebrew University
of Jerusalem (Master’s thesis).
Dickson, H.
1951 Arab of the Desert: A Glimpse into Bedouin Life in Kuwait. London, George Allen
and Unwin.
Diqs, I.
1969 A Bedouin Boyhood. New York, Praeger.
Doughty, C. M.
1979 Travels in Arabia Deserta (Reprinted Edition). New York, Dover. (originally
published in 1888).
Edelstein, G. and M. Avissar
1997 A Sounding in Old Acre. 'Atiqot 31: pp. 129–136.
el-Aref, A.
1938 Die Beduinen von Beerseba: Ihre Rechtsverhältnisse, Sitten, und Gebräuche. Luzern,
Verlag Räber und CIE.
1974 Bedouin Love, Law, and Legend: Dealing Exclusively with the Badu of Beersheva
(Reprinted Edition). Jerusalem, Cosmos Publishing Company (originally
published in 1944).
Eldar, I., Y. Nir and D. Nahlieli
1992 The Bedouin and Their Campsites in the Dimona Region of the Negev:
Comparative Model for the Study of Ancient Desert Settlements. In O.
205–217.
Evans, C.
1983 On the Jube Line: Campsite Studies in Kurdistan. Archaeological Review from
Cambridge 2: pp. 67–77.
Finkelstein, I.
1988 The Archaeology of the Israelite Settlement. Jerusalem, Israel Exploration
Society.
1995 Living on the Fringe: The Archaeology and History of the Negev, Sinai, and
Neighbouring Regions in the Bronze and Iron Ages. Sheffield, Sheffield Academic
Press.
Garlake, P.
1978 An Encampment of the Seventeenth to the Nineteenth Centuries on Ras
Abaruk, Site 5. In B. de Cardi (ed.), Qatar Archaeological Report: Excavations
1973. Oxford, Qatar National Museum: pp. 164–171.
Gazit, D.
1996 Map of Urim. Archaeological Survey of Israel 125. Jerusalem, Israel Antiquities
Authority.
Gilead, I.
1992 Farmers and Herders in Southern Israel During the Chalcolithic Period. In O.
T h e B e d o u i n T e n t      483

29–41.
Glueck, N.
1970 The Other Side of the Jordan. Cambridge, American Schools of Oriental
Research.
Goren-Inbar, N.
1993 Ethno-archaeology: The Southern Sinai Bedouin as a Case Study. In A.
Biran and J. Aviram (eds.), Biblical Archaeology Today: Proceedings of the Second
International Congress on Biblical Archaeology, Jerusalem, June–July 1990.
Jerusalem, Israel Exploration Society, Israel Academy of Sciences and
Humanities: pp. 417–419.
Haiman, M.
1989 Preliminary Report of the Western Negev Highlands Emergency Survey.
Israel Exploration Journal 39: pp. 173–191.
Herzog, Z.
1984 Beer-Sheba II: The Early Iron Age Settlement. Tel Aviv, Tel Aviv University,
Hole, F.
1979 Rediscovering the Past in the Present: Ethnoarchaeology in Luristan, Iran. In
C. Kramer (ed.), Ethnoarchaeology: Implications of Ethnography for Archaeology.
New York, Columbia University Press: pp. 192–218.
1980 The Prehistory of Herding: Some Suggestions from Ethnography. In M.-T.
Barrelet (ed.), L’Archéologie de l’Iraq: Du début de l’époque néolithique à 333 avant
notre ère: Perspectives et limites de l’interprétation anthropologique des documents.
Paris, Centre National de la Recherche Scientifique: pp. 119–127.
Homan, M. M.
2002 To Your Tents, O Israel! Leiden, E. J. Brill.
Insoll, T.
1999 The Archaeology of Islam. Malden, Blackwell.
Jabbur, J. S.
1995 The Bedouins and the Desert: Aspects of Nomadic Life in the Arab East. Albany,
State University of New York Press.
Katakura, M.
1977 Bedouin Village: A Study of a Saudi Arabian People in Transition. Tokyo, University
of Tokyo Press.
Kay, S.
1978 The Bedouin. New York, Crane, Russak.
Kirk, G. E.
1941 The Negev, or Southern Desert of Palestine. Palestine Exploration Quarterly
73: pp. 57–71.
Kletter, R.
2004 Jaffa, Roslan Street. 'Atiqot 47: pp. 193–207.
Kramer, C.
1979 Introduction. In C. Kramer (ed.), Ethnoarchaeology: Implications of Ethnography
for Archaeology. New York, Columbia University Press: pp. 1–20.
Kressel, G. M. and J. Ben-David
1995 The Bedouin Market-Corner Stone for the Founding of Be’er-Sheva: Bedouin
Traditions About the Development of the Negev Capital in the Ottoman
Period. Nomadic Peoples 36/37: pp. 119–144.
LaBianca, Ø. S.
1990 Sedentarization and Nomadization: Food System Cycles at Hesban and Vicinity in
Transjordan. Berrien Springs, Institute of Archaeology and Andrews University
Press.
Lancaster, W.
1997 The Rwala Bedouin Today. Prospect Heights, Waveland Press (second
edition).
Leary, W. E.
2000 Found: Possible Pre-Flood Artifacts. New York Times, 13 September: p. A4.
Lewis, N. N.
1987 Nomads and Settlers in Syria and Jordan, 1800–1980. Cambridge, Cambridge
University Press.
Marx, E.
1967 Bedouin of the Negev. ManchesterUniversity of Manchester.
Musil, A.
1928 The Manners and Customs of the Rwala Bedouins. New York, American
Geographical Society.
Perevolotsky, A., A. Perevolotsky and I. Noy-Meir
1989 Environmental Adaptation and Economic Change in a Pastoral Mountain
Society: The Case of the Jabaliyah Bedouin of the Mt. Sinai Region. Mountain
Research and Development 9: pp. 153–164.
Rogan, E. L.
1999 Frontiers of the State in the Late Ottoman Empire: Transjordan, 1850–1921.
Cambridge, Cambridge University Press.
Rosen, S. A.
1981 Observations on Bedouin Archaeological Sites near Maàleh Ramon, Israel.
Forum for Middle East Research in Anthropology 5: pp. 11–14.
1984 Kvish Harif: Preliminary Investigation at a Late Neolithic Site in the Central
Negev. Paléorient 10: 2, pp. 111–121.
of the American Schools of Oriental Research 287: pp. 75–85.
T h e B e d o u i n T e n t      485
1993 A Roman-Period Pastoral Tent Camp in the Negev, Israel. Journal of Field
1994 Map of Makhtesh Ramon. Archaeological Survey of Israel 204. Jerusalem, Israel
Antiquities Authority.
Rosen, S. A. and G. Avni
1997 The 'Oded Sites: Investigations of Two Early Islamic Pastoral Camps South of the
Ramon Crater. Beersheva, Ben-Gurion University of the Negev Press.
Rosen, S. A. and G. A. Goodfriend
1993 An Early Date for Gaza Ware from the Northern Negev. Palestine Exploration
Quarterly 125: pp. 143–148.
Saidel, B. A.
2000 Matchlocks, Flintlocks, and Saltpetre: The Chronological Implications for
the Types of Firearms Used by the Bedouin During the Ottoman Period.
International Journal for Historical Archaeology 4: pp. 191–216.
2001 Ethnoarchaeological Investigations of Abandoned Tent Camps in Southern
Jordan. Near Eastern Archaeology 64: pp. 150–157.
Schaefer, J.
1989 Archaeological Remains from the Medieval Islamic Occupation of the
Northwest Negev Desert. Bulletin of the American Schools of Oriental Research
274: pp. 33–60.
Schick, R.
1997–1998 The Archaeology of Palestine/Jordan in the Early Ottoman Period. ARAM
Periodical 9/10: pp. 563–575.
Simms, S. R.
Science 15: pp. 197–211.
Simms, S. R. and K. W. Russell
1997 Tur Imdai Rockshelter: Archaeology of Recent Pastoralists in Jordan. Journal
of Field Archaeology 24: pp. 459–472.
Ustinova, Y. and P. Nahshoni
1994 Salvage Excavations in Ramot Nof, Beèr Sheva. 'Atiqot 25: pp. 157–177.
Von Gernet, A.
1995 Nicotian Dreams: The Prehistory and Early History of Tobacco in Eastern
North America. In J. Goodman, P. E. Lovejoy and A. Sherratt (eds.), Consuming
Habits: Drugs in History and Anthropology. London, Routledge: pp. 67–87.
Ward, C.
2000 The Sadana Island Shipwreck: A Mid-Eighteenth-Century Treasure Trove. In
U. Baram and L. Carroll (eds.), A Historical Archaeology of the Ottoman Empire.
New York, Kluwer: pp. 185–202.
Weir, S.
1976 The Bedouin: Aspects of the Material Culture of the Bedouin of Jordan. London,
World of Islam Festival Publishing.
Wilford, J. N.
1999 Plumbing Black Sea for Proof of the Deluge. New York Times, 5 January: p.
F3.
2001 Scholars Find Further Signs of Big Flood Evoking Noah. New York Times, 1
October: p. A14.
Zarins, J.
1992 Pastoral Nomadism in Arabia: Ethnoarchaeology and the Archaeological
Record, a Case Study. In O. Bar-Yosef and A. M. Khazanov (eds.), Pastoralism
in the Levant: Archaeological Materials in Anthropological Perspectives. Madison,
Zarins, J., N. Whalen, M. Ibrahim, A. J. Mursi, and M. Khan
1980 Comprehensive Archaeological Survey Program: Preliminary Report on the
Central and Southwestern Provinces Survey 1979. Atlal 4: pp. 9–36.
Ziadeh, G.
1995 Ottoman Ceramics from Tiìnnik, Palestine. Levant 27: pp. 209–245.
Ch a p t e r 2 2
Naming the Waters

New Insights into the
Nomadic Use of Oases in
the Libyan Desert of Egypt
Alan Roe1
I n recent decades substantive progress has been made toward interpreting

the unique archaeological footprints of nomadic societies. Ethno-
archaeological methods have been valuable tools where the archaeological
record of such societies has proven more fragmentary or dispersed than that
of sedentary societies. While strides have been made in understanding the
material culture of nomads and, to a lesser extent, their social and economic
organization, however, we know much less about the ways in which this orga-
nization may find political and territorial expression. This chapter constitutes
an ethnographic study of pastoral nomadic resource use in the northern part of
the Libyan Desert of Egypt, with a specific focus on a group of small, currently
uninhabited, oases of the interior. The fieldwork for this study was undertaken
as part of an investigation of traditional natural resource use in the area, to
better inform the participatory protection and conservation of these oases.
Accordingly, it had a strong historical orientation, drawing heavily on oral
histories and corporate memories that extended back more than a century, well
before the introduction of modern technologies or the administrative impact
of the modern Egyptian State. The study describes a stable system of nomadic
resource use and political interaction that has spanned at least a century (some
secondary sources indicate even longer).
The objectives of this chapter are twofold. First, the case study of nomadic
ecology provides an insight into how this part of the Sahara has been historically
used by pastoral nomads and considers the distinct archaeological signatures of
this use. Second, the chapter describes an unusual nonconsensual relationship of
resource-sharing that historically existed between nomads and the population of
the oasis of Siwa. This historical relationship is utilized to demonstrate the poten-
tial difficulties of drawing conclusions from the archaeological record alone.
The research on which this chapter is based was undertaken under the auspices of
1.
the Egyptian-Italian Environmental Program.
487
488     A l a n R o e
Approaches to Old World Nomadic Pastor alism

With the realization that most pastoral nomads do indeed leave identifiable
archaeological residues (Hole 1978), there has followed considerable research
utilizing ethno-archaeological methods to examine the material cultures of
contemporary nomads to assist in the identification and interpretation of ancient
occupation sites (Simms 1988; Cribb 1991; Avni 1992; Banning 1993). While the
Near East and the Levant have been a major geographical focus for these studies,
ethno-archaeological investigations of nomadic campsites have also been under-
taken in a wide range of locations, including Iran (Evans 1983), southern Arabia
(Zarins 1992), Sudan (Bradley 1992) and Africa (Robertshaw and Collett 1983).
Although nomadic life is not normally consistent with a rich material culture
(Finkelstein 1995), ethno-archaeological investigations have identified a range
of architectural features and structures associated with pastoral camps and may
even allow researchers to distinguish between mobile and sedentary pastoral
populations (Cribb 1991). Researchers have further found that nomads exhibit
distinct preferences in the selection of occupation sites with respect to features
of landscape (Banning and Kohler-Rollefson 1992). Despite methodological
advances in recognizing and interpreting the archaeological signatures of
nomadic occupation, there are recognized limitations on the extent to which
archaeological materials alone allow for reconstructing nomadic societies
(Bar-Yosef and Khazanov 1992). In contrast to urban settlements, the archaeo-
logical record for nomads rarely allows investigation through a cross section of
society, to establish a wide socio-economic representation, or of longitudinal
development sequence at a single site. For this reason archaeologists have
looked increasingly to available ethnographic studies of nomadic societies to
provide possible analogies for societal organization and function (Hole 1979;
Smith 1985; Cribb 1991; Levy 1992).
Many researchers define pastoral nomadism as a distinct form of food-
producing economy. In its purest form pastoralism can be characterized as
subsistence-level livestock production in the absence of any form of supple-
mentary activity (Khazanov 1984). Economic specialization to this degree has
been associated with social and economic risks and instability (Cribb 1991),
however, and it is correspondingly scarce. More commonly, mobile pasto-
ralism has been practiced as the dominant activity within a wider portfolio of
economic activities, combining elements of agriculture, trade, waged labor and
raiding (Barfield 1993).
Early modern ethnographic studies of pastoral societies (Evans Pritchard
1940; Baxter 1954; Gulliver 1955) focused primarily on endogenous human
and social adaptation to remote marginal environments. Consequently, their
accounts tended to enforce precolonial and colonial perspectives of pastoral
Use of Oases in the L i b y a n D e s e r t      489
nomads as isolates sustained by their herds and socially structured by ecolog-

ical and environmental determinants. Such models tended to limit nomadic
articulation with wider society and structures such as markets. If anything,
this interaction was characterized by an unrelenting animosity between the
‘Desert and the Sown’ (Reifenberg 1950). Contemporary models of Old
World pastoral nomadism have been revised on the basis of later ethnographic
studies. Researchers now increasingly recognize relations of exchange between
nomads and sedentary populations (Khazanov 1984). Both groups have been
portrayed as mutually interdependent, where animals, animal products and
services are exchanged for nonpastoral products, such as grain and manufac-
tured commodities (Barth 1973). This new understanding of complexity in
nomadic-sedentary interactions, exchange and the possibility of transition
between productive modes comes as some assistance to archaeologists and
anthropologists working with archaeological materials. Some, however, have
questioned the extent to which modern ethnographies (often ‘development’
focused and thus highlighting socio-economic and sociopolitical change) may
be utilized as the basis for assumptions about the more distant past.
A great deal of ethno-archaeological research has been undertaken in the
Near East and Levant. In contrast, comparatively little is known about the
ecology and organization of pastoral nomadism among the Egyptian oases.
Although these ‘Islands of the Blessed’ have constituted an important pastoral
resource for millennia and have sustained diverse nomadic cultures, the scarcity
of contemporary analogues for ancient nomads and rapid rate of development
in the oases has severely restricted the potential for this kind of research.
Likewise, there has been only very limited investigation of the archaeology
of pastoral nomads in the Libyan Desert. In the absence of archaeological
evidence from the area of study, assumptions about nomadic societies are
currently based on incomplete and possibly ethnocentric Egyptian primary
sources and on ethnographic analogies drawn from other parts of North Africa.
The evidence offered in this chapter provides an ethnographic and ecological
baseline against which these assumptions may be reassessed.
The Physical Environment

The Libyan (or Western) Desert of Egypt (Figure 22.1), apart from the north
coastal fringe, is largely hyperarid.2 In this respect the ecology of the area
2.
The terms Libyan or Western Desert may be used interchangeably to describe the
hyperarid desert west of the River Nile. The term Libyan Desert is adopted here to
emphasize the biophysical integrity of the desert, crossing modern state borders in the
eastern part of the Sahara.
490     A l a n R o e
Figure 22.1. General location of the study area in Egypt.
differs significantly from the regions of the Near East and Levant, where much
of the archaeological study of nomadism has been centered. Negligible rainfall
over much of the desert interior means that life is largely concentrated in and
around the various oasis depressions, where the water table is proximate to
ground level. The ecology of the region is thus dominated by groundwater.
The principal geomorphic features of the north Libyan Desert are the
coastal plateau, the Qattara Depression, the oasis depressions and the Great
Sand Sea (Figure 22.2). The Qattara Depression encompasses an area of 19,500
km2 of playas, saltpans and sand plains at a mean elevation of about 60 m below
sea level (Masini 2001). Two small oases exist within the depression, fed by
brackish springwater: Al Qara in the west and Moghra in the east. Outside the
Qattara Depression to the south and east are other oasis depressions: (from
east to west) Shayata, Jirba, Siwa, Tabaghbagh, Al Araq, Bahrein, Nuweimisa
and Sitra. These are bounded to the south by the high dunes of the Great Sand
Sea. Only Siwa and Al Qara are currently populated, and all lie 300–350 km
south of the Mediterranean coast. The closest of these oases to the Nile Valley
lies about 400 km west of the Nile and the farthest about 600 km.
Although the Mediterranean coast receives up to 200 mm annual rainfall,
precipitation diminishes rapidly moving south across the coastal plateau. Rainfall
over the desert interior is negligible, amounting to no more than 9 mm annually at
Siwa. Mean maximal monthly temperatures range from 20°C in January to 38°C
in July (Kassas 1955). The oasis depressions support a diversity of vegetation,
Figure 22.2. The oases of the northern Libyan Desert.
and some 37 species (typical of the species found elsewhere in the Libyan Desert
oases) have been identified (Sinibaldi 2002). Around the brackish lakes grow
dense reed thickets, typically Phragmites australis. Low-lying moist areas affected
by saline soils are dominated by Arthrocnemum macrostachyum, Juncus rigidus and
Alhagi graecorum. Elsewhere, Nitraria retusa and Cressa cretica dominate. The
sandy margins of the oases support xerophytes such as Zygophyllum coccineum and
Cornulaca monocantha. Most oases also host communities of the date palm Phoenix
dactylifera. Evidence suggests that the northern Libyan Desert reached its present
condition of hyperaridity during the later Holocene (from about 5000 BP). A
series of moist intervals has since occurred during the Quaternary, however, the
most recent corresponding to the Greco-Roman period (332 BCE–395 CE) of
occupation in Egypt (Hassan and Gross 1987).
The Human Environment

There is evidence for a long period of resource use at the oases of the north
Libyan Desert.3 Lithic assemblages from the area indicate occupation by hunter-
gatherer cultures between 10,000–7000 BP (Hassan and Gross 1987). The
distribution of sites throughout the area suggests dry season occupation in the
Research in the area has hitherto been confined to Siwa, Al Qara and Al Araq oases
3.
(Hassan and Gross 1987). The full extent of prehistoric occupation will be determined
only through more extensive studies.
492     A l a n R o e
vicinity of oasis springs and marshes and the establishment of dispersed hunting
camps beyond the rim of the oasis depressions during the season of higher rain-
fall. Egyptian primary sources refer to the peoples of the Libyan Desert from
the middle of the third millennium BCE onward. Ancient Egyptians described
the peoples of areas west of the Nile as ‘Tjehenu’ and later also as ‘Tjemehu.’4
The former group perhaps occupied areas adjacent to the Nile Delta, whereas
the latter may have been distributed farther south (Osing 1980). Later, during
the second millennium BCE, the names of ‘new’ populations enter into the
Egyptian record, principally in declarations of military victories (Kitchen 1990).5
There is some evidence that the Tjemehu peoples of the second millennium
BCE were pastoral nomads or possibly seminomads. Records attest to their
cattle breeds being known and husbanded in Egypt during the 14th century BCE
(Kitchen 1990:16), while later Egyptian accounts, possibly exaggerated, describe
the capture of many thousands of livestock during New Kingdom wars and
skirmishes. In common with many nomadic cultures, the Tjemehu society left
little archaeological evidence (Knapp 1981), and most knowledge of its material
culture comes from Egyptian pictorial and textual sources.
Comparatively little is known about the geographic range or extent of these
Libyan tribes. Current assumptions, based on the archaeological evidence of
minor coastal trading sites (White 1990) and modern ethnographic studies
from neighboring Cyrenaica, have led to the belief that the ancient nomads
would have summered in better watered coastal areas and moved inland into
the desert during the cooler winter season. Ancient Egyptian inscriptions
describe the roaming bands moving into the western delta, Nile Valley, and to
the oasis of Bahariya, and possibly Farafra (Kitchen 1990:20). Fakhry (1974:61),
accepting this Egyptian evidence, viewed the occupation of the Egyptian oases
in terms of aggressive military incursions, although more recent interpreta-
tions suggest that more gradual, socio-economic processes may have been the
mechanism behind these population movements (O’Connor 1990). The settle-
ment of Libyan nomads along the western borders of Egypt was significant in
that, by the early first millennium BCE, Libyan chieftains not only controlled
the inner oases but, with the foundation of Dynasty XXII (945–715 BCE), had
assumed the government of the whole country. These developments, however,
lie outside the scope of the current discussion.
4.
The earliest reference to the Tjemehu comes from Dynasty VI (2325–2175 BCE),
the reign of Pharao Pepi I (Kitchen 1990).
5.
The ‘Meshwesh’ tribe is first mentioned during the reign of Amenhotep III
(1390–1353 BCE), while the ‘Libu’ are only mentioned by name on inscriptions dedi-
cated to Ramses II (1279–1213 BCE). Numerous other small bands or tribes are listed
by Bates (1914) and Rowe (1948).
Pastoral tribes are known to have utilized the interior of the Libyan Desert
throughout classical antiquity and are described in sources such as Herodotus
and Pliny. These Libyan or Berber groups remained in the desert interior long
after the Arab expedition of Amr ibn el Asi invaded Cyrenaica in 643 CE and
may not have been displaced until the ‘Beni Suleim’ Bedouin tribes moved west
from Arabia in the 11th century CE. At the end of the 18th century CE, ‘Awlad
Ali’ tribes from Cyrenaica migrated east across the coastal plateau to displace
the ‘Hanadi Arabs.’ Both of these groups were descendants of the original ‘Beni
Suleim’ (Murray 1935). Today, the Awlad Ali Bedouin remain the predominant
Arab tribe of the northern Libyan Desert and coastal plateau. Many are now
settled at coastal locations and in the western delta (a small community has also
settled on the outskirts of Siwa Oasis), where they are largely integrated into
Egyptian political and economic life ( Cole and Altorki 1998). The traditional
exploitation of desert resources through nomadic pastoralism was continued,
however, into the latter part of the 20th century CE, and a few households still
practice a modified form of this. As in other segmentary tribal systems, the
Awlad Ali are segmented into a number of clans and descent groups by gene-
alogy. The ‘beit’ is the minimal lineage group and is the basic camping and
economic unit, encompassing 20–50 people. The maximal lineage (or tribal
section) is known as the ashira and is headed by a senior sheikh.
The traditional nomadic economy of the Awlad Ali draws together a number
of elements. These include camel and ovicaprid (sheep and goats) herding,
some barley cultivation on the coastal plain, the provision of commercial and
transport services, and other forms of waged labor (Bujra 1973). Bedouin
livelihoods in the northern part of the Libyan Desert have historically been
constructed around these combined activities and were thus shaped not only by
ecological but also socio-economic determinants. The following information
describing nomadic ecology has been elicited on the basis of direct observation
of current pastoralism, discussion with individuals currently engaged in pastoral
production, and discussions with elders possessing personal recollections or
historical knowledge of production conditions spanning the 20th century CE.
Pastor al Migr ation

Traditionally, the Awlad Ali have herded both camels and ovicaprids, species
subject to very different management systems. While ovicaprid production has
been orientated to supply external markets, camels tend to be raised for more
subsistence functions (milk, transport and meat). Over recent decades, with
the growth of markets, there has been an increase in economic specialization
in production and overall growth in the size of the herds. Informants recall
that, at traditional subsistence levels of production, households would usually
494     A l a n R o e
manage stock at a ratio of about ten ovicaprids per camel. This chapter focuses
exclusively on ovicaprid management, since these have a much longer history
of domestication and management in the study area.6
The ovicaprid production ecology of the northern Libyan Desert is closely
linked to seasonality. The annual cycle begins on the coastal plateau, where
there is generally some rainfall during the winter months. In a good year this
may occur from October or November through March and stimulate the
growth of grasses and succulent annuals. These pastures provide a rich source
of nutrition for ovicaprids during the important lambing, kidding and milking
seasons. Synchronous with these first rains, households traditionally sow well-
watered areas (runoff channels and depressions) with barley. Except in years of
exceptionally high rainfall, these pastures are exhausted with the onset of the
summer heat, after which the barley is grazed in situ.7 In years when northern
pastures are poor (cited as two or three years out of five), all forage and fodder
resources are exhausted by early June. Under these conditions some sections
of the Awlad Ali tribe send their livestock deep into the Libyan Desert to the
oases of the interior, ranging from Moghra in the east to Jaghabub in the west.
This grazing migration has a long historical precedent, extending back many
generations to the arrival of the tribe in the area.
Being groundwater rather than rainwater dependent, the desert oases host
a largely perennial flora. Indeed, some of the key forage species in the oases,
such as Alhagi, appear to blossom during the summer months. The summer
availability of water and proximity of forage make the Libyan Desert oases
indispensable resources for pastoralists. Today herds are (and were historically)
kept in the vicinity of the desert oases for three to five months, until the return
of the rains on the northern fringes of the desert.
Ovicaprids require constant shepherding, both to direct the herd’s move-
ment and to protect them from predators like foxes, jackals and hyenas.
According to modern shepherds, the optimal management ratio is about one
shepherd to 200 animals. More animals, and the herd would become unman-
ageable, while a smaller herd results in an inefficient use of labor. Given that
even today, with generally larger herd sizes than in the past, many livestock
owners own considerably fewer than 200 animals, it is common practice for
the livestock of kinsmen to be combined to create larger management units for
6.
There is evidence that camels were not widely managed in Egypt until the end of
the first millennium BCE (Wilson 1984). A good overview of the available data for the
prehistoric spread of animal domestication across Saharan North Africa is presented
by Gautier (1987).
7.
When seasonal conditions were conducive to barley maturing full term, some was
generally held back from livestock for household supply of grain.
migration. Many households hire professional shepherds for herd management

as livestock owners may have other social and economic commitments over the
period of the migration. Until the recent past, migration herding was always
done by male kinsmen or family members, if not by the owner himself.
Throughout living memory, migration to the inner desert oases has been the
sole preserve of men with expert knowledge because of the scarcity of water fit
for human consumption, extreme summer temperatures and other dangers of
the long journey across the northern part of the desert.8 Individuals or small
groups of men carrying supplies by donkey or camel would escort the herds,
while the remainder of their household and kin groups remained at coastal
encampments. Tribal elders indicate that the migration of complete household
units would have been more common prior to the 20th century CE, especially
at times of drought on the north coast. Owing to the scarcity of forage and
water resources in the desert beyond the Qattara Depression, herds move
directly from oasis to oasis, stopping to drink and graze for a couple of days at
each water source before moving on. The final destination of the herds and the
duration of their visit to each oasis is partly determined by the tribal affiliation
of the owners and their grazing rights as recognized under traditional tribal law,
the 'urf. According to senior Awlad Ali informants, traditional grazing entitle-
ments to the oases of the northern Libyan Desert are divided among four tribal
sections: the ‘Asheibat,’ the ‘Arawa,’ the ‘Qatifa,’ and the ‘Samalus.’9
The Asheibat are recognized among Bedouin as holding primary grazing
rights to the oases of Shayata and Jirba, west of Siwa. The Arawa hold 'urfi
rights (by tribal law) to the spring of Tabaghbagh and oasis of Al Araq; the
Samalus have these entitlements to Bahrein oasis; and the Qatifa lay claim
to distant Nuweimisa and Sitra oases (Figure 22.3). According to tribesmen
encountered among these oases in summer 2002, these lineage groups would
traditionally restrict their herds to the areas over which they held 'urfi rights,
although there was open, short-term, access to water and grazing for moving
herds between oases. Under certain conditions tribal sections may also trade
longer-term grazing rights if mutually advantageous.
8.
The journey droving sheep and goats from the coast to the inner oases reportedly
took about 20 days (today the trip to Siwa is often made by truck) and was only possible
in stages utilizing water sources and sparse perennial forage en route. These locations
and routes were known to experienced drovers, but still animals were regularly lost in
transit.
9.
While the Asheibat, Arawa and Qatifa are each sections of the Awlad Ali, the
Samalus were originally a separate tribe (Murray 1935). Today the distinction is largely
overlooked and the Samalus act as a section of the Awlad Ali.
496     A l a n R o e
Figure 22.3. Traditional migratory destinations in the northern Libyan Desert.
Pastor al Ecology
The feeding behavior of sheep shows a clear preference for grazing on succu-
lent material, principally annual grasses and herbs, while goats combine
grazing with browsing on bushes and woody shrubs. In practice, however, the
Awlad Ali nearly always manage both species in combined groups (other than
small numbers of goats maintained permanently at the home encampment
to supply milk). The oasis depressions of the Libyan Desert offer a range of
forage opportunities for nomadic herds. On the basis of observations made at
the oases of Al Araq, Bahrein and Sitra, and the reports of Asheibat shepherds,
important forage species were identified and ranked in order of importance
(Table 22.1). These observations demonstrated that the reeds and rushes
(Phragmites and Juncus sp.) of the oasis wetlands constituted the most important
grazing resource. Not only do the thick stands provide the greatest density of
edible material, but it was reported that early in the life cycle of the plants they
are highly palatable and thirst quenching for ovicaprids.10 As mature stands,
however, the oasis reeds and rushes are unpalatable. The normal practice
adopted by desert herders is therefore to burn off mature stands to stimulate
This claim of the Asheibat has not been scientifically verified. Shepherds insist,
10.
however, that sheep and goats thrive on the young shoots of Juncus acutis, even where
standing water is too brackish for them to drink.
Table 22.1. Ovicaprid Oasis Forage Species Ranked by Importance:
Rank Species Location

1 Phragmites australis Oasis wetlands
2 Juncus acutis Oasis wetlands
3 Alhagi graecorum Oasis fringe
4 Nitraria retusa Oasis fringe
5 Cornulaca monocantha Sand
6 Zygophyllum coccineum Sand
regrowth. Young shoots rapidly emerge from burnt ashes and are ready for
grazing within three or four weeks.
Burning was observed at the oasis of Bahrein, and it was noted that this
was carefully managed and was, in this event, restricted to relatively small
areas of Juncus acutis. Herders must judge from experience how large an area
to burn to sustain their herds for a given duration. They must also time each
successive burning so that new shoots reach the required level of maturity just
as the previous area of grazing is exhausted. Ideally, each successive burning
should occur at different parts of the oasis, or at different oases, so that animals
cannot stray between grazing areas. According to Bedouin elders, the burning
of reeds and rushes has been practiced annually for generations and is a basic
management tool for the pastoral utilization of the desert oases. Asheibat
herdsmen reported that under this system of management, ovicaprids would
maintain condition without additional feeds. On the basis of observations,
and the reports of herders, it is estimated that the western oases of Jirba and
Shayata could support a population of 1500 ovicaprids. The larger eastern
oases (Al Araq, Bahrein, Nuweimisa, Sitra) were reported capable of supporting
about three times this number through the summer season. Migrant herds are
normally kept at the oases of the desert interior for about five months during
a migration year. They return to the north coast when the first winter rains
there stimulate pasture growth. Pastures of annual grasses are known to be of
superior nutritional value, highly desirable for lambing and milking, and (after
winter fattening) stock becomes available for sale to coastal markets.
During the period of annual sojourn at the desert oases, the living conditions
of herders are arduous. They endure extreme summer temperatures, often
without direct access to fresh water, forcing them to regularly fetch water from
neighboring springs or oases on pack animals. Furthermore, the Libyan Desert
oases are infested with flies and fever-bearing mosquitoes. The only protection
against these insects is to remain enshrouded in fire smoke. Food is largely
restricted to durable supplies of grain, brought from the coast, and immature
dates picked locally. This food is often mixed with the milk of lactating goats
or camels and supplemented by occasional hunting of gazelle (Gazella dorcas)
498     A l a n R o e
and hares (Lepus capensis). The activities of preparatory burning, regularly

accessing water potable to stock, and exploiting local food and fuel sources
results in a pattern of frequent localized movements within individual (and
between neighboring) oases as 'urfi rights allow. Under these difficult condi-
tions of occupation, herders have few of the regular chattels of the permanent
Bedouin camps of the coastal plateau. Commonly, they find shelter in caves or
build simple shelters from the most widely available material: bundled reeds
and rushes. Even today herders have very few material possessions, just water-
skins and maybe a jar, knives, old weapons and some basic utensils.
Relations with the Oaseans

At present only the oases of Siwa and Al Qara support permanent populations.
Historically and culturally these populations have been isolated from broader
Egyptian society and are believed to be the settled descendants of the original
Libyan peoples of the desert. Even today, they maintain their distinct Berber
language and cultural practices (Fakhry 1973). Since antiquity, these oaseans
have built their livelihoods around the cultivation of dates, olives and lesser
crops in their irrigated gardens (Kuhlman 1998). Consequently, the Siwans
have developed a sophisticated understanding of soil and water management. At
both Siwa and Al Qara, oasis populations have historically constructed fortified
mud-brick settlements, elevated high above the oasis depression floor with high
exterior walls and narrow entrances to facilitate their defense. This defensive
architecture was adopted partly owing to the feuds and battles that character-
ized the fractious medieval society of Siwa but also because the oases have been
historically subject to regular raids by desert nomads (Fakhry 1973).
The historical relationship between the Awlad Ali tribes and the sedentary
oaseans of the north Libyan Desert has been complex. On the one hand, the
oases towns have presented possibilities for plunder.11 On the other hand,
with their camels the desert tribes have largely controlled the transport of
oasis products to the markets of the Mediterranean coast and Nile Valley. The
ambiguity of this relationship is nowhere more apparent than in the respective
claims of both groups to the uninhabited oases of the study area. Senior sheikhs
of the Siwan and Al Qara tribes relate how, in the past, parties of oaseans would
travel with donkey caravans to collect the fruit of wild date palms (Azowi) from
the isolated desert oases during the annual winter harvest season.12 This prac-
tice ended in the early part of the 20th century CE, after which the extension
11.
According to Fakhry (1973:19), Bedouin predations on Siwa ended with the
conquest of that oasis by the troops of Muhammed Ali in the 19th century CE.
12.
The dates of the Azowi palm are deemed to be of the lowest quality of several vari-
eties harvested at Siwa. Only about a third is considered fit for human consumption.
of irrigation and intensification of date cultivation at Siwa itself made these

hazardous desert journeys redundant. This long tradition of resource manage-
ment is enshrined, however, in traditional Siwan lore and in the old historical
manuscripts of Siwa, which claim the uninhabited oases as the property of the
Siwan tribes. Specifically, the Al Qara people historically owned and harvested
the oases of Tabaghbagh and Al Araq, while the Siwa tribes harvested their
Azowi palms at the oases of Shayata, Jirba, Bahrein, Nuweimisa and Sitra.
While some Siwan elders maintain that Bedouin tribes have no rights to oases
resources, and therefore are prohibited from visiting them, other opinion posits
that Siwan ownership of the oases means exclusive control of their products
(dates and potable water) but that little could be done to actually prevent seasonal
Bedouin grazing. Consequently, visiting Bedouin herds should be tolerated. This
concession comes with a remarkable caveat: while nomads may drink from oasis
waters and (if necessary) clear and scrape out shallow water holes, they are not
permitted to assign names to water sources they dig. According to old Siwan
law manuscripts, ‘naming the waters’ would signify creation of a permanent well
and thus infringe on ownership rights. Through much of the 20th century CE
discrepancies existed between the respective resource claims to the uninhabited
oases (Figure 22.4). At the time of this study the Siwan claims expressed perceived
Figure 22.4. Overlapping territorial claims to the oases in the northern Libyan Desert.
500     A l a n R o e
historical and symbolic rights of ownership, which they no longer actually used,
while the Awlad Ali claims were related to actual management and use.
In practice, Bedouin and oasean groups have tended to overlook their
respective claims as there is little incongruency between these claims. As long
as the former do not challenge Siwan symbolic rights of ownership and the
latter are not in a position to prevent Bedouin herds from grazing, mutual
toleration of the status quo exists. In effect, ambiguity over resource owner-
ship serves the purposes of both groups; their sheikhs have never formally met
and refuse to do so.13 Apparent inconsistencies in the political relationship
between desert nomads and oaseans were explained thus: where co-operation
has occurred between groups (as historically in the transport of harvested oasis
products to the coast), these transactions had been arranged on personal bases
between individuals. It was reported that no wider structural or cooperative
agreements had ever existed between Bedouin and Siwans, which seems consis-
tent with historical evidence for almost simultaneous conflict and commerce
between the groups.
Archaeological Signatures
What does the information gathered during this study indicate about the
archaeological signatures of nomadic use and occupation of the Egyptian
oases? One of the most significant findings of the study, previously unre-
corded in the Libyan Desert, relates to the evidence for the long-standing
practice of managing key oasis forage species with fire. Evidence for past fire
management of this type should be observable in the strata of oasis sediments,
notably in wetland (or former wetland) areas of oasis depressions. At this
juncture it is worth referencing some findings of the Combined Prehistoric
Expedition in the southern Libyan Desert. Geo-archaeological investigations
conducted at remote oasis depressions have revealed layers of burnt reed
charcoal deposits resting on Aeolian sand at the base of ancient lake beds. The
accumulated evidence seems to suggest the periodic burning of Phragmites
and Typha (a species akin to Juncus) along the margins of the ancient lakes
(Haynes 1987).14 In the absence of other evidence it was speculated that this
13.
This has changed since a group of Asheibat Bedouin settled at Siwa about a century
ago. Siwan sheikhs still refuse to meet Awlad Ali sheikhs from outside the oasis but now
include the local Bedouin sheikhs within their own council. The Siwan Asheibat sheikhs
have links to both the Siwans and their kinsmen on the north coast.
14.
Burnt-reed charcoal deposits have been identified at the uninhabited oases depres-
sions of Oyo, Merga and Selima (Haynes 1987). In terms of geomorphology and ecology
these oases are similar to their counterparts in the northern Libyan Desert.
burning was undertaken “by Neolithic pastoralists to provide better access to

lake water” (Haynes 2001:129). Whether the modern practices recorded at Al
Araq, Bahrein, Sitra and Nuweimisa may have relevance to this conclusion is
not yet clear.
The material residues of modern herding camps are sparse, related to
the fact that at present only specialist male shepherds undertake migrations
through the desert. Shelters are usually temporary, owing to the necessity of
regularly relocating the herd between prepared grazing areas, and constructed
from local perishable materials such as reed stems and palm leaves. Exploration
of caves in the escarpment overlooking the oasis of Bahrein revealed evidence
of use including stores of old water jars, packsaddles and other equipment.
Heavy smoke-soot residues on the walls and roof of the cave suggested occupa-
tion, as well as storage functions, smoke providing some necessary relief from
swarming insects. Whether these caves were used by shepherds or represent
periods of occupation when entire household units migrated was not possible
to establish through superficial observation.
A further activity associated with oasis occupation is hunting, which brings
some variety to the diet of herders. Compared with coastal areas, the wildlife
of the oases is relatively abundant. Contemporary hunters utilize vintage
firearms together with a variety of more traditional techniques for trapping
birds and small mammals. Metal blades are useful for butchering, as well as
for cutting cords and ropes associated with packsaddles. Although sparse, all
the manufactured artifacts associated with migratory Bedouin seem to have
been acquired via the broader market economy, from population centers on
the north coast.
As reported by Siwan informants, the desert oases have also historically been
the destination of donkey caravans from the populated oases with the purpose
of collecting Azowi dates. The eldest living Siwans had participated in the last
of these expeditions in the early part of the 20th century CE. There are some
important differences between these visits and those of Bedouin to the oases.
Siwan caravans would stop only briefly at each oasis, requiring just a few days
to harvest dates from the palm groves at each site. Although it is possible
that past Bedouin migrations to the oases involved full household groups, it
is known with certainty that Siwan date-collecting camps were exclusively
male. Finally, the material culture of the Siwa oasis is highly distinct from that
available through regional markets and a majority of items were manufactured
within the oases.
Collectively, the above observations lead to a further point with respect to
interpretation of the archaeological record at the desert oases. While a future
investigator might identify the residues of regular occupation and resource
use by two contemporary (yet distinct) cultural groups over a long period of
502     A l a n R o e
time, it may be more difficult to accurately discern the duality in this use or the
complex social and political relationship between users. Although this hypo-
thetical investigator would find evidence for successive visits and occupations
through long historic periods, it may be less obvious that nomadic pastoralists
were usually in occupation during summer months, while Siwan campsites
were established only at the time of the date harvest during the winter months.
Despite the apparent archaeological proximity of residues, the two groups were
rarely, if ever, present simultaneously, and it may be misleading to assume any
kind of formalized relationship between them. As ethnographic research has
indicated, the historic relationship between oasis-using groups has effectively
been one of nonconsensual resource sharing manifest in a reluctant toleration
of respective activities. This reflects the current ambiguity, and the historic
volatility, of the relations between Bedouin and Oasis dwellers.
Nomadic Use of the Egyptian Oases

The system of traditional Awlad Ali migration described in this chapter consti-
tutes a journey that could exceed 400 km (each way) across arid desert terrain.
That such a pattern of regular pastoral migration with sheep and goats existed
historically is of itself noteworthy. It has long been known that camel herding
tribes made regular migratory journeys to and between the Egyptian oases
(Murray 1935), but hitherto there has been little evidence for similar migratory
use by herders of small stock. This study demonstrates that the interior oases
of the Egyptian desert have been and continue to be utilized as an important
reservoir of forage against conditions of drought or summer scarcity elsewhere.
That groundwater-dependent forage resources are available year-round indi-
cates that they could potentially complement pasture and migration regimes
at both northern (winter rainfall) and southern (summer rainfall) fringes of the
Libyan Desert. This study further suggests that the prospect of several months
of supply of forage is considered sufficient incentive to attract pastoralists over
long distances, enduring considerable hardship and risks.
The evidence of this study thus contributes additional data to help us
understand the behavior of past nomadic societies under comparable resource
conditions. It is worth noting that contemporary speculations about how
ancient Libyan pastoralists may have utilized natural resources through
migration describe nomads moving inland during the winter from summer
coastal camps. This model draws heavily on ethnographic observations made
in neighboring Cyrenaica and does not take into account the possibility of
summer migrations to inland oases or the forage resources available there.
While oasis resources, under the management system described, hold the
potential to support ovicaprid herds at maintenance for several months, it is
a normal practice for herds to return to rain-stimulated annual pastures for

the vital lambing, kidding and milking season, thus putting an upper limit on
the duration of pastoral occupation of oases over any given year; unless, of
course, coastal pastures fail completely. In this latter eventuality it is possible
that herding groups could occupy oasis sites as long as necessary until pasture
conditions improved elsewhere. It is probable that more-sustained periods of
occupation would be visible in the archaeological record. If sustained drought
in coastal areas drove large numbers of pastoral groups to seek permanent
forage resources in the desert interior, this would possibly exceed the capacity
of resources in the study area and require herders to spread further into the
Libyan Desert. This type of strategy seems consistent with primary Egyptian
sources (dated to around 1213–1204 BCE) that document the arrival of starving
Libyan bands at the Bahariya Oasis, apparently searching for productive land
(Magid, this volume; Kitchen 1990:20).
It is also noteworthy that, in the case of both Bedouin and (historically)
Siwan resource use, resources are divided into subunits under the manage-
ment of individual descent groups. General recognition of territorial rights
is of particular importance for Awlad Ali fire management: it enables herders
to plan and implement burning regimes within their own territorial unit, first
preparing and then setting aside areas for future use, without risk of other
resource users depleting these first. With respect to access rights, fire manage-
ment practices at the oases are closer to agricultural practice than the ‘open
access’ grazing of seasonal pastures. Most ethnographic studies note that where
Bedouin invest labor and other management inputs into the preparation of land
for cultivation, such as plowing and sowing, this investment confers exclusive
access to the product of these labors. Long-term sustained inputs can result in
recognition of territorial rights under traditional law.15 Similarly, the recog-
nized management inputs at the uninhabited oases seem to have conferred
exclusive access rights to resource-using Awlad Ali descent groups.
Discussion
The ethnographic work undertaken in the northern Libyan Desert may
contribute to our understanding of nomadic societies and pastoral resource
use in two distinct ways. The first relates to the production of ethnographic
analogy for historic and prehistoric pastoral resource use in the specific region
of study. The second relates to the practical constraints of interpreting the
archaeological record of nomadic occupations.
See Lancaster and Lancaster 1999 for a wider discussion of traditional Bedouin
15.
systems of access to resources and territoriality.

504     A l a n R o e
This study has described how the Arab Bedouin groups of the northern
Libyan Desert traditionally exploit and manage resources in the isolated and
currently uninhabited oases of the interior. It has provided the first description
of this form of migratory resource use and demonstrated both the significance
of groundwater-sustained oasis resources to pastoralists and their capacity to
access and utilize them. It has been postulated that these observations have
relevance not only to understanding the ecology of pastoral groups in the
northern part of the desert but also to those groups utilizing the southern
margins bordering the Sahel. Developmental change has now made it difficult
to observe traditional systems of resource management by pastoralists at other
Egyptian oases.
The case study further highlights the complexity of relationships that
may exist between nomads and other groups. At present, while the Awlad
Ali articulate effectively with the broader regional political economy, there
is considerable local variation in how specific relationships are managed.
Similarly, archaeological evidence may allude to temporary occupations
by small groups of Bedouin and Siwan resource users at the uninhabited
oases, but this does not necessarily represent co-residence, nor should it be
interpreted as the basis for any kind of broader structural relationship. The
mobility fundamental to the nomadic lifestyle will always present challenges
for the investigator (archaeological, ethnographic or other) in drawing general
conclusions from the data and will always require exhaustive explorations of
all potential interpretations.
References
Avni, G.
Implications for Archaeological Research. In O. Bar-Yosef and A. Khazanov
Perspective. Madison, Prehistory Press: pp. 241–254.
Banning, E.
1993 Where the Wild Stones Have Gathered Aside: Pastoralist Campsites in Wadi
Ziqlab, Jordan. Biblical Archaeologist 56: pp. 212–221.
Banning, E. and I. Kohler-Rollefson
1992 Ethnographic Lessons for the Pastoral Past: Camp Locations and Material
Remains Near Beidha, Southern Jordan. In O. Bar-Yosef and A. Khazanov
Perspective. Madison, Prehistory Press: pp. 181–204.
Barfield, T.
1993 The Nomadic Alternative. Englewood Cliffs, Prentice Hall.
Barth, F.
C. Nelson (ed.), The Desert and the Sown: Nomads in Wider Society. Berkeley,
University of California Press: pp. 11–22.
Bar-Yosef, O. and A. Khazanov
1992 Introduction. In O. Bar-Yosef and A. Khazanov (eds.), Pastoralism in the Levant:
Archaeological Materials in Anthropological Perspective. Madison, Prehistory
Press: pp. 1–9.
Bates, O.
1914 The Eastern Libyans: An Essay. London, Macmillan.
Baxter, P.
1954 The Social Organization of the Boran. University of Oxford (PhD
dissertation).
Bradley, R.
1992 Nomads in the Archaeological Record: Case Studies in the Northern Provinces of the
Sudan. Meroitica 13. Berlin, Akademie-Verlag.
Bujra, A.
1973 The Social Implications of Development Policies: A Case Study from Egypt.
In C. Nelson (ed.), The Desert and the Sown: Nomads in Wider Society. Berkeley,
Cribb, R.
Cole, D. P. and S. Altorki
1998 Bedouin, Settlers, and Holiday-Makers: Egypt’s Changing Northwest Coast. Cairo,
American University in Cairo Press.
Evans, C.
1983 On the Jube Line: Campsite Studies from Kurdistan. Archaeological Review
from Cambridge 2: 2, pp. 67–77.
Evans-Pritchard, E.
1940 The Nuer: A Description of the Mode of Livelihood and Political Institutions of a
Nilotic Tribe. Oxford, Clarendon Press.
Fakhry, A.
1973 The Oasis of Egypt. Volume 1: Siwa Oasis. Cairo, American University in Cairo
Press.
1974 The Oasis of Egypt: Volume 2: Bahariyah and Farafra Oases. Cairo, American
University in Cairo Press.
Finkelstein, I.
1995 Living on the Fringe: The Archaeology and History of the Negev, Sinai, and
Neighbouring Regions in the Bronze and Iron Ages. Sheffield, Sheffield Academic
Press.
506     A l a n R o e
Gautier, A.
1987 Prehistoric Men and Cattle in North Africa: A Dearth of Data and a Surfeit
of Models. In A. Close (ed.), Prehistory of Arid North Africa. Dallas, Southern
Methodist University Press: pp. 163–187.
Gulliver, P.
1955 The Family Herds. London, Routledge and Kegan Paul.
Hassan, F. and T. Gross
1987 Resources and Subsistence During the Early Holocene at Siwa Oasis,
Northern Egypt. In A. Close (ed.), Prehistory of Arid North Africa. Dallas,
Southern Methodist University Press: pp. 85–103.
Haynes, V.
1987 Holocene Migration Rates of the Sudano-Sahelian Wetting Front, Arba’in
Desert Eastern Sahara. In A. Close (ed.), Prehistory of Arid North Africa. Dallas,
Southern Methodist University Press: pp. 69–84.
2001 Geochronology and Climate Change of the Pleistocene–Holocene Transition
in the Darb al Arba’in Desert, Eastern Sahara. Geoarchaeology: An International
Journal 16: 1, pp. 119–141.
Hole, F.
1978 Pastoral Nomadism in Western Iran. In R. Gould (ed.), Explorations in
Ethnoarchaeology. Albuquerque, University of New Mexico Press: pp.
127–167.
1979 Rediscovering the Past in the Present: Ethnoarchaeology in Luristan, Iran. In
C. Kramer (ed.), Ethnoarchaeology: Implications of Ethnography for Archaeology.
New York, Columbia University Press: pp. 192–218.
Kassas, M.
1955 Rainfall and Vegetation Belts in Arid North-East Africa. In Plant Ecology:
Proceedings of the Montpellier Symposium. Strasbourg, UNESCO: pp. 49–59.
Khazanov, A.
Kitchen, K.
1990 The Arrival of Libyans in New Late Kingdom Egypt. In A. Leahy (ed.), Libya
and Egypt, c. 1300–750 BC. London, Centre for Near and Middle Eastern
Studies: pp. 15–27.
Knapp, A.
1981 The Theran Frescoes and the Question of Aegean Contact with Libya During
the Late Bronze Age. Journal of Mediterranean Anthropology and Archaeology 1:
pp. 249–279.
Kuhlman K.
1998 Roman and Byzantine Siwa: Developing a Latent Picture. In O. Kaper (ed.),
Life on the Fringe: Living in the Southern Egyptian Deserts During the Roman and
Early Byzantine Periods. Leiden, Research School CNWS: pp. 159–180.
Lancaster, W. and F. Lancaster

1999 People, Land, and Water in the Arab Middle East: Environments and Landscapes
in the Bilad ash-Sham. New York, Harwood.
Levy, T.
1992 Transhumance, Subsistence, and Social Evolution in the Northern Negev
Desert. In O. Bar-Yosef and A. Khazanov (eds.), Pastoralism in the Levant:
Archaeological Materials in Anthropological Perspective. Madison, Prehistory
Press: pp. 65–82.
Masini, A.
2001 Environmental Amelioration in Siwa: Technical Report. Gland, International
Union for the Conservation of Nature.
Murray, G. M.
1935 Sons of Ishmael: A Study of the Egyptian Bedouin. London, George Routledge.
Osing, J.
1980 Libyen, Libyer. In W. Helck and E. Otto (eds.), Lexikon der Ägyptologie, Volume
3. Wiesbaden, Otto Harrassowitz: pp. 1015–1033.
Reifenberg, A.
1950 The Struggle Between the Desert and the Sown [in Hebrew]. Jerusalem, Mosad
Byalik.
Robertshaw, P. and D. Collett
1983 The Identification of Pastoral Peoples in the Archaeological Record: An
Example from East Africa. World Archaeology 15: pp. 67–78.
Rowe, A.
1948 A History of Ancient Cyrenaica. In Cahiers des annales du service des antiquités
de l’Egypte 12. Cairo, Department of Antiquities.
Simms, S.
Science 15: pp. 197–211.
Sinibaldi, I.
2002 Proposal for the Establishment of the Siwa Protected Area. Siwa, Egyptian-Italian
Environmental Programme.
Smith, A.
1985 The Ethnoarchaeology of Pastoralism in Saharan and Sahel Zones of West
Africa. In M. Liverani, A. Palmieri and R. Peroni (eds.), Studi di paletnologia
in onore di Salvatore M. Puglisi. Università degli Studi di Roma “La Sapienza”:
pp. 57–70.
White, D.
1990 Provisional Evidence for the Seasonal Occupations of the Marsa Matruh Area
by Late Bronze Age Libyans. In A. Leahy (ed.), Libya and Egypt, c. 1300–750
BC. London, Centre for Near and Middle Eastern Studies: pp. 1–14.
508     A l a n R o e
Wilson, R.
1984 The Camel. Essex, Longman.
Zarins, J.
1992 Pastoral Nomadism in Arabia: Ethnoarchaeology and the Archaeological
Record: a Case Study. In O. Bar-Yosef and A. Khazanov (eds.), Pastoralism
in the Levant: Archaeological Materials in Anthropological Perspective. Madison,
Ch a p t e r 2 3
F rom Objects to Agents

The Ababda Nomads and the
Interpretation of the Past
Willeke Wendrich
T his chapter reflects nine years of interaction with the Ababda nomads
(singular: Abadi; feminine: Abadiyya), living in the southern part of the
Eastern Desert, between the Nile Valley and the Red Sea in Egypt (Figure
23.1). From 1994 to 2002 a group of up to 80 Ababda excavated the ancient
harbor town of Berenike, supervised by an international crew of approximately
40 archaeologists (Sidebotham and Wendrich 1995, 1996, 1998, 1999, 2000,
2007). The Ababda way of life is currently under pressure. Discussions during
and after the work often centered on the rapid change of Ababda culture in
recent times. These changes are the result of an increase of tourism along the
Red Sea coast and the fact that the Egyptian government would like to see
the Ababda settle. For this purpose several villages were built along the main
road along the Red Sea coast. The village layouts are spacious: the houses were
built in flat desert areas with much space between the buildings. Each house
contained two units, each with two rooms. Each Ababda nuclear family was
given a two-room unit, which one entered from a gated courtyard. Water tanks
were placed near areas where traditional Ababda dwellings were concentrated,
and the government provided drinking water for these communities. Apart
from these housing projects on the Red Sea coast, Ababda have settled in the
Nile Valley, mostly in the region around Kom Ombo, which was developed
to house the Nubians displaced by Lake Nasser in the 1960s. The younger
generation of Ababda whose parents had settled in Wadi Khareet expressed
a keen interest in saving a record of their culture. They asked if I could not
write a book about them, much like the books about the ancient inhabitants
of the region. I suggested that they should write such a book themselves, with
my assistance, a project that is still ongoing.
In relation to this writing project a group of Ababda and several members of
the excavation team started to record the Ababda culture and created a collec-
tion of Ababda material culture. As part of the process of awakening interest in
the cultural heritage of the Ababda, the Eastern Desert Antiquities Protection
Project (EDAPP) was granted a substantial contribution by the Cultural Fund
509
510     W i l l e k e W e n d r i c h
Figure 23.1. Map of a part of the Eastern Desert, between the Nile Valley and the Red Sea in the south
of Egypt and the north of Sudan, with the location of different groups of nomads indicated, as well as
the place names mentioned in the text (map prepared by H. Barnard).
of the Royal Netherlands Embassy in Cairo. This grant was used for training
in archaeological techniques and to build a site museum where the collection
could be housed. Part of the collection was displayed in this site museum;
another part was arranged into a small exhibit in one of the towers of the
restored Ottoman fort in Quseir. A third collection traveled to the Netherlands
and formed the core of an exhibition in the Wereldmuseum Rotterdam.1
The exhibit in the Netherlands contrasted two collections: one composed
by the Ababda, and one collected by the Egyptologist L. Keimer, purchased by
the Wereldmuseum in the 1950s. The ‘collection philosophies’ behind these
This exhibit was entitled Nomaden tussen Nijl en Rode Zee (Nomads Between the
1.
Nile and the Red Sea) and ran from April 2002 through March 2003.
T h e A b a b d a N o ma d s      511
two, and therefore also the objects, are very different. Keimer was interested in
making a ‘museum quality’ collection of the finest Ababda arts and crafts repre-
senting the life of ‘the other.’ The EDAPP collection is made by the Ababda
to represent their own cultural identity and to give a representative image of
what Ababda life is like to future Ababda, who will live in a changed world,
and to inform outsiders (Egyptians, foreign tourists) about Ababda culture and
heritage. In 2006 the Royal Netherlands Embassy in Cairo donated funds to
build an exhibition center that will form the core of the Beit al-Ababda (Ababda
House) in the Wadi al-Gemal National Park.
Keimer published extensive descriptions of the Eastern Desert cultures
(1951, 1952a, 1952b, 1953a, 1953b, 1954a, 1954b). His specific interest was to
outline ‘survival’ of the material culture of the Beja from the Pharaonic period
to the present. In a series of articles he highlighted a wide range of aspects such
as the use of headrests, amulets, kohl eye protection, makeup, hairstyles and
gazelle traps. During the work in 1998 it was clear that my interest in Ababda
culture was different from that of Keimer. As an archaeologist I was fascinated
by the recent history, the material culture and present way of life, of the Ababda
from an ethno-archaeological perspective. The strategies of the Ababda to cope
with periods of drought and to survive successfully in an arid region provided
important information on potential use of tactics to cope with similar circum-
stances by the ancient inhabitants of Berenike and by the nomadic Eastern
Desert Dwellers, for which we found evidence in the excavation. Both Keimer’s
and my interest in Ababda culture and history differ, therefore, from that of
the group of Ababda, working on the EDAPP collection.
What follows is a brief introduction into the Ababda way of life, highlighting
those aspects that are of interest to me as an archaeologist: their complex
social structure, their notion of personal ownership, and their impact on the
landscape. The physical traces that the Ababda way of life may leave for future
archaeologists may serve as a reminder that the sparse material traces of mobile
people can represent complex social systems and ideas of identity, belonging
and ownership. This chapter is partly descriptive because little information has
been published on the present-day Ababda. The discussion focuses on multiple
interpretations of the past, which can be identified in the various attempts that
have been made to describe and interpret Ababda history and culture.
The Ababda
The Ababda are perhaps best characterized as multiresource pastoral nomads
roaming the valleys (wadis) of the Eastern Desert and the Red Sea coast, although
a substantial number of them have permanently settled in the Nile Valley. Their
identity is strongly connected to the landscape they inhabit; the camels, sheep,
and goats they own; their material culture; and their social ceremonies. The
Ababda are generally classified as a subgroup of the Beja nomads, which occupy
an area stretching from Quseir, in Egypt, to Eritrea. The Beja include the
Bishareen, the Hadendowa, the Amarar, the Atman and the Beni Amr (Magid,
this volume; Murray 1923; Paul 1954; Salih 1980; Morton 1988; Hobbs 1990).
The Rashaida, who live around Shelateen and can be recognized by their purple
outfits, have migrated from Saudi Arabia relatively recently. The same is true for
the Ma'aza, who live north of Quseir (Hobbs 1990). In contrast to most of the
Beja, the Ababda speak a dialect of Arabic (De Jong 2002).
Several authors have claimed that the Beja have lived in the region of the
Eastern Desert of Egypt and the Red Sea Hills of Sudan since time immemo-
rial and that they are to be considered identical with the ancient Bulahau,
Blemmyes, Bega and Bougaites (Eide et al. 1998; Pierce 2001).2 These names
are known from Roman texts from the 4th to 6th centuries CE in which these
groups are invariably presented as untrustworthy, unruly and uncontrollable
(Updegraff 1988; Eide et al. 1998). This identification, however, is based on the
uncritical use of the ancient sources and a disregard of two millennia of devel-
opment (Burstein, this volume; Barnard 2005). The Greco-Roman sources
are internally incoherent, and it is obvious that for most ancient authors the
Blemmyes and other peoples were diffuse groups, perceived as ‘the other,’ and
representative of the diversity of inhabitants of the areas south of the Roman
border. Archaeological evidence for nomadic activities in the region seems to
indicate, in the late Roman period, a population well rooted in both the desert
and settled society. A distinctive type of pottery, which may have served as a
cultural or ethnic marker, was possibly produced in the desert (Barnard, this
volume; Barnard and Strouhal 2004; Barnard et al. 2005). There is no exten-
sive literature on the Ababda. The most comprehensive are the publications
of Louis Keimer (see above).
Social Organization
The Ababda are organized patrilineally and claim to descend from Abad, who
came to Egypt from Saudi Arabia and was the son of Zubeyr (ibn al-Awwam),
‘a friend’ of the Prophet Mohamed (Murray 1923). At the same time, they
consider themselves to belong to the Beja. In contrast to their closest southern
neighbors, the Bishareen, the Ababda call themselves ‘Arabs.’ This, however,
does not in the first place refer to Saudi Arabia as place of origin, and is not an
Bigeh, an ancient name for the harbor of Berenike appearing on several medieval
2.
maps, is thought to refer to the Bega who took control of the region in the 6th century
CE (Murray 1967; Eide et al. 1998).
ethnic designation, but instead refers to a lifestyle in contrast with the settled
inhabitants of the Nile Valley (Morton 1988). The Ababda lineage is subject to
mystification by the Ababda themselves. Identity is specified at different levels,
depending on the context, and a person can be Beja, Ababda, 'Ogada (one of
the Ababda ‘clans’), and his father’s and grandfather’s son. A young Ababda
participant in the project, who made an inventory of what several older persons
had to say on the subject, provided a chart in which the Ababda are divided into
16 qabila (tribes), each belonging to one of three family or clan groups that all
use a distinctive wasm, or clan sign (Table 23.1).
The family and tribal relations of the Ababda as recorded by Murray in
the early 20th century are very different (Murray 1923). The 'Ogada, who are
the most prominent Ababda group at present, do not feature in his overview
(Table 23.2). The Arabic terms 'aila (family, plural 'ayaal), qabila (tribe, plural
qabayil) and beit (house, plural buyut) can be utilized at different levels, to
indicate various group compositions. These words are used where English
would employ terms such as lineage, tribe, clan, family, house, extended family or
nuclear family. The terms as used by the Ababda differ from the way they are
employed by the Ma'aza, in the northern part of the Eastern Desert and the
Awlad Ali, a Bedouin group that moved from Saudi Arabia into the Sinai, the
Mediterranean coastal area and the Western Desert, probably around the 11th
century CE. The Ma'aza are divided into approximately 20 clans (aila, most
Table 23.1. Overview of Ababda Families and Tribes:
Family or clan (wasm = brand and Main center of ‘residence’ Qabila (‘tribe’)a
sign of clan ownership)
al-Habshee Arab Saleh Haranaab
Kerdjaab
'Ogada
Nuffa'
Saadalaab
'Amaraab
ar-Rasaan Manazig, Khuda', Hemeira Billalaab
Batranaab
Zeydaab
Timeen
Farhanaab
Kimilaab
Djahalaab
al-Khortaam Hamata, Abu Greya, Abu Khusun Greidjaab
Hamidaab
Farradjaab
Data recorded by Gamaa Hussein.

a
often translated as ‘family’), stemming from the same lineage (an often mythical
forebear), and divided into households (Hobbs 1990). The Awlad Ali have a
subdivision in qabila (tribe), which in turn is divided into other qabila (clan),
subdivided into aila (family, the maximal lineage), subdivided into beit (house,
minimal lineage), which in turn is subdivided into aila (extended family) and
beit (household) (Cole and Altorki 1998). These translations may perhaps lose
the most important organizational aspect of the nomadic interrelationships,
which are determined by real, mythical or invented lineage. In theory, mobility
does not affect these relationships, because the group identity is not linked
to a geographical region, nor is it defined by land ownership. In practice, the
Ababda tend to cluster in groups of the same families, clans or tribes, and
multiple groups are found to share the same region.
Land and Resource Ownership

The Eastern Desert is an arid region, with an annual rainfall between 20 and
200 mm occurring mostly in the winter and often causing flashfloods (seyl,
plural seyul). These floods create temporary lakes near the shoreline and fill up
the water holes, while a subterranean water flow fills the aquifers, providing the
region with enough water to last for at least two or three years without additional
rainfall. Aridity is determined by the lack of precipitation, by the rate of evapo-
ration and, most important, by the irregularity of the annual rain (Krzywinski
2001; Cappers 2006). This means that in some years the Ababda can embark
on opportunistic agriculture, but regular agricultural pursuits are impossible. In
the past the Ababda cultivated only durra (Sorghum sp.). Today corn (Zea mays,
also called durra in Arabic) or barley (Hordeum vulgare) is sometimes also sown.
Growing wheat, the staple of the Ababda diet, is not feasible. Dry farming of
barley requires about 200 mm of annual rainfall, just on the border of moist years,
and wheat needs approximately 300 mm rain per year (Cappers 2006). Stable
agricultural fields are therefore not a feature of the landscape. The notion of
ownership and claims on land are often considered to depend on a settled lifestyle
and the need to protect crops. Although this may sound like a plausible supposi-
tion, it is clear that mobile populations also have territorial claims and rights of
ownership. These claims are not focusing as much on a specific plot of land as
on resources. They are obtained and maintained by a complex social system of
reciprocal observation of rights, which enables mobile peoples (hunter-gatherers,
pastoral nomads or mobile agriculturalists) to leave property behind without the
risk of losing it to other groups or individuals.
Although landownership has not become an issue until recent infringements
of the Egyptian government into Ababda territory (Cole and Altorki 1998),
ownership of resources has always been recognized. A wasm is a symbol used
Table 23.2. Overview of the Ababda Tribe and Its Subdivisions (after Murray 1923:423):
Forebear Group Qabila (‘tribe’)

Muhammad, son of Ashab, son of Muammad-ab (northern section) 'Abdein-ab
Abad 'Amran-ab
'Atiya-b
'Adwalla-b
Edidan-ab
Faraj-ab = Farradjaab
Fisheij-ab
Hameid-ab = Hamidaab
Jâral-ab
Jubran-ab
Malak-ab
Rahal-ab
Seidan-ab
Shaf-ab
Shein-ab
Shuweim-ab
Bilal, Son of Jama, son of Ashab, Bilal-ab = Billalaab (southern 'Amîr-ab = Billalaab
son of Abad section) Batran-ab = Batranaab
Firhan-ab = Farhanaab
Hamud-ab
Jahad-ab
Jidal-ab
Kirj-ab
Rajab-ab
Saadalla-b = Saadalaab
Selîm-ab
Taman-ab = Timeen?
Zeid-ab = Zeydaab
Haran, son of Jama, son of Ashab, Haranaab Haranaab
son of Abad
Harein, son of Jama, son of Ashab, Harein-ab Harein-ab
son of Abad
Nafa-b = Nuffa'
Meleik, son of Abdalla, son of Meleik-ab Meleik-ab
Abad
Fuqara
Jimeil, son of Abad? Jimeiliyîn
'Ibud, son of Abad 'Ibudiyîn
Shanatir
‘Broken tribes’ Anqar-ab
Hamej
Heteimiya
Hukm
Kimeil-ab
Qireij-ab = Kerdjaab
as a burn mark for camels, but it can also be scratched in stone or laid out
with sticks and rocks as an explicit sign of land and resource ownership. It is a
‘signature’ recognized by all Ababda to secure an aila’s claim on the resources
of a particular area and form a material trace of social organization. Members
of the family are allowed to use the scarce water resources, hunt and gather
plants or wood. Outsiders have to ask permission and can be denied access in
times of resource stress. The wasm is also known from other nomadic groups,
such as the Ma'aza, who arrived in the Egyptian Red Sea Mountains around
1700 CE, and the Rashaida, who arrived in the border area between Egypt and
Sudan in the early 20th century CE (Hobbs 1990; Roe, in press).
Although the Ababda are usually described as pastoral nomads, whose
existence was based on the possession of camels, sheep and goats (Paul 1954;
Morton 1988), much of their activities, as known from descriptions dated to
the 19th–20th centuries CE, could perhaps better be described as hunting and
gathering. The women were responsible for the herds, while the men went out
into the desert to hunt gazelle or ibex and collect wood (for charcoal produc-
tion) or medicinal plants. The latter activity is still important, and the economic
importance of herbs sold in the Nile Valley is undiminished. For their own use
the Ababda collect the pods of the Acacia nilotica (used for tanning), the wood of
the Acacia tortilis (to make coffee mortars) and Lagenaria siceria, a gourd used as
a container (loza) for cosmetic hair grease (Christensen 2001; Cappers 2006).
Men traveled to the Nile Valley to sell their products and to assist with
the harvest, especially in Lower Nubia, where many men had left to work in
Cairo as cooks, guards and servants. The Ababda received sorghum or wheat
in exchange for their help and after the harvest the flocks were allowed to graze
on the abandoned fields. This activity came to an abrupt end with the closing of
the Aswan High Dam, and the subsequent filling of Lake Nasser, in the 1960s.
The Ababda along the Red Sea shore also procured fish and mollusks. Central
to the survival strategy of the Ababda is their superior knowledge of resources
that are not easily accessible.
At present, the Ababda women still mostly lead a traditional life. They set
up temporary residences in the wadis of the Eastern Desert with their children
and flocks. Women are not allowed to milk the sheep, goats or camels, nor
are they permitted to slaughter animals. If no male relatives are present, the
animals are not milked and the lambs, kids and young camels drink their full
share. When a man grows rich, meaning that his flock is getting too large to
be tended by his wife, he can marry a second, third or fourth wife, who then
takes part of the flock to a different area. There is a direct relationship between
the size of the flock and the carrying capacity of the land (Roe, in press). If the
amount of grazing needed cannot be found within a day’s walk, then part or
all of the flock is moved to a different area.
The Over night Bag and the Portable Residence

An ‘overnight bag’ of an Abadi, the minimal resources that a man will take on
a short trip, of two or three days’ walking, consists of two sheep’s skins, open
at the neck, the legs connected with a leather strap to form a carrying sling.
The smaller bag (girbi) contains water, while the other (giraab) holds a knife,
approximately a kilogram of flour, some salt, a plastic mixing bowl, matches,
a globular coffeepot (djabana) in a protective basket (kabuta), a metal coffee
roaster, a pestle and mortar, a cylindrical basket containing three or four tiny
Chinese cups, a wooden stand with hollows to hold the cups, raw coffee beans,
sugar, ginger root and a stick of meshwak (Salvadore persica: ‘toothbrush bush’).
An archaeologist, finding such a collection of coffee-production equipment
may conclude that we must be dealing with a permanent campsite because
no person in his right mind would bring all this stuff on a walking trip. An
inventory of what an American, making a similar trip, would bring shows
that ‘traveling light’ is a relative notion. He or she might carry two gallons of
water, in a plastic container, and a backpack with a Leatherman all-purpose
tool, a torch, spare batteries, clean underwear, clean socks, sleeping bag, disin-
fecting nonwater soap gel, toothbrush and toothpaste, deodorant, snakebite
kit, camera, notebook, pain killers, bandages, trail mix bars, Nescafe, a can
of tuna, a can of tomato paste, macaroni, processed cheese triangles and an
aluminum pot. Both travelers bring their essential equipment and supplies, but
the understanding of what is essential differs substantially.
For the Ababda, drinking coffee is an important, almost ceremonial, activity
any time of day. It is closely connected with basic hospitality. When meeting, or
taking a break, the Abadi will build a small fire and boil water. In the meantime
the coffee beans are roasted over the fire and crushed in the mortar with some
ginger root. The crushed coffee is carefully guided into the narrow opening of
the djabana, which is filled with hot water, and put back in the ashes to brew.
A small bung of date-palm fiber is stuffed in the opening as a strainer. A guest
is offered three, seven or nine cups of this strong, spicy coffee, which is served
in tiny cups, half filled with sugar. Part of the sugar is dissolved by stirring the
coffee with the end of a match. The second and third cups are poured onto
the remaining sugar. Figures 23.2–23.6 present an overview of the chaîne
opératoire of making djabana.
The djabana, and the importance of serving coffee as part of a hospitality
ritual, is not restricted to the Ababda. Similar use is known in a large region
encompassing the east of Sudan, Ethiopia and Eritrea. The Ababda form the
northernmost group that uses the globular coffeepots. Emphasis on the impor-
tance of serving coffee to guests is also found at the other side of the Red Sea,
in Yemen, but there the coffee is prepared in preferably heavy, silver coffeepots.
Figure 23.2. Preparation of djabana: pouring water into a tin can.
Figure 23.3. Preparation of djabana: roasting coffee beans in a tin can with a handle.
Figure 23.4. Preparation of djabana: pounding the roasted coffee beans in a wooden mortar with a
length of reinforcement steel.
Figure 23.5. Preparation of djabana: pouring the ground coffee into the narrow opening of the coffeepot
(also called djabana).
Figure 23.6. Preparation of djabana: the coffee with ginger is ready and the coffeepot (djabana) rests on
a ring stand. Small Chinese cups are half filled with sugar, which will last for at least three servings. The
strong, ginger-flavored coffee is poured and a new batch is started immediately. The guests drink either
three, seven or nine cups, depending on the length of the meeting.
Figure 23.7. Headrest (mutir'is), acacia wood with aluminum repair; the saddle is approximately 20
cm wide.
Figure 23.8. Headrest (mutir'is), leather over wooden frame; the height is 35 cm.
Figure 23.9. A milking basket (kahel) made of finely coiled doam palm leaf.
There are more striking correlations with Ethiopian and Eritrean material
culture, for instance the use of headrests (mutir'is; Figures 23.7–23.8), rather
than pillows, and the extensive use of finely coiled basketry, often decorated
with overlays of leather strips (kahel; Figure 23.9).
In contrast to the extensive inventory brought on short trips, the entire
family’s residence fits on two camels. Extended families live in small groups
of mat houses (beit al-burush) built by the women. The only contribution to
the house by the men is during the wedding, when the husband brings all the
materials needed to build the first mat house. The Ababda could be charac-
terized as ‘matrilocal,’ at least for the first year of the marriage. In the early
evening of the first day (al-farta) of the wedding ceremony, which usually
takes five days, the husband visits the family of his wife. With a stick (assaia)
he draws the outline of the place where the house should be built. The actual
building of the house is done by the older women of the wife’s family. First
they sprinkle sugar over the area where the house is to be built, to invoke a
life of sweetness and happiness. Then, using digging sticks, they excavate nine
postholes forming a roughly rectangular ground plan, oriented east-west, with
an extra post for the entrance at the long side of the house facing south, away
from the mostly northerly winds. Forked acacia branches are placed in these
postholes, and three or four curved roots of the acacia are tied to the forked
ends to form the skeleton of the roof (Figure 23.10). All branches are tied
together with a special plaited rope, made with eight white and eight black
bundles of wool (Figure 23.11). The effect is a decorative plait of white and
black V-shapes. The house skeleton is clothed with woven woolen carpets,
called shamla (Figure 23.12), on the inside, and with mats, sewn from plaited
doam palm leaf (Hyphaene thebaica), on the outside (Figure 23.13). Each mat
of the house is approximately 3x1.5 m and consists of 11 strips sewn together,
each strip plaited of 13 strands of doam palm leaf, approximately 20 mm wide.
These mats are molded over the frame and kept in place by wooden pegs used
much in the way that pins keep together fabric (Figure 23.14). These pegs are
approximately 300 mm long and pierce through the edges of two mats and
sometimes also the underlying carpet.
The entrance of the house is formed by the nicest looking carpet. The mats
forming the roof of the house form a slight overhang at the south, which gives
the entrance some protection from the sun. The entrance carpet, effectively the
south wall of the house, can be taken up, to open up the entire house (Figure
23.15). The house has no visible internal division, but the area to the east,
away from the entrance, is considered the women’s side, while the western side
is regarded as the male side (Saidel, this volume). The house is used mostly
by the women who cook, make coffee or bake bread in front of the entrance.
Women’s tasks include spinning wool and weaving carpets, both the heavy but
mostly undecorated shamla and more supple blanket-like carpets that are used
to sleep on (farsha) or under (hemel). Carpets are woven on simple horizontal
looms, consisting of four pegs in the ground to which two cross bars have been
tied to hold the warp under tension. The warp is of coarse wool twine; the weft
traditionally consists of natural colored wool. At present, a shamla is mostly
undecorated, whereas a farsha sometimes has a simple striped decoration.
The hemel (blanket) has patterns varying from simple stripes in natural colors
(white and shades of brown to almost black) to elaborate patterns of triangles.
These are now often made with very brightly colored wool, or synthetic fibers,
bought at the market.
Tanning leather, with acacia pods, is also a woman’s task, as well as manufac-
turing leather products, such as water-carrying bags (gurbah), leather buckets
to haul water from a well (dalwa), leather tanning vessels (garrad), decorative
bags in which to keep the household’s belongings (dabiya) and a large variety of
decorative camel gear. It is the children’s task to fetch water from wells (bir), or
natural rain collection basin (galt), and to take the animals out to browse.
Figure 23.10. Building a beit al-burush: forked acacia branches and roots are secured into the ground
and tied into a frame.
Figure 23.11. Building a beit al-burush: a special rope plaited of black and white string is employed for
the bindings.
Figure 23.12. Building a beit al-burush: the inner walls of the house consist of tightly woven goat-hair
carpets.
Figure 23.13. Building a beit al-burush: the outer walls of the house are made of large doam palm leaf
mats.
Figure 23.14. Building a beit al-burush: the mats are stitched together with sharp pegs of acacia wood.
Figure 23.15. Building a beit al-burush: the box-shaped mat house is oriented with its back to the north
to protect the entrance from the prevailing wind.
Foodways and Cooking Utensils

The traditional Ababda diet was mostly dependent on milk of the flocks and
cereals exchanged or bought in the Nile Valley. The staple food was aseeda, a
porridge made from water and flour (originally ground sorghum, more recently
wheat). If available, salt, butter or milk were added to the mixture that was
cooked for an hour in a burma (a vessel cut from hamr, steatite quarried at Abu
Gurdi, Rod al-Gamr and other places in the Eastern Desert). The porridge was
eaten with the right hand from a communal wooden bowl. These large wooden
bowls (gaddah) were a precious possession and hard to come by; therefore,
almost all gaddah show extensive repairs.
A fast way to prepare food was to make dough of water and ground sorghum
(durra) and bake bread (rudaaf) on top of heated stones in a shallow depres-
sion, approximately 20 cm in diameter, lined with flat stones. A fire was built
in this and then removed to reveal the hot stones onto which the dough was
poured, which was then covered with the hot charcoal. When sorghum was
replaced by wheat, around 1950, the dough was smoother and could be baked
directly in the sand. This bread, known as kaburri or gurs (disk), is baked for
15 min, after which the charcoal is removed, the bread turned and baked for
another 15 min. After having the sand and ashes wiped off, the bread is ready
to eat. Apparently an adoption from a Nile Valley tradition is the production
of ruqaaq, thin bread, rolled out on a table and baked on the metal lid of an oil
barrel, balanced on three stones over a fire.
Meat and vegetables were not regularly added to the meal. Occasionally
gazelle or ibex were hunted and consumed. At present, hunting is illegal,
although it still occurs, not only by the Ababda but also by mostly Arab tour-
ists (Cole and Altorki 1998). The flocks of female sheep and goats were kept
intact, and the male animals were sold off immediately after the rainy season.
Sheep, goats and camels are rarely slaughtered for consumption. Occasions
when this takes place are weddings and the karama, a festive gathering of the
aila at which each qabila slaughters one or more animals. The meat is prepared
much like the bread: large stones are heated, and the meat is spread out on top
and covered by charcoal. This procedure is called salaat and is considered a
highlight of celebrating the community. Outsiders are invited (karama is from
the root krm, ‘being generous’), but they are not allowed to slaughter animals.
They can contribute by giving money. Generosity is not just a positive char-
acteristic but a serious religious obligation.
Personal Care, Clothing and Ador nment

Ababda men wear wide cotton breaches with a very low crotch, over which they
have a simple shirt to their knees, also of white cotton, and a dark colored waist-
coat. The women wear very colorful flowery dresses, covered with equally colorful
but monochrome wraps. This is the modern equivalent of traditional Ababda-style
clothing, which existed of large sheets of cotton wrapped around the body for both
men and women (Keimer 1953b). Leather sandals are nowadays mostly replaced
by plastic flip-flops, although sturdy leather sandals with soles made of old car tires
are worn as well. When, however, the going gets tough, especially on mountain
paths with treacherous sharp rocks, shoes and sandals are usually taken off.
The Ababda and Bishari men were most famous for their hairstyles, which
consisted of long curly hair on top of the head and curled tresses or plaits
hanging down framing the face and neck, and for the wooden combs they wore
as adornment (Keimer 1954b). This gained them the derogatory nickname
‘fuzzy-wuzzies’ by the British in the 18th–19th centuries CE. The hair is rubbed
with fragrant sheep’s fat, which helps to prevent it from drying out in the wind
and sun. The combs, the main function of which is to decorate the hair, are still
made of acacia wood, but Afro-style plastic combs are bought at the market as
well. The use of headrests (Figures 23.7–23.8), rather than pillows, is possibly
intended to preserve the elaborate hairstyle.
Like that of the men, the hair of the women is carefully greased and plaited.
Gold-foil ‘coins’ are worked into the hairdo, lining the forehead. Colorful
beads, red berries and shells decorate the plaits. They make the hair heavy
and enhance dancing movements. Jewelry is considered very important by the
women. Heavy silver anklets (khul-khal) are a woman’s savings. Gold is used for
earrings and nose ornaments, the most traditional ones drilled through the top
of the nose and covering its largest part with a lozenge-shaped decorated plate
of gold. Necklaces and bracelets are usually made of cheaper materials such as
shells or glass, clay or plastic beads. A woman’s adornment concentrates on the
head. The men do not wear jewelry but may be seen wearing leather necklaces
or upper armbands containing an apotropaic amulet.
Gender Priorities
While creating the EDAPP collection, to be housed in the Beit al-Ababda in
the Wadi al-Gemal Nature Park, a group of men and women were asked to
list the objects that were essential to represent their culture. The priorities
were, perhaps not surprisingly, quite different, with one exception: everything
related to the djabana was considered the most important by both men and
women (Table 23.3).
Table 23.3. Priority List for the Collection of Ababda Objects:
Mena Womena
djabana globular coffeepot djabana globular coffeepot
muhmas coffee roaster muhmas coffee roaster
hohn mortar and pestle hohn mortar and pestle
kabuta l'il djabana basket for coffeepot kabuta l'il djabana basket for coffeepot
kabuta l'il findjaan basket with cups kabuta l'il findjaan basket with cups
masaanab coffee cup stand masaanab coffee cup stand
lauwaya ring stand for coffeepot lauwaya ring stand for coffeepot
hababa fire fan hababa fire fan
sarg camel saddle hegel ankle band
sarg sennaar back-bladed riding saddle zumam nose decoration
howiyya packsaddle loza hair fat container
feraya leather leg protector mukhala kohl (eye paint) holder
mukhlaya decorative saddlebag mutir'is headrest
hemel camel blanket amud forked wood for house
zumam nose ring for camel amud roof spans
rasaan. . head gear mereiya rope for house
mahakaba / front decoration with 1 or shamla carpet: inner house walls
m. fardateen 2 tassels
lebab back decoration birsh, burush mats: outer house walls
assaia stick (camel driving) farsh floor mat
gurda goat-hair saddle band hemel decorated floor carpet
auwkaal camel hobble rihaya mill
gurbadj whip gaddah wooden serving bowl
kahel basket for milking mufrak stirring utensil
daraga shield burma cooking pot, ceramic
seyf sword giddur cooking pot, stone or copper
khandjar, shutaal dagger dabya decorated leather bag
shutaal curved dagger gurbah undecorated leather bag
ganad fire maker gurbet el-moya leather water bag
dalwa leather bucket
hulaal hair comb gur leather tanning bag
loza hair fat container abreek metal or ceramic water jug
hegaab leather bound amulet mukhlaya decorative saddlebag
gurbah undecorated leather bag muhakkaba tasseled decoration
tambura musical instrument, lute hemel saddle blanket
tabla drum do'a wedding saddle
taar large drum
fustan, khalaga, toob dress, wrap
sirwal, budja long, baggy underpants shibshib sandals
aragi long overshirt
sawakni vest
shash white cotton turban
shibshib sandals
Data were collected in 1999 in Berenike, Arab Saleh and Manazig (Figure 23.1).
a
The men listed all different parts of camel gear as the second-most impor-
tant group of objects. Every tassel, belt and bag is identified by an individual
name. It is the mother of an adolescent boy who produces these items, to deck
out the camel of the proud youth. Most of the camel gear is purely decorative:
the saddlebag can hardly contain goods, but its long colored tassels enhance and
exaggerate the swinging motion of a running camel and make for a handsome
sight. The women’s equivalent, the wedding saddle, was not mentioned with
the same sense of urgency as the men’s camel decorations. Wedding saddles
are a spectacular sight, built up of a palm leaf basketry core, covered with red
velvet, and decorated with a carpet of kauri shells, waving ostrich feathers and
beads of glass and silver. The women did mention camel gear, the production
of which is their responsibility, but only after the household utensils. The
women concentrated on items for adornment, which were mentioned imme-
diately after the djabana gear. Jewelry is embellishment and at the same time
a woman’s capital.
For the men, shield, sword and dagger came immediately after the camel
adornments and saddles. Symbols of hunting, warfare and feasting, these
items are considered an integral part of Ababda male identity. Clothing was
initially not mentioned by men or women, even though the Ababda attire is
very different from what is worn in the Nile Valley. Hair combs, however, were
considered an important part of Ababda identity. The wooden, and at present
also plastic, combs are often hardly visible in the large shocks of hair; neverthe-
less, they are an integral part of the outfit. A drawing made by Saad Mansur
in 1998 summarizes the quintessential young Ababda male: on camelback
traveling through the wide desert landscape with mountains and acacia trees,
bearing shield and sword, comb in hair, riding a well-decked-out camel (Figure
23.16). The women, instead, mentioned all parts needed to build a house.
Immateriality of Ababda Culture

The strong social organization of the Ababda is not expressed by impressive
monumental architecture, representing social stratification in palaces for the
living or in mausoleums for the dead, but rather by feasting. The social integ-
rity is safeguarded by karama, irregular meetings at which several sheep or even
camels are slaughtered and consumed. At such a feast everybody shares in the
wealth; even those who could not afford to slaughter take part in the sumptuous
meal, consisting of sand-baked meat and bread. The location of these feasts is
not fixed, but some areas are considered more suitable than others. Karamas
are mostly celebrated on the wadi floor, where there is sand suitable for baking
bread and stones for roasting the meat. The remains of the feast, stacks of bones
(the skins are taken) and concentrations of charcoal, will leave a material trace.
Figure 23.16. Drawing by Saad Mansur of Ababda men and their camels in an Eastern Desert landscape.
Note the whips, shields, swords, feraya (leg protection covering the camel’s shoulders), tassels and
decorative hair combs.
One episode of rainfall and the subsequent flash flood, however, will probably
be sufficient to wash away most of the evidence of a feast.
The karama is not a religious phenomenon, although religious festivals
are sometimes combined with a karama. A good example is the maulid (birth
festival) of Sheikh Shazli, who reportedly discovered coffee. His tomb is hidden
deep in the Eastern Desert in a wadi surrounded by steep mountains. The
simple tomb has recently been adorned with a mosque and amenities for the
thousands of visitors that come and stay here every year during a pilgrimage of
several days. The architecture of the tomb of Sheikh Shazli is unlike the Ababda
tombs in the area. These are isolated rings of stone, often with an installation
indicating the direction of Mecca (qibla).
The Ababda are Muslims and trace their ancestry back to Saudi Arabia,
but they have their own take on Islam. Religion, magic and medicine are one
continuum. The great esteem in which deceased holy men are held, such as
Sheikh Shazli who died on his return trip from the hadj (pilgrimage to Mecca),
reflects their potential as mediators between humanity and God. In addition,
it is the local sheikh who is not only judge and adviser but also provider of
amulets against snakes, scorpions and other dangers. These amulets (higab) are
personal and nontransferable. Herbs play some role in medicinal treatment of

illness, but more important is the intervention of the sheikh, who can place kyet
bi-an-nahr (fire-signs), burn marks on the skin made with red-hot nail heads
(Barnard 2000). Protection against infant mortality is provided by putting a
miniature sword, in a miniature scabbard, in the crib. This allows the baby to
fight off the large mythical bird that preys on infants.
Swords used to be important in daily life and are still worn regularly by
almost every man who owns one. Sword and shield stand for masculinity and
protection. They are also used in dancing and form an important part of every
festivity, be it a birth, a wedding or a karama. The sword dance is performed
by two men, each holding a shield, made of elephant hide, in his left hand and
raising a sword in the air with his right hand. While stamping and jumping to
the rhythm of the accompanying instruments and songs, each man shakes the
flexible steel sword with short jerky movements so that it is kept in constant
vibration (Figure 23.17). The swords are never used for mock attacks. The dance
lasts for one to two minutes, after which the two dancers quickly put down their
accoutrements and return to the ring of onlookers. Two others speedily take
their place. When asked why the dancers were only allowed such a short period,
the answer was that everybody in the company wanted to have his turn.
Figure 23.17. The Ababda sword dance is performed by two men making their swords vibrate in the
air while accompanied by the clapping and singing of other men. They quickly give up their place to a
new pair of dancers.
The instruments used to accompany the singers are a small or a large drum
(tabla and taar respectively) and a kind of lute with four strings (tambura),
the resonance box of which is traditionally made of the shell of a turtle.
Alternatively, a plastic bucket or a broken jerry can functions as drum, and a
tambura can be made out of scraps of wood. Songs sung with the sword dance
are called hoseeb or simply seyf (sword). These are heroic songs in which the
sword is said never to rest in the scabbard. The bir (well) songs tell about clean
water in abundance. Adolescent boys have a different set of songs, such as the
bagraab, which accompany a different type of dance in which the dancers jump
straight into the air with both feet held together.
All dances are strictly segregated by gender and age. The girls sing harkaak
songs, which are mostly about the beauty of the camel. The women’s dance is
only performed in isolation, with no men present. One woman will move to
the center of the circle, while the others sing and clap the rhythm. She pulls
back the toob from her head and reveals her magnificent hairdo, decorated with
coins, shells and beads. Then she flicks her head and arms backward and curves
back slowly to her starting position. The accent of the music corresponds with
the repetitive backward jolt (Figure 23.18). Women also sing songs called
donub, which express the wish that there will be rain, or celebrate the prospect
Figure 23.18. The dance of Ababda women is performed by one woman rhythmically throwing back
her head and arms while accompanied by the clapping and singing of other women.
of getting together with others. These two subjects are closely related. The
women are often living an isolated life in separate wadis until the rain enables
or forces them to come together in one area. This has the added benefit of
enabling social interaction.
Mobility, Distance and Social Life

The rain in the Eastern Desert is usually concentrated in the winter and
in the south, in the Gebel Elba region near the Egyptian-Sudanese border.
Although the Ababda groups claim ownership of resources, there are no rules
on the grazing in a period in which rainfall is limited. If grazing is insufficient
to maintain the herds, when the rains have failed for a number of years, the
herds move to the south (Roe, in press). Mobility is triggered by natural
circumstances and does not follow a pre-established pattern, nor is it limited
to certain periods of the year (Morton 1988). Physical proximity of individuals
and groups is, therefore, mostly an effect of the carrying capacity of the land.
Families come together in the rainy season, in those regions that receive
the rain. This mobility prevents the dry regions from being overgrazed and
provides a period of recuperation of the local flora and fauna.
Because the rainfall is often very localized, knowledge of the situation in
the entire Ababda region is of the utmost importance. As recorded for the
Bishareen (Morton 1988), the Ababda exchange information from person to
person. Travelers are expected to take the time to provide news on rainfall, feed
growth and movements of people. The communication system, in a region that
until today is devoid of (cellular) telephones, functions with surprising accuracy
and speed. Morton (1988) recorded from oral history that in 1914 “a party
of Ababda had walked 500 km through mountain and desert on the accurate
information that work was available.” Our archaeological project, in the 1990s,
relied on this network to attract a group of up to 80 Ababda workmen. After
our unannounced arrival it would take no more than three days for groups
of Ababda to arrive at our campsite. At least 30 came from Wadi Khareet in
the Nile Valley. In contrast to 1914 they arrived by bus and pickup truck, but
they had received the ‘accurate information’ about the availability of work in
the same way.
The other opportunities that Ababda, especially the women, have to interact
are during special occasions, such as weddings and karamas. One woman, living
60 km from the nearest asphalt road, and about the same distance from any
other Ababda camp, informed me that she would travel “from the end of the
world” to participate in these occasions, especially if they involved a member
of her family.
Impact on the Landscape

The impact of the Ababda on the landscape is limited. Although all their economic
activities employ natural resources (grazing, charcoal burning, collecting medicinal
plants), the Ababda are selective in where and when they collect these resources.
During a collection trip to gather wood for a beit al-burush, two Ababda first went
to speak to the head of the family who owned the resources. He gave his consent
and joined the two others to help with (and control?) the wood collecting. Nine
forked branches of acacia were needed, to function as the ‘pillars’ of a house.
This involved visiting four different clusters of trees, spread out over a distance of
approximately 7 km in the same wadi. Only one branch was taken from each tree
and not more than two from the same cluster. This method of sustainable tree use,
which concentrates on browsing, is called ewak (Krzywinski 2001). The curved
pieces of wood for the roof of the house, actually tree roots, are only collected
opportunistically after being laid bare by a flash flood. They are kept and moved
from campsite to campsite until needed.
The Ababda will once in a while embark on hunting trips or will try to trap
animals, although bird hunting is increasingly rare. Part of the Ababda terri-
tory has recently been made a national park, and some of the Ababda have
been trained as rangers. The hunting of protected species, such as ibex and
gazelle, is now officially illegal in the entire region of the Eastern Desert, but
such activities, especially when instigated and financed by tourists from Saudi
Arabia (Cole and Altorki 1998), are very difficult to monitor.
Since the 1960s the Ababda have increasingly been able to purchase light
pickup trucks. With the exception of transport to some hard-to-reach areas,
the Toyota Hilux has mostly replaced the camel as beast of burden. Ababda
drivers are masters in negotiating the sandy wadis, knowing the types of sand
that require slow driving and the kinds that call for fast driving. There is a
distinct impact of these pickups on the wadi floor and flora, especially because
no tracks can be established: driving in the same path is a certain recipe to
get stuck. Increasingly, however, the Ababda are faced with churned-up wadi
floors, which challenge even the best of drivers, where tourist companies and
individuals have ventured out to enjoy a desert safari. The much heavier terrain
vehicles are far more destructive, not only because of the churning wheels but
also because they enable larger groups of people to reach more remote areas.
Ignorant of the cultural and natural rules and requirements of the region, these
visitors often behave like trespassers, cutting wood that ‘belongs to nobody’ in
an ‘empty land’ for a fuel-devouring fire that gives the city dweller a safe and
romantic ‘into the wild’ experience. Impact on the cultural and natural land-
scape of the Eastern Desert is, therefore, mostly the result of recent changing
modes of transportation and land use.
Ababda Material Tr aces

Our work with the Ababda has provided insights into the traces that are left
behind and how these can be used to understand ancient mobile populations.
The Ababda do not leave extensive garbage deposits, however, and much of the
refuse is eaten by sheep and goats. When a mat house is taken up and moved
to another location, a rectangular clearance, where stones and pebbles were
removed from the inside of the house, is clearly visible. Traces of the shallow
postholes can sometimes be discerned, but the precious wood is taken along
or, if no longer in usable condition, burnt as fuel. Concentrations of animal
dung are rare, as this too is used as fuel.
The most recognizable signs of Ababda presence are the things that have
been left behind on purpose. At wells and water holes leather buckets are left, to
enable other women and children tending the sheep, or thirsty travelers, to haul
up the water. In addition to communal property there are also personal belong-
ings. Such possessions are the things that can be carried, on foot or camelback,
or were left behind in a conspicuous place, often in trees. Similar behavior has
been attested elsewhere (Eerkens, this volume), but the personal belongings
of the Ababda are usually not hidden. Sometimes wooden platforms are built
on which individuals or families leave their possessions until they return to the
same area. These platforms are 1–2 m above the ground to keep the items safe
from sheep, goats and flash floods. The Ababda social code prevents anybody
from touching these unguarded ‘deposits,’ even if the person has died and will
not return to claim his or her belongings (Figure 23.19).
Tombs are another material sign of Ababda presence, but they are difficult
to interpret because they are widely dispersed and often consist of only a ring
of stones or a headstone. The tombs of important men are decorated with a
low stone enclosure and colorful flags. Green, the color of Islam, is dominant,
bleaching to a tattered white over time. Inside the enclosure a deposit of
personal belongings and provisions are placed: cigarettes, sandals and one or
two cooking vessels (Figure 23.20).
Even the immaterial aspects of Ababda life, such as celebrations, feasting
and religious beliefs, leave material traces in the form of bone scatters, amulets
and other apotropaic items, such as the miniature swords and scabbards. The
impact of the nomadic groups on the landscape in earlier periods was probably
equally limited and can be difficult to trace archaeologically. It is interesting to
note that at present the traces that are most obvious are those left on purpose,
such as the ones outlined here and the wasm resource ownership signs intro-
duced above.
Figure 23.19. An acacia tree with an Abadi’s belongings: a platform of sticks tied together holds a plastic
jerry can and several items wrapped in a checkered blanket. At the foot of the tree sits a metal can with
water; to the right lies a crate of the type in which fruits from the Nile Valley are transported.
Figure 23.20. An Ababda tomb, decorated with flags and a few grave goods. This tomb had two soapstone
vessels and a pair of sandals.
Discussion
This description of the Ababda culture was made with the help of our Ababda
team members. Although they participated in the process, the chapter at hand
is based on the features that I consider of importance to include and undoubt-
edly differs from what the Ababda writers’ team would have selected. The
purpose of their original request to me, to write a book about them, stemmed
from a notion that their way of life was under pressure and on the verge of
changing forever. The collection of Ababda artifacts was the starting point for
discussions on the less tangible aspects of Ababda way of life, including social
values and history. I will try to outline here at which points the various Ababda
histories diverge.
The Ababda refer to two types of history: the history of the Ababda origin
and the cultural memory of the Ababda way of life. The first is reflected in a
great interest in genealogical trees (Tables 23.1–23.2). The main purpose of
the genealogy is to trace the Ababda to Saudi Arabia and establish a firm link
with the birthplace of Islam and, preferably, a direct relation to the prophet
Mohamed. A secondary purpose of the virtual map of tribal relations is to
establish communality. Through common forebears, along patrilineal connec-
tions, every individual can define his or her relation to the larger group. The
Ababda consider themselves part of the Beja and related to the Bishareen, but
at the same time they clearly distinguish themselves. They also see themselves
as being fundamentally different from the Nubians and the Egyptians. The
latter term is not only used to indicate the settled population in the Nile
Valley, and now also along the Red Sea, but also to indicate power relations.
The ‘Egyptians’ with whom the Ababda have direct dealings are mostly repre-
sentatives of the army or the civil administration. The cultural memory of the
Ababda is related to landscape, location of resources, economic strategies and
cultural contacts. More recent oral history, of encounters with German and
British military during World War II or British and Italian representatives of
mining companies, are mingled with stories about fathers and grandfathers
who booked successes in the hunt. The recent history of the Ababda does not
consist of golden memories, remembering a blissful period. The drawbacks of
their poor existence without many resources are all too clear.
The historical interest of the Ababda differs greatly from that of Keimer,
the Egyptologist who visited the region in the late 1940s. His main interest in
the history of the Ababda was to emphasize a cultural continuity with ancient
Egypt. Keimer was certainly a keen observer, but the object of his studies was
the material culture of the Ababda rather than the Ababda culture as an entity
within its own specific context. His rather simplistic use of analogy between
ancient and modern cultures, based on random ethnographic parallels, failed
to take into account the cultural development and historical context of both
Pharaonic and Ababda culture and preserved a biased, static view of each civi-
lization. Keimer’s approach objectified and had no interest in giving a voice to
the Ababda and Bishareen.
My own approach of Ababda history originated in a working relationship
and discussions during our work. It was fueled by questions about the activities
of the ancient inhabitants of the region, the Ptolemaic and Roman settlers who
came from the Nile Valley, and the nomadic population of which we found
evidence in our excavations. Trying to gain an understanding of the factors that
determine Ababda mobility, while observing the ingenuity and improvisation
with which the Ababda make things work with a minimum of possessions, was
very helpful in thinking through our own interpretations of the past. At the
same time, observations of residual material traces, the expression of social
networks, and the material overrepresentation of certain aspects of the culture
(in the case of the Ababda all the equipment needed for the ‘coffee ceremony’)
gave important caveats to prevent too simple an explanation of archaeological
data. For instance, Eastern Desert Ware, a specific type of ceramics found in
the harbor town of Berenike in Late Roman deposits, was interpreted as a
cultural or ethnic marker, based on the shape, decoration, fabric, context and
distribution of the vessels. The importance of the djabana in Ababda culture
does not by any means constitute ‘proof’ for such an interpretation. It does,
however, provide an example of a mobile culture in which specific objects,
which may seem like ‘ballast’ to an outsider, constitute an essential part of the
material culture and are carried around at all times.
Thus we have at least four distinct interpretations of an Ababda past:
its Pharaonic roots, its Islamic base, its recent daily life history, and its
ethno-archaeological or ethnohistorical value. These interpretations are
not necessarily in conflict with each other, but it would be unsophisticated
to propose that they consist of four sets of ‘objective’ observations, coated
with different interpretations, which can be combined or harmonized to flesh
out the complete picture. At the outset each of these histories started with
particular explicit or implicit concerns and, more important, with different
definitions of what constitutes a valid method to reach conclusions. Multiple
interpretations of the past are an inherent effect of the scholarly discourse but
also of the recognition that scholars and scientists present a very specific set of
approaches. Even though the debate on its significance and consequences has
scarcely started, multivocality allows nonacademic voices to be heard. That is
why it is important that the Ababda write their own books.
References
Barnard, H.
2000 Geneeskunst geïnspireerd door armoede [Medicine Inspired by Poverty, in
Dutch]. Nederlands Tijdschrift voor Geneeskunde 144: pp. 949–951.
Data. In J. C. M. Starkey (ed.), People of the Red Sea: Proceedings of the Red Sea
Project II, Held in the British Museum, October 2004. Society for Arabian Studies
Monographs number 3. BAR International Series 1395. Oxford, Archaeopress:
pp. 23–40.
Barnard, H., A. N. Dooley and K. F. Faull
2005 New Data on the Eastern Desert Ware from Sayala (Lower Nubia) in the
Kunsthistorisches Museum, Vienna. Ägypten und Levante 15: pp. 49–64.
Barnard, H. and E. Strouhal
2004 Wadi Qitna Revisited. Annals of the Náprstek Museum Prague 25: pp. 29–55.
Cappers, R. T. J.
2006 Roman Foodprints at Berenike: Archaeobotanical Evidence of Subsistence and Trade
in the Eastern Desert of Egypt. Los Angeles, Cotsen Institute of Archaeology.
Christensen, A.
2001 Charcoal and Coffee. In K. Krzywinski and R. H. Pierce (eds.), Deserting the
Desert: A Threatened Cultural Landscape Between the Nile and the Sea. Bergen,
Alvheim & Eide Akademisk Forlag: pp. 109–132.
Cole, D. P. and S. Altorki
1998 Bedouin, Settlers, and Holiday-Makers: Egypt’s Changing Northwest Coast. Cairo,
American University in Cairo Press.
De Jong, R.
2002 Notes on the Dialect of the 'Ababda. In W. Arnold and H. Bobzin (eds.), Sprich
doch mit deinem Knechten aramäisch, wir verstehen es. 60 Beiträge zur Semitistik:
Festschrift für Otto Jastrow zum 60. Wiesbaden, Harrassowitz: pp. 337–359.
Eide, T., T. Hägg, R. H. Pierce and L. Török
1998 Fontes Historiae Nubiorum: Textual Sources for the History of the Middle Nile
Region Between the Eighth Century BC and the Sixth Century AD. Volume 3: From
the First to the Sixth Century AD. Bergen, University of Bergen, Department
of Greek, Latin, and Egyptology.
Hobbs, J. J.
1990 Bedouin Life in the Egyptian Wilderness. Cairo, American University in Cairo
Press.
Keimer, L.
1951 Notes prises chez les Bišarîn et les Nubiens d’Assouan: Première partie.
Bulletin de l’Institut d’Égypte 32, Session 1949–1950: pp. 49–101.
1952a Notes prises chez les Bišarîn et les Nubiens d’Assouan: Deuxième partie.
1952b Notes prises chez les Bišarîn et les Nubiens d’Assouan: Troisième partie.
1953a Notes prises chez les Bišarîn et les Nubiens d’Assouan: Quatrième partie.
Bulletin de l’Institut d’Égypte 34, Session 1951-1952: pp. 329–400.
1953b Notes prises chez les Bišarîn et les Nubiens d’Assouan: Cinquième partie (1).
1954a Notes prises chez les Bišarîn et les Nubiens d’Assouan: Cinquième partie (2).
1954b Notes prises chez les Bišarîn et les Nubiens d’Assouan: Sixième partie. Bulletin
de l’Institut d’Égypte 35, Session 1953-1954: pp. 471–533.
Krzywinski, K.
2001 Cultural Landscapes: An Introduction. In K. Krzywinski and R. H. Pierce
(eds.), Deserting the Desert: A Threatened Cultural Landscape Between the Nile
and the Sea. Bergen, Alvheim & Eide Akademisk Forlag: pp. 9–60.
Morton, J.
1988 Sakanab: Greeting and Information Among the Northern Beja. Africa: Journal
of the International African Institute 58: pp. 423–436.
Murray, G. W.
1923 The Ababda. Journal of the Royal Anthropological Institute of Great Britain and
Ireland 53: pp. 417–423.
1967 Troglodytica: The Red Sea Littoral in Ptolemaic Times. Geographical Journal
133: 1, pp. 24–33.
Paul, A. M.
1954 A History of the Beja Tribes of the Sudan. London, Cass.
Pierce, R. H.
2001 Past and Present in the Eastern Desert. In K. Krzywinski and R. H. Pierce
(eds.), Deserting the Desert: A Threatened Cultural Landscape Between the Nile
and the Sea. Bergen, Alvheim & Eide Akademisk Forlag: pp. 143–166.
Roe, A.
in press A Preliminary Investigation of the Parameters of Premodern Economic
Production and Exchange in the Desert Hinterlands of the Berenike
Archaeological Project Area. In S. E. Sidebotham and W. Z. Wendrich (eds.),
Report on the 2001 Excavation Season at the Greco-Roman Harbor of Berenike
(Egyptian Red Sea Coast) and Excavations at Abu Greiya. Los Angeles, Cotsen
Salih, H. M.
1980 Hadanduwa Traditional Territorial Rights and Inter-population Relations
Within the Context of the Nature Administration System (1927–1970). Sudan
Notes and Records 61: pp. 118–133.

CNWS.
CNWS.
1998 Berenike 1996: Report of the 1996 Excavations at Berenike (Egyptian Red Sea Coast)
and the Survey of the Eastern Desert. Leiden, Research School CNWS.
1999 Berenike 1997: Report of the 1997 Excavations at Berenike and the Survey of the
Egyptian Eastern Desert, Including Excavations at Shenshef. Leiden, Research
School CNWS.
2000 Berenike 1998: Report of the 1998 Excavations at Berenike and the Survey of the
Egyptian Eastern Desert, Including Excavations at Wadi Kalalat. Leiden, Research
School CNWS.
2007 Berenike 1999–2000: Report of the 1999 and 2000 Excavations at Berenike and the
Survey of the Egyptian Eastern Desert, Including the Survey of Sikait. Los Angeles,
Cotsen Institute of Archaeology.
Updegraff, R. T.
1988 The Blemmyes I: The Rise of the Blemmyes and the Roman Withdrawal from
Nubia Under Diocletian. In W. Haase and H. Temporini (eds.), Augstieg und
Niedergang der römische Welt, Volume 2 (10.1). Berlin, Walter de Gruyter: pp.
44–97.
Ch a p t e r 2 4
No Room to Move
Mobility, Settlement and Conflict
Among Mobile Peoples
R o g e r L . C r i bb
W hereas mobile people have usually been studied in terms of their

ecology and patterns of migration, I will focus on the implications
of mobility for the formation of temporary settlements. It is suggested that
mobile people are typically able to form communities of optimal size and
composition, thereby minimizing the level of interpersonal conflict: “So there
was not enough pasture land for the two of them to stay together, because
they had too many animals. So quarrels broke out between the men who took
care of Abram’s animals and those who took care of Lot’s animals. . . . Then
Abram said to Lot, ‘We are relatives and your men and my men shouldn’t be
quarreling. So let’s separate. . . . You go one way and I’ll go the other way’ ”
(Genesis 13:5–9). The consequences of permanent, fixed settlement for people
with a tradition of mobile lifestyles are considered here in relation to the stress
caused by unfamiliar housing and higher density living and the problems of
regulating conflict. Australian Aboriginals and Near Eastern pastoral nomads
are considered as examples.
Ethnographers and archaeologists dealing with mobile peoples have tended
to focus on those aspects of mobility connected with subsistence and to neglect
its sociopolitical dimensions (but see Tapper 1979 and Bates 1973). In this
chapter I explore the implications of a mobile lifestyle for social organization
and particularly for the management of conflict. Nomadic pastoralists and
those with a nomadic tradition, as well as mobile hunter-gatherers and those
with a hunter-gatherer tradition, are considered. While the economies and
logistics of mobility in nomadic-pastoralist and hunter-gatherer cases are very
different, there appear to be certain properties inherent in the practice of a
mobile lifestyle that cut across differences in subsistence, economy, ecology
and social organization.
Conflict Among Contempor ary
543
544    R o g e r L . C r i b b
Aboriginal Populations
In the 1970s James O’Connell wrote a paper called Room to Move (O’Connell
1979). It described an Aboriginal desert camp in Central Australia that had
been set up adjacent to a pastoral property, where some of the people worked.
These people, the Alyawara, had not been long in contact with white settlers;
two generations at most. The camp, consisting of various ‘humpies’ constructed
of tarpaulins and corrugated iron, was an informal affair, laid out according
to Alyawara custom in clusters of households arranged loosely according to
kinship distance. The camp was characterized by an open layout and low
density. O’Connell reported that the level of interpersonal conflict was low.
Disputes, mainly of a domestic nature, were usually solved by one of the
antagonists moving to a different household—often the woman moving into
an all-female household, for some time at least. The level of alcohol consump-
tion was low. Though partly integrated into the mainstream economy, these
people were living in a ‘tribal’ society. By virtue of voluntary associations and
appropriately spaced dwellings, however flimsy and dilapidated, they were
able to maintain a reasonably well integrated and peaceful community. They
had room to move.
In the city of Cairns, in north Queensland, Aboriginal people originally
from the communities in Cape York Peninsula are currently living an equally
‘tribal’ lifestyle. Kinship reckoning is paramount for them as well, but the
outcomes are quite different. They wander the streets in search of hidden
nooks and crannies where they can drink or sleep. Some find accommodation
at a designated night shelter or in one of the diversionary centers for those
too intoxicated to help themselves. Relatives with access to flats or houses in
the city are overwhelmed. There are large areas of the city center where they
dare not go. Ragged camps spring up on small lots of vacant land on the city’s
outskirts (Cribb 2003). People are hemmed in and crowded together. Disputes
escalate and frequently become violent. They have no room to move.
Horizontally and Vertically

Integr ated Societies
By ‘tribal society’ I mean one in which overlapping kinship ties determine the
allocation of resources, as opposed to one based on allocation through some
kind of division of labor. Tribal societies are a good example of what I will call
‘horizontal integration,’ where all people live much the same lifestyle and there
is little specialization in economic pursuits. In contrast, a vertically integrated
society is one in which there is a great range of lifestyles, and economic activity
is organized around a wide range of economic specializations. For the most
No Room to M o v e      545
part the allocation of resources does not occur along kinship lines but through
specialized institutions and markets. This is not to say that tribal people do not
interact with or participate in much wider economic and social systems. They
frequently do. Degrees of mobility and population density affect horizontally
organized societies far more than those that are vertically organized. Whereas
in the former case solutions are found in two dimensions, the latter allows
multidimensional solutions.
Mobile people are to a very large extent organized horizontally and live
in tribal societies; however, not all tribal people are mobile. There is a wide
range of tribally organized people who are settled horticulturalists, are agri-
culturalists, or practice some form of limited transhumance. Many tribes in
the Middle East are divided into mobile pastoral and settled sections with
some interchange of personnel between the two (Cribb 1991). The Yanomamö
horticulturalists in the Amazon basin are village dwellers who move their entire
communities in response to ecological factors and intervillage conflicts (the
two are often related), as do many other ‘slash-and-burn’ horticulturalists for
similar reasons (Chagnon 1976). The last two cases may be ‘mobile,’ but they
are not strictly speaking ‘nomadic’ because they do not move their productive
capital with them (Cribb 1991).
Whereas vertically organized societies offer many contexts for interpersonal
interaction (household, workplace, clubs and other voluntary associations)
with different sets of people, those organized horizontally allow far fewer such
contexts. Everyone is playing much the same game, doing much the same thing
within much the same context and within the same kin-related sets. This is
illustrated in Figure 24.1, which models a network of intersecting kin group-
ings. This horizontal organization has important implications in relation to
conflict avoidance and resolution. Whereas the urban dweller can escape family
stress at work and work stress at home, those who are tribally organized gener-
ally have access to far less variety in their choice of associations, the primary
ones being kin-based. Vertically organized societies, in short, offer not only
manifold domains of interaction but also more nooks and crannies in which to
hide from conflicts and potential conflicts.
Toler ance Thresholds Relating to Conflict

Although kinship networks can be a vital source of support, they can also
generate stress and conflict. Tribal societies handle this in various ways, for
example by means of tribal councils, groups of elders or elected headmen
presiding over spatially discrete groups or different levels of organization.
It has been suggested that optimal group sizes for different purposes may be
somehow wired into human interaction. After many years of research in the
546    R o g e r L . C r i b b
Figure 24.1. Aboriginal painting by Timmy Japangardi (Papunya, central Australia), showing the
relations between individuals in a group, mapped onto the paths between water holes in the landscape.
New Guinea highlands, Anthony Forge concluded that “Homo sapiens can only
handle a certain maximum number of intense social relationships, successfully
distinguishing between each. When the number of relationships he is involved
in rises above this figure he can only continue by classification of relationships
to cut down the total number of different relationships he has to act in and
carry information about around in his head” (Forge 1972:375).
There are optimal sizes for boards of directors, classrooms, councils and
military organizations (section, platoon, company, battalion), each of which

has its own tolerance threshold. It is precisely at these tolerance thresholds
that communication tends to break down and conflicts develop that will
frequently result in subgroups forming and going their separate ways. Where
such groups are engaged in the exploitation of resources and involve potential
population growth, other factors come into play. Those fracture points where
organizational breakdown occurs are often associated with breakdowns in the
management of resources. Tensions generated at these organizational thresh-
olds often coincide with ecological crises, as with the budding off of Yanomamö
villages as neighboring resources become exhausted, giving rise to intracom-
munal strife or the splitting of herding camps among nomadic pastoralists as
conflicts come to a head over pasture or migration tracks. Those tribal groups
who lead mobile lifestyles have the inherent capacity to manage such tensions
and information overload and to avoid or defuse any resultant conflicts by
simply translating social distance into physical distance.
Tapper (1979) has followed up the biogenic argument and considered the
implications of this important property of mobile territorial systems. He has
drawn attention to the way in which mobile people are able to map out their
social organization on the ground, by way of fluidity in their settlement systems,
in a way that settled people cannot, even if the latter are tribally organized: “It
will be among nomads rather than among settled peoples that we can expect the
emergence and evidence of any inherent social dynamic processes generating
interactional communities of a certain size and character . . . The same factors
may operate in settled societies . . . but it is sooner in mobile and flexible than
in stationary and fixed societies that we can expect them [biogenic factors] to
express themselves in social groupings” (Tapper 1979:61).
As I suggested earlier, mobility, whether among pastoral nomads or hunter-
gatherers, is usually treated by ethnographers in terms of economic objectives
and access to resources, but mobility has other vital implications. While atten-
tion is commonly paid to sociopolitical factors, researchers rarely explain how
these affect or are affected by mobility. Whereas those occupying permanent
settlements are frozen into a fixed pattern, be it kinship based or not, mobile
people may exercise the option of not only choosing the location of their settle-
ments but also of adjusting the density of these settlements so that households
and wider kin-based groups do not impinge unduly on each others’ activities
or privacy. They also have the option of choosing their immediate neighbors.
Whereas vertically organized societies achieve privacy through enclosure,
horizontally organized ones achieve it by spreading out.
Mobility between settlements therefore contains the possibility of mobility
within settlements in terms of the location of households over time. Mobile
people are able to choose their immediate neighbors on a seasonal basis so
548    R o g e r L . C r i b b
that temporary settlements closely reflect the existing pattern of social rela-
tions (Bates 1973). An excellent illustration of this, though on a small scale,
is an incident reported by Frederick Barth among the Basseri nomadic pasto-
ralists in southwest Iran. During an overnight stop on the migration trail,
two families were involved in a dispute over a choice tent site. The dispute
culminated in each family’s dragging its tents and belongings to opposite
ends of the camp, thereby literally defusing a conflict by means of physical
distance (Barth 1961).
Different cultures display different tolerance levels for crowding. Vertically
organized urban societies are able to withstand extremely high population
densities in contrast to those societies that are horizontally or tribally orga-
nized. In the latter, tolerance levels may be much lower for crowding by
strangers than by those who are bound by close ties of kinship. For tribally
organized or ‘segmentary’ societies even kin groups, or groups based on
kinship, may exhibit different degrees of density tolerance for different levels
of tribal organization.
Responses to Settlement Density

The obverse of the proposition that mobility acts to defuse conflict is what
happens when people are forced together in fixed settlements. Increases in
tension and potential conflict are accommodated by putting up walls, in both a
literal and figurative sense. This process is taken to its extreme in the case of the
modern city, where the privacy so obtained is reinforced through anonymity.
The beginnings of this process can be seen in the sedentarization of nomadic
pastoralists. First the settlement pattern becomes fixed. Next, storage structures
spring up, and eventually a permanent dwelling may be constructed, although
the family may continue to spend most of its time in the tent (Watson 1966),
what I have called elsewhere a ‘composite settlement’ (Cribb 1991:154–157). In
many cases courtyards with walls are constructed or houses begin to abut each
other as the settlement grows (Nissen 1968; Sweet 1974). In this way kinship
structures may still be preserved, though in a fixed spatial form unresponsive
to generational changes in household composition.
In Nomads in Archaeology I developed a model called the ‘domestic complex’
for examining intrasite spatial organization (Cribb 1991). This model examines
the internal organization of household units in which the pattern of structures
and activities tends to be remarkably consistent across households. For nomadic
Yörük, in southern Turkey, the complex occupies a rectangular area of about 10
by 20 m, aligned according to slope, terrain, aspect, prevailing wind direction
and location of water bodies. Crucially, no two domestic complexes overlap,
and they are usually separated by a distance over which a conversation can be
held. Tent sites and other features, such as external hearths and storage facilities
(fixtures), remain until re-occupation the following year and may be occupied
and used by different households. Such settlements are highly resistant to
crowding, which may partly account for the generally low levels of interper-
sonal conflict observed within nomadic communities (Swidler 1972; Watson
1979). It is of great interest that, while an increase in interpersonal conflict
has been observed among sedentarizing pastoral nomads (Swidler 1972), the
process occurs within the framework of a single culture.
In the Near East the internal structure of a tent site has much in common
with the living room of a village house occupied by members of the same ethnic
group, as has the etiquette of household life. As sedentarization proceeds, the
former is converted with relative ease into the latter. Indeed, settlement most
commonly occurs within the framework of a wider tribal entity in which the
majority of households are settled. It is true that some forcibly settled nomads
are known to abandon the houses that have been built for them because they
do not conform to cultural norms. This, however, presupposes the existence
of a blueprint for household organization. Pastoral nomads are part of a
wider tradition of which housing is a vital part. This is where the sedenta-
rization of nomadic pastoralists differs fundamentally from that of mobile
hunter‑gatherers.
Mobile people from a hunter-gatherer tradition undergoing sedentarization
possess no such architectural blueprints. They are part of a tradition in which
fixed and permanently occupied housing with a standardized floor plan has
no place. Although temporary shelters of various kinds were employed, the
organization of activities and space depended very much on the immediate
environment, including wind direction, shade patterns and ground surface.
We lack significant data on the settlement patterns of precontact or contact
Aboriginal societies. Excavations in Australia have been of insufficient extent
to throw much light on this question. The accounts of early historical ethnog-
raphers or observers tend to focus on aspects such as material culture, food
procurement, spirituality or the formal properties of kinship systems. Modern
regional archaeological studies have thrown some light on broad patterns of
seasonal movement, as have modern ethnographic and ethno-archaeological
studies. But neither gives very much attention to intrasettlement organization.
There are, however, a few clues from casual observations by settlers. A planta-
tion owner who settled on the west coast of Cape York Peninsula established
a working relationship with the local Aborigines, who formed a number of
camps along the adjacent beach (McLaren 1926). Over time the camps began to
merge into a single large camp, and this process was accompanied by a marked
increase in interpersonal tension, arguments and violence.
550    R o g e r L . C r i b b
The Impact of Fixed Housing

Social psychological studies of postcontact Aboriginal settlements have drawn
attention to the highly stressful nature of social interactions in which the fixed
nature of the European housing provided, together with its inadequacy, inhibit
the spatial expression of the realities of social relations, including constant
violation of explicit kin avoidance rules (Reser 1979). Household space,
designed along functional lines, breaks down as all rooms are used in much
the same fashion. In many cases the family and its visitors camp outside, using
the house itself for storage. After at least four generations of settlement many
houses in Aboriginal contemporary communities are abandoned because they
are incompatible with the social life of the inhabitants. Attempts to conform
to aspects of traditional culture frequently result in neglect and damage such
as chopping up furniture to be used as firewood. A good deal of research has
been done on the provision of housing in Aboriginal communities (Reser 1979).
Some radical designs have been put forward, with or without consultation with
Aboriginal people, and some have actually been implemented, with varying
degrees of success.
In a community in central Australia I came across a group of fiberglass
hexagonal huts, each with three alcoves for nuclear families, a door and storage
space. The central portion of the roof was open to the sky, enabling smoke to
escape from a central fireplace. The design had a certain logic. The singular
fact, however, was that all the complexes were empty and had been abandoned
some time ago. Presumably the former occupants were now part of a large
camp of people located in the dry streambed adjacent to the official settlement.
While I could not identify these people to ask why they had left the complexes,
their reasons were clear enough. Three nuclear families could not comfortably
occupy such a space. Dogs would be a nuisance, snakes could not move through
without being trapped and becoming dangerous, the enveloping walls would
prevent the wind from blowing through and, most important, the occupants
could not see what was happening in neighboring dwellings or notice the
approach of visitors until the latter were on top of them.
Contempor ary Communities

on Cape York Peninsula
Cape York Peninsula, lying in the extreme northeast of Australia, contains
about a dozen settlements of Aboriginal people who have been in contact with
European culture for approximately a century. In the early 20th century CE
the people were either encouraged or forced to live in settlements (missions)
and to reside in permanent dwellings laid out in lines or squares in European
fashion, with little or no attempt to accommodate kinship considerations. Since

the 1970s the missions have been transformed into local authorities run by
community councils, and one of the key functions of the community council is
to provide housing that varies little from the standard European plan of lounge
room, kitchen, bathroom and three bedrooms.
During the mission days households (read nuclear families) were generally
accommodated in small huts assigned by missionaries to families in an arbitrary
manner. An exception was the old settlement at Lockhart River, where housing
was organized in hamlets occupied by members of each separate clan group,
located at a distance from each other. This made for a reasonably harmonious
community. When the community was relocated after the Second World War,
however, housing was arranged in a radial pattern centered on the European
compound (council office, store, post office, European households) and
assigned on an arbitrary basis rather than in accordance with kinship distance
(Chase 1980; Figure 24.2). Tensions within the community increased dramati-
cally until in recent times Lockhart has come to be regarded as one of the most
dysfunctional communities on Cape York Peninsula.
It is remarkable that many of the inhabitants of Lockhart preferred, and
continue to prefer, camping on the nearby beach during the dry season.
Temporary makeshift dwellings of corrugated iron and canvas were arranged
in clusters in a line at regular intervals along the beach, each cluster occupied
Figure 24.2. Plan of the new Lockhart River settlement around 1973. Note the radial plan and
schematized arrangement of dwellings (after Chase 1980:Map 3).
552    R o g e r L . C r i b b
Figure 24.3. Dry season beach camp at Lockhart River (after Chase 1980:Figure 21).
by closely related people (Figure 24.3). Senior persons in adjacent clusters also
tended to be close relatives. Each cluster had a day camp or day shade on the
dunes above the beachfront. It is immediately obvious that such a community
conforms much more closely to the realities of kinship organization than did
the nearby permanent settlement and, while the overall population density of
the beach camps may be greater than that of the settlement, the mapping out
of kinship relationships made for a more harmonious settlement.
During fieldwork near Aurukun in 1985 our research party consisted of
myself, two anthropologists, a local couple, two middle-aged male traditional
owners and one elderly woman (Cribb 1986). Every evening we made camp in
much the same manner: a flat area on a salt pan, dune or sandy riverbank was
selected, and swags and blankets were laid out in a consistent linear pattern. The
couple camped together near the center with the elderly woman at one end and
the single males at the other end. Individuals were separated by a distance of at
least 2 m, and sometimes fires were maintained between adjacent individuals.
This pattern, even though not organized along kinship lines, was indicative of
the kinds of protocols that are of great significance in Aboriginal culture. Larger
numbers of people spending longer periods of time together would tend to
spread out in a low density settlement pattern composed of such modules.
The Camp at Chinaman Creek

Between 2001 and 2003, members of the Lockhart community and others
from adjacent areas staying in the regional city of Cairns, together with other
kin, constructed a similar kind of camp on the bank of Chinaman Creek,
a mangrove creek on the outskirts of Cairns. The area had been used as a
dumping ground and was littered with garbage and rusted car wrecks. Such a
location would not have been chosen by Aboriginal people given any choice in
the matter. There was no sanitation and the nearest water supply was approxi-
mately 100 m distant. The camp was located parallel to the mangrove fringe
and consisted of four clusters of makeshift dwellings constructed of timber,
canvas and plastic tarpaulins. It was bounded on the other side by stands of
long grass. One couple occupied an abandoned van. The area, approximately
10x40 m, was occupied by up to 20 people at any one time. People would come
and go, though there was a core of around 10 people. Related groups would
visit and congregate in large groups in which a good deal of heavy drinking
would occur. Such camps are known as ‘drinking camps’ and the people as ‘long
grass people.’ These are common on the outskirts of northern Australian cities
(Day 1999; Sansom 1980).
The singular feature of the camp at Chinaman Creek was the high level of
interpersonal conflict and frequent violence, certainly fueled by alcohol but
deriving partly from the fact that people were living virtually on top of one
another. As many as 20 adults were sharing the space that would normally be
occupied by a single family. Unlike the beach camps at Lockhart River, where
clusters were separated by a reasonable distance, the Chinaman Creek clusters
were closely juxtaposed with no room for activity areas or day shades. Frequent
arguments erupted through individuals invading another’s personal space or
sleeping area. Domestic disputes were frequent and often violent, leaving the
woman with nowhere to take refuge. Other arguments centered on kinship
issues with one person disputing the ancestry or relationships of another.
Public Space and Conflict

Whereas some anthropologists maintain that ‘traditional’ Aboriginal culture
is dying, I would counter that no culture is ‘traditional’ and that no culture
‘dies.’ What we are seeing is change, or metamorphosis, together with conti-
nuity (Chase 1980). Technologies are replaced and songs, stories, preferential
marriage systems and even whole languages may be lost; but the core of
the culture, the kinship system, remains strong and other aspects of life are
arranged around it. Touch this web at any point and one is quickly entangled.
Establish even a putative relationship with one person (as a ‘brother’) and one
suddenly has a house full of sisters, nephews and aunts.
One very important feature of Aboriginal culture is the tendency for all
disputes to become public, exposed to the full range of people who may have an
interest in them and tending to involve more and more people arraigned along
kinship lines. A certain amount of space is required for this. In contemporary
554    R o g e r L . C r i b b
communities such disputes can surge up and down a street and spill over into
other public areas, causing a great deal of noise but usually resulting in few
injuries. The Royal Commission of Inquiry into Aboriginal Deaths in Custody
(RCIADC) acknowledged the great difference between the ways in which
Aboriginal and Western societies cope with the issue of conflict and unaccept-
able conduct, particularly the intervention of certain kin in disputes, the public
nature of most disputes, and the general acceptance of fighting and swearing
as an integral part of disputes. Above all, ‘avoidance and mobility’ are solu-
tions that are far less available now than they were in earlier times, something
also acknowledged by the RCIADC. Under modern conditions, even within
Aboriginal communities, nearly all forms of overt conflict are likely to develop
into public-order issues.
Particularly for tribal people living in towns and cities, problems arise
through the expression of conflict in an urban setting. Disputes often take the
form of very loud shouting and swearing across a city street as if there were no
mainstream city life going on around them, reflecting the conduct of disputes
in tribal Aboriginal communities. Such disputes are typically aired in public,
where others may be aware of the nature of the dispute and the grievances
involved, so that potential allies can be mobilized or mediators brought in. The
disputes, which may appear chaotic, have an underlying structure and require
a great deal of space for their resolution.
Discussion
While there is no way that the degree of conflict can be measured through
archaeological evidence, there are ways in which archaeological data can have
a bearing on settlement density, which in turn relates to cultural factors such
as tolerance thresholds for crowding. On the basis of the above arguments
we would expect that tentlike house plans, arranged in a low density pattern,
could indicate the sedentarization of nomadic pastoralists accompanied by
certain changes in the degree and kind of social integration. Or it could reflect
marginalization and stress among nomadic people forced into diminishing life
space, similar to the present-day marginalization of nomad settlements on the
outskirts of a village or town. Or the clustering together of floor plans may
indicate a transition to a more permanent settlement.
Wherever the tolerance thresholds in Aboriginal social organization may
occur, they are not permitted full expression in the kinds of community or
urban situations dealt with above. While much of the technology and ideology
of past Aboriginal society has been irrevocably lost, this is still a horizontally
and tribally organized culture. The tendency in such a culture to give expres-
sion to the state of play of social relations has been severely constrained by
fixed settlements and loss of control over living space. In the past one may
imagine populations in a more or less continual state of flux, scheduling
their movements around broad changes in the availability of resources but
also responding to the vicissitudes of social life. The need for sociability was
constantly balanced by avoidance of those situations and those people likely to
give rise to interpersonal conflict. Settled life in fixed communities has severely
strained the tolerance thresholds of a traditionally mobile and flexible society,
and many of the dysfunctional features of contemporary Aboriginal society
(violent crime, domestic violence, communal strife) may derive in no small
measure from this source.
Perhaps clues may be found along the coast of Cape York Peninsula in the
linear distributions of occupation sites in the form of clearings on the coastal
dune systems and large, mounded shell heaps strung out along the salt pans and
silt plains (Cribb 1986). Such sites would have accommodated kinship groups
of various sizes moving from place to place in response to resource depletion
and the state of play of interpersonal relations with alternative settlement sites
being available close by.
References
Bates, D. G.
1973 Nomads and Farmers: A Study of the Yörük of Southeastern Turkey. Museum of
Anthropology Publications 52. Ann Arbor, University of Michigan.
Barth, F.
1961 Nomads of South Persia: The Basseri Tribe of the Khamseh Confederacy. Boston,
Little, Brown.
Chagnon, N.
1976 Yanomamö: The Fierce People. New York, Holt, Rinehart and Winston.
Chase, A. K.
1980 Which Way Now? Tradition, Continuity, and Change in a North Queensland
Aboriginal Community. University of Queensland, Department of Anthropology
and Sociology (PhD disseration).
Cribb, R.
1986 Archaeology in the Aurukun Region: A Report on Work Carried Out in 1985.
Queensland Archaeological Research 3: pp. 133–158.
2003 Getting to Know the Mob. (unpublished manuscript)
Day, B.
1999 Forgive Us Our Trespasses: Finding Space for Aboriginal Fringe Dwellers in
Darwin. Canberra Anthropology 22: 2, pp. 62–69.
556    R o g e r L . C r i b b
Forge, A.
1972 Normative Factors in the Settlement Size of Neolithic Cultivators. In P. J.
Ucko, R. Tringham and G. Dimbleby (eds.), Man, Settlement, and Urbanism.
London, Duckworth: pp. 363–376.
McLaren, J.
1926 My Crowded Solitude. New York, McBride.
Nissen, H. J.
1968 Survey of an Abandoned Modern Village in Southern Iraq. Sumer 24: pp.
107–114.
O’Connell, J. F.
1979 Room to Move: Contemporary Alyawara Settlement Patterns and Their
Implications for Aboriginal Housing Policy. In M. Heppell (ed.), A Black
Reality: Aboriginal Camps and Housing in Remote Australia. Canberra, Australian
Institute of Aboriginal Studies: pp. 97–120.
Reser, J. P.
1979 A Matter of Control: Aboriginal Housing Circumstances in Remote Com
munities and Settlements. In M. Heppell (ed.), A Black Reality: Aboriginal
Camps and Housing in Remote Australia. Canberra, Australian Institute of
Aboriginal Studies: pp. 65–96.
Sansom, B.
1980 The Camp at Wallaby Cross: Aboriginal Fringe Dwellers in Darwin. Canberra,
Australian Institute of Aboriginal Studies.
Sweet, L.
1974 Tell Toqaan: A Syrian Village. Museum of Anthropology Publications 54. Ann Arbor,
Swidler. W. W.
1972 Some Demographic Factors Regulating the Formation of Flocks and Camps
Among the Brahui of Baluchistan. In W. Irons and N. Dyson-Hudson (eds.),
Perspectives on Nomadism. Leiden, E. J. Brill: pp. 69–75.
Tapper, R. L.
1979 The Organization of Nomadic Communities in Pastoral Societies of the
Middle East. In L’Equipe écologie et anthropologie des sociétés pastorales
(ed.), Pastoral Production and Society/Production pastorale et société. Proceedings
of the International Meeting on Pastoralism. Paris, December 1976. Cambridge,
Watson, P. J.
1966 Clues to Iranian Prehistory in Modern Village Life. Expedition 8: pp. 13–23.
1979 Archaeological Ethnography in Western Iran. Viking Fund Publications in
Anthropology number 57. Tuscon, University of Arizona Press.
Ch a p t e r 2 5
NOMAD
An Agent-Based Model (ABM) of
Pastoralist-Agriculturalist
Interaction
L aw r e n c e A . K u z n a r a n d R o b e rt S e d l m e y e r 1
And Hamor and Shechem his son came unto the gate of their city, and communed
with the men of their city, saying, these men [the sons of Jacob] are peaceable
with us; therefore let them dwell in the land, and trade therein . . . And it came
to pass that on the third day . . . two of the sons of Jacob, Simeon and Levi . . .
took each man his sword, and came upon the city boldly, and slew all the males.
They took their sheep, and their oxen, and their asses, and that which was in
the city, and that which was in the field, and all their wealth, and all their little
ones, and their wives took they captive, and spoiled even all that was in the house
(Genesis 34:21–29).
Cathay’s great general, Prince Fu-hsing, advised his king . . . And if these cities
are forced to fight the Mongol army they’ll most likely surrender to them. I say
we should offer tribute to the Khan of the Mongol for now . . . Let’s give one of
your daughters to their Khan. Let’s give the men of their army heavy burdens
of gold, silver, satins, and other goods . . . And our soldiers carried off as much
satins and goods as their beasts could hold and went on their way, securing their
bundles with ropes of silk (Kahn 1998:147–148).
W hether the dramatic annihilation of the Shechemites by Jacob’s

nomadic sons or the smug satisfaction expressed by Mongol chroniclers;
propagandists, historians and anthropologists alike often portray contentious
pastoral-to-agricultural interactions (Goldschmidt 1979). The scenario reads
as follows: Pastoral nomads are free, violent and rapacious; agriculturalists
are wealthy and vulnerable, and inevitably, the martial nomads conquer the
hapless, effete city folk. The earliest written records concerning pastoralists,
1.
The authors gratefully acknowledge the indispensable computer programming
efforts of Allyson Kreft (who, at the time of this study, was Research Assistant at the
Department of Computer Sciences, Indiana University–Purdue University at Fort
Wayne).
557
558    L aw r e n c e K u z n a r and Robert Sedlmeyer
dating from 18th century BCE Mesopotamia, indicate that relations between
pastoralists and sedentary folk have been contentious since the birth of the state
(Simon 1981). But ancient writers also note that pastoral nomads and urban-
ized agriculturalists are symbiotically linked through trade, each providing
what the other needs or desires. Anthropologists reinforce this linkage by
documenting mutualism, pastoral domination or domination of pastoralists by
agricultural states (Asad 1973; Barth 1973; Gellner 1973; Dyson-Hudson and
Dyson-Hudson 1980; Simon 1981; Barfield 1989; McCorriston 1997). Sources
ancient and modern suggest counteracting forces between the need to trade and
the constant threat of domination by one’s trading partner in these societies.
The central problem for understanding cultural processes like these is that
anthropologists usually only observe the results of process, not the processes
themselves. Since many culturally interesting phenomena, such as cycles of
conquest, growth of family wealth and settling into village life (sedentariza-
tion), take place over decades or longer, it is not possible for anthropologists
to witness firsthand the processes that generate these phenomena. Researchers
must infer the mechanism that caused an observed set of affairs, much like an
archaeologist must infer the prehistoric processes that generate archaeological
remains (Cronk 1998). Relatively recent advances in computer science provide
anthropologists with a means of addressing this methodological difficulty.
Agent-based models (ABMs) use advances in object-oriented program-
ming and increased computing power to model the independent actions of
numerous individuals, allowing emergent phenomena to occur. ABM simula-
tions permit researchers to program theoretical processes and to test whether
or not these processes lead to observed phenomena in the context of simulated
social systems (Epstein and Axtell 1996). Anthropologists use ABM simula-
tions to explain the rise and fall of civilizations (Dean et al. 1999; Axtell et al.
2002; Lazar and Reynolds 2002), to model the emergence of ethnic groups
(McElreath et al. 2003), and to simulate complex ecological systems (Kuznar
2006). ABM simulations have the added benefit of forcing researchers to
specify all variables necessary for the functioning of a theory, enabling them to
appreciate where their empirical understandings of the world are inadequate
and suggesting future field research. In this chapter we demonstrate how ABM
simulations can be used to test theories of interest to anthropologists studying
pastoral nomads and their agricultural neighbors.
Pastor al Nomad–Sedentary
Agricultur alist Dichotomy
Many researchers have challenged the pastoral nomad–sedentary agricultur-
alist dichotomy, noting that there are many exceptions (Bates 1971; Salzman
NO M A D     559
1972; Spooner 1972). Even so, logistical incompatibilities exist often enough
between pastoral and agricultural production, the former leading to a focus
on mobile animal herding and the latter to sedentary agriculture, to allow
the distinctions to persist (Bernbeck, this volume; Barth 1973; Gellner 1973;
Goldschmidt 1979). Nomadic pastoralists specialize in animal husbandry that
requires mobility (Dyson-Hudson and Dyson-Hudson 1980:17). Furthermore,
pastoral production is particularly distinguished by its volatility (Barfield
1993:13). Even though pure pastoralists independent of agricultural society
do not exist, Barfield (1993:4) recognizes that enough people in many societies
do fit the description of nomadic pastoralists to be singled out for categorical
study. Historians have detailed clear and profound interactions between seden-
tary agricultural states and pastoral nomads, shaping the history of East Africa,
the Middle East and China (Lattimore 1951; Ibn Khaldun 1967; Simon 1981;
Barfield 1993). The persistence of the pastoral nomad–sedentary agriculturalist
distinction warrants current investigation of its use. Instantiating these two
general modes of production in computer models allows both the testing of
the validity of the concepts (do they work, or do these distinctions fall apart?)
and of theories based on these distinctions (do pastoralists and agriculturalists
interact in the ways and for the reasons hypothesized?).
Our aim in this chapter is to demonstrate how computer modeling can aid in
establishing the usefulness of these concepts and their associated theories. We
begin this discussion by reviewing theories of the cyclic history of pastoralist-
agriculturalist interaction. Then we turn to anthropological theories that
specify the mechanisms that may produce these historical cycles. We outline
a computer model designed to capture these mechanisms. The model is then
used to see whether the theorized mechanisms produce the historical cycles
and ethnographic patterns researchers propose. The chapter concludes with a
discussion of results and future directions for research.
Cycles of Conquest and Pastor al Nomads

The first theorist to examine the interaction of pastoral nomads and agricul-
turalists closely was the medieval scholar Ibn Khaldun. In the Muqaddimah,
published around 1381 CE, he proposed a systematic cycle of pastoral conquest,
sedentarization, social decay and subsequent conquest that largely held until
the modern era (Barfield 1989). He argued that because of the differences in
the material conditions faced by desert nomads and urbanites, they develop
different subsistence systems, psychologies and ideologies. Ibn Khaldun notes
a fundamental interdependence between nomads and city dwellers; “while [the
Bedouins] need the cities for their necessities of life, the urban population
needs [the Bedouins] for conveniences and luxuries” (Ibn Khaldun 1967:122).
He also argues that urban concerns, such as keeping up with inflationary prices
and engagement in commerce, distract agriculturalists from maintenance of
the martial skills necessary to defend their lifestyles (Ibn Khaldun 1967:276,
278, 285–286). The ultimate result is conquest by nomadic pastoralists who,
attracted to the ease of urban life, sedentarize. These recently sedentarized folk
are eventually conquered by another group of pastoral nomads, culminating in
historical cycles of conquest and settlement by pastoral dynasties.
Sedentary culture was always transferred from the preceding dynasty to the later
one. The sedentary culture of the Persians was transferred to the Arab Umayyads
and 'Abbâsids. The sedentary culture of the Umayyads in Spain was transferred
to the Almohad and Zanâtah kings of the contemporary Maghrib. That of the
'Abbâsids was transferred, successively, to the Daylam, to the Saljuq Turks, to the
Turks in Egypt, and to the Tartars in the two ‘Iraqs’ (Ibn Khaldun 1967:140).
Ibn Khaldun (1967:107) concludes with the same stereotypes echoed in antiq-
uity: “‘[D]esert life no doubt is the source of bravery, [and] savage groups are
braver than others. They are, therefore, better able to achieve superiority and to
take away the things that are in the hands of other nations . . . Whenever people
settle in fertile plains and amass luxuries and become accustomed to a life of
abundance and refinement, their bravery decreases to the degree that their
wildness and desert habits decrease . . . This is exemplified by dumb animals,
such as gazelles, wild buffaloes, and donkeys, that are domesticated.”
Lattimore (1951) described the rise of pastoral nomadism as an economic
specialty on the central Asian steppes, its characteristics and the cyclical rise
and fall of Chinese agrarian kingdoms in terms reminiscent of Ibn Khaldun.
He recognized that a continuum exists between pure pastoral and pure agrarian
societies but that, because of environmental differences in agrarian potential,
one or the other subsistence system would dominate a region, setting up a
dynamic interaction between nomadic pastoral societies of the steppe and
Chinese civilizations of the irrigated valleys that would permeate East Asian
history (Lattimore 1951; see also Barfield 1989; Khazanov 1994:60). This
interaction alternated among trade, extortion, raiding, conquest and vassalage
(Lattimore 1951:332, xlvii, 333, 77 and 547 respectively). Lattimore notes that
pastoral and agrarian societies are linked through interdependent trade rela-
tions, although he argues, contra Ibn Khaldun, that nomads needed Chinese
luxury goods more than Chinese needed pastoral products (Lattimore 1951:69).
Pastoral nomads enjoyed a military advantage by virtue of their mastery of the
horse and their constant practice of martial skills while protecting their flocks
(Lattimore 1951:63–66, 333). The successful domination of Chinese society
by nomads, however, undermined the very abilities that enabled the nomads’
success: “the protected oasis immobilized its protectors and, if it became too
NO M A D     561
wealthy and its protectors insufficiently mobile in the warfare of the steppe, it
was always overwhelmed by raiders” (Lattimore 1951:77, 333). And so, echoing
Ibn Khaldun, Lattimore demonstrates that a similar cycle of nomadic conquest,
transformation into agrarian society, and subsequent conquest characterized
Chinese history.
More recent scholarship largely reinforces these historic trends in pastoral-
agrarian interaction. Khazanov (1994) provides an encyclopedic comparison of
pastoral nomad histories from the Middle East, North Africa, East Africa and
the Asian steppe. He describes the modes of nomadic adaptation to agrarian
societies as sedentarization, trade, submission to states, and the subjuga-
tion of agrarian states. His thorough review of many historical examples
from throughout the Old World does justice to the diversity of pastoral
lifestyles, although certain basic patterns clearly emerge. Until the modern
era, pastoral nomads generally enjoyed a military advantage over sedentary
societies: “nomadic raids have been recorded in all regions where nomadism
is widespread” (Khazanov 1994:222, 263). The pattern persists even when
pastoralists have historically been encapsulated and forced to submit to states
(Burnham 1979; Irons 1979; Krader 1979; Lattimore 1979). One of the more
forceful denials of this pattern is Asad’s (1973) analysis of Bedouin history in
the Middle East, although in each of his examples state domination of nomads
is either incomplete or the result of modern military technology. While these
historians describe general patterns, and even provide some mechanisms, they
cannot provide detailed studies of the actual mechanisms by which individual
nomadic pastoralists would decide to settle. Ethnographic studies of contem-
porary pastoral societies are necessary to identify the actual mechanisms of
sedentarization and its effects.
Ethnogr aphy and Sedentarization

Ethnographies in the 20th century CE have both reinforced the basic trends
noticed by Ibn Khaldun and Owen Lattimore and exposed the diversity of the
ways that pastoral-agricultural interaction takes place and the local factors
that influence these contingent variations (Goldschmidt 1979; Dyson-Hudson
and Dyson-Hudson 1980; Barfield 1993). Despite this variation, a number of
basic generalizations appear to hold concerning pastoralists who rely on large
animals in arid lands. Herds are a volatile source of wealth, fluctuating wildly
from wealthy to poor levels based on environmental (weather, pasture) and
social (raiding, demography) factors (Goldschmidt 1979:20; Barfield 1993).
The key to long-term success is the building of large herds, fostering entre-
preneurial behaviors and attitudes among pastoralists (Goldschmidt 1979:23).
Citing Edgerton’s research on culture and personality, Goldschmidt argues that
these situational factors reinforce a set of personality traits, including aggres-

sion, use of violence and capacity to suffer physical hardship (Goldschmidt
1979:24). These generalities lead to the following tendencies that have influ-
enced the interactions of pastoral nomads and agriculturalists for millennia.
First, the pastoral combination of mobility and familiarity with weapons has
preserved pastoral nomads’ military pre-adaptation (Irons 1965, 1972; Salzman
1979). Second, pastoral nomadism as an economic specialization exists only in
conjunction with sedentary agricultural societies, so nomads must acquire the
primarily grain crops they need for food as well as other goods necessary for
their existence (Dyson-Hudson and Dyson-Hudson 1980). Third, this interac-
tion varies from raiding to extortion to trading, based on the relative military
and political strengths of nomads and villagers (Irons 1972; Asad 1973; Berntsen
1976; Swift 1979). Pastoral nomads and sedentary agricultural villagers, however,
are not forever locked in either their ways of life or battles. Instead, there has
always been a continuum between pure nomadic pastoralism (which has always
been a rare occurrence) and fully agricultural life, and there have always been
avenues for social mobility and trade between the two extremes.
Barth (1961) proposed two pathways that pastoralists might take in trans-
forming into sedentary villagers either through economic success or failure:
“radical departure from the middle range of wealth . . . cannot be accom-
modated within the [pastoral] organization” (Barth 1961:110). In the case of
economically successful herders, once their herd becomes too large for them
to manage, they must hire poor shepherds to watch them (Roe, this volume).
Despite contractual agreements, however, the productivity of these herds is
decreased because of a hired hand’s inattentiveness toward a flock that is not
his own (Barth 1961:13, 111). Ethnographers have noted similar problems in
pastoral societies in both the Middle East and East Africa (Barth 1961; Dahl
1979; Hannoyer and Thieck 1982). Consequently, wealthy herders realize
that investing their surplus in land and settling is the more profitable strategy,
and they become agrarian villagers, analogous to the transformation nomadic
elites often made after conquering agrarian states. The other pathway to
sedentism is poverty. Once a herder is so impoverished that his herd can no
longer sustain himself and his family, he has no choice but to drop out of the
pastoral game and attach himself to an agrarian village, finding work as a tenant
farmer or some other form of laborer (Barth 1961). Barth (1961) notes that
statistically, this is the more usual route to sedentism. Glatzer (1982), working
in Afghanistan, notes that sedentarization can also be a two-way street, with
particularly successful agriculturalists investing surplus wealth in livestock to
the point where they spin off pastoral families.
Barth’s general scheme, although subject to contingencies such as land
availability and the influence of state-level governments (Bates 1971; Irons
NO M A D     563
1972), has nonetheless been largely corroborated in the past 40 years. Beck
(1991, 1998) has recorded the sedentarization of most Qashqa'i pastoralists
in Iran, which has occurred as the result of a combination of legal changes in
land tenure that made nomadism difficult and the introduction of vehicular
transport that made moving with animals obsolete. In Afghanistan most seden-
tarization has occurred among pastoralists impoverished through poor herd
management, drought and warfare (Glatzer 1982; Barfield 2004). War, drought
and colonial domination have led to mass sedentarization of many pastoralists
worldwide (Asad 1973; Ben-David 1990; Fratkin and McCabe 1999; Kuznar
1999; Adano and Witsenburg 2003). While economic success is less often the
route pastoralists take to sedentism, Irons (1972) noted that this occurred in
the early 20th century CE among Yomut Turkmen in Iran, and Swift (1979)
documents such a shift in Somalia in the late 20th century CE.
This dual route to sedentarization is driven by the characteristic volatility of
pastoral production. Pastoralism is notably risky, leading to extremes of failure
and success in pastoral societies through the world (Barth 1961; Irons 1972;
Schneider 1979; Browman 1987; Cribb 1991; Kuznar 2001). Very successful
herders will be faced with unmanageable herds, and these herders are likely to
reinvest surplus stock in sedentary agrarian and urban enterprises. For the desti-
tute, pastoralism is no longer viable, and they must seek employment and support
from the larger settled agrarian and urban sectors. Because of pastoral volatility,
the interaction between agricultural and pastoral sectors is likewise uncertain
and dynamic. In summary, researchers have identified several relatively simple
principles that influence pastoral-agriculturalist interactions. Members of both
parties must acquire goods that the other has. Pastoralists have historically been
militarily superior to sedentary villagers. Pastoralism is riskier than agriculture
and the very wealthy and poor tend to sedentarize, becoming villagers. Variations
on these themes exist, owing to the contingencies of ecological particularities
and the presence or absence of a strong state. These few simple rules have the
potential to interact in very complex ways. Our aim is to instantiate these few
rules in a general but relatively realistic agricultural-pastoral scenario in order
to see if their interaction will produce the rich and varied historical trends
researchers have noticed. Given the historical dimension of these processes,
computer simulations are necessary to model how the behavior of individual
pastoralists and agriculturalists may lead to historical patterns.
Simulating Pastor alist-

Agricultur alist Inter actions
ABM simulations are computer programs that allow individual agents to
interact, in a virtual setting, with other agents and their environment according
to programmed rules (Epstein and Axtell 1996). Researchers can test theo-
ries of environmental influence or social interaction by programming these
conditions and rules into the model, and they can then verify if the model
produces lifelike results (Kuznar 2006). Compared with most mathematical or
statistical analyses, often based on aggregated or averaged data, these models
provide for more realistic results because the agents and their actions remain
discrete. Therefore, the possibility for nonlinear interactions exists, shifting
the model into varied and unexpected states and mirroring the complexity of
real-life systems (Axelrod 1997; Terna 1998). In our case we will employ ABMs
to capture the dynamics of pastoral nomad–sedentary agriculturalist trade
and conflict. Our ABM uses the Java-based ABM programming framework
called ‘Swarm’ (Terna 1998), which facilitates the programming of indepen-
dent agents, the assignment of behavioral rules, and the graphical display of
agent behavior.
Our primary aim is to demonstrate the potential usefulness of ABMs for
theory testing and the direction of field research. Therefore we have created
a simple scenario of pastoral-agriculturalist interaction that mimics the sorts
of environments and behaviors anthropologists have recorded in the arid
environments of the Middle East and East Africa. To instantiate a reasonable
simulation of pastoral-agricultural interactions, we need to create an environ-
ment, place resources in that environment, place pastoral and agricultural
agents in that environment, and provide the agents with decision-making and
behavior rules. For the sake of simplicity we have left out important factors,
such as the effect of a strong state government with the ability to garrison
troops or the ability of agriculturalists to invest in pastoral enterprises. We also
had to simplify some social interactions (such as inheritance rules and trade
scenarios) in order to focus on some essential aspects of pastoralist-agricul-
turalist interaction. In the future, simplified aspects of the simulation can be
sophisticated to more accurately reflect social realities. Also, more contingent
factors can be programmed into the simulation, testing the strength of their
effects in generating the diversity of pastoral lifestyles. Despite these simpli-
fications, our simulation still provides realistic interactions between relatively
militarized pastoral nomads and relatively vulnerable sedentary agriculturalists,
such as has recently been starkly demonstrated in the Darfur region of Sudan
(Kuznar and Sedlmeyer 2005). The various components of the model that
must be specified include time, environment, productive resources, the human
element and behavioral rules.
Since agrarian and pastoral rhythms occur ubiquitously on a seasonal basis,
each iteration of the simulation will represent one season: spring, summer, fall
or winter. The environmental setting is reminiscent of the typical arid envi-
ronment of the Middle East or East Africa in which agriculturalists inhabit a
NO M A D     565
watered valley while pastoralists migrate among highland and lowland pastures
scattered throughout the rest of the landscape. Some pastoralists migrate over
distances of 10–20 km (Shepardson and Hammond 1970; Kuznar 1995; Sidky
1995), others over hundreds of kilometers (Barth 1961; Cribb 1991). Because
migrations are usually between these extremes, as among Yomut Turkmen of
Iran (Irons 1972), we model an environment that is 50x50 km. A river cuts
across this environment, with agricultural land occurring along its course.
Pastures are scattered about the landscape, with summer pastures occurring in
the eastern portion of the environment (mountain pastures) and winter pastures
scattered nearer the river course (lowland pastures). Glatzer’s work (Glatzer
and Casimir 1983) in western Afghanistan provides a fairly representative
distribution of agriculture and summer and winter pastures and is the basis
for the geography of our model. Each land type needs a level of productivity,
a hazard rate and a hazard effect in order to model environmental fluctuations
that impact both agriculturalists and pastoralists.
Following experimental work, agricultural productivity for traditional Middle
Eastern fields growing emmer wheat would be 500 kg/ha primitive wheat,
providing 3300 Kcal/kg (Casimir 1988; Hillman 1990:169). We use a drought
hazard rate of 50%, given 40–60% recorded for Luristan, Iran (Hole 1978:141).
The effect of a hazard would be a 50% loss felt during autumn, reflecting the
decreased yield experienced during harvest time after a drought, consistent with
figures provided by Dean et al. (1999:187) for decreased yields in arid lands.
Pasture productivity varies considerably from extremely low, in Afghani and
Iranian deserts, to relatively high in the lush alpine pastures of South America
(Kuznar 1991a, 1995). A reasonable median would be 500 kg dry matter/ha/yr
as recorded for desert scrub in the American Southwest (Cook and Sims 1975).
Pastures beginning to grow in the spring typically do not achieve full produc-
tivity at this time, so spring productivity will be reduced by 20%. Summer
productivity will be 100%, while fall and winter productivity will amount to the
vegetation left over after grazing, following data from the Sahel in Mali (Diarra
and Breman 1975). The annual hazard rate (percent drought years) for pastures
in arid environments typically varies from lows around 10% (Dahl and Hjort
1976:116) to highs of 40% (Cook and Sims 1975) and ranges between 15–30%
in the Sahel in Mali (Diarra and Breman 1975). We use a median figure of 25%.
Since the key periods of stress are spring and summer, we divide this hazard
evenly between these two seasons. Hazard effects are often high, ranging from
50% animal loss in Iran (Barth 1961:7), to 80% dry matter loss for the U.S.
Western Range prairie (Cook and Sims 1975), and a moderate 51–57% dry
matter loss (severe drought compared to moderate drought) in Mali (Diarra
and Breman 1975). Given the more severe effect of drought on pasture, we use
65% dry matter loss for the pasture hazard effect.
We have already specified the agricultural resource levels in terms of

traditional wheat production in the Middle East. Herd animals present more
complex demographics, including dry matter assimilation rates, disease and
mortality rates, body mass, butchering percentage, meat and milk production,
and herd reproductive rate. We elaborate on each in turn. As a simplifica-
tion, we ignore fiber production because of the varied return rates for fiber
throughout the world and the volatility of the fiber market. It would be more
reasonable to include fiber production for a specific application. The amount
of food necessary to produce cattle, sheep and goats is about 3.5% of their
weight in dry matter/day (Devendra 1978; Iannelli 1984; Gatenby 1986).
Herd mortality rates (mostly from disease and predation) amount to 8% in
North Africa (Dahl and Hjort 1976:265) and the Peruvian highlands (Kuznar
1995). For herds in the Middle East, Hole (1978:144) reports rates of disease
of 30–50% of a herd every three to seven years. As we are modeling a herd
based on small stock, the standard live bodyweight per animal is roughly 40
kg, and the standard meat takeoff is 20 kg/animal (Devendra 1978). Dahl
and Hjort (1976:209) estimate that the takeoff rate for a flock is 8% per year,
providing 60,525 Kcal and 1500 g protein/head. Sheep, goats and other live-
stock also produce milk, however, and Dahl and Hjort (1976:212) estimate an
expected milk production of 45 kg per ewe/year for human consumption. In
a herd of 40% ewes this would produce 1100 Kcal/kg ewe and 65 g protein/
kg ewe (Dahl and Hjort 1976:212, 216). This equals, per head of herd, 19,800
Kcal/yr and 1690 g protein/yr. The herd reproductive rate, only assessed
during the spring, should average 18% (Dahl and Hjort 1976:231). Cribb
(1991:28) notes that this results in a comparable 0.7 to 0.9 births per ewe/yr
in the Middle East.
The Human Dimension

We have instantiated an environment with resources and have characterized the
seasonal and random variations in those resources. Now we must create human
agents who will populate and interact in this environment. Some agent attri-
butes will be common to all agents, such as basic nutritional demands. Others
will be specific to pastoral or agricultural ways of life, such as fertility rates,
mortality rates and life spans. Also, economic factors will be specific to ways of
life, including labor demands, access to resources and raiding mortality rates.
We adopt the basic nutritional demand figures used by Kohler et al. (1999:160)
in their ABM of the rise and fall of Anasazi civilization. Nutritional demands
are broken down to two basic requirements: Kcal and protein. Nutritional
protein demands are 63 g/day for males and 50 g/day for females and chil-
dren. Nutritional Kcal demands are 1872 Kcal/day for males, 1560 Kcal/day
NO M A D     567
for females and 1000 Kcal/day for children. Other biological and economic
attributes are different for agricultural and pastoral populations.
Sedentary agriculturalists tend to have higher fertility than mobile pastoral-
ists (Roth 1986). We use a total completed fertility rate of 6 children/female,
as commonly observed for sedentary agriculturalists (Thomas 1973:147).
Comparison of morbidity and mortality between agriculturalist and pastoral
populations in Mali indicates that the agriculturalist disease rate should be
about 3% while the childhood (age 1–5) mortality rate should be around 25%
(Fratkin et al. 1999:157). Recorded total completed fertility rates in pastoral
societies in Africa range from 3.2 among the Himba to 4.7 for the Pokot
(Fratkin and McCabe 1999:8). The fertility rate of the Basseri, in the Middle
East, was probably around 4 (Barth 1961:12). We use a fertility rate of 4 for
pastoral groups. Pastoral disease rates are probably closer to 2% (estimated
from the figure in Fratkin et al. 1999:157). A recurrent pastoral childhood
mortality rate would be around 40%, as recorded in surveys of pastoral groups
in the Sahel (Fratkin et al. 1999:154; Gray et al. 2003:S8; Roth 1986:71).
Each agriculturalist household operates as a corporate entity that farms
land to produce grain for consumption, tax payments and trade. A reason-
able labor rate for arid land farming would be 1 ha per adult and half that per
child, as suggested by figures presented by Hillman (1990:169) and consistent
with the amount of agricultural land typically cultivated in south central
Asian agriculturalist communities (Lewis 1965:100; Sidky 1995: 23). Labor is
a constant constraint on pastoral production (Swidler 1972; Dyson-Hudson
and Dyson-Hudson 1980; Kuznar 1991b). Basseri herding labor rates were
about 300 animals/adult and 150 animals/child (Barth 1961:6), and this figure
is reinforced cross-culturally (Cribb 1991:28).
Raiding mortality has been directly observed among East African pastoral
communities. Gray et al. (2003) provide detailed demographic data on raiding
mortality among the Karimojong during different periods of raiding intensity
and weapon use. In the 1950s an escalation of cattle raiding among Karimojong
led to violence, accounting for 22% of all male deaths (Gray et al. 2003:S15).
It was 12% in the two previous decades. Since the 1970s, and the introduction
of firearms, it has gone up to 33%. Unfortunately, they provide no data on
the number of raids a man does during his life. However, a man’s raiding life
span is probably not more than 20 years and is probably intense while trying to
acquire bride wealth (Gray et al. 2003:S18). Assuming one raid per year (prob-
ably more before and less after marriage), we divide the mortality rates by 20
to get the probability of death on a raid. The annual probability of dying on a
raid was therefore 0.0060 during the peaceful period while still using spears.
This would increase to 0.0110 during periods of intense raiding with spears.
The introduction of automatic rifles increases the mortality rate to 0.0165. A
reasonable, prefirearm long-term raiding death rate would be somewhere in

between intense and peaceful periods of raiding with spears: 0.0085. These
raiding data were derived from armed herders attacking other armed herders;
the rate for a herder raiding an unarmed agriculturalist would obviously be
lower. We will assume that the mortality rate for pastoralists would be twice
as low when attacking an agriculturalist, and the agriculturalist’s mortality rate
from being attacked would be twice as high. Therefore, the per capita death
rates per raid are adjusted to 0.00425 for pastoral nomads and 0.01700 for
agriculturalists.
The Rules of the Game

After instantiating an environment, resources and agents with essential attri-
butes, these various entities must interact and we must provide them with rules
that will govern their interactions. The rules can be divided into demographic
rules, movement rules and social rules (inheritance, trade and raiding). For
simplicity’s sake we use straightforward demographic rules. We already noted
the childhood mortality rates. Adulthood for agents begins at age 20. All agents
die at age 65, unless already dead. If an adult agent dies, his or her household
disappears as it is no longer viable. Women have a 20-year reproductive life
span, between age 20 and 40, and their probability of conceiving in any year
is their fertility rate divided by 20. When an agent reaches the age of 20, that
agent marries the next nearest agent who is of marriageable age. They then
form a new pastoral family herding unit.
Since pastoralists are the only agents who are mobile, the movement rules
will concern only them. Each pastoral agent has an adjustable vision of 10 grid
squares (roughly 5 km), allowing the agent to know what is near his or her loca-
tion. Once 80% of the pasture production is consumed, the agent moves in the
direction (determined by vision) of increased yield. If no direction is obvious
then the agent moves randomly. If an agent has a metabolic need (Kcal store is
at two seasons minimum yearly resource level for the family), the agent moves
toward an agriculturalist for trade. Agriculturalists don’t move, but they need
to acquire sufficient fields for their sustenance. When an agricultural family’s
needs exceed its field production, the family acquires new fields. It begins by
adding field plots contiguous to current fields. If that is not possible, it acquires
the nearest new fields. A new family acquires the nearest open agricultural land
to the adult male head of household. If all land is exhausted, then these new
families exit the program: we assume they migrate away.
Herds are collectively owned and pastured by families. Mirroring the
widespread tendency of pastoralists to practice patrilineal descent, we will
assume patrilineal inheritance (Barfield 1993). When a child matures, at age
NO M A D     569
20, inheritance occurs. The simulation calculates the parents’ family’s herd
size needs, plus adult child’s herd size needs. If the parent has a herd adequate
for the family needs plus that child’s, then the program calculates and awards
the basal herd needs for all and bequeaths the remaining portion to the child
proportionately (so if there are four children, each will receive 25% of the
excess herd). If the parent’s herd is not large enough to bequeath, then the
child remains stuck at home. It is customary for women in the Middle East
and Central Asia to receive a dowry, which is essentially her inheritance, given
to her, or her husband, at marriage, and is a form of portable wealth (Martin
and Voorhies 1975; Dyson-Hudson and Dyson-Hudson 1980; Barfield 1993).
While a woman may not typically receive animals, her dowry is nonetheless
wealth, and her parents need to be wealthy enough in animals to afford it. Our
model simply gives a woman her portion of the herd at marriage, which she
then combines with her husband’s herd as though they combined their wealth
assets. Children tend to stay at home in agricultural societies and often care
for their elderly parents (Handwerker 1983). When a parent dies, wealth and
field possessions are divided equally among children.
In order to determine whether two agents would trade, and at what rates of
exchange, we use the trade rule developed for the Sugarscape ABM (Epstein
and Axtell 1996). Trade occurs when mutually beneficial to both parties. If not,
either the pastoralist attempts a raid or, if unsuccessful in the attempt or execu-
tion, goes without. This models the risks that agents incur in holding out for a
better price, with the possibility of retribution. Agents will choose the nearest
neighboring potential partner, and each will estimate his own marginal rate of
substitution for the goods potentially traded. In our model the two goods at
issue are Kcal and protein. An agent’s marginal rate of substitution (MRS) of
Kcal for protein is simply calculated as Kcal stored/Kcal needed for a family
per year divided by protein stored/protein needed for the year (Epstein and
Axtell 1996). In our case, if a pastoralist’s marginal rate of substitution for Kcal
is greater than an agriculturalist’s marginal rate of substitution for Kcal, then
trade can occur. MRSs establish a range of feasible prices (quantities that can be
exchanged). If we have two agents, P (pastoral nomad) and A (agriculturalist),
the range of prices at which they will trade is [MRSp, MRSa]. While these
may be theoretically agreeable prices, they will not normally be considered
‘fair’ because an agent wealthy in one good may ‘gouge’ another who is needy
for that good. To limit the effect of such price gouging, we follow Epstein
and Axtell (1996) and use the geometric mean of the endpoints of the feasible
price range:
p ( MRS p , MRS a ) = MRS p MRS a

in which MRSp is the pastoralist’s marginal rate of substitution and MRSa is the
agriculturalist’s marginal rate of substitution. The amount exchanged is equal
to p units of Kcal if p>1, and 1/p units of Kcal if p<1. Altering MRSs by taking
this geometric mean was important in the model because during preliminary
model verification, price gouging emerged that would probably not be tolerated
in any culture, thus reinforcing Epstein and Axtell’s general intuition.
In Epstein and Axtell’s simulation, if mutually beneficial terms of trade are
not possible, agents seek other trading partners. In our simulation we more
realistically model the harsh constraints of satisfying need in the short-term:
sometimes trading partners cannot be found. Consequently, people either
suffer resource depletion and starve, or they use violence to obtain what they
need. Berntsen (1976) provides a description of just such raiding where Maasai
choose to raid agriculturalists too poor to trade. If a nomad has a nutritional
need yet cannot trade, then the nomad may raid an agricultural household.
In most tribal societies raids are considered unsuccessful if a member of the
party is killed (Chagnon 1992). If any raiders die, then the raid is unsuccessful
and nothing is gained. If the raid is successful, then the raiders divide the
spoils equally.
Following Barth’s (1973) hypothesis, there are two routes to sedentarization
for herders: success and failure. In the case of success a herder’s flock becomes
so large that he can no longer manage it profitably, even with hired help. In that
case the herder invests surplus wealth in fields and becomes an agriculturalist.
In a related way pastoral raiders gain so much sedentary wealth, such as fields,
that they likewise must settle down to enjoy the fruits of their success. To model
this route to sedentarization, we define a maximum herd size that a family can
manage. If herd size becomes twice this maximum herd size, pastoralists often
have mitigating strategies for very large herds; then the pastoral family becomes
agricultural and settles on unused agricultural land, purchased with their
animals. We estimate the purchase price of land by taking the average price
of protein in the previous iteration and multiplying this by the bodyweight of
the herder’s livestock. This wealth is used to buy the amount of land required
to grow the amount of grain that could be purchased with that herder’s liquid
capital. Failure also results in sedentarization. In this case pastoralists whose
wealth falls below the family’s minimum needs can become agriculturalists,
if they manage to migrate to unused land before their resources fail. These
pastoralists will also liquidate their livestock and buy land, as described above.
In this manner we expect to more realistically model the impoverished nature
of failed pastoral turned agriculturalist life.
NO M A D     571
Expectations
We have produced a model based on available empirical data on arid land
ecology as well as pastoral and agriculturalist demography, economics and
lifestyles. We have provided a set of rules that mirror historically important
theories of pastoral decision making and pastoralist-agriculturalist interac-
tions. If these theories can actually explain the phenomena they purport to
explain, then our simulation should produce those patterns. If they do not,
then either the theories are incorrect or important variables in the model
are wrongly specified, indicating future research directions to correct these
deficiencies. The primary research question we explore in this chapter is: “Do
the interactions among pastoral nomads and sedentary agriculturalists result
in cycles of raiding, pastoral aggrandizement, pastoral settlement and subse-
quent pastoral conquest as suggested by Ibn Khaldun, Owen Lattimore, and
Thomas Barfield?”
To test these propositions, incidents of raiding and sedentarization are
recorded per iteration, as well as the population numbers of pastoral nomads
and sedentary agriculturalists. If historical theories of cycles are correct, then
periods of intense raiding should be followed by periods of relative peace, until
pastoral numbers grow to a size where they again pose a more constant threat
to the agricultural population. If theories of sedentarization are correct, pasto-
ralists should run into problems with excess wealth and excessive poverty, and
poverty should be more common than enrichment. If these patterns emerge
from model runs, then not only will this confirm the validity of these theories,
but also, we can examine the model variables to gain a clearer understanding
of which were important in bringing about the results.
Model Runs and Results

Twenty model runs were performed to explore the range of resulting histories
and to test the sensitivity of the model’s parameters. An advantage of ABM
simulation is that every event is recorded so that when a particular outcome
emerges, such as the death of an agriculturalist family, the data log of the ABM
can be queried for the precise cause of the event. This capability allowed us to
derive the relationships among system variables we describe below. Overall,
the resulting pastoral-agricultural societies that we generated appeared similar
to ethnographically known societies, although the broad historical trends
researchers have recorded failed to emerge. We examine the possible reasons
for the model’s successes and failures. The following data illustrate a typical
run. Over a 200-season (50-year) period, the average nomad population was 43
(composed of families including a husband, wife and two immature children at
any time), and the agriculturalist population was 115 (with a family size of about
10). These figures capture the relative family size differences between nomads
and agriculturalists cited above, although the agriculturalists are perhaps
slightly more fertile than ethnographically observed. The average family herd
size was 113 sheep and goats. The average herd figures match the ethnographic
observations very well (Cribb 1991:35, Table 3.1a). Mobility patterns exhib-
ited a seasonal transhumance between summer pastures in highlands away
from sedentary villages and fall-winter settlement near agriculturalist trading
partners (Figure 25.1).
In a typical run, trading precedes raiding, but raiding begins to take prece-
dence over trading, being inversely related until both activities stabilize (Figure
25.2). In the run we are analyzing, the average trade/season/family was 0.09,
Figure 25.1. NOMAD screenshot: seasonal mobility graphics.

NO M A D     573
and the average raid/season/family was 0.15, or 0.36 trades and 0.6 raids/
year/family. Both activities typically took place during the harvest in the fall
(Figure 25.3). This was due to the availability of goods during the harvest
time of the year. It is important to note that nomads could trade or raid when-
ever a need arose, and indeed occasional off-season interactions took place.
However, an initial fall trade set a precedent for yearly seasonal trade and raid
Figure 25.2. NOMAD screenshot: trades and raids by season.

Figure 25.3. NOMAD screenshot: trade and raid graphics.
interactions as observed ethnographically (Barth 1961:99; Sweet 1969; Nielsen

2001). Interestingly, the average number of raids/year (0.6) was within the
order of magnitude of what Gray et al. (2003) indicate.
The runs provide insight into dimensions on which the model was particu-
larly sensitive. First, the productivities of labor in agricultural and pastoral
production easily tip prices to extremes. We found that agriculture was
particularly productive, devaluing Kcals and increasing the value of protein.
The price imbalance typically led to nomads practicing predatory pricing
and preventing agriculturalists from amassing wealth. Agriculturalists were
therefore constantly vulnerable to starvation. We had to allot agriculturalists a
modest protein supplement derived from their own wheat production, reason-
able enough, in order to keep their own Kcal/protein ratio from becoming
too imbalanced and driving down the value of their Kcals. This allotment
prevented the agriculturalists’ minimal rate of substitution from becoming
too large. Similarly, the Kcals derived from dairy production by nomads was
essential in preventing the nomadic Kcal/protein ratio from becoming overly
imbalanced. The need for not becoming too impoverished in what one’s
trading partner produces leads to the paradox that, although agriculturalists
and nomads are economic specialists, they must provide some small measure of
NO M A D     575
what the other produces so that they do not enter trade negotiations too weakly.
This necessary modest economic diversification could be one reason that the
myths of the carnivorous pure pastoralist and vegetable-eating agriculturalist
are untrue (Barfield 1993:3): real pastoralists have access to Kcal supplements,
and real agriculturalists have access to protein supplements.
Another implication of behavioral sensitivity to pricing is that, if prices
became overly favorable to one party (usually the nomads), then a pattern
of frequent raiding emerged. In runs where this took place, nomads raided
mercilessly and never traded with agriculturalists, eventually driving the agri-
culturalists to extinction. Our results were probably extreme, a result of the
artificial lack of economic alternatives or state protection for agriculturalists.
We think, however, that we have gained insight into one factor that might
lead to cycles of intense raiding by pastoral nomads. The most intense raiding
occurred when agriculturalists were unusually wealthy in Kcals, their primary
product, but very poor in stores of protein. In this situation nomads found
the most profitable course of action to be intense raiding, since agricultural-
ists could not afford pastoral products, despite their absolutely high wealth in
Kcals. We think that this result is reminiscent of Barfield’s (1989:9) observa-
tion that cycles of nomadic warfare in Asia coincided with periods of agrarian
prosperity. While food goods may not necessarily have been the economically
deciding factor, imbalances between agrarian stores of nomadic and agricul-
tural goods could have sparked a cycle of intense raiding. We think that this
proposition is worth further investigation.
Another implication of the sensitivity of behavior to pricing is the indication
that cultural norms like customary pricing, and markets, are probably necessary
for stable and profitable pastoral-agricultural specialization. We found much
more stable trading behavior when a customary set price was used instead
of the negotiated price based on marginal rates of substitution. Without a
market agriculturalists were occasionally, and in some runs frequently, wiped
out through predatory pricing, which in turn led to the eventual extinction of
nomads because of the lack of agriculturalist trading partners. A market would
permit agriculturalists to collectively dump their Kcals and stabilize a market
price, even a negotiated one, for nomad protein. In this way agricultural-
ists could sell their goods at a more stable price, reducing price fluctuations
and preventing their extinction, as well as that of the nomads. Interestingly,
archaeologists have noted that the fully nomadic pastoral specialization did not
emerge until after the emergence of towns and cities in Asia and the Middle
East (Lattimore 1951; Cribb 1991).
Although a pattern of intense raiding emerged in some runs, we failed to
detect major cycles of trading and raiding. Furthermore, pastoralists were not
reproducing quite fast enough to produce demographic pressure on resources
and to create new families. Without this pressure there could be no relative
impoverishment or enrichment of nomads. More fully specifying the model
will allow us to test whether raiding and trading cycles can emerge from the
interactions between nomads and agriculturalists or whether external political
influences drive such cycles, as some researchers suggest.
Discussion
ABMs can provide insights into the value of anthropological theories by
allowing researchers to test whether their theories could logically lead to the
phenomena they purport to explain. Using important theoretical concepts in
nomadic studies, we were able to generate a facsimile of pastoralist-agricul-
turalist interactions between individual families in an arid environment. Our
analysis of the ABM highlighted the need for modest diversification in pastoral
and agricultural economies, despite their respective economic emphases. The
sensitivity of behavior to pricing also highlights the potential long-term value
of markets and normative pricing for both agriculturalists and nomads. A
relationship between balancing complimentary goods, the volatility of pricing
and raiding was also observed, suggesting a possible influence on ancient
cycles. More research is necessary, however, to explore the basis for raiding
and trading cycles.
We have provided a generic example for purposes of illustrating the potential
use of ABMs for exploring theories of pastoralist-agriculturalist interactions. The
model’s sensitivity to some parameters, however, especially pastoral and agricul-
tural productivity, underscores the importance that local environmental, economic
and social context will have on the performance of any simulation. Therefore,
we encourage application of this tool to specific instances (see our application
to Darfur in Kuznar and Sedlmeyer 2005). General patterns may emerge from
such research, but they will be valid only after varied settings have been modeled,
providing a cross-simulation (analogous to a cross-cultural) analysis.
Finally, our experience in developing the NOMAD ABM reinforced the need
for quality empirical data collected through traditional ethnography. ABMs can
reflect the world only to the extent that their logic parallels real-world rela-
tionships and their parameters are grounded in actual data. Simulations are
not an alternative but a complement to traditional anthropological research.
Developing ABMs can enhance empirical data gathering as well. It becomes
clear, once one begins developing the logic for a simulation, what data are
required and what relationships need to be known in order to complete the
construction of a functioning simulation. Attempts at simulation before field-
work can guide researchers to the kinds of data that they need to gather to
address a particular research question.
NO M A D     577
References
Adano, W. R. and K. Witsenburg
2003 Surviving Pastoral Decline in Marsabit District. Amsterdam, Research Institute
for Global Issues and Development Studies.
Asad, T.
1973 The Beduin as a Military Force: Notes on Some Aspects of Power Relations
Between Nomads and Sedentaries in Historical Perspective. In C. Nelson
(ed.), The Desert and the Sown: Nomads in the Wider World. Berkeley, University
of California Press: pp. 61–73.
Axelrod, R.
1997 The Complexity of Cooperation: Agent-Based Models of Conflict and Cooperation.
Princeton, Princeton University Press.
Axtell, R. L., J. M. Epstein, J. S. Dean, G. J. Gumerman, A. C. Swelund, J. Harburger,
S. Chakravarty, R. Hammond, J. Parker and M. T. Parker
2002 Population Growth and Collapse in a Multiagent Model of the Kayenta
Anasazi in Long House Valley. In Proceedings of the National Academy of Sciences
99, Supplement 3, pp. 7275–7279.
Barfield, T. J.
1989 The Perilous Frontier: Nomadic Empires and China. Cambridge, Blackwell.
1993 The Nomadic Alternative. Englewood Cliffs, Prentice-Hall.
2004 Nomadic Pastoralists in Afghanistan: Reconstruction of the Pastoral Economy.
Washington, Bank Information Center.
Barth, F.
1961 Nomads of South Persia. New York, Holt, Rinehart and Winston.
C. Nelson (ed.), The Desert and the Sown: Nomads in the Wider World. Berkeley,
Bates, D. G.
1971 The Role of the State in Peasant-Nomad Mutualism. Anthropological Quarterly
44: 3, pp. 109–131.
Beck, L.
1991 Nomad: A Year in the Life of a Qashqa'i Tribesman in Iran. Berkeley, University
1998 Use of Land by Nomadic Pastoralists in Iran, 1970–1998. Bulletin of the Yale
School of Forestry and Environmental Science 103: pp. 58–80.
Ben-David, J.
1990 The Negev Bedouin: From Nomadism to Agriculture. In R. Kark (ed.), The
Land That Became Israel. Jerusalem, Magnes Press: pp. 181–195.
Berntsen, J. L.
1976 The Maasai and Their Neighbors: Variables of Interaction. African Economic
History 2: pp. 1–11.
Browman, D. L.
1987 Pastoralism in Highland Peru and Bolivia. In D. L. Browman (ed.), Arid Land
Use Strategies and Risk Management in the Andes. Boulder, Westview Press: pp.
121–149.
Burnham, P.
1979 Spatial Mobility and Political Centralization in Pastoral Societies. In Pastoral
Productivity and Society. Cambridge, Cambridge University Press: pp.
349–360.
Casimir, M. J.
1988 Nutrition and Socio-economic Strategies in Mobile Pastoral Societies in the
Middle East with Special Reference to West Afghan Pashtuns. In I. de Garine
and G. A. Harrison (eds.), Coping with Uncertainty in Food Supply. Oxford,
Clarendon Press: pp. 337–359.
Chagnon, N.
1992 Yanomamö. Fort Worth, Harcourt Brace.
Cook, C. W. and P. L. Sims
1975 Drought and Its Relationship to Dynamics of Primary Productivity and
Production of Grazing Animals. Paper presented at the Conference on Evaluation
and Mapping of Tropical African Rangelands, Bomako(Mali).
Cribb, R.
Cronk, L.
1998 Ethnographic Formation Processes. Social Science Information 37: 2, pp.
321–349.
Dahl, G.
1979 Ecology and Equality: The Boran Case. In L’Equipe écologie et anthropologie
des sociétés pastorales (ed.), Pastoral Production and Society/Production pastorale
et société. Cambridge, Cambridge University Press: pp. 261–282.
Dahl, G. and A. Hjort
1976 Having Herds: Pastoral Herd Growth and Household Economy. Stockholm Studies
in Social Anthropology 2. Stockholm, University of Stockholm.
Dean, J. S., G. J. Gumerman, J. M. Epstein, R. L. Axtell, A. C. Swedlund, M. T. Parker
and S. McCarroll
1999 Understanding Anasazi Culture Change Through Agent-Based Modeling.
In T. A. Kohler and G. J. Gumerman (eds.), Dynamics in Human and Primate
Societies. New York, Oxford University Press: pp. 179–205.
NO M A D     579
Devendra, C.
1978 Goats. In G. Williamson and J. W. Payne (eds.), Animal Husbandry in the
Tropics. New York, Longman: pp. 465–483.
Diarra, L. and H. Breman
1975 Influence of Rainfall on the Productivity of Grasslands. Paper presented at the
Conference on Evaluation and Mapping of Tropical African Rangelands, Bomako,
(Mali).
Dyson-Hudson, R. and N. Dyson-Hudson
1980 Nomadic Pastoralism. Annual Review of Anthropology 9: pp. 15–61.
Epstein, J. M. and R. L. Axtell
1996 Growing Artificial Societies: Social Science from the Bottom Up. Washington,
Brookings Institution Press.
Fratkin, E. and J. T. McCabe
1999 East African Pastoralism at the Crossroads: An Introduction. Nomadic Peoples
3: 2, pp. 5–15.
Fratkin, E., M. A. Nathan and E. A. Roth
1999 Health Consequences of Pastoral Sedentarization Among Rendille of
Northern Kenya. In D. M. Anderson and V. Broch-Due (eds.), The Poor Are
Not Us: Poverty and Pastoralism in Eastern Africa. Athens, Ohio University
Press: pp. 149–162.
Gatenby, R.
1986 Sheep Production in the Tropics and Sub-Tropics. New York, Longman.
Gellner, E.
1973 Introduction. In C. Nelson (ed.), The Desert and the Sown: Nomads in the Wider
World. Berkeley, University of California Press: pp. 1–9.
Glatzer, B.
1982 Processes of Nomadization in West Afghanistan. Studies in Third World Societies
18: pp. 61–86.
Glatzer, B. and M. J. Casimir
1983 Herds and Households Among Pashtun Pastoral Nomads: Limits of Growth.
Ethnology 22: pp. 307–325.
Goldschmidt, W.
1979 A General Model for Pastoral Social Systems. In L’Equipe écologie et anthro-
pologie des sociétés pastorales (ed.), Pastoral Production and Society/Production
pastorale et société. Cambridge, Cambridge University Press: pp. 15–27.
Gray, S., M. Sundal, B. Wiebusch, M. A. Little, P. Leslie and I. L. Pike
2003 Cattle Raiding, Cultural Survival, and Adaptability of East African Pastoralists.
Current Anthropology 44: pp. S3–S30.
Handwerker, W.
1983 The First Demographic Transition: An Analysis of Subsistence Choices and
Reproductive Consequences. American Anthropologist 85: pp. 5–27.
Hannoyer, J. and J.-P. Thieck

1982 Observations sur l’élevage et le commerce du Mouton dans la région de Raqqa
en Syrie.’ Production Pastorale et Société 14: pp. 47–63.
Hillman, G. and M. S. Davies
1990 Measured Domestication Rates in Wild Wheats and Barley Under Primitive
Cultivation, and Their Archaeological Implications. Journal of World Prehistory
4: 2, pp. 157–222.
Hole, F.
127–179.
Iannelli, P.
1984 The Principles of Pasture Improvement and Range Management and Their
Application in Somalia. Rome, FAO.
Ibn Khaldun (translated by F. Rosenthal)
1967 The Muqaddimah: An Introduction to History. Princeton, Princeton University
Press.
Irons, W.
1965 Livestock Raiding Among Pastoralists: An Adaptive Interpretation. Papers of
the Michigan Academy of Science, Arts, and Letters 1: pp. 393–414.
1972 Variation in Economic Organization: A Comparison of the Pastoral Yomut
and the Basseri. In W. Irons and N. Dyson-Hudson (eds.), Perspectives on
Nomadism. Leiden, E. J. Brill: pp. 88–104.
1979 Political Stratification Among Pastoral Nomads. In L’Equipe écologie et
anthropologie des sociétés pastorales (ed.), Pastoral Production and Society/
Production pastorale et société. Cambridge, Cambridge University Press: pp.
361–374.
Kahn, P. (ed.)
1998 The Secret History of the Mongols: The Origin of Chingis Kahn. Boston, Cheng
and Tsui.
Khazanov, A.
edition).
Kohler, T. A., J. Kresl, C. Van West, E. Carr and R. H. Wilshusen
1999 Be There Then: A Modeling Approach to Settlement Determinants and
Spatial Efficiency Among Late Ancestral Pueblo Populations of the Mesa
Verde Region, U.S. Southwest. In T. A. Kohler and G. J. Gumerman (eds.),
Dynamics in Human and Primate Societies. New York, Oxford University Press:
pp. 145–178.
NO M A D     581
Krader, L.
1979 The Origin of the State Among the Nomads of Asia. In L’Equipe écologie
et anthropologie des sociétés pastorales (ed.), Pastoral Production and Society/
231–234.
Kuznar, L. A.
1991a El medio ambiente y la capacidad de carga de la Sierra Alta de los Andes sur
central en el departamento de Moquegua, Perú [in Spanish]. Diálogo Andino
10: pp. 100–112.
1991b Herd Composition in an Aymara Community of the Peruvian Altiplano as a
Linear Programming Problem. Human Ecology 19: pp. 369–387.
1995 Awatimarka: The Ethnoarchaeology of an Andean Herding Community. Fort
Worth, Harcourt Brace.
1999 Traditional Pastoralism and Development: A Comparison of Navajo
and Aymara Grazing Ecology. In T. L. Gragson and B. G. Blount (eds.),
Ethnoecology: Knowledge, Resources, and Rights. Athens, University of Georgia
Press: pp. 74–89.
2001 Risk Sensitivity and Value Among Andean Pastoralists: Measures, Models,
and Empirical Tests. Current Anthropology 42: 3, pp. 432–440.
2006 Hi Fidelity Computational Social Science in Anthropology: Prospects for
Developing a Comparative Framework. Social Science Computer Review 24: 1,
pp. 1–15.
Kuznar, L. A. and R. Sedlmeyer
2005 Collective Violence in Darfur: An Agent-Based Model of Pastoral Nomad/
Sedentary Peasant Interaction. Mathematical Anthropology and Culture Theory
1: 4, pp. 1–22 (on-line at http://www.mathematicalanthropology.org/)
Lattimore, O.
1951 Inner Asian Frontiers of China. Boston, Beacon Press (second edition).
1979 Herdsmen, Farmers, Urban Culture. In L’Equipe écologie et anthropologie
des sociétés pastorales (ed.), Pastoral Production and Society/Production pastorale
et société. Cambridge, Cambridge University Press: pp. 479–490.
Lazar, A. and R. G. Reynolds
2002 Computational Framework for Modeling the Dynamic Evolution of Large-
Scale Multi-agent Organizations. Paper presented at the Computational
Analysis of Social and Organizational Systems Proceedings, Pittsburgh.
Lewis, O.
1965 Village Life in Northern India. New York, Random House.
Martin, M. and B. Voorhies
1975 Female of the Species. New York, Columbia University Press.
McCorriston, J.
1997 The Fiber Revolution: Textile Extensification, Alienation, and Social
Stratification in Ancient Mesopotamia. Current Anthropology 38: 4, pp.
517–549.
McElreath, R., R. Boyd and P. J. Richerson
2003 Shared Norms and the Evolution of Ethnic Markers. Current Anthropology 44:
1, pp. 122–129.
Nielsen, A. E.
2001 Ethnoarchaeological Perspectives on Caravan Trade in the South-Central
Andes. In L. A. Kuznar (ed.), Ethnoarchaeology of Andean South America:
Contributions to Archaeological Method and Theory. Ann Arbor, International
Monographs in Prehistory: pp. 163–201.
Roth, E. A.
1986 The Demographic Study of Pastoral Peoples. Nomadic Peoples 20: pp.
63–76.
Salzman, P. C.
1979 Inequality and Oppression in Nomadic Society. In L’Equipe écologie et
anthropologie des sociétés pastorales (ed.), Pastoral Production and Society/
429–446.
Schneider, H. K.
1979 Livestock and Equality in East Africa: The Economic Basis for Social Structure.
Bloomington, Indiana University Press.
Shepardson, M. and B. Hammond
1970 The Navajo Mountain Community. Berkeley, University of California Press.
Sidky, H.
1995 Hunza: An Ethnographic Outline. Jaipur, Illustrated Book Publishers.
Simon, R.
1981 Symbiosis of Nomads and Sedentaries on the Character of the Middle Eastern
Civilization. Acta Orientalia Academiae Scientiarum Hungary 35: 2–3, pp.
229–242.
Spooner, B.
1972 The Status of Nomadism as a Cultural Phenomenon in the Middle East. In
W. Irons and R. Dyson-Hudson (eds.), Perspectives on Nomadism. Leiden, E.
J. Brill: pp. 122–131.
Sweet, L.
1969 Camel Pastoralism in North Arabia and the Minimal Camping Unit. In A. P.
Vayda (ed.), Environment and Cultural Behavior. Austin, University of Texas
Press: pp. 157–180.
NO M A D     583
Swidler, W. W.
1972 Some Demographic Factors Regulating the Formation of Flocks and Camps
Among the Brahui of Baluchistan. In W. Irons and N. Dyson-Hudson (eds.),
Perspectives on Nomadism. Leiden, E. J. Brill: pp. 69–75.
Swift, J.
1979 The Development of Livestock Trading in Nomad Pastoral Economy: The
Somali Case. In L’Equipe écologie et anthropologie des sociétés pastorales
(ed.), Pastoral Production and Society/Production pastorale et société. Cambridge,
Terna, P.
1998 Simulation Tools for Social Scientists: Building Agent Based Models with
SWARM. Journal of Artificial Societies and Social Simulation. Volume 1, Number
2 (on-line at http://www.soc.surrey.ac.uk/JASSS/1/2/4.html).
Thomas, R. B.
1973 Human Adaptation to a High Andean Energy Flow System. Department of
Anthropology Occasional Paper No. 7. University Park, Pennsylvania State
University.
List of C ontributors
Abbas Alizadeh (Chapter 4)

Director of Iranian Prehistoric Project
Oriental Institute, University of Chicago (USA)
Hans Barnard (Chapters 1 and 19)

Research Associate, Cotsen Institute of Archaeology
University of California, Los Angeles (USA)
Reinhard W. Bernbeck (Chapter 3)

Professor, Department of Anthropology
Binghamton University (USA)
Alison Betts (Chapter 2)

Associate Professor and Chair, Department of Archaeology
University of Sydney (Australia)
Alexander V. Borisov (Chapter 10)

Senior Researcher,
Institute of Physical and Biological Problems in Soil Science
Russian Academy of Science (Russia)
David L. Browman (Chapter 7)

Professor of Archaeology, Department of Anthropology
Chair, Interdisciplinary Program in Archaeology
Washington University in Saint Louis (USA)
Giorgio Buccellati (Chapter 6)

Director, Mesopotamian Laboratory, Cotsen Institute of Archaeology
Professor Emeritus of Near Eastern Languages and Cultures
Professor Emeritus of History
University of California, Los Angeles (USA)
Stanley M. Burstein (Chapter 11)

Professor Emeritus of History
California State University, Los Angeles (USA)
Claudia Chang (Chapter 15)

Professor, Department of Anthropology and Sociology
Sweet Briar College (USA)
584
List of C o n t r i b u t o r s      585
Roger L. Cribb † (Chapter 24)

Adjunct Research Fellow, Department of Anthropology, Archaeology and
Sociology, School of Arts and Social Sciences at the Cairns Campus of James
Cook University (Australia)
Jelmer W. Eerkens (Chapter 14)

Associate Professor, Department of Anthropology
University of California, Davis (USA)
Michael D. Frachetti (Chapter 17)

Assistant Professor, Department of Anthropology
Washington University in Saint Louis (USA)
Eugeny I. Gak (Chapter 10)

Researcher, Department of Complex Archaeological Investigation
State Historical Museum, Moscow (Russia)
Margaret B. Holman † (Chapter 13)

Research Associate, Anthropology
Michigan State University Museum (USA)
Esther Jacobson-Tepfer (Chapter 9)

Maude I. Kerns Professor of Asian Art
Department of Art History, Architecture, and Allied Arts
University of Oregon (USA)
Lawrence A. Kuznar (Chapter 25)

Indiana University–Purdue University at Fort Wayne (USA)
William A. Lovis (Chapter 13)

Curator at the Michigan State University Museum
Michigan State University (USA)
Anwar Abdel-Magid (Chapter 20)

Academic Adviser for Sudan Cooperation
Senior Professor, Center for Africa Studies
University of the Free State (South Africa)
S. Brooke Milne (Chapter 8)

Assistant Professor Archaeology, Department of Anthropology
University of Manitoba, Winnipeg (Canada)
586    L i s t of Contributors
Alan G. Roe (Chapter 22)

Senior Researcher,
Afghanistan Research and Evaluation Unit (Kabul, Afghanistan)
Steven A. Rosen (Chapter 5)

Professor, Department of Bible, Archaeology, and Ancient Near Eastern
Studies, Ben-Gurion University of the Negev (Israel)
Benjamin A. Saidel (Chapter 21)

Assistant Professor, Department of Anthropology
East Carolina University (USA)
Robert Sedlmeyer (Chapter 25)

Associate Professor, Department of Computer Sciences
Indiana University–Purdue University at Fort Wayne (USA)
Natalya I. Shishlina (Chapter 10)

Senior Researcher, Archaeology Department
State Historical Museum, Moscow (Russia)
Andrew B. Smith (Chapter 12)

Associate Professor, Department of Archaeology
University of Cape Town (South Africa)
Stuart T. Smith (Chapter 16)

Associate Professor, Department of Anthropology
University of California, Santa Barbara (USA)
Jeffrey J. Szuchman (Chapter 18)

Post-Doctoral Scholar, Oriental Institute
University of Chicago (USA)
Willeke Z. Wendrich (Chapters 1 and 23)

Associate Professor of Egyptian Archaeology, Department of Near Eastern
Languages and Cultures, University of California, Los Angeles (USA)
List of Figures
1.1. World map showing the location of regions discussed in this volume
1.2. Schematic representation of mobility patterns
2.1. Map of Jordan, showing the position of sites mentioned in the text
3.1. Map showing the sites of the Halaf tradition mentioned in the text
3.2. Habitation by phase at Fıstıklı Höyük
3.3. Sherd jetons from Fıstıklı Höyük
3.4. Calcareous plates from Fıstıklı Höyük
3.5. Dynamics and structures of a Halafian multisited community on the upper
Euphrates
4.1. Map of Iran and Mesopotamia
4.2. Map of southwestern and south-central Iran
4.3. Satellite image of northern Fars
4.4. Semipermanent Qashqāii village and permanent local village, near
Firuzabad
4.5. Woman and children in a mobile pastoralist semipermanent Bakhtiyāri
village near Behbahan
4.6. Samples of ‘dot motif’ pottery vessels of the Late Susiana 1 phase
4.7. Map of KS-1626, showing the location of excavated areas
4.8. General view of KS-1626, looking south
4.9. Objects from Grave 3
4.10. Bakhtiyāri spring camp at Chogha Mish (KS-01)
4.11. Bakhtiyāri tent in eastern Khuzestan
4.12. Dar Khazineh (KS-1626); remnants of temporary camp fire pits and
Bakhtiyāri encampment.
5.1. Schematic graph of the agricultural boundary in the Negev
8.1. Map of the Canadian Arctic and Greenland, showing the route of the
Paleo-Eskimos
8.2. Examples of Paleo-Eskimo lithic artifacts illustrating their tiny size
8.3. Map of southern Baffin Island, in the Canadian Arctic
8.4. Map of Burwash Bay
8.5. Map of the northwest shore of Mingo Lake
8.6. Map of the Mingo River, showing the location of LeDx-42
9.1. Map of the Altai Mountain Region, northern Central Asia
9.2. View of Aral Tolgoi from mountains to the south
9.3. View looking over the Baga Oigor River up the Tsagaan Salaa River
587
588   L i s t of Figures
9.4. Aral Tolgoi: Pecked images of large birds and a horse

9.5. Aral Tolgoi: Pecked image of a wooly rhinoceros
9.6. Aral Tolgoi: Pecked image of an elk
9.7. Upper Tsagaan Gol complex: Image of pre–Bronze Age elk overlaid by a
smaller Bronze Age elk
9.8. Upper Tsagaan Gol complex: Pecked image of a rider preparing to shoot
his bow (detail)
9.9. Upper Tsagaan Gol complex: Pecked scene of hunters attacking an aurochs
9.10. Tsagaan Salaa/Baga Oigor complex: Drawing of the pecked image of a
mammoth
9.11. Tsagaan Salaa/Baga Oigor complex: Drawing of the pecked image of an
aurochs
9.12. Tsagaan Salaa/Baga Oigor complex: Drawing of overlaid images of elk and
other animals
9.13. Tsagaan Salaa/Baga Oigor complex: Drawing of a scene depicting a raid on
a family caravan
9.14. Tsagaan Salaa/Baga Oigor complex: Drawing of a scene with two wheeled
vehicles
9.15. Tsagaan Salaa/Baga Oigor complex: Drawing of a hunt scene
10.1. Map showing the position of the study area in the north Caspian Steppe
10.2. Site location map of Manych and Gashun-Sala
10.3. Manych: excavation trenches, a shallow pit, and copper
10.4. Finds from Manych
10.5. Finds from Gashun-Sala
10.6. Assemblage of handmade potsherds found at Gashun-Sala
10.7. Excavation trenches, handmade ceramics, animal bones at Gashun-Sala
10.8. Seeds of amaranth found at Gashun-Sala
10.9. Chemical properties of soils from Manych and Gashun-Sala
11.1. Map of southeast Egypt and northeast Sudan
12.1. Adam and Eve (engraving by Albrecht Dürer, 1504)
12.2. Khoekhoe family at Algoa Bay (woodcut by Hans Burgkmair, 1508)
12.3. Khoekhoe women interacting with colonists at the Cape
12.4. Estimated route of movement of domestic animals into southern Africa
13.1. The Great Lakes region on the border between Canada and the United
States
13.2. Locations of Native American people in the Great Lakes region during the
Historic Period
13.3. Regional environmental zones in the study area
13.4. Mackinac phase and other early Late Woodland sites mentioned in the text
13.5. Juntunen phase and other later Late Woodland sites mentioned in the text
14.1. Map of the border region between California and Nevada
LI s t of F i g u r e s      589
14.2. Cooking pot from Owens Valley, California

14.3. Distribution of pottery in Owens Valley, California
14.4. Owens River, near Big Pine
14.5. Owens Valley desert shrub landscape
14.6. Piñon-Juniper landscape of the White Mountains
14.7. High-altitude alpine desert landscape of the White Mountains
14.8. Density of the concentration of small seeds found on house floors in Owens
Valley
15.1. Map showing the location of the study area in southeast Kazakhstan
15.2. A Kazakh herdsman and his flock in front of an Iron Age kurgan
16.1. Ramesses III relief at mortuary temple in Medinet Habu
16.2. Nubian warriors on a painted box found in Tutankhamun’s tomb
16.3. Plan of the fortress at Askut (Sudanese Nubia)
16.4. Distribution of pottery found at Askut
16.5. Two figurines from Askut (Sudanese Nubia)
16.6. Plan of the Pyramid of Siamun and Weren at Tombos
16.7. Nubian-style burial of a woman at Tombos
17.1. Eurasian steppe on the border of Kazakhstan and China
17.2. Map of the Koksu River Valley
17.3. Seasonal environmental conditions in the Koksu River Valley
17.4. Excavated Bronze Age stone cist burial at Begash-2
17.5. Reconstruction of a Bronze Age semisubterranean house in Buguly (central
Kazakhstan)
17.6. Remains of an ephemeral settlement in the Koksu River Valley
17.7. Contemporary Kazakh settlement in the Koksu River Valley
17.8. Variable land-use pathways based on the quality and productivity of the
vegetation
17.9. Variable land-use pathways calculated to account for pasture quality,
terrain, and significant social locales
18.1. Map of border region between Anatolia and Mesopotamia
19.1. Examples of Eastern Desert Ware
19.2. Border area between Egypt and Sudan where Eastern Desert Ware has
been described
19.3. Places where clay and temper were collected for experiments
19.4. Preparation of potter’s clay
19.5. The shaping and smoothing of vessels
19.6. Saggars used in California and in Egypt
19.7. Temperature curves of three experimental firings
19.8. Firing vessels in California and in Egypt
19.9. Examples of vessels produced in California and in Egypt
20.1. Beja main tribal groups and the location of the study area
590   L i s t of Figures
20.2. A typical Hadendowa Badaigaw (mat tent-dwelling)

20.3. A euphorbia stem-dwelling in the north of the study area
20.4. Camels carrying palanquins in Nubia Sarracenorum (detail)
21.1. An early Iron Age I four-room house in Fields A and B at Tall al-'Umayri in
Jordan
21.2. Plan of a Bedouin tent in Wadi Fatima, Saudi Arabia
21.3. Plan of a Bedouin ‘clay house’ in Wadi Fatima, Saudi Arabia
22.1. General location of the study area in Egypt
22.2. The oases of the northern Libyan Desert
22.3. Traditional migratory destinations in the northern Libyan Desert
22.4. Overlapping territorial claims to the oases in the northern Libyan Desert
23.1. The Eastern Desert, between the Nile Valley and the Red Sea
23.2. Preparation of djabana: pouring water into a tin can
23.3. Preparation of djabana: roasting coffee beans
23.4. Preparation of djabana: pounding the roasted coffee beans
23.5. Preparation of djabana: pouring the ground coffee
23.6. Preparation of djabana: the coffee with ginger is ready
23.7. Headrest (mutir'is), acacia wood with aluminum repair
23.8. Headrest (mutir'is), leather over wooden frame
23.9. Milking basket (kahel), made of finely coiled doam palm leaf
23.10. Building a beit al-burush, forked acacia branches form a frame
23.11. Building a beit al-burush, special rope for bindings
23.12. Building a beit al-burush, the inner walls
23.13. Building a beit al-burush, the outer walls
23.14. Building a beit al-burush, mats and stitching
23.15. Orientation of the beit al-burush
23.16. Drawing of Ababda men and their camels
23.17. The Ababda sword dance
23.18. The dance of Ababda women
23.19. Acacia tree with an Abadi’s belongings
23.20. An Ababda tomb
24.1. Aboriginal painting showing relations between individuals in a group
24.2. Plan of the new Lockhart River settlement around 1973
24.3. Dry season beach camp at Lockhart River
25.1. NOMAD screenshot: Seasonal mobility graphics
25.2. NOMAD screenshot: Trade raid graphs
25.3. NOMAD screenshot: Trade and raid graphics
List of Tables
1.1. Overview of the Subject Matter of the Chapters in This Volume

1.2. Differences Between Hunter-Gatherers and Pastoral Nomads
1.3. Factors and Range of Mobility
2.1. Chronology of the Neolithic of the Near East
2.2. Radiocarbon Determinations for Sites Excavated by Betts in the East of the
Jordanian Steppe
2.3. Relative Numbers of Sheep-Goat Remains on Jordanian Steppe Sites
3.1. Traditional and Current Views of Mobile Groups
3.2. Phases and Their Absolute Dates at Fıstıklı Höyük
3.3. Structures and Installations in Use During Each Phase of Occupation at
Fıstıklı Höyük
3.4. Phytolith Samples with Wheat/Barley Husks per Phase
3.5. Occurrences of Jetons, Sickle Blades, and Calcareous Plates at Fıstıklı
Höyük
8.1. Flake-to-Tool Ratios Identified for Five Inland Paleo-Eskimo Sites on
Southern Baffin Island
8.2. Radiocarbon Dates for Five Inland Paleo-Eskimo Sites on Southern Baffin
Island
10.1. 14
C Data from Manych and Gashun-Sala
10.2. Seasonality of the Death of Animals Found at Gashun-Sala and Chilyuta
14.1. Population Levels, Average Precipitation and Estimated Degree of
Residential Stability for Six Regions of the Western Great Basin
14.2. Distribution of Pottery by Environmental Zone
19.1. A Possible Chaîne Opératoire for Archaeologically Recovered Sherds of
Eastern Desert Ware
21.1. Selected Examples of the Size of Bedouin Tents
21.2. Objects Found in the Bedouin Tent as Recorded by Burckhardt, Musil and
Dickson
21.3. Prices of Foodstuffs by ‘Load’
22.1. Ovicaprid Oasis Forage Species Ranked by Importance
23.1. Overview of Ababda Families and Tribes
23.2. Overview of the Ababda Tribe and Its Subdivisions
23.3. Priority List for the Collection of Ababda Objects
591
I ndex
Ababda (nomads) nutritional demands, 566 landscape of, 146

clothing of, 528 population in, 571 language of, 143
communication system raiding mortality in, 567, in Mari, 142
of, 534 571 military independence of,
in Eastern Desert, 509 results of simulations, 143
families/tribes, 513f 571–576 Palmyra and, 154
foodways of, 527 seasonal mobility, 572f in Syro-Mesopotamia, 141
gender priorities, 529f seasonality in, 573 trading system of, 153
genealogy of, 538 social rules in, 568 in Ur III period, 142
hairstyles of, 528 agricultural activities analogical reasoning
as hunter-gatherers, 516 Bakhtiyaari and, 90 failure of, in analyzing
hunting and, 535 mobile pastoralism and, mobile groups, 49,
interpretations of past, 539 89, 91 372–373
landscape impacted by, 535 pastoralism and, 373, 407 perception of integrated
material traces left by, 536 rainfall influencing, 514 communities
matrilocal social structure agriculturalists, 545, 557 influencing, 45–46
of, 522 agropastoralism, 7 Anatolia, 398f
medicine of, 531–532 Bedouin, 28 ancestor worship, 340
Mohamed connection to, in Central Andes, 162, anchoring, 61–62
512 166–167, 168 animal figurines, 164
multi-resource nomadism in Iron Age, 330 animal husbandry, 478
of, 511–512 ahiru, 151 animals. See aurochs (wild
music/dance of, 533f Ahlamu, 400 cattle); camel(s); gazelle;
overview of tribe, 515f Akhenaton, Pharaoh, 345 horses; sheep/goats;
sword dance of, 532f Aldenhovener Platte, 50 slaughtered animals
tombs of, 536, 537f alpacas, 163 anthrophages (man-eaters), 264
women social customs, 516 alpine desert, 316f anthropology, 576
ABMS. See agent-based models Altai Mountains, 200 Arabic language, 513–514
aboriginal painting, 546f herder-gatherers in, 224 Aral Tolgoi, 208, 209f, 212,
aboriginal populations, 544, map of, 203f 213f, 214f
550 rock art in, 206, 208 Aramaeans
acacia branches, forked, 524f semi-nomadic pastoralism Ahlamu and, 400
acacia tree, 537f in, 223 as multi-resource nomads,
Adam and Eve (Durer), 264, altiplano, 160, 170 403
265f Alyawara, 544 Roux on, 397
adaptive strategies, 286 Amadjuak, 186 sedentarization of, 398, 401
adornment, 530 Amadjuak Lake, 190f Tiglath-Pileser I on,
Agatharchides of Cnidus, amaranth seeds, 241f 399–400
253–258, 260 Amazon, 545 archaeobotanic determinations,
agent-based models (ABMS), American southwest, 50 240
557 Ammianus Marcellinus, archaeological excavations, 233
anthropology and, 576 258–259 archaeological fieldwork, 45
computer simulation of, Ammuru, 155 archaeological record
563––576 Amorite salt, 153 of Amorites, 155
demographic rules in, 568 Amorite wall, 155 of Baffin Island, 192t
environment in, 564 Amorites of Bedouin, 473
fertility rates in, 567 Ammuru and, 155 equality in, 337
herd size as factor in, 570 archaeological record of, ethnicity and, 343, 348–349
human agents on, 566 155, 156 Gashun-Sala, 236
labor rates in, 567 dynasties of, 143 hierarchy in, 337
movement rules in, 568 ethno-archaeology on, 156 historical record and, 250
592
I n d e x      593
of the Khoekhoen, 270 archaeological record of, el -Maqrizi on, 450

kurgans as, 230 473 material culture of, 511
lack of, for Amorites, 156 Asheibat, 500n13 milk/meat diet, 452
of Late Woodland people, Awlad Ali, 493 tribal groups of, 442f
285 campsites investigated, 465 Tuaregs dwellings and, 455
of oases, 500 in caves and rock shelters, Beni Amer, 441
from ritual structures, 1 467 Berber, Bonavia, 171n5
of steppe pastoralists, 155 caves as shelter for, 498 Berber language, 498
in Syro-Mesopotamia, 154 changes in encampments Berghe, Louis Vanden, 93
archaeological sites, 474 of, 479 Bering Strait, 178
archaeological survey, 4t clay house, 478f Berinike, Egypt, 358, 416
archaeozoological coffee’s importance to, 470 Berntsen, J.L., 570
identifications, 239–240, domination of, 561 berries, 288
388–389 economic relations with Besseri (nomads), 404
archaeozoological studies, 12 oaseans, 500 Blemmyes (nomads), 250, 253,
archery, 207, 212 ethno-archaeological 357
architecture, 121–122, 475 research on, 479 Beja and, 256
Arctic Small Tool tradition farming controlled by, 125 in Berinike, 358
(ASTt), 177 Gaza Ware and, 414–415, Eastern Desert Ware and,
argali, 212 472 260–261
Argentina, 170 Musil on customs of, 469 Eratosthenes on, 260
armature, 446 population of, 129 historical record of, 512
Asheibat, 500n13 post-Neolithic tents and, in Nile Valley, 258–259
Askut, Egypt, 343, 350f, 351f, 466 sedentarization of, 260
353f sedentarization of, 466 as sedentary community,
Assyrians, 401–402 Sinuhe and, 449 258
ASTt. See Arctic Small Tool tent camps of, 473, bofedales, 167, 168
tradition 476–477, 477f Bolivia, 170
aurochs (wild cattle), 212, 216f, tents, sizes based on tribe, bone dry stage, 424
219f 467f bone isotope analysis, 47n3
tobacco and, 471–472 bone tool, 237f
Baffin Island, 176, 185f, 188, tribal relationships and, 37 Boyr Ahmadi, 89
191, 192t Bedouin black tent Bronze Age, 202, 211, 247
Baga Oigor River, 210f, 211 construction of, 467 adaptive strategies in, 374
Bakhtiyaari ethno-archaeological value Assyrians in, 401–402
agriculture and, 90 of, 475 mobile pastoralism in, 379
encampment, 100, 100f history of, 466 rock art in, 207
multi-resource nomadism Insoll on men/women stone cist burial, 383f
in, 87 division in, 469 Talapty settlements in, 391
political organization of, 85 Israelite four room house Buguly, central Kazakhstan,
spring camp at Chogha and, 475 385f
Mish, 98, 98f men/women section of, 468 building, 145, 453, 523, 524f.
tent in Khuzestan, 99, 99f in North Africa, 453 See also house building
Bakun style, 85 objects found in, 468f Burckhardt, L., 452
Bambata, 270 beechnuts, 287 burial cult, 338, 340
Barfield, T.J., 370 Bega, 253 burial grounds, 382–384, 429
barley, 58, 100 Begash (Kazakhstan), 380, burial mounds, 115, 339
Barth, Frederick, 548, 570 382–384 burial pits, 97
basalt, 38 Behbahan, 84, 84f burin Neolithic sites, 32, 36
base camp, 5 beit al-burush (mat house), burning, 496–497
baskets, 313, 320 525f, 526f burnishing, 424
Basseri (nomads), 548 beit sha’ar (house of hair). See Burwash Bay, 188f
BCE, second millennium, 400, Bedouin black tent
406, 492 Beja, 253 caching, 313–314, 316–317,
BCE, third millennium, 144, Blemmyes and, 256 319
373 camel and, 443 calcareous plates, 59–60, 60f,
Bedouin. See also alyawara Hadendowa and, 441 60t
agropastoralism of, 28 history of dwellings, California, 308f, 431f
animal husbandry of, 478 460–461 camel(s)
594    I n d e x
Beja and, 443 Centre National de la Recherche Christides, V., 257

camelids, 160 Scientifique, 99 Christ’s thorn plant, 254
carrying palanquins, 459f ceramic technologies. See also chronology, 243t
domestication of, 124 pottery-making cisterns, 126
gear, 530 bone dry stage, 424 city-state model, 151
herding/migration, 502 burnishing, 424 clan leaders, 340
men and, 531f challenges for mobile clay, 420
military superiority communities, 310 climate change, 28, 204–205
through, 92 clay in, 420 clothing, 528
packsaddle for, 459–460 coiling, 422 clustered room architecture,
as subsistence strategy, 493 decorations, 422–423 121
camelid pastoralism development in, 318–319 coffee, 477, 531
Berber on, 171n5 experimental firings in, coffee ceremony, 517, 518f,
in Central Andes, 167, 170 428f 519f, 520f, 530, 539
Murra on, 161 fire clouds, 426 of Ababda, 517
Rowe on, 161 firing, 424, 426, 430f in Bedouin life, 470
Canada, 281f grog for, 432 ceramics for, 432
Canadian Arctic, 175f, 185f kiln, 426 djabana for, 517
Canadian Biotic Province, leather hard stage, 424 hess al-mihbas (mortar and
286–288, 299 limited production time pestle), 470
Cape York Peninsula in, 317 coiling, 422
(Australia), 550, 555 of Mata Ortiz, 420, 422 collective burial, 296
potters clay, 423f Colobi, 254
caravan stops, 171
saggar, 426, 427f, 428 colonization, 174, 178, 272
caravans, 217, 220, 221f
of Santa Clara, New communal burial sites, 94
caribou, 186, 188, 189, 192
Mexico, 422 communication system, 534
Carolinian Biotic Province,
seasoning in, 429 communities
286, 288
shaping/smoothing vessels, dispersed, 43
carpets, 523
425f imagined v. practiced, 65
Caspian Steppe, 231f
slipping, 424 integrated, 43, 45–46
cattle, 353, 492. See also
temper in, 420 mobile pastoralist, 79, 87
aurochs (wild cattle)
unglazed earthenware multisited, 44, 64f, 65, 66
caves, 456, 467n3, 498
through, 429 computer modeling, 559
Central Andes conflict, 543, 548, 553–554. See
ceramics. See also pottery
agropastoralism in, 162, in Central Andes, 165 also raiding
166–167, 168 for coffee ceremony, 432 conquest, 559–560
animal figurines in, 164 in East Africa, 270 cooperative buffering, 293, 300
camelid pastoralism in, experimental archaeology copper, 123, 235f
167, 170 and, 14 copper pin, 98, 98f
caravan stops in, 171 at Gashun-Sala, 236 Cotsen Institute of
ceramics in, 165 of Native Americans, 297 Archaeology, 2
dry-season sites in, 165 in South Africa, 273 Cribb, R., 6, 548
herding outposts in, 171 chaine operatoire (operational cuisine, 352. See also foodstuffs;
Inca period in, 166 sequence), 416, 418f foodways
livestock management in, chariots, 201n3 cultural biography, 52, 73
160, 164, 168, 169 charwa, 369 cuneiform texts, 156
pastoral nomadism in, 169 chert, 5–6, 38, 185, 186, 193 Cyrenaica, 492, 502
population of semi- Chicago, University of, 95
nomadic pastoralists chief (khan), 333–334, 340 Dafa’alla, Samia, 257
in, 163 Chilyuta, 244t dagger, 530
semi-nomadic pastoralism Chinaman Creek camp, Dar Khazineh, 97, 100, 100f
in, 160, 169 552–553 Darfur, Sudan, 564
tethered nomadism in, 167 chipped stone, 30, 35 date palm, 433n10, 433–434,
Wanka peoples in, 163 Chippewa, 283, 289, 291–292, 491
Wari Empire in, 166 294 decorations, 422–423, 433
wet-season sites in, 163 Chogha Mish (Susiana), 94, deer stones, 207
Central Plateau (Iran), 101, 98, 98f demographic rules, 568
104 chomur, 369 dental cementum analysis,
Central Sahara, 451 Christianity, 130 47n3
I n d e x      595
Desert and the Sown, 117, rainfall in, 534 environmental zones, 286f,
123, 126 Toyota Hilux in, 535 317t
desert kites, 120n3 traders in, 434 ephemeral settlements, 386,
desert shrines, 121, 130 Eastern Desert Antiquities 387f, 388
desert wells, 125 Protection Process ephemeral sites, 1, 15, 189
Dhuweila, 34 (EDAPP), 509 Epipaleolithic period, 35
diaspora, 44 Eastern Desert Ware, 415f, 539 equality, 337
dikes, 56 Blemmyes and, 260–261 Eratosthenes, 260
dimorphic society, 131, 139 chaine operatoire for, 416, essentialist approach, 359
dispositional practices, 62 418f ethnic dynamics, 343
djabana (coffee pot), 517, 518f, decorations of, 433 ethnicity
519f, 520f, 530, 539 demand-driven production agent-centered approach
DMAP. See Dzunghar of, 419 to, 357
Mountains Archaeology as ethnic marker, 419 archaeological record on,
Project ethnicity and, 358 343, 348–349
dog burials, 297 foodways maintained by, dynamic nature of, 346
domestic complex, 548 360 Eastern Desert Ware and,
domesticated animals, 271f in graves, 429 358, 419
donkey, 124, 443, 498, 501 household production of, essentialist approach to,
dot-motif, 86, 86f 419 359
dowry, 569 as information source for foodways expressing, 349
drinking camps, 553 desert dwellers, 416 funerary practice
dry-season sites, 165, 168 pastoral nomads and, 435 expressing, 349
Durer, Albrecht, 264, 265f household production
economic relations, 123, 129,
dwellings. See also tent- supporting, 435
489, 500
dwelling (maskan) Nubian cuisine and, 352
economic structures, 562
Bedouin tents, 466, 467f origin of, 344
evolution of, 151
Beja/Tuaregs, 455 pastoral nomadism and,
as factor in mobility, 47
caves, 456 357
food-producing, 488
history of, in Beja and Pharaoh Akhenaton on,
herd animals influencing,
Hadendowa, 460–461 345
120
of hunter-gatherers, 445 religious practice
of herd management,
litters, 458 expressing, 349
561–562
mat house, 447 ethnie, 344
mixed, 43 ethno-archaeology, 3
military tents, 448
in mobile pastoralism, 90 Amorites and, 156
naming schemes for, 445
palanquins, 458 in the Negev, 128 archaeological record and,
radial pattern of between pastoral nomads 487
arrangement of, 551 and settlements, 119 on Bedouin, 479
reed huts, 456 subsistence, 99–100 Bedouin black tent and,
of reed stems/palm leaves, EDAPP. See Eastern Desert 475
501 Antiquities Protection objective of, 465
rugs made into, 434 Process pitfalls of, 13, 117
snow houses, 179 Eerkens, J.W., 413 Ethnographic Gallery of the
stem-dwelling, 457 Egypt, 251f, 350, 417f, 421f, South African Museum,
tent-ring structures, 187 431f, 490f. See also Askut, 268
Dzunghar Mountains Egypt; Berinike, Egypt; ethnographies, 252, 366, 372,
Archaeology Project Libyan Desert of Egypt; 492, 502, 503
(DMAP), 377, 379f Tombos, Egypt analogical reasoning and,
Egyptians, 538 46
Early Bronze Age, 123, 128 elites, 344 of Middle East, 331
Early Intermediate Period, 165 elk, 206, 214f, 219f of nomadism, 368
Early Iron Age, 223 enclosed nomadism, 131, ethnohistoric records, 284–285
earthworks, 297 167–168 Eurasian Steppe, 230, 247,
East Africa, 270 endotribal marriages, 83n6 377f
Eastern Desert Enkidu, 48 Euro-American settlers, 280
Ababda nomads in, 509 environment, 176, 181, 204, Europe and the Peoples Without
historical record of, 252 370, 376, 380, 381, 382f, History (Wolf), 250
map of, 417f, 510f 447, 564 excavation trenches, 239f
596    I n d e x
experimental archaeology, 14 dog burials, 297 habitation, 54f

extortion, 562 ethnicity expressed habitus, 346
through, 349 Hadendowa, 441, 443, 444,
farmers, 273 among Nubians, 358–359 452, 454f, 460–461
farming, 118 stone cist burial, 383f hair tent-dwelling, 452–453
farming societies, 116 in Tombos, Egypt, 354 hairstyles, 528
farming villages, 85 tombs, 531, 536, 537f Hakalan, Lurestan, 93
Fars, 82 fur trade, 280, 284, 294–295 Halaf tradition, 52, 53f, 64f
Fattovich, R., 442 The Hamad, 29
Feast of the Dead, 296 gaddah (communal bowl), 527 hare, 29
fertility rates, 567 Gamshadai tribe, 369–370 The Harra, 29
“fierce horse riding warrior” Gash Delta Group, 442 hazard rate, 565
confederacies, 329 Gashun-Sala headrest, 521f
figurines, 353f amaranth seeds found in, Henry, Alexander, 284
fire clouds, 426 241f herd animal domestication,
fire management, 503. See also archaeobotanic 116, 119–120, 271
burning determinations in, 240 herd animals, 119, 566
firing, 424, 426, 430f archaeological record of, herd management, 495, 561
fishing, 287, 296 236 herd size, 570
Fistikli Hoyuk archaeozoological identifi herders, 496–498, 503, 562
calcareous plates in, 59–60, cations in, 239–240 herder-gatherers, 8, 14, 121,
60f, 60t bone tool in, 237f 170, 224, 269
case study of, 52–65 ceramics in, 236, 239f herding, 118, 171, 274
focal sites around, 63 chronology of, 243t Herodotus, 456, 493
grain cultivation in, 58 death of animals at, 244t Herzog, R., 257
habitation by phase in, 54f description of, 233
hess al-mihbas (mortar and
jetons in, 58 excavation trenches in, 239f
pestle), 470
multisited community map of, 232f
hierarchical society, 52, 339,
of, 44 paleosoil analysis of, 242f,
340
phases/dates of, 55t 243
hierarchy, 337
sherd jetons in, 60t pastoral nomadism from,
high density living, 543
sickle blades in, 60t 246
hippopatami, 259
site abandonment in, 57 potsherds in, 237f, 238f
historical record
stone tools in, 58 spindle whorl in, 237f
archaeological attention
stratigraphy of, 55 Gaza Ware, 414–415, 472
given to sites with, 474
structures in use at, 57t gazelle, 26, 29, 34, 126
fixed housing, 550 gender priorities, 529f archaeological record and,
flint knapping, 6 genealogy, 538 250
focal sites, 56, 63 genos/ethnos, 344–345 in BCE, second
foodstuffs, 471f geoglyphs, 13 millennium, 400
foodways, 349, 360, 527 geographic variability, 367 of Blemmyes, 512
forage management, 497 Gilgamesh, 48 of Iron Age, 337
forage species, 497f Glatzer, B., 565 of the Khoekhoen, 267
Forge, Anthony, 546 goats, 27, 120, 205, 444 military tents in, 448
four room house, 476f Gordon, Robert, 264 Holocene period, 204, 491
fragmentation, 145–146, 151 grain cultivation, 58 holy men, 531–532
frontier, 272 grasses, 100 horizontal integration, 544
fuel, 433n10 grazing resources, 496, 534 horse riding
functionalization, 144, 151 The Great Kalahari Debate, dependence on, 202
funerary feast, 200 14, 274 emergence of, 205, 225
funerary practice, 356f Great Lakes region, 286f, 298 images in Upper Tsagaan
burial cult, 338, 340 Great Lakes region (United Gol Complex, 216
burial grounds, 382–384, States), 280, 281f, 284f mobility from, 201
429 Great Sand Sea, 490 horses, 92, 124, 213f
burial mounds, 115, 339 Greek writers, 255 horticulturalists, 545
burial of, 356f Greenland, 175f house building
burial pits, 97 grog, 432 carpets for, 523
collective burial, 296 guns, 126 with forked acacia
communal burial sites, 94 branches, 524f
I n d e x      597
inner/outer walls, 524f, mobile pastoralism in, 102 archaeological record of,
525f pottery in, 102 270
mats for, 526f Tall-e Bakun, 79 family, 266f
by women, 522–523 Tell Abada, 79 historical record of, 267
human agents, 566 Tell Madhhur, 79 Jourdain on, 265–266
human ecodynamics, 375 Tepe Gawra, 79 languages of, 269
Human Relations Area Files, Iranian Cultural Heritage and material culture of, 267
46 Tourism Organization questions about, 275
hunter-fishers, 373 (ICHTO), 95 women/colonists, 268f
hunter-gatherers, 170, 282 Iron Age, 202, 211 Khuzestan, 99, 99f
Ababda nomads as, 516 agropastoralism in, 330 kibitka (braided mats), 245, 247
Chippewa as, 289 farmers, 273 kiln, 426
dwellings, 445 four room house, 476f kinship, 300, 544–545, 552
herders evolution from, 14 historical record of, 337 Kirghiz, 371
image in Upper Tsagaan social organization of, 341 Koksu River Valley
Gol Complex, 216f Iron Age settlements, 339 archaeozoological
images in Tsagaan Salaa/ Irons, W., 369 identifications in,
Baga Oigor, 222f Irving, W.N., 177 388–389
in Libyan Desert, 491 Islam, 130 environmental conditions
logistical mobility in, 7 Islands of the Blessed, 489 in, 382f
modern, 119 Israelite four room house, 475 ephemeral settlements in,
Natufian, 122 Isumkheb, 449 386, 387f
Ottawa as, 294 land use in, 389
pastoral nomads and, 6t Jebel Naja, 32 map of, 379f
in PPNB, 26 Joint Mongolian/American/ pastoralists in, 388
residential mobility in, 7 rock art in, 378
Russian Project, 208
sedentarization of, 549
Jones, S., 346 semi-subterranean houses
settled farmers and, 15
Jordan, 30f, 30–32, 35t in, 384
sheep/goats replacing
Jordanian Steppe, 31t small camps in, 384–386
economy of, 120
Jourdain, John, 265–266 vertical transhumance in,
in South Africa, 269, 273
Ju/’hoansi, 274 380
tethered mobility in, 7
juncus acutis, 497 water sources in, 381
hunter-herders, 36, 37
Juntunen phase, 295, 295f, Kopytoff, I, 52
hunting, 288, 501, 535
296, 298 Kroeber, Alfred, 264
Huron, 294
Juntunen site, 291 KS-1626, 96f–97f
Hymn to the Aton
kurgan, 115, 116, 230, 329,
(Akhenaton), 345
kaburri (bread), 527 334, 335f
ICHTO. See Iranian Cultural Kalmyk Steppe, 231 Kurzban, R., 347
Heritage and Tourism Kalmyks, 247
Organization karama (festive gathering), 527, labor rates, 567, 574
Inca period, 166 530–531, 534 land use, 389, 390f
Indian Culture and European Kazakh herder, 335f landscape
Trade Goods (Quimby), 284 Kazakhstan, 377f Ababda impact on, 535
Indicopleustes, Cosmas, 253 archaeological sites in, 335 of Amorites, 146
industrial nomadism, 153 “fierce horse riding analysis of, revealing
industrialization, 152 warrior” confederacies patterns, 11–12
inheritance, 568–569 in, 329 archaeologists learning
Insoll, T., 469 map of, 332f about, 13, 15, 374
The Instruction of Ani (New paleoclimatologists in, 380 ephemeral sites and, 15
Kingdom), 348 settlement in, 387f pastoral strategies and, 376
interregional marriage, 103 study area in, 331 variability and, 375
intracommunity agreement, 56 Keimer, L., 510–511, 538–539 landscape learning, 176
Inuit, 177, 187 Khaldun, Ibn, 559–560 land-use patterns, 194
Iran, 78 Khamseh Confederacy of Late Bronze Age, 215, 217f,
Central Plateau, 101 Basseri, 334 223
Kheit Kasim, 79 Khazanov, A., 561 Late Neolithic period, 25–26,
map of, 80f, 81f Kheit Kasim, Iran, 79 28, 33
Margueron, 79 The Khoekhoen, 264 Late Susiana 1 phase, 86, 86f
598    I n d e x
Late Woodland people, 281, sherds found in, 235f McGlade, J., 375
283 maple sap, 287 medicine, 531–532
adaptive strategies of, 286 maple sugaring, 292 Megabari, 254
archaeological record of, maps men, 531f
285 Altai Mountains, 203f Mesopotamia, 80f
cooperative buffering in, Amadjuak Lake, 190f Mesopotamian scribes, 49
293, 300 Anatolia, 398f meta-perception, 147n6, 152
earthworks of, 297 Baffin Island, 185f, 188 mgongo nuts, 122
ethnohistoric record of, Burwash Bay, 188f Michigan (United States), 280,
284–285 California, 308f 283
map of sites of, 290f, 295f Canada border, 281f Middle Assyrian period, 407
O’Neil site of, 298 Canadian Arctic, 175f, 185f Middle East, 330–331
Latin writers, 255 Caspian Steppe, 231f Middle Susiana phase, 104
Lattimore, O., 560 DMAP sites, 379f migration, 504
leather hard stage, 424 Eastern Desert, 417f, 510f camel herding and, 502
leather tanning, 523 Egypt, 251f, 421f, 490f destinations in Libyan
leather workers, 451 Egypt and Sudan, 417f Desert, 496f
legumes, 100 Eurasian Steppe, 377f for grazing resources, 534
Levant, 38, 465 Gashun-Sala, 232f to oases, 495
Libyan Desert of Egypt, 487, Great Lakes region, 284f of sheep/goats, 495n8
491f, 496f, 503 Greenland, 175f military potentials, 91–93,
limitational knowledge, 181, Halaf tradition, 53f 561, 563
184 Iran, 80f, 81f milking basket, 522f
lithic analysis, 307 Jordan, 30f Miller, Naomi, 100
lithic raw materials, 182–183, Juntunen phase, 295f Mingo Lake, 189f
186 Kazakhstan, 332f
Mingo River, 190f
lithic remains, 176, 177f, 189 Koksu River Valley, 379f
Minusinsk Basin, 205
litters, 458 KS-1626, 96f–97f
mnemonic devices, 59
livestock management, 160, Late Woodland people
mobile groups
164, 168, 169, 442 sites, 290f, 295f
analogical reasoning failure
livestock production, 488 Mackinac phase, 290f
with, 49
llama, 164 Manych, 232f
assumptions about, in
llameros, 171n4 Mesopotamia, 80f
research, 50
locational knowledge, 181, 184 Mingo Lake, 189f
material culture and, 46–47
Lockhart River settlement, Mingo River, 190f
pottery of, inferior to
551f, 552f Native Americans, 284f
logistic mobility, 7, 282, 294, Nevada, 308f sedentary, 414
299, 300 Southern California, 421f territorial claims of, 514
low population densities, 183 Sudan, 251f traditional and current
luxury goods, 338 Talgar alluvial fan, 336f views on, 44t
Lynch, Thomas, 161 Turkey, 53f mobile lifestyle, 543
world, 3f mobile pastoralism, 79
al-mabram (women’s section), el -Maqrizi, 450 agricultural activities in,
468f marginal rate of substitution 89, 91
Mackinac phase, 289–292, 290f (MRS), 569 architectural plans and, 475
macro-pastoral orbit, 370–371 Margueron, Iran, 79 in Behbahan, 84, 84f
macro-region, 147, 148 Mari, 142 in Bronze Age, 379
maize, 289–290 marriage alliances, 85 communities of, 87
Makran Baluch, 370–371 Mashkour, Marjan, 99 in Iran, 102
mammoths, 206, 218f mat house (badaigaw), 447 large population centers
Manych mat house (beit al-burush), 522 and, 103
archaeological excavations mat tent-dwelling, 447, military potentials in,
of, 233, 234f 454–455 91–93
chronology of, 243t Mata Ortiz (Chihuahua, nomadism and, 78n2
copper found in, 234f, 235f Mexico), 420, 422 politics of, 91–93, 105
map of, 232f material culture, 2, 11–12, sheep/goats and, 78, 101
paleosoil in, 241, 242f, 243 154, 267 in Susiana, 94–95
pastoral nomadism from, mating networks, 183, 187 urban centers and, 104, 105
246 matrilocal social structure, 522 variability in, 392
I n d e x      599
in Zagros Mountains, trade with urbanites, 557 of Libyan Desert of Egypt,

81, 83 nomad territories, 127, 152 491f
mobility patterns, 5f, 9t nomadic economy, 493 migration to, 495
Moghra, 490 nomadic pastoralism, 28. See resource use of, 487, 504
Mohamed, prophet, 512 also pastoral nomadism territorial claims in, 499f
Mojave Desert, 308 archaeologist’s dilemma occupational redundancy, 316,
Mongolian Altai. See Altai with, 366 319
Mountains ethnographic studies O’Connell, James, 544
Mosquito Ridge, 191 defining, 366 Old World archaeology, 2
movement rules, 568 geographic variability in, O’Neil site, 298
MRS. See marginal rate of 367 organizational breakdown, 547
substitution in PPNC, 28 ostrakon (writing surface), 432
mudbaru, 399–400 salient factors of, 366 ostrich eggshell beads, 272
mud-brick settlements, 498 Salzman on, 333 Ottawa, 283–294
multi-resource nomadism, 8, Siwan campsites and, 502 Ottoman period, 474, 477
128, 403 temporal variability in, 367 ovens, 57t
of Ababda, 511–512 tribal organizations in, 369 overnight bag, 517
in Bakhtiyaari, 87 typological categorizations ovicaprids, 493–494, 497f
in Gamshadai tribe, of, 372 ovicaprine bones, 29
369–370 variability as prime aspect Owens River, California, 314f
in North Arabian steppe, of, 367, 368, 372 Owens Valley, California
36 nomadism. See also pastoral desert shrub landscape in,
in Yaramadzai tribe, nomadism 315f
369–370 archaeology’s limitations pottery distribution in, 313f
mummification, 354 around, 405 pottery from, 312f
Murra, John, 161 enclosed, 80 seed density in, 320f
Murray, G.W., 513 ethnography of, 368 ownership, 514, 516
mushrooms, 6 material culture and, 488
music/dance, 533f mobile pastoralism and, packsaddle, 459–460
Musil, A., 469 78n2 Paiute, 103
musk ox hunting, 178 pastoralism and, 368 palanquins, 458, 459f
nomadization, 143 paleoclimatologists, 380
Nabateans, 124 Nomads in Archaeology (Cribb), Paleo-Eskimos, 193t
Nama, 267–268 6, 548 colonization path of, 178,
National Science Foundation, North Arabian steppe, 36–38 184
95 North Asia, 204 ephemeral sites of, 189
Native Americans, 252, North Jazira, 406 flake-to-tool ratios of, 192t
272, 284f, 296, 297. See Nubian(s), 255 landscape learning of, 180,
also Chippewa; Ottawa; burial of, 356f 181
Paiute; Paleo-Eskimos; cattle and, 353 limitational knowledge
Potawatomi; Shoshone cuisine of, role in ethnicity, of, 184
Negev, Israel, 125, 126f, 352 lithic raw materials of, 182,
127–128 funerary practice among, 183, 186, 191
Neilsen, A.E., 171–172 358–359 lithic remains of, 176, 189
Neo-Assyrian period, 399 pottery of, 351f locational knowledge of,
Neolithic period, 26t, 33 in Second Intermediate 184
Nevada, 308f Period, 350 low population densities
New World archaeology, 2 warriors, 349f of, 183
Nile Valley, 252, 258–259 nutritional demands, 566 mating networks of, 187
Nobatai, 260 nuts, 288 population of, 180n3
NOMAD, 576. See also agent- route of, 175f
based models (ABMS) oaseans, 498, 500 snow houses of, 179
nomad(s) oases social interaction of, 193,
economic relations with archaeological record of, 194
sedentists, 489 500 subsistence strategies of,
ephemeral sites of, 1 flora of, 494 181
etymology of word, 3 herders occupation of, 503 tent-ring structures of, 187
sedentarization of, 405, hunting in, 501 tool stone resources of, 185
475, 549 juncus acutis in, 497 toolmakers, 182
600    I n d e x
Tungatsivvik site, 177f Lattimore on, 560 distribution of, 317t

winter hibernation of, 179 nomadism and, 368 dot-motif, 86, 86f
Paleo-Indians, 174, 178, 183 strategies of, 376 in early Bronze Age, 128
Paleolithic Period, 218 pastoralist(s) Eerkens on, 413
paleosoil, 241, 242f, 243 environment influencing, in Egypt, 431f
Palmyra, 154 376, 381 experimental production
para-urban, 149 in Koksu River Valley, 388 of, 419–420, 422–432
Parchineh, Lurestan, 93 military superiority of, 563 interregional marriage
Pastner, S., 370 variability in, from Bronze spreading, 103
pastoral economy, 224–225, Age, 376 in Iran, 102
247 pasture land, 389, 390f Late Susiana 1 phase, 86,
pastoral ideology, 261 pasture productivity, 565 86f
based on herd ownership, pen-and-attached-room sites, Mackinac phase, in U.S.,
119 121–122 293
defining elements, 130 perceptual geography, 146n5 manufacture of, by mobile
hunters v. herders, 120–121 peripheral nomadism, 7 groups, 414
pastoral nomad(s) Peru, 165 Nubian, 351f
agriculture pursued by, 407 petroglyphs, 13, 155, 203 in Owens Valley,
classic stereotype of, 557 pharaoh, 10 California, 312f, 313f
economic structures with phytoliths, 58t, 61 Paiute, 103
settlements, 119, 123 pilgrimage, 11 repair of, 432
hunter-gatherers and, 6t Pinon-Juniper landscape, 315f sedentism and, 309
kurgan builders as, 116 plant species, 231, 245 Shoshone, 103
Middle Assyrian kings Pleistocene, 212 transhumance and, 309
influencing, 402 politics, 91–93, 105, 117, 148 in Western Great Basin,
military potentials of, 561 polygamy, 273
307, 311
in Negev, 125 population
women and, 85n8
population of, 129 in ABM simulation, 571
in Zagros Mountains, 93
resilience of, 6t, 115 aboriginal, 544, 550
Pottery Neolithic period. See
tribal identity retained of Bedouins, 129
Late Neolithic period
by, 404 density, 183, 318
pottery-making
pastoral nomadism, 7 density in Western Great
clay for, 423f, 424
bias against, in historical Basin, 311t
history of, in Western
data, 117 low densities, 183
Great Basin, 312
in Central Andes, 169 of Paleo-Eskimos, 180n3
population density and, 318
Christianity and, 130 of pastoral nomads, 129
degrees of, 118 of semi-nomadic residential mobility and,
in early Bronze Age, 123 pastoralists in Central 311
ethnicity and, 357 Andes, 163 pounders, 98, 98f
from Gashun-Sala, 246 portable house, 522 poverty, 562–563
Islam and, 130 postmolds, 292n2 PPNA. See Pre-Pottery
in Libyan Desert of Egypt, Post-Neolithic tents, 466 Neolithic A
487 Potawatomi, 283 PPNB. See Pre-Pottery
from Manych, 246 Potawatomi pattern, 301 Neolithic B
in Middle East, 330 potsherds, 237f, 238f, 416. See PPNC. See Pre-Pottery
sedentary agriculture also sherds Neolithic C
systems disinguished pottery. See also ceramic practical reason, 51n5
from, 118 technologies; Eastern precipitation, 162
Shaw on, 256 Desert Ware; Gaza Ware; prehistoric symbol systems,
variability in, 371, 374 potsherds 121
Wilkinson on challenges of in Askut, 351f Pre-Pottery Neolithic A
archaeology on, 397 Bakun style, 85 (PPNA), 25
pastoral nomad/sedentary in BCE, second Pre-Pottery Neolithic B
agriculturalist dichotomy, millennium, 406 (PPNB), 25–28
557–558 Bronze Age, 385 Pre-Pottery Neolithic C
pastoralism, 373, 407. See also caching of, 313–314, 316 (PPNC), 28
nomadic pastoralism in California, 431f pricing, 569, 574–575
Bronze Age, 377, 378 in Central Plateau, 101, production, demand-driven,
Herodotus on, 493 104 419
I n d e x      601
production, household, 245, in late Bronze Age, 223 of Blemmyes, 258

419, 432–433, 435 paintings in Central drive towards, in historic
production sequence, 152 Sahara, 451 literature, 49–50
productive capacity, 389 in Paleolithic Period, 218 economic relations of, 123
project, 62 pastoral economy pottery superior to mobile
protein, 574–575 representing, 224–225 groups, 414
protension, 62 as sign of ownership, 391 PPNB villages largest, 27
Ptolemaic explorers, 257 in Talapty, 389 sedentary culture, 560
public space, 553–554 in Tsagaan Salaa/Baga sedentary workers, 45
Oigor, 211, 218–223 sedentism, 10, 118
Qa’ Al-Ghirqa, 34 in Turkic Period, 218 Eerkens on, 413
Al Qara, 490, 498 Rockman, M., 176, 181, 199 from hunter-gatherer, 14
Qashqaaii territory, 83, 83n6, Roman skin tents, 449–450 pottery and, 309
84, 84f, 85, 87 Room To Move (O’Connell), 544 through poverty, 562
Qasr Burqu, 33 Roux, G., 397 Salzman on, 333
Qattara Depression, 490 Rowe, John, 161 sedentarization and, 90
Quimby, G.I., 284 Rowton, M.B., 131, 139 sedentists, 489
Royal Commission of Inquiry seed density, 320f
racial stereotyping, 347, 348 into Aboriginal Deaths in semi-nomadic pastoralism, 7,
radial pattern, 551 Custody (RCIADC), 554 161–162, 200n1, 202, 205
radiocarbon dating, 31t, 193t Royal Netherlands Embassy, in Altai Mountains, 223
raiding, 85, 562, 570, 572, 510, 511 in Central Andes, 160, 169
573f, 575 precipitation driving, 162
raiding mortality, 567, 571 saddle shape mill, 98, 98f temperature as variable
rainfall, 490–491, 514, 534 saggar, 426, 427f, 428 in, 162
Ramesses III, 347f Saka (Scythians), 338 semi-subterranean house(s),
RCIADC. See Royal Salzman, P.C., 333–334, 403 384, 385f, 388
Commission of Inquiry sam’al, 147 sequential organization, 55, 61
into Aboriginal Deaths in Santa Clara, New Mexico, 422 settled farmers, 15
Custody Saracens, 258–259 settled groups, 5, 10
Red Sea, 450 seasonal encampments, 26 settlement density, 548
reed huts, 456 seasonal migration, 88 settlement strategies, 307
reeds/rushes, 496–497 seasonal mobility, 572f Shahrani, M.N.M., 371
religious practice, 349 seasonality, 244t, 494, 565, 573 Shalmaneser I, 402
residential mobility, 282 seasoning, 429 Shaw, B., 256
in Canadian Biotic Second Intermediate Period, Shechemites, 557
Province, 299 350 sheep/goats, 26–27, 120, 201,
in hunter-gatherers, 7 security issues, 478 205, 269. See also ovicaprids
in Juntunen phase, 298 sedentarization Hadendowa raising, 444
in Ottawa, 294 of Aramaeans, 398, 401 in Late Neolothic period,
pottery-making and, 311 Barth on, 570 28
in Western Great Basin, of Bedouin, 466 in Middle Susiana phase,
310 of Besseri, 404 104
resource management, 547 of Blemmyes, 260 migration of, 495n8
rhinoceros, 212, 213f cultural results of, 560 mobile pastoralism and,
ringed seals, 186, 192 of herders, 562 78, 101
rish sefid (white beard), 89 of hunter-gatherers, 549 remains of, in Jordanian
rite of passage, 429 in Middle Assyrian period, steppe sites, 35t
ritual structures, 1 407 Sheikh Shazli, 531
Robertshaw, P.T., 267–268 in the Negev, 127 sherd jetons, 58–59, 59f
rock art, 203 of nomadic pastoralists, 554 sherds, 235f, 268
in Altai Mountains, 206, of nomads, 405, 475, 549 shield, 530
208 political conditions needed al-shigg (men’s section), 468f
of Aral Tolgoi, 212 for, 408 Shoshone, 103
in Bronze Age, 207 Salzman on, 403 Siamun, 355f
complexes of, 208 sedentism and, 90 sickle blades, 59, 60t
dating of, 206 sedentary agriculture systems, Sinai peninsula, 465
in Early Iron Age, 223 118 Sinuhe, 449
in Koksu River Valley, 378 sedentary communities site abandonment, 51–52, 57
602    I n d e x
site formation, 34, 52 in Carolinian Biotic mat house (badaigaw), 447,

Siwa, 490–491, 498 Province, 286–288 454–455
Siwan campsites, 502 of Chippewa, 292 Roman skin tents, 449–450
slashes, 164 as factor in mobility, 47 tensile structure, 446
slaughtered animals, 12 of herders, 497–498 women as builders of, 451,
slaughtering, 6 livestock management as, 453
slipping, 424 442 woven hair construction,
small camps, 384–386, 388 in Mackinac phase, 291 452–453
Smith, A. D., 344 nomad/urbanite differences tent-ring structures, 187
snow geese, 191 in, 559–560 Tepe Gawra, Iran, 79
snow houses, 179 of Ottawa, 293 territorial claims, 499f, 503,
snow leopards, 209 of Paleo-Eskimos, 181 514
social interaction, 193, 194, substitution, 569 territorial solidarity, 144, 145,
555 Sudan, 251f, 417f, 441 151
social knowledge, 181 Sudanese Nubia, 355f territoriality, 122
social organization, 513–514, Susiana, 94–95 tethered mobility, 7
543, 547 sustainable tree use (ewak), 535 tethered nomadism, 7, 167
social relationships, 546 Swarm, 564 tholoi, 57, 57t, 63
social rules, 568 sword dance, 532f Thutmose I, Pharaoh, 354
Society for American swords, 530 Tiglath-Pileser I, king,
Archaeology, 2 Syro-Anatolia, 399, 408 399–400
solstice shrines, 120 Syro-Mesopotamia Tjemehu, 492
Amorites in, 141 tobacco, 471–472
sorghum vulgare, 444
tombs, 531, 536, 537f
South Africa archaeological record in,
tomb fields, 122
ceramics in, 273 154
Tombos, Egypt, 343, 354, 355f,
domesticated animals in, material culture of, 154
356f
271f petroglyphs in, 155
tool stone resources, 184–185,
herders in, 269, 274 written sources in, 154
194
hunter-gatherers in, 269,
toolmakers, 182
273 tabla (drum), 533
tourism, 509
sherds in, 268 Talapty (Kazakhstan), 389, 391
Toyota Hilux, 535
Southern California, 421f Talgar alluvial fan, 336f, 338
trade, 562
spindle whorl, 237f Talgar region, 336, 339
in Eastern Desert, 434
Standard Cross-Cultural Tall-e Bakun, Iran, 79 as motivation for mobility,
Sample, 46 tambura (stringed instrument), 11
state ideology, 345 533 between nomads and
stationary periods, 5–6 taphonomy, 12 urbanites, 557
stem-dwelling (euphorbia), Tapper, R.L., 547 raiding and, 572, 573f, 574f
457f taxation, 153 rules, 569
steppe, 26, 142, 148 Tell Abada, Iran, 79 West African coast to Red
steppe kingdom, 156n11 Tell Madhhur, Iran, 79 Sea route, 450
steppe pastoralists, 155 temper, 420 transhumance, 7, 8, 65, 202
steppic sites, 31, 33, 35 temperature, 162 Eerkens on, 413
stone boiling, 320 temporal variability, 367 Lynch on, 161
stone cist burial, 383f tensile structure, 446 pottery and, 309
stone tools, 58, 298 tents, 125, 444 vertical, in Koksu River
storage structures, 57t tent camps, 55n6, 473, 476 Valley, 380
Straits of Mackinac, 281, 292 tent dwellers, 47 tribal identity, 404
stratigraphy tent-dwelling (maskan), 445. tribal organizations, 120, 122,
analysis of, 52 See also Bedouin black tent; 369
of Fistikli Hoyuk, 55 dwellings tribal people, 554
lack of, 13 armature, 446 tribal relationships, 37, 513
pitfalls of, 50–51 black tent in North Africa, tribal route (Il raah), 88
Walker on, 51n5 453 tribal social organization, 119,
subsistence strategies, 282 contemporary use, 448 544
camels and, 493 Hadendowa Badaigaw, 454f tribes, 120n2, 148, 150–151,
in Canadian Biotic of Isimkheb, 449 369–370, 467f, 513f, 515f
Province, 286–287 leather, 448–449 Trogodytes, 252, 254, 257
I n d e x      603
trucks, 126 variability wet-season sites, 163

truffles, 153n9 in Bronze age pastoralists, wheat, 58
Tsagaan Salaa River, 210f, 211 376 wheeled vehicles, 220, 221f,
Tsagaan Salaa/Baga Oigor, 210 in environment, 380 225
aurochs image in, 219f ethnographies supporting, White Mountains, 315f, 316f
caravan images in, 220, in nomadic pastoralism, wild date palm oases, 498, 501
221f 372 Wilkinson, T.J., 397, 406
elk image in, 219f landscape and, 375 winter hibernation, 179
hunter scene in, 222f in mobile pastoralism, 392 winter sites, 292
rock art in, 211, 218–223 in pastoral nomadism, Wolf, Eric, 250
wheeled vehicles images in, 367–368, 371–372, 374 women, 268f, 468–469
220, 221f temporal/geographic, 367 al-mabram (women’s
Tseganka 8, 336 Verity, David, 420 section), 468f
Tuaregs (South Africa), 92, vertical integration, 544–545 dance of, 533f
451, 455 vessels, breaking of, 429 hair tent-dwelling built by,
Tu-Bedawie language, 253 451, 453
Tungatsivvik site, 177f Wadi al-Gemal National Park, house building by, 522–523
Turkey, 53f 511 karama and, 534
Turkic Period, 218 Wadi Fatima, Saudi Arabia, leather tanning by, 523
Tutankhamun, Pharaoh, 349f 478f pottery and, 85n8
Tuzusai, 336 Wadi Jilat, 26, 32, 33 social customs of, in
twentieth century, 474 wadi systems, 26 Ababda life, 516
typological categorizations, 372 Walker, W.H., 51n5 tent-dwellings built by,
Wanka peoples, 163 451
University of Pennsylvania, Wari Empire, 166 writing, 145
100 wasm, 514, 516 written sources, 48, 154
Upper Tsagaan, 209–210 weaponry, 207
Upper Tsagaan Gol Complex Wereldmuseum Rotterdam, yak, 225
aurochs image in, 216f 510 yamina, 147
elk image in, 215f Weren, Pyramid of, 355f Yanamamo, 545
hunters image in, 216f, West African coast, 450 Yaramadzai tribe, 369–370
217f Western Great Basin Yomut Turkmen, 369
Late Bronze Age image baskets in, 313 yurt, 387f
in, 217f caching in, 317, 319
rock art in, 214–218 environmental zones in, Zagros Mountains, 81, 83,
Ur III period, 142 317t 92–93
urban centers, 104, 105, 474 population density in, 311t Zapotec, 346
urban perspective, 48 pottery in, 307, 311–312 Zimri Lin, 131
urbanites, 557 residential mobility in, 310 zor, 142

Wendrich y Barnard 2008-The Archaeology of Mobility

Uploaded by

Copyright:

Available Formats

Wendrich y Barnard 2008-The Archaeology of Mobility

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Wendrich y Barnard 2008-The Archaeology of Mobility

Uploaded by

Copyright:

Available Formats

The Archaeology of Mobility

Cotsen Institute of Archaeology

The Cotsen Institute of Archaeology at UCLA

Editorial Board of the Cotsen Institute of Archaeology

Editorial Advisory Board

Library of Congress Cataloging-in-Publication Data

Copyright © 2008 Regents of the University of California

1 The Archaeology of Mobility:

Part I: The Past at Present

4 Archaeology and the Question of Mobile

8 Colonization, Structured Landscapes and Seasonal Mobility:

12 Is the Absence of Evidence, Evidence of Absence? Problems in

Part II: The Present and the Future

17 Variability and Dynamic Landscapes of Mobile

Coffee, Tobacco, and Pottery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470

Public Space and Conflict . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553

List of Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584

List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587

List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591

Willeke Wendrich and Hans Barnard1

T hroughout history, groups and individuals have traveled the face of

stored or cached on purpose (Alizadeh, Chapter 4; Milne, Chapter 8; Eerkens,

Definitions: Moment, Movement and Motivation

Chapter/Author(s) Geographical Subject matter Time period

important aspect is that of a (residentially) mobile existence. This meaning

equally undefined. Many activities, varying from the slaughtering of animals

Hunter-gatherers Pastoral nomads

Ter minology Related to Hunter-Gatherers

Ter minology Related to Pastor al Nomads

snow forces the herds to low-lying meadows in winter. Horizontal transhumance

Table 1.3. Factors and Range of Mobility:

Length of period of movement Total no. of nights

Length of stay in each location Total no. of nights

Stability of mobility Repetition over multiple years, decades or centuries

Variation of mobility over time Inventory of multiple years, decades, centuries

Movement Range of movement Number of km per period of movement

Pattern of movement Fixed, flexible, random (specified with map)

Motivation Purpose of movement Hunting, gathering, herding (specified with

for each population separately according to very specific questions (Khazanov

combination of factors, partly dependent on the context and audience, and in

Relations Between Mobile and Settled Groups

As several of the contributions to this volume point out, there is the

Material Culture, Landscape and Fieldwork

mobile people are incapable of producing ceramics because of a lack of infra-

political circumstances. Central to understanding the archaeology of mobility

Acknowledgements and Dedication

Bar-Yosef, O. and A. Khazanov (eds.)

Period Age BP (uncalibrated) Age BCE (calibrated) Millennium BCE

the Mediterranean coast (Gopher 1993; Kuijt and Goring-Morris 2002:414). By

The End of the PPNB

of PPNB social structure has been addressed from a number of differing

Models of Steppic Economic Str ategies

the steppe where, in order to survive on reduced resources, they responded

Evidence from the Harr a and the Hamad

a similar interpretation, with a hunter-gatherer phase in the PPNB, followed

was a significant part of the cultural package. The chipped-stone repertoire is

associated with corresponding changes Dhuweila 1 5 2198

in economic strategies. Equally clearly, Mid/Late PPNB

recent sheep-goat pastoralists, indicating that herding needs strongly governed

economic success in a world of scattered and varied resources is a key skill in

17 Variability and Dynamic Landscapes of Mobile