COLLOQUIA PONTICA VOLUME 13
BEYOND THE STEPPE AND THE SOWN
COLLOQUIA PONTICA Series on the Archaeology and Ancient History of the Black Sea Area Monograph Supplement of Ancient West & East Series Editor
GOCHA R. TSETSKHLADZE (Australia) Editorial Board
A. Avram (Romania/France), Sir John Boardman (UK), O. Doonan (USA), J.F. Hargrave (UK), J. Hind (UK), M. Kazanski (France), A.V. Podossinov (Russia) Advisory Board B. d’Agostino (Italy), P. Alexandrescu (Romania), S. Atasoy (Turkey), J.G. de Boer (The Netherlands), J. Bouzek (Czech Rep.), S. Burstein (USA), J. Carter (USA), A. Domínguez (Spain), C. Doumas (Greece), J. Fossey (Canada), I. Gagoshidze (Georgia), M. Kerschner (Austria/Germany), J.-P. Morel (France), A. Rathje (Denmark), A. Sagona (Australia), S. Saprykin (Russia), T. Scholl (Poland), N. Theodossiev (Bulgaria), M.A. Tiverios (Greece), A. Wasowicz (Poland)
BEYOND THE STEPPE AND THE SOWN Proceedings of the 2002 University of Chicago Conference on Eurasian Archaeology EDITED BY
D.L. PETERSON L.M. POPOVA A.T. SMITH
BRILL LEIDEN • BOSTON 2006
All correspondence for the Colloquia Pontica series should be addressed to: Aquisitions Editor/Classical Studies or Gocha R. Tsetskhladze Brill Academic Publishers Centre for Classics and Archaeology Plantijnstraat 2 The University of Melbourne P.O. Box 9000 Victoria 3010 2300 PA Leiden Australia The Netherlands Tel: +61 3 83445565 Fax: +31 (0)71 5317532 Fax: +61 3 83444161 E-Mail:
[email protected] E-Mail:
[email protected] Illustration on the cover: Goblet from Trialeti kurgan 5 (background; after Miron and Orthmann 1995, fig. 10); detail of the Karashamb goblet (The Institute of Archaeology and Ethnography, Yerevan, Armenia). See for details about these goblets pages 251ff. of this book.
This book is printed on acid-free paper. Library of Congress Cataloging-in-Publication Data A C.I.P. record for this book is available from the Library of Congress.
ISSN 1389-8477 ISBN 10: 90 04 14610 5 ISBN 13: 978 90 04 14610 5 © Copyright 2006 by Koninklijke Brill NV, Leiden, The Netherlands Koninklijke Brill NV incorporates the imprints Brill Academic Publishers, Martinus Nijhoff Publishers and VSP. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Brill provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910 Danvers MA 01923, USA. Fees are subject to change. printed in the netherlands
CONTENTS
Preface .......................................................................................................... GOCHA R. TSETSKHLADZE
xi
Introduction .................................................................................................. DAVID L. PETERSON, LAURA M. POPOVA AND ADAM T. SMITH
xiii
List of Abbreviations ..................................................................................
xv
List of Figures ............................................................................................ xvii
I. LOCAL AND GLOBAL PERSPECTIVES
Chapter 1. The Early Integration of the Eurasian Steppes with the Ancient Near East: Movements and Transformations in the Caucasus and Central Asia .......................................................... PHILIP L. KOHL
3
Chapter 2. Three Deadly Sins in Steppe Archaeology: Culture, Migration and Aryans ............................................................................ DAVID W. ANTHONY
40
Chapter 3. General Migration Processes in the Aral Sea Area in the Early Iron Age ............................................................................ LEONID T. YABLONSKY
63
II. REGIONAL STUDIES
Chapter 4. Community Organisation among Copper Age Sedentary Horse Pastoralists of Kazakhstan .......................................................... SANDRA OLSEN, BRUCE BRADLEY, DAVID MAKI AND ALAN OUTRAM
89
vi
CONTENTS
Chapter 5. Perspectives on Chalcolithic Turkmenistan: The Global and the Local in the Kopet Dag ............................................................ 112 LAUREN ZYCH Chapter 6. The Dzhungar Mountains Archaeology Project: Reconstructing Bronze Age Life in the Mountains of Eastern Kazakhstan ................................................................................ 122 MICHAEL D. FRACHETTI Chapter 7. The Kura-Araxes ‘Culture’ in the North-Eastern Caucasus: Problems in its Identification and Chronology .................. 142 RABADAN MAGOMEDOV Chapter 8. The Middle Bronze Age Settlement Pattern of the Eastern Anatolian High Plateau in Light of New Evidence ................ 160 AYNUR ÖZFıRAT Chapter 9. Before Argishti: The Roots of Complex Societies in Caucasia. Notes from the Tsakahovit Plain, Armenia .......................... 172 ADAM T. SMITH Chapter 10. Survey and Settlement in Northern Mongolia: The Structure of Intra-Regional Nomadic Organisation ...................... 183 WILLIAM HONEYCHURCH AND CHUNAG AMARTUVSHIN Chapter 11. Recent Archaeological Research in the Khanuy River Valley, Central Mongolia .............................................................. 202 FRANCIS ALLARD, DIIMAAJAV ERDENEBAATAR AND JEAN-LUC HOULE
III. NEW DIRECTIONS IN THEORY AND PRACTICE
Chapter 12. Between the Steppes and the Sown: Prehistoric Sinop and Inter-regional Interaction along the Black Sea Coast .................. 227 ALEXANDER A. BAUER
CONTENTS
vii
Chapter 13. Over the Mountains and through the Grass: Visual Information as ‘Text’ for the ‘Textless’ ................................................ 247 KAREN S. RUBINSON Chapter 14. Innovation, Change, Continuity: Considering the Agency of Rusa II in the Production of the Imperial Art and Architecture of Urartu in the 7th Century BC .................................... 264 TUªBA TANYERI-ERDEMIR Chapter 15. Sequences of Signs: Eurasian Archaeology from a Perspective of Cultural Semiotics ...................................................... 282 GREGORY E. ARESHIAN Chapter 16. Writing the Landscape: Petroglyphs of Inner Mongolia and Ningxia Province (China) ................................................................ 296 PAOLA DEMATTÈ Chapter 17. Refining the Definition of Technology in the Southern Zone of the Circumpontic Metallurgical Province: Copper Alloys in Armenia during the Early and Middle Bronze Age ........................ 310 LAURA A. TEDESCO Chapter 18. The Samara Bronze Age Metals Project: Investigating Changing Technologies and Transformations of Value in the Western Eurasian Steppes ...................................................................... 322 DAVID L. PETERSON, PAVEL F. KUZNETSOV AND OLEG D. MOCHALOV (with a brief Appendix by AUDREY G. BROWN AND EMMETT BROWN) Chapter 19. Understanding the Productive Economy during the Bronze Age through Archaeometallurgical and Palaeo-environmental Research at Kargaly (Southern Urals, Orenburg, Russia) .................... 343 PEDRO DÍAZ DEL RÍO, PILAR LÓPEZ GARCÍA, JOSE ANTONIO LÓPEZ SÁEZ, M. ISABEL MARTINEZ NAVARRETE, ANGEL L. RODRÍGUEZ ALCALDE, SALVADOR ROVIRA-LLORENS, JUAN M. VICENT GARCÍA AND IGNACIO DE ZAVALA MORENCOS
viii
CONTENTS
Chapter 20. Why Have Siberian Artefacts Been Excavated within Ancient Chinese Dynastic Borders? ...................................................... 358 KATHERYN M. LINDUFF Chapter 21. Sensing the Shadow Empire: New Approaches to the Problem of the Khazar State .................................................................. 371 IRINA R. SHINGIRAY Chapter 22. Recent Xeroradiographic Analysis of Kura-Araxes Ceramics .................................................................................................. 382 MARYFRAN HEINSCH AND PAMELA VANDIVER Chapter 23. The Russian Far East and Cultural Interaction in Prehistoric Eurasia in Light of Ceramic Studies .................................. 395 IRINA S. ZHUSHCHIKHOVSKAYA Chapter 24. Trade Ceramics from Bohai Sites of Russian Primor’e .... 411 EVGENIA I. GELMAN
IV. PALAEO-ECOLOGY AND ENVIRONMENT
Chapter 25. Mediterranean Agriculture in South-western Crimean Palaeo-environments and Early Adaptations ........................................ 425 CARLOS E. CORDOVA AND PAUL H. LEHMAN Chapter 26. Cultural Transformations in the Black Sea Steppe between the Eneolithic and Bronze Age: Migrations or Economic Changes? ................................................................................ 448 YURI RASSAMAKIN Chapter 27. Pastoralism during the Late Bronze Age in Russia: Past Interpretations and New Goals for Future Research .................... 459 LAURA M. POPOVA
CONTENTS
ix
Chapter 28. Early Maritime Adaptation on the Southern Coast of the Far East of Russia in Ancient Times .............................................. 469 ALEXANDER N. POPOV AND DAVID R. YESNER Chapter 29. Pan-Regional Interaction vs Ecologically Induced Change: An Examination of the Rapid Transition from the Late Neolithic Era to that of Palaeometal among Maritime Populations of the Primor’e Region of the Russian Far East ............ 477 JIM CASSIDY AND NINA A. KONONENKO List of Contributors .................................................................................... 493 Index ............................................................................................................ 501
PREFACE
When Colloquia Pontica was established, it was always one of our intentions to publish volumes on Eurasian archaeology, not just on the Black Sea and its immediate environs – interaction between the peoples living around the Black Sea and those of the steppes and beyond was vital for the development of the whole of ancient Eurasia. In pursuit of this intent, nearly ten years ago we published as our third volume: J. Chapman and P. Dolukhanov (eds.), Landscapes in Flux. Central and Eastern Europe in Antiquity. This important work, the proceedings of a conference, was warmly received. The present volume marks another step towards fulfilling our aim. It offers a wide spectrum of new perspectives on the archaeology of Eurasia and Anatolia from earliest times down to the early Middle Ages. The papers originate from The 1st University of Chicago Conference on Eurasian Archaeology, organised by Adam Smith and his colleagues in the Department of Anthropology at Chicago in May 2002. The successor conference was held in April 2005; we hope to publish its proceedings too. I would like to offer my sincere thanks to Adam Smith and his collaborators for offering this volume to Colloquia Pontica. The preparation of such a large work for publication has taken some considerable time and effort. David Peterson and Laura Popova bore the brunt of this work in Chicago; James Hargrave has handled the copy-editing, reference checking, etc. My final thanks go to Ms Gera van Bedaf, our Desk Editor at Brill, to her colleagues in Leiden and to our typesetter in the Philippines. Gocha R. Tsetskhladze Series Editor Melbourne, December 2005
INTRODUCTION DAVID L. PETERSON, LAURA M. POPOVA AND ADAM T. SMITH
As archaeology struggled in the 20th century to define its disciplinary position and internal domains of specialisation, it crystallised regional traditions of research into a tightly bounded global geography. New World and Old were divided up into discrete territories that have rarely shifted their borders. The Near East, South Asia, the Andes, the Maya Lowlands, etc. all represent not only discrete places on the globe but also canonical spheres of professional expertise and scholarly tradition. With the globe largely parcelled into codified domains, it is rare in contemporary archaeology to be able to watch a new regional tradition of research come into being. And yet, since the collapse of the Soviet Union, archaeologists in the United States and Europe, in partnership with colleagues in the Newly Independent States, have been working to do just that. While archaeology in Eurasia has a deep and rich tradition within local academies, it has only been in the last two decades that the region’s archaeological record has gained prominence in contemporary debates in the United States and Europe. Building on the work of pioneers from V.G. Childe to C.C. Lamberg-Karlovsky and P.L. Kohl, who worked to establish scholarly relationships across the often tense Cold War boundaries of the Soviet era, a growing number of collaborative international projects are redefining our understanding of that vast continent from the Carpathians to the Caucasus to Central Asia and Kamchatka; and in so doing, they are posing important new historical questions for the regions on its periphery and anthropological questions with global ramifications. The 1st University of Chicago Conference on Eurasian Archaeology was held on 3–4 May, 2002 in order to provide a scholarly forum for researchers working in far-flung parts if the region to share current results and perspectives on entrenched problems. Despite the geographical immensity of Eurasia, the relation between and local scales of analysis has emerged in recent years as a pivotal point of contention in contemporary representations of the region’s archaeology. Two seemingly opposed perspectives have begun to dominate interpretative priorities. The first favours a cosmopolitan approach as a safeguard against the appropriation of archaeology by local forms of nativism and nationalism. The second,
xiv
D.L. PETERSON, L.M. POPOVA & A.T. SMITH
while sympathetic to aspects of the cosmopolitan agenda, is more concerned to develop local perspectives on social transformations as a means of bringing greater depth and rigour to the archaeological understanding of material cultural practices and socio-cultural processes. Accordingly, the theme of the 2002 Chicago Conference was ‘Beyond the Steppe and the Sown: Integrating Local and Global Visions’. We should like to thank all of the participants for their contribution both in the colloquia and in the informal communications between sessions. Together these interactions have helped to initiate a series of conversations on critical issues and problems that we hope will continue to shape growth of a tradition of Eurasian archaeology in the United States and beyond. The conference, and hence this publication, would not have been possible without the tireless contributions of the students in the University of Chicago’s Department of Anthropology; the conference organising committee included Tiffany Thompson and Lauren Zych, as well as the editors, with additional assistance from MaryFran Heinsch. Long after the conference was over, members of the organising committee for a 2nd University of Chicago Eurasian Archaeology Conference (held in April 2005) assisted with the gargantuan task of compiling the index to this volume. They included Alan Greene, Charles Hartley, Madeline McLeester, Maureen Marshall, Kraig Obadashian, Bilze Yazicioglu and Lauren Zych. Much needed assistance was provided by Paige Davis and David Forero. Thanks are also due to the faculty of the University of Chicago who served as session moderators during the two-day event: Michael Dietler, Nicholas Kouchoukos, Kathleen Morrison, Adam T. Smith and Tony Wilkinson. Our thanks to Gocha Tsetskhladze and James Hargrave for helping to shepherd the manuscript into publication. Lastly, we want to extend our appreciation to the Norman Waite Harris Memorial Fund and the Marion R. and Adolph J. Lichtstern Fund for providing the financial support that made this conference possible. D.L.P., L.M.P. and A.T.S. Chicago, November 2005
LIST OF ABBREVIATIONS
BAR BASOR KSIA
British Archaeological Reports. Bulletin of the American Schools of Oriental Research. Kratkie Soobshcheniya Instituta Arkheologii Akademii Nauk SSSR (Short Bulletins of the Institute of Archaeology, Academy of Sciences of the USSR), Moscow (in Russian). MAD Materialy po arkheologii Dagestana (Materials on the Archaeology of Daghestan), Makhachkala (in Russian). MIA Materialy i Issledovaniya po Arkheologii SSSR (Materials and Research on the Archaeology of the USSR), Moscow (in Russian). NZWKYOB Neimenggu Zizhiqu Wenwu Kaogu Yanjiusuo and Oerduosi Bowuguan (Archaeological Institute of Inner Mongolia Autonomous Region and the Ordos Museum). RosA Rossiiskaya Arkheologiya (Russian Archaeology), Moscow (in Russian). SA Sovetskaya Arkheologiya (Soviet Archaeology), Moscow (in Russian). SAPA Svod Arkheologicheskikh Pamyatnikov Severo-Vostochnogo Azerbaidzhana (Catalogue of Archaeological Monuments of North-East Azerbaijan) (in Russian). TUBA-AR Türkiye Bilimler Akademisi-Arkeoloji Dergisi. VDI Vestnik Drevnei Istorii (Bulletin of Ancient History), Moscow (in Russian). VDVORAN Vestnik Dal’nevostochnogo Otdeleniya Rossiiskoi Akademii Nauk (The Bulletin of Far Eastern Branch of Russian Academy of Sciences), Vladivostok (in Russian). VGMG Vestnik Gosudarstvennogo Muzeya Gruzii (Bulletin of the State Museum of Georgia), Tbilisi (in Georgian and Russian).
LIST OF FIGURES
Chapter 2 Fig. 1. Persistent economic and material culture frontiers in the Pontic-Caspian steppe region ............................................................................
46
Fig. 2. Contact between the Srubnaya-Andronovo horizon and the Bactria-Margiana complex, ca. 1800–1600 BC ................................................
59
Chapter 3 Fig. 1. Map of the region ......................................................................................
64
Fig. 2. Burials excavated in the Tumek-Kichidzhik (I) and Tarim-Kaya (II) cemeteries ....................................................................................................
68
Fig. 3. Burials excavated in the Sakar-Chaga cemetery. 1 – multiple; 2 – burned construction; 3 – inhumation ..........................................................
69
Fig. 4. Female skulls from the Sakar-Chaga 3 cemetery ....................................
71
Fig. 5. Male skull from the Sakar-Chaga 3 cemetery; and facial reconstruction by author ....................................................................................
72
Fig. 6. Multiple (1) and podboy (2) pits from the Sakar-Chaga 1 cemetery; podboy pit from the Gyur 4 cemetery (3) ........................................................
75
Fig. 7. Khorezmian ossuaries, 2nd century BC–2nd century AD ......................
78
Fig. 8. Male and female from the Sakar-Chaga 1 cemetery. Facial reconstructions by author ....................................................................................
79
Fig. 9. Male from the Gyur 4 cemetery; female from the Tumek-Kichidzhik cemetery. Facial reconstructions by author ......................................................
79
Fig. 10. Sculpted ossuary from the Yssi-gir 4 cemetery ....................................
82
Fig. 11. Male from Yssi-gir 4 cemetery. Facial reconstruction by author ........
82
Chapter 4 Fig. 1. Map showing site locations ......................................................................
90
Fig. 2. Remote sensing images of Krasnyi Yar: a.) magnetic field gradient image; b.) electrical resistance image ................................................................
97
Fig. 3. Remote sensing images of Vasilkovka IV: a.) magnetic field gradient image; b.) electrical resistance image ................................................................
98
Fig. 4. Krasnyi Yar I settlement plan ..................................................................
101
Fig. 5. Vasilkovka IV settlement
........................................................................
102
Fig. 6. Excavation plans for Krasnyi Yar I and Vasilkovka IV ..........................
103
xviii
LIST OF FIGURES
Chapter 5 Fig. 1. Map of major Namazga III sites discussed in the text ..........................
114
Chapter 6 Fig. 1. Eastern Eurasian steppe zone and study region ......................................
126
Fig. 2. Known sites in Semirech’ye ca. 1960 (after Arkheologiskaya Karta Kazakhstana [1960]) ..........................................................................................
128
Fig. 3. Landscape approach to mobile pastoralism ..............................................
131
Fig. 4. Satellite image of the Koksu river valley (source: USGS Landsat TM-7 2002) ........................................................................................................
133
Fig. 5. Survey region and polygons of Dzhungar Mountains Archaeological Project ..................................................................................................................
135
Fig. 6. Dzhungar Mountains Archaeological Project sites within each altitude zone ........................................................................................................ A: Percentage B: Number
137
Chapter 7 Fig. 1. Map of Velikent Kuro-Araxes culture sites ..............................................
145
Fig. 2. Ceramics: 1, female figurine (Karatsan); 2–5, andirons (2 – Kabaz-Kutan I; 3 – Galgalatly; 4–5 – Velikent II); 6–13, vessels (6 – Velikent III; 7–8 – Velikent II; 9 – Velikent V; 10, 12 – Kayakent VI; 11 – Chirkey; 13 – Torpakh-Kala); 14–15, frying pans (14 – Kabaz-Kutan I; 15 – Galgalatly) ..................................................................................................
147
Fig. 3. Bronze and stone artefacts (1 – Velikent I; 2–5, 7–11, 14–19, 21–27 – Velikent III; 20 – Velikent V; 6, 12 – Velikent – II; 13 – Bugdatepe; 28 – Velikent IV) ..................................................................
150
Fig. 4. Bronze, bone and paste artefacts (1–20, 22–36 – Velikent III; 21 – Manas) ........................................................................................................
151
Fig. 5. C14 Dates of sites in Daghestan ..............................................................
156
Chapter 8 Fig. 1. Map of Araxes painted ware settlements in eastern Anatolia ................
161
Fig. 2. Gümü¤pınar mound ....................................................................................
163
Fig. 3. Sarıveli mound and necropolis ..................................................................
164
Fig. 4. Çaygeldi fortress and necropolis ..............................................................
166
Fig. 5. Yılankalesi fortress ....................................................................................
167
Fig. 6. Haydarkalesi fortress ..................................................................................
168
LIST OF FIGURES
xix
Chapter 9 Fig. 1. Elevation map of Caucasia ........................................................................
174
Fig. 2. Late Bronze and Early Iron Age fortified sites: a comparison of the topography and architecture of three sites ........................................................
176
Fig. 3. An archaeological map of the Tsakahovit Plain, Republic of Armenia ................................................................................................................
177
Chapter 10 Fig. 1. The Egiin Gol region of northern Mongolia and the survey area ........
186
Fig. 2. Distribution of BEIA sites at Egiin Gol ..................................................
190
Fig. 3. Bivariate density estimation plot of BEIA sites. Nos. 1–4 and heavy lines show tributary valleys (contour lines are quartile contours at 10% intervals) ................................................................................................
191
Fig. 4. Summed probability distribution of the calibrated C14 determinations for Xiongnu cemetery site EGS 100 (BT refers to the local site name, Burkhan Tolgoi) ..................................................................................................
195
Fig. 5. Distribution of Xiongnu period sites at Egiin Gol (largest sites marked by squares) ............................................................................................
196
Fig. 6. BEIA and Xiongnu period site size comparison – one BEIA site is excluded for lack of size ................................................................................
197
Chapter 11 Fig. 1. Khanuy river and the research area ..........................................................
205
Fig. 2. Plan of the khirigsuur Urt Bulagyn (KYR) ............................................
210
Fig. 3. Plan of the khirigsuur KYR 2 ..................................................................
211
Fig. 4. Urt Bulagyn. Central mound in the background and small mounds in the foreground (photograph by Francis Allard) ..........................................
215
Fig. 5. Mound KYR 1–1 at Urt Bulagyn. Horse skeletal remains in situ (photograph by Francis Allard) ..........................................................................
220
Chapter 12 Fig. 1. Map of Sinop region and Black Sea sites mentioned in the text (after Doonan et al. 2001; base map by A. Gantos) ......................................
231
Fig. 2. Chalcolithic pottery from the Sinop region. A: Kiran Tepe; B: Mezarliktepe; C: Maltepe (photographs by A. Bauer and O. Doonan) ....
234
Fig. 3. EBA pottery from Kocagöz Höyük, Sinop region (photograph by O. Doonan) ..............................................................................
235
Fig. 4. EBA and MBA pottery from Güllüavlu, Sinop region (photograph by A. Bauer) ..................................................................................
235
xx
LIST OF FIGURES
Chapter 13 Fig. 1. Multistable image: rabbit and/or duck (after Gombrich 1969, fig. 2) ..................................................................................................................
249
Fig. 2. Bucket from Trialeti kurgan 17 (after Kuftin 1941, pl. LXXXVIII) ....
252
Fig. 3. Goblet from Trialeti kurgan 5 (after Miron and Orthmann 1995, fig. 10) ................................................................................................................
253
Fig. 4. Goblet from Karashamb (courtesy the State History Museum of Armenia) ........................................................................................................
254
Fig. 5. Goblet from Karashamb (courtesy the State History Museum of Armenia) ........................................................................................................
256
Fig. 6. Sealing Kt d/k 33, from Kültepe karum Kanesh (after Özgüç 1965, pl. XVII, 50) ......................................................................................................
257
Fig. 7. Sealing Kt a/k 494, from Kültepe karum Kanesh (after Özgüç 1965, pl. V, 15a) ............................................................................................................
258
Fig. 8. Sealing Kt c/k 1389, from from Kültepe karum Kanesh (after Özgüç and Tunca 2001, pl. C2) ..............................................................
258
Chapter 14 Fig. 1. Site and excavation information for cities built by Rusa II ..................
266
Fig. 2. Architecture and materials excavated from cities built by Rusa II ........
267
Fig. 3. Materials to be studied and their estimated values in answering key questions of research (5 highest-1 lowest) ................................................ 278f. Question A: Value in defining continuity and change in Urartian dynastic artistic tradition (internal development) Question B: Value in defining relations of Urartian art with the other artistic traditions of the Near East (external influence) Question C: Value in dating Question D: Importance in defining change in dynastic propaganda Question E: Importance in defining change in state religion Question F: Importance in defining change in governance strategies and practices Chapter 15 Fig. 1. 1, Horoztepe (after Özgüç and Akok 1958); 2–3, Nerkin Getashen (after Areshian 1988 and Kushnareva 1977); 4, Khanlar (after Gummel 1992); 5, Treli (after Abramishvili 1978); 6, Ust’ Labinskaya (after Guguev 1992); 7, Gundestrup (after Filip 1977) ............................................................
285
Fig. 2. Khanlar, barrow no. 149 (after Gummel 1992) ......................................
288
Fig. 3. Khanlar, barrow no. 150 (after Gummel 1992) ......................................
289
LIST OF FIGURES
Chapter 16 Fig. 1. ...................................................................................................................... A. Area map. B. Clusters of rock art sites in the Yinshan of Inner of Mongolia (1, Bu’erhan Shan; 2, Dishui Gou; 3, Ge-er’aobao Gou and Molehetu Gou) and Helanshan of Ningxia Province (4, Heishimao; 5, Daxifeng Gou; 6, Helankou; 7, Siyanjin; 8, Gujingou and Damaidi)
xxi 299
Fig. 2. (a) Bu’erhan Shan petroglyphs (Yinshan); (b) Siyanjin rock panel (Helanshan); (c) mask motifs at Helankou (Helanshan) ..................................
305
Fig. 3. Face or mask at Helankou with Xixia inscription ..................................
307
Fig. 4. Damaidi cart (Helanshan) ..........................................................................
308
Chapter 18 Fig. 1. Location of the Samara oblast on the Eurasian steppes ........................
323
Fig. 2. Simplified diagram of the interrelationship of technology, social practice and value ..............................................................................................
326
Fig. 3. Location of Kamyshla raion and the sites discussed in the text (prepared with the help of Laura Popova) ........................................................
329
Fig. 4. The mace from Kutuluk I ........................................................................
330
Fig. 5. The Kamyshla survey area and sites recovered by the survey: 1, Kibit I settlement; 2, Kibit II settlement; 3, Kibit III settlement; 4, Novoe Usmanovo I quarry; 5, Baitugan I quarry; 6, Baitugan II quarry; 7, Baitugan I settlement; 8, Baitugan II settlement; 9, Ermakovo I settlement; 10, artefact scatter ............................................................................
334
Fig. 6. Srubnaya-type pottery fragments (1–5, 7), an intact ceramic cup (6), and worked stone from Kibit I ..........................................................................
335
Fig. 7. Location of the 4 × 4 m excavation unit at Kibit I (left), and plan (upper right) and section (lower) of the edge of a house floor encountered in the excavation ................................................................................................
337
Fig. 8. Distribution of animal bones from the Kamyshla survey (prepared by Audrey G. Brown and Emmett Brown) ......................................
341
Chapter 19 Fig. 1. Regional context of Kargaly Project and Kargaly mining districts (after Chernykh, Lebedeva et al. 2002, 34, fig. 2.13; 49, fig. 3.2) ................
344
Fig. 2. ...................................................................................................................... 1. Bioclimatic phases: estimated chronology BP and palaeoclimate (after Khotinsky 1984) and archaeology (after Chernykh, Avilova et al. 2000; 2002). Arrows up and down (1–4) suggest the evolution of arboreal pollen (López et al. 2001; López-Sáez et al. 2002a; 2002b) 2. Representation of selected ecological groups. Each pollynomorph has been assigned to a single ecological group. The correlation is based
348
xxii
LIST OF FIGURES
on the floral catalogue of the Kargaly region and on the species identified during our field work. These pollynomorphs might correspond to several other ecological groups in other regions. 2.3. Cubic regression model: distance to forests explains .27646 of Betula pollen variance Fig. 3. Sampling model on digital terrain model: location of paleopalynological sequences and pollen rain samples at Kargaly (South Urals, Russia) ............ 349 Fig. 4. Map of distances between sampling points and forests at Kargaly (South Urals, Russia) .......................................................................................... Fig. 5. ...................................................................................................................... 1. Normalised Difference Vegetation Index (NDVI) and sampling model 2. Predictive model at Kargaly (South Urals, Russia)
353 354
Chapter 20 Fig. 1. Map showing Neolithic, Xia aand Shang archaeological sites ..............
361
Fig. 2. Bronze spatulas (bishou) from Huoshaogou, Yumena, Gansu (after Linduff 2000, fig. 10) ..............................................................................
363
Fig. 3. Zhukaigou, Inner Mongolia, Phase V, bronzes (after Kaogu xuebao 1988.3, 325) ........................................................................................................
365
Fig. 4. Bronzes of the northern type unearthed from the Fu Hao Tomb and their northern counterparts: 1–6, from Fu Hao Tomb; 7, from Yantoucun, Suide, Shaanxi; 8, Inner Mongolia; 9, Naimatai, Guinanxian, Qinghai; 10, Linzheyu, Baode, Shaanxi; 11, Transbaikalia, Russia; 12, Chirnokovo, Krasnoiarsk, Russia (after K.C. Chang 1986, fig. 51) ....................................
367
Fig. 5. Yuhuangmiao, tomb 86. Remains of wooden coffin and animal sacrifice (after Beifang kaogu sishininian 1990, 81, fig. 35) ..........................
367
Chapter 21 Fig. 1. The Khazar empire and the Islamic caliphate (ca. 900) ........................
374
Chapter 22 Fig. 1. Map of the Kura-Araxes horizon (modified after Kushnareva 1997) ....
383
Fig. 2. Orientation of pores (modified after Vandiver 1988, fig. 3) ..................
386
Fig. 3. Table of comparison of subtypes recovered from domestic contexts ....
387
Fig. 4. Photographs and corresponding radiographic images of textile impressions imbedded in the clay fabric ..........................................................
388
Fig. 5. Xeroradiographic images of select straw-tempered subtypes ..................
389
Fig. 6. Xeroradiographic images of select grit-tempered subtypes ....................
391
Chapter 23 Fig. 1. Research area, Primor’e region of the southern Russian Far East ........
397
LIST OF FIGURES
xxiii
Fig. 2. Area of the Zaisanovka culture. Pottery types of western and eastern types ........................................................................................................
398
Fig. 3. Table of radiocarbon dates of sites of the Zaisanovka culture ..............
399
Fig. 4. Area of the Andron culture, the Zaisanovka culture and of Neolithic sites of north-eastern China ..............................................................
403
Fig. 5. ‘Coarse’ and ‘fine’ pottery of the Andron and Zaisanovka cultures and from Neolithic sites of north-eastern China ..............................................
406
Fig. 6. Meander pattern of Andron and Zaisanovka pottery ..............................
407
Chapter 24 Fig. 1. Location map of Bohai and the type-sites of ceramic wares found in Bohai archaeological sites ..............................................................................
412
Fig. 2. Distribution of ceramic wares in the Bohai archaeological sites of Primor’e ..............................................................................................................
413
Fig. 3. Table of the chemical composition of the body of Yue ware ................
416
Fig. 4. Table of analyses of Yue glazes ..............................................................
417
Chapter 25 Fig. 1. (all after Cordova, Rybak and Lehman 2001) ........................................ A. Map of the Black Sea region. B. Vegetation zone in the south-western Crimea: 1 – Saline steppe; 2 – True steppe; 3 – pre-Montane forest-steppe; 4 – sub-Boreal forests of the northern slopes; 5 – Yaila steppes; 6 – Forests of the southern slopes; 7 – sub-Mediterranean woodlands. C. Cross-section through the mountains.
426
Fig. 2. The Heraklean peninsula: limits of the chora of Chersonesos and studied stratigraphic sections ..............................................................................
428
Fig. 3. Summary of palaeoclimatic data, pollen and soils ..................................
431
Fig. 4. NG2 core from the Chryornaya floodplain. Summary of pollen data ......................................................................................................................
433
Fig. 5. The Mediterranisation of the southern Crimean landscapes ca. 7500–3000 BP. The diagram shows the combination of factors that presumably led to the formation of shiblyak and phrygana, the two subMediterranean plant communities in the southern Crimea ..............................
442
Chapter 26 Fig. 1. Nine local variants of the Yamnaya Cultural-Historical Area (according to Merpert 1968, 14–39) ..................................................................
449
Fig. 2. Five local groups of the Yamnaya Cultural-Historical Unity (according to Shaposhnikova 1985, 347–9) for the northern Pontic steppe except the Dniester-Danube and Don river zones ............................................
450
xxiv
LIST OF FIGURES
Fig. 3. Economic models in the steppe (4th–2nd millennia BC) (according to Shilov 1975) ..................................................................................................
451
Fig. 4. Latest and Final Eneolithic steppe and forest-steppe cultures in the North Pontic area ................................................................................................
452
Chapter 28 Fig. 1. Radiocarbon dates from the Boisman 2 stratigraphic profile ................
474
Chapter 29 Fig. 1. Prehistoric radiocarbon dates for the Primor’e region of the Russian Far East ................................................................................................................
479
Fig. 2. Bronze Age radiocarbon dates for the Primor’e region of the Russian Far East ................................................................................................
480
Fig. 3. Spawning distributions of Cherry, Pink and Chum salmon in the eastern drainage systems of the Primor’e region of the Russian Far East ....
482
Fig. 4. Margarita culture site locations and radiocarbon dates in the Primor’e region of the Russian Far East ..........................................................
484
I. LOCAL AND GLOBAL PERSPECTIVES
CHAPTER ONE THE EARLY INTEGRATION OF THE EURASIAN STEPPES WITH THE ANCIENT NEAR EAST: MOVEMENTS AND TRANSFORMATIONS IN THE CAUCASUS AND CENTRAL ASIA PHILIP L. KOHL
There is a very long and rich tradition of archaeological research on the Eurasian steppes. Most of this work has been undertaken within what can be called the Russian and Soviet national archaeological tradition and during the past century has been conducted principally by outstanding Ukrainian and Russian scholars ever since the pioneering studies of V.A. Gorodtsov (Merpert 2001). Specific features of such national traditions of research condition what is known about the remote past: the questions asked and the explanations offered. The archaeological record compiled for a given area, even a vast one such as the Eurasian steppes, also often displays certain shared features that lend themselves to certain types of reconstructions and interpretations. It is not just a question of the tradition of research structuring our understanding, but also what has been found, the evidence itself. This paper begins by briefly and critically examining two related problematic concepts in Eurasian steppe archaeology: reliance on the concept of the archaeological culture to order materials and identify ethnically the makers of these cultures; and the use of migrations for explaining changes in material remains. It then argues for the need to integrate the evidence from the steppes with that compiled for more settled sown areas to the south in the greater ancient Near East and attempts to sketch some large-scale processes that affected pastoralists and agriculturalists, the steppe and the sown, during Bronze Age times. Finally, it concludes with a brief consideration of the most problematic, overused and potentially dangerous feature of steppe archaeology: the elusive quest to find the remains of Indo-Europeans, particularly the vaunted Aryans.
4
PHILIP L. KOHL
Two Problematic Concepts in Steppe Archaeology The Archaeological Culture: Splitting or Lumping, Purity or Hybridity, Stability or Change. In a thoughtful article A.N. Gei (1999) recently assessed current difficulties in interpreting the incredible amount of archaeological materials that have been recovered from the excavations of thousands of Bronze Age kurgans on salvage projects during the 1970s and 1980s, the so-called ‘golden age of kurgan archaeology’. Gei trenchantly criticised the current state of affairs, lamenting the bewildering confusion of names and inadequate conceptualisation of closely related archaeological remains; as an example, he listed 24 terms (1999, 37) used to describe Chalcolithic to Middle Bronze Age remains just within Rostov province in the lower Don region! According to Gei, the tendency to split or subdivide these remains into countless archaeological cultures is not just due to the recent tremendous accumulation of materials, but it is also a consequence of the inter-regional breakdown in communication and standardisation associated with the dissolution of the Soviet Union. He eloquently argued against these trends and for greater systematic treatment of these materials, insisting upon the necessity of creating computer data banks to process and prepare the materials, work that he and his colleagues have already begun at the Institute of Archaeology in Moscow. One can only concur with Gei’s assessment and applaud his recommendations. Personally, I even feel a bit relieved to learn that it is not only I who am confused by the formidable roster of archaeological cultures in ‘kurgan archaeology’, but even a scholar as intimately familiar with these materials as Gei. From this perspective, I feel more emboldened to offer my own, admittedly less informed interpretation of these materials. This reaction, of course, is not an excuse for shabby scholarship, but a recognition of the fact that no one can claim perfect understanding, that there is no final word to utter. Where Gei and I differ, however, is that implicit in his article and recommendations is the belief that it is just a question of doing more systematic work to get to that stage of understanding. Create a sufficiently comprehensive data bank and the ambiguities and uncertainties will be resolved; once this work is done, the cultural terminology will reflect past reality. His article is perfectly consistent with the positivist approach of the ‘explicitly scientific’ processual archaeology of the Anglo-American tradition in the late 1960s and 1970s. Unfortunately, reality is more complex. There will always be splitters and lumpers and proper times to stress differences or similarities. Such classifica-
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
5
tion problems cannot be circumvented through more refined numerical taxonomies or fancy computerised systems. The problem is conceptual. The fit between the archaeological culture and the culture of the ethnographer or ethnic culture is imprecise; we can never be certain that similar material remains relatively restricted in space and time signify a single or multiple groups. Similarly with different material remains. Are we dealing with two or more groups or the same group performing different activities? Even more problematic is the attempt to identify such archaeological cultures as ancestral to much later, historically mentioned ethnic groups. Some steppe archaeologists have offered such identifications as if they could be definitively established and were not always necessarily tentative. In reality, we all know that the archaeological culture is a problematic concept, sometimes nothing more than a convenient device for grouping together similar looking assemblages of artefacts and often employed unreflectively – assuming the fit between the defined culture and a past people. I wish to take this well-known critique a step further and question the implicit concept of the ethnic culture with which the archaeologist hopes his materials correspond. At times – the best of times – there may be a direct correlation between the archaeologically defined culture and some collective entity that distinguished itself and existed in the prehistoric past. Even then when this elusive correspondence is real, however, we should reflect more seriously on what contemporary social anthropologists mean by a culture or ethnic group. As Eric Wolf (1984) reminded archaeologists some years ago, the determination of the archaeological culture represents only the beginning, not the end point of the analysis. We must realise that cultures are dynamic entities that can change dramatically over very short periods of time, particularly as they get caught up in larger historical processes that can overwhelm and transform them. Such large-scale processes, involving the development of new technologies and economies and the large-scale movements of materials through various forms of exchange and of peoples through their migrations, were at work on the Eurasian steppes during the Bronze Age and profoundly affected the countless archaeological cultures that have been defined throughout this vast area. The cultures that ethnographers study are not pure, pristine entities developing in a vacuum. Rather, they are almost always hybrids, fissioning or coalescing, assimilating or modifying the customs of the neighbouring peoples with whom they constantly interact. Cultures are not primordial entities or essences once crystallised in time and then remaining forever the same; they are never made, but always in the making. Finally, ethnographic and historical sources
6
PHILIP L. KOHL
both make it patently clear that the same people can change its way of life, including its basic subsistence economy – more agricultural, more pastoral nomadic or whatever – within a single generation. As that is true, it means the material culture of a group or people can profoundly change as well quite quickly. Sudden shifts in the archaeological record need not necessarily involve the replacement of one people by another, but simply represent the transformation of the lifestyle of the same people. The point is not that cultures have to change so quickly, but just that they are capable of doing so. How do we interpret significant and sudden changes in the archaeological record: new peoples arriving on the scene or the same people adopting a new way of life? At times reasonable alternative explanations of the same evidence are possible. The Movements of Peoples: Singular Events or Protracted Patterned Processes, Anachronistic Invocations or Appropriate Models. Continental European and Soviet/Russian archaeologies have differed greatly from Anglo-American archaeology during the last few decades in the frequency in which sudden changes in the archaeological record are attributed to the appearance of new peoples. American processual archaeology, in particular, so categorically rejected the use of migrations for explaining cultural change that D. Anthony (1990) felt it necessary to plea-bargain for its reintroduction into Anglo-American archaeological discourse. He eloquently and sensibly urged his compatriots not to throw out the precious explanatory ‘baby’ of migrations with its dirty ‘bathwater’ just because the concept had so been misused and over-employed in the past. Why the discrepancy among these different traditions of archaeological research? Migrations that had profound social and economic consequences are well documented historically, and it is reasonable to expect that they occurred in some form in late prehistoric times as well. Why question what must have occurred with reasonable frequency? The processual critique or rejection of migrations to explain cultural change had several components. First, it was correctly perceived that migrations had been too frequently invoked to explain even minor changes in the archaeological record. Marauding mounted pastoral nomads sweeping across the steppes from the east or down from the north were easily imagined to explain shifts in mortuary practices or settlement patterns, even for periods long preceding the documentation of pastoral economies or hard evidence for the domestication and riding of horses (cf. Rassamakin’s nice caricature of this image: 1999, 59). More mundane, but at times more convincing explanations related to local adaptations to changing environmental conditions, and these
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
7
were too frequently overlooked in the search for the dramatic arrival of new peoples. Both those who proposed and those who rejected migrations frequently conceived of them as singular and, in that sense, accidental and highly contingent events. They often were not viewed as protracted, recurrent processes that could be explained structurally in terms of the ways of life and economies of the peoples who migrated. Migrations are not unitary phenomena. One migration does not necessarily resemble another either in the form it takes or in the reasons why the movement is occurring. Too frequently the type of migration invoked – the marauding nomad from the steppes – is described anachronistically, not in line with the abilities and technologies of the peoples who were then moving. One has to beware of the tendency to apply unproblematically to the Chalcolithic or Bronze Age types of migrations recorded for much later historical periods. Timur’s conquests may, to a certain extent, resemble Genghis Khan’s, but these were quite different from the forays of the Cimmerians and Scythians and both bore little, if any resemblance to the late Tripolye folks who were transforming their archaeological culture by developing a more mobile economy and becoming cattle-herders from the middle centuries of the 4th millennium BC onwards. In other words, peoples moved in different ways for different purposes, and we should be careful not to consider them the same and lump them all together. In most cases, we should conceive of ‘migrations’ not as single events – Genghis Khan storming across the steppes and wreaking havoc on the civilised world – but as protracted processes that are principally related to a people’s way of life and the natural and cultural environments to which they are adapting. The basic point is that when one is trying to model large-scale, ‘global’ interaction in the Bronze Age or in later prehistory in general, one cannot just talk about trade and the movement of materials. Most parts of the Bronze Age world were relatively sparsely populated, and that world was not divided up like it is today into well-demarcated states, each bordering neighbouring states to encompass literally the entire planet. During the Bronze Age there were vast zones that were essentially not occupied or whose productive potential had not yet been tapped, and the borders between polities or groups at different levels of development or doing different things were not well-defined and, consequently, quite porous. One cannot model large-scale processes in the Bronze Age without considering the fact that peoples were often moving in recurrent, patterned ways that can be detected archaeologically.
8
PHILIP L. KOHL
Culture-History on a Grand Scale: Connecting the Eurasian Steppes with the Ancient near East, ca. 3600–1900 BC . . . Just as a cart does not resemble a horse, the migrations of the third and second millennia BC pastoralists who were still involved with agriculture did not resemble the migrations of mounted nomads in later centuries. The former were markedly slower and more gradual, and land suitable for cultivation interested the migrators no less than new pastures did (Khazanov 1994, 94).
There is a long, misguided, and essentially unbroken tradition in American archaeology to contrast evolution with history. The former is admirable and scientific. It is what archaeologists as scientists should be doing – adopting a generalising comparative perspective, employing neo-evolutionary concepts such as stages of development or levels of socio-cultural integration to write conjectural accounts of how peoples in different, geographically distant and unrelated areas traversed similar evolutionary paths. Writing history or operating in a historicist mode, on the other hand, is conceived as a particularistic pursuit, concerned with getting one’s regional cultural areas defined and chronologies sorted out. At best, it is a necessary evil for other, perhaps local scholars to do, so that one can get on with the more ennobling task of learning how culture and evolution really work. Admittedly, there may be bit of a caricature in this portrait, but it is important to reject the false dichotomy between evolution and history. They are not opposed, but complementary concepts. Searching for timeless evolutionary parallels can be an enlightening exercise, but it also may be a misleading one. Thus, for example, I would question references to the ‘protourban’ gigantic settlements of late Tripolye times or to the ‘country of cities’ (Strana gorodov) for the fascinating and planned, but still quite tiny SintashtaArkaim Late Bronze Age settlements in the trans-Urals. The former are noteworthy for not exhibiting any sign of significant social differentiation despite their great size, and the latter remain, in my opinion, somewhat enigmatic, in need of a convincing explanation, but certainly not in terms of their undergoing an ‘urban revolution’ with all its concomitant features, including the practice of intensive irrigation agriculture. Such comparisons are misguided and impede, rather than promote, understanding. Properly conceived, culture history is concerned with writing a large-scale historical narrative that focuses on the interactions between areas and searches for patterns and processes that help explain what actually happened. Such a
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
9
narrative should be able to integrate the Bronze Age world of the Eurasian steppes with the more sedentary, sown world farther south – whether in the Caucasus, Central Asia, or the ancient Near East more narrowly conceived. Four large-scale interrelated and consecutive processes that helped integrate the western Eurasian steppe with areas farther south from ca. 3600–1900 BC will be briefly sketched here: 1. The break-up of the gigantic Tripolye settlements and the subsequent development of a more mobile economy, utilising oxen-driven wagons with heavy wooden tripartite wheels across Ukraine and into the preKuban region of the north-western Caucasus during the second half of the 4th and the first centuries of the 3rd millennium BC. More mobile, extensive agricultural and cattle-herding economies developed across the west Eurasian steppes in the wake of this substantial change in settlement patterns in the final or post-Tripolye period, and slightly later during the second half of the 4th and the first half of the 3rd millennium BC. The Pit-Grave, Novotitorovskaya, and Catacomb cultural formations archaeologically document these transformations. 2. The demise of earlier systems of exchange, which included particularly the movement of metals from the Carpatho-Balkans area across the steppes at least to the middle Volga area, and the nearly simultaneous or slightly subsequent appearance of the Early Bronze Age cultures of the Caucasus, particularly the Maikop and Maikop-related cultures of the north-western Caucasus and the Kura-Araxes and Kura-Araxesrelated Velikent cultures of Transcaucasia and the north-eastern Caucasus and the subsequent production and exchange of Caucasian arsenic bronzes. To adopt the formulation of E.N. Chernykh (1992), this pattern corresponds to the demise of the earlier ‘Carpatho-Balkan Metallurgical Province’ (CBMP) and the advent of the ‘Circumpontic Metallurgical Province’ (CMP). This shift and the appearance of these Early Bronze cultures of the Caucasus occurs approximately and not coincidentally at the same time as evidence for a substantial Mesopotamian presence along the upper Euphrates and in eastern Anatolia (the so-called Uruk expansion). 3. The movements into and out of the Caucasus, beginning at the end of the 4th millennium and continuing through the middle of the 3rd millennium BC. These movements included the dispersal of Kura-Araxesrelated peoples far to the south-east into the central Zagros and possibly
10
PHILIP L. KOHL
to Khuzistan and south-west as far as today’s northern Israel (the socalled Khirbet Kerak phenomenon), as well as subsequent movements from the steppes into Transcaucasia probably principally via the Caspian corridor beginning at the end of the first half of the 3rd millennium and continuing into the early 2nd millennium BC. These processes are documented by the disappearance of the Kura-Araxes settlements and the advent of the so-called ‘royal’ kurgans in eastern Georgia with oxendriven wagons with tripartite wooden wheels, precious jewellery and even possible evidence for human sacrifice (Bedeni, Martkopi, Tsnori and later Trialeti, Karashamb, etc.). 4. The movements of peoples from the steppes first into Khorezmia and then into the Zeravshan valley and onto the plains of Bactria and Margiana constituting one of the formative components of the so-called Bactria-Margiana Archaeological Complex (or BMAC). This process also was protracted but seems to have significantly affected cultural developments in Central Asia only towards the end of the 3rd and beginning of the 2nd millennium BC or somewhat later than in the Caucasus. The reasons for this are both historical and geographical: a) the development and spread of mobile cattle-herding economies proceeded initially west to east with the break up of the Tripolye culture; and b) east of the Caspian the steppe is separated from the irrigated plains of Margiana and Bactria by the Kyzyl and Kara Kum Central Asian deserts. Such deserts formed an arid barrier that did not exist farther west where the steppe merges imperceptibly with the plains of the north-western Caucasus. Let us examine them consecutively in more detail. The Collapse of the Tripolye Gigantic Settlements and the Emergence of More Mobile Economies across the Western Steppes, Utilising Oxen-Driven Wagons with Heavy Tripartite Wooden Wheels. The Cucuteni-Tripolye culture emerged in the second quarter of the 5th millennium BC with clearly established Neolithic roots extending back into the 6th millennium BC (Starcevo Cris, BugDniester, etc.). Gradually, the Cucuteni-Tripolye settlements expanded to the north-east from an original concentration along the numerous left-bank tributaries of the Danube and the middle Dniester (Monah and Monah 1997, 36) towards the Bug and middle Dnieper area near Kiev. Some later Tripolye settlements were even located east of the Dnieper (cf. Arkheologiya Ukrainskoi
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
11
SSR 1985, maps 5, 192; 6, 204; 7, 224). While some settlements, such as Vladimorovka (containing ca. 200 houses over a 34 ha area), seemed to represent fairly substantial towns, most sites were relatively small, conforming to the expectation of what Chalcolithic settlements should look like. This picture radically changed in the late 1960s when K. Siskin utilised aerial photos and discovered a series of extremely large settlements, mostly located in the foreststeppe region between the Bug and middle Dnieper rivers; these included the settlements of Maidanetskoe (270 ha), Dobrovody (250 ha), Talyanki (400 ha), Veselii Kut (150 ha), Miropolye (200 ha), Kosenkova (70 ha), Cicerkozovka (60 ha), Onoprievka (60 ha) and Pyanezkovo (60 ha). Extensive field investigations began in the early 1970s and included geomagnetic surveys to locate houses and plot more precisely the over-all plan of these gigantic settlements. Settlements increase in size from Tripolye A (20–60 ha) to Tripolye B1 (150 ha) or from the middle to the last quarter of the 5th millennium, but the largest settlements, such as Maidanetskoe and Talyanki, date from the last quarter of the 5th through the first several centuries of the 4th millennium BC, or, in general, one can refer to a more than half-millennium period of existence for these extremely large settlements, ca. 4200–3600 BC (compare Videjko 1995, 53 with Chernykh, Avilova and Orlovskaya 2000, 57–9). Most of the gigantic Tripolye settlements are clustered quite closely together in a rich black earth zone between the Bug and Dnieper rivers approximately 120 km due south of Kiev. Comparatively in terms of the overall size of the settlements, these gigantic Tripolye settlements are as large, or larger, than the earliest citystates of Sumer, precede them chronologically by more than half a millennium, and are more closely clustered together, although it is unclear whether they were simultaneously or successively occupied. The evidence for specialisation and internal social differentiation, on the other hand, is much less, and for this reason they should not be referred to as ‘cities’ or ‘proto-urban formations’ but simply as ‘giant sites’ or ‘gigantic settlements’. The aerial photos and geomagnetic surveys have shown that there were large areas within the settlements that were unoccupied, possibly for securing their herds of cattle. The two large ellipses of interconnected houses at Talyanki, for example, were separated from each other by an unbuilt area 70–100 m wide; within the inner ellipse radial rows of houses extend towards the centre in the northern part of the settlement, but as currently understood, the centre of the site, encompassing roughly 60ha, also comprised an open unbuilt area. Excavations at Maidanetskoe have also revealed that only about two-thirds of the investigated buildings were dwellings, a finding which necessitated, correspondingly,
12
PHILIP L. KOHL
a one-third reduction in the estimated population of the settlement (cf. below). In terms of their forms and dimensions, the houses of the contemporaneous smaller settlements are essentially identical to those found in the giant sites. There is a three-tiered settlement pattern: the gigantic settlements (100–400 ha); middle-sized (20–60 ha); and small (2–10 ha). The settlements are clustered in groups. Typically there are one or two middle-sized and/or two to three smaller settlements found within 3–10 km from one of the gigantic settlements (Videjko 1995, 66). Demographic reconstructions are based on the assumption of five to seven persons/house, and the houses within the settlements form groups (of related families?) of up to 20 houses or 100–140 people. The large settlement of Talyanki with 2700 houses was inhabited possibly by more than 15000 people, and, adding then the numbers for the satellite settlements within its cluster, the total population may have exceeded 30000 (Videjko 1995, 72). These calculations are, of course, rough and preliminary, but, taken all together, they show that tens of thousands of people occupied a relatively restricted area of the forest-steppe between the Bug and middle Dnieper rivers at the end of the 5th and beginning of the 4th millennium BC. Cucuteni-Tripolye settlements inherited the basic constellation of plant and animal domesticates that had diffused into the Balkan peninsula from Anatolia: emmer, einkorn and bread wheat (Triticum aestivum), naked and hulled barley, peas, vetch, lentils; and sheep, goat, cattle and pigs. They also grew buckwheat (Fagopyrum esculentum) and millet (Panicum miliaceum), the latter particularly in the eastern area where the giant sites are located. An extensive form of sowing summer wheat and barley interspersed with the growing of vegetables, such as peas and lentils, most likely occurred on the extensive meadows that surrounded the gigantic settlements. A strong indirect case for the use of a primitive plough or ard to create furrows in soils possibly already loosened by hoes can be made by the discovery of large elk antlers with clearly worked ends, such as that found on the settlement at Novyie Ruseshty I (Masson and Merpert 1982, 233; fig. 18, 234). Two types of ards, fashioned from deer and elk antlers, have been recovered: one with a vertically set blade for working recently cleared land; and the other with a horizontal blade for working intensively cultivated fields (Monah and Monah 1997, 80). A clay model of two yoked bulls (oxen?) drawing a sledge was recovered from Maidanetskoe, and traces of yoking and harnessing have been found on steer bones from Talyanki and other giant sites (Videjko 1995, 70, n. 68). It is difficult to imagine how these large sites with their thousands of inhabitants could have maintained themselves without harnessing oxen to draw these ploughs,
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
13
a development that would have increased several times the arable area (Videjko 1995, 70 with references). The gigantic settlements are thought to have been occupied successively for a relatively short period of time (a century or so?) on the basis of their thin cultural deposits. It is impossible, however, to establish this typologically through an analysis of their ceramics or through radiocarbon dates given the closeness in time in which the sites are presumed to have been occupied. Presumably, the short duration of these settlements is related to the type of agriculture that they pursued: an extensive form of shifting cultivation, involving the burning of vegetation and trees and the periodic movement of fields for re-nourishment of the soil. It is hypothesised with some supporting palynological evidence that such ecologically destructive practices led to deforestation and an anthropogenically-induced environmental crisis which was one of the factors responsible for the break-up of the giant sites and the fissioning of the culture into separate regional components by the middle of the 4th millennium BC. Animal husbandry also played an extremely important role in the subsistence economy of these gigantic settlements. Although some regional variation can be observed, cattle were clearly the most important domesticated species and kept for meat and milk production, as well as most likely for draft purposes. Large numbers of pigs were also raised, while sheep and goats, in general, assumed a much more subsidiary role (Masson and Merpert 1982, 234), though their numbers and significance seem to increase after the break-up of the settlements and occupation of the open steppe on Usatovo-related sites when pastoralism formed a more central role in the economy and when influences from the steppes to the east and the Caucasus to the south-east are more discernible. Rustling of cattle among the different Tripolye communities must have been a constant problem, and, as already suggested, the large open spaces within the gigantic settlements may have functioned, in part, to secure their herds at night. There is no commonly accepted explanation for the demise of the incredibly rich Balkan Chalcolithic settlements (or Kodzadermen-Gumelnitsa-Karanova VI Verband) at the end of the 5th millennium nor, subsequently, for the break up of these gigantic Tripolye settlements and the collapse of the CucuteniTripolye culture into numerous regional post-Tripolye variants or groups (cf. Parzinger 1998a, 465; 1998b). The shifts and dramatic reductions in settlements in both instances are clear; their explanation more difficult. If the demographic calculations for the gigantic Tripolye sites presented above are at all accurate, tens of thousands of people became engaged in adopting an even more
14
PHILIP L. KOHL
extensive, more mobile economy, relying principally on animal husbandry, both cattle-raising and increasingly the herding of sheep and goats (Masson and Merpert 1982, 237) during the second quarter to the middle of the 4th millennium BC. Extensive agriculturalists were becoming pastoralists, changing their way of life and, correspondingly, their material archaeological culture. Raised kurgans were beginning to appear on the steppes, and radical developments were taking place farther south in the Caucasus with the advent of the Maikop-Novosvobodnaya and Kura-Araxes and Velikent cultures. It is roughly at this same time or perhaps slightly later towards the middle of the 4th millennium that the earliest evidence for wheeled transport is documented, stretching across a vast interconnected region from northern Europe to southern Mesopotamia (Bakker et al. 1999). The precise determination of which area or which archaeological culture first developed wheeled vehicles may prove impossible to document archaeologically simply because the technology diffused as rapidly as it did across this vast contiguous area. Clay models of disk wheels have been found at the Late (or post-) Tripolye site of Velyka Slobidka on the Dniester, and two early pre-Pit-Grave kurgan burials with the actual remains of wooden wheels have been found respectively in the lower Don (Koldyri, burial 7, kurgan 14) and Kuban (Starokorsunskaya, burial 18, kurgan 2) areas, and Rassamakin (2002, 53) believes that their appearance in these latter areas was due to ‘the migration or resettlement of groups from the agricultural population’ farther west. The available evidence, in other words, seems to suggest that wheeled vehicles only appeared shortly after the collapse of the gigantic settlements and may indeed have been developed by these same peoples as they began to develop a more mobile economy. Additional evidence is needed to support this interpretation. Ultimately, however, I would argue that the question of the origins of this innovation is much less significant than the phenomenon of convergence, the almost simultaneous evidence for the early use of wheeled vehicles stretching from northern Germany and southern Poland south across Anatolia to southern Mesopotamia, around the middle of the 4th millennium BC or immediately after the collapse of the gigantic Tripolye settlements. Wheeled vehicles can be used for different purposes by different cultures (or different purposes by the same culture) across this interconnected area; they can serve military purposes, function to transport traded goods, such as semi-processed metal ores and ingots, and facilitate the development of a new more mobile way of life based principally on cattleherding.
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
15
The archaeological record very clearly demonstrates the subsequent massive utilisation of wheeled transport on the western Eurasian steppes, particularly in the Kuban area, during the first half of the 3rd millennium BC. Scores of kurgans containing wheeled carts with tripartite wooden wheels have been uncovered. Gei (2000, 176–7) calculates that over 250 burials with wheeled wagons have now been excavated on the western steppes, 115 of them alone being documented for the Novotitorovskaya culture of the Kuban steppes north-west of the Caucasus. These were not the chariots of a military aristocracy but the heavy, ponderous carts and wagons of cowboys who were developing a form of mobile Bronze Age pastoral economy that fundamentally differed from the classic Eurasian nomadism that is later attested historically and ethnographically. The opening up of the real steppes beyond the confines of narrow river valleys, a process archaeologically documented by the spread of kurgans across the steppes, involved two distinct phases: the first was associated with the introduction of heavy oxen-driven wheeled carts and wagons; and the second, which may have begun during the second half of the 3rd millennium BC, was associated with the riding of horses and their harnessing to lighter vehicles, developments which greatly enhanced the mobility of the pastoralists occupying the steppes from the trans-Ural region in the east to the Danube basin in the west. This latter process is first convincingly attested archaeologically across this area by the appearance of disc-shaped cheek-pieces first in the east and then rod-shaped cheek pieces farther west, beginning in late Middle Bronze Age times (Boroffka 1998). As based on the density of such finds, this development may have begun in the Don-Volga forest-steppe zone (Priakhin and Besedin 1999, 40), though this problem of origins may prove as insoluble as that for the development of wheeled vehicles. Once developed, the technology may have spread so rapidly that its precise origins will be incapable of archaeological determination, but again it is significant that this innovation was almost immediately adopted across the west Eurasian steppes and then rapidly diffused to the south and east. Shifts in the Exchange of Metals from the Carpatho-Balkans to the Caucasus and the Rise of the Early Bronze Age Cultures of the Caucasus. Chernykh’s concept (1992, 48–53) of a Carpatho-Balkan Metallurgical Province (CBMP) usefully elucidates the qualitatively distinctive production and exchange of metals that developed in south-eastern Europe during the 5th millennium BC. The
16
PHILIP L. KOHL
CBMP stretched from the northern Balkans to the middle Volga and encompassed five distinct ‘foci’: the northern Balkans; Transylvania and the middle Danube area; the northern Carpathians; the western Black Sea region; and the west Eurasian steppes between the Dnieper and Volga rivers. More recent analytical work (Pernicka et al. 1997, 134) has demonstrated that ‘the ores utilized during the Bulgarian Late Chalcolithic “metal boom” must have come from a variety of sources’ that were exploited at the end of the 5th millennium BC, and the same work has shown that much of the copper found in the Cucuteni-Tripolye culture to the north-east is better associated with the ores found in the Medni Rid area near the Black Sea coast than principally with the Ai Bunar mine (Pernicka et al. 1997, 141), as had been previously suggested. Such an identification, however, is still consistent with a presumed maritime trade in copper to the north-east that entered the same exchange network of other prestige goods, such as Spondylus shells. In short, the exchange of metals within the CBMP to the north-east, possibly as far as sites on the middle Volga, has clearly been demonstrated, though the nature of this exchange and its differential social effects on all the cultures participating in this network are more difficult to assess. The basic picture is one of sophisticated casting metallurgical production in the core resource-rich north Balkan area and less developed metalworking on the north-eastern periphery of the CBMP. This distinction implies that most of the copper that was being exchanged to the north-east must have arrived as a raw material or in semi-finished form where it was subsequently worked by local smiths (Chernykh 1992, 40). There is no evidence for the circulation of metals from the Caucasus or from the southern Urals during this period. Substantial amounts of copper had to move from the northern Balkans east as far as the Volga (Chernykh 1992, 46); some of this movement most likely was by sea, but some of it also had to go overland, at least to get to the lower Volga. Not surprisingly, a general fall-off in the accumulation of copper, gold and silver artefacts can be traced from the core northern Balkan area of production to the distant cultures on the steppes east of the Dnieper. By comparison with the CBMP, the Chalcolithic of the Caucasus or, more precisely, the immediately pre-Maikop and pre-Kura-Araxes horizons of the northern and southern Caucasus appears remarkably impoverished. Such underdevelopment, of course, contrasts sharply with what occurred during the Early Bronze Age when the Caucasus became one of the main suppliers of arsenical bronzes to the peoples of the steppes, particularly to the Pit and Catacomb grave communities. As Chernykh (1992, 159–62) has demonstrated, the north-
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
17
ern Caucasus from Maikop times through the Middle Bronze period may have functioned as the critical intermediary for receiving metals that originated in Transcaucasia and for producing and shipping bronze artefacts to the steppes. In other words, a major shift in intercultural relations began around the middle of the 4th millennium BC in which arsenical coppers and bronzes, many of which originated in the Caucasus, replaced the copper artefacts that had earlier been procured from the Balkans. The advent of the well-known Early Bronze cultures of the Caucasus – the Maikop and Novosvobodnaya cultural formations of the northern Caucasus and the Velikent and Kura-Araxes cultural formations of the north-eastern and southern Caucasus – marked a radical change in the production and exchange of metals throughout the entire interconnected area. Chernykh’s Circumpontic Metallurgical Province (CMP) emerged to replace the no longer functioning Carpatho-Balkan Metallurgical Province. Its emergence and the advent of these Early Bronze Caucasian cultures must be related also to roughly simultaneous developments that occurred farther south involving the greater integration of the Anatolian plateau and western Iran into a larger Mesopotamian world. Colonists and traders from southern Mesopotamia moved north and east far beyond the southern Mesopotamian alluvium, a movement now referred to as the Uruk expansion (Algaze 1993; Stein et al. 1996; Paléorient 1999; Avilova, Antipina and Teneishvili 1999; Rothman 2001; etc.), and greatly influenced the development of the Early Bronze cultures on the edge of the steppes. This paper cannot discuss in detail chronological issues concerning the appearance of the Early Bronze Caucasian cultures and their correspondences with the appearance of ‘Uruk’ settlers in eastern Anatolia. The fit is approximate with the former emerging around the middle of the 4th millennium BC. The high chronology favoured here is based on fairly extensive and consistent series of calibrated radiocarbon determinations (Rassamakin 1999, 163–4; Chernykh 2000, 74–5; Trifonov 2001, 76–7; also for Transcaucasia, cf. Kavtaradze 1983; 1999; and the partial uncalibrated list of Kushnareva 1997, 52, which, of course, must be corrected), as well as typological parallels linking Maikop materials, in particular, with the ‘Uruk’-related remains to the south. Our discussion will be limited to some observations concerning differences between the Early Bronze Maikop and Kura-Araxes cultural formations and will briefly describe some features of the metals produced by these Early Bronze cultures. Munchaev (1994, 178, 174) estimates that roughly 150 Maikop burial complexes have been excavated, while there are only some 30 or fewer known
18
PHILIP L. KOHL
Maikop settlements, only a few of which have been substantially excavated. Korenevskii (2001, 24), who considers the Meshoko-related fortified settlements in the piedmont and mountains with their distinctive ceramics with the so-called ‘pearl’ (zhemchuzhina) ornamentation to pre-date the Maikop formation, reduces the number of known Maikop settlements to 17 (Korenevskii 2001, 11). Whoever is correct, the mortuary assemblage to settlement ratio for Maikop remains is heavily weighted towards the former. This situation is almost the opposite of what is known for the contemporaneous Early Bronze Kura-Araxes cultural formation of Transcaucasia to the south where hundreds of settlements have been found, scores of which have been excavated; very few Kura-Araxes cemeteries have been located and investigated. As Chernykh (1992, 73) is at pains to point out, it is primarily this difference in the nature of the archaeological evidence that explains the apparent wealth of the Maikop metals relative to that of the Kura-Araxes culture. Both areas were working – and probably producing – metals on a large-scale, though we have more evidence from the Maikop culture just because more ‘royal’ kurgans have been uncovered. Indeed, the recent discovery of the grave of the ‘Signore di Arslantepe’ (Frangipane 1998; 2000) with its wealth of bronze weapons, bronze, silver and gold ornaments, and local Mesopotamian-related and ‘Transcaucasian’ (KuraAraxes) vessels underscores the degree to which our knowledge of KuraAraxes metallurgy and social differentiation is partial and distorted. The Maikop ‘culture’ clearly has multiple roots, a classic example of the hybrid nature of an archaeological culture. Many Russian archaeologists (for example, Munchaev 1994, 169) have documented striking ceramic parallels between early Maikop vessels and those found farther south in Syria and northern Mesopotamia (Amuq F and Gawra XII-IX), and it has been claimed that some of the spherical Maikop vessels may have been turned on a slow wheel, a technological development that may also reflect direct borrowing from the south. Perhaps most significantly, microlithic chipped stone tools were found in the great Maikop kurgan, and Munchaev (1994, 170, 189) relates their late presence there to the long-rooted Mesopotamian tradition of depositing such archaic artefacts beneath the floors of public buildings or temples (for example, in the earlier Yarim Tepe II and at Uruk itself); in other words, the fact that such a symbolic Mesopotamian practice is attested in the richest known ‘royal’ or chiefly Maikop burial must have significance not only for the earlier dating of the Maikop culture, but also for determining its cultural affiliation and formation.
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
19
Other scholars have focused on the northern steppe component of the Maikop culture. Most fundamentally, kurgan or raised earthen barrows are not characteristic of northern Mesopotamia, but pre-Maikop kurgans have been excavated in the central northern Caucasus and in the Kuban area. They also appear on the middle and lower Don on sites of the so-called Konstantinovka culture, some materials of which, such as the characteristic asymmetric flint arrowheads, show clear parallels with Maikop remains (Rassamakin 1999, 117–22). The Maikop settlements with their relatively thin cultural deposits, light-framed, clay-plastered wattle-and-daub houses, some of which were supported with wooden posts, and many of which contain numerous pits hardly recall typical Mesopotamian building traditions and techniques. Similarly, the subsistence economy of the Maikop culture, as understood from the excavations of a few of the settlements, seems to have focussed more on animal husbandry, cattle – and, if one considers the Meshoko-related fortified sites in the highlands as Maikop – pig-raising. A subsistence economy based on extensive agriculture and specialised animal husbandry also bespeaks more of a northern steppe connection (ultimately, to be related to the break-up of the Tripolye settlements?), than a southern Near Eastern heritage. The Maikop culture clearly has multiple origins or is syncretic in character, with local roots that extend naturally north onto the steppes and with surprisingly close and novel connections with northern Mesopotamia. The wealth of the metals – arsenical bronzes and silver and gold artefacts – found in the Maikop ‘royal’ kurgans and hoards is truly extraordinary, leading Chernykh (1992, 142–4) to reflect on the ‘problem of gold’ at this time. Indeed, if we trace the occurrence of gold in the area of our concern, we see a conspicuous shift from north to south that continues through Middle Bronze times: the early Chalcolithic florescence of gold consumption in the Balkans, particularly in the Varna cemetery; the abundance of gold (and silver) objects in the Maikop kurgans of the north-western Caucasus during the Early Bronze period; and the spectacular discoveries of precious gold and, to a lesser extent, silver objects in the monumental ‘royal’ kurgans of Transcaucasia and the famous hoards of Anatolia during the late Early and Middle Bronze periods. Although, undoubtedly, accidents of discovery play a part here, the trend is unmistakable and must reflect underlying historical processes. For example, Avilova, Antonova and Teneishvili (1999, 57–8) calculate that approximately 7400 gold and 1000 silver artefacts have been found in Maikop-related kurgans in the north-western Caucasus. Such precious artefacts practically disappear in the northern Caucasus
20
PHILIP L. KOHL
towards the middle of the 3rd millennium BC, while at the same time the number of gold and silver artefacts in Anatolia and Transcaucasia (and, not incidentally, in Mesopotamia, such as at the Royal Cemetery at Ur) sharply rises (calculated at about 32000 objects). This shift not only reflects changes in the production and supply of precious metals, but also the movements of peoples with their leaders or chiefs south, across or around the Great Caucasus range. The Maikop arsenical bronzes include not only ceremonial prestige weapons, which were potentially also useable, such as ribbed tanged daggers and shafthole axes, and ornaments, but also functional tools, such as substantial hoes found on Maikop settlements and chisels and awls, and large distinctive cauldrons; characteristic objects of uncertain significance include the so-called twisted circular ‘psalia’ (Munchaev 1994, 211; for a different interpretation, cf. Trifonov 1987) and the large, pitch-fork-like shafted hooks (or kryuki, Munchaev 1994, 207). Chernykh’s work has shown that the Maikop bronzes could be divided into two groups: copper-arsenic alloys; and copper-arsenic-nickel alloys, and he has postulated that the sources for the former were appropriate ore deposits in Transcaucasia, and for the latter, deposits located farther south, possibly in Anatolia, that were also utilised by the Mesopotamians. He (1992, 159–60) refers to the ‘North Caucasian Bridge’, which brought metals, presumably as ingots or in semi-worked form, across the Caucasus, and explains the wealth of the Maikop chiefs as associated with their unique role as intermediaries in this south-north metals trade, supplying vast areas of the steppes to the north and east with Caucasian-derived bronzes. This reconstruction probably simplifies a much more complex reality; at the very least, it requires more documentation based on systematic archaeo-metallurgical research in the northern Caucasus. What is striking is the difference between the contemporaneous Kura-Araxes and Maikop cultures in nearly all their material remains (from settlement patterns and domestic architecture to their subsistence economies and their metal assemblages). There is very little very evidence for direct contact between them (though some contact is possibly suggested in the Krasnodar reservoir settlements currently being investigated (Lyonnet, personal communication) and possibly also at the Lugovo settlement in Ingusheti. These two Early Bronze cultural formations are totally distinct phenomena, though part of the same overarching, interconnected system within the Circumpontic Metallurgical Province. It is only during this Early Bronze period when such a cultural divide so sharply separates the material remains of the northern and southern Caucasus. Later, the high Caucasus range appears to have been more porous with
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
21
greater evidence for connections on either side of the mountains (as evident in the close similarity between the later Colchidean Late Bronze culture of western Georgia and the Koban culture first documented in northern Ossetia and then south of the Caucasus in the Tli cemetery). This later Caucasian pattern of interaction had not yet been established during the Early Bronze period, and one can only speculate as to the nature of the relations – possibly hostile and/or competitive – between the bearers of the distinct Maikop and KuraAraxes cultural communities. The separation did not last. Some of the Kura-Araxes peoples dispersed out of Transcaucasia and settled far to the south-west and south-east, and other peoples with oxen-driven wheeled wagons came off the steppes and the north Caucasian plains and crossed the mountains and/or moved across the Caspian littoral plain to enter the rich valleys of eastern Transcaucasia. It is to these movements that we now turn. The Dispersal of the Kura-Araxes Peoples and the Advent of ‘Royal’ Kurgans in Transcaucasia. Whether it was the search for more arable land to support their burgeoning populations and/or their displacement with the arrival of new groups from the north with four-wheeled, oxen-driven wagons, the KuraAraxes peoples moved over some extended period during the late 4th and the early 3rd millennium BC far to the south-west across the Anatolian plateau to the Amuq plain and beyond to today’s northern Israel and to the south-east into north-western Iran and along the Zagros mountains and onto the Iranian plateau as least as far as Kermanshah. This spread of ‘Early Transcaucasian’ settlements has long fascinated archaeologists (see, for example, the map in Roaf 1990, 80), many of them speculating on the ethnic/linguistic identity of these migrants and interpreting them as ancestral to Hurrians or Hittites or other later historically attested peoples (for instance, Woolley 1953, 31–7). A. Sagona (1984) has published the most complete list of Kura-Araxes sites, arguing for the movement of these colonists first out of central Georgia (Kvemo and Shida Kartli) to the south followed by the development of distinctive regional traditions (Armenian, upper Euphrates, Khirbet Kerak) and then a subsequent spread to the north-east (Daghestan) and south-east into western Iran. Others have placed the beginnings of the Kura-Araxes cultures along the middle Araxes valley on the Ararat plain and in Nakhichevan with its subsequent spread first to the north and then south. Nearly everyone sees the ‘homeland’ of this culture in the area between the middle Kura and middle Araxes rivers in Transcaucasia, a view that implies
22
PHILIP L. KOHL
that Kura-Araxes materials found in other areas are intrusive in the local sequences. Indeed, many sites in eastern Anatolia (Norsuntepe, Arslantepe) and western Iran (Yanik Tepe, Godin Tepe) exhibit a sharp break in material remains, including new forms of architecture and domestic dwellings, and such changes support the interpretation of a spread from north to south. This general view may be true, but it useful to recall that modern political borders rarely define prehistoric culture areas, and regions immediately adjacent to southern Transcaucasia (i.e. south of the middle Araxes river) should also be seen as part of the formative area for this culture. While it is unclear precisely where the Kura-Araxes culture originated within Transcaucasia, the archaeological record does strongly support a movement of peoples north to south across a very extensive part of the ancient Near East during the late 4th to the middle of the 3rd millennium BC. That is, while migrations are notoriously difficult to document on archaeological evidence, these materials constitute one of the best examples of prehistoric movements of peoples available for the Early Bronze period. This dispersal needs to be restudied intensively; here only a few general observations can be made. First of all, calibrated radiocarbon dates are beginning to yield a consistent picture for the timing of this dispersal. The relevant VIB period at the extensively excavated site of Arslantepe near Malatya dates ca. 2900–2700 BC, a date which essentially is supported by evidence from neighbouring sites such as Norsuntepe (Di Nocera 2000, 75–6); it also apparently is consistent with sites in the Erzurum region farther east (Sagona 2000, 333). Related Khirbet Kerak materials from northern Israel have been dated roughly from 2700–2450 BC (de Miroschedji 2000, 258), suggesting an initial dispersal into the Upper Euphrates basin at the very beginning of the 3rd millennium (and after the collapse of the Uruk expansion), followed by a subsequent movement to the south-west in the second quarter of the 3rd millennium. The chronological gap between the appearance of Khirbet Kerak ware, also known as Red-Black Burnished ware (Braidwood and Braidwood 1960), ware on the Amuq plain and its appearance in the southern Levant may have been somewhat overestimated. New relevant radiocarbon dates from the southern Levant suggest that Khirbet Kerak ware may first have appeared ca. 2800–2700 BC or almost simultaneous with its appearance farther north (Philip and Millard 2000, 284). The overall pattern seems reasonably clear: an initial spread across eastern Anatolia to the upper Euphrates basin at the very end of the 4th and beginning of the 3rd millennium followed by a relatively rapid diffusion (during the course of a cen-
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
23
tury or so?) farther south-west ultimately to the eastern Mediterranean coast. Sites in the Urmia basin (Geoy Tepe and Yanik Tepe, for example) seem to have been occupied already in the last centuries of the 4th millennium (Voigt and Dyson 1992, I, 137). ‘Early Transcaucasian’ materials appear to be intrusive in this region; i.e. they represent a break with earlier Chalcolithic remains on these sites, but this movement appears to predate the spread into the upper Euphrates area. One can only speculate that the lack of an Uruk presence in north-western Iran may have facilitated this earlier movement to the east. Their spread farther south into central-western Iran occurred later, though precisely how much later is still unclear. Carbon dates are unavailable for the beginning of the relevant Godin IV period, though the excavators (Weiss and Young 1975, 2) believed that there was only a short break in the sequence between this period and the underlying Godin V period that can be dated to the last centuries of the 4th millennium. The assumption is that the Godin IV occupation began in the early 3rd millennium BC. Here too one wonders whether there is a causal relationship: the collapse of the ‘Uruk outpost’ at Godin V (Algaze 1993, 60), and the arrival of Transcaucasian colonists from the north. Secondly, these ‘peoples of the hills’ may have consciously avoided large settled areas controlled by stronger polities on the northern Mesopotamian plain. There were empty places they could settle and others that they could even possibly destroy. Still others were left alone presumably because they were environmentally alien and/or more densely occupied by polities more powerful than theirs. While our knowledge of the distribution of the sites containing Kura-Araxes materials throughout these different regions is obviously dependent upon the nature and extent of the surveys and excavations conducted throughout these different regions, which manifestly are not commensurate with one another, it also seems clear that not all contiguous zones were equally affected by these dispersals. The spread was not continuous and there are clear gaps in the distribution of sites containing these materials, such as in Lebanon between the Amuq plain and northern Israel. These gaps, to some extent, must reflect the historical reality that not all areas were settled. It is also obvious that for the most part these dispersals do not represent armed military invasions and that the movements involved considerable assimilation with pre-existing local traditions, exacerbating the archaeologists’ task of recognising them. Populations expanded and intermingled with one another. In these processes, social structures obviously must have changed. It is an
24
PHILIP L. KOHL
archaeological truism today to note that pottery styles do not equate with peoples, and the temptation to do so must be resisted. Nevertheless, the very frequency of distinctive, seemingly intrusive ceramics and other items of material culture, such as the highly specific andirons, suggest that this phenomenon, however short-lived, must have been reasonably substantial. At Beth Shean, for example, the Khirbet Kerak pottery constitutes more than 60% of the total ceramic assemblage in levels 11–9 before dropping off to 38% in level 8 and essentially disappearing in level 7 (as summarised in de Miroschedji 2000, 259). At the type site of Khirbet Kerak (Beth Yerah), these wares constituted 20–30% of the sherds found on the site. The site itself is 20–25 ha in size or considerably larger than any known Kura-Araxes site in Transcaucasia. Site size too, as we have seen with the gigantic Tripolye settlements, cannot simplistically be equated with social complexity. The data, however, are suggestive that the ‘peoples of the hills’ transformed themselves as they spread across large areas of the ancient Near East. It is unclear what was driving these dispersals. Possibly, they were in search of new sources of metal in Jordan or in Cyprus (cf. the recently excavated Kura-Araxes-related hearth stands and evidence for migrants from south-western Anatolia at the Early Bronze Age site of Marki Alonia: Frankel 2000; Frankel and Webb 2000; Webb and Frankel 1999). They may have been skilled metallurgists, but why leave a metalliferrous region like the Caucasus for unknown sources? Moreover, Khirbet Kerak materials are not found in the metal-bearing Wadi Feinan area south of the Dead Sea (de Miroschedji, 2000, 264). Perhaps they were simply in search of more and better arable land with natural population increases, replicating on a much larger scale the movements from the highlands to the plains that may have characterised the initial spread of Kura-Araxes settlements within Transcaucasia? Possibly, but why did they move and not others? Another factor may also have been at work. Some peoples were not only moving south out of the Caucasus, but others seem to have been moving into Transcaucasia from the north – at least at some point in the first half of the 3rd millennium (see the new calibrated dates for the ‘early kurgan cultures’ of Transcaucasia: Kavtaradze 1999, 81; and the discussion in Trifonov 2001, 79–80). It is hard to distinguish cause from effect here: did peoples move into the rich Alazani and Kura valleys because others had moved out or were the Kura-Araxes peoples moving south due to the incursions of peoples from farther north? Current evidence suggests that the dispersal to the south of the
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
25
Kura-Araxes culture predated the arrival of the peoples burying their leaders in large ‘royal’ kurgans, but we must remember that all these movements represented protracted processes, not events, that took place during several centuries. The higher dating of the earliest ‘royal’ kurgans in eastern Georgia suggests that these processes may partially overlap with one another and that, consequently, there may have been some synergistic relationship between them. Pit-Grave (Yamnaya culture) raised burial mounds or kurgans are documented across the western Eurasian steppes at least from the Dnieper to the Volga by the second half of the 4th millennium BC, and related cultures, such as Maikop/Novosvobodnaya and the later Novotitorovskaya (Gei 2000), with rich material inventories, are known primarily from their kurgan mortuary remains and are also found in the north-western Caucasus, particularly in the Kuban basin. Their spread to the north-eastern Caucasus is less well documented, and many kurgans, undoubtedly, have been levelled during the past 50 years with intensive modern settlement and agricultural exploitation. The Caspian littoral plain, which forms the primary north-south corridor linking the steppes to Transcaucasia and the Anatolian and Iranian plateaus, once was dotted with kurgans dating to various periods, but very few have been excavated. Maikoprelated materials have been found in a handful of kurgans in Daghestan, such as at the Large Miatli kurgan along the Sulak river and at Torpakh-Kala along the coastal plain south of Velikent. This evidence is hardly conclusive but has been cautiously interpreted as possibly suggesting a north-west to south-east movement of Maikop peoples during the final stages of this culture’s existence (cf. Magomedov 1991, 34–5). Currently available settlement pattern evidence suggests that the Caspian plain extending south from Velikent to northern Azerbaijan may have been largely abandoned at the end of the 3rd or beginning of the 2nd millennium BC, though it is unclear whether the local inhabitants of the plain retreated into the mountains to the west for ecological reasons and/or were displaced south or north due to the periodic movements of peoples off the steppes and down the coastal plain into Transcaucasia, movements that may have begun roughly 500 years earlier. These movements still require more extensive archaeological documentation, but the theory is potentially testable by systematically excavating several kurgans that still exist. It is likely that such postulated migrations did not represent armed invasions so much as cattleherders moving south with their families on oxen-driven wagons, as documented particularly on Novotitorovskaya culture sites to the north-west, in search of better pastures for their animals.
26
PHILIP L. KOHL
What is clear is that there was a sharp break in the distribution of settlements in Transcaucasia that occurred during the late Early Bronze period, possibly beginning in the second quarter of the 3rd millennium BC; specifically, the dense distribution of Kura-Araxes settlements was followed by a much more sparse distribution of known settlements and a sharp increase in burial sites during the late Early and Middle Bronze periods. The Kura-Araxes or Early Transcaucasian pattern of an abundance of known settlements and few known mortuary sites is followed by a precipitous decline in known settlements and the advent of essentially new, highly visible raised burial mounds. Moreover, the post Kura-Araxes settlements that have been investigated, such as the site of Uch-depe in Azerbaijan, are impressively fortified on a scale not characteristic of the earlier Kura-Araxes settlements, and there is more diversity evident in the material culture remains, as is evident in the establishment of five Middle Bronze overlapping archaeological cultures: Western, Trialeti, Karmirberd, Kizylvank and Sevan-Uzerlik (Kushnareva 1997, 84–5). These supplanted the earlier, more homogeneous Kura-Araxes remains. Large numbers of settlements only reappear much later in Transcaucasia during Late Bronze times, presumably in the second half of the 2nd millennia BC, and they are quite distinctive in form, size and location from the Early Bronze Kura-Araxes settlements. What is the significance of such a sharp shift in settlement patterns? Some raised burial mounds have been found in association with late Kura-Araxes settlements (for instance at Satkhe in Djavakheti), but they are not typical. Clusters of kurgans (or kurgan cemeteries) only appear subsequently in post KuraAraxes times, and some of these (for example, Tsnori and Martkopi) contain a wealth of precious materials (such as gold and silver vessels and ornaments), traces of oxen-driven wagons, and possibly even evidence for human sacrifice. Kushnareva (1997, 229–33) has convincingly reconstructed how the social structure of Transcaucasian Middle Bronze communities was transformed with these developments. The oxen-driven wheeled wagons and carts found in these early Transcaucasian kurgans clearly are related to the earlier vehicles documented in the north-western Caucasus and farther north on the Eurasian steppes, and the impressive numbers of gold and silver artefacts from these early ‘royal’ Transcaucasian kurgans also find earlier parallels to the northwest in the kurgans of the Maikop/Novosvobodnaya cultures. In short, while there are still gaps in the archaeological trail of evidence, a general movement of peoples migrating north to south across the Caucasus or around the western shore of the Caspian can be discerned. The discontinuity in the settlement
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
27
pattern evidence from Transcaucasia suggests that these movements were not sudden events, but processes that took place over some considerable period of time, continuing throughout the latter half of the 3rd millennium and well into the 2nd millennium BC. The final process we wish to adumbrate here concerns the movements of peoples north to south, from the steppes to the agricultural oases of Central Asia on the other side of the Caspian Sea. From the Eurasian Steppes to the Plains of Central Asia, the Steppe Component to the Bactria-Margiana Archaeological Complex (BMAC). Eurasian steppe nomadism has also been termed multi-animal nomadism (Bacon 1954, 46), because typically different species of animals – sheep, goats, horses, cattle, Bactrian camels – were herded together, some of whose value lay principally in transportation and riding. Eurasian nomadism had to adapt to severe continental climatic conditions: a short but very hot and intense summer and a much longer, bitterly cold winter during which time pastures were often covered with thick deposits of snow. Crops can be grown, though evidence for any form of intensive cultivation on the steppes during the Bronze Age – at least until the middle of the 2nd millennium BC when the climate temporarily became more humid – is contradictory (cf. Cernych et al. 1998; Gaiduchenko 1999; Gerskovic 1999, 82–9). Typically, whatever cultivation practices may have been adopted, the growing season had to have been very short, particularly relative to what was possible farther south in Transcaucasia and in southern Central Asia. These continental conditions became more severe, of course, as one crossed the Urals and moved deeper into Eurasia. Horses and, to a lesser extent, sheep were particularly valuable for the fact that they could uncover grass covered with snow by kicking at it with their hooves; apparently horses could reach grass in this manner that was buried up to a half metre in snow (Khazanov 1994, 50). Pastures so uncovered could, to some extent, be utilised by other less well-adapted species, thus eventually leading to the mixed herds characteristic of Eurasian nomadism. Another solution to this problem was to grow fodder and store it for the winter season when it was less available; this solution must have archaeological consequences that are still largely uninvestigated. Another problem on the open steppe was the availability of water; thus, not surprisingly, many of the migratory cycles of recorded nomadic groups followed the river valleys, moving meridionally north possibly into the forest-steppe zone in the summer and south into warmer, less snow-covered climes in the
28
PHILIP L. KOHL
winter. Such a pattern would bring herders in the Volga and Ural river basins annually down to the northern Caspian Sea. Ultimately, their increasing presence would successively displace other herders and cultivators farther south in a kind of chain reaction, until peoples began to traverse the Kyzyl and Kara Kum deserts to reach more hospitable conditions. N. Shishlina (2003, 364–5) believes that the open steppe was only progressively occupied. In Yamnaya (PitGrave) times at the end of the 4th and the beginning of the 3rd millennium, the movement of the herders was largely confined to the river valleys and immediately surrounding grasslands; the real open steppe was only occupied during subsequent Catacomb grave times out of necessity (increased aridity and overexploitation of the neighbouring grasslands) and due to their increased mobility and mastery of the horse (cf. also Shishlina and Hiebert 1998). Major rivers such as the Volga, and even the shallow Sea of Azov freeze during the winter, thus making it possible to cross them without difficulty. East of the Urals and the Ural river basin, the major rivers run south to north, and the climate becomes even more continental and severe, and the typical snow cover gets even deeper. It is easy to understand why the opening up of the eastern Eurasian steppe occurred centuries after the opening up of the western Eurasian steppe. The basic point to be emphasised here is that this classic form of Eurasian nomadism, as known to us historically and ethnographically, is not archaeologically attested during the Bronze Age. Morales-Muñiz and Antipina’s exhaustive review (2003) of the faunal evidence from the steppes from Chalcolithic through Bronze Age times records very little presence of sheep and goats, except at some post-Tripol’ye Usatovo sites, and a continuing focus on cattle that seems only to increase significantly during the Late Bronze Age, an increase that they intriguingly suggest may be related to the development of horse riding for cattle-herding purposes, the emergence of cowboys on the ‘Wild East’ of the Eurasian steppes. If Bronze Age steppe pastoralism is not the same as historic Eurasian mounted pastoral nomadism, then it is misleading to envision hordes of marauding nomads sweeping down off the steppes with their chariots and advanced bronze weaponry to invade and subjugate established agricultural societies to the south. Rather, more mobile semi-nomadic economies utilising oxen-driven carts and wagons, and herding principally cattle spread across the western Eurasian steppes during the second half of the 4th through the early 2nd millennia BC. At some point, probably fairly late in this process, they began to ride horses and develop lighter vehicles and new techniques for harnessing horses to them. Their way of life was never fully
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
29
nomadic, though obviously successful. As this way of life spread farther east into more continental, colder areas, these cattle-herders must have experienced a continuous pressure to move farther south into warmer climes more suitable for raising their livestock. This process would have been gradual, but relatively continuous leading to the movements or, perhaps better, successive displacements of cattle-herders north to south ultimately into areas where permanent settled agriculture had been practised for millennia. West of the Caspian this process seems to have been underway at least by the middle of the 3rd millennium BC with the advent of the impressively large kurgan burials in Transcaucasia. East of the Caspian this process was delayed for several centuries, a chronological discrepancy that can also be partially explained by the greater physical separation of the steppes from the cultivated oases of southern Central Asia. The deserts of the Kyzyl Kum and Kara Kum initially impeded this movement of herders north to south. In other words, while the Caucasus mountains form a sharp geographic boundary between the Eurasian steppes to the north and the highland plateaus of Anatolia and Iran to the south, the Central Asian deserts extend the range of transition between the steppes and the highlands and find no real parallel on the western side of the Caspian, though the dry Nogai and Kalmyk steppes also merge with each other and extend north around the Caspian. The Central Asian deserts functioned as a more effective cultural barrier until late in the 3rd millennium BC when the semi-sedentary cattle-herders in the Volga and Ural river basins and even farther east were able to traverse these extensive arid expanses by developing even more mobile economies with the help of horses and, most likely, Bactrian camels. These latter had earlier been domesticated in southern Central Asia and eastern Iran. It is therefore not surprising that contacts between the western Eurasian steppes and the ancient Near East developed earlier on the western side of the Caspian since the steppes and the Caucasus mountains, particularly in the north-west along the Kuban basin, were directly contiguous with one another, not separated by waterless deserts, the only problem being how to get around or over the mountains. One can visualise this important physical distinction between steppe and sown on either side of the Caspian as like a wedge opening west to east in which the north-west Caucasus mountains practically touch the Crimean peninsula and east Ukrainian steppes, while the oasis irrigation agricultural settlements of southern Central Asia are physically removed from the vast steppes of southern Russia and northern Kazakhstan by the formidable Kyzyl Kum and Kara Kum deserts.
30
PHILIP L. KOHL
While interaction with the steppes in terms of the gradual but continual movements of cattle-herders north to south took place later east of the Caspian Sea, its precise beginnings are uncertain due to the somewhat floating, unanchored chronology of Bronze Age steppe remains, particularly for sites located east of the Urals. The excavation of early Andronovo ceramics with Abashevo and Petrovo elements in association with Sarazm IV ceramics at Tugaï along the right bank of the Zeravshan river immediately east of Samarkand (Avanessova 1996) suggests that this process may have begun already by the middle of the 3rd millennium. Interestingly, the limited salvage work at this site also yielded considerable evidence for metal-working (crucibles, kilns, slag, etc.), a characteristic practice of the later Andronovo-related groups that moved south into Central Asia during the 2nd millennium and worked the tin mines of the Zeravshan valley (Alimov et al. 1998). The evidence for a steppe presence in southern Central Asia becomes more substantial with the emergence of the Bactria-Margiana Archaeological Complex (BMAC) at the end of the 3rd millennium BC, as documented by the discovery of scatters of ‘nomadic’ or ‘incised coarse ware’ sherds in the Murghab delta, as well as by their occasional occurrence in the excavations of the public buildings in the major centres in the Togolok, Gonur and Dashly oases. The origins of the BMAC are complex and cannot be satisfactorily addressed until important but inadequately investigated regions, such as Pakistani Baluchistan and south-western Afghanistan, are more systematically studied. Most archaeologists of the former Soviet Union initially seized upon the obvious parallels in ceramics and figurines between the BMAC materials, on the one hand, and those found on multi-period tells in the Kopet Dag piedmont strip of southern Turkmenistan, on the other, to argue for a strong formative influence from this latter region. Indeed, the very fact that BMAC remains were first described by Soviet scholars as Namazga VI underlies their belief in this basic genetic connection to the millennia-long developmental sequence established by Soviet archaeologists in southern Turkmenistan. Not surprisingly, Western scholars have observed similarities with other areas and in other materials, including earlier stone and metal seals and public architecture, from sites located to the south in eastern Iran, Afghanistan and Pakistani Baluchistan to suggest southern roots for the BMAC from these regions, if not also from the Indus valley itself. Undoubtedly, the BMAC had diverse origins and both these perspectives can be maintained; they are not mutually exclusive or, as B. Lyonnet (n.d., 48) observes, the BMAC represents ‘une culture nourrie des deux poles, “turkmène” et “indien”.’ Besides those two culture areas, there
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
31
is at least one other substantial influence on the BMAC: cattle-herding pastoralists from farther north settling down on the watered plains of Bactria and Margiana and changing their way of life and adopting the material culture of agriculturalists with whom they came into contact. Current archaeological evidence, as limited and problematic as it is, documents precisely the transformation described above; that is, the northern component of the BMAC, which constitutes only one of its substantial influences, has been detected archaeologically. What has been found is just what one would expect to show the gradual, but continuous infiltration and assimilation of cattle-herders into the established sown world of irrigation agriculturalists. Settlement pattern data for southern Central Asia, such as it is (cf. the calculations and caveats in Kohl 1984, 143–6, 151–4, 159–60; also now the more systematic work in the Murghab delta by Gubaev, Koshelenko and Tosi [1998]), supports the following conclusions: 1. There is scant evidence for occupation of the lowland plains of Margiana prior to the Middle Bronze period or datable to the second half of the 3rd millennium and later periods (Salvatori 1998, 52). 2. The total known occupied area for the plains of Margiana and northern and southern Bactria during Middle and Late Bronze Age times considerably exceeds the known occupied area for earlier Chalcolithic and Bronze Age remains from the piedmont strip of southern Turkmenistan or even also north-eastern Iran (excluding the Gorgan plain), a fact that essentially precludes the possibility of deriving the former exclusively from the latter, as has been postulated (by Biscione 1977, for example). 3. The most notable disjunction in the settlement pattern data in terms of location, size and nature of settlements from southern Central Asia from Neolithic through Iron Age times or throughout later prehistory occurs precisely during this initial major occupation of the lowland plains of Bactria and Margiana beginning in the last centuries of the 3rd millennium BC. The lowland plains watered by the lower Tedjen river and its terminal branches and the lower Murghab were occupied during Late Chalcolithic times (for example, the Geoksyur oasis and scattered Namazga III materials in the Kelleli oasis of Margiana) probably by settlers from southern Turkmenistan, but this occupation was restricted as the recent survey work has demonstrated. Evidence for an earlier occupation of the Bactrian plains west of the Kunduz valley or west of eastern Bactria: where such occupation has been documented
32
PHILIP L. KOHL
at Shortughaï and other settlements on the Ai Khanum and Taluqan plains (Lyonnet 1997, 68–71) is even less apparent, save for the scatters of ‘Early Neolithic’ lithic materials found in the sands north of the Bronze Age oases of northwestern Afghanistan (Vinogradov 1979). Salvatori believes that this picture of a relatively late Middle and Late Bronze occupation of the lowland plains of Margiana and Bactria is unlikely, and represents a product of insufficient research and the natural burial of earlier settlements: In our opinion only a wrong turning taken by the research or conditions of burial under alluvial deposits or the advancing sand dunes (in the distal areas of the delta) have so far stood in the way of the identification of the presence of an earlier settlement pattern. (Salvatori 1998, 52)
His view, of course, may be correct, though current negative evidence, which is based partly on the results of his own exhaustive and systematic survey work, suggests otherwise. The lack of substantial prior settlement on the lowland plains of Margiana and Bactria prior to the BMAC or initial Middle to Late Bronze occupation is consistent with the thesis of a gradual peopling of these plains by cattleherders coming from the north to settle down and practice a more secure and stable way of life. Salvatori adopts an evolutionary perspective: there simply should be earlier substantial settlement, even though it has yet to be documented and even though a more satisfactory and consistent historical explanation suggests itself. It is impossible to calculate precisely the increase in settled area and its overall extent during BMAC times. One major unresolved problem is the contemporaneity of the occupation of the various oases. If Russian archaeologists, like Sarianidi, are correct in arguing for the sequential occupation of the oases and the general movement of settlement in Margiana upstream over time (from Kelleli to Gonur to Togolok to Takhirbai), then the total occupation in any specific period correspondingly would be reduced (cf. Lyonnet n.d., 53–4). It is unclear to what extent this perspective is correct or to what extent it simplifies a much more complex reality (for example, Gonurdepe north was occupied during at least two periods: Hiebert 1994, 36–8). Nevertheless, when one compares the known settlement pattern data for southern Turkmenistan in the Early and Middle Bronze periods (traditionally Namazga III–IV) with that documented for the principally later BMAC settlements, there is a dramatic and marked increase in total settled area. It is difficult to consider such an increase as solely the product of a local developmental process.
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
33
The BMAC synthesis is highly original, even if some of its practices (such as diverse mortuary rituals or the use of seals) find their roots farther south in Baluchistan and eastern Iran or west in southern Turkmenistan. However one interprets the planned public architecture on BMAC sites, its distinctive, easily recognisable character is apparent. Most BMAC sites are at least to some extent fortified, a feature that was not characteristic for most earlier settlements in southern Turkmenistan, and the very presence of such fortifications suggests, of course, that the settling of these plains was not an entirely peaceful process. The conflict may not have been so much that between steppe and sown, as later arrivals ousting or displacing earlier peoples who had moved into the area. Sometimes the new immigrants may simply have occupied settlements that had been earlier abandoned. Many BMAC sites, such as the Dashly 3 ‘fortress’ (Sarianidi 1977) and the later ‘fort’ at Gonur South (Sarianidi 1998, 115), record later occupations by peoples who significantly altered the pre-existing architecture. These sites often contain considerable evidence for metalworking (Sarianidi, Terekhova and Chernykh 1977), a trait that also may reflect northern origins. Sites in the Kopet Dag piedmont strip of southern Turkmenistan typically form multi-period tells that were occupied in some cases for millennia. While the larger excavated sites in Margiana exhibit a cultural sequence and were occupied for some period of time, the total depth of cultural deposit on the BMAC sites never approaches that of the larger tells in southern Turkmenistan or north-eastern Iran. The ‘deep sounding’ of a stratified midden at Gonur North extended down 3.5 m before reaching virgin soil (Hiebert 1994, 30), while the total depth of cultural deposit at Altyn-depe was estimated as at least 30 m (Masson 1981, 75). The BMAC sites exhibit planned, highly symmetrically arranged architecture. By comparison, the proto-urban settlement at Altyn-depe with its winding streets and distinct residential and functional areas (Masson 1988, 5, fig. 1) seems to have evolved organically over time. It was not preplanned in the same way as the major BMAC settlements and as even smaller sites such as Sapallitepa obviously were. This contrast of BMAC features with other areas could be extended, the point being that the BMAC represents something original, a unique blend of diverse cultural elements. How convincing is the empirical support for this thesis of a gradual settling of Margiana and Bactria by northern cattle-herders? Do a total of 336 incised coarse ware sherds from 34 ‘nomadic camp sites’ (Cerasetti 1998, 67) constitute sufficient proof for such postulated movements? Evidence for the Indo-Aryans – if that is the correct ethnic/linguistic identification for these cattle-herders –
34
PHILIP L. KOHL
has never seemed so meagre and puny. Or does it? What would one expect to find archaeologically to document the process of gradual but continuous movements of cattle-herders onto the lowland plains of Margiana and Bactria? The plains these settlers entered were not totally empty, but sparsely occupied by peoples practicing a form of irrigation agriculture that had been developed over preceding millennia. As they entered these plains, their way of life changed; they focused more on irrigation agriculture, adopting and assimilating the preexisting material culture of the neighbouring peoples they encountered. The process is the inverse of that already adumbrated for the collapse of the gigantic Tripolye settlements. In the latter case, agriculturalists who also herded became pastoralists who also cultivated, adopting an ever more mobile and extensive economy. Here on the plains of Bactria and Margiana, cattle-herders settled down and began to cultivate crops more intensively than they had previously practiced farther north. This process of sedentarisation and assimilation undoubtedly was very complex. The peoples living in the Tazabag’yab culture settlements, which have been excavated in the Akchadarya delta of the lower Amu Darya south of the Aral Sea, clearly practiced irrigation agriculture (Itina 1977, 44–5). Their practice may or may not reflect influence or even some initial colonisation from the south (Itina 1977, 229–31), but this culture also exhibits clear links to the steppes farther north and represents a rapid development from the earlier ‘Neolithic’ Kelteminar culture documented in the area. The ‘steppe’ or ‘coarse incised’ ceramics of the Tazabag’yab culture are those most typically found in Margiana and Bactria, suggesting that these peoples were those first displaced farther south. They already knew how to survive in the new environment that they entered. Later movements may have been more direct with ever more mobile pastoralists moving from the southern Urals and other areas on the Eurasian steppes into the settled oases of southern Central Asia. The important point is that initially at least these movements did not represent armed invasions but protracted, largely peaceful infiltrations and encounters with earlier established cultures, the product being the emergence of a distinctive new archaeological culture or phenomenon, the BMAC, with its diverse multi-cultural roots. The Most Elusive Quest in Steppe Archaeology: The Search for Indo-Europeans/Aryans We have painted a very broad canvas chronologically ranging over more than two millennia and spatially extending from the Balkans to the plains of
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
35
Central Asia. Is it possible to identify linguistically and ethnically the peoples responsible for these interconnected processes? Is it even worthwhile to attempt such identifications on the basis of the mute prehistoric record? The gigantic settlements of the terminal Tripolye culture represent the culmination of ‘Old Europe’, an agricultural way of life, the origins of which can be traced back to the Balkans and, ultimately, to Anatolia. The spectacular Chalcolithic remains of the Balkans clearly emerged out of the earlier and equally spectacular Neolithic remains of Anatolia. During the first half of the 4th millennium BC Tripolye cultivators became cattle-herders and spread eastwards across the Eurasian steppes. Their focus on cattle-herding, as opposed to the mixed breed herding later characteristic of the Eurasian steppes, only increased during the Middle and Late Bronze periods, and their way of life, which fundamentally revolved around their cattle, must have been emphasised in all the material and spiritual aspects of their culture. Gradually and continuously they moved south and changed the ethnic compositions and, most likely, the languages spoken in the areas that they entered. The easily recognisable steppe elements in their material culture were transformed. Wheel-made pottery replaced the coarsely incised hand-made pottery, though these peoples continued to exploit metal resources and produce increasingly sophisticated and effective metal tools and weapons. If we extend our chronological horizon to encompass the 2nd millennium BC and our spatial parameters to include the Iranian plateau and the South Asian subcontinent, can we then paint a picture showing the Indo-Iranians entering Iran and the Indo-Aryans moving into the subcontinent? Unfortunately, the figures in our painting will never appear in such sharp focus – even if the essential identifying historical inscriptions are discovered. If cultures (and languages, peoples, ethnicities, etc.) are socially constructed and never made, but always in the making with diverse roots, then there is no single Indo-European or proto-Indo-European ‘homeland’. The story we have sketched can be extended spatially and chronologically, but it will always constitute an ever-unfolding process of development in which peoples continuously transform themselves, including their basic livelihoods. Sometimes they move and enter new areas, but they also continuously borrow and assimilate the technological innovations and cultural developments of other peoples with whom they always come into contact. The precise origin of developments in metal working, wheeled vehicles and horse riding may never be established precisely because of the rapid sharing and adoption of these incredibly useful and multifunctional innovations. What is more significant is the extent of the
36
PHILIP L. KOHL
interconnections suggested by the huge area – over which, for example, wheeled vehicles seem to appear simultaneously. It is inherently unlikely that all these developments were the products of a single, particularly gifted ethnic or linguistic group. The herders who initially left the Eurasian steppes were not mounted warriors led by a chariot-riding aristocracy, but impoverished cowboys, who also knew how to cultivate crops, seeking a better life. Archaeological cultures, like the ethnographic cultures that they are often assumed mistakenly to represent, are never pure, unmixed entities but always contain elements from neighbouring or earlier cultures. The Kura-Araxes Early Bronze ‘culture’ of Transcaucasia has no single point of origin. As bearers of this cultural formation moved farther south, they changed and adopted some of the features of the peoples with whom they came into contact. Similarly the peoples from the steppes, who moved into the southern Caucasus with their oxen-driven wagons and buried their dead in impressively large kurgans, also assimilated features of the Kura-Araxes folks who stayed behind. The origins of the Bactria-Margiana Archaeological Complex are not to be sought in southern Turkmenistan, eastern Iran, Baluchistan, the Indian subcontinent or the Eurasian steppes but in all these areas combined. Like other cultural phenomena, it was a hybrid, the product of a unique convergence of different cultural traditions. The same is true for the will-of-the-wisp Aryans. They did not miraculously emerge at one point in time in one circumscribed area, but the peoples who later left us early Indo-European texts had multiple roots and were constantly transformed and invigorated through the ever-ongoing processes of cultural assimilation and physical miscegenation. The argument for cultural diversity and hybridity is not based on political correctness, but on historical accuracy. Sometimes it is intellectually stimulating to paint broad canvases; at other times it may be dangerous. The quest for the Aryans, sadly, has a long history of falling into this latter category.
Bibliography Algaze, G. 1993: The Uruk Expansion: The Dynamics of Expansion of Early Mesopotamian Civilisation (Chicago). Alimov, K., Boroffka, N., Bubovna, M., Burjakov, J., Cierny, J., Jakubov, J., Lutz, J., Parzinger, H., Pernicka, E., Radililovkij, V., Ruzanov, V., Sirinov, T., Starsinin, D. and Weisgerber, G. 1998: Prähistorischer Zinnbergbau in Mittelasien: Vorbericht der Kampagne 1997. Eurasia Antiqua 4, 137–99.
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
37
Anthony, D. 1990: Migration in Archaeology: the Baby and the Bathwater. American Anthropologist 92, 895–914. Arkheologiya Ukrainskoi SSR 1985: Pervobytnaya Arkheologii vol. 1 (Kiev). Avanessova, N.A. 1996: Pasteurs et agriculteurs de la vallée du Zeravshan (Ouzbékistan) au début de l’Age du Bronze: relations et influences mutuelles. In Lyonnet, B. (ed.), Sarazm (Tadjikistan) Céramiques (Chalcolithique et Bronze Ancien) (Paris), 117–24. Avilova, L.I., Antonova, E.V. and Teneishvili, T.O. 1999: Metallurgicheskoe proizvodstvo v yuzhnoi zone tsirkumpontiiskoi metallurgicheskoi provintsii v epokhu rannei bronzy. RosA 1, 51–65. Bacon, E. 1954: Types of Pastoral Nomadism in Central and Southwest Asia. Southwestern Journal of Anthropology 10.1, 44–68. Bakker, J.A., Kruk, J., Lanting, A.E. and S. Milisauskas, S. 1999: The Earliest Evidence of Wheeled Vehicles in Europe and the Near East. Antiquity 73 (282), 778–90. Biscione, R. 1977: The Crisis of Central Asia Urbanization in II Millennium BC and Villages as an Alternative System. In Deshayes, J. (ed.), Le Plateau Iranien et l’Asie Centrale des Origines à la Conquete Islamique (Paris), 113–27. Boroffka, N. 1998: Bronze-und früheisenzeitliche Geweihtrensenknebel aus Rumänien und ihre Beziehungen. Alte Funde aus dem Museum für Geschichte Aiud Teil 2. Eurasia Antiqua 4, 81–136. Braidwood, R.J. and Braidwood, L.S. 1960: Excavations in the Plain of Antioch I: the Earlier Assemblages, Phases A-J (Oriental Institute Publication 61) (Chicago). Cerasetti, B. 1998: Preliminary Report on Ornamental Elements of ‘Incised Coarse Ware’. In Gubaev, A., Kosholenko, G. and Tosi, M. (eds.), The Archaeological Map of the Murghab Delta Preliminary Reports 1990–95 (Rome), 67–74. Cernych [Chernykh], E.N., Antipina, E.E. and Lebedeva, E.J. 1998: Produktionsformen der Urgesellschaft in den Steppen Osteuropas (Ackerbau, Viehzucht, Erzgewinnung und- verhüttung. In Hänsel, B. and Machnik, J. (eds.), Das Karpatenbecken und die Osteuropäische Steppe: Nomadenbewegungen und Kulturaustausch in den vorchristlichen Metallzeiten (4000–500 v. Chr.) (Munich), 233–52. Chernykh, E.N. 1992: Ancient Metallurgy in the USSR: The Early Metal Age (Cambridge/New York). Chernykh, E.N., Avilova, L.I. and Orlovskaya, L.B. 2000: Metallurgicheskie provintsii i radiouglerodnaya khronologiya/Metallurgical Provinces and Radiocarbon Chronology (Moscow) (in Russian and English). de Miroschedji, P. 2000: La céramique de Khirbet Kerak en Syro-Palestine: état de la question. In Marro, C. and Hauptmann, H. (eds.), Chronologie des Pays du Caucase et de l’Euphrate aux IVe–IIIe Millenaires (Paris), 255–78. Di Nocera, M.G. 2000: Radiocarbon Datings from Arslantepe and Norsuntepe: The fourth-third millennium absolute chronology in the Upper Euphrates and Transcaucasian Region. In Marro, C. and Hauptmann, H. (eds.), Chronologie des Pays du Caucase et de l’Euphrate aux IVe–IIIe Millenaires (Paris), 73–93. Frangipane, M. 1998: Nel Segno del Leone, Archeo XIV.4, 28–41. ——. 2000: The Late Chalcolithic/EBI Sequence at Arslantepe. Chronological and Cultural Remarks from a Frontier Site. In Marro, C. and Hauptmann, H. (eds.), Chronologie des Pays du Caucase et de l’Euphrate aux IVe–IIIe Millenaires (Paris), 439–72. Frankel, D. 2000: Migration and Ethnicity in Prehistoric Cyprus: Technology as Habitus. European Journal of Archaeology 3.2, 167–87. Frankel, D. and Webb, J.M. 2000: Marki Alonia: A Prehistoric Bronze Age Settlement in Cyprus. Antiquity 74, 763–4. Gaiduchenko, L.L. 1999: Osobennosti prirodopol’zovaniya drevnego Naseleniya ‘strani gorodov’ yuzhnogo Urala i Zaural’ya. In Zdanovich, D.G., Ivanova, N.O. and Predeina, I.V. (eds.), Kompleksnye obshchestva tsentral’noi Evrazii v III–I tys. do n.e. (Chelyabinsk), 309–12 (in Russian and English). Gei, A.N. 1999: O nekotorykh problemakh izucheniya bronzovogo veka na yuge evropeiskoi rossii. RosA 1, 34–49.
38
PHILIP L. KOHL
——. 2000: Novotitorovskaya Kul’tura (Moscow). Gerskovic, J.P. 1999: Studien zur spätbronzezeitlichen Sabatinovka-Kultur am unteren Dnepr und an der Westküste des Azov’schen Meeres (Archäologie in Eurasien 7) (Rahden, Westf.). Gubaev, A., Kosholenko, G. and Tosi, M. (eds.) 1998: The Archaeological Map of the Murghab Delta Preliminary Reports 1990–95 (Rome). Hiebert, F.T. 1994: Origins of the Bronze Age Civilization in Central Asia (American School of Prehistoric Research Bulletin 42) (Cambridge, Mass.). Itina, M.A. 1977: Istoriya stepnykh plemen yuzhnogo Priaral’ya (II – nachalo I tycyacheletiya do n.e.) (Moscow). Kavtaradze, G. 1983: K khronologii epokhi eneolita i bronzy gruzii (Tbilisi). ——. 1999: The Importance of Metallurgical Data for the Formation of Central Transcaucasian Chronology. In Hauptmann, A., Pernicka, E., Rehren, T. and Yalçin, U. (eds.), The Beginnings of Metallurgy (Der Anschnitt Beiheft 9), 67–102. Khazanov, A.M. 1994: Nomads and the Outside World, 2nd ed. (Madison, WI). Kohl, P.L. 1984: Central Asia: Palaeolithic Beginnings to the Iron Age (L’Asie Centrale des origines à l’Âge du Fer) (ERC Synthèse no. 14) (Paris). Korenevskii, S.N. 2001: Drevneishie Zemledel’tsy i Skotovody Predkavkaz’ya (Avtoreferat dissertatsii) (Moscow). Kushnareva, K.K. 1997: The Southern Caucasus in Prehistory: Stages of Cultural and Socioeconomic Development from the Eighth to the Second Millennium BC (Philadelphia). Lyonnet, B. 1997: Céramique et Peuplement du Chalcolithique à la Conquête Arabe: Prospections Archéologiques en Bactriane Orientale (1974–1978) vol. 2 (Paris). —— n.d.: Mari et la Margiane ou ‘la circulation des biens, des personnes et des idees’ dans l’Orient ancien à la fin du 3ème et au début du 2ème millénaire avant notre ère. Mémoire présenté dans le cadre d’une Thèse d’Habilitation, Avril 2001 (Paris). Magomedov, R.G. 1991: O kompleksakh maikopskoi kul’tury na territorii Dagestana. In Davudov, O.M. (ed.), Gory i Ravniny Severo-vostochnogo Kavkaza v Drevnosti i Srednie Veka (Makhachkala), 13–38. Masson, V.M. 1981: Altyn-depe During the Aeneolithic Period. In Kohl, P.L. (ed.), The Bronze Age Civilization of Central Asia: Recent Soviet discoveries (Armonk, NY), 63–95. ——. 1988: Altyn-Depe (Philadelphia). Masson, V.M. and Merpert, N.Y. (eds.) 1982: Eneolit SSSR (Moscow). Merpert, N.Y. 2001: Rol’ V.A. Gorodstova v sozdanii periodizatsii bronzovogo veka yuzhnoi poloviny vostochnoi evropy i voprocy istoriografii. In Kolev, Y.I. (ed.), Bronzovyi vek vostochnoi evropy: kharakteristika kul’tur, khronologiya i periodizatsiya (Samara), 7–10. Monah, D. and Monah, F. 1997. Cucuteni: the Last Great Chalcolithic Civilization of Europe (Bucharest). Morales-Muñiz, A. and Antipina, E. 2003: Srubnaya Faunas and Beyond: A Critical Assessment of the Archaeozoological Information from the East European Steppe. In Levine, M., Renfrew, C. and Boyle, K. (eds.), Prehistoric Steppe Adaptation and the Horse (Cambridge), 329–51. Munchaev, R.M. 1975: Kavkaz na zare bronzovogo veka (Moscow). ——. 1994: Maikopskaya kul’tura. In Kushnareva, K.K. and Markovin, V.I. (eds.), Epokha Bronzy Kavkaza i Srednei Azii: rannyaya i srednyaya bronza Kavkaza (Moscow), 158–225. Paléorient 1999 (25.1): The Uruk Expansion: Northern Perspectives from Hacinebi, Hassek Höyük and Gawra (Paris). Parzinger, H. 1998a: Kulturverhältnisse in der eurasischen Steppe während der Bronzezeit. In Hänsel, B. (ed.), Mensch und Umwelt in der Bronzezeit Europas (Kiel), 457–79. ——. 1998b: Der nordpontische Raum und das untere Donaugebiet in der späten Kupferzeit: das Ende des Kodzadermen-Gumelnita-Karanovo VI-Verbandes und die Cernavoda I-Kultur. In Hänsel, B. and Machnik, J. (eds.), Das Karpatenbecken und die Osteuropäische Steppe: Nomadenbewegungen und Kulturaustausch in den vorchristlichen Metallzeiten (4000–500 v. Chr.) (Munich), 123–34. Pernicka, E., Begemann, F., Schmitt-Strecker, S., Todorova, H. and Kuleff, I. 1997: Prehistoric Copper in Bulgaria. Its Composition and Provenance. Eurasia Antiqua 3, 41–180.
1. THE EURASIAN STEPPES AND THE ANCIENT NEAR EAST
39
Philip, G. and Millard, A.R. 2000: Khirbet Kerak Ware in the Levant: The Implications of Radiocarbon Chronology and Spatial Distribution. In Marro, C. and Hauptmann, H. (eds.), Chronologie des Pays du Caucase et de l’Euphrate aux IVe–IIIe Millenaires (Paris), 279–96. Priakhin, A.D. and Besedin, V.I. 1999: The Horse Bridle of the Middle Bronze Age in the East European Forest-Steppe and the Steppe. Anthropology and Archeology of Eurasia 38.1, 39–59. Rassamakin, Y. 1999: The Eneolithic of the Black Sea Steppe: Dynamics of Cultural and Economic Development 4500–2300 BC. In Levine, M., Rassamakin, Y., Kislenko, A. and Tatarintseva, N. (eds.), Late Prehistoric Exploitation of the Eurasian Steppe (Cambridge), 59–182. —— 2002: Aspects of Pontic Steppe Development (4500–3000 BC) in the Light of the New Cultural-Chronological Model. In Boyle, K., Renfrew, C. and Levine, M. (eds.), Ancient Interactions: East and West in Eurasia (Cambridge), 49–73. Roaf, M. 1990: Cultural Atlas of Mesopotamia and the Ancient Near East (Oxford). Rothman, M.S. (ed.) 2001: Uruk Mesopotamia and Its Neighbors (Santa Fe, NM). Sagona, A.G. 1984: The Caucasian Region in the Early Bronze Age (BAR International Series 214) (Oxford). ——. 2000: Sos Höyük and the Erzurum Region in Late Prehistory: A Provisional Chronology for Northeast Anatolia. In Marro, C. and Hauptmann, H. (eds.), Chronologie des Pays du Caucase et de l’Euphrate aux IVe–IIIe Millenaires (Paris), 329–374. Salvatori, S. 1998: The Bronze Age in Margiana. In Gubaev, A., Kosholenko, G. and Tosi, M. (eds.), The Archaeological Map of the Murghab Delta Preliminary Reports 1990–95 (Rome), 47–55. Sarianidi, V.I. 1977: Drevnie Zemledel’tsi Afganistana (Moscow). ——. 1998: Margiana and Protozoroastrianism (Athens). Sarianidi, V.I., Terekhova, N.N. and Chernykh, E.N. 1977: O rannei metallurgii i metalloobrabotke drevnei Baktrii. SA 2, 35–42. Shishlina, N.I. 2003: Yamnaya Culture Pastoral Exploitation: A Local Sequence. In Levine, M., Renfrew, C. and Boyle, K. (eds.), Prehistoric Steppe Adaptation and the Horse (Cambridge), 353–65. Shishlina, N.I. and Hiebert, F.T. 1998: The Steppe and the Sown: Interaction between Bronze Age Eurasian Nomads and Agriculturalists. In Mair, V.H. (ed.), The Bronze Age and Early Iron Age Peoples of Eastern Central Asia (Philadelphia), 222–37. Stein, G.J., Bernbeck, R., Coursey, C., McMahon, A., Miller, N.F., Misir, A., Nicola, J., Pittman, H., Pollock, S. and Wright, H. 1996: Uruk Colonies and Anatolian Communities: An Interim Report on the 1992–1993 Excavations at Hacinebi, Turkey. American Journal of Archaeology 100.2, 205–260. Trifonov, V. 1987: Nekotorye voprosy peredneaziatskikh svyazei maikopskoi kul’tury. KSIA 192, 18–26. ——. 2001: Popravik absolyutnoi khronologii kul’tur epokhi eneolita-srednei bronzy Kavkaza, stepnoi i lesostepnoi zon vostochnoi Evropy (po dannym radiouglerodnogo datirovaniya). In Kolev, Y.I. (ed.), Bronzovyi vek vostochnoi evropy: kharakteristika kul’tur, khronologiya i periodizatsiya (Samara), 71–82. Videjko, M.Y. 1995: Grosssiedlungen der Tripol’e Kultur in der Ukraine. Eurasia Antiqua 1, 45–80. Vinogradov, A.V. 1979: Issledovaniya pamyatnikov kamennogo veka v severnom Afganistane. Drevnyaya Baktriya 2, 7–62. Voigt, M.M. and Dyson, R.H. jr 1992: The Chronology of Iran, ca. 8000–2000 BC. In Ehrich, R.W. (ed.), Chronologies in Old World Archaeology 3rd ed. (Chicago), vol. 1, 122–78. Webb, J.M. and Frankel, D. 1999: Characterizing the Philia Facies: Material Culture, Chronology, and the Origin of the Bronze Age in Cyprus. American Journal of Archaeology 103, 3–43. Weiss, H. and Young, T.C. jr 1975: The Merchants of Susa: Godin V and Plateau-Lowland Relations in the Late Fourth Millennium BC. Iran 13, 1–18. Wolf, E. 1984: Culture: Panacea or Problem? American Antiquity 49.2, 393–400. Woolley, C.L. 1953: A Forgotten Kingdom (London).
CHAPTER TWO THREE DEADLY SINS IN STEPPE ARCHAEOLOGY: CULTURE, MIGRATION AND ARYANS1 DAVID W. ANTHONY
This publication is a sign of something new in Western archaeology. Twentyfive years ago, when I was trying to read Sovetskaya Arkheologiya as a graduate student at the University of Pennsylvania, I felt almost alone. The library cards in the backs of the Russian-language books and journals were blank – the only entry besides mine would be something like ‘Bindery 1968’. Only about ten libraries in the US even had these publications – Soviet archaeology journals were printed in Cyrillic script on cheap yellow paper, almost pure wood, that faded to a rich ochre brown within a few years. Soviet and Western archaeologists could not visit each others’ excavations – a Communist wandering around in rural Nebraska with a camera, topographic maps and sketchbooks would have had the same difficulty as a Western archaeologist in Kazakhstan. In many ways Soviet archaeology remained distinct from Western archaeology, a difference that first came into the open in a polite letter of rebuke written by V.G. Childe, the father of Western European prehistory, to the Institute of Archaeology in Moscow in 1956 (Merpert 1992). Really important innovations flowed back and forth between the USSR and the West – edge microwear analysis was invented by Semenov (1964) to determine the function of ancient flint tools, and Western archaeologists quickly adopted it; taphonomy was developed by Efremov (1940), an approach that still inspires Western archaeologists; while things like radiocarbon dating and 1970s Systems Theory were developed in the West and adopted in the USSR. But the two archaeologies remained apart in many mundane aspects of daily interpretation and method, and in Western texts on European archaeology the archaeological culture maps almost always stopped abruptly at the Soviet border. My dissertation adviser,
1
This paper was delivered as the keynote address at the conference that produced this volume. In revising it I have tried to retain something of the feel of the original talk.
2. STEPPE ARCHAEOLOGY: CULTURE, MIGRATION AND ARYANS
41
Bernard Wailes, a very patient man, specialised in Iron Age Ireland, about as far from the steppes as a European archaeologist could get. For field experience in Eastern Europe I excavated not in Russia but, like other Westerners at the time, in politically friendlier Yugoslavia. International excavation projects in the USSR were few and far between, and all of them were closely controlled by the Institute of Archaeology in Moscow. James Mallory was welcomed into the research institutes in Kiev, but did not get into the field. In the late 1970s, Phil. Kohl explored the possibilities in northern Iran and Afghanistan, at the other end of the steppes; ultimately he was invited to the Caucasus by the Institute of Archaeology in 1986 – one of the early signs of the coming change. Renate Rolle and other German archaeologists were permitted to collaborate on Scythian-Greek excavation projects in Ukraine and Crimea, but not in the steppes of Russia or Kazakhstan to the north and east. The city of Samara, now so familiar to several contributors in this volume, was closed to all foreigners; and the small rural fishing lodge that we have recently used as a field base was then the private retreat of the First Secretary of the Communist Party for the Samara oblast, an impossible place for an American to even know about. For a long time it was my fond hope simply to see enough Western archaeologists involved in steppe archaeology so that I could attend a steppe archaeology symposium at the Society for American Archaeologists (SAA) conference. It seemed an impossible dream. I became a North American prehistoric archaeologist and took a North American teaching position. But the situation was changing already in 1989, when D. Brown and I went on our first steppe-related research trip, to Kiev, to look at horse teeth in the collections of the Institute of Zoology – a trip that led to our first paper on bit wear and horseback riding (Anthony and Brown 1991; Anthony, Telegin and Brown 1991). Then in 1991 came the great change, the fragmentation of the Soviet Union, and suddenly the political barriers were gone. Russian archaeologists were invited by Antonio Gilman to a symposium at the American Anthropological Association (AAA) conference in Chicago in 1991, an event that set off a flurry of plans, collaborations, and very quickly a book, edited by Phil. Kohl and Clare Fawcett (Kohl and Fawcett 1995). In 1995 Brown and I helped to organise an international conference in Petropavlovsk, northern Kazakhstan, that was attended by some of the contributors to this volume, along with interested parties from 13 countries. Those who were in Petropavlovsk will never forget the conference field trip. It began in two buses at 6am and ended 21
42
DAVID W. ANTHONY
hours later at 3am in different buses, the original pair having expired several hours apart on an empty two-lane road in the middle of the northern Kazakh steppes. During the fourth feast that was served to us that day, an event in a lakeside pine forest that featured groaning tables full of champagne, vodka, kumyss and a variety of grilled meats and salads, it occurred to me that we were eating the funds that had been meant to pay for the conference publication. The trip was still one stop short of its final destination when, at 2am, as we finished the sixth and most incredible feast of the day, during a standing toast to the nobility of the horse, I realised that any possibility of a conference volume had been completely consumed. But it turned out that the day of Six Dinners and Four Buses was just a prelude. Victor Mair invited many of the Petropavlovsk conference attendants to repeat their papers at his larger conference at the University of Pennsylvania the following spring, in 1996, which quickly led to a two-volume publication (Mair 1998). Important EastWest conferences then occurred at Arkaim, Russia in 1999; at Cambridge University in January 2000; at the Carnegie Museum of Pittsburgh in October 2000; and finally in Chicago in the spring of 2002. Now, 12 years after the restructuring of the USSR, one-sixth of the land surface of the planet, formerly terra incognita to Western archaeologists, is open to Western researchers and collaborative projects. There were sessions on steppe archaeology at the SAA conferences in 1999, 2000, 2001 and 2002, organised by Claudia Chang, Sandra Olsen, Katherine Linduff, Irene Good and myself. Karen Rubinson started a Russian-language archaeological translation series funded by the American Institute of Archaeology. Russian archaeologists have organised sessions on steppe archaeology at the annual conferences of the European Association of Archaeologists in 1998, 1999, 2000, 2001 and 2002. International teams of archaeologists, Russian-American, Russian-German, Russian-French, Russian-Spanish, Kazakh-American and MongolianAmerican, are now working in the steppes. The Chicago conference was organised by a new generation of steppe archaeologists, graduate students at Chicago. And we had enough people to keep talking for days. I felt like a hermit who had come out of a cave. For so many years there was really almost no one that I could talk with about the things that interested me the most. And now there are so many of us that already we can divide into quarrelling factions – the first proud sign of a mature academic tradition. Now we face new problems – the problems of growth. How do we really bring this collaboration to maturity? How do we take advantage of two gen-
2. STEPPE ARCHAEOLOGY: CULTURE, MIGRATION AND ARYANS
43
erations of work by Soviet and post-Soviet archaeologists? Can their interpretations be joined profitably with ours? How do we convince our Slavic colleagues that Western archaeology has something substantive to contribute to the understanding of steppe prehistory? And as Western archaeologists who must find academic positions in Western universities, how do we convince more of our Western colleagues that steppe archaeology is ready to contribute new and crucial ideas to Western archaeological and historical thought? It is this last question that is the theme of this paper. Many Western academics recognise the potential value of steppe archaeology, but at times still seem suspicious of the field methods, the political agendas and the interpretive approaches often associated with it. The discussion period after Sandi Olsen’s symposium at the SAA conference in Philadelphia in 2000 veered unexpectedly into the subject of Aryan migrations and almost became a brawl. Dispute is the mother of discovery, but entrenched suspicions can be counter-productive. Why do they exist? This is of course a complicated question with no simple answer. I will try to approach it by isolating three problem areas, my ‘three deadly sins’ in steppe archaeology: 1) the equation of archaeological cultures with race and language; 2) the use of migration to explain culture change; and 3) the assignment of an Aryan, Indo-Iranian identity to archaeological cultures.
The First Deadly Sin: Linking Language, Race and Material Culture In Soviet and post-Soviet archaeology arguments over the meaning of the ‘archaeological culture’ continued for decades without real resolution, although the debate itself was positive. In practice many post-Soviet archaeologists equate archaeological cultures with ethno-linguistic and sometimes with genetic groupings. For example, my Russian colleagues in Samara published a book in 1995 – a very useful and important book, entitled Ancient Indo-Iranian Cultures of the Volga-Ural Region, 2nd Millennium BC (Vasiliev 1995). The title assumed the Indo-Iranian or Aryan identity of the Middle and Late Bronze Age societies of the middle Volga – without real criticism or disagreement. This identification is thought to be so firmly established that it is accepted by almost everyone working in the region. But Western archaeologists probably would not be convinced. The Soviet-era concept of the ‘ethnos’ as an historical cultural community that persisted through centuries or even millennia, in
44
DAVID W. ANTHONY
spite of changes in the material forms of technology or in domestic material culture, seems to underlie the linkage. I must quickly add that many contemporary Russian archaeologists, including Leonid Yablonsky, explicitly deny the equation of race with language and culture – Yablonsky has shown that the Iron Age ‘Saka’ archaeological culture of the Aral Sea area was produced by two quite different skeletal groups, Caucasoid and Mongoloid (Yablonsky 1996). But many others assign ethnic, linguistic and even racial identities to their archaeological cultures without much worry. Arguments over specific equations are intense, but they tend to be about which ethno-linguistic group – Iranian or Turkic, for example – should be equated with which archaeological culture (Chernykh 1995; Shnirelman 2001). In the West such equations are routinely rejected by most archaeologists, not just in specific cases, but as a general procedure. Most of the leading American, British and Western European archaeologists have accepted the proposition that race, language and culture are, first, elusive terms, almost impossible to define; and second, vary independently. Scott MacEachern restated the litany in a convenient form in a recent Current Anthropology article: African ethnic units are not bounded, homogeneous monoliths either frozen in place since before 1492, or caroming around the continent like culture-bearing billiard balls. They are dynamic entities, manipulated in response to changes in the natural and social environment, and their composition can change very quickly indeed. Differences in the temporal resolutions of genetic, linguistic, and archaeological analyses are one of the main difficulties in multidisciplinary studies . . . (MacEachern 2000, 370)
MacEachern says that ethno-linguistic identities are changeable, flexible and situational, something I think we would all accept. But like any generalisation this one is incomplete, and the very thing it disregards – cultural stability – is one of the most striking aspects of the prehistoric archaeology of the steppes. Soviet and post-Soviet-era beliefs in very long-lived cultural identities, even in pre-state, tribal conditions, are not just an artefact of a different academic tradition, but rather conform to the historical specificity of Russian experience. Persistent cultural frontiers were common in and near the steppes, particularly between pastoralists and farmers, the steppe and the sown – but also in other situations, less easy to explain from a Western perspective of situational and flexible ethnicity.
2. STEPPE ARCHAEOLOGY: CULTURE, MIGRATION AND ARYANS
45
For example, the first farmers in the Pontic-Caspian steppe region, north of the Black and Caspian seas, were Cri¤-culture pioneers who immigrated from south-eastern Europe and the lower Danube valley into the forested east Carpathian piedmont, in present-day Moldavia (Fig. 1). Their arrival created a cultural frontier north-west of the Black Sea in modern Ukraine between the Dnieper and Dniester rivers that persisted for 2500 years, from about 5800 to 3500 calBC. The frontier was not ecological – two distinct cultural systems existed side by side in the same forest-steppe vegetational zone. A robust bundle of differences in material culture clearly distinguished the immigrants and their cultural descendants – Cri¤, Linear Pottery, Triploye – from the indigenous societies and their cultural descendants – Dnieper-Donets, Mariupol, Sredni Stog and Yamnaya. They differed in house forms, settlement types, economy, ceramic form and decoration, ceramic technology, stone tool types, mortuary rituals, the presence-or-absence of female figurines and metallurgical techniques – in other words, they maintained distinctions in almost every aspect of material culture. And this was not the only robust, persistent cultural frontier in the region. Another frontier ran along the ecological divide between the northern broadleaf forests and the dry steppes to the south. In the northern forest zone people remained foragers, while in the steppe river valleys, people adopted domesticated animals and became cattle-and-sheep herders – and again, this economic frontier persisted for 2500–3000 years. It appeared about 5200/5000 calBC, when herding economies were widely adopted in the European steppes, west of the Urals; but were not adopted in the forest zone to the north. It persisted until about 2500/2000 calBC, when domesticated animals finally began to be kept by most societies in the southern forest zone. Like the Dnieper/Dniester frontier, the forest/steppe frontier was defined by a robust bundle of cultural distinctions – different house forms, pottery types, stone tools and burial practices – although at this frontier there was more interpenetration of traits, particularly in ceramic form and decoration. Finally, on the eastern edge of the European steppes there was yet a third persistent cultural frontier, a north-south line extending from the southern slopes of the Ural mountains to the deserts north of the Caspian Sea. This frontier was just as resistant to the spread of herding-based economies, but unlike the forest-edge frontier, it was not so clearly ecological. Grasslands extended unbroken from the VolgaUral region eastward into northern Kazakhstan. But long after herding was adopted in the Volga-Ural steppes (about 5200/5000 calBC), the societies of northern
46
DAVID W. ANTHONY
Fig. 1. Persistent economic and material culture frontiers in the Pontic-Caspian steppe region.
2. STEPPE ARCHAEOLOGY: CULTURE, MIGRATION AND ARYANS
47
Kazakhstan remained foragers – groups like Atbasar, Surtanda and TersekBotai. Like the first two frontiers, this third one persisted for at least 2500 years, until about 2500 calBC. In all three of these cases it is quite clear from published archaeological reports that the cultures on either side of the frontier knew and interacted with each other, but rather than assimilating, they remained distinct, and not for a few generations but for millennia. In some cases it seems that the frontier became even sharper with the passage of time – for example, the steppe-zone herding cultures of the Yamnaya period, 3300–2500 calBC, were more different from the nearby forest-zone foragers (Garin-Bor and Volosovo) than the earlier steppe zone herders of the Khvalynsk culture had been. Prominent Western historians and social theorists – Michael Mann, Eric Wolf, Anthony Giddens, Benedict Anderson – have described pre-state, tribal ethnic groups as unstable and unbounded – social clouds of steam, appearing and then disappearing without developing lasting regional boundaries. Modern states have borders, border police and passports, but these are recent inventions. They are thought to have replaced a protean pre-state flow of shifting tribal and clan alliances. In this Western view of the past, pre-state, pre-class group identities – Celt, Scythian, Cimbri, Teuton, Pict – are seen either as entirely modern inventions that never existed, created for modern political reasons; or as brief historical phenomena that were unable to persist for any length of time. Only the state is accorded the power to warp ethno-linguistic identity into a stable and persistent phenomenon, like the state itself. This interpretation of ancient ethnicity is applicable in many cases, but it is certainly incomplete. It does not help us to understand tribal cultural frontiers that were stable for hundreds or even thousands of years – on the contrary, it suggests that they should not have existed. But there they are, clearly documented by archaeologists in many places – including the steppes. Distinctive and persistent material-culture regions seem to have existed among Upper Palaeolithic foragers in Europe (Gamble 1999, 289–312), so it seems likely that they have always been with us. Recognising and understanding them is an essential problem in steppe archaeology. And it seems clear to me that one historical process – not the only one, but one possibility – that can create persistent cultural frontiers in a non-state context is, unfortunately, the second deadly sin in steppe archaeology – migration.
48
DAVID W. ANTHONY
The Second Deadly Sin: Migration as an Explanation for Culture Change Migration is under-used by Western archaeologists as an explanation for culture change, and it is over-used by Soviet and post-Soviet archaeologists. Both academic traditions have ignored most of the complexities of migration theory, as well as most of the recent historical literature on migration and its social effects. My Slavic colleagues have used migration too frequently as a convenient ‘default’ agent of culture change, without sufficient attention to other, internal mechanisms of change. For example, Chernykh (1992, 226–7) has explained the Late Bronze Age Seima-Turbino horizon, which is defined by unusually sophisticated cast tin-bronze weapon types that occur in cemeteries of the southern forest zone, from the Altai mountains to Moldavia, around 2000–1700 calBC, as the residue of a migration of a warrior elite from the east. Western archaeologists would be more inclined to interpret SeimaTurbino metals as the rapid spread of an elite style of metalwork among newly emerged chiefdoms competing for public prestige and power – driven by interregional trade, not migration. But Western archaeologists are wrong to dismiss migration as the random movement of enclosed cultures ‘caroming around the continent like culture-bearing billiard balls’, as MacEachern put it. In the West, long-distance migration by substantial populations often is regarded either as a density-dependent phenomenon largely confined to the modern era (Clark 1994); or as a random process of Brownian motion advancing perhaps 1 km per year, more like a gradual seepage than a wave or stream (Ammerman and Cavalli-Sforza 1979). This mechanistic, demographic explanation of migration is peculiar to Western archaeologists – even Western sociologists and historians would reject it, because it is woefully inadequate. People do not migrate, even in the modern world, simply because they are too crowded at home. Resource shortages and crowding are ‘push’ factors – negative conditions at home that raise the probability of out-migration. They are only one kind of push factor, and often not the most important kind. Many migrations have causes that are essentially social, not demographic. Wherever inheritance rules favour older siblings and condemn the younger ones to go out and find their own lands or clients or hunting grounds, as in ancient Rome, feudal Europe, Mayan Central America and many parts of modern Africa, the disfavoured siblings migrate (Kopytoff 1987; Anthony 1997). Wherever younger people are constrained and oppressed by the power of the elders, where the hand of the dead ancestors weighs heavily on social life, people who are dis-
2. STEPPE ARCHAEOLOGY: CULTURE, MIGRATION AND ARYANS
49
advantaged because they belong to the wrong clan or occupy an unfavourable position in the hierarchy are likely to leave. The persistent outward migrations and conquests of the pre-colonial east African Nuer were caused, according to Raymond Kelley (1985, 194–201), not by an absolute deficit of pastures within Nuerland, but rather by the internal pressures generated by a cultural system of lavish bride-price payments that made it very expensive for young Nuer men to obtain a socially desirable bride. This encouraged Nuer men to raid their non-Nuer neighbours for cattle (and pastures to support them) that could be used to pay the elevated Nuer bride-price. The Nuer cattle shortage was produced by a cultural practice, not by the simple ratio of cattle to pasture to humans. Migration cannot be explained just by absolute resource shortages – often it is a conscious social strategy that improves the migrant’s position in competition for status, privileges and even a good spouse. Social migrants do not necessarily leave forever – return migration is so common that modern demographers plan on it – because their real goal is to improve their status among old friends and family. Often they recruit clients and followers among the people who stayed behind, and convince them to migrate also. The demographic explanation of migration favoured by Western archaeologists depends on just one kind of push factor – population pressure – but that is not its worse flaw. It also ignores the other causes of migration entirely. The probability of migration is determined by an interplay between ‘push’ factors, ‘pull’ factors (the supposed attractions of the destination), the social pattern of networks of communication (which determine who has access to information about the destination), economic and social structures at home (which further constrain which elements of the population might consider migration feasible) and transport technologies (which can greatly increase or lower the cost of movement). Changes in transport costs or in communication networks or in objective conditions at the destination will raise or lower the threshold at which migration becomes an attractive option. Any model of migration that ignores pull factors, the flow of information, socio-economic constraints and the cost of transport is inadequate at the outset. The billiard-ball analogy is applicable only if the cause of migration is reduced to a single kick. The connection between long-distance migration and frontier formation processes becomes apparent only when the highly patterned nature of long-distance migration is recognised. In long-distance migrations, only those people who have access to favourable information about the destination will decide to leave. And they will go only to those places that they have heard good
50
DAVID W. ANTHONY
things about, ignoring other possible destinations. Long-distance migrants move in ‘streams’ to definite targets, not in broad ‘waves’ that wash randomly over the landscape. Information flows from the scouts who first explore the destination region, who could be disfavoured younger siblings, migrant labourers, mercenaries, fur trappers and so forth, to their co-residents and family back home. The social link between the scouts and their home group defines the potential pool of migrants. If social organisation at home is atomised – if people live in small, inward-looking communities like the Pueblo Indians in the south-west United States – then the potential pool of migrants is highly restricted. If social organisation at home is more open, with many links between socially heterogeneous communities that freely exchange information, the potential pool of migrants is larger. But in either case, migrants tend to be recruited from relatively restricted social groups – they are not drawn randomly from the home population at large. And this strongly affects their material culture and speech. Colonist speech generally is more homogeneous than that of the homeland. Dialectical differences were much smaller among colonial-era English-speakers in North America than they were in the British Isles. The Spanish dialects of colonial South America were more homogeneous than the dialects of southern Spain, which was the home region of most of the original colonists during the 1500s and 1600s; and the dialects of southern Spain were more homogeneous than the dialects of northern Spain – the region from which southern Spain was colonised after the defeat of the Iberian Islamic kingdoms in the 1400s (Hock and Joseph 1996, 362–5; Chambers and Trudgill 1980, 108). There are two reasons for the relative homogeneity of colonist speech. One is chain migration – colonists do not represent a random sample of all parts of their homeland, but instead tend to recruit family and friends from the same places and social groups that they came from. And when these linked groups of people arrive in a new place after a long-distance migration, they must depend on each other to survive. They then come under the social influence of the first effective settlers. The first group to establish a viable social system in a new place is often called a ‘charter group’ by historians (Nash 1984). The charter group, or the first effective settlers, generally get the best land. They might claim rights to perform the highest-status rituals, as among the Mayan royal lineages of Central America or the Pueblo Indians of the American south-west. In some cases (Puritan New England) their town councils choose who is permitted to
2. STEPPE ARCHAEOLOGY: CULTURE, MIGRATION AND ARYANS
51
join them. They therefore leave an inordinate cultural imprint on later generations, which tend to adopt their values, at least publicly. The advantage associated with first effective settlement explains why the English language, English house forms and English settlement types were retained in 19th-century Ohio although the overwhelming majority of later immigrants was German (Wilhelm 1992). It also explains why East Anglian English traits (typical of the earliest Puritan immigrants) continued to typify New England dialectical speech and domestic architecture long after the majority of later immigrants arrived from other parts of England or Ireland. Migration creates social dynamics that leave many of the later migrants indebted to or dependent on the first-comers, whose dialect and material culture provide the cultural capital for a new group identity. They exercise a kind of cultural hegemony over later migrants, something that they themselves recognise and may consciously exploit. The combination of chain migration, which restricts the pool of potential migrants at home, and the influence of the charter group or first effective settlers, which encourages conformity at the destination, produces a levelling of differences among colonists. Simplification (fewer variants than in the home region) and levelling (the tendency toward a standardised form) affect both the dialect and the material culture of migrants. In material culture, settlement organisation and domestic architecture particularly tend toward standardisation, perhaps because they are the most easily visible signals of identity in any agricultural landscape. Long-distance migration has a transforming effect on social identities – it creates communities that initially are relatively homogeneous, in language and material culture; it creates charter groups among the migrants; and the cultural power of these apex families tends to establish the rules for future generations’ inclusion in the elite. This chain of events has historically created very clear, persistent ethno-linguistic frontiers at the edges of colonised regions. For example, four folk-culture regions have been defined in colonial eastern North America by folklorists and historians, and four dialect regions by linguists (New England, MidAtlantic, Southern Tidewater and Southern Appalachian). Language and material culture frontiers coincide almost exactly in all four regions. Both dialect and culture are traceable in each case to a particular region in the British Isles from which the first effective European settlers came (Fischer 1989). The four ethno-linguistic regions of colonial eastern North America were created by four separate migration streams that imported people with distinctive ethno-lin-
52
DAVID W. ANTHONY
guistic identities into four different parts of the eastern woodlands (MaineConnecticut, south-eastern Pennsylvania, tidewater Virginia-Carolina and the central Appalachians) where their original linguistic and material differences survived, were elaborated, and persisted for centuries. In some ways, including modern presidential voting patterns, the ghosts of these regions survive even today. Similar persistent ethno-linguistic frontiers emerged at the edges of colonised regions in Wales (Welsh/English), Brittany (Breton/French), Switzerland (German/French) and the Hudson River valley in New York state (Iroquois/Algonquian).2 Migration does not always create stable ethno-linguistic frontiers, but it can, particularly when substantial numbers of immigrants move into a culturally foreign territory, or when they import a different religion and ideology, or when the migrants import a new economy – as the Cri¤ farmers did when they moved into the east Carpathian piedmont west of the Dniester about 5800 calBC, creating a frontier that lasted for millennia. In comparative cases including those just listed, robust material culture frontiers that persisted in one place for centuries were also language frontiers. By ‘robust’ I mean material culture frontiers that are defined by a bundle of trait
2
In Wales, the landsker, a named and overtly recognised ethno-linguistic frontier between Welsh and English regions, solidified about 1200, after the migration of Normans into Wales slowed. It persisted into the 20th century. It separated populations that spoke distinct languages (Welsh/English), built different kinds of churches (Celtic/Norman English), had different systems of politics and justice, managed agriculture in different ways with different tools, used different systems of land measurement, and maintained a wide variety of distinctions in dress, food and custom. A persistent ethno-linguistic frontier has existed since the post-Roman 5th or 6th century in Switzerland near Basle. After the fall of Rome, Germans migrated into the northern cantons of Switzerland, and the French kingdom of Burgundy occupied what had been Gallo-Roman western Switzerland. The frontier between them still separates ecologically similar regions that differ in language (German/French), religion (Protestant/Catholic), architecture, the size and organisation of landholdings, and the nature of the agricultural economy. Another post-Roman frontier, created by a migration during the 5th and 6th centuries, separated the Celts in Brittany from Gallic and Germanic tribes. After 1600 years it still separates Bretons from the French, a distinction expressed in dress, cuisine, house type, village organisation and language. In North America, the Hudson River valley was the boundary between Iroquoian and Algonkian language groups for at least 500 years before European contact, a stable linguistic frontier also expressed in different house types, settlement types, economies, ceramics, pipes, military organisation, and ritual calendars. Dean Snow (1995; 1996) has suggested that this ethno-linguistic boundary was created by a migration of pioneer Iroquoian farmers from what is now Pennsylvania into New York around 950 AD.
2. STEPPE ARCHAEOLOGY: CULTURE, MIGRATION AND ARYANS
53
contrasts – what Emberling (1997) calls ‘redundancy’ at a material culture frontier. Robust or redundant material culture frontiers that persist for centuries are, in most known cases, language frontiers as well. This is one way that language and material culture are related – a relationship that neither Western nor post-Soviet archaeologists have recognised, but one potentially very valuable to archaeologists. Migration can lead to the formation of such persistent ethnolinguistic frontiers, although it is not the only way they can come about. Understanding the interplay between social migration and frontier formation remains a challenge for steppe archaeology, whether practised by Western archaeologists or those from the former Soviet states. The challenge is perhaps sharpest when we try to apply migration theory to the problem of Aryan origins – the third deadly sin in steppe archaeology.
The Third Deadly Sin: The Assumption of Aryan Identity The term Aryan has at least two meanings. The historical Aryans were the Aryans of the Rig Veda and the Avesta, the seminal texts of Hinduism and Zoroastrianism. The Rig Veda, a collection of 1026 Sanskrit hymns and poems arranged in ten books, was compiled between 1500 and 1300 BC in the Punjab region of northern India/Pakistan, partly from older sources. The Avesta is younger. Its oldest sections, the Gathas, were composed in north-eastern Iran in an archaic dialect of Iranian, and probably date to about 1200–1000 BC. The Gathas were very similar to the Rig Veda in grammar, poetic meter, subject matter, deities and specific rituals. Both the Iranian-speakers of the Avesta and the Sanskrit-speakers of the Rig Veda called themselves Aryans. The two languages and ritual traditions were derived from a parental Indo-Iranian language and culture dated before 1500 BC. These two texts and their elusive Indo-Iranian parent define and identify the historical Aryans. Every aspect of this historical Aryan identity is open to historical and linguistic debate, and has been actively debated by scholars for 200 years (for a good review see Erdosy 1995). The second sense of the term Aryan is a political and racial fantasy, concocted to rationalise nationalist or racist claims to property and privilege. In a stupendous feat of the imagination, the historical Aryans became a genetically pure, morally superior race of northern European blonde supermen whose deeds lent support to a fascist political agenda. This second conception is propaganda, not
54
DAVID W. ANTHONY
history, and therefore is invulnerable to historical counter-arguments or archaeological evidence, which show very clearly that race had little or no connection with the historical Aryan identity. People with non-Aryan names and borrowed non-Aryan words appeared in the Rig Veda. Some of the Sanskrit-speaking chiefs who were heroes and subjects of hymns of praise in the Rig Veda had non-Sanskrit names and, presumably, non-Aryan local origins. So even the Aryans of the Rig Veda were not genetically ‘pure’ – whatever that means. The important and innovative quality of the Rig Veda was that it successfully established a single sacred canon of Aryan hymns and prayers, organised in a way that guarded it against the loss or addition of a single syllable. This almost incorruptible tradition created a clear and persistent ritual divide between those who were ‘proper’ and those who were not. If you sacrificed in the right way to the right gods, which required performing the right prayers in the right language, you were an Aryan. Otherwise you were not – a criterion that excludes almost all modern claimants to the name. The Rig Veda made the ritual and linguistic frontier clear. It did not imply or require racial purity. Even if the Rig Veda had tried to limit inter-ethnic mating it would not have worked. At almost any ethno-linguistic frontier in the world (Bantu/Khoisan, Swiss French/Swiss German, Northern Iroquois/Algonkian) people find mates among those on the other side. One of the most interesting things about ethnolinguistic frontiers is that they are not usually biological – they can persist for an extraordinarily long time, centuries, in spite of the regular movement of people across them. As Warren DeBoer (1990) noted in his study of native pottery styles in the western Amazon basin, ‘. . . ethnic boundaries in the Ucayali basin are highly permeable with respect to bodies, but almost inviolable with respect to style.’3 A similar dynamic explains the Germans who migrated 3
An example of this confusion between culture and biology can be found in Eric Wolf’s treatment of the Iroquois. Wolf asserted that the Iroquois did not really exist as a distinct tribe during the American colonial period – he called them a multi-ethnic trading company (Wolf 1982, 67; 1984, 394). The essential historical problem for Wolf was that Iroquoian communities were full of adopted non-Iroquois – Delaware, Nanticoke, Mohegan, etc. – who had joined the more powerful Iroquois. But if biology is independent of language and culture, then the simple movement of bodies into Iroquoia should not affect how we think about Iroquoian culture – what matters is how the immigrants acted. If they adopted Iroquoian language and culture and lived in Iroquoian houses and behaved according to Iroquoian rules, an Iroquoian identity was maintained. Wolf’s proposal that the Iroquoian tribal identity was a fiction created by French and English colonials can be defended only if you believe that biology equals culture – that cultures must be biologically ‘pure’ in some sense or they do not really exist. Wolf
2. STEPPE ARCHAEOLOGY: CULTURE, MIGRATION AND ARYANS
55
into Ohio but became culturally and materially Anglo-American. This concept of stable ethno-linguistic frontiers being permeable to people makes a lot of Western anthropologists uncomfortable. Many of us still seem to feel that if people move across an ethno-linguistic frontier freely, then the frontier does not really exist. But if biology is independent of language and culture, then the simple movement of bodies across a frontier should not affect how we think about its culture. What matters is how the boundary-crossers act. A distinctive cultural identity can exist for centuries in spite of the heterogeneity of the population. Edmund Leach (1954) redirected modern anthropology with the discovery that some people in Burma/Myanmar regularly adopted one ethnic identity on Sunday and a completely opposed one on Monday, depending on their business and social setting. But we have failed to investigate the opposite side of that behavioural frontier: while individuals could shift identities with astonishing ease, the neighbouring cultural systems that they moved between – upland Kachin hill farmers and lowland Shan paddy farmers or merchants – remained distinct. That stability in the local system of contrasts is just as interesting, and it is much more amenable to archaeological investigation than individual role-shifting. Aryan identity should be approached from this perspective. The archaeology of Aryan origins should recognise the possibility of long-lasting prehistoric ethno-linguistic identities – but there is no reason to expect skull shape to correlate. Instead, the reproduction and territorial expansion of the Aryan identity seems to have depended more on the public performance of a ritual drama, a kind of mythopraxis, in Sahlins’ (1985) terminology, but a drama shaped very much by political and economic contests for power. The historical Aryans believed that the relationship between gods and humans was based on the reciprocal exchange of gifts and favours. Humans gave a portion of their herds and wellcrafted verses of praise, and the gods in return provided protection from misfortune and the blessings of power and prosperity. ‘Let this racehorse bring us
did not believe this, but like many anthropologists he also did not address the contradiction. The Iroquois were distinctive culturally and linguistically, and the many adoptees in Iroquoian villages behaved for the most part like Iroquois. Moreover, the five ‘nations’ or tribes of the pre-European Northern Iroquois can be traced back archaeologically in their traditional five tribal territories to at least as early as AD 1300, more than 275 years before European contact. One could argue, in fact, that the five prehistoric ‘nations’ of the Northern Iroquois were older than most European nations at the time of contact, about 1575. This just doubles the irony in the idea that European nations ‘created’ the Iroquois in their own European image.
56
DAVID W. ANTHONY
good cattle and good horses, male children and all-nourishing wealth,’ pleaded a Sanskrit prayer in Book 1, hymn 162 of the Rig Veda. ‘Let the horse with our offerings achieve sovereign power for us.’ This relationship was mirrored in the mortal world when wealthy patrons among the Aryans sponsored the public feasts of cooked horse and beef flesh that were the concomitants of the proper sacrifice, providing the animals from their own large herds in return for the approval and loyalty of their clients. Another verse in the same hymn of sacrifice said, ‘Those who see that the racehorse is cooked, who say, “It smells good! Take it away!” and who wait for the doling out of the flesh of the charger – let their approval encourage us.’ The people received spectacle with their meat – they witnessed an elaborately scripted sacrifice punctuated by poems full of drama, rich in emotion, occasionally bawdy and earthy, and filled with clever metaphors and triple and double meanings. The poets believed that their words had an inherent transformative power because words could magically please or offend the gods, and because they were the stuff of verbal oaths and contracts, the legal glue of human social relations. Poetry in the Aryan language was central to the overall ritual performance. The best of these verbal displays were memorised, repeated and shared, and became part of the collective medium through which a variety of different people ended up speaking Indo-Iranian languages across most of the Eurasian steppes. ‘Let us speak great words as men of power in the sacrificial gathering,’ said the standard closing line attached to several different hymns in one of the oldest parts of the Rig Veda. This line expresses very well the linkage between the Aryan language, public ritual, verbal artistry and the projection of secular power: ‘Let us speak great words as men of power in the sacrificial gathering’ (RV 2. 12, 2. 23, 2. 28). There are many indications in the Aryan languages themselves that the Aryans did not come from India or Iran – they came from the north. Many words in the Rig Veda and Avesta were borrowed from a non-Aryan language, a local substrate. In fact the same substrate underlies both texts (Lubotsky 2001), which implies that the original contact was between the substratespeakers and very early proto-Iranian and proto-Sanskrit speakers. These early Aryans borrowed words for tiger, monkey, camel, ass, lentil, bread, canal, brick, permanent building and plough – familiar parts of the natural and cultural environment in eastern Iran and the Punjab, with which the original Aryan-speakers were apparently unfamiliar. They also borrowed the word for the soma plant, ançu, and almost certainly the soma ritual, from the same sub-
2. STEPPE ARCHAEOLOGY: CULTURE, MIGRATION AND ARYANS
57
strate. On the other hand, the poems of the Aryans show them to have had strongly pastoral values – the world was made from a sacrificed cow; the clouds were compared to dappled cows full of milk; milk and butter were the symbols of prosperity; and milk, butter, cattle and horses were the proper offerings to the gods. Horses were particularly important as ritual sacrifices and as the chargers that pulled chariots in war and in races. Horses were not native to India or Iran, and cattle-and-horse pastoralism was then, as now, at home on the Eurasian steppes and unusual in Iran or India. So the historical Aryan-speakers came from a natural environment different from that of eastern Iran and the Punjab, and not from an irrigation-based urban culture, but from a region where pastoral economies, horses and chariots were important. At the north edge of the Iranian and Afghan plateau was an ecological frontier that was a persistent cultural-economic boundary between 5000–1700 BC – and therefore probably a linguistic frontier as well. The Aryan languages had to originate north of this line, given their core vocabulary of ecological terms. The linguistic evidence points us pretty certainly toward the steppes, and in the centuries before 1500 BC. Can we identify the poetry-spouting, horse-sacrificing Aryans with any specific archaeological complex? Yes. The early stage in the spread of an Aryan identity across the eastern Eurasian steppes probably happened through the actual migration of people with a proto-Aryan language and culture from the Volga-Ural steppes (Poltavka and Abashevo cultures) into the steppes southeast of the Urals, a move that established the Sintashta-Arkaim culture in the upper Tobol river region. The Sintashta-Arkaim culture (ca. 2000–1750 calBC) was the first Eurasian steppe culture to display several traits later central to the Aryans, including compact fortified settlements, chariotry (which was much more than transport, for it provided many of the essential metaphors that coloured Vedic mythology and prayer); lavish expenditures of resources on the graves of the elite; and elaborate mortuary animal sacrifices that regularly involved the ritual slaughter of horses and cattle. Sintashta-Arkaim also witnessed another new behaviour – large-scale copper mining and bronze production. Steppe societies had been producing their own metals for almost 3000 years before 2000 calBC, but at about 2000 calBC, the dawn of the steppe Late Bronze Age, there was a significant increase in the scale and intensity of copper mining and in the technical sophistication of bronze weapons (Chernykh 1992). The subsequent spread of this ritual and technological complex from the SintashtaArkaim core region across the steppes of Kazakhstan to the edges of Iran and
58
DAVID W. ANTHONY
Afghanistan occurred under the material mantle of the Andronovo horizon (1800–1200 calBC: Fig. 2) and seems to have been accomplished through a mixture of small migrations and local emulation (Carpelan and Parpola 2001; Kuzmina 1994). Relatively small elite populations have encouraged widespread language shift toward their language, even in tribal contexts, where they introduced new religions or political ideologies and dominated regional wealth structures by controlling key trade routes, mines, or trade commodities (Atkinson 1989; Bentley 1981). Patron-client relationships, where patrons accepted responsibility for protection and wealth-sharing and clients returned service or tribute, have always been important vectors of language change, as long as vertical social mobility was a regular part of the system – so that clients could hope that, with luck, they or their children could become patrons themselves (Mallory 1992). Pashtun-speakers who were defeated or lost their homes in occasional inter-clan wars in southern Afghanistan tended to shift over to the Baluchi side and become ethno-linguistically Baluchi in the early 20th century for just this reason – the Baluchi patron-client system was open to the recruitment of outsiders, and offered vertical social mobility, unlike the Pashtun system (Barth 1981). Archaeologists might hope to find the symbols of authority, regalia and rituals that accompanied the spread of a new chiefly patron-client system in the steppes, but the spread of such elite material signals would have a very different appearance from the spread of a folk culture associated with a longdistance migration, such as the spread of Neolithic Cri¤ farmers. The Andronovo case seems to fall between these poles: while elite status symbols (metals, ornaments) and mortuary sacrificial rituals did spread with Andronovo, a regionally diverse set of domestic innovations also spread (a set of new ceramic forms and decorations, new settlement types, copper-mining complexes). The regional diversity in these domestic traits has been interpreted as a sign of local population diversity. Andronovo pastoralists lived on the outskirts of the fortified mud-brick towns of the Bactria-Margiana Complex (BMAC) on the northern fringes of Iran and Afghanistan between about 1800–1650 calBC – just before these towns were abandoned and the BMAC system collapsed. And Andronovo pots, at least, were used and discarded inside the BMAC temple rooms where ephedra, probably the drug called soma/haoma in the Aryan texts, was deposited – archaeological evidence that supports the linguistic evidence suggesting that Aryans learned the soma ritual from non-Aryans. Andronovo is in the right place, and
Fig. 2. Contact between the Srubnaya-Andronovo horizon and the Bactria-Margiana complex, ca. 1800–1600 BC.
2. STEPPE ARCHAEOLOGY: CULTURE, MIGRATION AND ARYANS
59
60
DAVID W. ANTHONY
at the right time, with public mortuary rituals, graves, economy, chariot technology, weaponry and social structures that parallel those described for the Aryans of the Rig Veda. The spread of Aryan languages southward across Iran or Afghanistan and into Pakistan/India probably occurred during the archaeologically confusing centuries that followed the collapse of the BMAC civilisation and related urban polities like Malyan in Iran in about 1700–1600 BC. We can argue about exactly how this happened, but the debate itself is not a sign of bigotry or naivety. It is perfectly legitimate to pursue Aryans archaeologically in the steppe cultures of the Late Bronze Age between 2000 and 1500 BC. Whether one wants to do this, or thinks it is a theoretically bankrupt romantic search for origins is another matter, but it cannot be said that an archaeology of Aryan origins is impossible, as many seem to feel (LambergKarlovsky 2002).
A Different Perspective I began by identifying three problem areas in steppe archaeology, three sources of misunderstanding between Western and steppe archaeological traditions. Rather than repeating the standard Western criticisms I have tried to show that steppe archaeology challenges both Western and traditional post-Soviet habits of thought. Long-lived contrasts in tribal cultural identities are a real and poorly-understood aspect of prehistoric cultural dynamics in the steppes; the study of migration is more complicated and useful than the billiard-ball analogy suggests, particularly when it is recognised as a social strategy used in competition for prestige and power; and the Aryan question can be approached responsibly through a careful combination of archaeology and linguistics. Neither Western nor post-Soviet archaeologists have yet realised the advantages that can be gained by looking at these old, familiar problems from a different perspective, with new assumptions and methods. That is without any doubt the reward that this still-new collaboration offers. Perhaps it is up to the next generation of archaeologists, those who organised this volume, to collect that reward.
2. STEPPE ARCHAEOLOGY: CULTURE, MIGRATION AND ARYANS
61
Bibliography Ammerman, A.J. and Cavalli-Sforza, L.L. 1979: The Wave of Advance Model for the Spread of Agriculture in Europe. In Renfrew, C. and Cooke, K.L. (eds.), Transformations: Mathematical Approaches to Culture Change (New York), 275–94. Anthony, D.W. 1997: Prehistoric Migration as Social Process. In Chapman, J. and Hamerow, H. (eds.), Migrations and Invasions in Archaeological Explanation (BAR International Series 664) (Oxford), 21–32. Anthony, D.W. and Brown, D. 1991: The Origins of Horseback Riding. Antiquity 65, 22–38. Anthony, D.W., Telegin, D. and Brown, D. 1991: The Origin of Horseback Riding. Scientific American 265.6, 94–100. Atkinson, R.R. 1989: The Evolution of Ethnicity among the Acholi of Uganda: The Precolonial Phase. Ethnohistory 36.1, 19–43. Barth, F. 1981: Ethnic Processes on the Pathan-Baluch Boundary. In Barth, F. (ed.), Features of Person and Society in Swat. Collected Essays on Pathans. Selected Essays of Fredrick Barth vol. 2 (London), 93–102. Bentley, G.C. 1981: Migration, Ethnic Identity, and State Building in the Philippines: The Sulu Case. In Keyes, C.F. (ed.), Ethnic Change (Seattle), 118–53. Carpelan, C. and Parpola, A. 2001: Emergence, Contacts and Dispersal of Proto-Indo-European, Proto-Uralic, and Proto-Aryan in Archaeological Perspective. In Carpelan, C., Parpola, A. and Koskikallio, P. (eds.), Early Contacts Between Uralic and Indo-European: Linguistic and Archaeological Considerations (Helsinki), 55–150. Chambers, J. and Trudgill, P. 1980. Dialectology (Cambridge). Chernykh, E.N. 1992: Ancient Metallurgy in the USSR: The Early Metal Age (Cambridge/New York). —— 1995: Postscript: Russian Archaeology after the Collapse of the USSR – Infrastructural Crisis and the Resurgence of Old and New Nationalisms. In Kohl and Fawcett (eds.) 1995, 139–48. Clark, G.A. 1994: Migration as an Explanatory Concept in Paleolithic Archaeology. Journal of Archaeological Method and Theory 1, 305–43. DeBoer, W.R. 1990: Interaction, Imitation, and Communication as Expressed in Style: The Ucayli Experience. In Conkey, M.W. and Hastorf, C.A. (eds.), The Uses of Style in Archaeology (Cambridge), 82–104. Efremov, I.A. 1940: Taphonomy: A New Branch of Paleontology. Pan-American Geologist 74, 81–93. Emberling, G. 1997: Ethnicity in Complex Societies: Archaeological Perspectives. Journal of Archaeological Research 5.4, 295–344. Erdosy, G. (ed.) 1995: The Indo-Aryans of Ancient South Asia (Indian Philology and South Asian Studies 1) (Berlin). Fischer, D.H. 1989: Albion’s Seed: Four British Folkways in America (Oxford). Gamble, C. 1999: The Paleolithic Societies of Europe (Cambridge). Hock, H.H. and Brian, D. 1996. Language History, Language Change, and Language Relationship (Berlin). Kelley, R. 1985: The Nuer Conquest. The Structure and Development of an Expansionist System (Ann Arbor). Kohl, P.L. and Fawcett, C. (eds.) 1995: Nationalism, Politics, and the Practice of Archaeology (Cambridge). Kopytoff, I. 1987: The Internal African Frontier: The Making of African Political Culture. In Kopytoff, I. (ed.), The African Frontier. The Reproduction of Traditional African Societies (Bloomington), 3–84. Kuzmina, E.E. 1994: Otkuda prishli Indoarii? (Moscow). Lamberg-Karlovsky, C.C. 2002: Archaeology and Language: the Indo-Iranians. Current Anthropology 43.1, 63–88. Leach, E. 1954: Political Systems of Highland Burma (Boston).
62
DAVID W. ANTHONY
Lubotsky, A. 2001: The Indo-Iranian Substratum. In Carpelan, C., Parpola, A. and Koskikallio, P. (eds.), Early Contacts Between Uralic and Indo-European: Linguistic and Archaeological Considerations (Helsinki), 301–17. MacEachern, S. 2000: Genes, Tribes, and African History. Current Anthropology 41.3, 357–84. Mair, V. (ed.) 1998: The Bronze Age and Early Iron Age Peoples of Eastern Central Asia (Washington, DC). Mallory, J.P. 1992: Migration and Language Change. Universitetets oldsaksamlings skrifter ny rekke 14 (Oslo), 145–53. Merpert, N.Y. 1992: K publikatsii pisma V.G. Childa. RosA 4, 184–96. Nash, G.B. 1984: Social Development. In Greene, J.P. and Pole, J.R. (eds.), Colonial British America (Baltimore), 233–61. Sahlins, M. 1985: Islands of History (Chicago). Semenov, S.A. 1964: Prehistoric Technology (Bath). Shnirelman, V.A. 2001: The Value of the Past: Myths, Identity and Politics in Transcaucasia (Osaka). Snow, D.R. 1995: Migration in Prehistory: the Northern Iroquoian Case. American Antiquity 60.1, 59–79. —— 1996: More on Migration in Prehistory: Accommodating New Evidence in the Northern Iroquoian Case. American Antiquity 61.4, 791–6. Vasiliev, I. (ed.) 1995: Drevnie Indoiranskie Kul’tury Volgo-Ural’ya (II tys. do n.e.) (Samara). Wilhelm, H.G. 1992: Germans in Ohio. In Noble, A.G. (ed.), To Build in a New Land. Ethnic Landscapes in North America (Baltimore), 60–78. Wolf, E.R. 1982: Europe and the People Without History (Berkeley). —— 1984: Culture: Panacea or Problem? American Antiquity 49.2, 393–400. Yablonsky L.T. 1996: Saki Iuzhnogo Priural’ya (arkheologia i antropologia mogilnikov) (Moscow).
CHAPTER THREE GENERAL MIGRATION PROCESSES IN THE ARAL SEA AREA IN THE EARLY IRON AGE LEONID T. YABLONSKY
From a geographical point of view, the southern Aral Sea region is limited by the modern and ancient (now dry) system of channels of the Amu Darya river. The area occupied by the Syr Darya river belongs to the eastern Aral Sea area (Fig 1). Since very ancient, at least Neolithic, times the southern Aral Sea area archaeologically and anthropologically has essentially been part of the Eurasian steppes (Vinogradov et al. 1986). Because of the specific ecology that is connected with the changeable behaviour of the Amu Darya’s water flow, this territory always had its own cultural peculiarity. The western Sari Kamish delta of the Amu Darya now represents an open alluvial plain formed mainly by multiple delta channels. Some of the channels are clearly visible even today. Takirs (a Turkmen word denoting flat, sandless and very dry naturally barren spaces) constitute the geomorphological bases of the delta. Multiple hills formed from limestone and covered by sand make a notable part of the local landscape. They are situated mainly in the western part of the ancient delta, close to the border of the Kara Kum desert and the Sari Kamish basin which was periodically flooded or dry (depending on the main direction of the Amu Darya). The multiple asynchronous kurgan cemeteries belonging to the Early Iron Age (the 1st millennium BC and beginning of the 1st millennium AD) are located on the top of these hills. The deltaic channels of Amu Darya and Syr Darya rivers united periodically, ensuring latitudinal cultural-genetic connections. It was also thus in the Early Saka era. The southern Aral Sea area, especially the Sari Kamish delta of the Amu Darya, was relatively isolated geographically: to the south was the almost impassable region of the Kara Kum desert; the Usturt plateau which is rocky and covered by alkali soil and almost always waterless is situated to the northeast; the Sari Kamish lake and trans-Uzboy plateau are located to the west. Life here depended on the changeable water flow of the Uzboy river. Further to the
Fig. 1. Map of the region.
64 LEONID T. YABLONSKY
3. GENERAL MIGRATION PROCESSES IN THE ARAL SEA AREA
65
west there is the big waterless space of the trans-Caspian lowland. The very salty Aral Sea and the periodically dry or waterlogged channels of the Akcha Darya and northern Amu Darya deltas are situated to the north. These borders were never impenetrable, and populations, heterogeneous from both the biological and cultural points of view, appeared here periodically. They took an active part in local ethnogenesis and shaped the specific direction of local ethnic processes. In this paper, we would like to briefly discuss the principal points concerning the influence of migrations on these processes.
Pre-Achaemenian Times (Prior to the 6th Century BC) The migration of heterogeneous nomads at the beginning of the Early Iron Age (the late 8th and 7th centuries BC) became the starting point of the local ethnogeneses. At the end of the 2nd and beginning of the 1st millennium BC, global climatic changes took place in the Volgo-Urals, northern Kazakhstan and Mongolian steppes. From a palaeo-ecological point of view, these changes were marked by the displacement of the natural climatic zones, with changes in soil compositions and in the degree of atmospheric moisture. Set in historical context, these changes caused numerous crises for the cultures that dated back to the Late Bronze Age. Movements of forest-steppe and steppe populations as well as cultural and genetic diffusion finally gave rise to the formation of the Saka cultures in the eastern part of the steppe zone. Apparently, it is possible to connect the almost absolute absence of archaeological sites dating to the 8th–7th centuries BC in the huge territory of the steppes located from the Danube up to the southern Urals area with this ecological crisis (Zhelezchikov et al. 1995). At the same time, the aridisation and intensification of the continental climate that began in the territory of central Mongolia was accompanied by changes in the landscape: the floodplains and foothill valleys were transferred to the steppes (Demkin 1997). Similar processes took place in the forest-steppe in the west Siberian area. At the same time the beginning of a more moist climate is marked in the northern Kazakhstan region that, in part, resulted in a departure to the south of the population whose cultural-economic type had been formed in the stable forest-steppe zone and at the southern border of the forest. Expansion of the steppe culture area is traced to the north, up to territory of the southern trans-Ural, southern
66
LEONID T. YABLONSKY
Urals and Bashkortostan (Khabdulina and Zdanovitch 1984). This new beginning of the pre-Aral Sea ethnogenesis in the Early Iron Age was closely connected with these climatic changes. In the Neolithic epoch (5th–4th millennia BC), the bearers of the Kelteminar culture settled the Sari Kamish delta of the Amu Darya. During the Bronze Age, drainage (Vainberg 1991a, 126) and swamping (Itina 1977, 32) of the delta occurred in relation to the active fluctuating of the Akcha Darya’s channels. Because of this, the territory of the Sari Kamish delta became, at this time, unfit for constant habitation. Local settlements belonging to the Late Bronze Age Amirabad culture arose here not earlier than the 8th century BC and apparently disappeared at the end of the century. At the end of the 8th century BC, the settlements and cemeteries of the new Kuyusay culture and the synchronous cemeteries established by the Saka appeared (Vainberg 1979a; Yablonsky 1986; 1996). It is absolutely clear that since that time, after the animal-breeders had arrived, the Sari Kamish delta once again became the centre of the ethno- and cultural genesis. At the beginning of the 1st millennium BC, the water in the Akcha Darya channels began to dry up, resulting in the desolation of the region. Archeological finds of Saka arrowheads were made on the surface of the takirs but Saka regular settlements and cemeteries are unknown here. The settlement at Yakke-Parsan 14 dates to the Early Saka period and possesses transitional signs of the Late Bronze Amirabad culture, but the latter did not itself continue into later times (Itina 1977, 171–2). Thus the territory of the Akcha Darya delta should be excluded from the geographical zone of the Saka ethnogenesis in the Aral Sea area. In contrast, in the Sari Kamish region, we have multiple settlements and cemeteries belonging to that time. The cattle-breeder settlement at Kuyusay 2, the cemeteries of Tumek-Kichidzhik and Tarim-kaya 1 (Vainberg 1979a; 1979b) and the cemeteries of the Sakar-Chaga type (Yablonsky 1991a; 1996) also belong to pre-Achaemenian times. The discovery of the Sakar-Chaga type cemeteries in the Sari Kamish region contradicted descriptions of the local population as monocultural in character during the pre-Achamenian period (Vainberg 1979b; 1992) and challenged suggestions that the region did not belong to the Saka world (Yablonsky 1986; 1996). It was recognised that, at the beginning stage of colonisation, complex interactions between different cultures stimulated the development of local populations in this relatively small region (Vainberg 1991a, 126; 1996, 10; Yablonsky 1991b). The majority of graves in the cemeteries of Tumek-Kichidzhik and Tarim-
3. GENERAL MIGRATION PROCESSES IN THE ARAL SEA AREA
67
Kaya 1 were placed in narrow and shallow rectangular burial pits (Fig. 2). A western or eastern orientation of the deceased predominates. Bodies were placed in a supine position, but a considerable number of skeletons in the Tarim-Kaya 1 cemetery had been placed on their left or right sides in a flexed position. Burials of dismembered bodies were also found, as well as multiple burials in the same tomb chamber. In the latter case, the earliest skeletons were partially destroyed or replaced. Such burials had, on the whole, a poor set of grave goods; more often there was only a hand-made pot placed at the head of the deceased. Construction of the burial chambers, orientation of the deceased, a combination of the burials in extended and flexed position, and the poverty of grave goods are all typical signs of the Srubnaya culture burials located in the western part of the steppes and dating back to the Late Bronze Age (Dvornichenko and Koreniako 1989, 149–50). Burials of dismembered bodies are also well known in the materials of the Srubnaya culture. In part, they are present in the Late Bronze cemeteries located in the Caucasus and Central Asia which are connected with the southward migration of the Srubnaya tribes of the steppe. The main features of the funeral rite remind us of the old hypothesis of B. Grakov (1954, 167) who dated the last wave of Srubnaya migration from the Volga river area southward to the end of the 8th or beginning of the 7th century BC. This hypothesis accords well with current ideas about Srubnaya migration not only to the Black Sea and pre-Caucasus areas (Kozenkova 1982) but also to Central Asia (Kuzmina 1994). At the same time, general features of the funeral rite at the cemeteries of the Sakar-Chaga type show similarities with Late Bronze and Early Iron Age sites in the eastern part of the Eurasian steppes and Saka cemeteries in the Syr Darya delta. Their burial rite included: 1) a combination of cremations and inhumations; 2) burials built on both the ancient soil surface and in deep pits; 3) incomplete cremation due to construction of the kurgan mound while the funeral pyre still blazed; 4) the kindling of several fires around the funeral places; 5) pole constructions inside the burial chamber and on the ancient soil surface;1 6) layers of cane plants inside and on the bottom of the burial chambers; etc. (Fig. 3). The craniological type found in the 1
During the excavations we found around the burial place or inside the burial pits a system of smaller pits having a diameter of 10–25 cm and depth of 20–50 cm, each marking the spot where a wooden pole had actually stood. Sometimes there were only imitations of the wooden constructions in which no traces of wood were found, and sometimes funeral goods had been placed in the pits instead.
68
LEONID T. YABLONSKY
Fig. 2. Burials excavated in the Tumek-Kichidzhik (I) and Tarim-Kaya (II) cemeteries.
3. GENERAL MIGRATION PROCESSES IN THE ARAL SEA AREA
69
Fig. 3. Burials excavated in the Sakar-Chaga cemetery. 1 – multiple; 2 – burned construction; 3 – inhumation.
70
LEONID T. YABLONSKY
cemeteries belongs to the Kuyusay culture and is specific and unique within the territory of Central Asia (Trofimova 1974a; 1979; Yablonsky 1996). The skulls are characterised by the protomorphic sharply Europoid (Caucasoids) features. They are dolichocranic, and the cranial height is very large. The facial skeleton is very wide but the height of face is low. The facial skeleton is well profiled. The skulls on the whole are very big and very massive. The skulls that came from the Sakar-Chaga cemeteries demonstrate an absolutely different craniological complex. The male skulls are brachicranic. They have relatively small cranium sizes and an average size of the facial skeletons. The female series, according to form and relative cranium sizing are similar to males, but their facial skeleton is flat and the angle of the nose bones is small (Figs. 4–5). A complex of craniological signs testify to their oriental origin. According to their general features, the skulls join well to the craniological variations typical for the Sakas of Kazakhstan. Lastly, the burials excavated in the cemeteries of the Sakar-Chaga type consisted of the so-called ‘Scythian Triad’. This special complex of funeral goods is represented by weapons, horse furnishings and items decorated in the so-called ‘animal style’ (Yablonsky 2000a). Thus both archaeological and anthropological data testify that at least two different population groups settled the Sari Kamish area in the Early Saka period. One of these groups had a Volgo-Ural origin and the other one came to the Amy Darya delta from the east. Both components created the cultural and genetic basis of the local population. Nevertheless, we need to stress that in pre-Achaemenian times the sum of the cultural differences between the groups was greater than the sum of the similar features of their material cultures. It possible that later such close interaction between these groups made the formative basis of the southern Aral population that had formed an early state system of archaic Khorezm and gave the principles of the Khorezmian ethnos.
Achaemenian Period The Saka funeral rite described above was practised up to the middle of the 6th century BC. We do not believe that the crisis of the Saka culture took place in the Sari Kamish area any later than the middle of the 6th century BC. Evidence for this is reflected in both the archaeology of funeral rites and in the
3. GENERAL MIGRATION PROCESSES IN THE ARAL SEA AREA
Fig. 4. Female skulls from the Sakar-Chaga 3 cemetery.
71
72
LEONID T. YABLONSKY
Fig. 5. Male skull from the Sakar-Chaga 3 cemetery; and facial reconstruction by author.
3. GENERAL MIGRATION PROCESSES IN THE ARAL SEA AREA
73
available osteological data. A very specific craniological type residing in the Early Saka population disappeared after that time. We have no reason to suggest that ecological changes provoked this crisis. We have both direct and indirect evidence for this opinion. The direct evidence is that after the Sari Kamish delta was supplied with water at the beginning of the 1st millennium BC, the water supply remained stable during the remainder of this millennium (Vainberg 1997, 25). Indirect evidence is provided by the ethno-cultural situation in the Syr Darya delta, most notably the close similarities between the funeral rite of the local Saka tribes and those of the Sari Kamish region. These similarities were found not only in the construction of the burial chambers, but also in the set of grave goods. In the middle of the 6th century BC, the types of these grave goods changed as a result of the reorientation of the region to a different metallurgical centre and a change in military tactics. Nevertheless, there is no doubt that the Saka culture, even in this transformed state, inhabited the region under ecological conditions similar to Amu Darya up to 5th century (Itina and Yablonsky 1997). The crisis of the Saka culture in the Sari Kamish area began as a result of circumstances that did not take place in the Syr Darya area. It is important to point out that after the Achaemenian conquest of Central Asia in the last third of the 6th century BC, Khorezm was subdued by Persia. This event could have been accompanied by a loss of grasses that led to the destruction of the Saka animal-breeders. However, the crisis occurred in pre-Achaemenian times and most probably was connected with the rise of the Ancient Khorezmian state. This state was formed not later than the end of the 7th century BC (Rapoport 1996, 56; Itina et al. 1996, 24; Vishnevskaya and Rapoport 1997, 150). From that time up to the middle of the 6th century BC, the Saka culture coexisted with the Khorezmian state in the limited region of the delta (not more than 100 km). It is well known from the archaeological data that this coexistence caused the development of the classic Khorezmian culture throughout the Achaemenian period. At the same time, the culture of the Saka completely disappeared from the Khorezmian territory. It is most probable that the main part of the cattle-breeders were forced to leave the delta because they were dissatisfied with the policies imposed by the government that had seized leadership in the territory.
74
LEONID T. YABLONSKY
Early Hellenistic Period At the turn of the 5th and 4th centuries BC, Khorezm achieved political independence from Achaemenian power, marking a new Hellenistic stage of its historical development. The upper chronological limit of the period is not well determined but the 1st century BC is the most probable date. This date seems to be right because the time between the 2nd and 1st centuries BC was a period marked by the devastation and ensuing desolation of many Khorezmian fortresses, settlements and temples. Scholars have different points of view on the reasons for these events. Some (Itina et al. 1996, 25) connect the events with pressure from the nomadic tribes who participated in the devastation of the Graeco-Bactrian kingdom; but another (Vainberg 1991b, 39–40) believes that they are connected to religious reforms which in part led to the final destruction of cult buildings with secular ground plans. It is important to stress that the minting of the most ancient Khorezmian coins and the use of the Khorezmian calendar began just in this time as a new dynasty came to rule (Itina et al. 1996, 25). At the beginning of this period (the end of the 5th century BC), following a long hiatus, new kurgan cemeteries appeared again in the Khorezmian territory. This archaeological fact should be connected with the general historical situation in the region. After the Achaemenians lost power, the Khorezmian state probably instituted protectionist politics with reference to the cattle-breeders who had settled in the periphery of the state, particularly in the Sari Kamish area (Vainberg 1991a, 136; 1991b, 46). The appearance of a new wave of animal-breeders in this area is evidenced by the collective tombs (Fig. 6.1) made from pits with dromos (Yablonsky, 1998) and by burial pits with side niches ( podboy burials)2 (Yablonsky 1999) (Fig. 6.2). It is possible that this wave of breeders arrived as a result of nomadic migrations caused by external ecological factors. A period of aridisation in the Ural steppes falls just between the 4th and 2nd centuries BC (Demkin and Riskov 1996a; Demkin 1997, 158). This was the reason for the mass outflow of nomads in that part of the Ural steppes (Tairov 1995). Nevertheless, the process of aridisation and desertification of the steppes did not have a global character. In the region of the southern Ural steppes, the dynamics of the natural situation 2
The Russian term for a special burial pit construction, in which a rectangular pit was first dug, and then a niche was dug in the side of the pit (lengthwise) in which the body of the dead was placed (Fig. 6).
3. GENERAL MIGRATION PROCESSES IN THE ARAL SEA AREA
75
Fig. 6. Multiple (1) and podboy (2) pits from the Sakar-Chaga 1 cemetery; podboy pit from the Gyur 4 cemetery (3).
76
LEONID T. YABLONSKY
did not exceed the limits of steppe and dry-steppe conditions and remained favourable for the nomadic way of life (Demkin and Riskov 1997, 144). This was one of the reasons for the concentration of the nomadic Sarmatian tribes in this region at the first stage of their cultural formation. Most probably, this increased concentration of nomads led to the demographic crisis that caused the migrations of the Sarmatians during the 3rd–2nd centuries BC to move not only to the west (Zhelezchikov 1983) but also to the south, to the oases of Central Asia (Skripkin 1984; 1990). The notable similarities in some features of the classic Sarmatian and Khorezmian funeral rite make it possible for us to believe that the Sari Kamish territory was also open for Sarmatian migration at this time. Osteological data provide clear evidence of a change in the physical type of the local population as compared with the previous Saka period. Some representatives of the Kuyusay culture could have supplied their own contribution to the process of the formation of the Khorezmian population’s anthropological features in the second half of the 1st millennium BC, but this contribution was not too substantial. Craniological data gives graphic evidence of the common (on a high taxonomic level) origin of the Sarmatian and Khorezmian populations in this period. The geographical motherland of the common physical type is probably located in the trans-Ural steppe and forest-steppe during the Sauromatian period (6th–5th centuries BC). Archaeological data also support this presumption. Hand-made pots with rounded bottoms and talc admixture in the clay first appear in the Aral Sea area together with collective and podboy burials during this Sauromatian period. At the same time, there are no necropoleis in the area which, according to the complex of archaeological data, could be considered to belong to the classic Sarmatians. The different orientation of the dead, the absence of funeral goods typical of the classic nomads (such as weaponry and horse harnesses), the custom of multiple burials in the same grave pit, and the tradition of replacing previous skeletons inside the pits (Fig. 6.1–2) all testify to the undulating migrations that led to the syncretism of the Khorezmian funeral rite. This syncretism arose out of the heterogeneity of the migrants themselves and was based on local cultural and ideological traditions. From an historical point of view, the habitation of the Sari Kamish area by the culturally heterogeneous groups of the eastern nomads was the result of the general changes caused by military-political (Dandamaev 1963) and ecological (Demkin and Riskov 1997) events. In the course of these events, the early Sarmatian
3. GENERAL MIGRATION PROCESSES IN THE ARAL SEA AREA
77
archaeological culture was formed. Its heterogeneous physical anthropological base is without doubt (Yablonsky 1997; 2000b). At the same time and on a similar cultural-genetic base, the formation of a separate group of Khorezmian cattle-breeders took place. The specificity of this process is emphasised by the custom of artificial cranial deformation (Fig. 8). The classic Sarmatians never used this custom. The notable syncretism of the funeral rite of the Central Asian animal-breeders was marked in the archaeological literature by the strange term ‘Sarmatoids’. This might be explained by the multi-component basis, which has original features among the populations that had formed the proper Sarmatian ethnos. This tendency is outlined in the Sari Kamish funeral rite not later than the middle of the 4th century BC and marked the irreversible process of the cultural and ethnic divergence of the Aral Sea and Ural animal-breeders. At this time, the transition to the custom of burying de-fleshed human bones in special ossuary vessels took place in the Khorezmian territory. Rare cases of the reburial of human bones in everyday vessels relate to this period. The grave pits where de-fleshed bones were buried have also been investigated. These features marked the first steps in the transition from inhumation to the Zoroastrian burial custom. The transition to this funeral rite occurred, according to current archaeological data, not earlier than the 1st century BC or the middle of the 2nd century BC when the global historical changes described above took place Khorezm. At this time, the local population, including Khorezmian cattle-breeders, began to use a funeral rite that placed de-fleshed bones in very different types of ossuaries (Fig. 7). Simultaneously, the burials having classic nomad goods (a rich complex of weaponry which in part included the composite bow of the ‘Hun’ type) appeared in the Khorezmian territory in this period (Fig. 6.3). The group of nomads that left such cemeteries also brought the new type of ceramics. One main feature characterises these ceramics: all the pots are alien to Khorezm. Vessels of Parthian origin were found among the pots (Maslov and Yablonsky 1996). The burial chambers also have different types. The simple pits, podboys, catacombs and pits that have wood covers were all excavated in the same cemetery. It is important to mention that catacombs appeared in the Khorezmian territory now for the first time. The different types of grave pits and ceramics testify to the cultural heterogeneity of this population. The anthropological data also give the idea about physical specificity of the population compared with the proper Khorezmian population (Fig. 9).
78
LEONID T. YABLONSKY
Fig. 7. Khorezmian ossuaries, 2nd century BC–2nd century AD.
3. GENERAL MIGRATION PROCESSES IN THE ARAL SEA AREA
79
Fig. 8. Male and female from the Sakar-Chaga 1 cemetery. Facial reconstructions by author.
Fig. 9. Male from the Gyur 4 cemetery; female from the Tumek-Kichidzhik cemetery. Facial reconstructions by author.
80
LEONID T. YABLONSKY
Obviously, these grave pits belong to the groups of nomads who returned to the northern steppes after the conquest of the Graeco-Bactrian state. Cemeteries that had typologically the same set of grave goods appeared on the Uzboy plateau at the same time (Usupov 1986). Cemeteries of a similar cultural type but relatively late (1st–3rd centuries AD) number kurgans in the tens and hundreds, having both podboys and catacombs (Lokhovits and Khazanov 1979, 111). There are no such cemeteries within the limits of the Uzboy region, nor in the Usturt plateau. These territories became the arenas of nomadic migrations from the south and a concentration of nomads took place just in the Khorezmian territory. This fact is not accidental. The excavators of the sites noted their close similarity to the synchronous kurgan cemeteries located to the south of Khorezm, in the oases of Bukhara and Fergana (Lokhovits 1968, 156–67). They believed that Khorezmian necropoleis belonged to the immigrants who came from the south and who included a Sarmatian component. Nevertheless, according to the archaeological data, newcomers settled in the Sari Kamish area and very soon entered into an active interaction with the local agricultural populations. The skulls coming from these cemeteries have similarities to both classic Sarmatian burial and the local craniological series from the Zoroastrian necropolis excavated in the fortress of Kalali-gir 2 (Trofimova 1959, 30–79). They are also similar to the craniological series excavated at the cemeteries of SakarChaga 1 ( podboys) and Tarim-Kaya 2 (ossuary) which belonged to the local Khorezmian population. It is important to stress that these last sites are dated earlier than cemeteries of the newcomers. The cultural and physical similarities described above suggest that these cemeteries were left by the population of Khorezmian animal-breeders who participated in the Graeco-Bactrian war, and who returned to their homeland after the state’s defeat. This is why it was easy for them to settle in the Khorezmian territory and come into contact with both local farmers and administrators. Nevertheless, they retained the funeral traditions that they had acquired while living outside Khorezm.
Kushan and Afrigidean Periods (4th–5th Centuries AD) The Khorezmian necropoleis of this period allow us to trace another parallel ethno-genetic line connected with an unbroken cultural-historical development
3. GENERAL MIGRATION PROCESSES IN THE ARAL SEA AREA
81
of the local cattle-breeders. This line is clearly shown archaeologically and anthropologically in the materials that came from the ossuary cemeteries. Of particular significance among the sites of this time are the burials that contained the de-fleshed bones of people with artificially deformed heads (although now a distinct circular form of deformation). The cemetery of Yassi-gir 4 is a type site. The sculptural anthropomorphic ossuary vessel that came from the cemetery is unique (Yablonsky 1989). It was produced from good clay and the surface of the vessel was covered with red angob. It presents a human figure (Fig. 10) that below the waist transforms into the base of a cylindrical pot. The general height of the vessel is 58 cm. Features of the head and face were made in a realistic manner. Artificial cranial deformation of a circular type was stressed. The nose is long and has a high bridge. The relief of the frontal bone is smoothed. This is usual for females and children. The opening in the back of the anthropomorphic vessel was used to put the de-fleshed bones inside. It was closed with a disk-like cover which was fastened to the vessel’s body with alabaster. The left hand of the figure was turned palm up. The composition of the figure reminds us of a woman who has a baby in her hands. The eyes of the women are directed to the left hand where the baby’s head should be. The surface of the figure has no sign of clothes, hat, or headdress – all of these elements differ from other examples of anthropomorphic ossuary vessels known from Khorezmian sites. We believe that the clothes that covered the sculpture were made from organic material that was not preserved. According to the disposition of the head and hands, it is possible that a doll symbolising a baby was placed in the outstretched hands. This suggestion is supported by the fact that bones were found belonging to a 6 or 7-yearold girl inside of the pot. It is interesting that the skull of the interred girl was artificially deformed in the same manner as the figure’s head. As was mentioned above, this ossuary had no similarity with the examples represented in the collection of the Khorezmian ossuaries. It is especially interesting that technological, morphological and semantic analogues to our figure can be found far from Khorezm, in the eastern part of the Asia. In this sense, the clay figures known from the Mongolian territory are also interesting. These were found in burials dating to AD 386–534. Both figures represent warriors that have their hands pulled apart (Kessler 1995, Fig. 50). It is important to mention that the production technology of the Yassi-gir figural vessel had similar features: the hands were made separately from the body and connected using the ‘shank-bushing’ technique. From a semantic point of view, the figures
82
LEONID T. YABLONSKY
Fig. 10. Sculpted ossuary from the Yssi-gir 4 cemetery.
Fig. 11. Male from Yssi-gir 4 cemetery. Facial reconstruction by author.
3. GENERAL MIGRATION PROCESSES IN THE ARAL SEA AREA
83
of warriors appear to be standing guard. These warrior figures were presented in the Chinese archaeology exhibition at the Los Angeles Museum. The author of the catalogue mentions that the tomb and grave goods suggest the influence of the Chinese culture in the area of Inner Mongolia at the beginning of the 1st millennium AD (Kessler 1995, 86). Archaeological, osteological, numismatic and other data testify to the Pre-Syrdarya and probably more eastern origin of the migrants who were found in the Yassi-gir 4 cemetery (Fig. 11). The archaeological evidence indicates important changes in the life of the Khorezmians during the 4th–5th centuries AD that led to the formation of a culture very different from the preceding local traditions of the Hellenistic period. These changes cannot be connected with the movement of Huns that took place in the 4th century BC. From the archaeological and osteological points of view, it is possible to suggest that the materials coming from the Yassi-gir 4 cemetery give us evidence of one of the Hun populations that came to the Khorezmian territory settled there and assimilated some features of the Khorezmian culture and ideology (Yablonsky and Bolelov 1994, 32–3). The processes shaping the creation of the Khorezmian ethnos were never straightforward. Periodically they were interrupted by internal factors – ecological, military-political, social, religious and others – that acted in combination or separately. This process was periodically corrected by the appearance of new nomadic groups in the territory of the Aral Sea area. Some of these groups who settled in the periphery of Khorezm made their cultural and genetic contribution to the development of the ethno-historical and anthropological situation in the region at each chronological stage. The general direction of the ethno-genesis of the populations that settled in the animal-breeding periphery of Khorezm may serve as a model of the interactions between farmers and nomads in Central Asia. But such models should be modified according to the specificity of the concrete processes of ethnogenesis that passed into the local regions of the area.
Bibliography Dandamaev, M.A. 1963: Pohod Daria protiv skifskogo plemeni tigrahauda. Kratkie Soobshenia Instituta Narodov Azii AN SSR 61, 17–187. Demkin, V.A. 1997: Paleopochvovedenie i Arkheologia (Pushino). Demkin, V.A. and Riskov, Y.G. 1996: Pochvi i prirodnaya sreda suhih stepei Uzhnogo Urala v epohi bronzi i rannego zheleza (Pushino). Dvornichenko, V.V. and Koreniako, V.A. 1989: Predshestvenniki savromatov v Volgo-Donskom mezhdurechie, Zavolzhie i Uzhnom Priuralie. In Ribakov, B.A. (ed.), Arkheologia SSSR: Stepi Evropeiskoi chasti SSSR v skifo-sarmatskoe vremya (Moscow), 148–52.
84
LEONID T. YABLONSKY
Grakov, B.A. 1954: Kamenskoe gorodishche na Dnepre (Moscow). Itina, M.A. 1977: Istoria stepnih plemen Uzhnogo Priaralia (II-nachalo I tisycheletia do n.e.) (Moscow). Itina, M.A., Levina, L.M., Nerazik, E.E. and Rapoport, J.A. 1996: K 60-letiu Khorezmskoi arheologoetnograficheskoi expeditsii. Etnographicheskoe Obozrenie 6, 19–33. Itina, M.A. and Yablonsky, L.T. 1997: Saki Nizhnei Syrdar’i (po materialam mogilnika Uzhnii Tagisken) (Moscow). Kessler, A.T. 1995: Empires Beyond the Great Wall. The Heritage of Genghis Khan (Los Angeles). Khabdulina, M.K. and Zdanovich, G.B. 1984: Landshaftno-klimaticheskie kolebania golotsena i voprosi kul’turno-istoricheskoi situatsii v Severnom Kazahstane (Chelyabinsk). Kozenkova, B.I. 1982: Obryd krematsii v kobanskoi kul’ture. SA 3, 14–34. Kuzmina, E.E. 1994: Otkuda prishli Indo-Arii? (Moscow). Lokhovits, V.A. 1968: Novie dannie o podboinih pogrebeniah mogilnika Tuz-gir. In Vinogradov, A.E. (ed.), Istoria, arkheologia i etnographia Srednei Azii (Moscow), 157–64. Lokhovits, V.A. and Khazanov, A.M. 1979: Podboinie i katakombnie pogrebenia mogilnika Tuzgir. In Itina, M.A. (ed.), Kochevniki na granitsah Khorezma (Moscow), 111–33. Maslov, V.E. and Yablonsky, L.T. 1996: Mogilnik Gyur-IV v Severnoi Turkmenii. RosA 2, 168–81. Rapoport, J.A. 1996: Religia drevnego Khorezma. Nekotorie itogi issledovania. Etnographicheskoe Obozrenie 6, 53–77. Skripkin, A.S. 1984: Nizhnee Povolzhie v pervie veka n.e. (Saratov). —— 1990: Aziatskaya Sarmatia (Saratov). Tairov, A.D. 1995: Istoriko-ekologicheskaya situatsia v Uzhnom Zauralie vo vtoroi polovine I tis. do n.e. i rasselenie rannih sarmatov. In Kochkina, A.F. (ed.), Rossia i Vostok: problemi vzaimodeistvia (Chelyabinsk), 23–36. Trofimova, T.A. 1959: Cherepa iz ossuarnogo nekropolya kreposti Kalali-gir 1 (raskopki 1953 g.). In Debets, G.F. (ed.), Materiali Khorezmskoi expeditsii 2 (Moscow), 30–79. —— 1974a: Kraniologicheskie materiali iz mogilnika Tumek-Kichcidzhik. Sovetskaya Etnographia 5, 139–49. —— 1974b: Cherepa iz podboinih i katakombnih zahoronenii mogilnika Tuz-gir. In Zolotareva, I.M. (ed.), Rasogeneticheskie protsessi v etnicheskoi istorii (Moscow), 154–79. —— 1979: Cherepa iz pogrebenii kuusaiskoi kul’turi v mogilnikah Tumek-Kichidzhik i Tarimkaya. In Itina, M.A. (ed.), Kochevniki na granitsah Khorezma (Moscow), 77–93. Usupov, H. 1986: Drevnosti Uzboya (Ashkhabad). Vainberg, B.I. 1979a: Pamyatniki kuusaiskoi kul’turi. In Itina, M A. (ed.), Kochevniki na granitsah Khorezma (Moscow), 7–76. —— 1979b: Kurgannie mogilniki Severnoi Turkmenii. In Itina, M.A. (ed.), Kochevniki na granitsah Khorezma (Moscow), 167–77. —— 1991a: Drevnyaya istoria obvodnenia Prisarikamishskoi delti Amudar’I v svete arkheologicheskih rabot poslednih desytiletii. In Andrianov, B.V. (ed.), Aral’skii Krizis (Moscow), 123–41. —— 1991b: Izuchenie pamyatnikov Prisarikamishskoi delti Amudarii v 70–h–80–h godah. In Vainberg, B.I. (ed.), Skotovodi i zemledeltsi levoberezhnogo Khorezma (Moscow), 5–108. —— 1992: Pamyatniki skotovodcheskih plemen v levoberezhnom Khorezme. In Ribakov, B.A. (ed.), Arkheologia SSSR. Stepnaya polosa Aziatskoi chasti SSSA v skifo-sarmatskoe vremia (Moscow), 116–21. —— 1997: Ekologia Priaralia v drevnosti i srednevekovie. Etnographicheskoe Obozrenie 1, 23–41. Vinogradov, A.V., Itina, M.A. and Yablonsky, L.T. 1986: Drevneishee naselenie nizovii Amudar`i (arheologo-antropologicheskoe issledovanie) (Moscow). Vishnevskaya, O.A. and Rapoport, J.A. 1997: Gorodishe Kuzeli-gir. K voprosu o rannem etape istorii Horezma. VDI 2, 150–73. Yablonsky, L.T. 1986: Mogilnik rannesakskogo vremeni Sakar-Chaga 3. Sovetskaya Etnographia 5, 45–54.
3. GENERAL MIGRATION PROCESSES IN THE ARAL SEA AREA
85
—— 1989: Karakumskaya madonna. Izvestia Akademii nauk Turkmenskoi Sovetskoi socialisticheskoi respubliki, seria obshestvennih nauk 1, 14–20. —— 1991a: Formirovanie kul’turi sakov Uznogo Priaralia. Sovetskaya Etnographia 1, 72–89. —— 1991b: Naselenie Uzhnogo Priaralia v rannem zheleznom veke (po materialam mogilnikov) (Avtoreferat dissertatsii) (Moscow). —— 1996: Saki Iuzhnogo Priural’ya (arkheologia i antropologia mogilnikov) (Moscow). —— 1997: Sotsialnaya stratifikatsia stepnih populytsii v fiziko-antropologicheskom aspekte. In Maksimenko, V.E. (ed.), Donskie Drevnosti 5 (Azov), 154–63. —— 1998: Mnogoaktnie zakhoronenia v podzemnih sklepah Prisarikamishia epohi zheleza. In Mishkin, V.N. (ed.), Ufimskii Arkheologicheskii Vestnik 1 (Ufa), 8–24. —— 1999: Nekropoli drevnego Khorezma (Moscow). —— 2000a: ‘Scythian Triad’ and Scythian World. In Davis-Kimball, J., Murphy, E., Koryakova, L. and Yablonsky, L.T. (eds.), Kurgans, Ritual Sites, and Settlements. Bronze and Iron Age (BAR International Series 890) (Oxford), 3–9. —— 2000b: Antropologicheskie aspekti formirovania rannesarmatskoi kul’turi. In Mishkin, V.N. (ed.), Rannesarmatskaya kul’tura. Formirovanie. Razvitie. Khronologia (Samara), 29–40. Yablonsky L.T. and Bolelov, S.B. 1991: Mogilnik Yassi-gir 4 v Prisarikamishie: pogrebalnii obryad i antropologia. In Vainberg, B.I. (ed.), Novie otkritia v Priaralie 2 (Moscow), 4–72. Zhelezchikov B.F. 1983: Ekologia i nekotorie voprosi hozyistvennoi deyatelnosti sarmatov Uzhnogo Priuralia i Zavolzhia v IV v. do n.e. – IV. n.e. In Istoria i kul’tura sarmatov (Saratov), 48–59. Zhelezchikov, B.F., Sergatskov, I.V. and Skripkin, A.S. 1995: Drevnaya istoriya Nizhnego Povolzhya po pismennim i istoricheskim istochnikam (Volgograd).
II. REGIONAL STUDIES
CHAPTER FOUR COMMUNITY ORGANISATION AMONG COPPER AGE SEDENTARY HORSE PASTORALISTS OF KAZAKHSTAN SANDRA OLSEN, BRUCE BRADLEY, DAVID MAKI AND ALAN OUTRAM
Introduction Recent investigations of the Copper Age Botai culture (4th millennium BC), in northern Kazakhstan, document an unusual economy focused very heavily on horses. This subsistence strategy was so successful that the Botai were able to maintain stable villages ranging in size from 44 to 158 houses or more. Four such settlements have been identified to date: Botai, Roshchinskoe, Krasnyi Yar I and Vasilkovka IV. Of these, the largest village, Botai, has received the most attention, with over half of its houses having been excavated or tested (Zaibert 1993). Our current research examines the two smaller villages of Krasnyi Yar and Vasilkovka, just 14 km apart, near the city of Kokshetau (Fig. 1). Remote sensing, including magnetic gradient imaging and electrical resistivity, were employed to map the settlements. One complete pit-house and its immediate surroundings were excavated at both sites and test trenches were dug through two other houses. The goal was to understand the nature of sedentary horse pastoralism during the Copper Age of the Eurasian steppe. The Botai culture of northern Kazakhstan was part of a larger cultural entity characterised by pit-house settlements, a significant reliance on domestic animals, bell-shaped geometrically-impressed pottery, a biface technology, and scant examples of copper tools and ornaments. The broad territory of the ‘Geometric Impressed Pottery Province’ (GIPP) runs along the Urals and across western and northern Kazakhstan. This group of related ‘tribes’ appears to be primarily adapted to the forest-steppe. Sites are generally distributed along rivers or near lakes. Other Copper Age cultures subsumed under this large horizon include the Chuzh’yael’sk, Atym’insk, Volvonchinsk, Iorkutinsk, Pernashorsk, Amnya I and Sartyn’ya I cultures of the
90
S. OLSEN, B. BRADLEY, D. MAKI & A. OUTRAM
Fig. 1. Map showing site locations.
4. SEDENTARY HORSE PASTORALISTS OF KAZAKHSTAN
91
northern Urals and lower pre-Ob river area; the Ayatsk, Lipchinsk, Sosnovoostrovsk and Shapkul’sk cultures of the central Ural region (Shorin 1999, 25, tabl. 2); the Surtanda (Matyushin 1982) and Kysykul’sk (Shorin 1999) cultures of the southern Urals; and the Tersek in western Kazakhstan (Kalieva and Logvin 1997). Those along the Urals are purported to have received domestic sheep and cattle from the Near East as early as the 6th millennium BC (Matyushin 2003), but these species are not known in northern Kazakhstan until the Bronze Age, ca. 2000 BC. The Geometric Impressed Pottery Province seems to emerge out of a group of Neolithic hunting cultures that similarly have been assigned copious local names (Kislenko and Tatarintseva 1999; Shorin 1999; Zaibert 1992). Most are characterised by a pressure blade technology, sparse pottery of similar geometric designs, and ephemeral camps or tiny settlements concentrated along rivers or adjacent to lakes and ponds. In the region where the Botai settlements are found, the preceding Neolithic Atbasar sites are primarily strewn along the Ishim and Chaglinka rivers (Zaibert 1992, Kislenko and Tatarentseva 1999). In 2001, our team excavated the Atbasar campsite of Zhusan located on the bank of a prehistoric pond, just 600 m north-east of the settlement of Krasnyi Yar (Fig. 1). According to Kislenko and Tatarintseva (1999), the Atbasar people typically built aboveground shelters that preserved very poorly. Other than the shift from a blade to a biface technology, the most outstanding differences between the Neolithic and Copper Age are the size of their settlements and their economy. Neolithic sites range from artefact scatters in the plough zone, like Zhusan, to a few shallow houses. In contrast, the Botai villages consist of large groups of pit-houses aligned in rows along avenues and arranged around plazas. Faunal assemblages for the Neolithic are sparse because sites are small and shallow, but there appears to be a mixed group of hunted wild ungulates like aurochs, red deer, elk, saiga and horse, as well as smaller game like beaver and marmot. The Copper Age site of Botai is best known for its enormous collections of horse bones, estimated at 300,000 fragments. Only a tiny portion of the assemblage is derived from domestic dogs and wild game.
92
S. OLSEN, B. BRADLEY, D. MAKI & A. OUTRAM
The Transition from Hunting to Equine Pastoralism The faunal collection from the site of Botai has been studied in part by several scholars (Anthony 1996; Brown and Anthony 1998; Levine 1999a; 1999b; Akhinzhanov et al. 1992; Olsen 1996; 2000a; 2000b; 2003). Currently there remains some debate over the status of the horse at the time the Botai settlements were occupied (3600–3100 BC) (Levine 1999b), but considerable secondary evidence supports domestication. The faunal assemblages provide the richest indications for horse domestication among the Botai. A remarkable 99% of the faunal material from the eponymous site of Botai has been identified as horse. Dogs are next in prevalence after the horse, and wild game is limited to one or a few elements per species. The percentage of horses at Krasnyi Yar is roughly 90%, with aurochs being the second most common species. Mortality patterns reflect the keeping of large numbers of adult stallions at Botai, probably primarily for transporting humans and products, as well as hunting. Rituals include the burial of 14 horse carcasses with four humans in a mass grave at Botai (Rikushina and Zaibert 1984), as well as numerous pit deposits in which horse skulls are placed with other ritual items, sometimes on stone altars (Olsen 2000a). Ritual deposits link domestic dog sacrifices with certain elements of horses, particularly the skull and mandibles (Olsen 2000a; 2000b). This affiliation is common in Eurasia and likely reflects the use of dogs for both horse herding and in the hunt alongside the horse (Olsen 2000b). Skeletal element frequencies can provide valuable information about the distance between large game kill sites and home bases. The Schlepp Effect (Perkins and Daly 1968) is the bias in element frequencies at the home base caused by pedestrian hunters minimising their carrying load through efficient field processing of large game. People who have horses to ride and use as packhorses should be much less concerned with meat weight and distance. Olsen (2003) has reconstructed the relative value of the various skeletal parts for the horse, based on meat distribution, marrow content, ritual deposits and bone artefact-manufacture at Botai. Based on this study, the vertebrae would have been the elements least likely to have been transported long distances from the kill to the settlement by hunters on foot. Experimental butchery of a horse by Bradley and Olsen demonstrated that processing at the kill site could quickly remove most of the meat so that the heavy vertebral column could be abandoned there with little economic impact. Outram and Rowley-Conwy’s
4. SEDENTARY HORSE PASTORALISTS OF KAZAKHSTAN
93
(1998, tabl. 1) experiments show that the vertebral column and ribs, when stripped of meat, weigh a massive 32.5 kg on average and do not yield marrow. Although it is relatively easy to strip the larger muscles off the back, the intercostal muscles on the ribcage are better removed by the consumer after cooking (Olsen 2003). Further, although the ribs were used frequently for making bone tools, the vertebrae were not. Brown and Anthony (1998, 344) have identified wear on five lower second premolars of horses from Botai as having been caused by a soft bit of rawhide or hemp. As tempting as it is to use this evidence to document early horse domestication, there are problems with these particular designations. Olsen has identified extremely similar variations in occlusal surfaces on numerous Pleistocene horse teeth from North America, dating to many millennia before the arrival of humans. This suggests that the ‘soft bit wear’ on the Botai teeth, which differs significantly from that caused by a metal bit, could just as likely be natural attrition. Another line of evidence does indicate that the Botai people were making rawhide thongs, suitable for a wide range of horse trappings. The most common bone artefacts at Botai are thong-smoothers (implements for stretching and straightening rawhide strips) (Olsen 2001). Made on horse mandibles, these unusual tools exhibit polish and striations in their notches compatible with the reciprocal movement of narrow strips of hide across their edges. Although thongs could serve several purposes, they would have been extremely useful for controlling horses by means of even basic bridles, pole snares, lassoes, whips and hobbles. Rawhide bridles, like those employed by the American Indians soon after they acquired horses, need not have cheek pieces or bits. ‘War bridles’ and ‘racing bridles’, as they were called (Wilson 1978), consisted of a plain thong twisted around the mandible at the diastema, or bar. The Schlepp Effect has been observed in the stone tool assemblages, as well as among the horse bones. Application of the Schlepp Effect to materials other than bones is totally unexplored in Kazakhstan, but extremely important, nonetheless. At the Neolithic site of Zhusan, our excavations produced finished blades and blade tools made on local red and green medium-grained quartzite known to be available at a quarry less than 20 km from the camp. The quartzite quarry of Zhartas, documented by Bradley, revealed at least 30 shallow pits surrounded by flaking debris. Blade precores and waste products indicate that blades, probably of Neolithic age, were prepared at the quarry. Small bifaces may also indicate a later use of the source.
94
S. OLSEN, B. BRADLEY, D. MAKI & A. OUTRAM
The Copper Age Krasnyi Yar and Vasilkovka sites produced a wide array of raw material from unknown sources, as well as the local Zhartas quartzite. At Krasnyi Yar, according to Bradley’s analysis, it is evident that some imported stone was transported to the village in fairly large pieces and was then knapped into bifaces, scrapers and other flake tools in and around the houses. A cache pit full of chert flakes was found outside the house, and although they appeared to have come from a single flaking episode, refitting attempts were unsuccessful. The dissimilarity between the Neolithic and Copper Age lithic transport patterns is striking. Given the supportive evidence from the fauna, our interpretation for the cause of the variance through time is the introduction of domestic horses for transport of raw materials over long distances. During the Neolithic, preparing pressure blade precores at the quarry greatly reduced the carrying load. Pressure blade precores are lightweight and the blades have a high ratio of cutting edge to volume. In contrast, during the Copper Age, distance and carrying weight seem to have been much less significant. Despite much effort, archaeologists have only located sources for medium-grained quartzite in the nearby outcrops. Jasper, chert and especially a fine-grained quartzite, are plentiful in the Botai assemblages, but have not been found in the immediate region. Detailed surface geology maps do not indicate potential source areas either. Perhaps all of these materials were derived from more distant sources. Jasper is common in the Urals far to the west (Matyushin 2003). The large quantity of debitage found in clusters at Krasnyi Yar shows that much of the manufacture occurred at home. Generally, biface technology results in heavier tools than the Neolithic pressure blade technology, and it is therefore a less efficient use of the available raw material. It would also require the transport of greater volumes of stone into the settlements. Another aspect of the Botai stone tool assemblages has an important bearing on their subsistence practices. Although bifaces that resemble projectile points are fairly common, few show the distinctive impact breaks we would expect if they were being used to hunt large animals such as horses. Even when sediments are carefully screened, as at Krasnyi Yar and Vasilkovka, projectile points are relatively rare in Botai collections. Only a couple of bone points have been found near the surface at the site of Botai, and these are more likely to date from the Bronze Age. Bone harpoons, also used for hunting large game, are not particularly common either (at Botai, N = 13). If the economy was based strictly on hunting, then considerably more weaponry should have been present.
4. SEDENTARY HORSE PASTORALISTS OF KAZAKHSTAN
95
Instead, the most common stone tools are scrapers. Their ubiquitous distribution and heavy use indicate extensive and intensive hide working. Given all of this evidence, it becomes extremely difficult to argue against horse domestication, especially in light of the dramatic shift to larger settlements at the Botai sites compared to the preceding Neolithic. How else can one justify large permanent villages when the diet is so heavily biased toward horsemeat and there is no evidence for domestic crops? Pollen analysis by Robert Scaife yielded only native wild species typical of the forest-steppe and there are no artefacts clearly associated with the production of domestic grain. The lifestyle of sedentary horse herders was considerably different from that of their predecessors who lacked a dependable source of meat and milk, rapid transit for hunters and pack animals to carry heavy supplies. In order to understand better the nature of sedentary pastoralism among the Botai, our team embarked on a programme of remote sensing at Krasnyi Yar and Vasilkovka. In 2001, approximately 66% of the total site of Krasnyi Yar was surveyed using magnetic field gradient and electrical resistance instruments. The main part of the site was investigated, as well as a small outlier to the south-east of the site proper. In 2002, the entire site of Vasilkovka was mapped with remote sensing. The results indicate a surprising degree of village planning and standardisation in house shape, size and orientation. Based on this level of social organisation, it becomes exceedingly difficult to envision the Botai as strictly horse hunters with no domestic animals other than the dog, especially given that there were no domestic plants.
Recent Excavations In the 1980s, teams of archaeologists from the Petropavlovsk Pedagogical Institute excavated one whole house and a second partial one at Krasnyi Yar I and a house at Vasilkovka IV (Zaibert 1993). Their work demonstrated that these sites were roughly coeval with the larger site of Botai and derived from the same culture. In 2000, a joint Kazakh-American team from the Carnegie Museum of Natural History and the Presidential Cultural Centre of Kazakhstan excavated a total area of 240 m2 at Krasnyi Yar, including a third house and two large test trenches. The southern trench bisected an additional house. Nearly all of the
96
S. OLSEN, B. BRADLEY, D. MAKI & A. OUTRAM
artefacts from the excavations at Krasnyi Yar appear to be from the Copper Age. Krasnyi Yar is estimated to have a total area of 5 ha, compared to 9 ha for the site of Botai. In 2001, the Neolithic campsite of Zhusan, located near Krasnyi Yar, was excavated and the stone quarry of Zhartas was located, mapped and evaluated. In 2002, at Vasilkovka IV, one complete pit-house and its surrounding ground surface were excavated and a large trench was dug through an additional pit-house. Our excavations there also produced midden deposits from late 19th-early 20th century Kazakh herders’ camps filling the pit-house depressions, as well as a small scatter of Neolithic material. The site of Vasilkovka covers over 3 ha.
Remote Sensing at Krasnyi Yar and Vasilkovka One of the objectives of this investigation was to identify and map subsurface archaeological features using non-invasive prospecting methods. The geophysical data collected by Maki and our team during this investigation were targeted at houses, pits and other domestic structures in order to determine the degree of residential planning involved in village construction. The remote sensing investigation also looked for evidence of communal areas like plazas, since these relate to social organisation. Linear arrangements of post moulds, possibly representing stockades or corrals, were sought, because they may provide evidence for defence and early horse domestication. The general procedure that was followed began by dividing each survey area into a series of survey grids with dimensions of 30 × 30 m. Each grid was surveyed using the two different instruments (a gradiometer and a resistance meter), taking readings along regularly spaced transects. The value and position of each data point was automatically recorded in digital format, downloaded to a computer, and processed to produce subsurface images of the site. The remote sensing study of Krasnyi Yar consisted of magnetic field gradient and electrical resistance surveys over 24300 m2, or 66% of the total area of the site (Fig. 2a–b). No serious problems for geophysical data collection were encountered because of the terrain or vegetation. The soils at Krasnyi Yar consist of a relatively uniform clay-silt loam. The entire site of Vasilkovka, some 29400 m2, was covered by remote sensing (Fig. 3a–b). Gravel containing igneous or metamorphic stone, iron concretions,
Fig. 2. Remote sensing images of Krasnyi Yar: a.) magnetic field gradient image; b.) electrical resistance image.
4. SEDENTARY HORSE PASTORALISTS OF KAZAKHSTAN
97
Fig. 3. Remote sensing images of Vasilkovka IV: a.) magnetic field gradient image; b.) electrical resistance image.
98 S. OLSEN, B. BRADLEY, D. MAKI & A. OUTRAM
4. SEDENTARY HORSE PASTORALISTS OF KAZAKHSTAN
99
and modern agricultural disturbance were sources of signal clutter that sometimes complicated intrepretation of the archaeological features. Magnetic field gradient surveys measure the deviation from magnetic uniformity and report it as positive data when the deviation is in the direction of the earth’s magnetic field and as negative data when the deviation is in the direction opposite the earth’s magnetic field. In these surveys, the archaeological features that possess a larger magnetic susceptibility contrast appear with greater contrast in the survey image. The magnetic gradient data image from Krasnyi Yar provides very clear outlines of many of the houses, as well as revealing possible hearths within the houses, intramural and extramural pits, post moulds and modern features like roads and paths in great detail. The results from Vasilkovka were somewhat less impressive, but houses, hearths and some pits, as well as modern features, do show up relatively well. Electrical resistance surveys introduce an electrical current into the soil and measure the ease or difficulty with which this current flows through the soil. Variance in measured resistance values across a site can be interpreted as variance in the relative resistivity of materials in the vicinity of each reading. Resistance surveys respond to a combination of soil moisture, soluble ion concentration and physical soil type. Moist soils have lower resistivity than dry soils. Fine soils, such as clay, have lower resistivity than coarse soils, such as sands or gravels, and high salinity soils have low resistivity. Twin electrode surveys, like the ones conducted at Krasnyi Yar and Vasilkovka, respond to the soil resistivity in the immediate vicinity of the sampling probes. Archaeologically useful surveys result when the resistivity contrast between the archaeological features and the surrounding soil matrix is great enough to be detected. For the electrical resistance, lighter shades on the resistance map record low resistance data. These areas may be features such as pits and trenches with high moisture content, clays, soils with high salinity, and so forth. Darker shades record areas with relatively high resistance. These may be features such as stone architecture, sand or gravel filled pits, and so forth. The horizontal resolution of the electrical resistance survey in some cases approached 1 m. In terms of the Krasnyi Yar and Vasilkovka images specifically, the electrical resistance images show probable houses as high resistance anomalies, but usually do not reveal their shapes as distinctly as the magnetic gradient images do. For Krasnyi Yar, about 43 of the houses are indicated by electrical resistance, compared to 54 in the magnetic gradient image. For Vasilkovka, 38
100
S. OLSEN, B. BRADLEY, D. MAKI & A. OUTRAM
houses were detected using electrical resistance and 44 with magnetic gradient imaging. Using just electrical resistivity, most of the houses could only be identified as anomalies of an unknown nature. The coordination of the two remote sensing methods increased the degree of certainty in house identification, particularly for the more ambiguous magnetic anomalies. Electrical resistance shows geological features well, but does not reveal most of the smaller pits. In addition to locating houses and smaller features, remote sensing produces a view of whole settlements prior to digging and with a relatively short investment of time. The non-destructive aspect of remote sensing means that archaeologists can learn much about village organisation without large-scale excavations.
House Structures House Locations. On the ground surface, house depressions typically show up as circular areas of darker, denser vegetation compared to surrounding areas. In the early summer, at the large site of Botai, pit-house depressions are quite visible, both in aerial photographs and on the ground. Krasnyi Yar and Vasilkovka, however, have much more subtle indications of houses on the surface. The vegetation is often somewhat greener and denser, and, occasionally, there are slight depressions, but in many cases there is either no surface indication or it is ambiguous. Compared to the 54 possible pit-houses identified by remote sensing at Krasnyi Yar (Fig. 4), only 22 were mapped from visual inspection of the ground surface. At Vasilkovka, the maximum number of houses estimated from visual observation was 34, compared to the 44 revealed by remote sensing (Fig. 5). In addition, the confidence in house identification is tremendously improved with remote sensing compared to observation of ground surface fluctuations and vegetation patterns. Of course, only excavation can confirm whether the anomalies are actually houses. Our excavations at Vasilkovka have borne out the accuracy of remote sensing to a remarkable extent. Using the magnetic gradient image, stakes were positioned in the ground surface immediately above the actual corners of the house prior to digging, enabling the team to plan better an excavation strategy. Hearths and several pits were accurately located, as well.
4. SEDENTARY HORSE PASTORALISTS OF KAZAKHSTAN
101
Fig. 4. Krasnyi Yar I settlement plan.
House Construction. Materials and building methods for Botai culture houses have been important topics for investigation and experiments have been conducted in order to attempt to reconstruct these dwellings (Kislenko 1993). It was once thought that the superstructures of these semi-subterranean houses were built by log cribbing, but, according to Bradley, Botai stone tool assemblages are curiously depauperate of tree-felling tools. Axes are rare and those that are found are usually quite diminutive. The houses recently excavated at Krasnyi Yar and Vasilkovka are square in outline and like those from previous excavations at all the Botai sites, lack post moulds around the perimeter (Fig. 6). For this reason, it is unlikely that vertical supports were used in a wattle-and-daub type of structure. The absence of post moulds and the common use of log houses in the region today led archaeologists to believe that cribbing was a viable method of construction for Botai houses.
102
S. OLSEN, B. BRADLEY, D. MAKI & A. OUTRAM
Fig. 5. Vasilkovka IV settlement.
Fig. 6. Excavation plans for Krasnyi Yar I and Vasilkovka IV.
4. SEDENTARY HORSE PASTORALISTS OF KAZAKHSTAN
103
104
S. OLSEN, B. BRADLEY, D. MAKI & A. OUTRAM
In 2002, a breakthrough was finally made in understanding house construction. Fortunately, the two houses excavated at Vasilkovka had burned, consolidating and preserving evidence for wall and roof construction. The aboveground wall foundations of the burned houses appear to have been built of a local clay with quartz pebble inclusions that were either naturally occurring or were added to increase stability. Abundant charcoal in the layers representing the collapsed roofs demonstrated that the houses were covered with a lightweight wooden framework that was sealed with sediments. French and Kousoulakou’s (2003) micromorphological study of a profile through a house excavated at Botai in 1995–96 revealed a deposit of horse manure just above the fallen roof. This could be interpreted in two ways, either as dung cleared from a stable or corral and discarded in a collapsed structure, or as roofing material placed on the intact house for insulation. Today, Kazakh herders use dung to cover their outbuildings to keep the animals warm in winter and cool in summer. Recognition of house shapes by means of excavation can be difficult because of frost heaving and wall collapse. Tracking the outline of a house requires a great deal of skill since walls and roofs generally collapsed prior to the infilling of the structure with wind or water-transported sediments. For this reason, plans of houses from older excavations are often inaccurate (Zaibert 1993), making it difficult to determine the shapes, sizes and orientation of Botai houses. The magnetic gradient data images often delineate the outlines of houses fairly clearly, providing a valuable guide for excavations. Although it is not possible in every case to determine the shape of individual houses, most are rectangular. A few appear to be pentagonal. The sizes of Botai houses are somewhat variable, ranging from approximately 5 m to 8 m on a side, with most being about 6 m across. In the north-western part of the village of Krasnyi Yar, two houses appear to overlap in the magnetic image (Fig. 2a, located at 980 North/970 East). This is interesting because it suggests that one house was built partially on top of the foundation of an earlier house. This is the only clear case of demolition and rebuilding of houses, but open-ended arrangements of possible post moulds representing barriers or enclosures may also indicate remodelling of the village layout. House Orientation. House orientation may be closely linked to ethnicity in this region. The predominant orientation of houses in the Botai culture positions the
4. SEDENTARY HORSE PASTORALISTS OF KAZAKHSTAN
105
corners at the cardinal directions. Orientation is evident in the remote sensing images, as well as the excavations. Fig. 6 shows the floor plans for the houses excavated at Krasnyi Yar and Vasilkovka. This is a pattern that is already apparent in the Neolithic from the Urals east across northern Kazakhstan (Aleksashenko and Viktorova 1991; Chemyakin 1991; Kerner 1991; and Stefanof 1991). Neolithic houses in many of the region’s cultures are square, about 6 m on a side, and have the same orientation as those dominant at the Botai sites. Although it is clear that cardinal directions were taken into consideration in the building of houses for both the Neolithic and Copper Age Botai cultures, it is not known whether this recurrent practice was for pragmatic or sacred purposes. House orientation may have had some advantage in terms of thermal regulation. During the winter months, the sun remains quite low in the southern sky in this region. By having both a south-east- and a south-west-facing wall, the sun could shine on one surface of the house in the morning and on another in the afternoon. This orientation might maximise the absorption of solar radiation during the winter months by providing warmth for most of the daylight hours. Regardless of whether solar heating was a factor, it is clear that cardinal directions were important in house orientation to the Botai, as well as other GIPP cultures and their predecessors in the region.
The North-West/South-East Axis of Botai Settlements Aerial photos of Botai show multiple rows of houses lining up along a northwest/south-east axis. The house rows at Krasnyi Yar ramble a bit, but also exhibit this same basic north-west/south-east orientation (Fig. 4). At Vasilkovka, the houses are in lines strung along a natural ridge like beads in a necklace (Fig. 5). The ridge coincidentally is angled along a north-west/south-east axis. Perhaps the very location was chosen because the natural lay of the land coordinated so well with this tradition. It has already been demonstrated through the orientation of houses that cardinal directions were important to the Botai people. Although their village plans may have had ceremonial meaning, it is also simply more efficient to coordinate the orientation of the settlement with that of the houses. The alignment of dwellings in rows produces avenues to walk or ride along and facilitates the efficient grouping of structures close together without making passage between them difficult.
106
S. OLSEN, B. BRADLEY, D. MAKI & A. OUTRAM
Plazas or Communal Areas At both Krasnyi Yar and Vasilkovka, the remote sensing shows open areas with rows of houses grouped around them. The clearest examples are in the northern parts of both sites (Figs. 4 and 5). It should be pointed out that the plazas may not have been completely open, however, as there are some curved arrangements of probable post moulds visible in the northern plaza at Krasnyi Yar. If these are plazas, then they could have been centres of communal activities of various sorts. Excavation transects through these open spaces in the future could provide clues to their functions, as well as shedding light on the culture’s social organisation.
Enclosures In the magnetic images of Krasnyi Yar (Fig. 2a), alignments of small spots are hypothesised to be post moulds representing aboveground structures or barriers. The visibility of such features is particularly pertinent for locating possible horse corrals and defensive stockades. The appearance of such fine anomalies demonstrates the high quality of the magnetic images, especially given the more typical 1m resolution of the instrument. Many of the alignments at Krasnyi Yar are semicircular. The fact that most are not completely enclosed may mean that there was considerable demolition and reconstruction around the village. One particular structure is nearly perfectly round and measures 22 m in diameter. It may originally have been a complete circle that was obliterated by a later house, although it is also possible that it abutted an existing house and was only a partial circle when intact. Because we are interested in identifying possible corrals for domesticated horses, it is important to pursue subsurface investigations of some of these alignments. Soil chemical tests, as well as soil micromorphological and palynological analyses, should determine if there was a concentration of manure in some of these areas indicating that they were animal enclosures.
Hearths and Pits Remote sensing revealed a number of features measuring about a meter in diameter both inside and outside houses (Figs. 2a, 3a, 4, 5). Those in the centre of
4. SEDENTARY HORSE PASTORALISTS OF KAZAKHSTAN
107
pit-houses are most likely hearths. Excavations in two houses at Vasilkovka confirmed the locations of hearths seen in the magnetic image. High-temperature magnetic susceptibility enhancement in the clay floor explains their visibility. Excavations reveal that the hearths were shallow depressions where the fuel was piled rather than complex architectural features. Typically, Botai hearths are located near the centre of the dwelling (Zaibert 1993, 26) (Fig. 6). House 2 at Vasilkovka had two off-centre hearths. The first was to the north-east but it was sealed by floor levelling before the second was added to the south-east of centre. Extramural and intramural pits averaging about a meter in diameter are readily detectable in the magnetic images of Krasnyi Yar and to a lesser degree at Vasilkovka. Unfortunately, these features cannot be readily distinguished from hearths in the images, but their placement relative to houses can provide clues. Many such pits have been excavated at Botai and they are common at the other sites, as well. They generally contain large quantities of animal bones, a few stone scrapers and other tools, very few potsherds and some charcoal. At Krasnyi Yar, there were seven pits distributed evenly around the periphery of the house we excavated. It is likely that most of these pits first functioned as storage containers for fresh products and subsequently for refuse, but some also served religious purposes. Although most of the final pit deposits appear to be discarded waste materials, strong patterning of some animal remains clearly represents ritual behaviour. Most notably, four extramural pits at Krasnyi Yar contained the tops of horse skulls that had been worked. The skulls, which were reduced to only the frontals, parietals and nasals, may have been used as masks or plates of some kind. These horse skull ‘plates’ were closely associated with articulated cervical vertebrae from horse necks or, in one case an aurochs neck. The vertebral units may represent offerings of large pieces of meat, since the neck provides a great quantity of muscle tissue, and the cervicals lack cut-marks indicative of filleting (meat removal) prior to burial. This ritual closely resembles one performed at a Bronze Age religious site in the of Khanuy valley, Mongolia (Francis Allard, personal communication), where dozens of horse skulls were buried under stone cairns beside their articulated neck vertebrae. One dog skull was found in each of two pits to the west of House 3 at Krasnyi Yar. Observed in 15 cases at Botai (Olsen 2000b), the interment of whole dogs and dog skulls in pits west of houses is such a prevalent trait that the finding of dogs in these pits was even predicted prior to excavation at Krasnyi Yar. The burial of dogs or dog skulls on the west sides of houses reiterates the importance of cardinal directions among the Botai (Olsen 2000a). Numerous
108
S. OLSEN, B. BRADLEY, D. MAKI & A. OUTRAM
pairs of pits appear to the west of houses in the magnetic gradient image of Krasnyi Yar. In the future, it would be valuable to test some of these pits to increase our understanding of the dog sacrifice ritual. Dogs are often seen as spiritual guardians because in life they protect their owner, his family and the home from intruders (Olsen 2000b). Throughout Indo-European and Indo-Iranian religious texts, there is a recurrent theme of dogs as guardians (Green, 1997). A vivid reference to guard dogs in the home can be found in a Vedic hymn entitled ‘Varuna Provoked to Anger’ (Rig Veda 7. 86). The hymn describes the sage Vasistha being attacked by a guard dog when he entered the house of Varuna, the sky god who watched over people’s deeds. The dog barked, ran toward the man, and tried to bite him (O’Flaherty 1981, 212). Vasistha put the dog to sleep with the words of a sleeping spell (Rig Veda 7. 55). The burial of dogs just outside Botai houses may be a way of protecting the entrance to the home from dangerous spirits. At this point, it is unclear where the thresholds of the Botai houses are, however. At the Ukrainian Copper Age village of Dereivka a dog was also buried near a house (Telegin 1986). In the Indo-European Rig Veda and Indo-Iranian Avesta religious texts, dating to the 1st millennium BC, dogs were considered the guardians of the gate to the Afterlife, which lay to the west. Why was the west considered the direction for the afterlife in Eurasia in later times? A common theme across many cultures was that the Sun God traversed the sky each day from east to west, growing older as he went until he expired in the west at sunset. The spiritual connection between dogs and the west appears to date to at least the Copper Age in this region.
Post-Occupation Features and Geology Remote sensing reveals a wide variety of features, not all of which are strictly related to the occupation of the site. Geological features caused by changes in soil type or compaction are expressed more clearly in the electrical resistance image than in the magnetic gradient image. The magnetic gradient images show surface and subsurface features that postdate habitation, such as the path along the north-west, the field road along the east, and the recent trash pit to the west of Krasnyi Yar. At Vasilkovka, a
4. SEDENTARY HORSE PASTORALISTS OF KAZAKHSTAN
109
deep trench dug along the east side of the site during this last century to act as a fire-break shows up vividly in the remote sensing images.
Conclusions The two remote sensing techniques reveal different things, so both are therefore important. Comparisons between the two images provide greater confidence in identification of features when those anomalies are expressed in both images. This information has already aided significantly in the selection of areas to be excavated and will do so even more in the future. Ground truthing, that is, verification through excavation, in sample areas will allow us to interpolate for similar anomalies in the images that show up in other parts of the site. Results from remote sensing combined with excavation indicate that the Botai people built their houses in the same way and with the same orientation as preceding Neolithic people. The pottery of these two ages is also very similar, but the Neolithic people had very small settlements and probably had no domestic livestock. The large size of Botai sites, the deliberate planning of the settlements and remodelling indicate a long-term stability of the population. Whether or not some members of the population, or even most, periodically left the settlement to go on hunting forays, to search for stone resources, or to graze their herds, there must have been a commitment to the village that lasted for many decades, at least. At Krasnyi Yar remote sensing survey revealed at least 54 probable houses. If we assume that only half of the houses were occupied at one time, which is a modest estimate given that only two houses appear to overlap, then we can do a very rough population estimate. Using five as the average number of inhabitants in a given house, then probably at least 135 people lived there at a time. For Vasilkovka, this crude minimal estimate is about 110 people. For the site of Botai, the population could have risen to 400 or more. It is difficult to imagine a population density of hundreds of people on 3–9 ha over a long period of occupation if the economy was based almost completely on the hunting of wild horses. Additionally, these three Botai sites appear to be roughly coeval and are in proximity to each other. Vasilkovka and Krasnyi Yar are only about 14 km apart. Such population density and permanence would seriously
110
S. OLSEN, B. BRADLEY, D. MAKI & A. OUTRAM
stress local populations of wild game, particularly when food procurement focused so heavily on one species. Stone artefacts indicate that long-distance transportation was likely, Botai people engaged in quarrying raw materials, and processing hides was a dominant activity. There is also little evidence for the hunting of large game with stone tipped weapons, and although present, wood-working tools were not designed for forest clearing or log house production. Remote sensing has provided important insight into Botai settlement patterns. Village planning, with its avenues, plazas, rows of houses and enclosures, indicates a relatively high level of social organisation and coordination within the community. Cardinal directions were important in terms of ritual sacrifices, as well as both house and village layout. The results of this study support the conclusions derived from Botai faunal assemblages that these communities had a stable economy based primarily on sedentary horse pastoralism
Acknowledgements We sincerely wish to express our gratitude to the National Science Foundation, grants BCS-9816476 and BCS-0414441, the National Geographic Society, and the Carnegie Museum of Natural History for funding our research through the years. Our special appreciation goes out to Maral Ghabulina, Valerii Voloshin, Shagban Bektasov, our field team which includes many students from the Kokshetau University, and the staff of the Presidential Cultural Centre of Kazakhstan for all of their help and support. In memory of Dr Kimal Akishev, who facilitated our international collaboration so much, we are most grateful.
Bibliography Akhinzhanov, C., Makarova, L. and Nurumov, T. 1992: K istorii skotovodstva i okhoty v Kazakhstane (Almaty). Aleksashenko, N.A. and Viktorova, V.D. 1991: Neoliticheskie zhilishcha Yuao V. In Neoliticheskie Pamyatniki Urala (Sverdlovsk), 161–83. Anthony, D. 1996: Bridling Horse Power: The Domestication of the Horse. In Olsen, S.L. (ed.), Horses Through Time (Boulder, CO), 57–82. Brown, D. and Anthony, D. 1998: Bit Wear, Horseback Riding and the Botai Site in Kazakstan. Journal of Archaeological Science 25.4, 331–47. Chemyakin, Y.P. 1991: Neoliticheskaya stoyanka barsova gora II/17 v Surgutskom Priob’e. In Neoliticheskie Pamyatniki Urala (Sverdlovsk), 184–96. French, C. and Kousoulakou, M. 2003: Geomorphological and Micromorphological Investiga-
4. SEDENTARY HORSE PASTORALISTS OF KAZAKHSTAN
111
tions of Palaeosols, Valley Sediments and a Sunken Floored Dwelling at Botai, Kazakhstan. In Levine, M., Renfrew, C. and Boyle, K. (eds.), Prehistoric Steppe Adaptation and the Horse (Cambridge), 105–14. Green, M. 1997: Dictionary of Celtic Myth and Legend (London). Kalieva, S. and Logvin, V. 1997: Skotovody Turgaya v tret’em tysyacheletii do nashei ery (Kustanay, Kazakhstan). Kerner, V.F. 1991: Poselenie Isetskoe Pravoberezhnoe. In Neoliticheskie Pamyatniki Urala (Sverdlovsk), 46–67. Kislenko, A.M. 1993: Opyt rekonstruktsii eneoliticheskovo zhilishcha. In Problemy Rekonstruktsii khozyaistva i tekhnologii po dannym arkheologii (Petropavlovsk, Kazakhstan), 117–36. Kislenko, A. and Tatarentseva, N. 1999: The Eastern Ural Steppe at the End of the Stone Age. In Levine, M., Rassamakin, Y., Kislenko, A. and Tatarintseva, N. (eds.), Late Prehistoric Exploitation of the Eurasian Steppe (Cambridge), 183–216. Levine, M. 1999a: Botai and the Origins of Horse Domestication. Journal of Anthropological Archaeology 18.1, 29–78. —— 1999b: The Origins of Horse Husbandry on the Eurasian Steppe. In Levine, M., Rassamakin, Y., Kislenko, A. and Tatarintseva, N. (eds.), Late Prehistoric Exploitation of the Eurasian Steppe (Cambridge), 5–58. Matyushin, G. 1982: Eneolit Yuzhnogo Urala (Moscow). —— 2003: Exploitation of the Steppes of Central Eurasia in the Mesolithic-Eneolithic. In Levine, M., Renfrew, C.and Boyle, K. (eds.), Prehistoric Steppe Adaptation and the Horse (Cambridge). O’Flaherty, W. 1981: The Rig Veda (London). Olsen, S.L. 1996: Prehistoric Adaptation to the Kazak Steppes. In Afanas’ev, G., Cleuziou, S., Lukacs. J. and Tosi, M. (eds.), The Colloquia of the XIII International Congress of Prehistoric and Protohistoric Sciences, Forlì, 8–14 September 1996 vol. 16: The Prehistory of Asia and Oceania (Forlì), 49–60. —— 2000a: Expressions of Ritual Behavior at Botai, Kazakhstan. In Jones-Bley, K., Huld, M. and Volpe, D. (eds.), Proceedings of the Eleventh Annual UCLA Indo-European Conference (Journal of Indo-European Studies Monograph Series 35) (Washington, DC), 183–207. —— 2000b: The Sacred and Secular Roles of Dogs at Botai, North Kazakhstan. In Crockford, S. (ed.), Dogs Through Time: An Archaeological Perspective (BAR International Series 889) (Oxford), 71–92. —— 2001: The Importance of Thong-smoothers at Botai, Kazakhstan. In Choyke, A. and Bartosiewicz, L. (eds.), Crafting Bone: Skeletal Technologies through Time and Space (BAR International Series 937) (Oxford), 197–206. —— 2003: The Exploitation of Horses at Botai, Kazakhstan. In Levine, M., Renfrew, C. and Boyle, K. (eds.), Prehistoric Steppe Adaptation and the Horse (Cambridge), 83–104. Outram, A. and Rowley-Conwy, P. 1998: Meat and Marrow Utility Indices for Horse (Equus). Journal of Archaeological Science 25, 839–49. Perkins, D. and Daly, P. 1968: The Potential of Faunal Analysis: An Investigation of the Faunal Remains from Suberde, Turkey. Scientific American 219, 96–106. Rikushina, G.B. and Zaibert, V.F. 1984: Predvaritel’noe soobshchenie o skeletnykh ostatkakh lyudei s eneoliticheskovo poceleniya Botai. In Bronzovyi vek Uralo-Irtyshskovo mezhdurech’ya (Chelyabinsk), 121–34. Shorin, A.F. 1999: Eneolit Urala i Copredel’nykh Territoriii: Problemy Kul’turogeneza (Yekaterinburg). Stefanof, V.I. 1991: Neoliticheskoe pocelenie Duvanskoe V. In Neoliticheskie Pamyatniki Urala (Sverdlovsk), 144–60. Telegin, D. 1986: Dereivka: A Settlement and Cemetery of Copper Age Horse Keepers on the Middle Dnieper (BAR International Series 287) (Oxford). Wilson, G.L. 1978: The Horse and Dog in Hidatsa Culture (Reprints in Anthropology 10) (Lincoln, NA) (reprinted from Anthropological Papers of the American Museum of Natural History 15 [2], 1924). Zaibert, V.F. 1992: Atbasarskaya kul’tura (Chelyabinsk). —— 1993: Eneolit Uralo-Irtyshskovo (Almaty).
CHAPTER FIVE PERSPECTIVES ON CHALCOLITHIC TURKMENISTAN: THE GLOBAL AND THE LOCAL IN THE KOPET DAG LAUREN ZYCH
The fall of the Soviet Union, and the era of international collaboration that followed, has opened many doors for Western scholars interested in exploring the archaeology of North and Central Asia. In the last few decades, Central Asia generally and Turkmenistan in particular have significantly altered current understandings of complex civilisations and Old World archaeology. While increasing numbers of Western investigators in this region have published on the general character and sequence of the ancient cultures of Central Asia, there is still much to be explored. To date, research interests in Turkmenistan have been predominantly global in orientation; Western scholars, in particular, have produced an entire discourse on the impact of Central Asia on the greater Near East and South Asia. This global emphasis is evident in both the periods selected as objects of study and in the theoretical frameworks constructed in their interpretation. In this paper I will describe the earliest chronological period that has drawn the attention of international-minded scholars, the Namazga III period. It will serve as an example of the exceedingly global character of Western research and ultimately demonstrate a critical need for research on a local scale. At the end of the 4th and beginning of the 3rd millennium BC, the inhabitants of modern Turkmenistan were living in small agricultural towns. These towns were located in two distinct ecological settings in the Late Chalcolithic. The bulk of the population inhabited the fertile strip of land located between the peaks of the Kopet Dag mountains to the west and the sand dunes of the Kara Kum in the east. A smaller portion resided in the desert itself, in an oasis environment formed by the delta of the Tedjen river. Despite differences in setting, these modest settlements represent a single archaeological complex, barring some regional variations, which is described in the following section.
5. PERSPECTIVES ON CHALCOLITHIC TURKMENISTAN
113
The Namazga III Archaeological Complex In an effort to avoid the implications inherent in the term ‘culture’ or ‘archaeological culture’, I have decided to follow the example of F. Hiebert (1994) and adopt the ‘archaeological complex’ for use in the following discussions. An archaeological complex represents a series of stylistically associated artefacts or other features that appear repeatedly at multiple sites. As such, it may represent an archaeological ‘culture’, a trade route, or a recurring combination of status items. In this case, it simply refers to the architecture, burial practices, ceramics and small finds that characterise Central Asian sites at the beginning of the 3rd millennium BC (Fig. 1). The archaeological data discussed below data were compiled from a variety of sources including site reports, regional overviews and theoretical syntheses (Biscione 1973; Gupta 1979; Khlopin 1974; Kircho 1988; 1994; Kohl 1981; 1984; 1992; Lamberg-Karlovsky 1973; Lisitsina 1969; Masson and Masson 1959; Masson 1961a; 1961b; 1968, 1981; 1988; 1992; 1993; Masson et al. 1974; Masson and Korobkova 1989; Masson and Sarianidi 1972; Muller-Karpe 1984). The architecture characteristic of the Late Chalcolithic or Namazga III period in Turkmenistan consists of multi-roomed mud-brick buildings separated by narrow streets or alleys. Within each residential unit is a central court surrounded by five to seven rooms, including storage facilities and occasionally a sanctuary, identified by thick walls and a disc-hearth or ‘fire-altar’. Ceramic kilns, wasters and flint drills provide evidence for household manufacturing. Simple pit burials in abandoned portions of the settlement were standard practice in the foothills of the Kopet Dag, but in the desert oasis a distinct burial tradition was in place. Here the dead were interred off-site, in a cemetery separated from the village by a partial wall, and the remains were placed in subterranean mud brick constructions or tholoi. Instead of the individual burials seen elsewhere, these tholoi contained multiple successive burials. The ceramic assemblage of this period is characterised by regional variation in manufacturing technique and decorative style. The oasis sites, referred to collectively as the ‘Geoksyur Oasis’, represent an ‘eastern zone’ which produced high quality, hand-made painted ware with a sand temper. Vessels are decorated with complex geometric patterns executed in polychrome combinations of red and black on a buff surface. The most striking motifs are constructed around ‘T’ shapes and crosses, designs which are found on more than half of the total
114
LAUREN ZYCH
Fig. 1. Map of major Namazga III sites discussed in the text.
5. PERSPECTIVES ON CHALCOLITHIC TURKMENISTAN
115
assemblage. The deep bowl is the most common form, usually decorated in a wide band along the upper portion of the vessel. In the ‘central zone’, represented by Kara-depe (Masson 1961b), the pottery contains a chaff temper instead of sand and is painted in monochrome instead of polychrome. In addition, the decorative motifs known from Kara-depe include a variety of zoomorphic representations notably absent from the oasis sites. They have been compared, historically (Kohl 1984, 101; Gupta 1979, 93), to motifs on painted ware from Tepe Hissar in north-eastern Iran, as goats and leopards are the most common subjects. In the ‘western zone’, Ak-depe represents a final regional variant characterised by the presence of Burnished Grey ware (Cleuziou 1986). This pottery was fired in a reducing atmosphere and decorated with incision and appliqué snakes, revealing important differences in firing technique as well as decoration and surface treatment. Unlike the pottery, the abundant anthropomorphic figurines are similar in all three areas. The figurines are predominantly female and are often divided into two basic types, sharing a seated aspect but representing different degrees of stylisation. One type is highly stylised, the torso and limbs fused to create a single smooth line. Facial features consist solely of pinched noses and large S-shaped curls. The other category is more realistic in appearance, with distinct arms and shoulders, moulded breasts and almond-shaped eyes. At times, necklaces and clothing are represented with brown paint or appliqué knobs. A brief list of small finds completes the archaeological complex. Among metal objects, a ‘shovel-shaped’ copper pin with a flat triangular tip and thickened butt is relatively common. In stone, such as alabaster or steatite, pink and white vessels were skilfully carved and decorated with fluting or incised ornamentation. Other stone artefacts include plaques or amulets carved into the form of a stepped cross.
Colonies, Merchants and Civilising Influences: Central Asia in the Global Context About 30 years ago, Central Asia made its debut in Western archaeological literature with the publication of a major work in the English language (Masson and Sarianidi 1972). This occurred as a consequence of the discovery of Central Asian features outside Turkmenistan. Select elements of the archaeological complex described above have been identified in the Zerafshan river valley of
116
LAUREN ZYCH
Tadjikistan, at the site of Sarazm; in the Helmand valley of Iranian Sistan, at Shahr-i Sokhta; and throughout Baluchistan in the Quetta and Pishin valleys. Sarazm. At the site of Sarazm, more than 600 km east of the oasis, several characteristic elements of the Namazga III archaeological complex have been identified, including aspects of architectural design, ceramic assemblages and small finds. It is important to note that Sarazm is quite a distance, not only from settlements in Turkmenistan, but also from other known areas of settlement. Given the current state of research it appears to be an extremely large, isolated urban development. Architecture in Late Chalcolithic levels is very similar to contemporary architecture in Turkmenistan. Multi-room domestic units separated by small streets or alleys contained courtyards, rooms for storage, and ritual rooms. The latter contained ‘fire altars’ identical in construction to those described above (Isakov 1981, 274–6). The ceramics at Sarazm have been extensively studied by A. Isakov and B. Lyonnet (1988; Lyonnet 1996). Although they have been unable to identify a local ceramic type, analysis indicates that ceramics with Namazga III parallels (including some with the classic T-shaped motif) make up approximately 40% of the total assemblage (Isakov and Lyonnet 1988, 31–5, tabl. 1). Further analysis indicated that some of the Geoksyur-style pottery was imported (Besenval and Isakov 1989, n. 68). Numerous small finds exhibit Central Asian parallels, among them metal objects, numerous agricultural implements, grinding stones, mortars and pestles, vessels, spindle whorls and other objects made of stone (Isakov 1981, 279). However, the most relevant items for comparative purposes are terracotta figurines. Three seated female figurines and a fourth with an elongated neck and bird-like head all find parallels within the Geoksyur Oasis (Isakov 1994, 4; 1988, 120). Due to the clear Central Asian influences visible in the late 4th and the early 3rd millennium BC levels at the site, archaeologists have identified Sarazm as a Central Asian colony whose origins lay in the late Namazga II/Yalangach Period (Lyonnet 1996, 58). Several scholars also position Sarazm as an active participant in the global trade networks of the 3rd millennium BC. Besenval and Isakov (1989), for example, interpret Sarazm as a manufacturing centre, dedicated to metallurgy and lapis lazuli, and a locus of commercial activity. Three lines of evidence appear to support this picture of Sarazm as a centre of manufacturing and commerce: the availability of raw materials, evidence for on-site production, and the abundance of finished products on the site. Gold, silver,
5. PERSPECTIVES ON CHALCOLITHIC TURKMENISTAN
117
copper, mercury, lead, lapis lazuli, turquoise, carnelian and even tin have been cited as potential commodities (Besenval 1987, 445; Besenval and Isakov 1989, 18). Evidence for production can be seen in crucibles and moulds, slag, and debitage from stonework (Besenval 1987, 454–5; Isakov 1994, 8). Finished metal objects include 39 knives, 23 daggers and countless other objects, while more than 200 lapis beads were recovered from a single burial (Isakov 1994, 6, 8). However there are problems with interpreting Sarazm as a commercial centre. The evidence for metallurgy suggests that production occurred at the household level, in domestic contexts rather than specialised craft areas (Besenval 1987, 445). Furthermore, metallographic analysis indicates that different tools were cast from the same smelt, a practice that does not support existing arguments for large-scale production (Isakov et al. 1987, 101). In addition, the quantity of debitage from lapis lazuli is insufficient to support a claim for significant stone-working industries. Moreover, the lapis beads were recovered from a burial which may predate the establishment of the site (Hiebert 2002), which is more than 500 km away from the nearest lapis source (Besenval and Isakov 1989, n. 71). Although it is tempting to view Sarazm as a key player in a ‘permanent circuit’ of materials between Mesopotamia and the East (Amiet and Tosi 1978, 24), the evidence in support of this identification is not incontestable and alternative possibilities warrant exploration. Shahr-i Sokhta. The excavation of Shahr-i Sokhta in Iran also brought Central Asia to the attention of a more general audience. The discovery of a ProtoElamite tablet on the last day of excavation in 1975 guaranteed that the site would capture the attention of Near Eastern archaeologists (Tosi 1976, 168). Even more provocative, however, is the presence of Geoksyur ceramics in the very same levels as the tablet and other administrative items, presenting a potential link between Proto-Elamite Iran and Turkmenistan (Amiet and Tosi 1978, 10). No architectural parallels have been identified due to restricted Period I exposures, but other connections with Namazga III sites exist in the form of seated female figurines (Tosi 1983, 148–9) and a single collective tholos burial (Piperno and Salvatori 1983, 175; Piperno 1986, 261). Tosi has suggested that Shahr-i Sokhta is a colony from Turkmenistan (LambergKarlovsky and Tosi 1973; Tosi 1973, 439–43; 1974, 14; Lamberg-Karlovsky 1993, 63). This argument was originally based on quantitative ceramic studies which indicated that 40% of the Shahr-i Sokhta remains were derived from the Geoksyur Oasis (Lamberg-Karlovsky and Tosi 1973, 25). However, this number
118
LAUREN ZYCH
was later revised by Biscione (1984), whose study indicated that Baluchistan ceramics are considerably more common than those from Turkmenistan. With only 11% of the total assemblage clearly related to Namazga III ceramics, Shahr-i Sokhta is no longer convincing as a Central Asian colony. But the presence of a significant quantity of Geoksyur ceramics at Shahr-i Sokhta is extremely important, particularly in light of its proto-Elamite connections. Lamberg-Karlovksy and Tosi (1973) have proposed the existence of two large interaction spheres at the beginning of the 3rd Millennium BC: the Jemdet Nasr sphere to the west and the south Turkmenian sphere to the east. Shahr-i Sokhta, then, is unique as the only site participating in the activities of both. While the interaction sphere (Caldwell and Hall 1964) may be an appropriate descriptive term for the archaeological data, it cannot be attributed an explanatory function for which it was never intended. We understand neither the motivation nor the cultural mechanisms underlying the formation and maintenance of these two interaction spheres. Baluchistan. The third and final region with connections to Central Asia is Baluchistan, where ‘Quetta Ware’ is distributed throughout the Quetta and Pishin valleys (Fairservis 1956). This wheel-made pottery is decorated with black paint in geometric, zoomorphic and floral designs. Among the geometric motifs are crosses, stepped lozenges, chequered panels and the T-shape – some of the more common designs found on Geoksyur polychrome. Seated female figurines also link Baluchistan to Turkmenistan (Fairservis 1956, fig. 17a). Piggott (1950) originally believed that the Central Asian ceramics were derived from Quetta ware, but recent publications have suggested the reverse. Lamberg-Karlovsky and Tosi (1973), as well as Biscione (1973; 1984) and Sarianidi (1983), have argued that the origins of Quetta ware lie in the Geoksyur polychrome, noting that the abrupt appearance of Quetta ware coincides with additional cultural changes, including the appearance of stamp seals, clay figurines and alabaster vessels, the adoption of more elaborate architecture and collective burials, and the emergence of a more complex social and economic organisation. This perspective assumes that a ‘civilising influence’ from Central Asia spread south through the Helmand river valley to Baluchistan. However, J.F. Jarrige (1988) has emphasised several problems with the ceramic comparison. First and foremost is the difference in production technology. Quetta ware is thrown on a wheel, while Geoksyur ceramics are hand made (Jarrige 1988, 96). In addition, close examination of the painted motifs
5. PERSPECTIVES ON CHALCOLITHIC TURKMENISTAN
119
indicates that Quetta ware did not evolve directly from the Geoksyur polychrome, but underwent significant modification (Masson and Sarianidi 1972, 96). In general, the Quetta material is of a much higher quality than that from Turkmenistan. Jarrige has suggested that textiles, wood and baskets may have been responsible for the diffusion of the shared painted motifs (1988, 99).
Conclusions To summarise, sites in Tadjikistan, Iran and Baluchistan share certain aspects of Central Asian material culture. Excavations at Sarazm, Shahr-i Sokhta and sites in the Quetta-Pishin valley have encouraged scholars to explore the nature and extent of interaction as manifested in the Late Chalcolithic period. Although this is a rewarding and worthwhile task, the result is that Turkmenistan enters into archaeological discourse only as it relates to interaction spheres, migration, colonisation and long-distance trade. The subtle implication is that the archaeology of Central Asia is of peripheral interest, being explored only as it intersects the established disciplines of Near Eastern and South Asian archaeology. It is merely one stop along the Silk Road, rather than a final destination. With international collaboration on the rise, new methodological and theoretical approaches are being negotiated between distinct archaeological traditions. At this point, it is important that we remain conscious of research priorities in the coming years. Most of the Western literature on Central Asia is overwhelmed by ‘global’ scale analyses and it seems that we must work to strike a balance between global and local research, adding questions about social, economic and political processes to those being posed of long-distance trade and interregional contact. In this manner we can establish the archaeology of Turkmenistan as interesting and informative in its own right, and we can achieve a more thorough and developed understanding of the wealth and complexity of Central Asian prehistory.
Bibliography Amiet, P. and Tosi, M. 1978: Phase 10 at Shahr-i Sokhta: Excavations in Square D and the Late 4th Millennium Assemblage of Sistan. East and West 28, 9–31. Besenval, R. 1987: Découvertes Récentes à Sarazm (R.S.S. du Tadjikistan): Attestation des Relations au IIIe Milléniare entre l’Asie Centrale, l’Iran du Nord-est et le Baluchistan. Comptes Rendues des Séances, 442–56.
120
LAUREN ZYCH
Besenval, R. and Isakov, A. 1989: Sarazm et les Débuts du Peuplement Agricole dans la Région de Samarkand. Arts Asiatiques 44, 5–20. Biscione, R. 1973: Dynamics of Early South Asian Urbanization: First Period of Shahr-i Sokhta and its Connections with Southern Turkmenia. In Hammond, N. (ed.), South Asian Archaeology (London), 105–18. ——. 1984: Baluchistan Presence on the Ceramic Assemblage of Period I at Shahr-i Sokhta. In Allchin, B. (ed.), South Asian Archaeology, 1981 (London), 69–80. Caldwell, J.R. 1964: Interaction Spheres in Prehistory. In Caldwell, J.R. and Hall, R.L. (eds.), Hopewellian Studies (Springfield, IL). Cleuziou, S. 1986: Tureng Tepe and Burnished Grey Ware: A Question of Frontier? Oriens Antiquus 25, 221–56. Fairservis, W.A. 1956: Excavations in the Quetta Valley, West Pakistan (New York). Gupta, S.P. 1979: Archaeology of Soviet Central Asia and the Iranian Borderlands vol. II (Delhi). Hiebert, F.T. 1994: Origins of the Bronze Age Oasis Civilization of Central Asia (Cambridge, Mass.). ——. 2002: Bronze Age Interaction between the Eurasian Steppe and Central Asia. In Boyle, K., Renfrew, C. and Levine, M. (eds.), Ancient Interactions: East and West in Eurasia (Cambridge), 237–48. Isakov, A. 1981: Excavations of the Bronze Age Settlement of Sarazm. In Kohl, P.L. (ed.), Bronze Age Civilizations of Central Asia: Recent Soviet Discoveries (Armonk, NY), 273–86. ——. 1994: Sarazm: An Agricultural Center of the Ancient Sogdiana. In Litvinskii, B.A. and Bromberg, C.A. (eds.), The Archaeology and Art of Central Asia: Studies from the Former Soviet Union (Detroit), 1–12. Isakov, A., Kohl, P.L., Lamberg-Karlovsky, C.C. and Maddin, R. 1987: Metallurgical Analysis from Sarazm, Tadjikistan SSR. Archaeometry 29.1, 90–102. Isakov, A. and Lyonnet, B. 1988: Céramiques de Sarazm (Tadjikistan, URSS): problèmes d’échanges et peuplement à la fin du chalcolithique et au début de l’age du bronze. Paléorient 14.1, 31–47. Jarrige, J.F. 1988: Les Styles de Geoksyur et de Quetta at la Question des Rapports entre les Régions au Nord et au Sud de l’Hindu Kush à la Fin du 4e et au debut du 3e millénaires. In L’Asie Centrale et ses Rapports avec les Civilsations Orientales des Origines à l’Âge du Fer (Paris), 95–100. Khlopin, I.N. 1974: Ancient Farmers of the Tedzen Delta. East and West 24, 51–87. Kircho, L.B. 1988: The Beginning of the Early Bronze Age in South Turkmenia on the Basis of Altyn-depe Material. East and West 38, 33–64. ——. 1994: New Studies of the Late Chalcolithic at Altyn-depe, Turkmenistan. In Kozintsev, A.G., Masson, V.M., Solovyova, N.F. and Zuyez, V.Y. (eds.), New Archaeological Studies in Asiatic Russia and Central Asia (St Petersburg), 39–43. Kohl, P.L. (ed.) 1981: Bronze Age Civilizations of Central Asia: Recent Soviet Discoveries. (Armonk, NY). ——. 1984: Central Asia: Paleolithic Beginnings to the Iron Age (Paris). ——. 1992: Central Asia (Western Turkestan): Neolithic to Early Iron Age. In Ehrich, R.W. (ed.), Chronologies in Old World Archaeology 3rd ed. (Chicago), vol. 1, 179–95. Lamberg-Karlovsky, C.C. 1973: Prehistoric Central Asia: A Review. Antiquity 47, 43–7. ——. 1993: The Biography of an Object: The Intercultural Style Vessels of the Third Millennium BC. In Lubar, S.D. and Kingerey, W.D. (eds.), History of Things: Essays on Material Culture (Washington, DC), 270–92. Lamberg-Karlovsky, C.C. and Tosi, M. 1973: Shahr-i Sokhta and Tepe Yahya: Tracks on the Earliest History of the Iranian Plateau. East and West 23.1–2, 21–58. Lisitsina, G.N. 1969: The Earliest Irrigation in Turkmenia. Antiquity 43, 279–88. Lyonnet, B. (ed.) 1996: Sarazm (Tadjikistan) Céramiques (Chalcolithique et Bronze Ancien) (Paris). Masson, M.E. and Masson, V.M. 1959: Archaeological Cultures of Central Asia of the Aeneolithic and the Bronze Age. Cahiers d’Histoire Mondiale 5, 15–40.
5. PERSPECTIVES ON CHALCOLITHIC TURKMENISTAN
121
Masson, V.M. 1961a: The First Farmers in Turkmenia. Antiquity 35, 203–13. ——. 1961b: Kara-depe u artika. Trudy X, 319–463. ——. 1968: The Urban Revolution in South Turkmenia. Antiquity 42, 78–187. ——. 1981: Altyn-depe During the Aeneolithic Period. In Kohl, P.L. (ed.), Bronze Age Civilizations of Central Asia: Recent Soviet Discoveries (Armonk, NY), 63–95. ——. 1988: Altyn Depe. (Philadelphia). ——. 1992: Ilgynly-depe A New Center of Early Farming Culture in South Turkmenia. Jarrige, C. (ed.), South Asian Archaeology, 1989 (Madison, WI), 195–200. ——. 1993: À l’origine des villes en Turkmenie méridionale. Découvertes des civilizations d’Asie Centrale. Les Dossiers d’Archéologie 185, 22–7. Masson, V.M., Berezkin, Y.E. and Solovyova, N.F. 1994: Excavations of Houses and Sanctuaries at Ilgynly-depe Chalcolithic Site, Turkmenistan. In Kozintsev, A.G., Masson, V.M., Solovyova, N.F. and Zuyez, V.Y. (eds.), New Archaeological Studies in Asiatic Russia and Central Asia (St Petersburg), 18–26. Masson, V.M. and Korobkova, G.F. 1989: Eneolithic Stone Sculpture in South Turkmenia. Antiquity 63, 62–70. Masson, V.M. and Sarianidi, V.I. 1972: Central Asia. Turkmenia before the Achaemenids. (London). Müller-Karpe, H. 1984: Neolithisch-kupferzeitliche Siedlungen in der Geoksjur-Oase, Süd-Turkmenistan (Munich). Piggot, S. 1950: Prehistoric India: to 1000 BC (Harmondsworth, Mx). Piperno, M. 1986: Aspects of Multiple Ethnicity Across the Shahr-i Sokhta Graveyard. Oriens Antiquus 25, 257–70. Piperno, M and Salvatori, S. 1983: Recent Results and New Perspectives from the Research of the Graveyard at Shahr-i Sokhta, Sistan, Iran. Annali. Istituto Orientale di Napoli 43.2, 173–91. Sarianadi, V.I. 1971: Southern Turkmenia and Northern Iran. Ties and Differences in Very Ancient Times. East and West 21.3–4, 291–310. ——. 1983: The Pottery of Shahr-i Sokhta and Its South Turkmenian Connections. In Tosi, M. (ed.), Prehistoric Sistan I (Rome), 183–90. Tosi, M. 1967: Excavations at Shahr-i Sokhta, a Chalcolithic Settlement in Iranian Sistan: Preliminary Report on First Campaign. East and West 18.1–2, 9–66. ——. 1969: Excavations at Shahr-i Sokhta: Preliminary Report on the Second Campaign, SeptemberDecember 1968. East and West 19.3–4, 283–386. ——. 1973: Early Urban Evolution and Settlement Patterns in the Indo-Iranian Borderland. In Renfrew, C. (ed.), Explanation of Culture Change: Models in Prehistory (London), 419–46. ——. 1974: The Lapis Lazuli Trade Across the Iranian Plateau in the 3rd Millennium BC. In Gururajamanjarika. Studi in Onore de Giuseppe Tucci I. (Naples), 3–22. ——. 1976: Shahr-i Sokhta. Iran 14, 167–8. ——. 1982: The Development of Urban Societies in Turan and the Mesopotamian Trade with the East: the Evidence from Shahr-i Sokhta. In Nissen, H. and Renger, J. (eds.), Mesopotamien und Seine Nachbarn (Berlin), 57–78. ——. 1983: Development, Continuity, and Cultural Change in the Stratigraphical Sequence of Shahr-i Sokhta. In Tosi, M. (ed.), Prehistoric Sistan I (Rome), 127–82.
CHAPTER SIX THE DZHUNGAR MOUNTAINS ARCHAEOLOGY PROJECT: RECONSTRUCTING BRONZE AGE LIFE IN THE MOUNTAINS OF EASTERN KAZAKHSTAN MICHAEL D. FRACHETTI
Introduction This conference is devoted to current and new research across Eurasia. Though in light of the past 60 or more years in which Soviet scholars had been studying the archaeology of Central and Inner Asia, the ‘newness’ of the research being presented today should be placed in perspective. From the 1930s, Soviet archaeologists conducted large-scale survey expeditions, such as those around Khorezm (Tolstov 1952), as well as the Altai and Semirechye surveys. For their time, these large-scale surveys were comparatively unrivalled in the West, especially in terms of their overwhelming sponsorship and ambition. Many of these Soviet expeditions culminated in the early 1960s, and Soviet archaeology went into a period of synthesis and material studies. Throughout the 1970s and 80s, Soviet archaeologists and ethnologists made strides toward developing a distinct theoretical platform, which can be compared in some way to the ‘culture concept’ of Western anthropology. Essentially, they constructed the archaeological culture-history of ‘ethnos’ according to Soviet styled macro-history (Bromlei 1983; Bromlei and Kozlov 1989), itself theoretically underpinned by Marxist-Leninist historical materialism. Unfortunately, the reality of the cold war meant that both Western and Soviet scholars were largely isolated from one another. Eurasian archaeology behind the Iron Curtain was consequently lost to the general consciousness of most Western scientists. Yet by 1989 and 1990, the windfall of perestroika and the dismantling of the USSR opened the door to ‘new’ research frontiers for both post-Soviet and Western scholars in places like Uzbekistan, Turkmenistan and Kazakhstan, and across the former USSR. Today, Eurasian archaeological research is making headway in a number of
6. THE DZHUNGAR MOUNTAINS ARCHAEOLOGY PROJECT
123
exciting areas of inquiry – such as the development of Central Asian urbanism (Hiebert 1994), the social organisation of Chalcolithic tribal societies (Kislenko and Tatarintseva 1999), origins of agriculture and animal domestication (Anthony and Brown 2000; Harris et al. 1996), and the trajectory of prototypical languages (Mallory 1989). New approaches are recasting these archaeological problems, while our historical picture of Eurasia as a region is rapidly changing as well. In fact, today’s collaborative research programmes between Russia, the Newly Independent States, Europe and the USA, offer new opportunities for archaeological theory as well as the exposition of new data. Another exciting aspect of the collaborative research in the former Soviet Union is the exchange of recent developments in modelling techniques and methods such as GIS, remote sensing and ecological landscape modelling. So, the ‘newness’ of Eurasian studies today should be presented as a respective ‘eye-opening’ on the part of scholars from Western and Newly Independent States. It is in this manner that today’s research provides a context for new data and new ideas to reinforce one another in a collaborative fashion.
Current Debates in Steppe Archaeology As noted, the current debates emerging within archaeological studies of Central and Inner Asia provide exciting lines of inquiry that may reframe our understanding of more general historical and anthropological processes. For example, recent archaeological studies of the Eurasian steppes have fruitfully enriched our understanding of horse and animal domestication, the origin of chariots (Brown and Anthony 1998; Anthony and Brown 1991), the development of long-distance networks (Chen and Hiebert 1995), the spread of Indo-European languages (Mallory 1989; Renfrew 1987), and the proliferation of metallurgy across Asia (Chernykh 1992; Mei J. and Shell 1999; Mei and Shell 1998). In reality, this is only a sampling of some dynamically changing research themes. Yet at the heart of all these issues lies a largely misunderstood and complex history of adaptation and social strategy concerning Bronze Age ‘tribes’, often ambiguously lumped into the ‘Andronovo culture’. Anyone who has worked in the vast steppe region invariably recognises that the steppe is neither a uniform social context, nor a consistent ecological setting. However, rough similarities in ceramic styles, burial types and technology have tethered archaeologists to
124
MICHAEL D. FRACHETTI
the established paradigm of cultural-historical social groups proposed more than 40 years ago. Unfortunately, these ‘cultural groups’ (such as the Andronovo culture or Fedorovo culture, etc.) fail to account for the differences in ecology, adaptation, organisation and identity of the populations that are lumped within them. Thus the ‘Andronovo paradox’ might best be summarised in the following way: 1) Since we have a cursory understanding of the details of Bronze Age life across the steppes, we can safely make generalisations and speculations about the nature of domestic life, language transfer, diffusion of material culture, horse riding and so on . . . but; 2) Because of the lack of real analytical details concerning Bronze Age life on the steppes, our generalisations preclude making substantial headway toward reconciling these very problems – without falling into traps of ethno-historical back-projecting or dubious oversimplification of the potential strategies of prehistoric steppe societies. One may observe that, simple, materially based culture-histories of Eurasia fail to convey the richness and complexity of the political economy of the Bronze Age, which is necessary for modelling the complex socio-economic innovations that occurred from the end of the 4th millennium BC in the steppe regions. More specifically, our understanding of the innovations on the steppe depends on the clarification of one the enduring problems in Eurasian archaeology – the emergence and nature of mobile pastoralism in prehistory. Unchallenged ethnographic models of pastoral nomadism have provided the basic picture of steppe societies in prehistory, giving us an easy explanation for the spread of everything from horse riding technology to ceramics, language and religion on the backs of ‘mobile’ societies. Were Bronze Age steppe societies ‘nomadic’ in the way we understand the term ethnographically? How can archaeologists escape ethnographic pigeonholes to build a ‘ground-up’ model of mobile pastoral life-ways in prehistory? For a more directed research question, perhaps we should be asking ourselves whether pastoral activities, and the anthropogenic landscapes these activities create, are actually reflected in the archaeological record of the ‘Andronovo Bronze Age’, or sufficiently represented in our recreations of Bronze Age lifeways? If not, then we need a better approach to understand how comparable
6. THE DZHUNGAR MOUNTAINS ARCHAEOLOGY PROJECT
125
components of material culture, such as ceramics, metals and even burial forms, can be found in disparate regions, while the populations of these different regions likely employed unique economic and political strategies. The discrepancy between relatively uninformative archaeological leftovers and the assumed richness of (agro?) pastoral life in the Bronze Age leaves us with a daunting project. The challenge is to model those strategies of prehistoric populations that delimit pastoral adaptations, while not excluding the possibility that the societies of the Bronze Age may have developed a way of life that was unique from our preconceptions of ‘nomadic society’. In fact, many scholars today avoid the term nomadism altogether because of this problem. But as I hope to show, we can recover evidence of a mobile pastoral way of life, at least in certain areas of the steppe, but we must make attempts to model the dynamics of communication across this vast region by first understanding the practicalities and limitations that local populations faced in their daily lives, and within their local ecological contexts. As I will discuss in detail below, a landscape approach represents a useful analytical method for identifying mobile adaptations in that it broadly relates archaeological features to ecological factors as well as the spatial and temporal dynamics of human investment in the land. In fact, landscape archaeology can improve our understanding of mobile strategies by modelling the ecological and physical constraints of the environment, as well as the agentive ways humans activate and use their landscape in spite of these natural limitations. Site based archaeological methods are inadequate for discerning prehistoric mobile pastoralism because they represent techniques designed for large-scale sedentary populations, used to study a mobile network of small ones.
Regional Environment of the Eastern Eurasian Steppes Semirechye means ‘seven rivers’; it refers to the rivers that flow north-west from the Tian Shan range (altitude ~ 7000 m) toward Lake Balkhash (Fig. 1). To the north-west of the Tian Shan, river drainages of the Dzhungar-Alatau range offer low grasses and a generally harsh climate. A wide plateau, known as the Dzhungarian Gate, provides a corridor between the Dzhungar-Alatau and Altai mountains, connecting the steppes of Kazakhstan, China and Mongolia (Lattimore 1940). The valleys are moist, but the soils are generally poor. Grassy
Fig. 1. Eastern Eurasian steppe zone and study region.
126 MICHAEL D. FRACHETTI
6. THE DZHUNGAR MOUNTAINS ARCHAEOLOGY PROJECT
127
river drainages lead to upland pastures, which provide summer grazing for sheep, cattle, horses and camels. The mobile pastoralists of this region traditionally practice seasonal vertical transhumance – going up the valleys to high pasture in the summer, and down to lower altitudes during the harsh winters (Fedorovich 1973; Vainshtein 1991; Dakhshleiger 1978).
Previous Archaeology The most comprehensive summary of the regional archaeology, the Arkheologicheskaya Karta Kazakhstana, published by the Kazakhstan Academy of Sciences in 1960 (Fig. 2), presents only a fraction of the sites and monuments from the region. The known monuments from this region are only roughly dated, and the Soviets mostly used typological or relative dating techniques. In addition, the majority of the sites are unexcavated, with many not even registered on the map. Recent research by Alexei Mar’yashev, of the Institute of Archaeology in Almaty, Kazakhstan, suggests that the drainage valley of the Koksu river is host to hundreds of burial kurgans and thousands of rock-art panels – likely dating to the Bronze Age or earlier (Mar’yashev and Goryachev 1993). Currently, the burials are considered to be a regional variant of the Andronovo type, with common ‘Fedorovo’ and ‘Alakul’ type ceramics, and simple metal grave goods (Goryachev and Mar’yashev 1998). Of great interest is the abundant rock art in the Koksu valley, studies of which have only been under serious study in the last ten years (Mar’yashev 1994; Mar’yashev 1994; Mar’yashev and Goryachev 1998). In addition, recent excavations at the settlement site Talapty (in the Koksu valley) have revealed interesting domestic artefacts such as grinding stones and large storage ceramics, which entice new investigations of Bronze Age Dzhungaria (Mar’yashev and Goryachev 1996). The archaeology of the Koksu valley is still at a preliminary stage, with the majority of the sites documented in the 1960s, representing rock art or large Saka (Iron Age) kurgans. Surprisingly, few of the kurgans have been systematically excavated. The upper reach of the Koksu river valley is completely unstudied, making the known Bronze Age sites in the lowlands difficult to interpret or contextualise within a broader scope. For this reason, a new collaborative project, the Dzhungar Mountains Archaeology Project (DMAP) (Frachetti and Mar’yashev forthcoming), was designed to provide a comprehensive data set from the Koksu river valley. The innovation of the DMAP is the use of a
Fig. 2. Known sites in Semirech’ye ca. 1960 (after Arkheologiskaya Karta Kazakhstana [1960]).
128 MICHAEL D. FRACHETTI
6. THE DZHUNGAR MOUNTAINS ARCHAEOLOGY PROJECT
129
landscape approach for the collection of consistent data across proposed landscapes and taskscapes (such as possible pastoral orbits), thereby illustrating a cohesive relationship between the ecological, social and ideological strategies of Bronze Age populations.
A Landscape Approach ‘Landscape’ oriented studies have been part of archaeology long enough that the term alone does not convey a singular approach (cf. Ashmore and Knapp 1999). Yet, most agree that landscape archaeology represents an approach that situates people in an environmental and social milieu, where they create and negotiate the ecological, political, ideological and ritual boundaries of their way of life (Anshuetz et al. 2001). Likewise here, a landscape approach assumes that various forms of economic activity and social interaction produce discernable anthropogenic ‘footprints’ that correspond to specific adaptive strategies in time, and that humans actively change the natural and social environment within which they live (cf. Erickson 2000). Tim Ingold points out that ‘landscapes’ reflect the impact of agents situated in time and space (Ingold 1993, 152), 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 are largely patterned and repetitive. Ethnographically, nomadic groups often demonstrate a unique spatio-temporal perception of social and environmental interaction (Evans-Pritchard 1940; Barth 1969; Beck 1991; also see Sarkozi 1976 on Mongolian ‘geomancy’). 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. As such, archaeologists should be able to regenerate the anthropogenic and natural footprints of those landscapes from prehistory. By documenting the rich archaeological resources of the Koksu valley, groundup modelling of various kinds of landscapes (ecological, ideological, political, etc.) more accurately depicts the complex factors that conditioned pastoral strategies during prehistory. Accordingly, the process of uncovering the conditions and reality of emerging ‘mobile pastoralism’ in prehistory is based in middle range analysis, rather than in back-projected ethnography. Specifically, the landscape approach applied here considers the habitation ecology, the geogra-
130
MICHAEL D. FRACHETTI
phy of ritual and ideological features, and the association between the built landscape and marking of group or individual identity, as fundamental components of the Bronze Age political economy. Conceived in this way, the landscape approach serves as the vital link between theoretical questions concerning pastoral subsistence strategies and the analytical techniques available to archaeologists for discerning mobile pastoralism in prehistory (Fig. 3). Modelling Habitation Ecology. Habitation ecology encompasses the economy and living strategies employed by societies over time and space, within the limits and opportunities of specific environments. The annual conditions of the Koksu valley’s sub-alpine environment structure specific pastoral activities both spatially and temporally. For example, the variation of altitude in the Koksu valley acts as a seasonal indicator, as higher elevations are generally uninhabitable during winter months. Viewing the location of archaeological features within a seasonal landscape allows for correlations between material forms and group movements or activities. For example, if seasonal transhumance was the predominant model, we would expect to find more substantial group settlements in winter lowland areas and shorter lived herder’s camps in the summer high pastures (cf. Fedorovich 1973). Yet vertical transhumance is only one model; the spatial organisation of the sites in the valley should lead the archaeologist to test other possible adaptations. Second, the environment has constraints in terms of seasonal biomass (Kalinina 1974; Goloskokov 1984) that can be assessed both through archaeobotanical methods and from analysis of colour satellite imagery. Using archaeobotanical and palaeoclimatic information, some demographic limits can be deduced for both human and animal populations. How many animals could be supported and moved in various locations at different times? Based on accessibility of grasses, how early in the year could a group begin to migrate? Can we detect traces of pathways from satellite imagery? What alternative routes existed during seasons of unusual flooding, or long winters? The key to the ecology aspect of landscape archaeology is to recognise how various strategies demand different temporal or seasonal schedules, and to link the archaeological ‘footprint’ with ecological and seasonal factors. Ideological Landscapes and Marked Identity. Of course, the landscape incorporates more than merely economic and biotic scheduling; ritual, ideology and
Fig. 3. Landscape approach to mobile pastoralism.
6. THE DZHUNGAR MOUNTAINS ARCHAEOLOGY PROJECT
131
132
MICHAEL D. FRACHETTI
group affinity are also impressed and build into it. On the Eurasian steppe, the most obvious anthropogenic fixtures in the landscape are burials and rock art. The location of burials and their association with other features such as rock art, settlements, pastures, spectacular vistas and other loci, is largely understudied. These anthropogenic features can also be socially contextualised by detailing the anthropological data (for example, the sex of the interred), the associated material culture (ceramics, metal, etc.), and by fixing the monuments within various temporal and spatial schedules of pastoral activity. Mapping formal relationships between ritual spaces and domestic spaces (for instance, burials and pastures), and using GIS to understand the creation of anthropomorphic or celestial associations through rock art, earthworks, or kurgan burials, are productive techniques for uncovering how the landscape was imbued with meaning. How do landscape features communicate conditions of social affinity or political alliances? How do landscape constructions correlate with other structures, or with naturally occurring landscape elements? Does the location of specific monuments relate to the movement of herds, seasonal hunting, or other factors? These and other questions are provocative for recasting life in the Bronze Age.
The Dzhungar Mountains Archaeology Project The Dzhungar Mountains Archaeology Project (DMAP)1 began in 1999, and is designed to address the issue of emerging pastoralism and social interaction in the eastern Eurasian steppe zone. The practical aim of the DMAP is to record and model the social landscape of mountain populations in prehistory – which entails mapping the ecological and anthropogenic environment of the Koksu river valley (Fig. 4), within a framework of proposed activities of human agents in space and time. As discussed above, the landscape approach enables us to discern how the Bronze Age population of this region took advantage of available resources, and how they used and constructed a landscape to live beyond their ecological limitations. Thus, we are interested in determining the reality of various subsistence strategies, social interactions and ideological frameworks, using grounded and methodologically consistent archaeological data recovery.
1
The DMAP is funded by the National Science Foundation and the George F. Dales Foundation for archaeological research.
Fig. 4. Satellite image of the Koksu river valley (source: USGS Landsat TM-7 2002).
6. THE DZHUNGAR MOUNTAINS ARCHAEOLOGY PROJECT
133
134
MICHAEL D. FRACHETTI
The 2002 archaeological survey was a collaborative effort with Kazakh archaeologists (Dr Alexei Mar’yashev) from the Institute of Archaeology in Almaty, and with geologists (Dr Bulat Aubekerov) and botanists (Dr Saida Nigmatova) from the Kazakh National Academy of Science (also in Almaty). In addition, a number of other specialists are working to provide a variety of additional data. Our methodology builds on a number of well-known field techniques. These include: 1) surface survey and mapping; 2) palaeo-environmental sampling; 3) archaeological excavation; and 4) computer assisted spatial modelling using GIS. Before going to the field, a comprehensive map and satellite imagery study was conducted, making the project’s survey design as streamlined, accurate and informed as possible. For example, Corona images available from the US Geological Survey enable the identification of modern agricultural development, towns, roads and other important features. These images are also being used in conjunction with Landsat-7 colour imagery for detailed modelling of environmental conditions and archaeological features. The DMAP represents a collaborative research effort, bringing together American and Kazakh archaeologists, as well as Kazakh geologists and paaleobotanists, toward the production of a coherent archaeological, palaeo-ecological and geomorphological reconstruction of the Koksu valley during the Bronze Age. Archaeological Survey. The survey component of the DMAP was focused within the Koksu river valley. The main objective of the land survey was to make a detailed database record and digital map of the archaeological monuments (cist burials, kurgans, settlements, rock art, megaliths, etc.) located in the river valley. Conducted in May 2002, the surface survey accounted for more than 1500 km2 of total landscape analysis, and 106.7 km2 (10671 ha) of field walked polygons. For archaeological recovery, the Koksu river valley and floodplain was divided into six topographic landscape polygons: two lowland polygons, two mid-elevation polygons, two upland elevation polygons (Fig. 5). Within each polygon, the actual area walked was determined by contextual conditions; many of the sub-alpine areas were treacherous and/or inaccessible on foot. Nonetheless, the estimated coverage within the survey polygons is 85–90% of the total surface area. For each polygon, the archaeological survey team made a detailed record of all features and locales indicating human manipulation. These feature types included settlements, stone configurations, rock art, burials and general anthropogenic features. A feature is defined as a discrete archaeological element. Thus a group
Fig. 5. Survey region and polygons of Dzhungar Mountains Archaeological Project.
6. THE DZHUNGAR MOUNTAINS ARCHAEOLOGY PROJECT
135
136
MICHAEL D. FRACHETTI
of five burials was recorded as five features. Rock-art features were grouped into panels, based on proximity. Examples of ‘anthropogenic features’ include dense rectangular patches of phosphate rich grass growth (indicative of a corral), or animal terraces, formed by the repetitive tracking of herds across hill slopes. In the Koksu river valley, settlements and burials are generally visible on the surface as stone configurations, or visible in profile where erosion ravines have cut across sites. Rock art is naturally exposed, and was photo documented in all cases. All features were recorded using a standardised record form, and mapped and/or photographed. Of course, feature locations were marked using hand-held GPS units.2 In addition to the archaeological team, a small team of geologists also recorded and studied the geomorphology of each survey polygon, and produced an overall assessment of the morphological development of the valley. Preliminary Survey Results. Within the scope of the archaeological survey, we recorded 382 archaeological features. For the purpose of classification, sites were categorised as low, middle or high altitude features (Fig. 6a–b). Lowland areas are categorised by elevations less than 800 m. Mid-altitude areas range from 800 m to 1200 m, and high-elevation zones are above 1200 m. The altitude designations correspond with ecological factors, such as snowfall threshold, as well as vertical boundaries between various botanical compositions of the rangelands (Goloskokov 1984). The Dzhungar mountains are located between 43° and 46° N latitude, which is typically represented by a semi-arid desert climate. However, due to the steep rise of elevation in the Dzhungar mountain range (from 600 m to over 4500 m), the botanical composition is similar to grasslands regions, located north of 50° latitude. In the Dzhungar mountains, rangeland above 1200 m is characteristically ‘steppic’, while zones below 1200 m are considered semi-steppe, or semi-desert (below 800 m) (Goloskokov 1984). Thus, categorising the features according to elevation can reveal preliminary associations between ecological conditions, social pathways and human strategies. In the lowland polygons, our primary feature categories were burials (kurgans and stone arrangements) and settlement structures. The largest numbers of burials are classified as kurgans – generally considered to date to the Saka 2
Garmin Global Positioning Systems with accuracy from 4 to 10 m were used. The average error was between 4 and 6 m.
Fig. 6. Dzhungar Mountains Archaeological Project sites within each altitude zone. A: Percentage B: Number
6. THE DZHUNGAR MOUNTAINS ARCHAEOLOGY PROJECT
137
138
MICHAEL D. FRACHETTI
period and later (i.e. later than the 7th century BC). Two notable Bronze Age cemeteries at Talapty and Kuigan had been previously excavated, revealing stone lined cists and small hand-thrown ceramic vessels. (Mar’yashev and Karabaspakova 1988; Mar’yashev and Goryachev 1993). These cemeteries contain more than 60 individual burials, which were not individually recorded in the DMAP database. In addition to burials, settlement structures were found to be more abundant in lowland areas than previously recorded. In fact, nearly 35% of the total sites in lowland areas were settlements. In association with these monuments, a considerable number of petroglyphs are found in the immediately adjacent hillsides. However, the elevation of many of the actual rock art is often higher than 800 m, thus they fall into the statistical set of midland elevations. In the midland polygons, the percent composition of features is generally comparable, but the total number of sites is much higher. Noticeable increases in the number of burials (both kurgans and stone arrangements), petroglyphs and settlements at midland altitudes, suggests increased activity in this zone. Ethnographic models of seasonal transhumance explain that midland altitudes are productive in both spring and fall, whereas high and low altitudes are only used in the summer and winter, respectively (Chekeres 1973). This pattern, or seasonal orbit, may account for the increased percentages of specific feature types. However, the results of analysis currently underway concerning seasonal vegetation, palaeozoology and prehistoric land-use, will surely provide more informative explanations. Also situated within the midland polygons, the settlement site of Begash was selected for excavation, as well as burials from the large Bronze Age cemetery of Begash II nearby (briefly discussed below). There are considerably fewer sites in the highland polygons. The most notable change is an increase in percentage of kurgans. This may reflect a conscientious utilisation of highlands as burial grounds in the Iron Age, or perhaps merely reflects the difficulty in locating settlements in sub-alpine meadows. Due to the logistical difficulties in mounting an expedition to some of these remote sub-alpine meadows, the collection of data in high altitude zones of the Dzhungar mountains demands special techniques, which will be employed in future surveys there. Excavations. In order the better to contextualise some of these monuments, we initiated and directed excavations of a settlement site (Begash) and burial complex (Begash II) dating to the beginning of the 2nd millennium BC. For the settlement site, the excavation strategy was designed to recover both eco-
6. THE DZHUNGAR MOUNTAINS ARCHAEOLOGY PROJECT
139
logical data as well as cultural material. Working with palaeoclimatologists, botanists and geomorphologists, our strategy included botanical sampling, soil sampling, micro-morphological sampling,3 and the collection of organic material suitable for C14 dating. From excavated contexts we collected over 1000 ceramic fragments, and over 50 kg of faunal remains, as well as spinning and weaving artefacts, grindstones, and bone implements. In addition to the settlement excavation, three Bronze Age burials were excavated, revealing human remains as well as rare bronze and gold earrings. With the permission of the Kazakh authorities, the human remains were brought to the University of Pennsylvania for studies of DNA and physical characteristics. This is one of the few instances since the fall of the Soviet Union that a collection of Central Asian human remains is being studied within the United States. Also significant, Electron Spin Resonance was successfully used to date one of the burials to the 17th century BC, making it one of the oldest dated Bronze Age burials in the region.4 The analysis of the above-mentioned samples is conducted in part by scientists in Kazakhstan, and in part by American and European specialists, making the DMAP a completely collaborative and international research effort. Complete publication of this material is forthcoming. GIS modelling. Data synthesis and computer modelling is the final phase of the project, which entails using GIS to understand the distribution of archaeological features and ecological conditions within the study zone. Computer simulations allow for the regeneration of past landscapes, as well as an understanding of how sites are statistically situated in the valley, by correlating the actual monument types with various factors such as proximity to springs or grassy fields. GIS also allows the use of time simulation algorithms, with which various mobile strategies can be related to factors such as geographic impedance, economic viability, social interaction, ritualised space and human decision-making. This modelling phase will be used to explore various temporal and spatial hypotheses concerning mobile pastoralism, as well as other subsistence models (such as mixed agro-pastoralism). The goal of this phase of analysis is to use archaeological and ecological data to expose the reality of mobile pastoral 3
Micro-morphology will be useful in discerning seasonal or periodic renewal of floors, as well as the natural processes that affect site formation. 4 ESR calculates the amount of electrons that accumulates in bone samples, according to a known rate of accumulation.
140
MICHAEL D. FRACHETTI
strategies during the Bronze Age, and subsequently propose how various groups may have interacted across broader regions given the scale of their social landscape. Results from this stage of synthesis will be available soon.
Conclusions The DMAP is a collaborative project bringing together state of the art technology and survey techniques, while in many ways falling in line with more than 60 years of research in the Semirechye region. Since we are still in the early stages of analysis of this material, I have chosen to wait to make interpretive statements. The productivity of the Dzhungar Mountains Archaeology Project, and in fact the success of any archaeology project in this vast region, will depend on our ability as archaeologists to work toward intelligible reconstructions of prehistoric life-ways from a variety of analytical techniques and grounded data. These results will form part of a solid database through which we can better compare regional developments across Eurasia.
Bibliography Arkheologicheskaya Karta Kazakhstana 1960: (Almaty). Anshuetz, K.F., Williams, R.H. and Scheick, C.L. 2001: An Archaeology of Landscapes: Perspectives and Directions. Journal of Archaeological Research 9, 157–211. Anthony, D.W. and Brown, D.R. 1991: The Origins of Horseback Riding. Antiquity 65, 22–38. ——. 2000: Eneolithic Horse Exploitation in the Eurasian Steppes: Diet, Ritual and Riding. Antiquity 74, 75–86. Ashmore, W. and Knapp, A.B. 1999: Archaeologies of Landscape (Malden, Mass.). Barth, F. 1969: Ethnic Groups and Boundaries (Boston). Beck, L. 1991: Nomad: A Year in the life of a Qashga’i Tribesman in Iran (Berkeley). Bromlei, I.V. 1983: Ocherki teorii etnosa (Moscow). Bromlei, I.V. and Kozlov, V. 1989: The Theory of Ethnos and Ethnic Process in Soviet Social Science. Comparative Studies in Society and History 31, 425–38. Brown, D.R. and Anthony, D.W. 1998: Bit Wear, Horseback Riding and the Botai Site in Kazakhstan. Journal of Archaeological Science 25, 331–47. Chekeres, A.I. 1973: Pogoda, klimat i otgonno-pastbishchnoe zhivotnovodstvo (Leningrad). Chen, K.T. and Hiebert, F. 1995: The Late Prehistory of Xinjiang in Relation to its Neighbors. Journal of World Prehistory 9, 243–300. Chernykh, E.N. 1992: Ancient Metallurgy in the USSR: The Early Metal Age, (Cambridge/New York). Dakhshleiger, G.F. 1978: Settlement, and Traditional Social Institutions of the Formally Nomadic Kazakh People. In Weissleder, W. (ed.), The Nomadic Alternative (Paris), 361–9. Erickson, C. 2000: The Lake Titicaca Basin: A Precolumbian Built Landscape. In Lentz, D. (ed.), Imperfect Balance: Landscape Transformations in the Precolumbian Andes (New York), 311–56.
6. THE DZHUNGAR MOUNTAINS ARCHAEOLOGY PROJECT
141
Evans-Pritchard, E.E. 1940: The Nuer (Oxford). Fedorovich, B.A. 1973: Prirodnye usloviya aridnykh zon SSSR i puti razvitiya vnikh zhivotnovodstvo. In Abramzon, S.M. and Orazov, A. (eds.), Ocherki po istorii Khozyaistva narodov Srednei Azii i Kazakhstana (Leningrad), 207–22. Goloskokov, V.P. 1984: Flora Dzhungarskogo Alatau (Almaty). Goryachev, A.A., and Mar’yashev, A.N. 1998: Nouveaux sites du Bronze Recent au Semirech’e (Kazakhstan). Paleorient 24, 71–80. Harris, D., Gosden, C. and Charles, M. 1996: Jeitun: Recent Excavations at an Early Neolithic Site in Southern Turkmenistan. Proceedings of the Prehistoric Society 62, 423–42. Hiebert, F.T. 1994: Origins of the Bronze Age Oasis Civilization in Central Asia (Cambridge, Mass.). Ingold, T. 1993: The Temporality of the Landscape. World Archaeology 25, 152–74. Kalinina, A.W. 1974: Osnovnye tipy pastbishch Mongol’skoi Narodokhoi Respubliki: ikh struktura i produktivnost’ (Leningrad). Kislenko, A. and Tatarintseva, N. 1999: The Eastern Ural Steppe at the End of the Stone Age. In Levine, M., Rassamakin, Y., Kislenko, A. and Kislenko, T.N. (eds.), Late Prehistoric Exploitation of the Eurasian Steppe (Cambridge), 183–216. Lattimore, O. 1940: Inner Asian Frontiers of China (Boston). Mallory, J.P. 1989: In Search of the Indo-Europeans (London). Mar’yashev, A.N. 1994: Petroglyphs of South Kazakhstan and Semirechye (Almaty). Mar’yashev, A.N. and. Goryachev, A.A. 1993: Voprosy tipologii i khronologii pamyatnikov epokhi bronzy Semirechiya. RosA 1, 5–20. ——. 1996: Otchet: o polevykh issledovaniyakh otdela arkheologicheskoi tekhnologii v 1995g (Delo # 2316/sv 181) (Almaty). ——. 1998: Naskal’nye Izobrazheniya Semirech’ya (Almaty). Mar’yashev, A.N. and Karabaspakova, K.M. 1988: Novye pamyatniki Epokhi Bronzy vostochnogo Semirech’ya. In Medvedev, V.E. and Khudyakov, Y.S. (eds.), Drevnie pamyatniki severnoi Azii i ikh okhrannye raskopki (Novosibirsk), 24–39. Mei, J. and. Shell, C. 1998: Copper and Bronze Metallurgy in Late Prehistoric Xinjiang. In Mair, V. (ed.), The Bronze Age Peoples of Eastern Central Asia (Philadelphia), 581–603. ——. 1999: The Existence of Andronovo Cultural Influence in Xinjiang during the 2nd Millennium BC. Antiquity 73, 570–8. Renfrew, C. 1987: Archaeology and Language: The Puzzle of Indo-European Origins (Cambridge). Sarkozi, A. 1976: A Mongolian Manual of Divination by Means of Characteristics of the Land. In Sinor, D. (ed.), Tractata Altaica (Weisbaden). Tolstov, S.P. 1952: Khorezmskaia arkheologo-etnograficheskaia ekspeditsiia: Trudy Khoremzskoi arkheologo-etnograficheskoi ekspeditsii (Moscow). Vainshtein, S.I. 1991: Mir Kochevnikov Tsentralnii Azii (Moscow).
CHAPTER SEVEN THE KURA-ARAXES ‘CULTURE’ IN THE NORTH-EASTERN CAUCASUS: PROBLEMS IN ITS IDENTIFICATION AND CHRONOLOGY RABADAN MAGOMEDOV
Soon 40 years will have elapsed since B.A. Kuftin defined a very clear and distinctive archaeological culture, which he initially termed on the basis of its geographical distribution ‘the Kura-Araxes Aeneolithic’ or ‘the Kura-Araxes Aeneolithic culture’ (Kuftin 1944). From the beginning investigators were fascinated by the striking originality of this culture and even spoke about the possibility of seeing in it the basis for the former Caucasian-Iberian ethno-linguistic unity (Munchaev 1961, 164–5; Krupnov 1964, 26). However, already in 1959 Prof. G.A. Lomtatidze at a scientific session of the Institute of History, Languages and Literature of the Daghestan Branch of the Soviet Academy of Sciences had cautioned his colleagues ‘not to be too attracted to this question since the more that is uncovered of Aeneolithic sites each year, the more basis there will be to observe cultural areas not only in the Caucasus but within Transcaucasia and within the Kura-Araxes Aeneolithic. The latter perhaps seemed like an integral unit when it was defined by the late B.A. Kuftin, but today it possibly can be divided into separate units’ (Lomtatidze 1961, 297). As was expected, further investigations of the Kura-Araxes culture took place in different republics, then within republics and within gorges and narrowly circumscribed groups of sites. For a long time specialists have recognised that its chronological parameters relate to the Early Bronze Age, not to the Aeneolithic. The territory over which it is distributed now includes part of the central Caucasus, the north-eastern Caucasus, almost all of Transcaucasia (save for western Georgia or Colchis and the Black Sea coast), eastern Anatolia, the north-western part of the Iranian plateau, the lands of north-eastern Syria, Lebanon, and almost all of historical Palestine, where the Kura-Araxes culture and the culture of the Khirbet Kerak ware is seen as nearly identical. The principal aim of this paper is to attempt a new conceptual identification of the Kura-Araxes sites of the north-eastern Caucasus and an examination of
7. KURA-ARAXES ‘CULTURE’ IN THE NORTH-EASTERN CAUCASUS
143
these sites in terms of their chronology and periodisation. Following the work of M.G. Gadzhiev (1991), under the term north-eastern Caucasus we understand the basic physical-geographic and historical ethnocultural region of the Caucasus that includes part of Ingushetia, Chechnya, Daghestan and north-eastern Azerbaijan (Khachmas, Kuba, Kusar and Divichi administrative regions or what is frequently now referred to as the Khachmas-Kuba zone). One additional clarification: we will not consider aspects of the origins of the KuraAraxes culture, in general, or of its separate variants, in particular, nor will we discuss the reasons for its final decline. There exist different opinions in the Caucasian specialist literature on the question of the cultural affiliation of the Kura-Araxes sites in the north-eastern Caucasus; essentially, these opinions can be divided into two groups. The most common view is to emphasise the similarity of the sites situated in the northeastern Caucasus, including the Khachmas-Kuba zone, and to group them into a single local variant of the Kura-Araxes culture. Earlier A.P. Kruglov (1958, 29), B.B. Piotrovsky (1949a, 38–9; 1949b, 179, 183), R.M. Munchaev (1955, 18; 1961, 155), E.I. Krupnov (1964, 37) and K.K. Kushnareva and T.N. Chubinishvili (1970, 60–61) referred to the Daghestan sites (Kayakent, Velikent and others) as the northernmost sites of the Kura-Araxes culture with clearly expressed local features. They did not know about the antiquity of the Early Bronze Age sites in north-eastern Azerbaijan. From the 1970s on, numerous Kura-Araxes sites in this area have been investigated, and today the thesis is widely accepted to group Daghestan and north-eastern Azerbaijan into a separate local variant of the Kura-Araxes culture (Khalilov and Kesamanly 1973, 55–63; Kushnareva and Dzhaparidze 1978, 274; Gadzhiev 1990, 68; 1991, 230–1; Khalilov, Koshkarly and Arazova 1991, 125–6). One must also note that recently Munchaev (1994, 25) somewhat changed his previous interpretation of this problem and asserted that ‘the settlements and cemeteries of Daghestan appear as a distinctive local group of sites . . . of the [Kura-Araxes] cultural-historical community.’ One would have to await further investigations to speak about their affinity with sites in the Khachmas-Kuba zone. A second group of investigators, principally archaeologists from Azerbaijan, examine the Kura-Araxes sites of north-eastern Azerbaijan as a separate variant of the Kura-Araxes culture, functioning as a link between similar sites in Daghestan and the northern Caucasus, on the one hand, and the basic southern group of sites of this culture, on the other (Akhundov 1987; Musaev 1991, 14; 1992, 114–6). G.S. Ismailov (1983, 34; 1987, 74–80) proposes a somewhat different position, considering it
144
RABADAN MAGOMEDOV
possible to define a series of distinctive cultures within the Kura-Araxes cultural-historical community (or CHC), including the northern Caucasian and Azerbaijanian (occupying ‘approximately today’s territory of Azerbaijan’). At present approximately 90 Kura-Araxes culture sites are known in the north-eastern Caucasus (Fig. 1). The geographical distribution of these sites shows that the greatest concentration of them are situated in the lowland plain of Daghestan and north-eastern Azerbaijan. Very few are recorded in Chechnya or in many regions of internal or mountainous Daghestan. Both the degree of research on these Kura-Araxes sites and the availability of the publications of their materials vary greatly. The sites that are most thoroughly studied include: Serzhenyurt I and II in Chechnya, Galgalatly I, Chirkey (Tad-Shob), Kayakent (Geme-tyube I–II), Velikent I–II, Mamai-Kutan, TorpakhKala and Kabaz-Kutan I in Daghestan, and Serkertepe (formerly Borispol’tepe) in Azerbaijan. Peculiarities in their architectural plans and building methods vary according to their location. Essentially, they can be divided into two large groups: firstly they appear in mountainous parts of Daghestan; secondly, in the piedmont and valley sections of Chechnya, Daghestan and north-eastern Azerbaijan. The ‘mountain’ settlements are situated principally on rises and remains of river terraces (Chirkey, Mekegi) and on crests and steep slopes of mountains (Sigitma, Galgalatly I). Circular houses typically made of stone are most characteristic of these mountain sites. The houses have flat roofs, clay shelves and benches along the walls, and central circular hearths set into the floors; sometimes two-chambered kilns with arched roofs are found next to the houses near the walls. The cultural deposits on these mountain settlements are typically thin, rarely reaching 2 m in depth. The ‘coastal plain’ sites are located on natural hills, frequently on the edge of river and marine terraces. As a rule, the cultural deposits on these sites vary between 2–3 m in depth and contain several building levels. On some coastal plain sites of Daghestan (such as Velikent I–II, Torpakh-Kala and Kabaz-Kutan I) the cultural deposit is 4–6 m thick. Their houses are made with wickerwork frames (or wattle-and-daub), mud bricks, and, more rarely, stones, but also include pit-houses and semi-subterranean houses. In most cases the houses have circular plans. As in the mountain settlements, houses in these coastal plain sites typically have circular hearths set in their centres. One must note that both the plain and mountain sites of this north-eastern Caucasus variant lack the internal projections or ‘petals’ for these circular hearths that are characteristic of the Transcaucasian
7. KURA-ARAXES ‘CULTURE’ IN THE NORTH-EASTERN CAUCASUS
Fig. 1. Map of Velikent Kuro-Araxes culture sites.
145
146
RABADAN MAGOMEDOV
Kura-Araxes settlements. Both the plain and mountain sites exhibit a general tendency for the houses to develop from circular to rectangular forms. Very few mortuary remains of this north-eastern Caucasian variant of the Kura-Araxes culture have been uncovered and studied. We only have three Early Bronze Age cemeteries that have been investigated in the mountains and piedmont regions of the north-eastern Caucasus; these are Shebokha, Gono and Karatsan II. Stone chambers with collective burials are typical. Likewise it is possible to trace a line of development from circular to rectangular burial chambers. These latter continue to develop during Middle Bronze Age times. A whole series of mortuary sites have been uncovered along the coastal plain of Daghestan; these include interred earthen burials (Karabudakhkent, Kayakent VI, Velikent, III, IV and V, and Dzhemikent II) and kurgans (Manas, KataragachTapa and Shakh-Nasib Tapa). Catacomb-shaped chambers with collective burials are characteristic for both the raised kurgans and interred earthen burials. As is well known, a distinguishing feature of the Kura-Araxes culture is the great variety of mortuary practices. Except for the coastal plain of Daghestan, catacomb burials are not found anywhere throughout the broad territory over which Kura-Araxes sites are distributed. Catacomb burials, which have been uncovered in Armenia, relate to the Middle and Late Bronze Age, and some examples of Kura-Araxes burials in Georgia are similar in form to the catacomb burials but are still not the same as those from the coastal plain of Daghestan. Questions concerning the origins of the catacomb burial constructions and the kurgans in the north-eastern Caucasus are still not resolved. The earliest excavated kurgan seems to be that of Torpakh-Kala (Kotovich 1985), in which was found a burial complex of ‘Maikop type’. Unfortunately, KuraAraxes burial sites are still not known in north-eastern Azerbaijan, and we cannot evaluate the features of their Early Bronze Age mortuary rites in comparison with the materials from Daghestan. The cultural similarity of ‘Kura-Araxes culture’ sites in the north-eastern Caucasus that allow us to distinguish them from other variants of the ‘KuraAraxes cultural-historical community’ is most clearly evident in their material inventories, particularly in their ceramics. The ceramics of the north-eastern Caucasian Kura-Araxes culture sites (Fig. 2) show an indisputable similarity in the technology of their manufacture and their shapes and decorations which allow us to differentiate them from the entire corpus of Kura-Araxes ceramics. At the same time, they exhibit local distinguishing features that are related to their chronology and regional distribution. For example, high quality, primarily
7. KURA-ARAXES ‘CULTURE’ IN THE NORTH-EASTERN CAUCASUS
147
Fig. 2. Ceramics: 1, female figurine (Karatsan); 2–5, andirons (2 – Kabaz-Kutan I; 3 – Galgalatly; 4–5 – Velikent II); 6–13, vessels (6 – Velikent III; 7–8 – Velikent II; 9 – Velikent V; 10, 12 – Kayakent VI; 11 – Chirkey; 13 – Torpakh-Kala); 14–15, frying pans (14 – KabazKutan I; 15 – Galgalatly).
148
RABADAN MAGOMEDOV
wheel-made burnished wares (Fig. 2) with highly specific decorations (toothed and combed impressions) are found on a series of early settlements in Chechnya (Serzhenyurt I–II), Daghestan (Velikent II, Kabaz-Kutan I) and Azerbaijan (Serkertepe, Dashlytepe). Conventionally, we refer to this ware as a Velikent II type (Munchaev 1975, 359, fig. 76.9–11; Gadzhiev et al. 1997, 198–9, fig. 16; Gadzhiev et al. 2000, 51, 65, 67–8, figs. 7a.1, 7a.10, 21). This ware is completely uncharacteristic for Transcaucasian Kura-Araxes culture sites. Similarly, roughly coated or coarse-slipped wares (Fig. 2) are not found on the Transcaucasian sites, except for some in Kobystan and Kakheti, comparable to those which first appear on sites in the north-eastern Caucasus sometime around the middle of the 3rd millennium BC, and which become extraordinarily widespread in this area during the Middle and Late Bronze Age. In general, a lower percent of decorated wares and the general impoverishment of their ornamental motives and compositions distinguishes the north-eastern Caucasian ceramics from other Kura-Araxes ceramics. Transcaucasian ceramics very rarely exhibit the method of decoration with different applied bands or cordons that is such a characteristic feature of the north-eastern Caucasian ceramics (Fig. 2.6). Conversely, the concentric spiral decoration, which is one of the most specific distinguishing decorative features of classic Kura-Araxes ceramics, only appears rarely on late coastal plain sites of the north-eastern Caucasus. Also types of vessels, such as decorated goblets, cups, and vases with hollow stands (Fig. 2.10, 12–13), large narrow-based vessels with high cylindrical necks (Fig. 2.11), and flat braziers with sides sometimes pierced with holes opened at their edges (Fig. 2.14–15), are very characteristic for many of the sites in the north-eastern Caucasus, while they are only infrequently encountered in other areas of the Kura-Araxes culture. One must also note that different types of bowls occupy a significant place in the general mass of ceramics from the north-eastern Caucasus in all phases of their development. This feature is not observed in Transcaucasia. Almost all the different types of hearth supports (Fig. 2.2–5) so characteristic for the KuraAraxes are found in the north-eastern Caucasus. The horseshoe-shaped hearth supports, however, which are so broadly distributed throughout the other areas occupied by Kura-Araxes sites, are only extremely rarely encountered in the north-eastern Caucasus (for example, at the Sharakun settlement). Finally one must note that the clay anthropomorphic (typically female) figurines (Fig. 2.1) from the north-eastern Caucasus can be typologically distinguished from the terracotta materials from other Kura-Araxes culture sites (Gadzhiev et al. 1995, 144, pl. 24c; 2000, 51, 65, fig. 7b.9, 7b.19–20).
7. KURA-ARAXES ‘CULTURE’ IN THE NORTH-EASTERN CAUCASUS
149
A comparative examination of the non-ceramic inventories also reveals peculiarities of the north-eastern Caucasian sites from other Kura-Araxes culture sites. If the distinction is not so obvious in terms of stone and bone artefacts (although the numerous graceful, polished perforated stone axes, as are found on sites in Daghestan and in north-eastern Azerbaijan, are not encountered on the Transcaucasian sites), one cannot say this about the metals, especially those relating to the late stage of the development of the Kura-Araxes culture in the north-eastern Caucasus (Gadzhiev et al. 1997, 90–1, 93, 101, figs. 49–53, 57). The north-eastern Caucasian metal is distinguished by its ore sources, particularly by the composition of its base metals (as opposed to Transcaucasia, the north-eastern Caucasus have relatively few tin-bronzes), and by the limited types of characteristic metal objects. Very rich collections of bronze tools, weapons and ornaments (Figs. 3–4) have been found in the late catacomb collective tombs of Velikent that are undoubtedly specific to the north-eastern Caucasus region. The Velikent bronzes, such as shaft-hole axes (Fig. 3.27), chisel-shaped, socketed objects of uncertain use (Fig. 3.10), toggle-pins with circular heads and perforated mid-sections (Fig. 4.18), crook-shaped, intricately decorated toggle pins (Fig. 4.14–16), rod-shaped toggle pins (Fig. 4.11–12), anchor-shaped pendants (Fig. 4.29, 30, 34–35), bracelets with thickened sections (Fig. 4.19, 20, 22–23, 25), long beads with triangular sections (Fig. 4.6), and other types, essentially provide no direct or broad analogies with objects from other Early and Middle Bronze Age Caucasian sites. Likewise double-volute pendants (Fig. 4.21) that are very well known on Early and Middle Bronze Age sites of the north-eastern Caucasus are absolutely not characteristic for the vast majority of Kura-Araxes Caucasian sites. In our opinion, based on the level of unity between characteristics of the north-eastern Caucasian sites, and differences in objects from them and those associated with Kura-Araxes sites, it is possible to refer to them as belonging to separate archaeological cultures. Our attempts to think about the KuraAraxes culture in terms of different taxonomic units of one archaeological formation (separating an archaeological cultural-historical community into several independent cultures) is not new. The opinion of G.S. Ismailov, an Azerbaijani archaeologist, concerning this problem has already been mentioned. Even earlier, G.E. Areshyan (1982, 256) suggested that it is possible to separate out the Shengavit archaeological culture from the larger community of the Kura-Araxes cultural sphere. Finally, of special importance to this argument is the opinion of Gadzhiev (1991, 233–4) who said that ‘the culture which is represented in
150
RABADAN MAGOMEDOV
Fig. 3. Bronze and stone artefacts (1 – Velikent I; 2–5, 7–11, 14–19, 21–27 – Velikent III; 20 – Velikent V; 6, 12 – Velikent – II; 13 – Bugdatepe; 28 – Velikent IV).
7. KURA-ARAXES ‘CULTURE’ IN THE NORTH-EASTERN CAUCASUS
Fig. 4. Bronze, bone and paste artefacts (1–20, 22–36 – Velikent III; 21 – Manas).
151
152
RABADAN MAGOMEDOV
the Early Bronze Age sites of the north-eastern Caucasus (the north-eastern Caucasian culture) . . . belongs to the vast Kura-Araxes cultural community’ (see also Gadzhiev 1990, 69–70, 72). We argue that this separate culture (the northeastern Caucasian culture), framed within the Kura-Araxes archaeological culture complex, should be called the Velikent culture. Unlike other terms that exist in the literature, which describe local variants of the Kura-Araxes culture in the north-eastern Caucasus (such as ‘north-eastern Caucasian’, ‘western Caucasian’, or ‘Daghestani’), this name is easier to use and does not carry any known ethnic baggage. Besides this, we should keep in mind that Velikent is not only a well studied complex of different kinds of sites, which reflect the whole chronology and periodisation of this culture, but also that Velikent was practically the first Kura-Araxes site in the north-eastern Caucasus, discovered in 1880 by A.A. Rusov. It should be noted here that the term ‘Velikent culture’ is actively used by Gadzhiev when referring to post-Kura-Araxes complexes associated with the Middle Bronze Age in the coastal plain of Daghestan, such as the upper layers of the settlement of Velikent I, Mamai-Kutan, Kayakent I, and also late Velikent catacomb burials (Gadzhiev 1991, 237–8; Gadzhiev, Davudov and Shikhsaidov 1996, 75–7). Attempts to designate this region as ‘the Velikent centre of metallurgy’ of the Middle Bronze Age (Gadzhiev and Korenevsky 1984, 25–6) and similarly to organise sites of this region into a ‘Velikent group of sites of the Middle Bronze Age’ (Magomedov 1985, 221–3) were undertaken earlier. It is interesting to note that before the discovery of the Velikent catacomb burials, all specialists agreed that the abovementioned settlements of the coastal plain of Daghestan belonged to sites of the Kura-Araxes culture. Metallurgical inventories related to the Middle Bronze Age, discovered in a majority of the catacomb burials, are not characteristic of classical sites of the Kura-Araxes culture. On the other hand, the Catacomb ceramics found in the upper layers of the Velikent I settlement and at other sites are undoubtedly analogous. These differences need to be explained. The explanation of this situation was discovered in the transformation of the Kura-Araxes culture in its late stage into a whole series of cultures of groups of sites in the Middle Bronze Age. In some of the Velikent catacomb burials and in the upper layers of the Velikent I settlement imported Bedeni ceramics were discovered. These artefacts help explain one more factor (a particular moment) in the creation of the cultural differentiation of sites in the north-eastern Caucasus in the Middle Bronze Age. A similar process happened in Transcaucasia in relation to Post Kura-Araxes sites of the Martkopi-Bedeni-Tsnori type.
7. KURA-ARAXES ‘CULTURE’ IN THE NORTH-EASTERN CAUCASUS
153
It is clear that the fading of the Kura-Araxes culture in the north-eastern Caucasus was a long process and happened differently in the mountains and valleys, as Gadzhiev (1990, 72–3) has noted. In the mountains of Daghestan and southeastern Chechnya, the Ginchi culture followed the Kura-Araxes culture. In the foothills of Daghestan, in the valley of the Sulak river, there was one more culture in the Middle Bronze Age that sharply differed from the Kura-Araxes culture. This separate culture is called the Prisulakskaya culture. In the coastal plain of Daghestan the Kura-Araxes culture had not disappeared by the second half of the 3rd millennium BC, at least that is how many of us imagined this period until now, although this idea has undergone certain transformations. One of the most important signs of these changes (the evolution of structures from circular to rectangular, the appearance of ceramic vessels with coated bodies, bronze ornaments with corded design, etc.) can be explained fully within the framework of the existence and development of one archaeological culture, if the separation of archaeological cultures is not tied to strict frameworks of archaeological periodisation, as is stereotypically done. The stratigraphy of sites of the coastal plain of Daghestan, such as Kayakent I, Mamai-Kutan and Velikent I, attests to the fact that there is no break between the occupational layers from the Early and Middle Bronze Age. The analysis of materials from these sites has shown that there was an undoubted succession between the inventory in the lower and upper layers of the settlement. At the same time, ceramics and other products from the late catacomb period (Fig. 3.1–5, 7–11, 14–27; Fig. 4.1–20, 22–36), associated with the Velikent culture of the Middle Bronze Age (Gadzhiev et al. 2000, 87–105, figs. 43–54, 57; Magomedov 2000, 31–59, figs. 9–20), have numerous analogies among the inventory of artefacts from the upper layers at the previously mentioned sites. Moreover, at Velikent, three catacombs, with collective graves and numerous artefacts, were studied in three cemeteries (Gadzhiev 1991, 134–5, 170, fig. 26.59; Gadzhiev et al. 1995, 144–5, pl. XXVc–f ). These three catacombs are very similar to the rest of the group of late catacombs, but they are dated to an earlier period and belong to the Kura-Araxes culture. Besides Kura-Araxes ceramics in these early catacombs, there is also a small amount of bronze ornaments, analogous to the ornaments from the late catacombs. We have briefly described the periodisation and chronology of Kura-Araxes sites in the north-eastern Caucasus. Currently, there are only a few known different outlines of the periodisation and chronology of sites of the Kura-Araxes culture as a whole, but none of these theories is accepted by all. For a long
154
RABADAN MAGOMEDOV
time, the chronological outline proposed by Munchaev (1975, 191–5) was popular. According to this outline, sites of the Kura-Araxes culture dated from the 3rd millennium BC, in general. However, Munchaev has recently changed his opinion on this, it seems, inasmuch as he now recognises that ‘the emerging tendency to make the early type of Kura-Araxes culture older – no later than the second half of the 4th millennium BC, is correct. . . . Evidently, the upper limits of the Kura-Araxes culture should be limited to the time around the middle of the 3rd millennium BC (2500–2400 BC)’ (Munchaev 1994, 17). At the same time, he admitted that he has relied on the well known periodisation of Kushnareva and Chubinishvili (1970, 61–2), in which Kura-Araxes sites are divided into three periods: early (3000–2700/2600 BC); middle (2700/2600–2400/2300 BC) and late (2400/2300–2000 BC). In our opinion, the periodisation of Kushnareva and Chubinishvili should be reconsidered. A variation of the periodisation and chronology of this period, offered by G.L. Kavtaradze (1981, 139–42; 1983, 82), is based on calibrated radiocarbon dates and pushes the Kura-Araxes culture farther into the past, relegating this culture almost entirely to the 4th millennium BC, which has brought about a stream of objections from several other specialists (see, for example, Andreeva 1987, 273–83). The periodisation and chronology of Kura-Araxes sites in the north-eastern Caucasus is naturally always related to the general periodisation of the KuraAraxes culture as a whole. According to Munchaev (1975), the earliest north-eastern Caucasian sites of this culture are located in the coastal plain of Daghestan (of the middle phase of the Kura-Araxes culture during the middle of the 3rd millennium BC). In his opinion, the late phase of the development of the KuraAraxes culture can only be found in sites in the mountain region. According to Munchaev, this kind of ‘zonal’ chronology relates the Kura-Araxes sites in the north-eastern Caucasus to the migration of Kura-Araxes tribes from Transcaucasia to the coastal plain of Daghestan, and from there to the mountains. A principally new system of periodisation for the sites of the Kura-Araxes culture in the north-eastern Caucasus has been put forward by Gadzhiev (1987a, 13–7, figs. 1–2; 1987b, 62–8, tabls. XXII–XXIII; 1990, 66–74, fig. 282; 1991, 127–50, figs. 27–28), dividing them, as Munchaev did, into two stages: early and late. By absolute chronology Gadzhiev proposed a dating for these sites between the last quarter of the 4th and the last quarter of the 3rd millennium BC, as well as a boundary between the two greater stages of periodisation in the middle of the 3rd millennium BC. New research on Kuro-Araxes sites, especially the Velikent complex of sites, has also received a whole series of
7. KURA-ARAXES ‘CULTURE’ IN THE NORTH-EASTERN CAUCASUS
155
radiocarbon dates from the laboratory of the University of Arizona (21 dates for the four Velikent sites and Kabaz-Kutan I – Fig. 5), making it possible for us to bring a few essential corrections to Gadzhiev’s scheme. Based on C14 dates, we propose a general dating for sites of the Velikent Kuro-Araxes culture within the limits of 3500–1900 BC. This new periodisation consists of four phases: 1) 3500–2900 BC; 2) 2900–2500 BC; 3) 2500–2200 BC; and 4) 2200–1900 BC. To the first, earliest phase of the Velikent culture we attribute the settlement sites of Serzhenyurt, Velikent II and the latest Eneolithic materials from the Chinna settlement, and, possibly, Ginchi (?). The lowest level of the Kabaz-Kutan I settlement may be synchronised with the late period of this same phase. The next two phases roughly coincided with Gadzhiev’s periodisation and include within them the basic range of KuroAraxes sites in the region. At roughly the end of the third quarter of the 3rd millennium BC the area of the Velikent Kuro-Araxes culture sharply contracts and continues to do so right up until the first centuries of the 2nd millennium BC, when the Kuro-Araxes culture is recorded only in the territory of the coastal plain of Daghestan. To the fourth phase of our classification we attribute the upper level of lowland coastal sites (Kayakent I, Mamai-Kutan and Velikent I), as well as the later group of Velikent and Manas catacombs. It is not possible to strictly assign the poorly investigated sites in Azerbaijan to the foregoing phases. High quality ceramics occur in a few of these sites (Serkertepe, Dashlytepe), similar to Velikent II, and at Serkertepe there are also painted ceramics. All this allows us to consider that in this part of the area of the Velikent Culture there are sites of a very early time, dating to the end of the 4th–beginning of the 3rd millennium BC. Ceramics are found in many sites in Azerbaijan that are identical to ceramics of the high levels of coastal plain sites in Daghestan. While we do not know whether it is possible to assign a few of these sites to the later phase of the Velikent culture, perhaps in the future with the excavation of cemetery sites in north-eastern Azerbaijan, as well as the receipt of a series of C14 dates, it will become possible to chronologically differentiate the sites in Azerbaijan. Thus, we identify Kuro-Araxes sites of the north-eastern Caucasus with the Velikent culture within the limits of a single Kuro-Araxes cultural-historical community, with a general dating from the middle of the 4th millennium BC to the beginning of the 2nd millennium BC. The Velikent culture is distinguished from the other Kuro-Araxes cultures by a series of traits, among which the most important appear to be: 1) stone building structures – in the mountains, unfinished,
Fig. 5. C14 Dates of sites in Daghestan.
156 RABADAN MAGOMEDOV
7. KURA-ARAXES ‘CULTURE’ IN THE NORTH-EASTERN CAUCASUS
157
framed-wattled constructions, in lowlands and semi-lowlands, stone structures – on balance; the general line of the development of architecture is from a circular plan to a rectangular construction; 2) present in the central part of the dwellings sunk into the floor are circular hearths (tondyri) without internal projections or ‘petals’; 3) the collective character of burials in crypts and catacombs; 4) in the mountains, stone crypts of circular or rectangular forms, reflecting developments in building practices; 5) high quality ceramics of the Velikent II type with incised ornamentation, made with the use of a potter’s wheel (characteristic of the first phase); 6) typical forms of grey-red-black burnished ceramics (deep bowls and basins, cups, beakers, and vases on footed bases, etc.), rarely found in Transcaucasia; 7) vessels with slipped bodies (characteristic for the end of the second phase); 8) terracotta female figurines of the Kataragach-Tapa type; 9) drilled polished stone axes; 10) double-volute bronze pendants and a whole series of other bronze objects.
Bibliography Akhundov, T.I. 1996: Istoricheskaya topografiya poseleniy i sistema rasseleniya v SeveroVostochnom Azebaidzhane (seredina III – seredina II tysyacheletiya do n.e.). (Avtoreferat dissertatsii) (Leningrad). Andreeva, M.V. 1987: Retsenziya na kn.: Kavtaradze G.L. K khronologii epokhi eneolita i bronzy Gruzii. SA 4, 273–83. Areshyan, G.E. 1982: Sovremennoe sostoyanie izucheniya architektury shengavitskoi archeologicheskoi kul’tury (kuro-arakskaya kul’turnaya obschnost’). In V respublikanskaya nauchnaya konferentsiya po problemam kul’tury i iskusstva Armenii (Yerevan). Chubinishvili, T.N. 1971: K drevnei istorii Yuzhnogo Kavkaza 1 (Tbilisi). Gadzhiev, M.G. 1979: Severo-Vostochnyi Kavkaz i kuro-arakskaya kul’tura. In IX Krupnovskie chteniya po archeologii Severnogo Kavkaza: Tezisy dokladov (Elista), 6–7. ——. 1980: K periodizatsii kuro-arakskoi kul’tury na Severo-Vostochnom Kavkaze. In X ‘Krupnovskie chteniya’ po archeologii Severnogo Kavkaza: Tezisy dokladov (Moscow), 1–2. ——. 1985: Severo-Vostochnyi Kavkaz kak geograficheskaya i etnokul’turnaya oblast. In Drevnie kul’tury Severo-Vostochnogo Kavkaza (Makhachkala), 10–26. ——. 1987a: Razvitie kul’tury Daghestana v epokhu rannego metalla (Voprosy periodizatsii). In Islammagomedov, A.I. (ed.), Etnokul’turnye protsessy v drevnem Daghestane (Makhachkala), 5–21. ——. 1987b: Severo-Vostochnyi Kavkaz v epokhu rannei bronzy. In Kavkaz v sisteme paleometallicheskikh kul’tur Evrazii: Materialy I simpoziuma ‘Kavkaz i Yugo-Vostochnaya Evropa v epokhu rannego metalla’, Telavi-Signakhi – 1983 (Tbilisi), 62–8, tabls. XXII–XXIII. ——. 1990: Severo-Vostochnyi Kavkaz v epokhu rannego metalla (Voprosy kul’turnoi preemstvennosti). In Esayan, S.A. and Areshyan, G.E. (eds.), Mezhdistsiplinarnye issledovaniya kul’turogeneza i etnogeneza Armyanskogo nagor’ya i sopredel’nykh oblastei (Yerevan), 66–74. ——. 1991: Rannezemledel’cheskaya kul’tura Severo-Vostochnogo Kavkaza (Epokha eneolita i rannei bronzy) (Moscow). Gadzhiev, M.G., Davudov, O.M. and Shikhsaidov, A.R. 1996: Istoriya Daghestana s drevneishikh vremen do kontsa XV v. (Makhachkala).
158
RABADAN MAGOMEDOV
Gadzhiev, M.G., Kohl, P.L., Magomedov, R.G. and Stronach, D. 1995: The 1994 Excavations the Daghestan-American Archeological Expedition to the Velikent in Southern Daghestan, Russia. Iran XXXIII, 139–47. ——. 1997: The 1995 Daghestan-American Velikent Expedition: Excavations in Daghestan, Russia. Eurasia Antiqua 3, 181–222. Gadzhiev, M.G., Kohl, P.L., Magomedov, R.G., Stronach, D. and Gadzhiev, S.M. 1997: Daghestan-American Archeological Investigations in Daghestan, Russia 1997–1999. Eurasia Antiqua 6, 47–123. Gadzhiev, M.G. and Korenevsky, S.N. 1984: Metall Velikentskoi katakomby. In Mammaev, M.M. (ed.), Drevnie promysly, remeslo i torgovlya v Daghestane (Makhachkala), 7–27. Ismailov, G.S. 1983: Rannebronzovaya kul’tura Azerbaidzhana (Avtoreferat dissertatsii) (Tbilisi). ——. 1987: Rannebronzovaya kul’tura na territorii Azerbaidzhana. In Kavkaz v sisteme paleometallicheskikh kul’tur Evrazii: Materialy I simpoziuma ‘Kavkaz i Yugo-Vostochnaya Evropa v epokhu rannego metalla’, Telavi-Signakhi – 1983 (Tbilisi), 74–80. Kavtaradze, G. 1983: K khronologii epokhi eneolita i bronzy Gruzii (Moscow). Khalilov, D.A. and Kesamanly, K.O. 1973: Gyafle-tepeleri – poselenie epokhi rannei bronzy v Azerbaidzhane. In Munchaev, R.M. and Markovin, V.I. (eds.), Kavkaz i Yugo-Vostochnaya Evropa v drevnosti (Moscow), 55–9. Khalilov, D.A., Kesamanly, K.O. and Arazova, R.B. 1991: Archeologicheskie pamyatniki SeveroVostochnogo Azerbaidzhana (SAPA 1) (Baku). Kotovich, V.G. 1959: Novye arkheologicheskie pamyatniki Yuzhnogo Daghestana. In MAD I, 121–56. ——. 1982: Problemy kul’turno-istoricheskogo i khozyaistvennogo razvitiya naseleniya drevnego Daghestana (Moscow). Kotovich, V.M. 1985: Kurgan Torpakh-Kala. In Drevnie kul’tury Severo-Vostochnogo Kavkaza (Makhachkala), 59–80. Kruglov, A.P. 1958: Severo-Vostochnyi Kavkaz vo II–I tysyacheletiyakh do n.e. (Moscow/ Leningrad). Krupnov, E.I. 1964: Drevneishaya kul’tura Kavkaza i kavkazskaya etnicheskaya obschnost’ (K probleme proiskhozhdeniya korennykh narodov Kavkaza). SA 1, 26–43. Kuftin, B.A. 1944: Urartskiy ‘kolumbariy’ u podoshvy Ararata i kuro-arakskiy eneolit. VGMG 13–B (Tbilisi), 1–171. ——. 1948: Archeologicheskie raskopki 1947 goda v Tsalkinskom rayone (Tbilisi). Kushnareva, K.K. 1993: Yuzhnyi Kavkaz v IX–II tys. do n.e. Etapy kul’turnogo i sotsial’no-ekonomicheskogo razvitiya (St Petersburg). Kushnareva, K.K. and Chubinishvili, T.N. 1970: Drevnie kul’tury Yuzhnogo Kavkaza (Leningrad). Kushnareva, K.K. and Dzhaparidze, O.M. 1978: Retsenziya na knigu: Munchaev R.M. Kavkaz na zare bronzovogo veka (Moskva 1975). SA 1, 268–79. Lomtatidze, G.A. 1961: Vystuplenie na Zaklyuchitel’nom zasedanii sessii 14 maya 1959 g. MAD II, 293–8. Magomedov, R.G. 1985: K vydeleniyu pamyatnikov velikentkoi gruppy epokhi srednei bronzy. In Vsesoyuznaya archeologicheskaya konferentsiya ‘Dostizheniya sovetskoi archeologii v XI pyatiletke’: Tezisy dokladov (Baku), 221–3. ——. 2000: Materialy k izucheniyu kul’tur epokhi bronzy v Primorskom Daghestane (Makhachkala). Munchaev, R.M. 1955: Kayakentskoe poselenie i problema kavkazskogo eneolita. SA XXII, 6–20. ——. 1961: Drevneishaya kul’tura Severo-Vostochnogo Kavkaza (Moscow). ——. 1975: Kavkaz na zare bronzovogo veka. Neolit, eneolit, rannyaya bronza (Moscow). ——. 1994: Kuro-arakskaya kul’tura. In Archeologiya. Epokha bronzy Kavkaza i Srednei Azii. Rannyaya i srednyaya bronza Kavkaza (Moscow), 8–57. Munchaev, R.M. and Smirnov, K.F. 1956: Pamyatniki epokhi bronzy v Daghestane. (Kurgannaya gruppa u stantsii Manas). SA XXVI, 167–203. ——. 1958: Archeologicheskie pamyatniki bliz sela Karabudakhkent (Daghestanskaya ASSR). In Piotrovsky, B.B. (ed.), Drevnie plemena i narodnosti Kavkaza (Moscow), 147–75.
7. KURA-ARAXES ‘CULTURE’ IN THE NORTH-EASTERN CAUCASUS
159
Musaev, D.L. 1991: Pamyatniki epokhi rannei bronzy Severo-Vostochnogo Azerbaidzhana (Avtoreferat dissertatsii) (Baku). ——. 1992: Lokal’nye raznovidnosti keramiki kuro-arakskoi kul’tury. In Materialy nauchnoi konferentsii, posvyaschennoi poslednim rezul’tatam archeologicheskoi i etnograficheskoi nauk v Azerbaidzhane (Baku), 114–6. Piotrovsky, B.B. 1949a: Archeologiya Zakavkaz’ya s drevneishikh vremen do I tysyacheletiya do n.e. (Leningrad). ——. 1949b: Poseleniya mednogo veka v Armenii. SA 23, 171–84.
CHAPTER EIGHT THE MIDDLE BRONZE AGE SETTLEMENT PATTERN OF THE EASTERN ANATOLIAN HIGH PLATEAU IN LIGHT OF NEW EVIDENCE AYNUR ÖZFIRAT
The Eastern Anatolian plateau, which lies to the east of the Euphrates river and to the north of the south-east Taurus mountains. Unlike the Early Transcaucasian period (Kura-Araxes), does not have a homogenous culture in the Middle Bronze Age. Two different cultures, which relate to different geographical areas, can be seen in eastern Anatolia in this period. The first area, the Malatya-Elazıª plain, is in the western extreme of eastern Anatolia, consisting of the Euphrates valley and the area known as Altınova, which is now situated in the region of the Keban reservoir. While the evidence shows that the early Middle Bronze Age (MBAI) occupation of this area had local roots, a relationship developed with Northern Mesopotamia during the Assyrian Colony period (Di Nocera 1998; Oguchi 1998; Sevin 1998). From the beginning of the Old Hittite period, this area began to come under the influence of central Anatolia and, a little later, during the Hittite empire, it took on a completely Hittite character. The other geographical and cultural area of eastern Anatolia lies to the east and north-east of a line drawn roughly through Mu¤, Erzurum and Artvin (Çilingiroªlu 1990; 1994; Özfırat 2001a; 2001b; 2001d; forthcoming b; see also Yakar 2000, 410–3) (Fig. 1). This area is characterised by a high plateau, interspersed with the extinct volcanoes of Nemrut, Süphan, Tendürek and Aªrı. This area appears to have had very little influence from Mesopotamia and central Anatolia, but instead seems to have kept closer ties with north-western Iran, Transcaucasia and the Caspian Sea region. This groups of this region, which includes Georgia, Armenia, Azerbaijan, Nakhichevan and Iranian Azerbaijan, experienced major socio-economic changes towards the end of the Early Transcaucasian period (Kura-Araxes) at the end of the 3rd millennium BC, which resulted in a new culture being formed in the Middle Bronze Age (Martirosyan 1964, 47–70; Rubinson 1976; Aliev 1991; Lordkipanidze 1991, 54–70; Abibullaev 1982, 145–78; Edwards 1986;
Fig. 1. Map of Araxes painted ware settlements in eastern Anatolia.
8. THE EASTERN ANATOLIAN HIGH PLATEAU
161
162
AYNUR ÖZFIRAT
Djaparidze 1995; Edens 1995; Khanzadıan 1995, 5–38; Simonyan 1996; Bah¤aliyev 1997, 105–10; Kushnareva 1997, 81–150). Groups during this period changed from living on the plains and agricultural regions to the mountains and high pastures The evidence for this change is as follows: there is an obvious reduction in the number of agricultural settlements; cemeteries appear on the high pastures, which are unrelated to any settlement; a new tradition in pottery styles develops; and finally a new type of burial mound (kurgan) occurs. The pastoralist, semi-nomadic and nomadic lifestyles associated with this culture, which were the norm for this period, were quite different to the agriculturally based settled lifestyle of the Early Transcaucasian period. Throughout the whole region, apart from the mounds at Uzerlik Tepe in Azerbaijan, Kültepe II and ⁄ahtahtı (Gavurkale) in Nakhichevan, Haftavan Tepe and partly at Geoy Tepe in north-western Iran, there are no striking Middle Bronze Age settlements with heavy occupational layers. At the same time, we cannot say that the old settlements on the valley plains were completely abandoned during this time.1 The high plateau of eastern Anatolia is the area exhibiting the least amount of evidence for the Trancaucasian culture in the Middle Bronze Age. This Middle Bronze Age of this region, which was almost entirely unknown until a few years ago, developed in a way that is comparable to the other regions of the Transcaucasian culture. In this article, we are dealing only with the settlement system of this period, which is still open to debate. The eastern Anatolian mounds such as Karagündüz (Sevin et al. 1998, 578; 2000a, 411; 2000b, 859–60; Özfırat 2001a, 83–5), Van Kalesi (Tarhan 1994, 47; Tarhan and Sevin 1994, 852; Özfırat 2001a, 85–6), Tilkitepe (Korfmann 1982, 179–83), Dilkaya, Sos (Sagona et al. 1995: 205–6; 1996a, 153–4; 1996b, 32–3) and even Çelebibaª contain little material dating to the Middle Bronze Age occupation of the region. Only a small amount of Middle to Late Bronze Age pottery was found in them, but not in stratigraphical order, except at Sos. At Karagündüz, the Early Transcaucasian level appeared immediately below the Urartian occupation level. Although mounds such as Gümü¤pınar (Fig. 2), Nurettin, Sarıveli (Fig. 3) and Adaksu are established to the west of Lake Van on the high plains of the Mu¤ and Bitlis regions, the amount of Middle to Late Bronze
1
There are many stratified settlements (mounds) with strata from the first half of the 2nd millennium BC in the dry valleys and on the steppes of Transcaucasia. But the strata are not very thick, indicating that in the Middle Bronze Age the mounds were only used for a short period. See Kushnareva (1997, 207–8).
163
Fig. 2. Gümü¤pınar mound.
8. THE EASTERN ANATOLIAN HIGH PLATEAU
Fig. 3. Sarıveli mound and necropolis.
164 AYNUR ÖZFIRAT
8. THE EASTERN ANATOLIAN HIGH PLATEAU
165
Age pottery collected from them is insignificant (Özfırat 1994, 363; 2000, 194; 2001a, 75–9; 2001b, 31–40; 2001c, 124; 2001d).2 This suggest that the agricultural valleys of the eastern Anatolian high plateau were not heavily inhabited in the Middle Bronze Age.3 However, the widespread cemeteries found on the high pastures imply the existence of a large population (Özfırat 1994; 1999, 1–3; 2000, 194–6; 2001a, 71–83; 2001b, 29–40; 2001c, 124; 2001d; forthcoming a). The main feature of these mountain slopes and high volcanic terrain, which lie roughly in the range of 1800–2500 m above sea level, is their fertile pastures. Besides the cemeteries, another interesting feature established on these high pastures are settlements surrounded by defensive structures such as stone walls. For example, Çaygeldi (Fig. 4), Yılankalesi (Fig. 5), Haydarkalesi (Fig. 6) on the slopes of Mount Süphan to the west of Lake Van, and Eski Norgüh on the Gürpınar plain are all walled settlements (Özfırat 1999, 2–3; 2000, 195; 2001a, 87–9; 2001b, 37–42; 2001d).4 However, these surface finds alone do not indicate that these settlements were heavily inhabited during the Iron Age, or the period in which the defensive walls were established. It appears that in the extended period from the beginning of the Middle Bronze Age up until into the Early Iron Age, the nomadic population slowly began to gather itself around certain organisational and economic centres (Mikaelyan 1968; Mnatsakanyan 1967; Kleiss and Kroll 1980; Badaljan et al. 1992; 1993; 1994; Smith and Kafadarian 1996; Smith 1998; 1999; Avetisyan et al. 2000; Schachner 2001, 298–320; Sevin 2004). It is possible that these primitive looking fortresses in the Van basin belong to the Late Bronze-Early Iron Age transition. The cause of the reduction in the number of settlements and the change to a pastoral lifestyle, which is seen throughout the region towards the end of the 2
The other sites in Fig. 1 (Soªkom, Erentepe, Mollakendi and Hino) were found in Burney’s (French and Summers 1994, 79; Özfırat 2001a, 134, n. 288) and Rothmann’s (1995, 285–6) surveys, but there are only one or two undecorated potsherds in these mounds. 3 There is also evidence for a break in the settled village lifestyle, based on agriculture, at the beginning of the Middle Bronze Age, from the western end of eastern Anatolia and the Erzurum region. For example, at the end of the Early Bronze Age, Karaz and Güzelova in Erzurum, Deªirmentepe and Pulur-Sakyol in the Keban region, and Köflkerbaba and Imamoªlu in the Malatya region, were all more or less abandoned. The painted sherds identified as ‘Transcaucasian ware’ from building Level 9 (Old Hittite) at the mound of Imikuflaªı, on the Euphrates near Malatya (Sevin and Derin 1987, 183–4) and in the Arslantepe EB3B stratum (Palmieri and Frangipane 1990, 192) must be imports. 4 We also found a new fortress and cemetery in Iªdır province (Gre Her¤e, see Fig. 1).
Fig. 4. Çaygeldi fortress and necropolis.
166 AYNUR ÖZFIRAT
8. THE EASTERN ANATOLIAN HIGH PLATEAU
Fig. 5. Yılankalesi fortress.
167
Fig. 6. Haydarkalesi fortress.
168 AYNUR ÖZFIRAT
8. THE EASTERN ANATOLIAN HIGH PLATEAU
169
3rd millennium BC, has naturally been a subject of much debate. The two reasons that have been proposed for this shift are changes in environmental-climatic conditions and the political situation at this time. The most acceptable reason for this shift in settlement and economy seems to be that an immigrant population was assimilated into the native population of the Middle Bronze Age culture at the end of the 3rd millennium BC. It is at least clear that in southern Transcaucasia in this period a new ethno-cultural group and lifestyle appeared, which is primarily known by the kurgans these people left behind. A second ethnic group (indicated by painted ware) which does not seem to be related to the culture represented by the kurgans, has far more problematic roots. The system of settlement on the high plateaus of eastern Anatolia appears to have continued into the Early Iron Age, but with a different monochrome pottery tradition. According to the results of excavations at the cemeteries of Karagündüz (Sevin and Kavaklı 1996), Ernis (Sevin 1996), Hakkari (Sevin et al. 2001; Sevin and Özfırat 2001) and from our survey, there were no major changes in the socio-economic system based on pastoralism and transhumance until the time of the Urartian kingdom.
Acknowledgments Our surveys in the eastern Anatolian high plateau have been carried out with the permission of the General Directorate for Monuments and Museums. I wish to express my appreciation to them for granting me permission to undertake the surveys since 1995. I wish also to thank Yüzüncü Yıl University of Van for its support, and I am especially grateful to Veli Sevin for providing considerable and valuable assistance. My thanks also go to Margareth Payne for the translation of the Turkish text into English.
Bibliography Abibullaev, O.A. 1982: Eneolit i Bronza na Territorii Nahıçhevanskoii ASSR (Baku). Aliev, V.G. 1991: Kul’tura Epokhi Sredneii Bronzi Azerbaiidjana (Baku). Avetisyan, P.S., Badalyan, R.S. and Smith, A.T. 2000: Preliminary Report on the 1998 Archaeological Investigations of Project ArAGATS in the Tsakahovit Plain, Armenia. Studi Micenei ed Egeo-Anatolici 42.1, 19–59. Badaljan [Badalyan], R.S., Edens, C., Kohl, P.L. and Tonikjan, V.A. 1992: Archaeological Investigations at Horom in the Shirak Plain of Northwestern Armenia. Iran 30, 31–48.
170
AYNUR ÖZFIRAT
Badaljan, R.S., Edens, C., Gorny, R., Kohl, P.L., Stronach, D., Tonikjan, V.A., Hamayakjan, S., Mandrikjan, S. and Zardarjan, M. 1993: Preliminary Report on the 1992 Excavations at Horom, Armenia. Iran 31, 1–24. Badaljan, R.S., Kohl, P.L., Stronach, D. and Tonikjan, V.A. 1994: Preliminary Report on the 1993 Excavations at Horom, Armenia. Iran 32, 1–29. Bah¤aliyev, V.B. 1997: The Archaeology of Nakhichevan (Istanbul). Çilingiroªlu, A. 1990: Van ve Urmiye bölgeleri arasındaki kültürel ili¤kinin Van-Urmiye boyalıları ı¤ıªında deªerlendirilmesi. X. Türk Tarih Kongresi, 169–73. ——. 1994: Van Gölü havzasında üretilen M.Ö. II. binyıl boyalı çanak çömleªi. Arkeoloji Dergisi 2, 1–31. Di Nocera, M.G. 1998: Die Siedlung der Mittelbronzezeit von Arslantepe. Eine Zentralsiedlung von Beginn des zweiten Jahrtausends v. Chr. in der Ebene von Malatya (Türkei) (Risultati delle ricerche e scavi della Missione Archaeologica Italiana nell’Anatolie Orientale VIII) (Visceglia/ Rome). Djaparidze, O.M. 1995: Mittlere Bronzezeit. In Miron, A. and Orthmann, W. (eds.), Unter Wegs zum Goldenen Vlies: Archaeologische Funde aus Georgien (Stuttgart), 79–94. Edens, C. 1995: Transcaucasia at the End of the Early Bronze Age. BASOR 299–300, 53–64. Edwards, M.R. 1986: Urmia Ware and its distribution in North-Western Iran. Iran 24, 57–77. French, D. and Summers, J. 1994: Pre-Urartian and Urartian Pottery From the Mu¤ Region. Anatolian Studies 44, 77–84. Khanzadıan, E.V. 1995: Metsamor 2, La Necropole: Les Tombes du Bronze Moyen et Recent 1 (Académie Nationale des Sciences d’Arménie, Civilisations du Proche-Orient Hors Série 1) (Neuchâtel/Paris). Kleiss, W. and. Kroll, S. 1980: Die Burgen von Libliuni (Seqindel). Archäologische Mitteilungen aus Iran 13, 21–61. Korfmann, M. 1982: Tilkitepe: Die ersten Ansätze prähistorischer Forschung in der ostlichen Turkei (Istanbuler Mitteilungen Beiheft 26) (Tübingen). Kushnareva, K.K. 1997: The Southern Caucasus in Prehistory: Stages of Cultural and Socioeconomic Development from the Eighth to the Second Millennium BC (Philadelphia). Lordkipanidze, O. 1991: Archäologie in Georgien von der Altsteinzeit zum Mittelalter. (Quellen und Forschungen zur prähistorischen und provinzial römischen Archaeologie 5) (Weinheim). Martirosyan, A.A. 1964: Armeniya v Epokhi Bronzı i Rannego Zheleza (Yerevan). Mikaelyan, G.A. 1968: Ziklopitscheskie Kreposti Sevanskogo Basseina (Yerevan). Mnatsakanyan, A.O. 1967: Treasures From an Armenian Lake. The Illustrated London News 6663 (15 April), 27–9. Oguchi, H. 1998: Notes on Khabur Ware from Sıtes Outside Its Main Distribution Zone. AlRafidan 19, 119–33. Özfırat, A. 1994: M.Ö. II. Binyıl Doªu Anadolu Boyalı Seramik Kültürleri Üzerine Ara¤tırmalar. XI. Ara¤tırma Sonuçları Toplantısı, 359–77. ——. 1999: 1997 Yılı Bitlis-Mu¤ Yüzey Ara¤tırması: Tunç ve Demir Çaªlar. XVI. Ara¤tırma Sonuçları Toplantısı II, 1–22. ——. 2000: 1998 Yılı Bitlis-Mu¤ Illeri Yüzey Ara¤tırması: Tunç ve Demir Çaªları. XVII. Ara¤tırma Sonuçları Toplantısı 2, 193–210. ——. 2001a: Doªu Anadolu Yayla Kültürleri (M.Ö. II. Binyıl) (Istanbul) (with summary in English). ——. 2001b: Erste Betrachtungen zum ostanatolischen Hochland im 2. Jt. v. Chr. Istanbuler Mitteilungen 51, 27–60. ——. 2001c: 1999 Yılı Mu¤ Ili Yüzey Ara¤tırması: Tunç ve Demir Çaªları. XVIII. Ara¤tırma Sonuçları Toplantısı 2, 123–40. ——. 2001d: Research on the Cultures of East Anatolia in the 2nd Millennium BC. In Belli, O. (ed.), Istanbul University’s Contributions to Archaeology in Turkey (1932–2000) (Istanbul), 326–30. ——. forthcoming a: Van-Urmia Painted Pottery from Hakkari. Archäologische Mitteilungen aus Iran und Turan 34.
8. THE EASTERN ANATOLIAN HIGH PLATEAU
171
——. forthcoming b: New Investigations on Pastoral Life of the Eastern Anatolian High Plateau during the second millennium BC. Proceedings of the International Congress on the Archaeology of the Ancient Near East (Paris 15–19 Nisan 2002). Palmieri, A. and Frangipane, M. 1990: The 1988 Campaign at Arslantepe, Malatya. XI. Kazı Sonuçları Toplantısı 1, 191–201. Rothmann, M.S. 1995: The Pottery of The Mu¤ Plain And The Evolving Place of a High Border Land. XII. Ara¤tırma Sonuçları Toplantısı, 281–304. Rubinson, K.S. 1976: The Trialeti Culture (University Microfilms) (Ann Arbor). Sagona, A.G., Sagona, C. and Özkorucuklu, H. 1995: Excavations at Sos Höyük 1995: First Preliminary Report. Anatolian Studies 45, 193–218. ——. 1996a: Excavations at Sos Höyük 1994. XVII. Kazı Sonuçları Toplantısı 1, 151–5. Sagona, A.G., Erkmen, M., Sagona, C. and Thomas, I. 1996b: Excavations at Sos Höyük 1995: Second Preliminary Report. Anatolian Studies 46, 27–52. Schachner, A. 2001: Azerbaycan: Eine terra incognita der Vorderasiatischen Archäologie. Mitteilungen der Deutschen Orient-Gesellschaft zu Berlin 133, 251–332. Sevin, V. 1996: Van/Ernis (Ünseli) Nekropolü Erken Demir Çaª Çanak Çömlekleri. Jahbruch für Kleinasiatische Forschung 14, 439–67. ——. 1998: Imiku¤aªı Kazılarının I¤ıªında Habur Seramiªinin Kuzey Yayılımı. XXXIV ème Rencontre Assyrıologique Internationale, 383–91. ——. 2004: Pastoral Tribes and Early Settlements of the Van Region, Eastern Anatolia. In Sagona, A.G. (ed.), A View From the Highlands: Archaeological Studies in Honour of Charles Burney (Ancient Near Eastern Studies Suppl. 12) (Herent), 179–203. Sevin, V. and Derin, Z. 1987: Imiku¤aªı Kazıları, 1985. VIII. Kazı Sonuçları Toplantısı, 181–204. Sevin, V. and Kavaklı, E. 1996: Van/Karagündüz: An Early Iron Age Cemetery (Istanbul). Sevin, V. and Özfırat, A. 2001: Die Stelen aus Hakkari: Steppennomaden in Vorderasien. Istanbuler Mitteilungen 51, 11–26. Sevin, V., Kavaklı, E and. Özfırat, A. 1998: Karagündüz Höyüªü and Nekropolü 1995–1996 Yılı Kurtarma Kazıları. XIX. Kazı Sonuçları Toplantısı 1, 571–89. Sevin, V., Özfırat, A. and Kavaklı, E. 2000a: Karagündüz Höyüªü 1998 Yılı Kurtarma Kazıları. XXI. Kazı Sonuçları Toplantısı 1, 409–20. Sevin, V., Kavaklı, E and. Özfırat, A. 2000b: Van-Karagündüz Höyüªü Kazıları (1997 Yılı Çalıªmaları). Belleten 43.238, 847–81. Sevin, V., Özfırat, A. and Kavaklı, E. 2001: 1997–1999 Hakkari Kazıları. XXII. Kazı Sonuçları Toplantısı 1, 355–68. Simonyan, A.E. 1996: Pasteurs et chefs de guerre au Bronze Moyen (XXIII–XXII Siècle Avant J.C.). In Santrot, J. (ed.), Arménie: Tresors de l’Arménie ancienne, des origines au I er siècle (Paris), 54–69. Smith, A.T. 1998: Late Bronze/Early Iron Age Fortresses of the Ararat and Shirak Plains, Armenia: Typological Considerations. Ancient Civilizations from Scythia to Siberia 5.2, 73–97. ——. 1999: The Making of an Urartian Landscape in Southern Transcaucasia: A Study of Political Architectonics. American Journal of Archaeology 103.1, 45–71. Smith, A.T. and Kafadarian, K. 1996: New Plans of Early Iron Age and Urartian Fortresses in Armenia: A Preliminary Report of the Ancient Landscapes Project. Iran 34, 23–37. Tarhan, M.T. 1994: Recent Research at the Urartian Capital Tushpa. Tel Aviv 21.1, 22–57. Tarhan, M.T. and Sevin, V. 1994: Van Kalesi ve Eski Van ⁄ehri Kazıları 1990 Yılı Çalı¤maları. Belleten 57.220, 843–61. Yakar, J. 2000: Ethnoarchaeology of Anatolia: Rural Socio-economy in the Bronze and Iron Ages (Tel Aviv University, Sonia and Marco Nadler Institute of Archaeology, Monograph Series 17) (Jerusalem).
CHAPTER NINE BEFORE ARGISHTI: THE ROOTS OF COMPLEX SOCIETIES IN CAUCASIA. NOTES FROM THE TSAKAHOVIT PLAIN, ARMENIA ADAM T. SMITH
Since 1998, the joint Armenian-American archaeological project ArAGATS (Archaeology and Geography of Ancient Transcaucasian Societies) has conducted a multi-year regional study of the emergence and reformation of the socio-political landscape of the Tsakahovit Plain, a small, elevated plateau in western Armenia (Avetisyan et al. 2000; Badalyan et al. 2003; Smith et al. 1999; 2005). Our investigations had three primary goals: 1. to document the shifting patterns of landscape production in the region from the earliest times to the Soviet era. 2. to attend closely to the regional organisation of political life before Argishti. That is, during the Late Bronze and Early Iron Ages of the late 2nd and early 1st millennia BC, the eras that preceded the invasion of the region by the kings of Urartu. 3. to examine the sociological outlines of 2nd millennium BC complex societies in order to theorise both the organisation of early south Caucasian polities and their historical connection to better known successor kingdoms, such as Urartu. In this short paper I would like to focus on the latter two goals by providing an overview of the unexpectedly precocious Late Bronze Age political communities that we have documented in central Armenia. In doing so, I am in effect speaking against two ingrained intellectual traditions in the archaeology of the Caucasus (for an extended discussion, see Smith and Thompson 2004). The first is an historical tendency to invest explanation in the diffusion of cultural traits from Mesopotamia and other centres of ‘high civilisation’. The results of our study suggest that historical transformations in early Caucasia will never be intelligible in reference to its southern neighbour but rather must be teased out of a serious effort to define local social practices. The second tradition is an anthropological one that focuses an understanding of the social in the
9. THE ROOTS OF COMPLEX SOCIETIES IN CAUCASIA
173
identification of the ‘ethnos’. Our investigations have attempted to shift the interpretive burden away from essentialised visions of enduring cultural formations and towards an account of the social and political institutions that actively produce distinct visions of community. Let me begin by providing some geographic and temporal orientation to the archaeology of the southern Caucasian Late Bronze Age (Fig. 1).
The Southern Caucasian Late Bronze Age The beginning of the Middle Bronze Age in southern Caucasia was marked by the disappearance of the Kura-Araxes horizon and the large-scale abandonment of settled village communities, such as that recorded at the site of Shengavit near Yerevan. As a result, the vast majority of the archaeological record for the Middle Bronze Age comes from mortuary sites (see, for example, Avetisyan et al. 2000; Gogadze 1972; Kuftin 1940; Kushnareva 1993; 1997; Puturidze 2003; Rubinson 1977). The rich inventory of the enormous kurgans of the Middle Bronze Age signify a profound departure in social relations indicated by the collective burials of the Kura-Araxes phase (such as those documented by Kohl and Gadzhiev at Velikent: Gadzhiev et al. 1997). Even more dramatic expressions of this inequality are visible in the Middle Bronze II period when a great part of highland Caucasia was enveloped in the so called Trialeti-Vanadzor (Kirovakan) horizon, which, despite local territorial and chronologically distinct groupings, still presents burial complexes of unprecedented wealth (cf. Devedzhian 1981; Gogadze 1970). The monumental construction and rich mortuary goods of tombs from Vanadzor, Karashamb and Lori Berd, as well as the iconography of elite privilege portrayed on the metal vessels from Karashamb and Korukh Tash, testify to profound changes in the social orders of southern Caucasia (Oganesian 1992). There remains considerable disagreement over the causes and extent of the Middle Bronze Age social transformations. What is most critical for this discussion is the observation that Middle Bronze Age horizons in southern Caucasia indicate an increasing degree of social inequality, predicated upon an elite culture of martial heroics (Smith 2001). Yet there are few indications at present that this social stratification was formalised into an institutional matrix during the early 2nd millennium BC. It is not until the Late Bronze Age that we see the first clear evidence for territorial polities centred in a crystallised set
174
ADAM T. SMITH
Fig. 1. Elevation map of Caucasia.
9. THE ROOTS OF COMPLEX SOCIETIES IN CAUCASIA
175
of enduring institutions. The Late Bronze Age is marked most conspicuously by the reappearance of numerous permanent settlements in the form of variably sized stone-masonry fortresses built atop hills and outcrops (Fig. 2). With the opening of the Late Bronze Age, the social inequalities visible in the kurgans of the early 2nd millennium appear to have been formalised into a tightly integrated socio-political apparatus where critical controls over resources – economic, social, sacred – were concentrated within the cyclopean stone masonry walls of powerful new centres. It is this fluorescence of socio-political complexity 700 years before Argishti’s forces crossed the Araxes that has provided the central focus for our investigations on the Tsakahovit plain of western Armenia.
Late Bronze Age Politics on The Tsakahovit Plain The Tsakahovit plain, 10–12 km across at its maximum extent, lies between the northern alpine slopes of Mount Aragats (4090 m) and the dry, rugged hills of the Pambakh range (Fig. 3). During the summers of 1998 and 2000, our investigations centred on a systematic archaeological survey of approximately 85 km2 of the highlands surrounding the plain (to our knowledge the first intensive walking survey conducted in the Caucasus). Test excavations at five fortresses – Hnaberd, Tsakahovit, Gegharot, Tsilkar and Poloz-Sar – and four cemeteries – Tsakahovit 1, Hnaberd 14 and 18, Mantash 8 – complemented the results of the survey, providing depth to the sequences detected on the surface. During the 2002 field season, our research turned toward more localised investigations as we began extensive excavations at the fortresses of Tsakahovit and Gegharot. Both survey and excavations indicate that the Tsakahovit plain was intensively and extensively occupied during the Late Bronze Age. Surface ceramics and the morphology of fortification architecture suggest that five small outpost fortresses were built during the Late Bronze Age. Excavations at the three largest fortresses in the region – Gegharot, Tsakahovit and Hnaberd – confirm this assignment of the earliest period of fortress-centred socio-political communities to the Late Bronze Age. Calibrated AMS dates from these three fortified sites confirm the ceramic sequences, with Late Bronze Age occupations beginning at Gegharot in the early 15th century BC and ending at Tsakahovit sometime between the late 14th and early 12th centuries BC. Our systematic survey was designed to provide a sense of the Late Bronze Age landscape that lay beyond the restricted confines of the large fortresses that
Fig. 2. Late Bronze and Early Iron Age fortified sites: a comparison of the topography and architecture of three sites.
176 ADAM T. SMITH
Fig. 3. An archaeological map of the Tsakahovit Plain, Republic of Armenia.
9. THE ROOTS OF COMPLEX SOCIETIES IN CAUCASIA
177
178
ADAM T. SMITH
had long been thought to exhaust the archaeological record of the Late Bronze Age. To this end, we walked linear transects across four major survey zones between fortified centres: the north slope of Mount Aragats, the south-west foothills of the Pambakh range, Mount Kolgat and Mount Vardablur. In the course of this survey we discovered several new fortified settlements, including a cluster of small outposts in the Pambakh slopes (Tsilkar, Ashot-Yerkat, and Poloz-Sar) and two fortresses on the Aragats slope (Gekhadzor and Sahakaberd). As we have reported elsewhere (Avetisyan et al. 2000), the remains encountered in the hinterland beyond the fortresses were primarily architectural, including small settlements, extensive irrigation networks, carved stone markers, and cemeteries. Our survey recorded 193 discrete cemeteries composed of cromlechs typical of Late Bronze Age mortuary architecture on the slopes of Mount Aragats and the Pambakh range, a density of 4.6 cemeteries per km2. The mountain slopes surrounding the Tsakahovit plain appear to have been transformed into a vast necropolis contemporaneous with the emergence of complex polities in the region. Most interestingly, the crisp boundaries of cemetery distribution on the slope of Mount Aragats suggest the possibility that Late Bronze Age polities in the region were not small fortress-states limited to a single settlement administering its immediate hinterland. Rather, these polities seem to have flourished as multi-centred political entities. The appearance of small fortified outposts in the hinterlands beyond the large centres provides an indication of the ability of these early southern Caucasian polities to project authority into the hinterlands. The series of outposts discovered in the Pambakh foothills above Gegharot suggest not only a polity able to police well-defined territorial boundaries, but also one concerned to closely monitor the local population as well. If the pivotal transformation from the Middle to Late Bronze Age socio-political communities lies in the increasing formalisation of social inequalities that emerged upon the collapse of the Kura-Araxes horizon, then what were the sources of elite power and the structure of governmental institutions? Excavations in the cemetery complexes of the Tsakahovit plain and elsewhere suggest that during the Late Bronze Age, the ostentatious displays of social inequality known from the grand Middle Bronze Age kurgans moderate considerably. This is not to suggest that social inequality ameliorated. Quite the contrary, the demands of rulers quartered in fortified settlements likely intensified the social distance between elites and subjects. However, the expression of this distance in mortu-
9. THE ROOTS OF COMPLEX SOCIETIES IN CAUCASIA
179
ary contexts took on less outwardly demonstrative forms, suggesting that the legitimacy of rule no longer rested as heavily on conspicuous displays of wealth but rather had been regularised into an enduring institutional order. While the institutional order of Late Bronze Age regimes on the Tsakahovit plain is as yet not well assayed – in preparation – exploratory soundings at Tsakahovit fortress have provided hints regarding the overall outlines of preUrartian political institutions. Soundings excavated across the site encountered a thick burned level (complete with the remains of burned beams) capping a well-preserved occupation level. In the west terrace excavations we uncovered several large storage jars (pithoi). Indeed, the range of wares recovered within and below the burnt level was tightly restricted to large jars and pithoi. Two complimentary soundings on the citadel excavated in 1998 also uncovered substantial evidence of a burnt layer capping the Late Bronze Age occupation, although the destruction level was not as thick or consistent as on the terrace. In contrast to the terrace operation, the Late Bronze Age levels on the citadel produced fewer large storage jars compared to a robust assemblage of fine wares and bowls. My colleagues and I have suggested, albeit provisionally, that the division between a set of storage vessels on the terrace and more delicate wares on the citadel indicates a basic segregation of functional areas at the site – a spatial differentiation of functions that echoes the results of investigations at other Late Bronze Age sites such as Horom South and Metsamor (Smith et al. forthcoming). Our excavations in 2002 have moved us closer to fleshing out this model as we appear to have uncovered the remains of a large institutional structure (i.e. monumental ‘public’ architecture).
Discussion Overall, the preliminary findings of the archaeological investigations of Project ArAGATS in the Tsakahovit plain point to the early fluorescence of fortressbased polities in southern Caucasia whose authority was predicated upon both the projection of governmental power into the hinterlands via fortified outposts and irrigation facilities and the reproduction of the state in reference to a martial political culture first visible in the tombs of the Middle Bronze Age. By developing an account of the sociological organisation of these polities, it is also possible to theorise the historical interconnections between these early fortress-based
180
ADAM T. SMITH
communities and later expansive polities, such as Urartu (cf. discussion in Burney 2002). We can outline the fundamentals of what has recently been termed an ‘early southern Caucasian political tradition’ in reference to the following key dimensions. • A multi-centred, if not multi-regional polity with its centres of power located within elevated plains. • Polities were assembled and held together in significant part by a political culture predicated upon violence and the military heroics of the ruler. • Political centres were constructed along the margins of these plains rather than at their centre, effectively surrounding the area of large-scale arable land within an effective apparatus of surveillance. • Multiple fortified centres in a single area created an integrated network of coordinating sites of governmental authority. • Political authority was exercised from fortified centres which drew together the dominant political institutions into a tightly integrated, coherent built environment. • Political control over subject populations and economic production was thrust outwards from the fortress centres through a broader re-ordering of the landscape focused on the construction of large-scale irrigation networks and military outposts. What has come to intrigue us are the numerous points of overlap that we can define through a similar thumbnail sketch of Urartian political authority (discussed in further detail in Smith and Thompson 2004). In brief, we can describe the fundamentals of Urartian political authority in reference to the following key elements • An archipelagic polity with its centres of authority located within elevated plains. • The polity was assembled and held together in significant part by a political culture predicated upon violence and the military heroics of the king. • Political centres were constructed along the margins of these plains rather than at their centre, effectively surrounding the area of large-scale arable land within an effective apparatus of surveillance. • Multiple fortified centres in a single area created an integrated network of coordinating sites of governmental authority.
9. THE ROOTS OF COMPLEX SOCIETIES IN CAUCASIA
181
• Political authority was exercised from fortified centres which drew together political, economic, and religious institutions into a tightly integrated, coherent built environment. This is not to suggest that the origins of the Urartian state apparatus itself lie in southern Caucasia but rather that we can see in the Late Bronze Age polities of the region the emergence of a set of institutional forms and strategies for rule that reverberated into the Iron Age and profoundly shaped Urartian political life. Despite these highly suggestive formal parallels, a great deal of work remains to be done in defining Caucasia’s socio-political traditions before Argishti. Future investigations in the region must work to re-centre archaeological research on an understanding of the scales of analysis and interpretation that emerge out of an understanding of the sets of practices that interweave the local and the global. In doing so, archaeology in the Caucasus can not only present us with a profoundly interesting new historical tradition, but can also provide an anthropological lens through which to view it.
Bibliography Avetisyan, P.S., Badalyan, R.S., Gevorkyan, A. and Khnkikyan, O. 2000: The 1998 Excavation Campaign at the Middle Bronze Age Necropolis of Sisian, Armenia. Studi Micenei ed EgeoAnatolici 42.2, 161–73. Avetisyan, P.S., Badalyan, R.S. and Smith, A.T. 2000: Preliminary Report on the 1998 Archaeological Investigations of Project ArAGATS in the Tsakahovit Plain, Armenia. Studi Micenei ed Egeo-Anatolici 42.1, 19–59. Badalyan, R.S., Smith, A.T. and Avetisyan, P.S. 2003: The Emergence of Socio-Political Complexity in Southern Caucasia. In Smith, A.T. and Rubinson, K.S. (eds.), Archaeology in the Borderlands: Investigations in Caucasia and Beyond (Los Angeles), 144–66. Burney, C.A. 2002: Urartu and Its Forerunners: Eastern Anatolia and Trans-Caucasia in the Second and Early First Millennia BC. Ancient East and West 1.1, 51–4. Devedzhian, S.G. 1981: Lori-Berd I (Yerevan). Gadzhiev, M., Kohl, P., Magomedov, R. and Stronach, D. 1997: The 1995 Daghestan-American Velikent Expedition: Excavations in Daghestan, Russia. Eurasia Antiqua 3, 1–40. Gogadze, E.M. 1970: K voprosu o genezise i periodizatsii kurgannoy kul’tury Trialeti. Voprosy drevney istorii (Kavkasko-Blizhnevostchnyy sbornik) 3, 221–51. ——. 1972: Periodizatsiya i genezis kurgannoy kul’tury Trialeti (Tbilisi). Kuftin, B.A. 1940: K voprosu o rannykh stadiyakh bronzovoy kul’tury na territorii Kavkaza. Kratkiye soobshcheniya o dokladakh i polevykh issledovaniyakh Instituta Istorii 8, 5–35. Kushnareva, K.K. 1993: Iuzhnyi Kavkaz v IX–II tis. do n.e.: Etapy kul’turnogo i sotsialno-ekonomicheskogo razvitiya (St Petersburg). ——. 1997: The Southern Caucasus in Prehistory: Stages of Cultural and Socioeconomic Development from the Eighth to the Second Millennium BC (Philadelphia). Oganesian, V.E. 1992: A Silver Goblet from Karashamb. Soviet Anthropology and Archeology (New York) 30.4, 84–102.
182
ADAM T. SMITH
Puturidze, M. 2003: Social and Economic Shifts in the South Caucasian Middle Bronze Age. In Smith, A.T. and Rubinson, K.S. (eds.), Archaeology in the Borderlands: Investigations in Caucasia and Beyond (Los Angeles), 111–27. Rubinson, K S. 1977: The Chronology of the Middle Bronze Age Kurgans at Trialeti. In Levine, D. and Young, T.C.J. (eds.), Mountains and Lowlands: Essays in the Archaeology of Greater Mesopotamia (Malibu, CA), 235–50. Smith, A.T. 2001: The Limitations of Doxa: Agency and Subjectivity from an Archaeological Point of View. Journal of Social Archaeology 1.2, 155–71. Smith, A.T., Badaljan [Badalyan], R.S and Avetissian [Avetisyan], P.S. 1999: The Crucible of Complexity. Discovering Archaeology 1.2, 48–55. Smith, A.T., Badalyan, R.S. and Avetisyan, P.S. 2005: Southern Caucasia During the Late Bronze Age: An Interim Report on the Regional Investigations of Project ArAGATS in Western Armenia. In Çilingiroglu, A. (ed.), Anatolian Iron Ages: The Proceedings of the Fifth Anatolian Iron Ages Colloquium Held at Van 2001 (Ankara) 175–85. Smith, A.T. and Thompson, T.T. 2004: Urartu and the Southern Caucasian Political Tradition. In Sagona, A.G. (ed.), A View From the Highlands: Archaeological Studies in Honour of Charles Burney (Ancient Near Eastern Studies Suppl. 12) (Herent), 557–80.
CHAPTER TEN SURVEY AND SETTLEMENT IN NORTHERN MONGOLIA: THE STRUCTURE OF INTRA-REGIONAL NOMADIC ORGANISATION WILLIAM HONEYCHURCH AND CHUNAG AMARTUVSHIN
Introduction In A Study of History, Arnold Toynbee (1934, 16) suggests of nomadic civilisation that it ‘. . . is essentially a society without history. Once launched on its annual orbit, the nomadic horde revolves in [that orbit] thereafter and might go on revolving forever.’ Toynbee considers the pastoral nomadic adaptation a life way that is not likely to show change or growth, let alone socio-political development (Lattimore 1962, 242). This conception of nomadic society as having a timeless resistance to political, economic and social development lingers in the historical research on nomadic peoples (Humphrey and Sneath 1999). One region that resists these persistent ideas is the Eurasian steppe, an area which has repeatedly witnessed the emergence of highly organised nomadic confederations and empires such as those of the Scythians, Xiongnu, Turks, Uighurs and imperial Mongols (Sinor 1970). Recent explanations for steppe nomadic polities emphasise the limitations and instability of the nomadic economy (Khazanov 1994), migration and internal conflict caused by the expansion of early neighbouring states (Di Cosmo 1994), and opportunities for political finance through resource appropriation from sedentary societies (Kradin 2000). In this paper we evaluate two comprehensive and compelling models by Thomas Barfield (1989; 2001) and Nicola Di Cosmo (1999; 2002) that address the formation and structure of steppe polities. Using archaeological survey data from Mongolia, we compare these models according to their expectations for change at the local level in relation to the emergence of a regional polity. Barfield originally published his model for the finance and structure of the Xiongnu nomadic polity (ca. 200 BC–AD 100) in 1981 and went on to elaborate his ideas by addressing the rise and fall of Inner Asian political systems in relation to the neighbouring regions of China and Manchuria (Barfield 1989). Barfield argues that nomadic complex polities differ from sedentary ones in being structured primarily by external relations consisting of the direct appropriation
184
WILLIAM HONEYCHURCH AND CHUNAG AMARTUVSHIN
of economic subsidies from neighbouring states. The nomadic polity, accordingly, was not financed by pastoral or other kinds of indigenous production, but by the extraction of resources from adjacent agricultural economies. Barfield describes the political organisation of such an extractive system as ‘simultaneously containing both a tribal and state hierarchy, each with separate functions . . . autocratic and state-like in foreign affairs, but consultative and federally structured internally’. Nomadic polities were characterised by at least three tiers of administrative hierarchy including local leaders, appointed regional governors, and the imperial leader and his court. Barfield (1989, 8) has further argued that at the local level, the tribal structure remained intact under the rule of chieftains whose power was derived from their own people’s support, not from imperial appointment.1 Therefore, the rise of a regional confederation is expected to change little at the local level and articulation between local and regional spheres of authority would have been relatively informal. In contrast, Di Cosmo (1999) places emphasis on endogenous processes of political development and on the long-term traditions of statecraft upon which aspiring political leaders depended. While recognising the importance of external influences, he suggests that in order to understand the emergence of largescale political integration across the steppe, external modulations must be viewed primarily in the context of existing economic and political relationships. Di Cosmo argues that ‘not only did change occur, but . . . its driving force should be sought exactly in those structural limitations often attributed to pastoral nomadism’ (Di Cosmo 1999, 7–8). The steppe region, therefore, was not a passive recipient stimulated by change in surrounding areas, nor was it dependent on external arenas for political opportunity and finance, but instead generated structural change as the result of indigenous processes that might have been triggered by internal or external events. The mechanisms Di Cosmo identifies as important to the emergence of a steppe polity are: (a) a series of destabilising crises; (b) military build up and centralisation around a charismatic leader; and (c) ideological and political consolidation which effectively ‘. . . reoriented social and political relations from the horizontal to vertical and from semi-egalitarian to hierarchical (Di Cosmo 1999, 20–1). Di Cosmo envisages a restructuring of political, economic, and ideological relationships as the basis for 1
Barfield uses the terms ‘tribe’ and ‘tribal’ not in the sense of evolutionary models but instead as the term is typically used in the ethnographic literature on nomadic groups, referring to an organisation more similar to a chiefdom (see Cribb 1991, 52; Khazanov 1994, 119–22).
10. SURVEY AND SETTLEMENT IN NORTHERN MONGOLIA
185
regional integration. A primary feature of this process is the increasingly structured and contested articulation of the regional and local levels of authority which would likely have produced organisational changes at the local level (Di Cosmo 1999, 21).
Archaeological Survey at Egiin Gol The Egiin Gol Survey (EGS) was designed to evaluate current hypotheses for the development and structure of nomadic confederations using archaeological evidence. Survey has been carried out along the lower Egiin Gol river (N49°27’ E103°28’: Fig. 1) which is a major tributary within the Selenge watershed of northern Mongolia. The lower Egiin Gol meanders along a flood plain varying in width from 0.8 to 1.5 km and receives five tributary streams that form side valleys opening onto the main Egiin Gol river. Elevations within the survey area range between 840 and 1250 m above sea level and the local forest-steppe environment consists of valley grasslands intermixed with higher elevation stands of larch, pine and birch. Annual precipitation and soil quality is sufficient to support extensive dry-farmed agriculture currently while canal-irrigated agriculture was practiced in the area during the 19th century. A full-coverage, pedestrian survey was carried out during a five-year period (1996–2000) and encompassed a 310 km2 area consisting of the main Egiin Gol valley, tributary valleys and several low-lying ridges. In excess of 500 archaeological sites have been recorded dating from the Early Upper Palaeolithic to the mid-20th century. A primary goal of the survey project has been to demonstrate the effectiveness of full coverage, intensive survey methods for the recovery of artefact scatters identified as nomadic habitation sites. Using a variety of survey techniques including close interval walking, shovel testing and a GIS locational model, survey crews have been successful in locating over 70 sherd and sherdlithic scatters. Transects along ridges revealed no sites and while our survey technique may have missed low density and small scatters along the valley bottoms, we obtained enough information to provide patterns of settlement at Egiin Gol during the Bronze and Early Iron Age (BEIA) and the period of the Xiongnu confederation. In conjunction with evidence for ritual, mortuary and special purpose sites, the Egiin Gol habitation data can be used to evaluate Barfield’s and Di Cosmo’s expectations for the local level during the period that the Egiin Gol valley
186
WILLIAM HONEYCHURCH AND CHUNAG AMARTUVSHIN
Fig. 1. The Egiin Gol region of northern Mongolia and the survey area.
10. SURVEY AND SETTLEMENT IN NORTHERN MONGOLIA
187
became articulated within the regional system of the Xiongnu polity. Barfield suggests that local political and economic organisation remains essentially unchanged despite the addition of administrative hierarchy at higher levels of a regional confederation. Limited articulation between the steppe political system and internal production, and between the local and regional levels of organisation are central parts of Barfield’s model since external income from China is thought to have provided both a source of finance and a basis for political cohesion. Accordingly, we expect the Egiin Gol region to show a high degree of organisational continuity over the time periods in question. In particular, hierarchical and functional relationships between sites should be maintained and the largest settlements and the range of site size categories should remain relatively stable. A consistent spatial relationship between site location and subsistence resources is likewise expected if in fact local economic production was not necessary for the finance of the regional organisation. Di Cosmo’s model predicts very much the opposite scenario such that a higher degree of centralisation at the regional level, stronger articulation between the local and regional levels, and internal and external sources of finance suggests that local site and resource relationships may not have been continuous. The location, size and patterning of Egiin Gol survey sites dating to the Bronze and Early Iron Age and the Xiongnu period have been compared in order to discover just how much stability or change occurred in the valley as the Xiongnu polity emerged.
Survey Results: The Bronze and Early Iron Age According to chronologies developed by Mongolian and Russian archaeologists, the BEIA spans the late 2nd through to the mid-1st millennium BC, corresponding to the Karasuk and Tagar periods of the Siberian record (Novgorodova 1989, 236). This period is represented at Egiin Gol by four site types which are common throughout Mongolia and Buriatiya and include khirigsuur complexes, slab burials, habitations and associated special purpose sites. Khirigsuurs consist of a central stone mound surrounded by a circular or rectangular stone enclosure and having small, circular rock heaps as satellites. These satellites often cover a shallow pit containing horse crania, sherds and occasional bronze objects. The khirigsuur mounds are often found in associated groups forming complexes having from two to ten separate features at Egiin Gol. There remains some question as to the function of khirigsuurs and their relationship to
188
WILLIAM HONEYCHURCH AND CHUNAG AMARTUVSHIN
other site types, however most researchers consider them to be ritual monuments with some evidence for mortuary use (Danilov and Konovalov 1988). The latest publications on khirigsuur chronology date their construction period to the end of the 2nd and beginning of the 1st millennium BC, based on stratigraphic and artefactual evidence (Tsybiktarov 1998, 141). A total of 174 sites having 383 individual khirigsuur features have been recorded in the Egiin Gol survey area. Slab burials consist of stone slabs arranged in a rectangular formation, usually oriented east-west, and covering a shallow burial pit (Navaan 1975). In the Egiin Gol valley, these mortuary features appear individually or in small groups of up to four to nine, and very often in proximity to khirigsuurs or khirigsuur complexes. The Egiin Gol slab burials show differentiation by size according to three groups, the largest of which contains burials up to 10 m in length. In addition to human and animal skeletal remains, slab burials contain a wide range of artefacts including bronze, stone, ceramic and bone items which are most commonly found in disrupted contexts (Dikov 1958, 57). Slab burial assemblages have been stylistically dated from the late 2nd to the mid-1st millennium BC and this proposed date range has been supported by several recently published radiocarbon dates (Tsybiktarov 1998, 103–4). A total of 44 sites with 86 slab burials have been discovered at Egiin Gol and excavations have been carried out at several of these sites by Mongolian archaeologists (Amartuvshin and Erdenebat 2000; Erdenebaatar 1997). A single radiocarbon date2 made on fragments of a horse skull excavated from slab burial EGS-7J and having a sigma 2 range of 940–800 calBC falls nicely within Tsybiktarov’s periodisation for the slab burials of Mongolia and Buriatiya. Bronze and Early Iron Age habitation sites are represented by surface and subsurface artefact scatters usually having some microlithic tools and debitage and coarse red or brown ware ceramics used to make vessels with flat or bevelled rims. Vessel decorations include ‘pie-crust’ appliqué and simple punctuates or incisions usually directly below the rim. Such artefact assemblages are similar to those found in Egiin Gol slab burials and have been documented for both slab burials and contemporary habitation sites in Siberia and Mongolia (Semina 1985, 113–21; Lbova and Grechishchev 1995). While ceramic diagnostics provide initial evidence for the chronology of early ceramic and ceramic-lithic scatters, a forthcoming series of radiocarbon dates on samples excavated from 2
Beta-155036: 2710 ±50 BP, 2ó calibrated date range is 940–800 BC at 95% probability.
10. SURVEY AND SETTLEMENT IN NORTHERN MONGOLIA
189
Egiin Gol habitation contexts will test this periodisation. Fifteen artefact scatters with a mean site size of 2459 m2 have been assigned to the BEIA in the Egiin Gol survey area. Additionally, a small number of rock art and earthwork sites have also been discovered by the survey in association with prominent khirigsuur complexes and slab burial groups. Rock art sites include two areas with ochre painted surfaces, a single area with pecked images on a granite face, and one site with a highly eroded ‘deer stone’ stele (see Volkov 1981). A single earthwork site in the form of a rectangular enclosure with eroded earthen walls (50 × 42 m) and having finds of microliths and ceramics was also located in association with a large khirigsuur group. Stylistic elements, site associations and the artefacts from these ‘special purpose’ sites suggest a Bronze and Early Iron Age date. Examining the distributional pattern of BEIA sites (Fig. 2), the focus of activity during this period was clearly centred in tributary river valleys while areas along the Egiin Gol river have comparatively lower densities of sites. Each tributary valley has roughly the same complement of site types. Site clustering shows a consistent pattern characterised by two groups of sites in each side valley, one in the middle to upper elevations and the other at the valley mouth (Fig. 3). This pattern closely resembles that of current seasonal habitation of Egiin Gol herding families who use the upper valley regions for winter and spring campsites and the mouths of tributary valleys for summer and fall campsites (Erdenebaatar 2000). Each surveyed valley thus has a settlement structure similar to that of its neighbours. The distribution of BEIA sites points to a pattern of discrete, small-scale groups inhabiting side valleys, each with sufficient pastoral resources and having uninhabited boundary areas between them, potentially due to spatial buffering. Viewed at a larger scale, however, it is clear that despite some segregation within the lower Egiin Gol, side valley groups likely joined together as a collaborative social unit on occasion. Reconnaissance beyond the survey area up to distances of 20–30 km shows a complete drop-off of highly visible khirigsuur complexes and slab burials with the re-occurrence of these sites in river valleys similar to that of the Egiin Gol, at some distance. Such larger scale patterning indicates that during the BEIA, the Egiin Gol valley groups may have formed a single cluster among a number of clustered groups dispersed over several river valleys in the forest-steppe region. At this point, it is unclear as to the type and amount of interaction carried out between these larger spatial groups. A preliminary assessment of the Egiin Gol patterning is one of nested
190
WILLIAM HONEYCHURCH AND CHUNAG AMARTUVSHIN
Fig. 2. Distribution of BEIA sites at Egiin Gol.
10. SURVEY AND SETTLEMENT IN NORTHERN MONGOLIA
191
Fig. 3. Bivariate density estimation plot of BEIA sites. Nos. 1–4 and heavy lines show tributary valleys (contour lines are quartile contours at 10% intervals).
192
WILLIAM HONEYCHURCH AND CHUNAG AMARTUVSHIN
modular structure with buffer zones that we might expect with classic segmentary organisation. Egiin Gol BEIA site locations suggest a landscape of semisegregated side valley units with an emphasis on proximity to local pastoral resources and having periodic inter-group association at the scale of the lower Egiin Gol valley as a whole.
Survey Results: The Xiongnu Period The name ‘Xiongnu’ [Hsiung-nu] is taken from the Chinese historical records in which it first appears in texts from the 3rd century BC referring to a steppe nomadic group practicing mounted warfare and located to the north of the Chinese Central Plain (Pulleyblank 1983, 450). The Xiongnu polity is historically dated from 209 BC to AD 93 inclusive of several major political and territorial transformations. The association of a specific archaeological culture with the historical Xiongnu polity was first proposed at the end of the 19th century based on the burial excavations of Talko-Gryntsevich in the Zabaikal region of Siberia. In the 1920s, mortuary research conducted by the Kozlov expedition at the Noyon Uul (Noin Ula) site of Central Mongolia and by Sosnovskii in Zabaikal helped to strengthen the hypothesis that a specific set of well-documented burial types could be related to the Xiongnu nomads of the Chinese histories (Konovalov 1975, 17–8; Rudenko 1962, 6). Later excavations at walled sites by Davydova (1995) in Siberia and Perlee (1961, 17–39) in Mongolia recovered artefact assemblages similar to that known from Xiongnu period mortuary sites and began to add a much needed settlement perspective to Xiongnu archaeology. These excavations also confirmed reports in the Han dynasty records of sedentary, agricultural centres in use among the Xiongnu. Additional evidence has accumulated from field research in Mongolia, Buriatiya and Inner Mongolia to link the historical and archaeological records (Wu 1990). An increasing number of samples excavated from burial and settlement contexts belonging to the Xiongnu archaeological culture have provided radiocarbon date ranges inclusive of the historically known period of the Xiongnu polity (discussed below). The C14 analyses are supported by epigraphic, numismatic and stylistic dates from artefact assemblages excavated from mortuary and habitation sites (Konovalov 1976, 173–206; Trever 1932, 13–5). The regional distribution of this archaeological culture approximates the historically attested size and extent of the Xiongnu polity, and a significant degree of social
10. SURVEY AND SETTLEMENT IN NORTHERN MONGOLIA
193
differentiation is evidenced by the mortuary record in accordance with historical reports (Tseveendorzh 1985, 51; Minyaev 1985, 72–3). Finally, specific elements of material culture and mortuary practice described in historical texts have been archaeologically recovered in Mongolia and Buriatiya (Minyaev 2002). Based on these associations between the archaeological and historical records, Egiin Gol sites including cemeteries and settlements, indicate that the valley was integrated into the Xiongnu polity. Egiin Gol, therefore, provides an interesting case study for observing local changes brought about as the result of inclusion in a regionally organised steppe system. Xiongnu habitation sites at Egiin Gol have been identified by scatters of diagnostic ceramics similar to those found from Xiongnu period burials in the valley and also documented at major Xiongnu settlement and mortuary sites across the eastern steppe region (Davydova 1995; Minyaev 1998; Tseveendorzh and Batsaikhan 1994). These sites are distinguished by both coarse brown ware and fine grey ware ceramics, the latter turned on a slow wheel, used to make a variety of jar type vessels, some having perforated bases. Decorations include scraped-and-polished vertical lines, paddling with a thong wrapped paddle, or an incised ‘wave-and-line’ motif usually appearing on the lower neck, shoulder, or upper body of the vessel. Based on these diagnostics, 14 of the Egiin Gol artefact scatters have been assigned to the Xiongnu period and have a mean site size of 12641 m2, a substantial increase over BEIA habitation size. Two radiocarbon determinations have been made on samples excavated from Xiongnu habitation sites at Egiin Gol. At the lowest levels of site EGS 110 carbonised grain samples recovered from a hearth feature provided a sigma 2 date range of 390–185 calBC (95% probability).3 The second C14 analysis was made on carbonised wood excavated from beneath the fragments of a large Xiongnu style jar from a trash pit at site EGS 486. The sigma 2 range for this determination is 170 calBC–calAD 230 (95% probability).4 Both time spans overlap the historical period of the Xiongnu polity and are well within the range of combined radiocarbon determinations for the central Xiongnu cemetery at Egiin Gol (see below). Six Xiongnu cemeteries have been identified in the Egiin Gol valley based on the presence of distinctive stone ring surface features which mark pit burials having individual or multiple interments. The majority of cemeteries contain one 3
Beta-152047, 2220 ±40 BP: EGS 110, carbonised grain (Triticum aestivum; Honeychurch et al. 2001). 4 Beta-152046, 1970 ±80 BP: EGS 486, carbonised wood.
194
WILLIAM HONEYCHURCH AND CHUNAG AMARTUVSHIN
to six burials with the exception of a single large cemetery, central to the lower valley, comprising over 100 burials (EGS 100). From 1992 to 1999 the cemetery was excavated in its entirety and a series of radiocarbon determinations were made by the Egiin Gol survey project and the Mongol-French expedition (Hall, Batsaikhan and Honeychurch 1999; Erdenebaatar et al. 1998). Based on seven C14 determinations5 and using the OxCal calibration program (version 3.5) which is able to sum the probability distributions of calibrated C14 dates, a floruit (Bronk Ramsey 2000) was calculated for the cemetery ranging from 450 BC to AD 250 (Fig. 4). This chronological range includes the historical dates for the Xiongnu confederation and overlaps the radiocarbon determinations made for the Xiongnu habitation sites EGS 110 and 486. In addition, a single special purpose site (EGS 131) having an earthen wall and ditch system (110 × 185 m) has been dated to the Xiongnu period based on radiocarbon analysis and diagnostic artefacts including a Chinese bronze mirror fragment. Extensive shovel testing and screening have exposed a relatively small number of artefacts and given the distinctive topographical location of this site overlooking the entire lower valley, we interpret EGS 131 as a potential defensive structure or ceremonial enclosure used on a temporary basis. The Xiongnu period at Egiin Gol shows a substantially different locational pattern from that of the preceding Bronze and Early Iron Age. With the exception of a few sites in the south-eastern tributary valleys, habitation and cemeteries are concentrated in the main Egiin Gol valley, as is the single special purpose site. The location of habitations having large surface areas during the Xiongnu period fall at opposite ends of the valley (Fig. 5). The large habitation cluster to the north-west is made up of a tightly spaced group of three sites which are close enough to be continuous (90–190 m separation) and add up to roughly the same surface area as the single large site to the south-east, approximately 4 ha. The mean site size for habitation increases substantially over that of BEIA and the range of site size categories expands as well (Fig. 6). The Xiongnu period sites do not conform to a clear centralising pattern but instead are located in the areas of entrance to and exit from the main valley and along the Egiin Gol river. If a line were to be drawn connecting each Xiongnu habitation, it would approximate the major route of travel through the Egiin Gol 5
Determinations used are TX-9315, TX-9316, TX-9317, Beta-103071 (Hall, Batsaikhan and Honeychurch 1999), Ly-6857 (Cubezy et al. 1996), LE-6062 (Burial 88, coffin wood: 2040 ±45 BP, 2ó calibrated date range is 160 BC–AD 50 at 95% probability), and LE-6060 (Burial 73, charcoal: 1720 ±40 BP, 2ó calibrated date range is 230 BC–AD 420 at 95% probability).
10. SURVEY AND SETTLEMENT IN NORTHERN MONGOLIA
195
Fig. 4. Summed probability distribution of the calibrated C14 determinations for Xiongnu cemetery site EGS 100 (BT refers to the local site name, Burkhan Tolgoi).
196
WILLIAM HONEYCHURCH AND CHUNAG AMARTUVSHIN
Fig. 5. Distribution of Xiongnu period sites at Egiin Gol (largest sites marked by squares).
0
14
N
0
20000
N
Std Dev
Mean
10000
Settlement Size (m2)
50000
40000
14
12,385
12,641 m2
30000
XIONGNU
Fig. 6. BEIA and Xiongnu period site size comparison – one BEIA site is excluded for lack of size.
1909
40000
Std Dev
30000
2459 m2
20000
Mean
10000
BEIA
50000
10. SURVEY AND SETTLEMENT IN NORTHERN MONGOLIA
197
198
WILLIAM HONEYCHURCH AND CHUNAG AMARTUVSHIN
valley in use for at least the past 200 years. Cemetery and habitation sites located in the south-eastern upper tributary valleys are likewise located along routes leading to major mountain passes still in use today. These patterns suggest that Xiongnu sites were positioned not in relation to local resources, as during the BEIA, but in relation to lines of mobility through the valley with emphasis on areas of entrance and exit having potential for interaction with external regions.
Conclusions Returning once again to Barfield and Di Cosmo, the hypothesis that nomadic societies at the local level form stable political and economic units that resist change is not supported by the preliminary evidence from the Egiin Gol valley. Instead, the local perspective provided by full-coverage survey offers a more complicated and informative scenario. The survey site data show that during the transition from the BEIA to the Xiongnu period, the Egiin Gol region witnessed changes in site size and size categories as well as locational emphases which likely indicate political and economic changes in the local organisation of the valley. As a working hypothesis requiring further testing, we propose that the site patterns at Egiin Gol show a transition in emphasis from control of local and presumably pastoral resources in tributary valleys to the control of points of articulation and lines of movement connecting the valley to external regions. These preliminary patterns favour the expectations of the Di Cosmo model over those of the Barfield model, which de-emphasises changes at the scale of the local area during the formation of a regional nomadic polity. The work at Egiin Gol demonstrates a process of local re-organisation that potentially facilitated higher levels of integration among nomadic groups and made possible centralised and spatially extensive Inner Asian political systems (Fletcher 1986, 21). We should emphasise that while the Egiin Gol evidence suggests that local areas were more dynamic than would be predicted by Barfield, his model does provide one explanation for the pattern of site changes in the valley during the Xiongnu period. Barfield views the extraction and redistribution of Chinese products through the nomadic royal court as the primary means of integration and support of the nomadic polity and therefore, the increased political importance of circulated goods might explain the expansion of external relationships
10. SURVEY AND SETTLEMENT IN NORTHERN MONGOLIA
199
at Egiin Gol. The appearance of Chinese bronze mirrors (EGS 100, 131) and Han dynasty Wushu coins (EGS 36) in Egiin Gol Xiongnu burials and at habitation sites provides some evidence for the importance of such products. While the circulation of desirable objects of foreign production was certainly one important integrative strategy, Claudia Chang (with Tourtellote 1998, 275) provides historical and archaeological evidence to suggest that nomadic steppe polities may have also developed more complex economic systems involving a variety of sources of income and political finance. These included differentiated and specialised subsistence economies having both agricultural and pastoral sectors and interaction with a wide range of neighbouring groups through exchange and tribute relationships. The organisational changes observed at Egiin Gol, therefore, might be related to either a regionally integrated and increasingly differentiated subsistence system or perhaps to the development of macroregional trade and transport routes, a possibility that Barfield has argued for in his most recent work (2001, 21). These and other competing hypotheses remain to be tested against the archaeological evidence from the Egiin Gol valley. As analysis of our survey database progresses, an improved understanding of nomadic polities, their organisation, and the nature of inter-regional interaction in north-east Asia will emerge. One conclusion that is certain at this point is the evident success of archaeological survey in recovering minimal habitation sites and a number of other site types related to early nomadic groups. Fullcoverage, regional survey at Egiin Gol has added a local and indigenous perspective on nomadic society that is rarely accessible through the historical records of 2000 or more years ago. Archaeological surveys similar to that carried out at Egiin Gol, combined with mortuary and settlement excavation at a number of sites across Inner Asia have resulted in a heightened appreciation for the complex nature of nomadic society and its potential for structural change. These large-scale, state-like, and often imperial political systems of nomadic Inner Asia have important lessons to teach those of us seeking to better understand the development of human organisation through its remarkable diversity (Drennan 1996).
Acknowledgments The Egiin Gol Survey has received generous support from the Wenner-Gren Foundation for Anthropological Research, the Mongolian Academy of Sciences
200
WILLIAM HONEYCHURCH AND CHUNAG AMARTUVSHIN
and the University of Michigan Museum of Anthropology. We would like to thank Henry Wright, Jeffrey Parsons, Carla Sinopoli, Mark Hall and Damdinsuren Tseveendorj for useful comments and other assistance on this paper.
Bibliography Amartuvshin, C. and Erdenebat, U. 2000: Mongol-Amerikiin khamtarsan ekspeditsiin tailan, 2000 ond Bulgan aimagiin Khutag-Ondor sumyn nutag Egiin Gold. Unpublished field report, Academy of Sciences Institute of History, Ulaanbaatar. Barfield, T.J. 1981: The Hsiung-nu Imperial Confederacy: Organization and Foreign Policy. Journal of Asian Studies 41.1, 45–61. ——. 1989: The Perilous Frontier: Nomadic Empires and China (Cambridge). ——. 2001: The Shadow Empires: Imperial State Formation along the Chinese-Nomad Frontier. In Alcock, S.E., D’Altroy, T.N., Morrison, K.D. and Sinopoli, C.M. (eds.), Empires: Perspectives from Archaeology and History (Cambridge), 10–41. Bronk Ramsey, C. 2000: OxCal Program Manual (Available from http://www.rlaha.ox.ac.uk/oxcal/ oxcal.htm). Chang, C. and Tourtellote, P. 1998: The Role of Agro-pastoralism in the Evolution of Steppe Culture in the Semirech’e Area of Southern Kazakhstan during the Saka/Wusun Period (600 BCE–400 CE). In Mair, V. (ed.), The Bronze Age and Early Iron Age Peoples of Eastern Central Asia vol. 1 (Washington, DC), 264–79. Cribb, R. 1991: Nomads in Archaeology (Cambridge). Crubézy, E., Martin, H., Batsaikhan, Z., Erdenbaatar, D., Maureille, B. and Verdier, J. 1996: Pratiques funéraires et sacrifices d’animaux en Mongolie à la période proto-historique. Paléorient 22.1, 89–107. Danilov, S.V. and Konovalov, P.B. 1988: Novye materialy o kurganakh-kereksurakh Zabaikal’ya i Mongolii. In Konovalov, P.B. (ed), Pamiatniki epokhi paleometalla v Zabaikal’e (Ulan-Ude), 61–79. Davydova, A.V. 1995: Ivolginskii arkheologocheskii kompleks: Ivolginskoe gorodishche (St Petersburg). Di Cosmo, N. 1994: Ancient Inner Asian Nomads: Their Economic Basis and Its Significance in Chinese History. Journal of Asian Studies 53.4, 1092–1126. ——. 1999: State Formation and Periodization in Inner Asian History. Journal of World History 10.1, 1–40. ——. 2002: Ancient China and Its Enemies: The Rise of Nomadic Power in East Asian History (Cambridge). Dikov, N.N. 1958: Bronzovyi vek Zabaikal’ya (Ulan-Ude). Drennan, R. 1996: One for All and All for One: Accounting for Variability without Losing Sight of Regularities in the Development of Complex Society. In Arnold, J.E. (ed.), Emergent Complexity: The Evolution of Intermediate Societies (Ann Arbor), 9, 25–34. Erdenebaatar, D. 1997: Selenge Mornii sav dakhi’ khurel ba tomor zevsgiin turuu ueiin dursgal. (PhD thesis, Academy of Sciences Institute of History, Ulaanbaatar). ——. 2000: Bulgan Aimagiin Khutag-Ondor sumyn Khantai bagiin nutag Egiin Golyn khondiid yavuulsan etnografiin ekspeditsiin sudalgaany tailan. Unpublished field report, Department of Archaeology and Ethnology, Ulaanbaatar University (on file at the Mongolian Institute of History). Erdenebaatar, D., Erdenebat, U., Torbat, T., Crubézy, E., Maureille, P. and Khai, L. 1998: Burkhan Tolgoin khunnu bulshny sudalgaa. Arkheologiin sudlal 18.8, 92–107. Fletcher, J. 1986: The Mongols: Ecological and Social Perspectives. Harvard Journal of Asiatic Studies 46.1, 11–50.
10. SURVEY AND SETTLEMENT IN NORTHERN MONGOLIA
201
Hall, M., Batsaikhan, Z. and Honeychurch, W. 1999: Radiocarbon Dates From Northern Mongolia. Radiocarbon 41.1, 103–10. Honeychurch, W., Trigg, H., Nelson, A., Rainville, L., Amartuvshin, C. and Erdenebat, U. 2001: Report of the Mongol-American Egiin Gol Survey, 2000. Unpublished field report, Institute of History, Mongolian Academy of Sciences, Ulaanbaatar. Humphrey, C. and Sneath, D. 1999: End of Nomadism? Society, State and the Environment in Inner Asia (Durham, NC). Khazanov, A. 1994: Nomads and the Outside World (Madison, WI). Konovalov, P.B. 1975: Pogrebal’nye sooruzheniya khunnu. In Sibir’, Tsentral’naya, i Vostochnaya Aziya v srednie veka (Novosibirsk), 17–46. ——. 1976: Khunnu v Zabaikal’e (Ulan-Ude). —— (ed.). 1995: Kul’tury i pamyatniki bronzovogo i rannego zheleznogo vekov Zabaikal’ya i Mongolii (Ulan-Ude). Kradin, N.N. 2000: Kochevniki i zemledel’cheskii mir: Khunnskaya model’ v istoricheskoi perspektive. Vostok 3, 5–16. Lattimore, O. 1962: Studies in Frontier History (Paris). Lbova, L.V. and Grechishchev, E.P. 1995: K probleme poselenii epokhi metalla stepnoi zony Zapadnogo Zabaikal’ya. In Konovalov, P.B. (ed.), Kul’tury i pamyatniki bronzovogo I rannego zheleznogo vekov Zabaikal’ya i Mongolii (Ulan-Ude), 7–17. Minyaev, S. 1985: K probleme proiskhozhdeniya syunnu. Information Bulletin of the Association for the Study of the Cultures of Central Asia 9, 70–8. ——. 1998: Derestuiskii mogil’nik (St Petersburg). Minyaev, S. and Sakharovskaya, L. 2002: Soprovoditel’nye zakhoroneniya ‘tsarskogo’ kompleksa No. 7 v mogil’nike Tsaram. Arkheologicheskie Vesti 9. Navaan, D. 1975: Dornod Mongolyn khurliin ue (Ulaanbaatar). Novgorodova, E.A. 1989: Drevnyaya Mongoliya (Moscow). Perlee, K. 1961: Mongol Ard Ulsyn ert, dundad ueiin khot suuriny tovchoon (Ulaanbaatar). Pulleyblank, E. 1983: The Chinese and Their Neighbors in Prehistoric and Early Historical Times. In Keightley, D. (ed.), The Origins of Chinese Civilization (Berkeley), 411–65. Rudenko, S.I. 1962: Kul’tura khunnov i Noinulinskie kurgany (Moscow). Sinor, D. 1970: Central Eurasia. In Sinor, D. (ed.), Orientalism and History (Bloomington), 93–119. Stuiver, M., Reimer, P., Bard, E., Beck, J., Burr, G., Hughen, K., Kromer, B., McCormac, G. van der Plicht, J. and Spurk, M. 1998: INTCAL98 Radiocarbon Age Calibration, 24,000–0 cal BP. Radiocarbon 40, 1041–85. Toynbee, A. 1934: A Study of History vol. 3 (Oxford). Trever, K.C. 1932: Excavations in Northern Mongolia (Leningrad). Tseveendorzh, D. 1985: Novye dannye po arkheologii khunnu. In Vasil’evsk, R.S. (ed.), Drevnie Kul’tury Mongolii (Novosibirsk), 51–87. Tseveendorzh, D. and Batsaikhan, Z. 1995: Khunnugiin vaarny tukhai. Arkheologiin sudlal 14.5, 76–107. Tsybiktarov, A.D. 1998: Kul’tura plitochnykh mogil Mongolii i Zabaikal’ya (Ulan-Ude). Volkov, V.V. 1981: Olennye Kamni Mongolii (Ulaanbaatar). Watson, B. 1961: Records of the Grand Historian of China (New York). Wu, E. 1990: Lun Xiongnu kaogu yanjiu zhongde jige wenti. Kaogu Xuebao 4, 409–37.
CHAPTER ELEVEN RECENT ARCHAEOLOGICAL RESEARCH IN THE KHANUY RIVER VALLEY, CENTRAL MONGOLIA FRANCIS ALLARD, DIIMAAJAV ERDENEBAATAR AND JEAN-LUC HOULE
Carried out in conjunction with the Institute of History of Mongolia, the Khanuy Valley Project on Early Nomadic Pastoralism in Mongolia completed its first field season over a period of five weeks during the months of July and August 2001. The project, which aims to clarify the timing and circumstances of the early stages of nomadic pastoralism in central Mongolia, focuses on a portion of the Khanuy river valley, a region characterised by large numbers of stone built archaeological sites located on the valley floor and hill slopes. Aside from an earlier iconographic survey of some of its ‘deer stones’, no systematic study of the valley’s archaeological landscape had been attempted prior to the establishment of this collaborative effort with Mongolian archaeologists. The project differs from other archaeological field studies in Mongolia in its regional focus, its multidisciplinary nature, and in the fact that it includes an ethnographic component. Because so little was known of the valley’s archaeology when the project was initiated, the objective of the first field season was for us to familiarise ourselves, through interviews and the excavation and mapping of a small number of sites and structures, with the main features of both the archaeological and present-day cultural landscape. We also hoped to collect materials that could be subsequently analysed and chronometrically dated. The few interviews conducted during the first field season gathered data on land use patterns, information which in turn is sure to provide important insights into past subsistence systems. This paper reviews the results of the first field season, briefly considers how the data may shed light on past socio-political processes, and discusses the future directions to be followed in order to address such issues in a more systematic manner. The mapping survey conducted by the project at Gol Mod-2, a large nearby Xiongnu cemetery, is discussed in another publication (Allard et al. 2002).
11. ARCHAEOLOGICAL RESEARCH IN THE KHANUY RIVER VALLEY
203
The Bronze Age of Mongolia Many different types of stone built sites have been identified within the territory of present-day Mongolia (Erdenebaatar 2004).1 Mongolian archaeologists date many of these sites to the Bronze Age, a period which they suggest extends from the 2nd to the 1st millennium BC. Although a few radiocarbon dates are now available, this chronology relies mainly on similarities between artefacts excavated in Mongolia and material recovered from chronometrically dated contexts in surrounding regions. Importantly, the sites identified so far all appear to have had a ritual and/or funerary function and there is as yet no clear evidence of substantial surface remains of Bronze Age settlements. Many archaeologists believe that full nomadic pastoralism was adopted in Mongolia around the beginning of the 1st millennium BC, and thus that many of these Bronze Age sites – many of which are very substantial constructions – were built and used by mobile herders. In contrast to such sites – which continue to be the focus of investigations by both Mongolian and foreign archaeologists –, there has been little interest in locating and excavating the campsites of past herders. The most prominent archaeological sites in the research area are stone built structures known in Mongolian as khirigsuurs. Widely distributed throughout the western two-thirds of Mongolia (and even beyond its borders), the khirigsuurs consist of a central stone mound surrounded by a stone ‘fence’, which is either a circle or a four-sided perimeter of ground level stones. The larger sites often also include specific areas dedicated to small stone mounds or stone circles, various paths, as well as other structures. Up until two years ago, a total of about a dozen khirigsuurs had been completely or partially excavated in Mongolia. Possibly as a result of looting, nothing was found under many of the central mounds, while some did contain the skeletal remains of humans placed inside stone cysts on the original ground surface, with their heads pointing west. The few artefacts associated with the interred apparently did not include weapons. Deer stones, the second type of widespread archaeological remains in the western two-thirds of Mongolia (and beyond in Eurasia), are stone stelae whose surfaces are decorated with a range of images. These include groups of
1
General information on the Bronze Age of Mongolia discussed in this paper relies on the paper by Erdenebaatar cited here, as well as on the many conversations we have had with him and Bill Honeychurch.
204
F. ALLARD, D. ERDENEBAATAR & J.L. HOULE
stylised deer with a bird like muzzle and long flowing antlers, circles (variously interpreted as the sun, moon, or earrings), bows, an object thought to be a shield, as well as metal weapons and tools, some of these suspended from what appears to be a belt. The placement of these various images in horizontal bands and on different sides of the stelae suggests to some that they represent tattoos on the human body. Both the types of weapons depicted on the deer stones and the few instances of a physical association between khirigsuurs and stelae suggest that the latter also date to the Bronze Age. Distributed throughout the eastern two-thirds of Mongolia, the so-called ‘slab burials’ are a third prominent element of the archaeological landscape during the Bronze Age. Although the burial pits themselves vary, consistent features of slab burials include the orientation of the body toward the east and the use of stone slabs, which stand vertically on end and whose top portions usually extend above the ground surface to mark the tomb’s perimeter. Mongolian archaeologists report that almost every slab burial excavated so far appears to have been looted in the past, although excavations have nevertheless yielded a range of artefacts. Along with bone arrowheads, these include various bronze artefacts, including arrowheads, helmets, axe heads, daggers, ornaments in the ‘animal style’ and small belt pendants. On the basis of a few available radiocarbon dates and comparisons with material excavated outside Mongolia, archaeologists suggest that the slab burials were common from the 13th to the 3rd/2nd centuries BC.
The Research Area: Geography, Archaeology and Objectives Located in Ondor Ulaan sum of Arkhangai aimag, some 430 km west-northwest of Mongolia’s capital Ulaan Baatar, the research area within which archaeological investigations have been carried out extends along a 30 km stretch of the Khanuy river, a stream that flows in a general north-easterly direction from its source in the Khangay mountain range in central-west Mongolia (Fig. 1). The size and boundaries of the 330 km2 research area were determined after carefully considering available resources available over the projected five year project, the known location of significant sites, as well as the need to include within the area as broad a range of ecological and land-use zones as possible. It is felt that the study of such an area will permit the recovery of the kind and amount of data needed to address issues such as the dynamics of human-land
11. ARCHAEOLOGICAL RESEARCH IN THE KHANUY RIVER VALLEY
Fig. 1. Khanuy river and the research area.
205
206
F. ALLARD, D. ERDENEBAATAR & J.L. HOULE
relationships, the spatial patterns displayed by habitation and non-habitation sites, as well as chart changes in these and other variables. Within the research area, the Khanuy river meanders through a valley floor whose altitude ranges between 1600 and 1700 m. During the summer months (prior to the rainy season), the Khanuy is a slow moving stream whose maximum width does not exceed about 40 m. A wetland area extends in some places up to a few kilometres from the river. The extent of the present and past floodplain has yet to be determined. A limited number of streams, most of these small and some seasonal, flow into the Khanuy from the hills that border the valley. The topography is typical of the steppe landscape found throughout much of Mongolia. The terrain is mostly gentle and there are few steep cliffs. Within the research area, the hills reach a maximum altitude of about 2000 m and most of the side valleys are short and contribute little water to the area’s hydrological regime. Vegetation consists of extensive grasses that cover all of the valley bottom as well as much of the hill slopes. Stands of larch cover large portions of the hilltops and hillsides above 1800 m throughout most of the research area. Like many other regions of Mongolia, this section of Khanuy valley remains economically and geographically isolated. The only permanent settlement is small and the majority of the valley’s inhabitants are herders who practice mobile pastoralism and live in gers (more commonly known in the west as ‘yurts’). Since the early 1990s, a small number of families have begun growing vegetables on small plots that, in the case of mobile herders, are abandoned for a part of the year. The herders move from two to four times a year, although there is evidence that present-day migrations are much less extensive than in the past. Typically, the herders locate their summer camps along the banks of Khanuy river and their winter camps near the bottom of the hills. All three types of Bronze Age sites discussed above are present in large numbers in the research area, with many easily observed from the dirt road that runs along the western side of the valley. The most visually prominent of these are the khirigsuurs, the largest of which are distributed along the valley floor while smaller ones are located on hilltops. Slab burials are also found in significant numbers, with many in fact located within the perimeter of khirigsuurs. Deer stones, which are also well represented, are found as isolated stelae or in groups. Located in the northern sector of the research area next to a khirigsuur, a concentration of impressive deer stones – identified as the ‘deer stone site’ in Fig. 1 – has been studied by Mongolian archaeologists interested in the iconography of their surface designs. Unfortunately, most of these now
11. ARCHAEOLOGICAL RESEARCH IN THE KHANUY RIVER VALLEY
207
lie on their side at the site, having been moved from their original (but unrecorded) location. A non-systematic survey of the research area has revealed the presence of other stone built sites on hillsides and hilltops on both sides of the river. These include small single mounds, mounds surrounded by a fence whose corners are marked by standing stones, and long lines of surface stones. According to Mongolian archaeologists, some of these structures date to periods other than the Bronze Age, such as some of the small single mounds, which they believe date to the Turkish period (6th–7th centuries AD). Importantly, all of the remains mentioned so far appear to have played a ritual and/or funerary role. We have not yet come across any surface evidence of camp sites or more permanent habitation sites in the research area, an absence which, along with other data presented later, points to the likely association between the Bronze Age sites and a nomadic pastoralist way of life, with such sites possibly representing some of the early stages in the development of this subsistence system in central Mongolia. The Khanuy valley’s rich archaeological landscape encourages us to gather as much information as possible from the research area. Rather than concentrate on the excavation of a few selected burials in the valley and the comparison of grave goods with material from other regions, our research plan consists of studying, in a more intensive manner than is common in Mongolian archaeology, the research area as a whole, as well as single sites, so that new information and insights may be gained regarding ritual behaviour, the internal organisation of sites, the spatial patterning of sites within the valley, as well as the chronometric dating of a sufficient number of sites and structures to permit the charting of temporal change in all of these variables. Combined with the recovery of such data, a settlement pattern study aimed at the identification of difficult to locate campsites will contribute to an understanding of ancient subsistence systems and the spatial organisation of power in the valley. In our attempt to study the pre-existing circumstances, emergence and early development of early nomadic pastoralism in central Mongolia, we rely on the experience of Mongolian archaeologists in other regions of Mongolia to restrict our study to the valley’s Neolithic, Bronze Age and Iron Age occupation. The latest period of specific interest to this project is the Xiongnu period, which extends from the 3rd century BC to the 2nd century AD, and which is represented by a significant cemetery located in the south-east portion of the research area (Allard et al. 2002).
208
F. ALLARD, D. ERDENEBAATAR & J.L. HOULE
Spatial Patterning in the Research Area No full-coverage or probabilistic survey of any portion of the valley has yet been carried out, although we believe sufficient time has been spent driving and walking through the research area to permit us to make a few general observations and suggest possible spatial patterns to be tested by survey over the course of the next field seasons. For example, we observe that the largest khirigsuurs tend to be located to the west of Khanuy river, usually on level ground between the hill slopes and the westernmost extent of the floodplain, close to the line of the present-day dirt road. Some of the large sites are also found at slightly higher elevations, sometimes on promontories that allow for a good view of the valley. The location of a selection of large khirigsuurs in the research area was recorded using a GPS (Global Position System) and is illustrated in Fig. 1. No khirigsuurs have yet been identified at the lowest elevations of the valley or in the wetland area, a finding that suggests (as in the case of the modern road) a preference to build sites on firm ground beyond the reach of floodwaters. The apparent absence of large khirigsuurs to the east of the river is possibly explained by a lack of sufficiently large expanses of level land between the river and hill slopes and/or the presence of ponds or lakes in that sector. The first field season concentrated on an area midway up the research area where a number of khirigsuurs have been located. These include what appear to be the two largest khirigsuurs in the research area: Urt Bulagyn (KYR1) and another, slightly larger one, located 1.3 km to its south. The presence of many khirigsuurs in this sector of the valley (only four of which are identified in Fig. 1), including the two largest sites, points to its possible socio-political importance at the time of construction and use. A survey of a roughly 15–20 km2 area centred on these two sites and extending from the river to the western hilltops identified a total of over 100 sites, including smaller khirigsuurs (a few on hilltops and promontories), lines of stones on the valley floor, isolated stone mounds and stone circles, and mounds surrounded by a square or round perimeter of stones. Many of the smallest sites are located on the hill slopes. No clear surface evidence of ancient settlements or even temporary campsites was identified in any portion of the research area during the first field season.
11. ARCHAEOLOGICAL RESEARCH IN THE KHANUY RIVER VALLEY
209
Urt Bulagyn The site of Urt Bulagyn (KYR1) is located at N48°05’35” E101°03’25”, 2 km west of the Khanuy river, less than 1 km west of the valley’s marshy bottom, and roughly 1km east of the valley’s western foothills. Built entirely of unaltered – and apparently quarried – stones, it measures 390 m (north-south) × 390 m (west-east) and is the second largest khirigsuur identified in the research area (Fig. 3). Much of the 2001 field season was dedicated to investigating this impressive site, which consists, like other large khirigsuurs in the valley, of a number of well-defined areas, each characterised by a distinctive type of structure. These structures are easily visible among the short grass that covered the site at the time of investigation. This section provides a description of the site’s spatial organisation, of each type of structure found at the site, and of the excavations carried out during the first field season. The Central Mound. Urt Bulagyn’s most prominent structure is its central mound, which is 4.9 m tall at its highest point and about 30 m in diameter at ground level (Fig. 2). A central depression extending toward the east suggests that the mound may have been disturbed in the past from that side. Although the disturbance does not reach the ground surface below the mound, it is possible that any man-made depression – such as that caused by looting – may have since then partially filled up with stones rolling down from its sides. No attempt to excavate the central mound was made during the 2001 field season. The Plaza and Fence. Extending from the mound to the eastern fence, an extensive scatter of surface stones suggests the possible presence of various structures. Such types of stone scatters have been identified at other large khirigsuurs in Khanuy valley. At Urt Bulagyn, we note the presence of disturbed and poorly defined ‘horns’ extending from the central mound, a feature identified so far at no less than one other site in Mongolia. Although the horns might in the past have defined areas with special functions, their lack of symmetry and poorly defined shapes may simply indicate that they are the remains of looting activity into the central mound’s eastern side. Better defined is an apparent path that is 40–45 m long. On the surface, the mostly disturbed scatter of stones makes it difficult to distinguish any pattern or evident structure(s), although it is possible to detect what appear to be two rows of stones – many of these white – that define the northern and southern edges of the path.
210
F. ALLARD, D. ERDENEBAATAR & J.L. HOULE
Fig. 2. Plan of the khirigsuur Urt Bulagyn (KYR).
11. ARCHAEOLOGICAL RESEARCH IN THE KHANUY RIVER VALLEY
Fig. 3. Plan of the khirigsuur KYR 2.
211
212
F. ALLARD, D. ERDENEBAATAR & J.L. HOULE
A trench measuring 11 × 6 m was opened and excavated to a depth of about 20 cm, revealing the surface upon which the majority of the stones in the path rest (the excavation trench is shown in Fig. 3). The overall distribution of the stones revealed by the excavation did little to clarify the internal structure of the path, although the stone rows (especially the northern line of stones) now stood out more clearly than prior to the excavation. Just within the southern edge of the path, a small mound measuring 1.3m in diameter became apparent once the top 10 cm of soil had been removed. The mound was excavated to a depth of 23 cm and yielded a number of animal bones (possibly sheep or goat). Two other possible small stone mounds were identified in the unexcavated portion of the path. Aside from the central mound and the possible features mentioned above, the plaza, whose perimeter is defined by the rectangular fence, also includes a few well-defined structures, as well as numerous vague scatters of stones. The structures include one burial (indicated in Fig. 3 as a rectangle), an arc-shaped line of stones, two stone circles, four small stone mounds, one larger mound next to the plaza’s north-west corner mound, and an even larger one placed atop an earthen base in the north-east sector. Undoubtedly, many other structures within the plaza have yet to be identified, having been disturbed and/or covered by sediments since the construction of the site. The rectangular stone fence, which measures about 65 × 90 m, consists of rows of stones that generally protrude from the ground by no more than 15 cm. It is important to point out the marked irregularity of the rectangle – opposing sides differ in length and are not perfectly parallel – and the fact that the rows of stones are themselves not straight. It is suggested that this absence of linearity is not the result of soil movement but due to the fact that the builders simply did not require the sides of the fence to be straight. A total of four large stone mounds are located at the corners of the stone fence perimeter. The south-east corner mound is the only component of the fence complex that was excavated. The mound is about 1.4 m tall, 7–8 m in diameter, and was found to consist of about 3000 stones. A number of scattered horse bones and teeth recovered as the stones were removed are believed to be recent deposits, or to at least postdate the original construction. A layer of disturbed earth lay at the bottom of the stone mound. Sterile soil was reached at a depth of 15 cm below this disturbed layer. No artefacts or faunal remains were recovered from these bottom layers and it is suggested that the south-east corner mound may have functioned primarily as an architectural feature.
11. ARCHAEOLOGICAL RESEARCH IN THE KHANUY RIVER VALLEY
213
The Small Stone Mounds. A total of 1752 small stone mounds occupy the area between the stone fence and the ring-shaped sector of stone circles that surrounds the site (Fig. 2). Of these, 315 are labelled ‘satellite mounds’ (see below). The majority of the non-satellite mounds are located on the southern and eastern sides of the fence. The northern sector is no more than a narrow band of mounds, while very few are present on the western side. Measurements of a sample of mounds permit the following observations. The mounds are generally placed some 4–6 m from one another. Their average height (above present-day ground level) and diameter are 45 cm and 4.0 m respectively, with recorded heights ranging between 24 and 67 cm, and diameters ranging from 2.2 to 6.3 m. Although some of the mounds present a sparse surface appearance of fewer than 20 stones, over 100 visible stones have been recorded in the case of a few large mounds. It is important to note that the mounds are typically larger and taller than their surface appearance suggests since a portion of each mound remains buried under the surface. Also significant is the fact that, together with the placement and shape of the fence, the stone mound area appears to provide the khirigsuur with an orientation toward the west-northwest. If we consider for a moment that the scarcity of mounds along the western fence represents the site’s ‘opening’, the heading of Urt Bulagyn can then be determined to be roughly 285°. The mound dimensions do not differ significantly among different sectors of the mound area, except for the sparse band of mounds on the western side, where the largest and tallest mounds tend to be concentrated. The southern mound sector is itself marked by two features not found on the other three sides. First, many of the mounds appear to be arranged in rows that display a south-east to north-west orientation. Second, the satellite mounds mentioned above are found only in the southern sector. These mounds are typically small (average diameter: 1.7 m; average height: 23 cm) and placed to the immediate south-west of the larger mound to which they appear to be attached. Two nonsatellite stone mounds were excavated, both of which are located in the southern sector of the stone mound area. The stone mound KYR1–1 measures 3.8 m in diameter and is 47 cm tall (maximum height above the present-day surface). An area of 18 m2 (4.0 × 4.5 m) was excavated and the stones removed and counted. The mound consists of a total of approximately 400 stones (stones less than 10 cm long were not counted). The heaviest of these, reaching almost 1 m long, required two people
214
F. ALLARD, D. ERDENEBAATAR & J.L. HOULE
to carry out of the trench. The base of the mound was found to lie 10–12 cm below the modern ground surface, on what, it is suggested, may have been the original surface before wind blown sedimentation began covering the base. The bottom stones at the centre of the mound formed an irregular stone cyst that offered some degree of protection to a set of carefully placed horse bones found at a depth of 8–10 cm below this original surface. The bones include six front teeth pointing upward and six (as yet unidentified) vertebrae, one of these vertebrae placed on its own and five placed in sequence to the south-west of the teeth and single vertebrae (Fig. 4). The vertebrae and teeth display an orientation of 117°. It is worth pointing out that the highly selective nature of the remains, the short distance (8–10 cm) separating the row of vertebrae from the other bones, and the fact that the vertebrae abut one another in an anatomically correct manner, together make it highly likely that vertebrae were first removed from the neck and then carefully positioned on the ground while still held together by their ligaments. Excavation, which was continued to a depth of 40 cm below the original surface, located no further faunal material or artefacts. A second stone mound (KYR1–9), measuring 3.5 m in diameter, yielded a complete horse skull pointing east (90°). The skull, recovered from the bottom of the mound near its centre, appeared to be partially (if not completely) buried below the soil surface on which the mound’s bottom stones lay. Although the top of the skull was crushed when found, it appears as if it had been carefully positioned to protect it from the weight of the overlying stones. A molar was removed from the skull and radiocarbon dated to 1040–850 BC (calibrated date; 2 sigma = 95% probability). The Northern Path/Platform. This is a long structure located immediately north of the stone mound sector. Measuring 4 × 125 m, it appears as a well-defined scatter of surface stones. It displays no clear surface pattern of internal organisation. No excavations of this structure were carried out during the first field season. The Human Burials. A total of 15 structures believed to be human slab burials have been identified at the site. All burials except for one are located in the northern part of the site between the stone mound area and stone circle area (the other, mentioned earlier, is located in the southern sector of the plaza). On the surface, the burials generally appear as a central scatter of stones surrounded by a rectangular perimeter consisting of a varying number of corner stones and standing stones set end to end.
11. ARCHAEOLOGICAL RESEARCH IN THE KHANUY RIVER VALLEY
215
Fig. 4. Urt Bulagyn. Central mound in the background and small mounds in the foreground (photograph by Francis Allard).
The frequent absence of corner stones and the irregular surface appearance of the burials – very few sides are parallel – do not permit for more than a rough estimate of the dimensions and orientations of the burials. The following general patterns can nevertheless be distinguished. The size of the burials, calculated as the area within the perimeter defined by actual or estimated corner stones, ranges from 4.9 m2 (2.7 × 1.8 m) to 28.5 m2 (5.7 × 5.0 m). As mentioned earlier, local archaeologists suggest that the head of the interred in slab burials tends to point toward the east. If that remains true at Urt Bulagyn, and if we use the longer side of the burial to define its orientation, we note that all 15 of the burials are therefore oriented in a south-west to north-east direction, and that 13 of the 15 burials in fact point more specifically toward the east – north-east (between roughly 49° and 85°). Two burials were excavated during the first field season, both of these located near the western end of the long rectangular path. The burial KYR1–3 consists of a conglomerate of stones clearly marked by four cornerstones. The somewhat irregular perimeter marked by the cornerstones
216
F. ALLARD, D. ERDENEBAATAR & J.L. HOULE
measures approximately 2.7 × 3.2 m, with the longest (non-parallel) sides oriented toward an average heading of 64°. A 5 × 5 m area was excavated A total of about 350 stones were removed during the excavation. At a depth of 55–70 cm below the present-day surface, the distribution of the stones began to show the outline of a smaller inner burial. At a depth of 75–80 cm below the surface, a human jawbone was discovered within the inner burial at a distance of 5 cm from its northern wall. Further skeletal remains were found at a depth of 84 cm. These included portions of a human skeleton that pointed toward the north-east. The bones included parts of the upper and lower leg, of the pelvis and a portion of a lower arm bone. Missing bones included the skull, ribs, vertebrae, the feet and anklebones, and most of the pelvis and arm bones. A sheep/goat jaw and horse teeth were found in the area of the (missing) head. The fact that the distal ends of the lower leg bones abutted against a stone suggests the possibility of a secondary burial, a hypothesis supported by the apparent placement of sheep/goat and horse remains where the human head would have been. On the other hand, it is also possible that the discovery of a human jaw near the northern wall of the burial (and its placement higher than the skeleton) indicates rodent activity or possibly looting. The burial KYR1–5 is situated 20 m north-east of KYR1–3. Marked by four cornerstones, the central conglomerate of surface stones measures approximately 1.7 × 2.0 m, with the longest sides oriented somewhere between 60° and 70°. If we include the scattered stones found beyond the central conglomerate, the burial site measures approximately 4.6 × 4.1 m. A 4 × 4 m area was excavated. A total of 225 stones were removed during the excavation. At a depth of 50–65 cm below the modern surface, the distribution of the stones revealed the outline of an inner burial. At a depth of 1.0 m, an area of soil discoloration became evident near the centre of the burial pit. No artefacts or bones were found within or near this area. It is suggested that this area of discoloration represents the traces of a decayed buried child. The discoloration was followed to a depth of 1.40 cm, at which point the soil was found to be sterile. The Stone Circles. The outermost area of the khirigsuur is a ring-shaped section consisting entirely of ground level stone circles. A total of 1021 stone circles were counted. Circles range in diameter from roughly 1 to 3 m, with many of the largest circles found in the western part of the stone circle area. Stones typically protrude above the ground surface by no more than about 15 cm, with many lying flat on the ground and others buried just below the surface. The
11. ARCHAEOLOGICAL RESEARCH IN THE KHANUY RIVER VALLEY
217
number of stones protruding above the ground ranges from 1 in the case of circles that are almost entirely buried, to about 20 in the case of the large circles whose stones are all exposed. A small percentage of the ‘circles’ are actually more square than circular. The western sector of the stone circle area is the widest, reaching a maximum width of about 100 m. In many parts of the stone circle area, stone circles are arranged in rows that are roughly parallel to one another and to the side of the stone fence nearest to them. However, this arrangement of parallel rows often breaks down, for example when rows merge together or end abruptly, as well as near the corners of the site. The distance between circles throughout much of the stone circle area ranges from about 5 to 7 m, with circles in the western sector separated by as much as 10 m. The distance between the rows themselves averages 6–7 m. Three stone circles were excavated: two in the western sector of the stone circle area, and one in the eastern sector. Stone circle KYR1–2 is located at the eastern edge of the western sector. The circle is in reality more square than circular. The orientation is roughly 25°, with each side measuring 1.5–1.9 m. The circle is made up of ten large stones, all of them protruding above the ground surface (by as much as 14 cm) and ranging in length (at ground level) from 50 to 75 cm. A total of 9 m2 was excavated by trowel, following the natural stratigraphy – this includes roughly 3 m2 within the circle and an area of about 6 m2 outside the circle. The natural stratigraphy consists of a uniform, loose, rootless topsoil layer whose depth ranges between 6 and 18 cm, and which lies on a deeper and uneven layer of small pebbles (gravel). Many of the stones that make up the circle appear to rest on this gravel layer. A total of 1328 whitened animal bone fragments ranging in size from 0.1 cm to 1.5 cm were recovered from the excavation. Parts of some of the fragments were bluish or blackened. The sediments were not screened and all of the bone fragments were picked up by hand when found. The majority of the bone fragments appear to lie on top of the gravel layer, although some were also recovered down to a depth of 8 cm in that layer. Owing in part to the highly fragmentary and altered nature of the bone fragments, it is not possible to determine whether their distribution reflects any significant patterning. The fragments were, however, concentrated near the centre of the circle. No bone fragments or artefacts were recovered from the area outside the circle. The gravel layer within the circle was excavated to a depth of about 35 cm (below the ground surface), at which point the matrix was found to consist of greyish sterile loose clay.
218
F. ALLARD, D. ERDENEBAATAR & J.L. HOULE
Stone circle KYR1–10 is located within the western sector of the stone circle area, about 80 m west of KYR1–2. It is also more of a square with rounded corners than an actual circle. The orientation is roughly 6°, with each side measuring 1.7–1.8 m. The circle is made up of 11 large stones, most of them protruding above the ground surface. A few stones appear to be missing from the north-east section of the circle. A similar excavation trench and technique to that used for KYR1–2 revealed a gravel layer beginning at a depth ranging between 8 cm (near the edge of the excavation trench) and 16 cm (near the middle of the circle). The stones, which range in height from 13 to 29 cm, are embedded in the gravel layer, in some cases by as much as 12–15 cm. A total of 33031 whitened animal bone fragments ranging in size from 0.1 cm to 4.0 cm were recovered from the excavation. A large number of bone fragments appear to lie on top of the gravel layer, although many are present within that layer as well to an additional depth of about 4 cm near the centre of the circle and up to 16 cm near the stones. Again, no spatial patterning was apparent. Although bone fragments are clearly concentrated in the centre, they are found throughout the circle (some of them next to the stones), as well as beneath some stones (fewer than 20 were found beneath the stones of the northern side) and even outside the circle (especially in the north-east sector of the trench, where the absence of stones could have permitted the movement of bone fragments into that area). Measuring 1.6 m in diameter, stone circle KYR1–6 is located along the western edge of the eastern sector of the stone circle area. Seven of the eight larger stones that make up the circle protrude from the ground surface, one of these by 12 cm. A total of 4 m2 was excavated by trowel, including 1.8 m2 within the circle. Most of the few hundred fragments of animal bone – some partly bluish or blackened – and charcoal recovered during the excavation were found on the surface of the gravel layer (lying at a depth similar to that of the other two circles), although some are present within that layer as well. White, bluish and dark burnt bone fragments were found throughout the circle. A large amount of dark burnt material was found in the south-west quadrant beside one of the stones at a depth of about 6 cm in the gravel layer. A cursory analysis of a sample of diagnostic bone fragments recovered from stone circle KYR1–10 suggests that they may represent cattle, sheep/goat and possibly horse. The small size of the sample and the highly fragmentary nature of the bone pieces underscore the need to carry out a more thorough analysis of a larger sample before we are able to determine the animal species repre-
11. ARCHAEOLOGICAL RESEARCH IN THE KHANUY RIVER VALLEY
219
sented. The colour of the bone fragments suggests the burning of bones at very high temperatures, with the whiter pieces heated to higher temperatures than the darker fragments. The absence of clear evidence of burning or of any digging within the stone circles themselves, combined with the highly scattered distribution of the bone fragments, together suggest the following possible scenario. Bones of various animal species were first cremated at a location outside the circle. The resultant small fragments were then collected and scattered within the circle. It is possible that at the time the (apparently ritual) activity was performed, the stones lay on the ground surface and that no effort was made to cover the bone fragments, with wind dispersing any ash present and subsequent slow sedimentation leading to the gradual burial of the bone fragment layer and the bases of the stones. Other Structures. The only structure identified as part of the site but located beyond the stone circle sector is a stone mound situated 20 m east of the eastern edge of the sector. The mound’s diameter is 10 m.
Khirigsuur KYR2 The khirigsuur KYR2 is located at N48°05’30” E101°02’56”, a few hundred metres east of the valley’s western foothills, about 600 m west-south-west of Urt Bulagyn, and almost 3 km west of the Khanuy river (Fig. 1). Smaller than, but structurally similar to, Urt Bulagyn, KYR2 measures approximately 140 m (north-south) × 130 m west-east (Fig. 5). No excavations have yet been carried out at KYR2. The central mound measures 16 m in diameter at ground level. The mound and plaza are enclosed within an irregularly shaped stone fence whose dimensions are roughly 37 × 57 m. Four stone mounds ranging in diameter from 3.0 to 3.4 m are located at the corners of the stone fence perimeter. One interesting similarity with Urt Bulagyn is not readily apparent. At both sites, the eastern fence is 8% longer than the western fence. The stone mound area consists of 110 mounds, which range in diameter from 1.3 (a ‘satellite’ mound) to 5.0 m. The largest mounds are distributed along the eastern fence. The stone mound area displays a number of similarities with Urt Bulagyn. First, the stone mounds are concentrated – in this case exclusively – along the northern, eastern and southern sides of the stone fence perimeter. Second, the shape of the stone mound area, combined with the placement of the fence, suggests a site orientation of 283°, which is almost identical to the
220
F. ALLARD, D. ERDENEBAATAR & J.L. HOULE
Fig. 5. Mound KYR 1–1 at Urt Bulagyn. Horse skeletal remains in situ (photograph by Francis Allard).
orientation of Urt Bulagyn (285°). Third, many rows of mounds display a south-east to north-west orientation. Fourth, a number of mounds in the southern sector appear to be accompanied by single small ‘satellite’ mounds located to their immediate south-west. A total of 110 stone circles have been identified in the ring-shaped circle area, with others probably hidden from view or disturbed by the marmot burrows in the eastern sector. The circles, which range in diameter from 0.8 to 2.1 m, appear to be distributed in two complete rings and one incomplete ring (along the western side). While the high degree of structural similarity between Urt Bulagyn and KYR2 is easily recognised, we also note the apparent absence at KYR2 of a path to the east of the central mound, a long northern path, ‘horns’ extending from the central mound, and slab burials.
11. ARCHAEOLOGICAL RESEARCH IN THE KHANUY RIVER VALLEY
221
Concluding Thoughts and Future Directions The first field season recovered sufficient data to permit the detection of a number of behavioural and spatial patterns, as well as the formulation of hypotheses to be tested over the course of future field seasons. To begin, all of the evidence collected so far points to the likely ritual and funerary nature of the stone built sites in the valley and their association with a nomadic pastoralist system of subsistence. Only human burials and apparently ritual deposits of animals have been identified to date at the sites, while no surface traces of permanent or even temporary habitations have yet been located anywhere in the valley (except for the remains of more recent sites). If, as Mongolian archaeologists suggest, every small mound at the khirigsuurs is likely to contain selected anatomical parts of horses, we are faced with the intriguing possibility that such behaviour may lie at the root of the horse’s continued widespread use and near reverence by Mongolia’s mobile pastoralists. The absence of agricultural tools in burials is further indication that the khirigsuurs were built and used by early nomadic pastoralists. The single radiocarbon date placing Urt Bulagyn around the turn of the 1st millennium BC is also consistent with the chronology proposed by many scholars for the spread of nomadic pastoralism in the eastern Eurasian steppes. The elucidation of the khirigsuurs’ ritual function(s) will require the collection of more datable material and the excavation of other structures at the sites, including the long path located in the northern sector of many khirigsuurs. It is also possible that the different sectors of each khirigsuur were constructed at the same time. Owing to its associated error interval, radiocarbon dating might not be helpful if the period of construction did not exceed one or two generations. Experimental labour studies might help us calculate the effort expended in building a large khirigsuur such as Urt Bulagyn. Whatever that value might be, it is likely that construction required a concerted and sustained effort by a large number of people. This in itself is intriguing since that level of energy expenditure and labour recruitment might appear incompatible with the low population densities and significant mobility of domestic units that characterise nomadic pastoralism. If, as suggested, the builders of the khirigsuurs were mobile herders, how were leaders able to recruit a sufficiently large labour force to help build such large sites? While the recovery of weapons from slab burials points to the existence of conflict during the Bronze Age, it is unclear whether coercion played any significant role in the recruitment and management of the labour force needed to build the
222
F. ALLARD, D. ERDENEBAATAR & J.L. HOULE
khirigsuurs. As can be seen in Fig. 3, the orientation of the slab burials contrasts with that of the khirigsuurs and it has been suggested that they are the products of a later period of occupation of the valley. The significant nondomestic component clearly evident at the khirigsuurs, along with the fact that weapons and the products of economic or trading activity are typically absent from their central mounds, together point to the possibility that ideology and its attendant rituals played a crucial role in promoting social cohesion and concerted action at the time the khirigsuurs were built, whether such efforts resulted from fear of a leader widely recognised as a god or as the principal conduit to the gods, or from strongly held religious beliefs shared by members of a more egalitarian society. A comparison of the khirigsuurs Urt Bulagyn and KYR2 may shed light on some of the issues discussed above, as well as help clarify the details of ritual behaviour at the time of their use. To begin, a look at the site maps clearly brings out the close structural similarity between the two khirigsuurs, suggesting the existence of shared ideas about the appropriateness of a distinct ground plan and ritual programme. KYR2 appears to be a smaller version of Urt Bulagyn, without the slab burials (which in any case may not have been contemporary with the main portion of the site), northern path, or structures to the east of the central mound. Although we do not yet have precise data to support this, our brief reconnaissance of the research area was able to identify widespread similarities among all the khirigsuurs, something we plan to test more systematically over the course of the next field seasons. One further prominent shared feature is the orientation of the sites. Not only do Urt Bulagyn and KYR2 display an almost identical orientation, the other sites in the valley also appear to be oriented toward the west-northwest, or east-southeast (depending on which side is considered the ‘opening’). This is highly significant since the khirigsuurs are clearly not pointing toward any single prominent feature in the landscape. The similar orientation of another khirigsuur in northern Mongolia again suggests that the sites in Khanuy valley were not built to face the western or eastern side of the valley. The only remaining explanation is that the khirigsuurs are oriented toward some celestial body or event that displays regularity. Site plans of other khirigsuurs will permit us to clarify further the range of orientations displayed by the sites in the research area, while we also investigate their likely astronomical associations. The remarkable consistency in spatial organisation and ritual practice displayed by the khirigsuurs, along with their high visibility from afar, are typi-
11. ARCHAEOLOGICAL RESEARCH IN THE KHANUY RIVER VALLEY
223
cal elements of established religions. However, this consistency also presents the possibility that the khirigsuur phenomenon may in fact have been relatively short lived. In contrast to the 16th–7th century BC range proposed by Mongolian archaeologists, one of us suggests a period of construction and use of no more than 200 years, since it might be argued that greater variability in both structure and practice would be expected over a longer time span. If the period of khirigsuur construction and use was in fact relatively short, one can also propose that this is what we might expect from a system characterised by power relations and collective action driven by ideology rather than by coercion and/or economic incentives. Whether religious power is believed by the populace to reside in individual leaders or channelled through a specialised priesthood in a non-hierarchical society, such a system is always at risk of being destabilised once belief becomes compromised by an accumulation of unfulfilled promises and predictions. Most certainly, the next step will be to date chronometrically more sites so that the hypothesis of a short-lived khirigsuur phenomenon may be put to the test. It is interesting to note the spatial dimensions of the khirigsuur phenomenon. The limited carrying capacity of the land makes it all too clear that the sites are situated too close to one another for each to be the centre of a competing polity and demographic base. Possibly, the spatial patterning displayed by the khirigsuurs – with regard to location, structural complexity and size – reflects kinship ties among a ruling elite, with continued building activity throughout the region associated with the funerary and ritual demands of generational succession rather than the appearance of new competitors. The full coverage survey of the research area that we plan to conduct over future field seasons will permit the further clarification of spatial patterning among all types of sites, information that we can use to try and understand the details of the socio-political structure during the Bronze Age. Finally, the high density of labour intensive sites in the valley does bring up the question of where the workers required to build the khirigsuurs lived, let alone the fact that we should expect the presence of temporary habitation sites even without the existence of large stone built constructions. The identification of campsites and other settlements is also important as a means of providing essential insights into the valley’s early demographics, since population pressure on limited resources – itself possibly the result of climatic change – might certainly be a central factor in the emergence, development and decline of the khirigsuur phenomenon. Khirigsuurs may have served as territorial markers on
224
F. ALLARD, D. ERDENEBAATAR & J.L. HOULE
preferred land, while rituals and religious leadership may have emerged as a means of countering the negative impact of an increasingly large population and/or worsening climate. The few interviews we have conducted so far with local families have helped us better understand the potential impact of flooding and erosion on the preservation of ancient campsites located on the valley floor. We are also learning about present-day land-use patterns in the valley, information that will prove invaluable when we design a field methodology designed to locate the remains of campsites and estimate the size of the population at the time of the khirigsuurs. Possibly, our investigations into settlement patterns may even reveal the presence of a semi-nomadic sector of society, one that could have been more closely associated with the flowering of the khirigsuur phenomenon than fully mobile herders.
Bibliography Allard, F., Erdenebaatar, D., Batbold, N. and Miller, B. 2002: A Xiongnu cemetery found in Mongolia. Antiquity 76, 637–38. Erdenebaatar, D. 2004: Burial materials related to the history of the Bronze Age on the territory of Mongolia. In Linduff, K.M. (ed.), Metallurgy in Ancient Eastern Eurasia from the Urals to the Yellow River (Lewiston NY), 189–222.
III. NEW DIRECTIONS IN THEORY AND PRACTICE
CHAPTER TWELVE BETWEEN THE STEPPE AND THE SOWN: PREHISTORIC SINOP AND INTER-REGIONAL INTERACTION ALONG THE BLACK SEA COAST ALEXANDER A. BAUER
Introduction: The Black Sea as a Periphery This paper is neither about the steppe nor the sown. Rather its focus is on one of the areas that lies between these two landscapes: this is the large body of water known as the Black Sea, a region largely considered marginal to those surrounding it. The reasons for this ‘marginality’ are numerous, the first being the fact that, following the boundary conventions between modern states, archaeological boundaries assigned to past inland cultures are often assumed to extend to the coast, with bodies of water such as the Black Sea then acting as dividing lines. When the cultures of peoples located along these coasts differ from that of their inland counterparts, as they often do, these differences are usually explained as ‘peripheral variants.’ A second reason for the marginal treatment of the Black Sea is due to its geopolitical situation: throughout almost the entire history of archaeology as a discipline, a series of political standoffs has made the Black Sea a physical and intellectual boundary line between the regions surrounding it. A third reason may be attributed to the biases of scholarship: there is a tendency for archaeologists and historians to regard the Black Sea as a backwater of the Mediterranean, whether it is said explicitly (for example, by Braudel 1972 [1949]), or implicitly by analyses that focus on influences external to the region (Hiller 1991; Tsetskhladze 1994). These and perhaps other reasons have prevented the Black Sea from being treated as a unit of analysis in its own right (Özveren 2001). But identities in the past as well as now do not always follow the lines drawn by modern states (Anderson 1983; see also Curtin 1984). Water may be considered a natural boundary, but more often than not it facilitates, rather than inhibits, interaction. This perspective suggests then that we shift our understanding of maritime regions, and consider water as connecting rather than dividing.
228
ALEXANDER A. BAUER
There are historical reasons, too, to regard the Black Sea as a ‘world’ unto itself. The north–south divide between the Russian and Ottoman empires (and the Soviet Union and NATO-aligned Turkey in the 20th century) often masked underlying processes of coherence across the region, that did nevertheless emerge from time to time as power shifted between the two powers (see Özveren 1997). And although in the earlier, classical periods, the region’s structure was controlled by Mediterranean powers, we cannot forget that it was maritime access that enabled them to infiltrate the region, and the area under control traced the Black Sea littoral. In spite of these indications that there may be some coherence to the Black Sea world as a whole, the region has nevertheless been neglected in archaeological analyses. The important studies that have been done almost all begin with the period of Greek colonisation (for instance, Tsetskhladze 1996), a time when the region is generally regarded ‘peripheral’ to and acted upon by dominating influences from the south. Comparatively little research has focused on the earlier periods in the Black Sea, and those areas that have been studied are often defined in terms of their inland counterparts: their inhabitants are thus, unsurprisingly, regarded as ‘cultural variants’. In the period I am focusing on in this paper, the Early Bronze Age, the ‘variants’ that appear in Ukraine are the Usatovo and Kemi-Oba groups (Manzura 1994; Zbenovich 1973), while in the Caucasus, the Dolmen culture may be awarded that distinction (Markovin 1997; Trifonov 1994). In the Balkans, the sites of the Varna lakes region to the north and Burgas bay to the south are considered to display traits distinguishing them from inland ‘type’ sites (Nikolova 1995; Tonœeva 1981). And at √kiztepe on the Turkish coast, much of the material is called ‘early Hittite’, and the site itself has been identified as Zalpa, the capital of the Kashka, named in the Hittite textual sources as an unusual tribe located in these northern regions (Alkım et al. 1988; Gurney 1992; Macqueen 1980). The neglect of the prehistoric periods is particularly acute along the Black Sea coast of Turkey, where √kiztepe represents the only systematic excavation to have taken place (Alkım et al. 1988; Bilgi 1994; 1998). Ceramic typologies for the Turkish Black Sea region reflect this situation and divide the pottery phases into Chalcolithic, Early Bronze, Archaic (colonisation period), etc. (Burney 1956; French 1991). The pre-Greek hand-made coarse wares ubiquitous in the region are all conventionally labelled ‘Early Bronze’ presumably based on its very general similarity with EB pottery from elsewhere in north-western and central Anatolia (see Yakar 1975). What this has resulted in is a typological
12. PREHISTORIC SINOP AND INTER-REGIONAL INTERACTION
229
‘gap’ of about 2000 years, leading some scholars to suggest that the region was uninhabited during this time (Burney 1956, 181–2; I¤ın 1998).1 Such a scenario seems highly unlikely, however, as the regional environment is extremely hospitable: it easily supports a diverse economy and may be dry-farmed. Moreover, the region is rich in natural resources, such as timber (Doonan 2002). Even at √kiztepe, the establishment of a chronology has been difficult (see Thissen 1993; Yakar 1975), and most of the pre-Classical material is dated to the Early Bronze Age. And while this site has produced some very important information regarding the Black Sea region in prehistory, its location at the mouth of the Kizilirmak, the main river running from the central plateau through the Pontic mountain chain into the Black Sea, connects it to the inland regions more readily than sites in more isolated areas along the coast. Being thus uniquely situated, not to mention its being the only site investigated in the region, it cannot by itself satisfactorily explain the history of settlement in the Black Sea region as a whole. Recent research in the vicinity of Sinop, just west of √kiztepe, has ameliorated the situation somewhat. In 1996 the Black Sea Trade Project began research in that region to investigate patterns of maritime exchange in the Black Sea and the economic relationships of its ports and hinterlands (Hiebert et al. 1997a; 1998). It is also aimed at developing a systematic and integrated multi-component research program across terrestrial, shallow and deep-water environments (Ballard et al. 2001; Hiebert et al. 1997b). One of these components is the Sinop Province Regional Survey, which over four seasons from 1996 to 1999 was focused on documenting the settlement history of the Sinop region from the inland valleys and mountains to the sea, by employing both extensive reconnaissance of its numerous ecological zones and intensive techniques of systematic survey in selected zones (Doonan et al. 2001; Doonan 2004). Through these investigations, some 74 pre-Classical sites were documented. A preliminary study of the ceramics from these sites was aimed at developing a typology based on ware and technology, in addition to macroscopic features such as form and decorative style (Bauer 2001). Results of this work suggests
1
Even I¤ın (1998, 110), who acknowledges the presence of a Middle Bronze Age in the Sinop region, following the discoveries of such at √kiztepe, nonetheless continues to postulate a gap from the 18th to the 8th century BC.
230
ALEXANDER A. BAUER
that the prehistoric Sinop material is for the most part conservative and local in character, and is quite distinct from contemporary material from inland on the Anatolian plateau. Interestingly, its closest parallels may be found in the √kiztepe assemblage, as well as in pottery traditions across the Black Sea. How are we to interpret this unusual situation? Are we dealing with simple trade, the movements of peoples, or something harder to define? As our knowledge of the prehistoric Black Sea is limited, to begin answering these questions requires a more careful analysis of the material that we do have. Naturally, any conclusions that can be drawn must be seen as provisional and requiring further investigation. This paper, then, has two parts. First, I will enumerate the prehistoric pottery types found in the Sinop region, offering comparisons to other Black Sea material where appropriate. Following this, I will discuss the general picture that emerges, and consider whether the current state of trade studies in archaeology can provide an adequate framework for interpreting the prehistoric situation in the Black Sea.
Sinop Pottery in Prehistory The northernmost point of Anatolia, the Sinop promontory is a region of rolling hills and agricultural plains that juts out into the Black Sea halfway along its southern shores (Fig. 1). Rainfall is plentiful and the climate is moderate – more like the eastern European shores of the Black Sea than the dry, steppe environment of the central Anatolian plateau. The region is thus densely vegetated, home to deciduous oak and pine forests that have been famed since antiquity (Doonan 2002). This region, rich in agricultural and timber resources, is effectively cut-off from the central Anatolian landmass and from points east along the Black Sea coast by the rugged Pontic mountain chain. While the mountains reach heights of 4000 m only in the far south-eastern part of the coast, the steep edge of the plateau is cut by few natural river valleys and intermontaine passes, thus making movement difficult. An ethnographer working in the region noted that villagers living on the northern slopes practically live in a different world from those on the southern side, and ‘[i]n some areas a complete change occurs within a few miles in the design and materials used in house building, the style of peasant dress, agricultural techniques and field usage, village settlement patterns, accents, kinship terms, and many other details of peasant life’
12. PREHISTORIC SINOP AND INTER-REGIONAL INTERACTION
231
Fig. 1. Map of Sinop region and Black Sea sites mentioned in the text (after Doonan et al. 2001; base map by A. Gantos).
(Meeker 1971, 319). Moreover, the Pontic mountains often extend right up to the coastline, restricting movement east and west along the shore. This situation makes communities located along the southern shoreline like Sinop effectively islands along the Black Sea coast, most accessible by sea. This situation undoubtedly had a significant effect on the region’s historical and cultural identity, as M.I. Maksimova (1951) noted over 50 years ago. In a preliminary study, all of the ceramics from the Sinop Province Regional Survey that could be identified as pre-Classical were analysed. Due to the typological problems discussed earlier, this meant that they could be dated somewhere within the five millennia between the Ceramic Neolithic and the Iron Age. One of the primary aims of the ware study thus was to discern any chronologically relevant groupings among the assemblages. Of course, an absolute chronology for this material will not be possible without the results of more comprehensive stratigraphic investigations. Aside from chronology, factors determining assemblage
232
ALEXANDER A. BAUER
variation may include any combination of economic, environmental and social processes (Arnold 1985; Rice 1984; Skeates 1998; van der Leeuw 1991; Yentsch 1991). Through macroscopic examination it was possible to define eleven distinct wares, which may be grouped into four more general types: chaff and mineraltempered wares, which appear across all sites and phases; dark-burnished wares, which seem to characterise earlier parts of the sequence; buff wares, which seem to occur during later phases; and shell-tempered wares. Ongoing study is aimed at confirming the uniformity of these types and more fully understanding the techniques employed in their manufacture. The pre-eminent feature of the ceramic assemblage of the Sinop region is its distinctly local character. While some broader connections can be made to the traditions of elsewhere in Anatolia and the Black Sea, the pottery generally seems to be conservative in both ware and vessel form. This is undoubtedly one of the reasons that a clear chronological sequence has been so difficult to establish. Common forms include holemouth pots, shallow bowls and large, flaringrim jars. Such shapes are not very distinctive, and variation in these forms seems to be as great within sites as among them. This may indicate that ceramic production is not only regional, but local from site to site, although the presence of a large quantity of kiln debris at one site (Mezarlıktepe) may suggest that specialised production was occurring there. Decoration is rare, usually limited to surface treatments such as slip and burnish. While clear chronological indicators are lacking in the assemblage due to this regionalism, the appearance of a variety of horizontal handles places the majority of these sites chronologically in the Bronze Age. The earliest period represented in the Sinop regional ceramic sequence appears to be the Chalcolithic, as exemplified by dark burnished ceramics similar to wares elsewhere in Anatolia (see also Thissen 1993). At one of the surveyed sites, Mezarlıktepe, incised sherds similar to pottery from Ilıpınar VA in the Marmara region (see Roodenberg 1999) and Yarımburgas Cave (level 3) in Turkish Thrace (see Özdoªan et al. 1991) may indicate an early Chalcolithic (late 6th millennium BC) date for some material (Fig. 2, B). The later Chalcolithic (late 5th/early 4th millennium) is better attested to at sites like Maltepe (Fig. 2, C) where there appeared an abundance of an unusual oystergrey burnished ware (also noted by Burney 1956, 183), which bears some similarities with the so-called ‘Büyükkaya Ware’ of this period (Parzinger 1993). Here and at the site of Kıran Tepe (Fig. 2, A) some of the forms also show
12. PREHISTORIC SINOP AND INTER-REGIONAL INTERACTION
233
affinities with types from Chalcolithic √kiztepe and Büyük Güllücek, most notably a variety of ‘Karanovo type’ horn handle. While it is difficult at this point to clearly define a Chalcolithic phase in the Sinop region, preliminary indications such as these do seem to place Sinop within a broader tradition that, as at √kiztepe, looks west to Turkish Thrace and the Balkans, possibly resulting from coastal contact (Makkay 1993; Özdoªan 1993; Thissen 1993). Interestingly, the white painted ceramics noted in the Samsun region by Thissen (1993) and others have not yet been discovered in our survey of Sinop, although this could be a product of the weathered nature of surveyed ceramics.2 The Early Bronze Age seems much better represented in the Sinop region, with horizontal handles and slipped and burnished wares appearing in abundance. The one excavated prehistoric site in the region, Kocagöz Höyük, which had been investigated briefly by Akurgal and Budde in the 1950s (Erzen 1956, 71–2), produced a distinctive highly polished black pottery, sometimes decorated with incised lines, filled with white paint (Fig. 3). While parallels for some of these techniques can be found at √kiztepe, it is unusual on black-polished sherds and suggests contacts with Troy I–II in north-west Anatolia (Erzen 1956, 72), Karanovo VI–VII and Sitagroi IV–V (Bailey 2000, 251–3), rather than central Anatolia. In fact, parallels in manufacture and forms can be drawn with Troy II pottery at the Hermitage Museum in St Petersburg, particularly the pieces that were found with the Troy II treasure hoards (F. Hiebert, personal communication). Similar ceramics, with white paint over incised decoration, can be found elsewhere in the Black Sea region as far north as Usatovo on the Ukrainian coast (Lazarov 1984, 65–6; Nikolova 1995; Rassamakin 1994, 47; Zbenovich 1973, 516), suggesting that seaborne contacts were widespread by this time (see also Anthony 1986, 299). This conclusion is strengthened by the appearance of askos-type juglets at Kocagöz, a type common at the Bulgarian site of Ezerovo, a settlement submerged in the Varna lakes on the Black Sea coast (Tonœeva 1981). These pots also bear some similarity to the Yortan ‘beakspouted’ jugs that are found in western Anatolia at this time (see Kâmil 1982), and which may be considered part of a Balkan-influenced assemblage. 2
Alkım et al. (1988, 174) mention the fact that the white paint is almost invisible, perhaps due to the passage of time. Thissen (1993, 222–5) futher discusses this and similar observations in an appendix, arguing that in fact the decoration was made that way. Whether we are just missing this type in Sinop or the lines have just weathered away still cannot be determined with any certainty.
234
ALEXANDER A. BAUER
Fig. 2. Chalcolithic pottery from the Sinop region. A: Kiran Tepe; B: Mezarliktepe; C: Maltepe (photographs by A. Bauer and O. Doonan).
Both the Early and Middle Bronze Ages are represented at the site of Güllüavlu overlooking the central Karasu valley (Fig. 4). This small mound, recently cut by bulldozers widening the Sinop-Erfelek road, produced similar polished, incised wares, as well as inverted-rim bowls similar to those of Troy I. Some more unusual pieces were found as well, including relief decoration similar to that on Trojan ‘face pots’, an impressed cylindrical piece that bears some resemblance to figurines from Usatovo (Zbenovich 1973, 515, fig. 1), and the foot of a larger anthropomorphic figurine. The appearance of finer Middle Bronze Age red-slipped wares, including a ‘teapot spout’ similar to those at Kültepe (Özgüç 1959, 61 and pl. XXXVIII), provide the first indication of contact between the Sinop region and central Anatolia.3 The fact that these items were 3
Although it should be noted that MBA wares have also been found in the Sinop region at the coastal site of Kö¤k Höyük in Gerze. While our survey found this site to contain buff wares common to the period, I¤in (1998, pl. 10.12–13) also recorded teapots of central Anatolian type similar to the one we recorded at Güllüavlu.
12. PREHISTORIC SINOP AND INTER-REGIONAL INTERACTION
235
Fig. 3. EBA pottery from Kocagöz Höyük, Sinop region (photograph by O. Doonan).
Fig. 4. EBA and MBA pottery from Güllüavlu, Sinop region (photograph by A. Bauer).
236
ALEXANDER A. BAUER
clustered at the top of the scarp left by the bulldozing activity may suggest, however, that they come from a single pit and do not necessarily indicate the presence of a Middle Bronze Age, central-Anatolian-related settlement. The best evidence for the Late Bronze and perhaps early Iron Ages, which are very difficult to define for this region, interestingly comes from two sites located closest to the coast. The first, Kö¤k Höyük, comprises the mound of eroding earth located beneath the lighthouse in the town of Gerze, south of Sinop. In a strikingly parallel situation, the second site was identified as an erosional scarp with several identifiable strata located on the kale of Sinop town itself, just below the bus station, and barely 50 m from the water. As previous studies have failed to identify a Late Bronze Age phase in the region, the situation of these two sites may support Özdoªan’s (2003) hypothesis that tectonic or sea level changes may have submerged or destroyed many of the settlements of this period. Moreover, since these disappearing sites are potentially very important, the Sinop bus station site was the subject of a short rescue operation in the summer of 2000 (see Hiebert et al. in preparation). Surveyed ceramics from the site included numerous hand-made sherds with an unusual rope-like decoration applied to the vessel exterior. Similar rope decoration frequently appears during the Middle Bronze to Iron Ages of the north Pontic steppe (Sava 1994, 153, fig. 6, 12–16), and might indicate continuing overseas contact in the period just prior to Greek colonisation. To summarise, then, the prehistoric material from the Sinop region is for the most part hand made, conservative and local in character. In all periods, however, it can be seen as part of a broader tradition connected to the Balkans and elsewhere around the Black Sea. During the earliest phases of Sinop’s ceramic sequence, connections look toward the Marmara region and Thrace, perhaps suggesting contact westward along the Turkish Black Sea coast. In subsequent periods contacts are more widespread – down the coast to Troy and Thrace, but also across the sea to points northward along the Balkan coast and the Pontic steppe. Only rarely are there clear parallels to central Anatolian types. This situation is presumably due to the location of this peninsula on the Black Sea coast, isolated from the Anatolian plateau by steep mountains. Interestingly, overall comparison of the Sinop material to that from √kiztepe in the neighbouring Samsun province reveals more differences than one might expect. For while they both display obvious Balkan features, the √kiztepe Early Bronze Age material maintains closer connections to central Anatolian assemblages (see
12. PREHISTORIC SINOP AND INTER-REGIONAL INTERACTION
237
Alkım et al. 1988, 196) than the Sinop pottery does.4 This may be the result of that site’s location at the mouth of the Kizilırmak (ancient Halys river), which provides the best north-south route through the Pontic mountains. It should be noted that there is no natural coastal road between Sinop and that region, so that the easiest connection between them would have been by sea. The most interesting development occurs during the Late Chalcolithic and Early Bronze Ages, when the region’s orientation shifts from being more internally directed to one focused towards broader Black Sea cultural traditions. Both the assemblages from Sinop as well as from √kiztepe generally look different from contemporary material from inland on the Anatolian plateau, and instead take as their closest parallels artefact types from coastal Bulgarian sites, such as Ezerovo II in the Varna lakes region (Tonœeva 1981). Connections can also be made with material of the Usatovo group in the north-west Pontic steppe (Zbenovich 1973). While the details of chronology have yet to be satisfactorily worked out for the many cultural ‘variants’ along the Black Sea coast, similar trends may be observed across the region. Scholars working in Bulgaria have noted that strong links begin to develop between coastal sites in the western Black Sea at this time (Leshtakov 1995). And evidence for connections along the northern Black Sea coast, between Usatovo and Maikop in the Late Eneolithic, and the Kemi-Oba and Dolmen cultures in the Early Bronze Age, is continually growing (Rassamakin forthcoming).
Inter-Regional Interaction and the Prehistoric Black Sea World The picture that emerges from this study is that the Black Sea may not always be best understood as a ‘periphery’, but in fact may have been a conduit for cultural interaction in different periods and at different scales. What are the 4 The early chronology of √kiztepe has been reassessed by Thissen (1993), who raises the dates of the material from mound II originally dated ‘EBA I’ at least two thousand years earlier (to the end of the 6th millennium BC [all dates calibrated], making it ‘early Chalcolithic’. He agrees that √kiztepe I, sounding A may properly be dated to the EBA (early 3rd millennium BC), as assumed by the excavators, but he also suggests that soundings C and F should be ascribed to the ‘late Chalcolithic’ or late 4th millennium BC. While I do not agree with all of his specific criticisms, his general reasoning for these revised dates seems sound, and I am accepting them for the present analysis.
238
ALEXANDER A. BAUER
implications of this for our understanding of the Black Sea during prehistoric and other periods? Following recent conceptualisations of culture as locatable within contexts of social interaction and discourse, I want to consider how trade, as a social activity, may itself engender new cultural forms. The idea is that increasing maritime trade and communication may have promoted the development of a distinctive ‘Black Sea culture’ (Hiebert 2001). To a certain extent, this intersects with some of the conclusions drawn by Chernykh (1992), although the correspondence is yet to be fully explored. To investigate this problem, however, requires not only a systematic analysis of Black Sea prehistory, but necessitates some rethinking about trade studies and the interpretation of material culture more broadly. Trade studies have a long history in archaeology, but few new archaeological perspectives on trade have developed since the 1980s. This is presumably due to trade’s close association with the classic ‘ecological’ models of positivist approaches: neighbouring cultures could be seen as one part of the ‘environment’ with which a given cultural system interacted in a measurable way (Chang 1975) and producing tangible results (Renfrew 1975). In the changing academic atmosphere, trade studies are passé, seeming to be a macroscopic and ‘static’ topic, unrelated to issues of contemporary interest such as agency and identity (Meskell 2002). But before we write it off as a topic that is no longer relevant to, or cannot be improved by, contemporary theorising, depending on one’s perspective, there are clearly social aspects to the practice of trade that have been overlooked – for one, it is fundamentally a social act (Bauer and Doonan 2002). And if we accept that trade is a social phenomenon, then its study can be enriched by recent approaches that focus on the contexts within which social action, such as trade, is played out. While any given trade interaction occurs within a unique context, each interaction is only one part of a longer chain of meaning. When we choose to interpret material remains as ‘imported’ or ‘trade items’, we are choosing to highlight or focus on certain moments or kinds of meanings at the expense of others (Bauer 2002). But this ‘contextual embeddedness’ is not limited to our academic enterprises but is what allows for all meaning to be interpreted and communicated (Preucel and Bauer 2001). Transformations in material culture meanings result when conflicting value systems are brought into harmony through the negotiation of value between two or more contexts. But, as Appadurai (1986a) and others (Fotiadis 1999; Kopytoff 1986; Bloch and Parry 1989; etc.) have
12. PREHISTORIC SINOP AND INTER-REGIONAL INTERACTION
239
pointed out, we must not simply oppose ‘utilitarian’ and ‘symbolic’ – meanings that do not neatly fall into such categories (see also Baiburin 1997). Objects convey multiple meanings, and these meanings pattern and are patterned by our ongoing engagements with them. Thus we must develop an approach to material culture meaning that recognises this variability while at the same time can be successfully applied to the study of past economic and spatial relationships among communities (Preucel and Bauer 2001). Classic studies of trade focus on questions regarding the presence of imported materials across a site or region. For the most part, these kinds of studies focus on indexical meanings by attempting to document inter-regional trade and interaction by interpreting some artefacts in an assemblage as ‘foreign’. Studying trade like this is certainly one way to investigate the development of broader Black Sea maritime interconnections. When we look at the material, however, it is clear that evidence for such trade is slight. In the prehistoric pottery assemblage from the Sinop region, there is nothing yet to suggest the presence of wares actually imported from elsewhere in the Black Sea. On the contrary, the presence of kiln debris and wasters in the Bronze Age forms and clay fabrics provides some evidence for local ceramic production. While more thorough analytical work is yet to be done, preliminary indications point to a local origin for most if not all of this pottery. There are, however, many kinds of indexicality, all of which can reveal something about social relations. Thus a second way to investigate interregional connections would be to look at material culture features that are shared among neighbouring groups. This type of diffusionist analysis focuses on the distribution of specific cultural traits across a region and takes style instead as signalling allegiance to a community or group (see Childe 1925; Dixon 1928). But the utility of this approach for the study of cultural interaction and trade is obviously limited. It can only be used to suggest the most general linkages among regions, lest we fall into the trap of making equations between pots and people like those of the ‘Bell-Beaker’ culture and, perhaps more pertinent to this context, the ‘Pit-Grave’ culture. What this kind of interpretation may address, however, is the lesser claim of signalling cultural orientation. And it is this sort of meaning that guided my thoughts about the development of a Black Sea ‘culture’ (and I use that term advisedly) in the introduction. For in the Early Bronze Age of Sinop, it is clear that the material is much more closely aligned with other Black Sea traditions and is distinctly separate from those of central Anatolia.
240
ALEXANDER A. BAUER
But is there a way to investigate cultural connections without making simplistic linkages? One way might be to shift our focus to ‘communication’ rather than ‘trade’ as the proper object of inquiry in studies of inter-regional relationships. Anthropologists examining the importance of discourse to community building point out that the exchange of objects may be secondary to the circulation of the idea of trade, the idea that social relations exist, and actual trade may need to occur only occasionally to maintain the circulation of discourse about it (Urban 1996, 162). The movement of less tangible commodities, such as information and technological knowledge (Kohl 1987; Wobst 1977), may be as important here. And we must consider that the interaction itself may produce its own culture and cultural traditions (Artzy 1997; 1998; Sherratt 1998). In short, the movement of objects is only phenomena, and does not necessarily reflect the reality of the underlying situation. Thus we must consider ways to investigate connections such as information exchange that do not leave the archaeological patterns that the trade in tangible commodities does. In this regard, a fruitful way forward may be to investigate the process of pottery manufacture. Ceramic studies based on manufacturing techniques, in addition to form and decoration, enable us to go beyond the one-dimensional classifications commonly used for the region, and which over-emphasise differences among assemblages (see Vandiver 1988). Analysis of these stages of production, tempering, forming, firing and also methods of decoration, allows us to get at the ‘habitual’ action of the manufacturing process, the steps of decision-making and motor movements (see Arnold 1981; 1985, 233 ff.; Reina and Hill 1978; Spier 1967), which may be considered socially-learned and habitual practices of an individual (see Daniel 1984; Singer 1984; see also Dietler and Herbich 1998). Similarities in these ‘habits’ as well as in any idiosyncratic aspects of the manufacture between Sinop and other Black Sea regions (and not with other regions) may suggest that even if pots themselves are not being traded, information is being exchanged and broader community bonds are forming on some level. Analyses along these lines may thus provide an alternative way to assess similarities among Black Sea ceramics and investigate the movement of information, ideas and behaviour, which may be considered the ‘value added’ meanings of objects (see Appadurai 1986b; Kopytoff 1986; Orser 1992): in other words, such objects may become ‘valuable’ for the social meanings embedded within them, and because of the social relations that they help facilitate among communities. While style is typically taken as the prime indicator
12. PREHISTORIC SINOP AND INTER-REGIONAL INTERACTION
241
of these social values in pottery studies (Weissner 1989) an approach that draws on multiple kinds of evidence, including technology, provides for a more holistic perspective. Some preliminary observations along these lines may be made, although systematic work needs to be done. The prehistoric Sinop ceramics are hand made, and were probably constructed using a coil-building technique. The extensive use of shell temper, burnish, and painted, incised decoration in the Late Chalcolithic and Early Bronze Age assemblages relates to its situation in the Black Sea region, and connects it to a broader tradition of dark-burnished and incised shell-tempered wares found in coastal Bulgaria (Draganov 1995; Leshtakov 1995; Tonœeva 1981), at Usatovo and Kemi-Oba sites in coastal Ukraine (Rassamakin 1999; Zbenovich 1973), and in parts of the Caucasus (Markovin 1997; Trifonov 1994). While firing temperature and atmosphere have yet to be determined, it is clear that some of the pottery was fired in distinctive ways as well. In his argument for Balkan-Anatolian connections across the Black Sea, Thissen (1993, 214) dwelled at length about the firing technology that produced a distinctive red and black burnished pottery which he considered ‘local’ to the western and southern Black Sea coasts. While I have no doubt that the production itself was local, the technique’s connection to the characteristic pottery of the Kura-Araxes region to the east is unmistakable (Kushnareva 1997; Sagona 1984), and seems to indicate the presence of broader connections along the Black Sea coast to the southern Caucasus.
Conclusions Situated as it is between the well-studied regions of the Near East and Europe, the Mediterranean and the steppe, the Black Sea has been surprisingly understudied. This is particularly the case when it comes to the prehistoric periods, when for many of the coastal areas, barely a ceramic chronology has been established. Recent work in the vicinity of Sinop, Turkey has been aimed at remedying this situation in some small way, and the picture that is emerging from these studies is an interesting one. The pottery traditions are quite distinct from those of neighbouring regions, and at least during some periods, they are closer in many respects to traditions across the sea. Yet at the same time, evidence for the existence of imported materials, usually the hallmark of inter-regional
242
ALEXANDER A. BAUER
trade, is lacking. Should we thus conclude that there was no such interaction, or might there indeed be connections, but just not the sort determined by traditional analyses? Thus it is here suggested that a broader understanding of trade as a social process might help us to understand better these connections and the shifting maritime orientation of Black Sea communities, such as that of the Sinop region at the very beginning of the Bronze Age. After all, trade in its various forms has long been recognised as an important factor in establishing and maintaining social relationships among groups and individuals (Lévi-Strauss 1969 [1947]; Malinowski 1922; Mauss 1990 [1950]; Weiner 1992). And in one of the few recent discussions of trade, Renfrew (1993, 9) remarks that ‘[i]f we seek to gain an insight into the range of interactions, it is more important to do so under the rubric of “interaction” than of “trade,” since the underlying motivation and functional role may not primarily be the acquisition of goods.’ Indeed, for the Black Sea, one productive avenue of future research would be to consider the role of fishing communities and the knowledge sharing such communities facilitate (see Knudson 1995; McGlade and McGlade 1989). Future work on the Black Sea coast of Turkey is aimed at sharpening our understanding of both the prehistoric Black Sea and of the varied nature of its inter-regional relationships. Between the steppe and the sown lies the sea. And it is only through more careful study that we may come to understand the potentially important role it played in the interactions between the Eurasian steppe and the settled lands of the greater Near East.
Acknowledgments This paper developed out of preliminary investigations for my dissertation, Fluid Communities: Interaction and Emergence in the Bronze Age Black Sea (Department of Anthropology, University of Pennsylvania 2006). I would like to thank Owen Doonan, Fred Hiebert, Bob Preucel and Pam Vandiver for their guidance and support during various stages of my research.
12. PREHISTORIC SINOP AND INTER-REGIONAL INTERACTION
243
Bibliography Alkım, U.B., Alkım, H. and Bilgi, Ö. 1988: √kiztepe I (Ankara). Anderson, B. 1983: Imagined Communities: Reflections on the Origin and Spread of Nationalism (London). Anthony, D.W. 1986: The ‘Kurgan Culture’, Indo-European Origins, and the Domestication of the Horse: A Reconsideration. Current Anthropology 27, 291–313. Appadurai, A. 1986a: Introduction: Commodities and the Politics of Value. In Appadurai (ed.) 1986, 3–63. —— (ed.) 1986b: The Social Life of Things: Commodoties in Cultural Perspective (Cambridge). Arnold, D.E. 1981: A Model for the Identification of Non-Local Ceramic Distribution: A View from the Present. In Howard, H. and Morris, E. (eds.), Production and Distribution: A Ceramic Viewpoint (BAR International Series 120) (Oxford), 31–44. ——. 1985: Ceramic Theory and Cultural Process (Cambridge). Artzy, M. 1997: Nomads of the Sea. In Swiny, S., Hohlfelder, R. and Swiny, H.W. (eds.), Res Maritimae: Cyprus and the Eastern Mediterranean from Prehistory to Late Antiquity (Alpharetta, GA), 1–16. ——. 1998: Routes, Trade, Boats and ‘Nomads of the Sea’. In Gitin, S., Mazar, A. and Stern, E. (eds.), Mediterranean Peoples in Transition: Thirteenth to Early Tenth Century BCE (Jerusalem), 439–48. Baiburin, A. 1997: The Functions of Things. Ethnologia Europaea 27, 3–14. Bailey, D. 2000: Balkan Prehistory (London). Ballard, R.D., Hiebert, F.T., Coleman, D., Ward, C., Smith, J.S., Willis, K., Foley, B., Croff, K., Major, C. and Torre, F. 2001: Deepwater Archaeology of the Black Sea: The 2000 Season at Sinop, Turkey. American Journal of Archaeology 105, 1–17. Bauer, A.A. 2001: The Prehistoric Pottery of Sinop Province, Turkey: Observations on Pre-Greek Interaction in the Black Sea. Paper presented at the Annual Meeting of the Archaeological Institute of America, San Diego. ——. 2002: Is What You See All You Get? Recognizing Meaning in Archaeology. Journal of Social Archaeology 2, 37–52. Bauer, A.A. and Doonan, O.P. 2002: Buying a Table in Erfelek: A New Look at the Social Nature of Trade. Paper presented at the Annual Meeting of the Society for American Archaeology, Denver. Bilgi, Ö. 1994: √kiztepe Kazılarının 1993 Dönemi Sonuçları. Kazı Sonuçları Toplantası 16, 141–60. ——. 1998: √kiztepe Kasızı 1996 Dönemi Sonuçları. Kazı Sonuçları Toplantası 19, 323–56. Bloch, M. and Parry, J. 1989: Introduction: Money and the Morality of Exchange. In Parry, J. and Bloch, M. (eds.), Money and the Morality of Exchange (Cambridge), 1–32. Braudel, F. 1972 [1949]: The Mediterranean and the Mediterranean World in the Age of Phillip II (London). Burney, C. 1956: Northern Anatolia before Classical Times. Anatolian Studies VI, 179–203. Chang, K.C. 1975: Ancient Trade as an Economics or an Ecology. In Sabloff, J. and LambergKarlovsky, C.C. (eds.), Ancient Civilization and Trade (Albuquerque), 211–24. Chernykh, E.N. 1992: Ancient Metallurgy in the USSR: The Early Metal Age (Cambridge/New York). Childe, V.G. 1925: The Dawn of European Civilisation (London). Curtin, P.D. 1984: Cross-Cultural Trade in World History (Cambridge). Daniel, E.V. 1984: Fluid Signs: Being a Person the Tâmil Way (Berkeley). Dietler, M. and Herbich, I. 1998: Habitus, Techniques, Style: An Integrated Approach to the Social Understanding of Material Culture and Boundaries. In Stark, M.T. (ed.), The Archaeology of Social Boundaries (Washington, DC), 262–3. Dixon, R.B. 1928: The Building of Cultures (New York). Doonan, O.P. 2002: Production in a Pontic Landscape: The Hinterland of Greek and Roman Sinope. In Faudot, M., Fraysse, A. and Geny, E. (eds.), Pont-Euxin et Commerce: actes du IXe Symposium de Vani (Besançon), 185–98.
244
ALEXANDER A. BAUER
——. 2004: Sinop Landscapes: Exploring Connection in a Black Sea Hinterland (Philadelphia). Doonan, O.P., Gantos, A., Hiebert, F., Besonen, M. and Yaycioªlu, A. 2001: Sinop Regional Archaeological Survey 1998–99: The Karasu Valley Survey. TUBA-AR 4, 113–35. Draganov, V. 1995: Submerged Coastal Settlements from the Final Eneolithic and Early Bronze Age in the Sea around Sozopol and the Urdoviza Bay near Kiten. In Bailey, D.W. and Panayotov, I. (eds.), Prehistoric Bulgaria (Madison, WI), 225–42. Erzen, A. 1956: Sinop Kazısı 1953 Yılı Çali¤maları. Türk Arkeoloji Dergisi 6, 69–72. Fotiadis, M. 1999: Comparability, Equivalency, and Contestation. In Robb, J.E. (ed.), Material Symbols: Culture and Economy in Prehistory (Southern Illinois University, Carbondale Occasional Paper) (Carbondale, IL), 385–98. French, D.H. 1991: The Iron Age in the Black Sea. Thracia Pontica 4, 237–40. Gurney, O.R. 1992: Hittite Geography: Thirty Years On. In Akurgal, E., Ertem, H. and Süel, A. (eds.), Hittite and Other Anatolian and Near Eastern Studies in Honor of Sedat Alp (Ankara), 213–21. Hiebert, F.T. 2001: Black Sea Coastal Cultures: Trade and Interaction. Expedition 43, 11–20. Hiebert, F.T., Doonan, O.P. and Smith, J.S. in preparation: Sinop Kale Northwest Excavations, 2000. Hiebert, F.T., Smart, D., Doonan, O. and Gantos, A. 1997a: Black Sea Trade Project. American Journal of Archaeology 101, 377. ——. 1997b: From Mountaintop to Ocean Bottom: A Holistic Approach to Archaeological Survey along the Turkish Black Sea Coast. In Tancredi, J. (ed.), Ocean Pulse (New York), 93–108. ——. 1998: Foreland and Hinterland: Sinop, Turkey. In Fossey, J. (ed.), Foreland and Hinterland (Montreal). Hiller, S. 1991: The Mycenaeans and the Black Sea. In Laffineur, R. and Basch, L. (eds.), Thalassa. L’Egée Préhistorique et la Mer (Aegaeum) (Liège), 207–16. I¤ın, M.A. 1998: Sinop Region Field Survey. Anatolia Antiqua 6, 95–139. Kâmil, T. 1982: Yortan Cemetery in the Early Bronze Age of Western Anatolia (BAR International Series 145) (Oxford). Knudson, S. 1995: Fisheries along the Black Sea Coast of Turkey: Informal Resource Management in Small-scale Fishing in the Shadow of a Dominant Capitalist Fishery. Human Organization 54, 437–48. Kohl, P.L. 1987: The Ancient Economy, Transferable Technologies and the Bronze Age World System: A View from the Northeastern Frontier of the Ancient Near East. In Rowlands, M., Larsen, M. and Kristiansen, K. (eds.), Centre and Periphery in the Ancient World (Cambridge), 13–24. Kopytoff, I. 1986: The Cultural Biography of Things: Commodotization as Process. In Appadurai, A. (ed.), The Social Life of Things: Commodoties in Cultural Perspective (Cambridge), 64–91. Kushnareva, K.K. 1997: The Southern Caucasus in Prehistory: Stages of Cultural and Socioeconomic Development from the Eighth to the Second Millennium BC (Philadelphia). Lazarov, M. 1984: La Navigation le Long du Littoral Trace du Pont Euxin Avant la Colonisation Grecque. In Peshev, A., Popov, D., Jordanov, K. and von Bredov, I. (eds.), Dritter Internationaler Thracologischer Kongreß (Sofia), 63–8. Leshtakov, K. 1995: The Detachment of the Early Bronze Age Ceramics Along the South Bulgarian Black Sea Coast. In Lazarov, M. and Angelova, C. (eds.), Les Ports dans la Vie de la Thrace Ancienne (Thracia Pontica V) (Sozopol), 23–38. Lévi-Strauss, C. 1969 [1947]: The Elementary Structures of Kinship (Boston). McGlade, J. and McGlade, J.M. 1989: Modelling the Innovative Component of Social Change. In van der Leeuw, S.E. and Torrence, R. (eds.), What’s New? A Closer Look at the Process of Innovation (London), 281–99. Macqueen, J.G. 1980: Nerik and Its ‘Weather God’. Anatolian Studies 30, 179–87. Makkay, J. 1993: Pottery Links between Late Neolithic Cultures of the NW Pontic and Anatolia and the Origins of the Hittites. Anatolica 19, 117–28. Maksimova, M.I. 1951: Hittites in the Black Sea Region. Journal of Near Eastern Studies 10, 74–81. Malinowski, B. 1922: Argonauts of the Western Pacific (London).
12. PREHISTORIC SINOP AND INTER-REGIONAL INTERACTION
245
Manzura, I. 1994: Cultural Groups from the Steppes of Eastern Europe in the Eneolithic Period and Early Bronze Age. Oxford Journal of Archaeology 13, 265–77. Markovin, V.I. 1997: Dol’mennye pamiatniki Prikuban’ia i Prichernomor’ia (Moscow). Mauss, M. 1990 [1950]: The Gift. (New York). Meeker, M.E. 1971: The Black Sea Turks: Some Aspects of their Ethnic and Cultural Background. International Journal of Middle East Studies 2, 318–45. Meskell, L. 2002: The Intersections of Identity and Politics in Archaeology. Annual Review of Anthropology 31, 279–301. Nikolova, L. 1995: Data About Sea Contacts During the Early Bronze Age in South-Eastern Europe (c. 3500/3400–2350–2250 BC). In Lazarov, M. and Angelova, C. (eds.), Les Ports dans la Vie de la Thrace Ancienne (Thracia Pontica V) (Sozopol), 57–86. Orser, C.E. jr 1992: Beneath the Material Surface of Things: Commodoties, Artifacts and Slave Plantations. Historical Archaeology 26, 95–104. Özdoªan, M. 1993: Vinça and Anatolia: A New Look at a Very Old Problem (or Redefining Vinça Culture from the Perspective of Near Eastern Tradition). Anatolica 19, 173–93. ——. 2003: The Black Sea, The Sea of Marmara and Bronze Age Archaeology: An Archaeological Predicament. In Wagner, G.A., Pernicka, E. and Uepermann, H.-P. (eds.), Troia and the Troad: Scientific Approaches (Berlin), 105–20. Özdoªan, M., Miyake, Y. and Dede, N.Ö. 1991: An Interim Report on Excavations at Yarımburgaz And Toptepe in Eastern Thrace. Anatolica 17, 59–121. Özgüç, T. 1959: Kültepe-Kani¤ (Türk Tarih Kurumu Yayınlarından V 19). (Ankara). Özveren, Y.E. 1997: A Framework for the Study of the Black Sea World, 1789–1915. Review of the Fernand Braudel Center 20, 77–113. Özveren, Y.E. 2001: The Black Sea as a Unit of Analysis. In Aybak, T. (ed.), Politics of the Black Sea: Dynamics of Cooperation and Conflict. (London), 61–84. Parzinger, H. 1993: Zur Zeitstellung der Büyükkaya-Ware: Bemerkungen zur Vorbronzezeitlichen Kulturfolge Zentralanatoliens. Anatolica 19, 211–29. Preucel, R.W. and Bauer, A.A. 2001: Archaeological Pragmatics. Norwegian Archaeological Review 34, 85–96. Rassamakin, Y.Y. 1994: The Main Directions of the Development of Early Pastoral Societies of Northern Pontic Zone: 4500–2450 BC (Pre-Yamnaya Cultures and Yamnaya Culture). In Ko¤ko, A. (ed.), Nomadism and Pastoralism in the Circle of Baltic-Pontic Early Agrarian Cultures: 5000–1650 (Baltic-Pontic Studies 2) (Poznan), 29–70. ——. 1999: The Eneolithic of the Black Sea Steppe: Dynamics of Cultural and Economic Development 4500–2300 BC. In Levine, M., Rassamakin, Y., Kislenko, A. and Tatarinteseva, N. (eds.), Late Prehistoric Exploitation of the Eurasian Steppe. (Cambridge), 59–182. ——. forthcoming: The Chalcolithic and Early Bronze Age Hydrostrategy in the Black Sea Steppe Area. In Gheorgiu, D. (ed.), The Hydrostrategies of Chalkolithic Cultures (Oxford). Reina, R. and Hill, R.M. 1978: The Traditional Pottery of Guatemala (Austin). Renfrew, C. 1975: Trade as Action at a Distance: Questions of Integration and Communication. In Sabloff, J.A. and Lamberg-Karlovsky, C.C. (eds.), Ancient Civilization and Trade (Albuquerque), 3–59. ——. 1993: Trade Beyond the Material. In Scarre, C. and Healy, F. (eds.), Trade and Exchange in Prehistoric Europe (Oxford), 5–16. Rice, P.M. 1984: Change and Conservatism in Pottery-Producing Systems. In van der Leeuw, S.E. and Pritchard, A.C. (eds.), The Many Dimensions of Pottery: Ceramics in Archaeology and Anthropology (Amsterdam), 233–93. Roodenberg, J. 1999: Ilıpınar, An Early Farming Village in the Iznik Lake Basin. In Özdoªan, M. (ed.), Neolithic In Turkey (Istanbul), 193–202. Sagona, A.G. 1984: The Caucasian Region in the Early Bronze Age (BAR International Series 214) (Oxford). Sava, E. 1994: Epoca Bronzului – Perioada Mijlocie ¤i târzie. Thraco-Dacica 25, 141–58. Sherratt, S. 1998: ‘Sea Peoples’ and the Economic Structure of the Late Second Millennium in the Eastern Mediterranean. In Gitin, S., Mazar, A. and Stern, E. (eds.), Mediterranean Peoples in Transition: Thirteenth to Early Tenth Century BCE (Jerusalem), 292–313.
246
ALEXANDER A. BAUER
Singer, M. 1984: Man’s Glassy Essence: Explorations in Semiotic Anthropology (Bloomington). Skeates, R. 1998: The Social Life of Italian Neolithic Painted Pottery. In Bailey, D. (ed.), The Archaeology of Value: Essays on Prestige and the Processes of Valuation (BAR International Series 730) (Oxford), 131–41. Spier, R.F.G. 1967: Work Habits, Postures and Fixtures. In Riley, C.L. and Taylor, W.W. (eds.), American Historical Anthropology: Essays in Honor of Leslie Spier (Carbondale, IL), 197–220. Thissen, L. 1993: New Insights in Balkan-Anatolian Connections in the Late Chalcolithic: Old Evidence from the Turkish Black Sea Littoral. Anatolian Studies 43, 207–37. Tonœeva, G. 1981: Un Habitat Lacustre de l’Âge du Bronze Ancien dans les Environs de la Ville de Varna. Dacia 25, 41–62. Trifonov, V.A. 1994: The Caucasus and the Near East in the Early Bronze Age (Fourth and Third Millennia BC). Oxford Journal of Archaeology 13, 357–60. Tsetskhladze, G.R. 1994: Greek Penetration of the Black Sea. In Tsetskhladze, G.R. and De Angelis, F. (eds.), The Archaeology of Greek Colonisation: Essays Dedicated to Sir John Boardman (Oxford), 111–35. —— (ed.) 1996: New Studies on the Black Sea Littoral (Colloquia Pontica 1) (Oxford). Urban, G. 1996: Metaphysical Communities (Austin). van der Leeuw, S.E. 1991: Variation, Variability, and Explanation in Pottery Studies. In Longacre, W.A. (ed.), Ceramic Ethnoarchaeology (Tucson), 11–39. Vandiver, P.B. 1988: The Implications of Variation in Ceramic Technology: The Forming of Neolithic Storage Vessels in China and the Near East. Archaeomaterials 2, 139–74. Weiner, A.B. 1992: Inalienable Possessions: The Paradox of Keeping-While-Giving (Berkeley). Weissner, P. 1989: Style and Changing Relations Between the Individual and Society. In Hodder, I. (ed.), The Meanings of Things (London), 56–63. Wobst, H.M. 1977: Stylistic Behavior and Information Exchange. In Cleland, C. (ed.), Papers for the Director: Museum of Anthropology, Anthropological Papers (Ann Arbor). Yakar, J. 1975: Northern Anatolia in the Early Bronze Age. Tel Aviv 2, 133–45. Yentsch, A. 1991: The Symbolic Divisions of Pottery: Sex-related Attributes of English and Anglo-American Household Pots. In McGuire, R.H. and Paynter, R. (eds.), The Archaeology of Inequality (Oxford), 192–230. Zbenovich, V.G. 1973: Chronology and Cultural Relations of the Usatovo Group in the USSR. In Kalousek, F. and Budinsky-Kriœka, V. (eds.), Symposium Über die Entstehung und Chronologie der Badener Kultur (Bratislava), 513–24.
CHAPTER THIRTEEN OVER THE MOUNTAINS AND THROUGH THE GRASS: VISUAL INFORMATION AS ‘TEXT’ FOR THE ‘TEXTLESS’ KAREN S. RUBINSON
Yoffee and Sherratt (1993, 6) propose that ‘Since the nature of culture is very different along the spectrum of human societies, it follows that archaeological theory must vary commensurately with the societies and problems being investigated.’ In the study of material culture in Eurasia, where the only first-hand textual records about ancient peoples were made by literate societies at Eurasia’s periphery (Greece, Assyria and China, for example), change in artefact subjects and styles is most often attributed to contact with a ‘dominant’ culture outside of Eurasia. At the local level such change has been less studied, and when it is, a very much more complex explanation must be found, one that challenges the above stated model, which is essentially a core-periphery construct. I present here an approach specifically appropriate to the steppe and the Caucasus in prehistoric times, for peoples without written texts in times when others around them have writing. I propose that iconography changes more quickly than artistic style in areas such as the steppe and the Caucasus,1 and that those changes in iconography can often reflect changes in the society in which they take place,2 as well as interchange with external areas. The observation of these iconographic changes requires precise analysis of details of the images being
1
This is in sharp contrast to the situation in the ancient Near East in the 1st millennium BC, where Winter (2001) has cogently laid out the importance of style in defining groupings of art objects. I would argue that when one has a very large data set, in a cultural-historical setting where subject matter is characteristically long-lived, such an approach is both possible and appropriate. In the case of the steppe and the Caucasus, with a small data set of ‘art objects’, iconography assumes greater significance. It is interesting that Winter is arguing in her context for greater rigour and attention to detail, as I do herein. See also Marcus 1991 (especially 559), where she too, in making arguments about style, makes a case for the importance of details. 2 Schapiro (1994, 59) in his 1953 classic and still stimulating discussion of style makes this point. See also Wöfflin, as cited by Winter (1998, 65), and Winter (1998, 67), where she sees style shifts as reflecting historical/cultural change.
248
KAREN S. RUBINSON
studied.3 Sources of new imagery are meaningful in historic, economic, or other socio-cultural terms. I begin with the premise that if two visual images from different geographical areas in Eurasia look similar to each other and date within more or less the same timeframe, that visual similarity can be explained by some socio-cultural phenomenon, characterised by contact, either direct or indirect.4 Behind that premise lies the assumption that what you know influences what you see. This assumption can be illustrated by the multistable images, that is, images that can be read in more than one way, now used by anthropologists to illustrate a variety of ideas (for example: Hodder 1991, 17, fig. 1; Shanks and Tilley 1987, 36–7, fig. 2.1; Preucel and Bauer 2001, 92, fig. 6). In the context of understanding illusion, Gombrich brought this kind of image from psychology to the study of art history (1969, 4–6, fig. 2) (Fig. 1). Read as either a rabbit or a duck, it demonstrates the assumption behind the premise – that what you know does in fact shape or frame what you see.5 In this image, if one ‘expects’ a duck, the rabbit becomes effectively invisible (and vice versa); what is in the viewer’s mind affects what the eye comprehends. Likewise, one must look at the objects of artistic production within the local context in which they were found, but must also know something about nearby contemporary cultures, so that the objects can be understood in all possible ways. It can be meaningful when art objects share resemblances with objects found outside the area where they are found as well as when they do not have commonalities of appearance (see Marcus 1996, 2, 16–7, 75–8). Keeping that in mind, let us consider two concepts in the study of art history – style and iconography. Art historical style is not the same as the multi-
3
In the context of the ancient Near East, Postgate (1994, 180–1, figs. 17.4–17.5) argues, inter alia, for the importance of noting details of iconography within sealings of the same style. His arguments there relate to artistic representation and written text and in the first case argues that greater precision of observation of details can amplify texts and illustrate cultural-historical change and circumstance. 4 This premise not only underlies a great deal of archaeological theory but is a fundamental assumption in the study of what can be called art objects found in archaeological contexts. See, for example, Marcus 1991; Winter 1980; 2001. 5 Panofsky (1982 [1955], 39) notes ‘our identification and interpretation will depend on our subjective experience’ as well, which means more than just knowledge of cultures, an issue more recently addressed explicitly and in detail by Hodder (1990; 1992) and others.
13. OVER THE MOUNTAINS AND THROUGH THE GRASS
249
Fig. 1. Multistable image: rabbit and/or duck (after Gombrich 1969, fig. 2).
faceted styles discussed in the anthropological literature.6 Winter recently said of art-historical style: For my [that is, Winter’s] purposes I would take the position that once there is anything in the work we can call form, then there is also style, but that it is also important to keep what is intrinsic to the work distinct from what is extrinsic to it, by consistently referring to posthoc determinations as the products of ‘stylistic analysis.’ In that way, style is a function of a period, place, workshop or hand; it is inherent in the work, and it is thus what is apparent to the perceiver. Stylistic analysis then introduces the conscious observation, selection, and articulation of manifest properties to the act of perception. (Winter 1998, 57)
One could usefully replace ‘stylistic analysis’ with ‘iconographic analysis’ in the above statement. ‘Iconography’ is the study of subject matter displayed by the ‘art object’, and, on a deeper level, its meaning (Panofsky 1982, 26). To make these distinctions plain, one could look, for example, at 1st millennium BC ivories of the clearly different North Syrian and Phoenician styles depicting the same subjects, as Winter does in discussing, among other examples, winged sphinxes (1998, figs. 1–2). She characterised the Phoenician as a delicate style and the North Syrian as a blocky, more massive one (Winter 1998, 58–62). Another example is images of Buddha from Gandhara (3rd century AD) and Mathura (5th century AD) where elements of the iconography of the
6
The substantial literature on this subject includes the many articles in Conkey and Hastorf 1990. A recent summary by Hegmon (1998, 264–6) lists additional bibliography and raises many important theoretical and methodological issues.
250
KAREN S. RUBINSON
Buddha figure transcend both the chronological and spatial differences. The differences are seen most clearly in the treatment of body and drapery, the heavy drape hinting at the body below in earlier example and with the body clearly showing through the sheer drapery indicated on the surface of the body in the later sculpture. Yet both images clearly represent the Buddha, who is defined by the hairstyle with topknot and long ear lobes (Elsen 1972, 45–6, figs. 59–60). Another illustration of distinctions of style can be seen in representations of a modern icon, the Brooklyn Bridge, including those by Georgia O’Keefe and John Marin. Despite the different stylistic approaches of the two artists, O’Keefe’s more blunt and solid and Marin’s more lively and linear, the bridge’s towers clearly define the monument in both renderings (www.brooklynexpedition.org/structures/buildings/bridge/bl_bridge_exp.html).7 Any object of artistic production is characterised by both style and iconography. I suggest that, at least in prehistoric Eurasia, elements of iconography were more readily and rapidly borrowed, transferred, or exchanged than elements or characteristics of style.8 Given this hypothesis, a rigorous analytical approach to elements of iconography could yield information about socio-cultural interaction. On the steppe and in the Caucasus, in prehistoric times, we have objects that can be labeled art. Such objects are produced by societies that do not have writing and so we, as modern observers, must try to understand these objects without written texts to help explain their meaning. In this context, objects with figural or pictorial representations serve archaeologists as texts for these societies (cf. Marcus 1991, 538). I am not addressing here material culture in general as text for the past, as discussed by Hodder (1991, 125–7) and others. Rather I am addressing objects with pictorial representation, narrative or otherwise.9 Hodder (1991, 126–7) did say, ‘In so far as material culture is a lan7
Both of these images are in the Brooklyn Museum. John Marin, Brooklyn Bridge #6, 1913, etching. Georgia O’Keefe, Brooklyn Bridge, 1929, oil on masonite, 77.11. 8 The most basic elements of what can be defined as style and iconography may in some cases overlap. Schapiro’s ‘form elements’ (1994 [1953], 54) have been interpreted by Marcus (1991, 538–9, 545) to include such markers as ‘tassels between legs’. In Marcus’s context (1991, fig. 25), where subject matter is widely distributed in time and space, these can be seen as stylistic traits. However, in the steppe/Caucasian context, they can instead be characterised as part of the iconography. In the process I describe herein, we might consider that we are comparing iconography elements within the steppe/Caucasus to style elements in surrounding areas. 9 When such objects are published, they are usually first described. Then the images are interpreted as stories and the sources of the motifs sometimes explored. My hypothesis about why
13. OVER THE MOUNTAINS AND THROUGH THE GRASS
251
guage, it is a simple one when compared with spoken or written language.’ Although Hodder had in mind much less complex objects of material culture than those with figural or pictorial representation in that statement, it reflects one aspect of the appeal of such objects, that they appear simple to understand. Indeed, overall, pictorial objects seem less ambiguous and seem to provide more complete information than other categories of material culture. As a result, in my view, scholars run a greater risk of misinterpretation of such objects, since they seem more explicit. In areas and times without texts, description and interpretation is especially perilous (see Rubinson 2003). When such objects are rare, as for example in the Caucasus in the Bronze Age, they are intensely mined for meaning, both intrinsic and extrinsic. I propose here that the process of iconographic exploration be made explicit and suggest ways in which rigour can be infused into the process. The potential value of the analysis of iconography is certainly not limited to the case study used here, but it is sufficient to make clear the necessary approach, process and procedures. The case study comes from the Trialeti culture of the Southern Caucasus, dating to the end of the 3rd and beginning of the 2nd millennium BC. The culture is best known by kurgan burials, primarily of the elites of this culture, since settlements are rare and fugitive (Kuftin 1941; O. Japaridze 1969; Gogadze 1972; Kushnareva 1997, 84–114; Puturidze 2003). Among the many objects excavated from the Trialeti culture are three silver vessels with embossed figural ornamentation. Two of these, a small bucket from kurgan 17 (Fig. 2) and a goblet from kurgan 5 (Fig. 3), were excavated in the late 1930’s from Trialeti itself, in Georgia (Kuftin 1941, figs. 88, 91), and a second silver goblet was excavated in 1987 at the site of Karashamb in Armenia (Fig. 4) (Oganesian 1988; 1992a; 1992b). The three silver objects are distinctive in Middle Bronze Caucasia for many reasons. The two goblets both display human figures, which are neither characteristically represented in the preceding period in Caucasia nor widespread during the Trialeti culture. Both goblets are divided into registers and the figures on the vessels are shown standing on ground lines, a new feature in Caucasian representational art (N. Japaridze
scholars are drawn to objects with imagery and so eagerly explain them is that the visually/artistically presented stories appear more complete and seem to be more quickly accessible than other ways of understanding and interpreting a culture, such as the laborious accretion of data from survey and excavation.
252
KAREN S. RUBINSON
Fig. 2. Bucket from Trialeti kurgan 17 (after Kuftin 1941, pl. LXXXVIII).
1988, 16). The hunting scenes on the Karashamb goblet and Trialeti bucket are also new to Caucasia in this period. There is no question that these objects, found in elite burials, bore meanings within the culture where they were found and that they were items of prestige and value, but these are not issues which I will address here (but see Rubinson 2003; Smith 2001; Helms 1993, 149). What specifically inspired the individuals who crafted these objects and what local dynamics specifically existed which caused the new images to be produced are, in my view, unknowable at this time. Here we will be looking at the images themselves, to delineate the steps necessary to define the iconography of these vessels and to demonstrate the process by which iconographic detail can contribute to the knowledge and understanding of specific aspects of the culture in which they were found.
13. OVER THE MOUNTAINS AND THROUGH THE GRASS
253
Fig. 3. Goblet from Trialeti kurgan 5 (after Miron and Orthmann 1995, fig. 10).
I argue below that the silver bucket, although found in a burial at Trialeti, has its origin elsewhere, so I begin by looking at the two silver goblets, which are related in iconography, shape, size and technique of manufacture. Both goblets are made of thin silver plate, with hollow foot and flat bottom, and are decorated by bands of ornament separated by narrow plain bands. The lower parts of both goblet bowls bear similar rosette ornaments. The Karashamb goblet is 13.2 cm high, 10.5 cm in diameter and weighs 215 g (Santrot et al. 1996, 65), slightly larger and heavier than the Trialeti example, which stands 11.3 cm high, is 9.8 cm in diameter and weighs 163.39 g (Miron and Orthmann 1995, 238). The human figures on both share large eyes, prominent noses and receding
254
KAREN S. RUBINSON
Fig. 4. Goblet from Karashamb (courtesy the State History Museum of Armenia).
chins, as well as footgear with turned-up toes. However, the humans on the Trialeti goblet have hair and beards indicated, unlike those from Karashamb, and the puffy style of the Karashamb figures contrasts with the more angular Trialeti style. The complex imagery of the Karashamb goblet provides more elements for analysis, so we will look at a few levels of imagery on that vessel in detail: the first, a single iconographic detail, the second a grouping of more than one specific iconographic detail within a single figure, and lastly a complex scene with many details. I am not proposing that these levels are sufficient for a complete iconographic investigation, for they are not, but they will serve to demonstrate the possibilities.
13. OVER THE MOUNTAINS AND THROUGH THE GRASS
255
Let us consider first a single detail – shoes with turned up toes, which appear on the human figures. If we look at human representations from Caucasia from earlier periods, which are rare, we do not see such shoes. For example, the human figure on a vessel from the Sunzhenski cemetery of the Maikop culture of the north Caucasus is quite summary in presentation (Koronevskii, 1993, fig. 32.1). Human figurines from Kura-Araxes contexts do not have shoes with turned up toes (for example, Kushnareva and Markovin 1994, pl. 14.12–16). They also do not occur on figural art from greater Mesopotamia in contexts earlier than the Trialeti culture. Nor are shoes with turned up toes shown in contemporary objects from greater Mesopotamia, such as cult basin fragments from Ebla, in north Syria (Amiet 1980, figs. 448–451), approximately contemporary with the Trialeti culture and depicting some similar subject matter. In fact, if one looks at the largest corpus of imagery, cylinder seals, of the early 2nd millennium BC from all of the ancient Near East, where there are many kinds of human representations, one does not find such a detail. But shoes with turned up toes are important forms in the repertoire of drinking vessels from the Assyrian trading centre of Kültepe karum Kanesh in central Anatolia (Özgüç 1986, 69, pl. D.1–2, pl. 119.4), a site contemporary with the Middle Bronze Trialeti culture. In fact, to my knowledge, it is only in Anatolia that such a specific image is found in a context earlier than or contemporary to its appearance on the Trialeti culture goblets. But the tentative nature of this comparison is clear. It demonstrates the danger of seeking iconographical comparisons in single details. In my view, the potential relationship suggested here has insufficient authority to be used to make a case for contact, inspiration, or any other kind of interaction. A combination of two or more details in an image should be a more reliable marker of contact/interaction. As an example, an image with a combination of details is one of the three headless figures on the Karashamb goblet (Fig. 5). The rightmost figure has, like the other two, the neck clearly shown. In addition, the figure has one arm extended up and the other downwards. The same two iconographic details occur on a sealing from Kültepe karum Kanesh, in the style which is characterised by N. Özgüç as the local style (Özgüç 1965, pl. XVII.50) (Fig. 6). Above the man with the spear is a headless figure shown with a stalk-like neck and with the arms akimbo. In the period of the Trialeti culture – and any time before, in the greater Near East and the steppe, this sealing and the Karashamb goblet are the only two objects which I have identified that have this exact motif. That is not to say that headless figures do not occur
256
KAREN S. RUBINSON
Fig. 5. Goblet from Karashamb (courtesy the State History Museum of Armenia).
in the art of other cultures (for instance, a later example of a headless figure on the relief of the Battle of Til-Tuba, from the time of Assurbanipal in Assyria: Curtis and Reade 1995, 73, bottom), but the particular combination of elements, the neck being shown together with arms oppositionally flung up and down, is apparently preserved only in these two examples. An even stronger indicator of contact/interaction/inspiration would be a scene made up of a complex of specific elements. In this case study the central seated figure next to a table is an example. This image is not found in any Caucasian culture prior to the appearance on the Karashamb and Trialeti goblets (Figs. 3–4). The central scenes on both the Karashamb and Trialeti goblets share
13. OVER THE MOUNTAINS AND THROUGH THE GRASS
257
Fig. 6. Sealing Kt d/k 33, from Kültepe karum Kanesh (after Özgüç 1965, pl. XVII, 50).
many details. The main figures sit on stools with crossed struts, there are animals next to the tables before the main figures, the tables and stands share many details, such as the hoofed feet. These shared elements of iconography are again not part of the iconography of earlier Caucasia. Although such scenes, usually called presentation scenes, are common in the ancient Near East, they are not part of, for example, the iconography found on the steppe, so the directionality of interaction/inspiration/source is not to the north of the Trialeti culture. When one investigates the contemporary art of the ancient Near East, one finds the details seen on the goblets frequently in the seals of the Anatolian local style of the Assyrian Trading colony period: the stool with crossed struts, the hoofed altar with discrete objects on top, the animal near the altar, exemplified by the impressions of sealings found at Kültepe karum Kanesh (N. Özgüç 1965, 77, fig. 15a) (Fig. 7). In this instance, I name the hoofed table an altar, for within the context of Anatolia and known ancient Near Eastern practice, this is the customary identification. Although an object with similar form occurs on the Trialeti silver work, we can only note the similarity in form. There is no basis for inferring that the image represents an altar in the Caucasian context. Until more data are collected, precise identification and interpretation of the visual images remains elusive. That the Trialeti culture goblets reflect in some way the imagery specifically grounded in the Anatolian local style may be demonstrated by comparing the details of an impression of a typical contemporary seal in Assyrian style, also from Kültepe, where, for example, the principal figure sits on a chair rather than a stool (Özgüç and Tunca 2001, pl. C.21) (Fig. 8). It is in such small details that iconographic analysis reveals the kinds of precise visual relationships from which or through which it may be possible to propose more macro interactions.
258
KAREN S. RUBINSON
Fig. 7. Sealing Kt a/k 494, from Kültepe karum Kanesh (after Özgüç 1965, pl. V, 15a).
Fig. 8. Sealing Kt c/k 1389, from Kültepe karum Kanesh (after Özgüç and Tunca 2001, pl. C2).
13. OVER THE MOUNTAINS AND THROUGH THE GRASS
259
Above, I noted that certain iconography did not occur in Caucasia earlier than the Trialeti culture. In the process of defining the meanings of this distinctive iconography, it is necessary to look at how the Trialeti culture goblets compare to art in the preceding Kura-Araxes and Early Kurgans periods, as well as the nearby northern Caucasian Maikop culture, to verify the novelty in the iconography, to explore what visual elements remain the same from the earlier to later periods, and to define local characteristics further in comparison to extrinsic areas. Some stylistic characteristics seen on the Trialeti and Karashamb goblets are found in earlier Caucasian representations. For example, all four legs are shown of all animals standing on the ground lines on both goblets, and, with the exception of the feline and boar in the top register of the Karashamb goblet and one animal on the lower register of the Trialeti goblet, are laid out without overlap. These stylistic characteristics are seen, for example, on a copper diadem from Kwazchela, which is decorated by both birds and animals, in profile, with all legs shown, discretely laid out in separate areas (Lordkipanidze 1991, fig. 22.5). N. Japaridze (1988, 16) has noted that prior to the Trialeti culture in Caucasia, figures did not stand on ground lines and that this characteristic is an influence from the Near East. I would agree that it is specifically the standing on the ground line that is new, since if we look at the north Caucasian Maikop culture, on two of the well-known silver vessels, the animals are arranged in rows, although they do not, in fact, stand directly on ground lines, even when a ground line is indicated (Chernyk 1992, fig. 22). The previously mentioned ceramic vessel from the Sunzhenski cemetery, also of the Maikop culture, is ornamented by a row of figures on the shoulder (Korenevskii 1993, figs. 31–33). Birds are illustrated in a row on the inside of a vessel of the Kura-Araxes culture (Kushnareva and Markovin 1994, pl. 3.37) and deer are shown in a row along the outside of another Kura-Araxes vessel (Kushnareva and Markovin 1994, pl. 3.38). Although animals in rows are not the dominant decoration of metalwork and ceramics in the time preceding the Trialeti culture, when animals are shown, their placement in rows is characteristic, although the placement of animals standing directly on a ground line is not. The idea that elements of the decoration are bounded, as the figures are within the bands on the goblets, can also be seen in objects from Georgia in the Early Kurgans period, which just precedes the period of the Trialeti culture. Examples where design elements are bounded include the lion from Tsnori,
260
KAREN S. RUBINSON
where the dots which decorate the front of the animal are bounded by lines (Miron and Orthmann 1995, fig. 51). The braided bands on a pendent from Ananauri are bounded by smooth bands (Miron and Orthmann 1995, fig. 50) and the bands of circles on a pin from Bedeni are contained within hatched lines (Miron and Orthmann 1995, fig. 8). Further, much of the ornament on pottery of the Kura-Araxes, Early Kurgans and Trialeti periods are arranged in horizontal bands (Kushnareva and Markovin 1994, pls. 3, 15, 18). Based on the appearance of animals in rows and bounded patterns in Caucasia previous to the time of the Trialeti culture, I would tentatively suggest that although the iconography on the Trialeti and Karashamb goblets is new and derived from external visual formulas, that the style of the representations shares important elements with earlier local artistic representation. The organisation of the goblets, contrasts with the organisation of the decoration on the silver bucket from Trialeti kurgan 17 (Fig. 2). On the bucket, the animals are arranged in an all-over pattern, with only those at the base of the bucket standing on a ground line. It is this aspect of the ornament, a space densely filled with figures at mutliple levels and orientations, so similar to the Anatolian style seals of Kültepe karum Kanesh, which suggests to me that perhaps the bucket is an import from Anatolia. Although further research will be needed to confirm this, I venture to make this proposal because of specific characteristics of the style of the object, which is distinctive from the style of the goblets. At the outset of this paper, I suggested that iconography can change more quickly and more easily than style in the cultures of the areas of the steppe and the Caucasus. In this case study, I have demonstrated that elaborate new imagery is seen in the Trialeti culture, yet some earlier stylistic tendencies continue in the style of the goblets. The bucket has both imagery new to Caucasia and a different style and thus may represent an object which was not made locally. One might query why one should bother to ask such questions of these objects. I concur with Hodder (1991, 1, 11–2) that archaeological information can have historical value. I want to know not only the locally based questions of why such imagery newly occurs in this place and how it is similar or different to visual imagery from the same place in earlier times or who within the society made the object and for what audience, but also from where, precisely, it is borrowed. In this specific case, I suggest that it is indeed Anatolia, and specifically the settlements of the Assyrian trading colonies, that is the source
13. OVER THE MOUNTAINS AND THROUGH THE GRASS
261
of the new iconography. There is other evidence which suggests contact between the southern Caucasus and central Anatolia in this period that I have discussed elsewhere (Rubinson 2001), and I have previously suggested that all of these similarities are a faint reflection of an ongoing interaction between these two areas (Rubinson 2003). Marina Puturidze instead suggests that such contact was only intermittent (Puturidze 2003). With more specific examination of details which can be compared between objects of the Trialeti culture and the ancient Near East, perhaps, even with such a small data set, some quantitative measure can be defined to bear on the argument. In the meantime, I have attempted above to make a case that new and different iconography plays an important role when using art to understand the underlying culture, society, and cultural interactions in cultures where there are no texts and other kinds of direct historical knowledge. In the study of the art objects of the literate ancient Near East, for example, artistic images are named in the minds of the modern viewer as the ancient viewer and their place in the local cultural systems implicitly understood (Postgate 1994). Thus, generally, style becomes the art-historical focus of investigation except for images not described in the texts (Winter forthcoming; Marcus 1996). In the steppe and the Caucasus, where objects with pictorial/representational imagery remain rare and texts do not exist, the iconography of the art objects, through detailed and precise description and observation, can amplify information which is visible through the rest of the archaeological record or even shed new light on the societies from which they come.
Acknowledgments My thanks to Prof. Katheryn Linduff for her cogent comments on an earlier version of this manuscript. My study of the Trialeti culture materials in Georgia was supported by a grant from the American Philosophical Society and graciously facilitated by the late Tamaz Kiguradze. My analysis of the Karashamb goblet has been facilitated by photographs generously provided by Ruben Badalyan, Stephan Kroll, Adam T. Smith and Laura Tedesco. Irene J. Winter provided both published and pre-publication versions of her work. Alex Bauer and Michelle Marcus provided useful bibliography. I am grateful for all those who provided input and support.
262
KAREN S. RUBINSON
Bibliography Amiet, P. 1980: Art of the Ancient Near East (New York). Chernykh, E.N. 1992: Ancient Metallurgy in the USSR: The Early Metal Age (Cambridge/New York). Conkey, M.W. and Hastorf, C.A. (eds.) 1990: The Uses of Style in Archaeology (Cambridge). Curtis, J.E. and Reade, J.E. 1995: Art and Empire: Treasures from Assyria in the British Museum (London). Elsen, A.E. 1972: Purposes of Art, 3rd ed. (New York). Gogadze, E.M. 1972: Trialetis korganuli kulturis periodizatsia da genezisi (Tbilisi). Gombrich, E.H. 1969: Art and Illusion: A Study in the Psychology of Pictorial Representation (Princeton). Hegmon, M. 1998: Technology, Style, and Social Practices: Archaeological Approaches. In Stark, M.T. (ed.), The Archaeology of Social Boundaries (Washington, DC), 264–79. Helms, M.W. 1993: Craft and the Kingly Ideal (Austin). Hodder, I. 1990: Style as Historical Quality. In Conkey and Hastorf 1990, 44–51. ——. 1991: Reading the Past, 2nd ed. (Cambridge). Japaridze, N. 1988: Yuvelirnoye Iskusstvo epokhi bronzy v Gruzii (Tbilisi). Japaridze, O.M. 1969: Arkheologicheskiye Raskopki v Trialeti (Tbilisi). Kévorkian, R.H. 1996: Arménie entre Orient et Occident (Paris). Korenevskii, S.N. 1993: Drevneishee osedloe nacelenie na Srednem Tereke (Moscow). Kuftin, B.A. 1941: Arkheologicheskie Raskopki v Trialeti I (Tbilisi). Kushnareva, K.K. 1997: The Southern Caucasus in Prehistory: Stages of Cultural and Socioeconomic Development from the Eighth to the Second Millennium BC (Philadelphia). Kushnareva, K.K. and Markovin, V.I. 2001: Arkheologiya. Epokha bronzy Kavkaza i Srednei Azii. Rannyaya I srednyaya bronza Kavkaza (Moscow). Lordkipanidze, O. 1991: Archäologie in Georgien. Von der Altsteinzeit zum Mittelalter. (Quellen und Forschungen zur prähistorischen und provinzialrömischen Archäologie 5) (Weinheim). Marcus, M.I. 1991: The Mosaic Glass Vessels from Hasanlu, Iran: A Study in Large-Scale Stylistic Trait Distribution. The Art Bulletin 73, 536–60. ——. 1996: Emblems of Identity and Prestige: The Seals and Sealings from Hasanlu, Iran: Commentary and Catalog (University Museum Monograph 84; Hasanlu Special Studies 3) (Philadelphia). Miron A. and W. Orthmann 1995: Unterwegs zum Goldenen Vlies (Saarbrücken). Oganesian, V.E. 1988: Serebryanyi kubok iz Karashamba. Istoriko-filologicheskii zhurnal 4 (123), 145–61. ——. 1992a: A Silver Goblet from Karashamb. Soviet Anthropology and Archeology (New York) 30.4, 84–102. ——. 1992b: Raskopki karashambskogo mogil’nika v 1987g. Arkheologischeskiye Raboty na Novostroikakh Armenii 1 (Yerevan), 26–36. Özgüç, N. 1965: Kültepe Mühür Baskılar da Anadolu Grubu (Ankara). Özgüç, N. and Tunca, O. 2001: Kültepe-Kani·. Mühürlu ve yazıtlı kil bullalar (Ankara). Özgüç, T. 1986: Kültepe-Kani· II (Ankara). Panofsky, E. 1982 [1955]: Iconography and Iconology: An Introduction to the Study of Renaissance Art. In Panofsky, E. (ed.), Meaning in the Visual Arts (Chicago), 26–54. Postgate, J.N. 1994: Text and Figure in Ancient Mesopotamia: Match and Mismatch. In Renfrew, C. and Zubrow, E.B. (eds.), The Ancient Mind: Elements of Cognitive Archaeology (Cambridge), 176–84. Preucel, R.W. and Bauer, A.A. 2001: Archaeological Pragmatics. Norwegian Archaeological Review 74.2, 85–96. Puturidze, M. 2003: Social and Economic Shifts in the South Caucasian Middle Bronze Age. In Smith, A.T. and Rubinson, K.S. (eds.), Archaeology in the Borderlands: Investigations in Caucasia and Beyond (Los Angeles), 111–27.
13. OVER THE MOUNTAINS AND THROUGH THE GRASS
263
Rubinson, K.S. 2001: Metal Vessels with Basket Handles at Trialeti. In Maisuradze, B. and Rusishvili, R. (eds.), Caucasus: Essays on the Archaeology of the Neolithic-Bronze Age (Dedicated to the 80th Birthday of Prof. Otar Japaridze) (Dziebani Suppl. 6) (Tbilisi), 123–4. ——. 2003: Silver Vessels and Cylinder Sealings: Precious Reflections of Economic Exchange In the Early Second Millennium BC. In Smith, A.T. and Rubinson, K.S. (eds.), Archaeology in the Borderlands: Investigations in Caucasia and Beyond (Los Angeles), 128–43. Santrot, J. et al. 1996: Arménie. Trésors de l’Arménie ancienne (Paris). Schapiro, M. 1994: Theory and Philosophy of Art: Style, Artist, and Society: Selected Papers (New York). Shanks, M. and Tilley, C. 1987: Re-Constructing Archaeology (Cambridge). Smith, A.T. 2001: The Limitations of Doxa: Agency and Subjectivity from an Archaeological Point of View. Journal of Social Archaeology 1.2, 155–71. Winter, I.J. 1980: A Decorated Breastplate from Hasanlu, Iran: Type, Style, and Context of an Equestrian Ornament (University Museum Monograph 39; Hasanlu Special Studies 1) (Philadelphia). ——. 1998: The Affective Properties of Styles: an Inquiry into Analytical Process and the Inscription of Meaning in Art History. In Jones, C.A. and Galison, P. (eds.), Picturing Science, Producing Art (London/New York), 55–77. ——. forthcoming: Establishing Group Boundaries: Toward Methodological Refinement in the Determination of Sets as a Prior Condition to the Analysis of Cultural Contact and/or Innovation in 1st Millennium BCE Ivory Carving. In Uehlinger, C. (ed.), Form and Meaning, Style and Content (Fribourg). Yoffee, N. and Sherratt, A. 1993: Introduction: The Sources of Archaeological Theory. In Yoffee, N. and Sherratt, A. (eds.), Archaeological Theory: Who Sets the Agenda? (Cambridge), 1–9.
CHAPTER FOURTEEN INNOVATION, CHANGE, CONTINUITY: CONSIDERING THE AGENCY OF RUSA II IN THE PRODUCTION OF THE IMPERIAL ART AND ARCHITECTURE OF URARTU IN THE 7TH CENTURY BC TUŸBA TANYERI-ERDEMIR
A Case of Testable Agency: The Kingdom of Urartu and Rusa II A living hegemony, as Raymond Williams (1977, 112) states, ‘is always a process’ which ‘has continually to be renewed, recreated, defended and modified.’ In any political entity, change is an inherent and inseparable component of governance strategies and is particularly central to establishing legitimacy of the ruling authority. Although studying and defining changes in material culture has always been a vital component of archaeology, the agency of the rulers in bringing about this change has mostly been assumed, and rarely tested. The archaeological record of the kingdom of Urartu, particularly the imperial art and architecture of the reign of Rusa II (ca. 685–645 BC) – an exceptionally dynamic and productive Urartian king – offers us an apt dataset for such an investigation. Combining this dataset and agency theory with an applicable methodology, we can investigate how Rusa II might have used material culture as an active component in the renewal, reproduction and modification of his rule. Marcia-Anne Dobres and John Robb (2000b, 3–4) have recently directed our attention to a need for further theoretical elaboration and methodologically sound applications of agency theory in archaeology. In contemporary archaeology, different aspects of agency were successfully applied to ancient societies (Smith 2001; Chapman 2000; Sassaman 2000; Joyce 2000). I am hoping that this project will be a contribution to the study of agency of ruling authorities in imperial constructs. I aim to present here a potentially testable hypothesis of how one might define and discuss the way in which the agency of a ruling authority can be investigated and tested in the archaeological record. It is an opportune time to study this problem in the Urartian context for three reasons. First, excavated data from controlled archaeological contexts both at
14. THE IMPERIAL ART AND ARCHITECTURE OF URARTU
265
sites built by Rusa II and his predecessors have accumulated over the years. It is thus possible to investigate and define change in full depth and on an empirically sound basis. Second, the main dataset that ought to be investigated, namely the art and architecture from the reign of Rusa II, is well-defined temporally and archaeologically. Finally, the Urartian imperial context allows us to consider the possibility of active involvement of the monarchs, both in general and in the particular case of Rusa II, in the production of material culture, especially in architecture and art. The Urartian material record of the 7th century offers an extraordinary opportunity to define change and test agency in bringing about this change. All of the Urartian citadels excavated to date were violently destroyed at the end of the 7th century BC, never to be reoccupied. This widespread destruction allowed for the archaeological recovery of a large number of artefacts in situ from exceptionally well preserved contexts from several excavated Urartian citadels. During the excavations of the temple of Haldi at Ayanis, for example, bronze shields and weapons that were once hung on the pillars of the courtyard of the temple were excavated on the floor, right at the spot where they fell during the final destruction of the citadel (Çilingiroªlu 2001, 45–6). In most cases, this violent destruction brought about the preservation of organic remains and precious objects. The absence of later occupation, by contrast, has allowed the excavators to expose very large sections of the sites. The archaeological evidence recovered from such promising contexts offer an invaluable chance to scrutinise the idea of change in material culture, provided that the dataset is empirically sound. To date, five citadels have been identified through building inscriptions and material culture as constructions of Rusa II. Evidence collected in the past few decades from cities built by Rusa II (Figs. 1–2) – namely Karmir-Blur, Adilcevaz, Bastam and Ayanis – along with a better understanding of the previously excavated material from Toprakkale suggests that the material culture produced during the reign of this particular monarch was considerably different from that of his predecessors, and to a large extent more innovative. Although some scholars (Çilingiroªlu 1998, 45; Salvini 1995, 104–5) have realised that there was a significant change of some sort under the rule of Rusa II, no systematic study has yet been conducted on the art and architecture to define and investigate this change through the medium of material products of his reign. Once this change is systematically and clearly defined, we can then proceed to see whether or not we can test the assumption that this change was brought about
266
TUŸBA TANYERI-ERDEMIR
Toprakkale
Karmir-Blur
Ayanis
Eastern Turkey, Armenia, northnorth-east of west of Yerevan Citadel of Van
Eastern Turkey, on the north-western shore of Lake Van, 6 km north of Adilcevaz
North-western Iran, in western Azerbaijan
Eastern Turkey, on the eastern shore of Lake Van, 35 km north of Van
Citadel located on a natural rock spur, connected to the Zimzim mountain
Citadel located on the north-western corner of a partly artificial hill, overlooking Lake Van
Citadel built on top of a hill in the Aq Chay valley, north of Aq Chay river and overlooking the Qara Ziyaeddin plain
Citadel located on top of a hill overlooking Lake Van
Citadel located on a hill next to the banks of Zanga river The lower town extends to south and west of the Citadel
Burneyʼs survey identified a lower town occupation (1957:50–51), however subsequent excavations failed to locate it
The fortress is built in three levels on the bedrock (OB, MB, UB), covering the eastern ridge of the hill
The lower town extends to north (Pınarbaşı), east (Güneytepe) and south of the citadel
Landscape/ Type of Site
Bastam
Location
Adilcevaz/ Kef Kalesi
The lower town extends towards the east and northeast
1898, Prussian team, LehmannHaupt
1939–1950s, 1964 by a joint team Istanbul University from the Academy of Sciences of the Öğün and Bilgiç Armenian SSR and the State Hermitage Museum, Leningrad
1968–1978, German Institute of Archaeology at Teheran
1989–present, Ege University Çilingiroğlu
Kleiss
Excavations
1880s, British team, Rassam, Clayton, Raynolds
Piotrovsky
1911–12, two Russian expeditions, Orbeli and Marr 1959–61 and 1976, Turkish team, Erzen Piotrovsky (1950; 1952; 1955; 1969; 1970), Oghanesyan (1955)
Öğün (1982)
Kleiss (1979; 1988)
Çilingiroğlu and Salvini (2001)
Fig. 1. Site and excavation information for cities built by Rusa II.
Major Publications
Barnett (1950), LehmannHaupt (1907), Öğün (1961)
267
14. THE IMPERIAL ART AND ARCHITECTURE OF URARTU Karmir-Blur
Adilcevaz/ Kef Kalesi
Bastam
Ayanis
Fortification walls, rock cut stairway leading up to the citadel
Fortification walls, city gates
Fortification walls
Fortification walls, 2 city gates
Fortification walls, city gate
Temple of Haldi
A susi Temple, dedicated to Haldi
2 Pillared halls
A large structure with an open court, surrounded by a number of smaller buildings
A large administrative complex (palace?) with 120 rooms organised around a central courtyard
A large administrative complex with 140 rooms
A large administrative complex
A large administrative complex
Storage facilities Cisterns
Storage facilities
Storage facilities
Storage facilities
Storage facilities
N/A
Domestic units, streets and workshops
N/A
Domestic units
Domestic units
Large buttressed structure (administrative?)
Large buttressed structure (administrative?)
Keşiş Göl inscription, found nearby, but not associated with architecture on the site
Two long inscriptions associated with the foundations of the temple
Inscribed basalt blocks with relief sculpture found in one of the pillared halls
Inscription giving the name of the city found on the site, but not associated with architecture
A gate inscription excavated from the city gate
Metal armour, decorative objects and furniture pieces (throne?)
Metal armour, decorative objects and furniture pieces
Metal armour, and furniture pieces
Metal armour
Metal armour, decorative objects and furniture pieces
Figurines (wood)
Figurines (wood) Stone wall mosaics
Stone wall mosaics
Temple of Haldi in a pillared hall
Architecture (Citadel)
Toprakkale
A platform, to the south-west of temple
Building Inscriptions Major finds
Stone wall mosaics
Architecture (lower town)
Stone wall mosaics
A long inscription on the outer wall of the temple
Decorated altar Stone sculpture
Stone sculpture
Stone vessels
Stone vessels
Stone vessels
Seals and bullae
Seals and sealings
Seals and sealings
Seals and sealings
Tablet
Tablets and bullae
Tablets and bullae
Tablets and bullae
Red polished ware
Red polished ware
Red polished ware
Red polished ware
Red polished ware
Fig. 2. Architecture and materials excavated from cities built by Rusa II.
268
TUŸBA TANYERI-ERDEMIR
through the agency of the king and the possibility that it might have been an active component of his strategies of governance.
The Kingdom of Urartu The kingdom of Urartu, located within the present day borders of eastern Turkey, Armenia and north-western Iran, was one of the most prominent powers of the Iron Age in the Near East, from the 9th century to the end of the 7th century BC. Although its frontiers shifted through the course of its existence, the heartland of the kingdom was located in the Lake Van basin. At the zenith of its power, the lands under Urartian domination extended as far as Lake Urmia in the east, Lake Sevan in the north-east and nearly to Erzincan in the west. Urartu seems to emerge as a political power sometime in the earlier part of the 1st millennium BC around the Van basin (Zimansky 1985, 28–31; Erzen 1979, 5). The creation of a solid political entity in this region seems to be a response to the increased Assyrian military campaigns in the area, which might have lead to the unification of dispersed groups for protection and defence. In the absence of archaeological data, it is not surprising that our primary source of information on the formative period of the Urartian kingdom (ca. 13th–9th centuries BC) is the Assyrian textual record. Urartu’s relationship to Assyria as its southern neighbour and major rival had been a very important one in both designating the nature of the Urartian state, as well as its affairs with other political powers of the Near East and the Caucasus. Because of this continuous interaction, the Assyrian record supplies significant insights to the course of events in the Urartian kingdom. In addition to the textual record, the reliefs that adorn the walls of Assyrian palaces contain several narrative scenes of Assyrian military campaigns to the kingdom of Urartu (Gunter 1982). The archaeological and textual sources suggest that by the 9th century BC, the Urartian kingdom flourished as a formidable power that mirrored Assyria in several aspects: in its military power, in the implementation of writing, in its art and even in the definition of ‘kingship’. Starting with the 9th century BC, the known Urartian material culture is outstandingly imperial in nature. This uniformity is usually associated with a strong central authority (Zimansky 1995, 111) and, by implication, royal control over choices in the creation and production of materials.
14. THE IMPERIAL ART AND ARCHITECTURE OF URARTU
269
The Urartian kings were involved in various extensive building projects that required employment of massive amounts of labour. Archaeologically the most significant edifices left to us are citadels that were built on strategic and defendable locations.1 Excavated Urartian citadels show us that storage facilities and temples were prominent components of the Urartian citadels signifying their importance as economic and ritual foci. Urartian temples, like the ones in Kayalıdere, Altıntepe, Anzaf and Ayanis were located at the heart of many citadels (Belli 1997; Burney 1966; Çilingiroªlu 2001; Özgüç 1966). These temples were decorated lavishly. Imperial armour, such as shields, helmets and quivers dedicated to Urartian deities, adorned the exterior facades and sometimes the pillared courtyards of the temples. Temple furnishings also included large bronze cauldrons, thrones, footstools and candelabra. The walls of the cella and the walls of the courtyards were sometimes painted, as at Altıntepe (Özgüç 1966), or covered with stone mosaic inlays as in Ayanis (Çilingiroªlu 2001). Many Urartian buildings and imperial bronze objects bear inscriptions that state the name of the king who built the edifice or dedicated the object to a particular deity. These dedicatory inscriptions constitute the majority of the Urartian epigraphic sources (König 1955; 1957; Melikishvili 1960). In these inscriptions, the kings conventionally identify themselves in their royal titles with the names of their fathers, such as ‘king Ispuini, son of Sarduri’ or ‘king Rusa, son of Argishti’. This tradition enabled the construction of a patrilineal pedigree for the Urartian dynasty. The implementation of Urartian art in imperial ruling strategies and in establishing the legitimacy of the rulers has been addressed (Smith 2000; Tanyeri 2001). A recent study by Adam Smith (2000) on the representations of the built environment in Urartian art raises the question of utilising these images for purposes of legitimacy. One important contribution of this study is the incorporation of affective qualities of representation into an understanding of how legitimacy might have been established in the Urartian kingdom. Tanyeri’s (2001) article, on the other hand, presents an historical sketch of the mechanisms through which Urartian ruling dynasty might have implemented art in order to define the concept of kingship and establish its legitimacy. 1
For example, for the 9th century BC, see Belli 1998; Tarhan 1994. For the 8th century BC, see Erzen 1978; Özgüç 1966; Oghanesyan 1961. For the 7th century BC, see Çilingiroªlu and Salvini. 2001; Piotrovsky 1950; 1952; 1955; 1970; Oghanesyan 1955; Öªün 1961.
270
TUŸBA TANYERI-ERDEMIR
Conceptualisation of Change in Urartian Studies The specific concept of change through time and/or across geography has attracted little attention in Urartian studies. Indeed, the appearance of the Urartian material culture as a relatively uniform entity may be one of the reasons why most scholars in the past had a strong tendency to perceive Urartian art and material culture as static. This understanding is evident, for instance, in van Loon’s (1966, 168) survey of Urartian art: ‘One of the most striking facts about the Urartian court style is its continuity. During the two centuries of its existence (circa 800–600 BC) very little change is visible.’ Perhaps because of this apparent uniformity, scholars opted for highlighting continuities rather than changes and innovations in Urartian material culture through time. Most of the general surveys of material culture have favoured a classification according to categories such as metalwork, architecture, pottery, etc., rather than chronology. The general overviews of art presented by van Loon (1966), Akurgal (1968) and Çilingiroªlu (1998), for example, are organised in this fashion. While this form of categorisation is helpful in presenting a whole array of artefacts executed in a particular medium, such studies risk falling short on two accounts. First, unless a discussion of chronology within the medium under study is presented, a study might give a misleading view of the material culture as uniform over time. Second, such studies may obscure chronological trends that do not stand out strongly in any given category of artefacts, which may be evident only when looked at the broader spectrum of several contemporary categories. Overall, such studies may well have strengthened and perpetuated the view of Urartian art and architecture as changing only in minor and inconsequential ways through time. A similar trend can also be observed in studies of Urartian iconography. Eichler’s (1984) study of representations of supernatural beings in Urartian art, for instance, fails to address the question of whether there is change in the iconographic repertoire employed in Urartian iconography through time. Ta¤yürek’s (1978) investigation of images of the god Haldi presents similar problems. He too, fails to address whether or not there is any change in the representations of Haldi through time. One further problem that these two studies suffer is that they rely almost exclusively on illicitly excavated artefacts, for which there is no contextual information. In terms of addressing change through time in imperial art, Azarpay’s 1968 book Urartian Art and Artifacts: A Chronological Study is unique. It is organised chronologically. Azarpay was able to define four phases of artistic pro-
14. THE IMPERIAL ART AND ARCHITECTURE OF URARTU
271
duction in the Urartian imperial tradition, an early phase, a transitional phase, a second phase and a later phase. In these phases, she was able to see and define several iconographic and stylistic developments that took place during the reigns of various rulers. Her work, however, predates several very important Urartian archaeological excavations, and needs to be reassessed with the new data uncovered in the last three decades. Some scholars working with a variety of categories of material culture have recognised several changes in the material record in the broad history of Urartu. Kleiss (1976, 35–6; 1994) identified a change in building style of the fortress architecture between the 8th and 7th centuries BC. Azarpay and Akurgal made certain comments on the stylistic changes in Urartian imperial art. Azarpay (1968, 72–3) noted a number of changes in iconographic elements and style from the 8th into the 7th century BC. Mostly based on her observations on the decorated two-dimensional metal artefacts, such as bronze belts, she stated that framed friezes and panelled compositions disappeared, and there was a marked decrease in narrative scenes. The incised decorations on metal artefacts in general, and the lion and bulls on metal shields in particular, were less detailed in their depiction. She also suggested that the stepped patterns on the nose wrinkles of the lions were a stylistic development of the 7th century BC (Azarpay 1968, 73). In respect to metal works of art, Akurgal’s stylistic observations are similar to Azarpay’s. Akurgal defined two styles in Urartian art, namely ringenstil (circular stylisation) and buckelstil (embossed stylisation). He suggested that ringenstil was used on artefacts of the 8th and 7th centuries BC, and buckelstil was seen in artefacts of the 7th and 6th centuries BC (Akurgal 1959a; 1959b; 1968). The changes recognised by these scholars, however, have been defined in terms of centuries, without chronologically pinpointing these to the reign of any particular monarch. The archaeological data amassed from previous and ongoing excavations portray a strong contrast between the sites founded by Rusa II (Toprakkale, KarmirBlur, Adilcevaz, Bastam, Ayanis) and the earlier ones of the 9th (Anzaf, Van Kalesi) and 8th (Altıntepe, Çavu¤tepe, Erebuni, Argishtihinili, Menuahinili) centuries BC. Several scholars have noted a change in the quantity of the production of prestigious artefacts and monumental architecture (Çilingiroªlu 1998, 45; Zimansky 1995a, 95). The most visible indication of this is the ambitious building program of Rusa II. Nearly all Urartian kings appear to have involved themselves in massive building projects, including citadels, temples and hydraulic works. It was also common for an Urartian king to name a major fortress he
272
TUŸBA TANYERI-ERDEMIR
had founded after himself, such as Sardurihinili or Argishtihinili. Rusa II seems to have gone to an extreme in both of these traditions; the magnitude of his building activity surpassed the constructions of all the previous rulers. To date, there are five known and excavated sites founded by Rusa II, and of those, three – Ayanis (Rusahinili Eiduru-kai), Toprakkale (Rusahinili) and Bastam (Rusa-i URU.TUR) – bear his name (Zimansky 1995a). It is also noteworthy that two fortresses built by Rusa II are located in close proximity to earlier citadels: Karmir-Blur in the immediate vicinity of the 8th-century citadel of Erebuni, and Toprakkale near Van Kalesi. The innovations in material culture associated with Rusa II that have been noted include cuneiform writing on clay tablets and the use of sealed bullae (Salvini 1995, 103). Another innovation is architectural decorations in stone, both with relief sculptures such as the examples from Adilcevaz and Kef Kalesi (Burney and Lawson 1958; Öªün 1982) and with stone inlay mosaics from Toprakkale, Bastam and Ayanis (Çilingiroªlu 2001, 41–4; Kroll 1979, 160, pl. 49.2). Moreover, Çilingiroªlu (1998, 144–5) notes that the stone vessels excavated from Ayanis are seen only in this period. An increase in the Assyrian influence on art of Rusa II has been observed. Half a century ago, Barnett recognised a strong Assyrian influence on several metal artefacts from Toprakkale and concluded that they predated Sargon II’s Eighth Campaign, 714 BC (Barnett 1950, 35–7). Based on the assumption that the campaign would have severed relations between Urartu and Assyria, Barnett (1950, 39) thought that the Assyrian influence could not have been strong in the Urartian art of the 7th century BC. Azarpay (1968, 64) later dated these artefacts to the 7th century based on their stylistic and epigraphic properties. The epigraphic evidence showed that the inscriptions on some of these objects were written in an Assyrian type of cuneiform that was used starting with Rusa II, but not attested during the reign of Rusa I. These yet sketchily defined changes in material culture can further be refined precisely by conducting a systematic survey of the material culture. My primary aim is to think of ways of seeing how Rusa II might have used material culture as an active component in his governance. In this pursuit, I follow Harvey (1990, 226–7) in the understanding that ‘ideological and political hegemony in any society depends on the ability to control the material context of personal and social experience.’ In order to have a clear idea about how Rusa II might have changed the material context of personal and social experience for his subjects in an effort to secure his legitimacy, we should first define innovations, changes and conti-
14. THE IMPERIAL ART AND ARCHITECTURE OF URARTU
273
nuities in the material record. The hypothesis that Rusa II introduced new categories of material culture which were not present prior to his reign, such as clay tablets, seals or, stone vessels, needs to be addressed. We also need to define the new stylistic developments and iconographic additions to the artistic repertoire. Change in material culture should be documented both quantitatively and qualitatively. Continuities in the stylistic and iconographic traditions, both in the already existing and new categories of material culture need to be identified and documented. Once we clearly define innovations, changes and continuities in imperial art and architecture, then we can aim to test the agency of Rusa II in bringing about any change, and gain insight into the mechanisms of expression of kingship that might have been employed by Rusa II to mark his own reign as distinct from the earlier kings.
Agency, Structure and the Dynamics of Change Theoretical approaches developed within the last few decades in the fields of anthropology and sociology have led to a better understanding the role of agency within dynamic social structures (Bourdieu 1990; 1997; Giddens 1979). Unlike methodological individualism or methodological structuralism, the approaches developed by Bourdieu and Giddens incorporate both structure and agency, while emphasising their interdependence. They can account for the choices and strategies of agents, as well as the conditions of the socio-political and economic contexts. In these theories, ‘every social actor knows a great deal about the conditions of reproduction of the society in which he or she is a member’ (Giddens 1979, 5). The theories of Bourdieu and Giddens, which have significant similarities, can be applied in this study in the following manner: 1) the Urartian ruler can be considered an active agent, operating in an already established and dynamic structure of his imperial realm; 2) within this context, the ruler has significant, but not necessarily complete knowledge pertaining to the conditions of reproduction of systems and structures of power and authority; 3) we can thus consider the possibility of his exercising conscious choices. In this case, the choices pertaining to the creation of an imperial art are particularly important to consider. I argue that through making deliberate choices Rusa II was capable of modifying and restructuring the context within which he was operating. This also implies a dynamic and reciprocal interaction between the agent and the structure.
274
TUŸBA TANYERI-ERDEMIR
These choices would be related to the construction and assertion of legitimacy in two ways. First, they define a dynast’s power and authority via culturally accepted parameters, and second, they illustrate and underline the difference of one king’s authority and governance from that of his predecessors. For the definition of authority, we can turn to Bourdieu’s concept of symbolic power. Bourdieu (1991, 166) defines symbolic power as a ‘power of constructing reality’ which is implanted through art, religion and language. These structures would enable a person, a group, or institution to exert power by ‘structuring the perception which social agents have of this world’ (Bourdieu 1991, 105). Following Irene Winter (1998, 57), I consider style to be an active component of this dynamic. By the time of Rusa II, the structure of art in the imperial realm of Urartu had already embodied a certain repertoire of symbols and styles for the definition and communication of ‘kingship’. This being the case, we can investigate how Rusa II situated himself in this structure, by retaining some of these styles and symbols, altering others and creating wholly new ones. In this respect, understanding how a ruler differentiates himself from his predecessors becomes very important. For the issue of differentiation, I build upon Hobsbawm (1997, 10) for whom the past is ‘a permanent dimension of the human consciousness, an inevitable component of institutions, values and other patterns of human history’. The vital component in defining change in any given ‘present’ requires an illustration of how that change is situated in relation to a certain ‘past’. The changes taking place in the reign of Rusa II are only meaningful in the context of their relation to what existed before, in the ‘past’. In the creation of that ‘controlled past’, style is a powerful tool, which can manipulate the acts of ‘remembrance’ and ‘forgetting.’ The ruling authority can either choose to establish visual and mental links with, or to disassociate itself from, a selected portion of the past. One way of achieving this is to manipulate material culture, using style instrumentally to establish visual and mental links with the material culture expressions of a particular period in history. Zanker (1990, 5–31, 37–43) demonstrated that Augustus consciously selected themes and styles from the art of his predecessor Caesar, while he excluded the iconography and style of the ‘Dying Republic’ from his new artistic repertoire. Rusa II might have carried a similar program to Augustus’s conscious selection of iconography and style in order to highlight specific connections to some of his predecessors, while pronouncing a break from some others. This can be demonstrated by looking at the patterns of change and continuity in the dynastic art, and analysing its meaning, as Zanker has done, by investigating these changes in the light of socio-political affairs of the period.
14. THE IMPERIAL ART AND ARCHITECTURE OF URARTU
275
The question that remains is whether or not we can envisage the agency of Rusa II in the production of material culture. To what extent can we attribute the changes observed in the material record to the king’s conscious decisions? Is it possible that the changes taking place resulted from increased contacts with Assyria? Were these changes a societal response to a cultural trauma such as Sargon II’s attack on Urartu in 714 BC? In order to answer these questions, and to test the hypothesis presented, we could conduct the following steps of investigations. 1) The first step should be an exploration of whether there is uniformity in the imperial art and architecture through space during the reign of Rusa II. If the changes in the material culture are results of increased contacts with Assyria during this period, we would expect the material culture of the immediate contact areas to be most strongly affected, and consequently to show more Assyrian influence. Thus, an uneven distribution of Assyrian influence over geography, with a strong concentration of material culture showing strong Assyrian influence in the contact areas, and possibly even a visible diffusion of this influence over space and time would indicate that these changes could have resulted from increased contacts with Assyria. If a strong Assyrian influence in the mid-7th century works can be detected, as has been noticed by Azarpay (1968, 63–5) and Barnett (1950, 35–7) previously, and if this Assyrian influence is uniform through space, it needs to be considered along with questions pertaining to change, in terms of what this means vis-à-vis the ruling authority and the general population of Urartu. 2) If there is a considerable uniformity in material culture throughout a vast area, it will be considered as an indicator of the presence of a strong central authority closely involved in its production. For this, we should compare the data from Karmir-Blur (Armenia), Bastam (Iran), Ayanis, Toprakkale, Adilcevaz (Turkey), and see whether there are striking similarities through which we can observe a certain level of uniformity. 3) If a strong central control over production can be observed, we can then conduct a study in order to see whether it can be linked to the monarch. The agency of the king in the creation of an imperial artistic tradition has been successfully investigated in several recent studies of Assyrian art (Russell 1987; 1991; Winter 1981; 1997; 1998). The methodology used in these studies employs textual evidence in both establishing the involvement of the king in the production of material culture, and in assigning meaning to changes in art. Russell (1987, 537), for instance, demonstrated how Sennacherib’s selections in decorating Court VI of his palace at Nineveh might have played an active role in his governance strategies, particularly in terms of ‘maintenance of the boundaries
276
TUŸBA TANYERI-ERDEMIR
of empire and the creation of a center’ by analysing the correlation between the monumental display inscriptions and themes represented in art, and looking at the observed changes in the historical context. Winter (1997, 367) illustrated the active involvement of Esarhaddon in reviewing the options in the production of his sculpted image from textual references. On a more theoretical level, the concept of agency has been investigated recently by Winter (1996; 1998) in terms of both production and perception of art. Her inquiry on style and agency are particularly appropriate for my research, as she explored the possibility of how style might have been employed actively as a carrier of meaning in governance strategies. The abovementioned studies are particularly important for this project, as the nature of the material and textual evidence in Urartu is comparable to Assyria. The written record has the potential to provide important evidence for an investigation of the involvement of the ruler. Through a study of the Urartian epigraphic evidence, it is possible to see to what extent the rulers claim their involvement with particular edifices, artefacts or events. In this respect, we can regard the presence of dedicatory inscriptions as a sign of Rusa II’s involvement in the production. We should also check for correlations between the written word and art. The historical context will allow us to see whether a correlation in changes in art, architecture and textual evidence might be meaningful in terms of serving for a specific agenda related to the king. This then should be considered as a strong indication of the king’s agency. The Urartian epigraphic sources are helpful in defining change through time. There are over 400 known monumental inscriptions from all known Urartian rulers, starting with Sarduri I (ca. 832–825 BC) up to the very last Urartian kings of the early 6th century BC in an unbroken sequence. All these inscriptions, almost without exception, are written as words from the king’s mouth. The continuities and discontinuities in the epigraphic record are particularly important in designating the changes in ruling strategies and practices. 4) If we are to think that these changes are caused by a social trauma, such as Assyrian king Sargon II’s attack on Urartu in 714 BC, then one may expect the changes to appear in material culture within a fairly short time following the traumatic event. If indeed Sargon II’s attack can be considered as a likely candidate, we should expect to see changes in the material record either in the last decade of the 8th century, or in the first decades of the 7th. In shedding light on this problem the establishment of a chronology of the changes becomes very important. We should investigate the chronological connections of the changes in the material culture with the artistic traditions of
14. THE IMPERIAL ART AND ARCHITECTURE OF URARTU
277
Assyria, Neo-Hittite states, north-western Iran and the Caucasus to see whether we can anchor some of the changes by establishing chronological parallels with other cultures. We can then assess the evidence to see whether the scarcity of materials from Argishti II’s reign is the result of a considerably lower rate of production, or is archaeologically underrepresented for some reason. 5) If a level of certitude can be established with respect to the involvement of the monarch in the production of material culture through the investigations in steps two and three, we further need to illustrate that it is Rusa II who was responsible for the changes taking place in this period, and not his immediate predecessor Argishti II. In this respect, the results of the investigation in step four will be decisive. In conducting these steps of investigations, we should keep in mind that different types of material culture are apt for answering different questions. In Fig. 3, a set of six key questions and differing values of various types of material culture is presented. Once these steps of investigations are conducted, we can then weigh the evidence to see whether they indicate that the changes in the material culture were brought about by the agency of Rusa II, or they occurred independent of the king’s involvement.
Conclusion and Summary Giddens’s (1979) and Bourdieu’s (1990; 1997) theoretical approaches to agency can account for the choices and strategies of agents, as well as the conditions of the socio-political and socio-economic structures in which agency is manifested. This study aims to define how a particular agent in an ancient society was acting in the already established and dynamic structure of his imperial realm, and how agency can be investigated in the archaeological record through the material remains of his reign. In this paper, I present a potentially testable hypothesis to investigate the possibility that changes in the material culture of the 7th-century Urartian kingdom might have been brought about through the agency of the ruler, and consciously employed by him in manipulating his ruling ideology. The archaeological data from the reign of Rusa II not only presents an ideal case to test the agency of this particular ruler in the production of royal art and architecture, but also offers the possibility of investigating how he might have used material culture actively in his governance strategies.
278
TUŸBA TANYERI-ERDEMIR Category Metal
Estimated number 300
Sub categories
Inscription
Decoration
Shields
Y
Y
5
4
5
5
4
2
N
4
2
5
5
4
2
N Helmets
Y
N
Quivers
Y
N
C
D
E
F
Y
4
4
3
3
3
1
N
2
1
1
2
1
1
Y
5
4
5
5
4
2
N
2
2
5
5
4
2
Y
4
4
3
3
3
1
N
2
1
1
2
1
1
Y
5
4
5
5
4
2
N
2
2
5
5
4
2
Y
4
4
3
3
3
1
N
2
1
1
2
1
1 1
N
N
1
2
2
1
1
Spearheads
Y
N
2
2
4
4
2
1
Nails
N Y
N N
1 2
2 1
1 4
1 4
1 2
1 1
Furniture attachments
N
N
1
1
1
1
1
1
Y N Y
Y Y Y
4 2 5
4 4 5
5 3 5
4 3 4
3 2 5
2 2 2
N
2
2
4
4
5
2
Y
4
5
4
3
4
2
N
2
2
1
2
3
1
N
Cauldrons
Y
N
Y
4
5
5
4
5
2
N
3
3
3
4
4
2
Y
3
4
3
4
4
2
N
2
2
1
1
1
1
Ornaments
Y N N
Y Y Y
4 3 4
4 4 4
3 2 4
2 1 3
4 4 3
2 1 2
Vessels
Y
Y N Y N Y Y Y Y N
5 4 3 3 3 3 3 3 2
4 2 4 1 5 5 5 5 3
5 5 3 2 5 4 4 4 2
4 4 2 1 5 5 5 4 2
3 2 2 1 5 5 5 4 2
2 2 2 1 2 2 2 2 1
Figurines
N 30
B
Arrowheads
Wall plaques
Stone
A
Sculpture in relief Wall mosaics Stone vessels
Y N N N
Fig. 3. Materials to be studied and their estimated values in answering key questions of research (5 highest – 1 lowest). Question A: Value in defining continuity and change in Urartian dynastic artistic tradition (internal development) Question B: Value in defining relations of Urartian art with the other artistic traditions of the Near East (external influence) Question C: Value in dating Question D: Importance in defining change in dynastic propaganda Question E: Importance in defining change in state religion Question F: Importance in defining change in governance strategies and practices
279
14. THE IMPERIAL ART AND ARCHITECTURE OF URARTU
Seals
100
Y
N
Clay Tablets
30
Cuneiform
Y
Hieroglyph
Bullae
Ivory
Y
100
15
Furniture attachments
Y
4
5
5
5
5
5
N
3
3
4
4
4
5
Y
4
5
4
4
4
5
N
1
1
1
1
1
4
Y
5
5
5
5
3
5
N
1
1
5
5
3
5
Y
4
4
3
3
2
5
N
1
1
3
3
2
5
Y
Y
5
5
5
5
3
5
N
N Y
1 4
1 4
5 4
5 5
2 3
5 5
N Y
1
1
1
1
1
4
Y
4
5
4
3
5
1
N N
Y N
Figurines
Y
5
5
5
3
5
2
N N Y
Y N Y N Y Y Y
3 3 4
3 3 4
2 2 4
1 1 4
2 2 3
1 1 1
N
Y
4
4
4
2
2
1
N Bone Wood Ceramics
5 5 15
Acknowledgments I am grateful to Paul Zimansky, Irene Winter, Mary Beaudry, Susan Allen, Ömür Harman¤ah, Elif Denel, Andrew Cohen, Heidi Luchsinger, Volkan Rodoplu, David Stone and Aykan Erdemir for reading and commenting on the earlier versions of this paper, and Jülide Aker and Stephan Brighton for bringing several references cited here to my attention. Any possible errors, however, are my responsibility. I would also like to thank the organisers and participants of the University of Chicago Conference on Eurasian Archaeology, and Özgür, Çaªatay and Duru Güney for being my hosts in Chicago.
Bibliography Abay, E. 2001: Seals and Sealings. In Çilingiroªlu and Salvini 2001, 320–53. Akurgal, E. 1959a: Urartu Medeniyeti. Anatolia 4, 67–75. ——. 1959b: Urartäische Kunst. Anatolia 4, 77–114. ——. 1961: Die Kunst Anatoliens von Homer bis Alexander (Berlin). ——. 1968: Urartäische und Altiranische Kunstzentren (Ankara). Azarpay, G. 1968: Urartian Art and Artifacts: a Chronological Study (Berkeley/Los Angeles). Barnett, R.D. 1950: The Excavations of the British Museum at Toprak Kale, near Van. Iraq 12, 1–43. Belli, O. 1998: Anzaf Kaleleri ve Urartu Tanrıları (Istanbul).
280
TUŸBA TANYERI-ERDEMIR
Bourdieu, P. 1990 [1980]: The Logic of Practice (Stanford). ——. 1991: Language and Symbolic Power (Cambridge, Mass.) ——. 1997 [1977]: Outline of a Theory of Practice (Cambridge). Burney, C.A. 1966: A First Season of Excavations at the Urartian Citadel of Kayalıdere. Anatolian Studies 16, 55–111. Burney, C.A. and Lawson, G.R.J. 1958: Urartian Reliefs at Adilcevaz, on Lake Van, and a Rock Relief from the Karasu, near Birecik. Anatolian Studies 8, 211–8. Chapman, J. 2000: Tension at Funerals: Social Practices and the Subversion of Community Structure in Later Hungarian Prehistory. In Dobres and Robb 2000a, 169–95. Çilingiroªlu, A.A. 1983: Mass Deportation in the Urartian Kingdom. Anadolu Ara¤tırmaları 9, 319–23. ——. 1994: Urartu Tarihi (Bornova). ——. 1998. Urartu Krallıªı Tarihi ve Sanatı (Izmir). ——. 2001: Temple Area. In Çilingiroªlu. and Salvini 2001, 36–65. Çilingiroªlu, A. and Salvini, M. (eds.) 2001: Ayanis I: Ten Years’ Excavations at Rusahinili Eiduru-kai 1989–1998 (Rome). Dobres, M.-A. and Robb, J.E. (eds.) 2000a: Agency in Archaeology (London/New York). ——. 2000b: Agency in Archaeology: Paradigm or Platitude? In Dobres and Robb 2000a, 3–17. Earle, T. 1990: Style and Iconography as Legitimation in Complex Chiefdoms. In Conkey, M.W. and Hastorf, C.A. (eds.), The Uses of Style in Archaeology (Cambridge), 73–81. Eichler, S. 1984: Götter, Genien und Mischwesen in der Urartäischen Kunst (Berlin). Erzen, A. 1978: Çavu¤tepe I (Ankara). ——. 1979: Eastern Anatolia and Urartians (Istanbul). Forbes, T.B. 1983: Urartian Architecture (Oxford). Giddens, A. 1979: Central Problems in Social Theory Action, Structure and Contradiction in Social Analysis (Hong Kong). Gunter, A. 1982: Representations of Urartian and Western Iranian Fortress Architecture in Assyrian Reliefs. Iran 20, 103–12. Harvey, D. 1990: The Condition of Postmodernity: An Enquiry into the Origins of Cultural Change (Cambridge, Mass./Oxford). Hobsbawm, E. 1997: On History (New York). I¤ık, C. 1976/77: Adilcevaz Urartu Mezarlıªında Bulunan Bir Mühür – A Cylinder Seal from Adilcevaz. Anatolia 20, 83–92. Joyce, A.A. 2000: Agency and the Founding of Monte Albán: Sacred Propositions and Social Practices. In Dobres and Robb (eds) 2000, 71–91. Kleiss, W. 1976: Urartäischer Architektur. In Keller, H.-J., (ed.), Urartu: Ein wiederentdeckter Rivale Assyriens (Munich), 28–44. ——. 1979: Bastam I: Ausgrabungen in den Urartäischen Anlagen 1972–1975 (Berlin). ——. 1988: Bastam II: Ausgrabungen in den Urartäischen Anlagen 1977–78 (Berlin). ——. 1994: Notes on the Chronology of Urartian Defensive Architecture. In Çilingiroªlu, A. and French, D. (eds.), Anatolian Iron Ages 3 (Ankara) 6–12. König, F.W. 1955: Handbuch der chaldischen Inscriften I (Graz). ——. 1957: Handbuch der chaldischen Inscriften II (Graz). Kroll, S. 1979: Die Kleinfund. In Kleiss, W. et al., Bastam I: Ausgrabungen in den Urartäischen Anlagen 1972–1975 (Berlin), 151–82. Lehmann-Haupt, C.F. 1907: Materialen zur älteren Geschichte Armeniens und Mesopotamiens (Berlin). Melikishvili, G.A. 1960: Urartskie klinoobraznye nadpisi (Moscow). Oghanesyan, K.L. 1955: Karmir Blur V: Arkhitektura Teisebaini (Yerevan). ——. 1961: Arin-Berd I. Architektura Erebuni: Po materialam raskopok 1950–1959 (Yerevan). Öªün, B. 1961: Kurze geschichte der Ausgrabungen in Van die Türkischen Versuchsgrabungen auf dem Toprak-kale 1959. Zeitschrift der Deutschen Morganländischen Gesellschaft 111, 254–82. ——. 1967: Urartu Ara¤tırmalarının Anadolu, Yunanistan ve Etrüsk Tarihi ve Arkeolojisi Bakımından Önemi. VI. Türk Tarih Kongresi, 65–71. ——. 1982: Die urartäischen Paläste und die Besattungsbräuche der Urartäer. In Papenfuss, D. and Strocka, V.M. (eds.), Palast und Hütte (Mainz), 217–36.
14. THE IMPERIAL ART AND ARCHITECTURE OF URARTU
281
Özgüç, T. 1966: Altıntepe I: Architectural Monuments and Wall Paintings (Ankara). ——. 1969: Altıntepe II: Tombs, Storehouses and Ivories (Ankara). Piotrovsky, B.B. 1950: Karmir-Blur I: Archeologiceskie Raskopki v Armenii 1: 1939–1949 (Yerevan). ——. 1952: Karmir-Blur II: Archeologiceskie Raskopki v Armenii 2 (Yerevan). ——. 1955: Karmir-Blur III: Archeologiceskie Raskopki v Armenii 5: 1951–1953 (Yerevan). ——. 1969: Urartu (Munich/Geneva/Paris). ——. 1970: Karmir Blur (Leningrad). Russell, J.M. 1987: Bulls for the Palace and Order in the Empire: The Sculptural Program of Sennacherib’s Court VI at Nineveh. Art Bulletin 69.4, 520–39. ——. 1991: Sennacherib’s Palace without a Rival at Nineveh (Chicago). ——. 1998: The Program of the Palace of Assurnasipal II at Nimrud: Issues in the Research and Presentation of Assyrian Art. American Journal of Archaeology 102, 655–715. Salvini, M. 1995: Geschichte un Kultur der Urartäer (Darmstadt). ——. 2001: Monumental Stone Inscriptions. In Çilingiroªlu. and Salvini 2001, 251–70. Sassaman, K.E. 2000. Agents of Change in Hunter-Gatherer Technology. In Dobres and Robb 2000a, 148–68. Sayce, A.H. 1925: The Kingdom of Van (Urartu). In Cambridge Ancient History vol. III: The Assyrian Empire (Cambridge), 169–86. Seidl, U. 1979: Die Siegelbilder. In Kleiss, W. et al., Bastam I: Ausgrabungen in den Urartäischen Anlagen 1972–1975 (Berlin), 137–49. ——. 1988: Die Siegelbilder. In Kleiss, W. et al., Bastam II: Ausgrabungen in den Urartäischen Anlagen 1977–1978 (Berlin), 145–54. Smith, A.T. 1999: The Making of an Urartian Landscape in Southern Transcaucasia: a Study of Political Architectonics. American Journal of Archaeology 103, 45–71. ——. 2000: Rendering the Political Aesthetic: Political Legitimacy in Urartian Representations of the Built Environment. Journal of Anthropological Archaeology 19, 131–63. ——. 2001: The Limitations of Doxa: Agency and Subjectivity from an Archaeological Point of View. Journal of Social Archaeology 1.2, 155–71. Tanyeri, T. 2001: Urartu Krallıªı’nda √ktidar ve Sanat √li¤kileri. Sanat Dünyamız 80, 213–8. Tarhan, T. 1994: Recent Research at the Urartian Capital Tushpa. Tel Aviv 21.1, 22–7. Ta¤yürek, O.A. 1978: Darstellungen des Urartäischen Göttes Haldi. In Studien und Kultur Kleinasiens II: Festschrift für Friedrich Karl Dörner, zum 65. Geburtstag am 28. Februar 1976 (Leiden), 940–55. van Loon, M.N. 1966: Urartian Art: Its Distinctive Traits in the Light of New Excavations (Istanbul). Wartke, R.-B. 1990: Toprakkale. Untersuchungen zu den Metallobjecten im vorderasiatischen Museum zu Berlin (Berlin). Williams, R. 1977: Marxism and Literature (Oxford/New York). Winter, I.J. 1977: Perspective on the Local Style of Hasanlu IVB: Study in Receptivity. In Levine, L.D. and Young, T.C. (eds.), Mountains and Lowlands: Essays in the Archaeology of Mesopotamia (Malibu), 371–86. ——. 1981: Royal Rhetoric and the Development of Historical Narrative in Neo-Assyrian Reliefs. Studies in Visual Communication 7, 2–38. ——. 1996: ‘Agency’ An Alternative to Subjectivity. In Garber, M., Franklin, P.B. and Walkowitz, R.L. (eds.), Field Work: Sites in Literary and Cultural Studies (New York/London), 196–203. ——. 1997: Art in Empire: The Royal Image and the Visual Dimensions of Assyrian Ideology. In Parpola, S. and Whiting, R.M. (eds.), Assyria 1995 (Helsinki), 359–81. ——. 1998: The Affective Properties of Styles: An Inquiry into Analytical Process and the Inscription of Meaning in Art History. In Jones, C.A. and Galison, P. (eds.), Picturing Science Producing Art (New York/London), 55–77. Zanker, P. 1990: The Power of Images in the Age of Augustus (Ann Arbor). Zimansky, P.E. 1985: Ecology and Empire: The Structure of the Urartian State (Chicago). ——. 1995a: An Urartian Ozymandias. Biblical Archaeologist 58.2, 94–100. ——. 1995b: Urartian Material Culture as State Assemblage: An Anomaly in the Archaeology of Empire. BASOR 299, 103–15. ——. 1998: Ancient Ararat: A Handbook of Urartian Studies (New York).
CHAPTER FIFTEEN SEQUENCES OF SIGNS: EURASIAN ARCHAEOLOGY FROM A PERSPECTIVE OF CULTURAL SEMIOTICS GREGORY E. ARESHIAN
The reconstruction of the prehistory of speakers of Indo-European languages and of their cultures has been and will remain one of several major problems of Eurasian archaeology. For more than a century scholars have attempted to identify a specific region as the homeland of the Proto-Indo-Europeans and to pinpoint an archaeological culture that can be attributed to them. Depending on the suggested localisation of the homeland, subsequent migrations of IndoEuropeans have been tentatively reconstructed both linguistically and archaeologically. However, several scholars (Trigger 1989, 244, 288, 383–4), including the author of this paper, have demonstrated years ago the methodological inadmissibility of straightforward identifications of previously described archaeological cultures with speakers of specific languages and with specific ethnic groups. The total scholarly experience of the 20th century has come to confirm Edward Sapir’s conclusion expressed almost 90 years ago: That differences in culture ever neatly correspond to differences of race and language can not be maintained, but I wish to point out that the numerous homologies are of at least as great historical importance as the discordances . . . The units of distribution of these three sciences [i.e. physical anthropology, archaeology/ethnology and linguistics], while never coinciding throughout, do nevertheless show significant lines of accord. (Sapir 1994, 42–3)
Therefore, what indeed it is possible to do is to study the variety of probable correlations between linguistic models of the Proto-Indo-European culture and systems of archaeological assemblages reflecting similar levels of technology, social organisation and ideology. In actual practice it would mean that several synchronic and diachronic sets of archaeological cultures could be outlined as potential correlates of the Proto-Indo-Europeans. Within this broad perspective the validity of archaeological reconstructions of Indo-European migrations must be touched upon as well. The history of archaeology has clearly indicated that
15. EURASIAN ARCHAELOGY AND CULTURAL SEMIOTICS
283
the process of archaeological reconstructions of prehistoric and historic migrations in the Old and New Worlds must follow the paths of groups of migrating populations step by step, which gives initial preference to the study of intra-regional migrations and relegates interregional and intercontinental reconstructions to consequent exploration. Naturally, such research should elaborate on the variety of factors and models of migrations (Anthony 1997, 22–9). One of the major sub-fields of Indo-European studies in which archaeology has been almost completely neglected is the reconstruction of myths and rituals of Indo-European origin. Until very recently the study of Indo-European myths and rituals had been viewed as an exclusive domain of philology and linguistics (Puhvel 1987, 20). The examination of pertinent data from archaeology and art history has rarely been integrated with an analysis of written texts. Among few exceptions from this general condition are Garret Olmsted’s book devoted to the study of the Gundestrup cauldron (1979), one of the last papers of Marija Gimbutas (1990) and only a handful of other publications. This has been natural development of research because the whole idea of Indo-European prehistory had emerged gradually within the framework of comparative linguistics and comparative philology. The linguistic material was overwhelming in quantity and all the different theoretical approaches were first and foremost applied to that material. Meanwhile archaeology was coming of age. Since then more and more ancient cultic sites, tombs and objects of art depicting mythological and ritual scenes have been excavated and meticulously documented. While juxtaposing linguistic and archaeological data, one would be looking for a methodology for their integration. Here appears the advantage of an application of basic principles of cultural semiotics, that allows us to reveal hidden links between archaeological and linguistic evidence. Semiotics would perceive and analyse the data from linguistics, philology, folklore, archaeology and art history as components of an integral cultural metatext communicating specific messages through a variety of codes. The archaeological records comprised of cultural topography, depositional evidence (stratigraphy, planigraphy), artefactual typology, etc., function as one of such codes. Since the quality of field work has been consistently improving during the 20th century, more and more archaeological evidence can be studied within four interconnected dimensions of semiotic research: semantics, syntactics, text semiotics and pragmatics. Various semiotic ideas implicitly or explicitly underlie the sub-field of symbolic archaeology. Yet archaeological records are discussed rarely and only implicitly from
284
GREGORY E. ARESHIAN
the perspective of syntactics and text semiotics. Semantic interpretation of nonverbal components usually shed light on socio-cultural environment (usually referred to as ‘social context’) of a metatext under consideration. At the same time, the study of relationships between different units composing a metatext, which is a goal of textual semiotics, reveals the specificity of cultural messages. In other words, syntactic analysis allows us to achieve a closer understanding of individual and collective creators of cultural messages, imposing also more objective restrictions on our interpretations of the past, and, therefore, increasing the probability of reliable conclusions. The more we find specific syntactic connections between components of a metatext, the closer we get to correct interpretation of the cultural message from its creators and perpetuators. The more differentiated is the system of interconnected components, the more specific is its origin. This may be clarified by a simple analogy. A cultural metatext being subjected to interdisciplinary semiotic study looks like a parallel multilingual inscription written in languages of archaeology, linguistics, art history, folklore, etc. None of the languages is unknown, although the understanding of each language varies substantially. But unfortunately the inscription is damaged. Several parts of the text are missing from all versions. Others are preserved only in one of the languages and therefore need to find the right place in the original text. But fortunately several identical sections of the text are preserved in different languages, making the task of reconstructing and interpreting the whole text a plausible and thrilling interdisciplinary enterprise. Within the framework of interdisciplinary Indo-European studies, data pertaining to the Middle and Late Bronze Age of Caucasia and Anatolia offer a unique opportunity for semiotic exploration. One of the zoomorphic pictoral compositions recurrently appearing on objects of ancient art from the region presents a mediated pursuit, in which stags are chased by wild goats that in their turn are followed by a pair of male wolves. To the best of our current knowledge, the earliest figurative composition of this kind is found on a bronze sistrum from Horoztepe (Özgüç and Akok 1958, 48, fig. 12) in north-central Anatolia near the coast of the Black Sea in modern Turkey. The sistrum originates from a tomb dated within the time frame of 2400–2000 BC according to different chronologies (De Jesus 1980, 131; Yakar 1985, 31). Ten bronze figurines decorate the external perimeter of the rectangular frame of the sistrum (Fig. 1.1). They are arranged into three levels along the vertical axis of the sistrum. Two dogs or wolves, the species of which can-
15. EURASIAN ARCHAELOGY AND CULTURAL SEMIOTICS
285
Fig. 1. 1, Horoztepe (after Özgüç and Akok 1958); 2–3, Nerkin Getashen (after Areshian 1988 and Kushnareva 1993); 4, Khanlar (after Gummel 1992); 5, Treli (after Abramishvili 1978); 6, Ust Labinskaya (after Guguev 1992); 7, Gundestrup (after Filip 1977).
286
GREGORY E. ARESHIAN
not be identified more accurately, occupy the lower register. From there each of them chase a mountain goat placed in the middle level. In their turn the goats are running toward the upper register where figurines of one male stag and five female (or young) deer are located. At this point, one should mention the correspondence of each of the three different species of animals to one specific level on the vertical axis. The next object of art presenting a similar figurative composition is a clay jar found at Nerkin Getashen on the shore of Lake Sevan in Armenia (Kalantarian and Khachatrian 1969). Dated to ca. 1900–1750 BC (Areshian 1988a, 84), it belongs to one of the archaeological cultures representing the Trialeti cultural sequence, thus being contiguous with the scene from Horoztepe from both chronological and spatial perspectives. The jar is covered with red slip, while decorations are drawn in black, forming a horizontal frieze (Fig. 1.2–3). The latter is composed of six running animals painted with great attention to details. The first animal is a female deer, followed by a male stag, then runs a male mountain goat with a she-goat behind it, which is chased finally by two male wolves. Above each animal a waterfowl (probably a goose, or a swan) is depicted, signifying according to one possible interpretation the souls of animals (Areshian 1988a, 85–6). A disk above the back of the she-goat might signify the full moon. Thus, the Getashen jar provides us with supplemental information, necessary for a more accurate reconstruction of the composition from Horoztepe. From now on, one can assume that the predators pursuing goats are twin wolves, and the whole mythological metatext is expanding with the inclusion of birds. A third artefact which extends the evidence of existence of the pictorial scene under consideration into the Late Bronze Age of the same region is a deep clay bowl with a brownish black surface found in the barrow (kurgan) no. 13 near Khanlar in the middle part of Kura river valley in Azerbaijan (Gummel 1940, 28, 31, 61–2, fig. 11IV-IA, IV-IB, 26IA, IB; Piotrovsky 1949, 90). The barrow dates back to ca. 1500–1200 BC according to our current perception of Caucasian chronology. The inner surface of the bowl is covered by an incised decoration with an inlay of white paint (Fig. 1.4). It depicts the same sequence of animals: ‘wolf-goat-stag’, but with an addition which is very important for understanding the probable variety of nonverbal codes presenting the mythological metatext under reconstruction. In this case, each species of animals is depicted in conjunction with a specific ornamental symbol: the canine is linked to meanders, the goat to a sequence of hatched triangles, and the stag to a tree
15. EURASIAN ARCHAELOGY AND CULTURAL SEMIOTICS
287
branch or herringbone ornament, which can be interpreted as a signifier for tree. Wherefore this metatextual structure may have been identified by the sequence ‘meander-hatched triangle-tree’ and, if developed further, it could acquire a hieroglyphic expression. A number of other artefacts of the Middle and Late Bronze as well as of the beginning of the Iron Age of Caucasia indicate the widespread character of this mythologema. They either depict the same animals combined with other personages and symbols (Middle Bronze Age clay vessels from Akhchia and Aruch, Late Bronze Age bronze standard from Lori-Berd) or present those animals individually, as it has been attested by bronze figurines found in funerary assemblages (Lchashen, Shirakavan, Artik, Tolors, etc.) (Areshian 1988a, 86, 88–91). The fact that all of the aforementioned objects of art were found in tombs establish the direct relationship of our mythological metatext with funerary rituals. Yet another piece of evidence, coming from outside of art history, independently confirms this conclusion. Archaeological planigraphy, i.e. situational records of the tombs near Khanlar provide the opportunity for such a verification. The excavation was carried out in 1941, yet its results were published only recently (Gummel 1992). Underneath barrow no. 149, a woman’s skeleton was found, buried on top of an extinguished funeral pyre (Gummel 1992, fig. 2). Together with her, complete skeletons of a stag and a snake, as well as bones of a goat or sheep and of a canine were uncovered in the large grave pit. The typology of the pottery and an imported Mittanian cylinder seal found in this assemblage allow us to date the grave ca 1500–1300 BC. Barrow no. 150 of this funeral field was more diverse in its faunal remains (Fig. 3). Two sacrificed stags were laid down in that grave, one on each side of the beam of a wooden vehicle, accompanied by skeletons of several snakes and of a horse, together with bones of (a) dog,1 (b) goat or sheep, (c) bird, (d) bull, (e) pig, and (f) cat. One of the sacrificed snakes was laid down around an unbaked clay figurine of a quadruped. Remains of a pyre with calcinated human bones evincing cremation of the principal deceased were found on the vehicle, accompanied by the skeleton of a sacrificed driver. 1
The dog sacrificed in his barrow could have been either a semantic substitute for a wolf, or, more likely, an archaeological evidence of a ritual attested in Indo-European written records, during which dogs were sacrificed to a werewolf-like god-warrior. Both interpretations fit well into the metatextual sequence ‘wolf-goat-stag’.
288
GREGORY E. ARESHIAN
Fig. 2. Khanlar, barrow no. 149 (after Gummel 1992).
15. EURASIAN ARCHAELOGY AND CULTURAL SEMIOTICS
Fig. 3. Khanlar, barrow no. 150 (after Gummel 1992).
289
290
GREGORY E. ARESHIAN
The evidence from the graves of Khanlar not only supports the validity of the zoomorphic sequence of ‘wolf-goat-stag’ in connection with funerary rituals, but it also essentially expands our metatextual narrative. We now know with a high degree of certainty that, according to the metatext, the dead was moving to the afterlife through a sacred pyre in a wagon driven by stags. The mythological serpent had a prominent position in this new realm. It was signified in two different codes: as physical remains of snake skeletons and as images of snakes found on pottery from barrow no. 149 (Fig. 2.21). Birds, first being present in the Getashen composition, reappear in this ritual. A boar and a bull come forth as additional personages of the metatext. Finally, horses could become substitutes for stags. So the metatextual sequence has expanded becoming ‘snake-wolf-goat-stag’ with collateral additions of ‘bird’, ‘pig’ or ‘boar’, ‘bull’, ‘horse’. Its relationship to ritual passage from the world of the living to the realm of the dead also becomes evident, indicating the connection of this metatext with the cyclical idea of ‘birth-death-resurrection’. Yet the importance of the mythologema under consideration to the ancient inhabitants of Caucasia (including eastern Anatolia) extended far beyond rituals performed on burial grounds. As one of several examples supporting this conclusion, the sanctuary of Dvin, which collapsed in conflagration during the Urartian conquest in the first quarter of the 8th century BC, must be mentioned. Several large black burnished storage jars (pithoi), decorated with scenes depicting stags, goats, wolfs or dogs, and snakes in different combinations, have been uncovered in the shrine (Kushnareva 1977, 16–21, fig. 20; 26–9, figs. 9–11). Some of the pictorial patterns from Dvin present stags in connection with radiant solar signs, while goats and wolves are associated with circles interpretable as full moons.2 It is noteworthy that mythological ideas which had formed the core of this metatext survived the destruction of the sanctuary at Dvin by almost a millennium. Pictorial patterns reminiscent of this mythologema are found on a ritual vessel from Artashat in Armenia (Jores D. Khachatrian, personal communication)
2
It must be mentioned here that philological and archaeological data from Anatolia and Caucasia indicate that during the Middle and Late Bronze Ages the mytho-ritual image of the stag semiotically was associated with both the sun and the new moon crescent, being opposed at the same time to the full moon. The author indents to further explore these structual-mythological relations on another occasion.
15. EURASIAN ARCHAELOGY AND CULTURAL SEMIOTICS
291
and on bronze openwork belt buckles from Treli (Fig. 1.5) and another unknown location in Georgia (Abramishvili 1978, fig. 111; Burney and Lang 1971, ill. 71), all of them dated around the beginning of the Christian era. Proceeding in our reconstruction beyond the variety of codes studied by archaeology and art history, one can find the same metatextual sequence of semantic units in the written records belonging to speakers of the Anatolian group of Indo-European languages. The widely known Illuyanka myths (usually associated with Hattic mythological tradition) already occupied a central place in Hittite mythology of the Old Kingdom (ca. 1750–1500 BC). They locate the mythological snake at the bottom of the vertical axis of the universe: either in the netherworld or in the sea (Hoffner 1990, 12–3). A Luvian magic ritual performed during years of famine by the head of a clan included the sacrifice of a goat, wine and fruits, as well as a procession by eight youths symbolising warriors. The leader of the procession was wearing the skin of the sacrificed goat and howled like a wolf (Otten 1953, 14–7, the so called ‘Zarpija-Ritual’ from Kizzuwatna), instead of bleating like a goat, as one would expect. The mytho-ritual attribution of magic, werewolf-like power to Hittite warriors become evident from this text. Finally the metatextual sequence ‘snake-wolf-goatstag’ in Hittite mythology will be complete, if one consider the fact that the stag was regarded as sacred animal symbolising the sun with a special sacrificial ceremony being dedicated to them (Haas 1994, 432, 818). An amazing, almost identical parallel to the Hittite expression of this metatext is found in a Slavic myth reflected in the ancient Russian Vol’kh-Vseslav (Jakobson and Szeftel 1966, 341–53, 363–8) and Serbian Zmaj Ognjeni Vuk epics (Jakobson and Ruziœic 1966). The central hero of the epos, prince-warrior bearing the name of Vuk (‘wolf’ in the Serbian tradition) and Vol’kh (‘magician’ in the ancient Russian version) was born from the marriage of the snake from the netherworld to a human princess. According to the myth, he was born in coal, for which the ancient Russian word jazno is used. It find its cognates in Old-Indic ajínam ‘skin’, derived from ajáh ‘he-goat’, ajá ‘she≈ goat’, and in Lithuanian oz÷ys ‘goat’, ozinis ‘goatish’ (Jakobson and Szeftel 1966, 341). Thus the substitution of ‘coal’ for initial ‘goat-skin’ in the myth is probable. The myth focuses on the werewolf-like nature of the warrior-prince who becomes wolf by night, being connected to the moon. At the end VukVol’kh attacks the sun, which, in Hittite mythology, is identified with the stag. The Slavic epos not only provides the opportunity to verify the reality of the sequence ‘snake-wolf-goat-stag’ in the Indo-European mythological tradition,
292
GREGORY E. ARESHIAN
but it also places the warrior-werewolf endowed with magic capabilities on the centre-stage of our cyclical metatext of ‘birth-death-resurrection’. Being completely preserved in the philological and folkloric codes of the Anatolian and Slavic traditions, the Indo-European cultural metatext under consideration can also be revealed through the perusal of written sources concerning Roman and Vedic myths and rituals. Reflecting the ‘birth’ component of the cycle, the upbringing of the Roman eponymic twin-warriors depended on a shewolf, while within the ‘death/passage to another world’ component, Romulus’ final miraculous disappearance was associated with a goat and a swamp (Capra Mares), the swamp being a place of disappearance of Hittite Storm Gods (Areshian 1988a, 95). The goat plays the role of being messenger of gods in Vedic myths, representing the god Pushan, who had the appearance of a goat himself and was worshiped as ‘God of the Path’ securing the transfer of sacrificial offerings between the world of men and the realm of gods (Ivanov 1974, 117). Semantic shifts between individual elements of the metatext and, possibly, transformational qualities envisioned in its personages could have prompted the creation of intermediary mythological characters. Those intermediary mythic beings have blended together adjacent images in the sequence. In such a manner, Greek mythological thought had created the image of a tragélaphos ‘goat-stag’; for the Old Persian, the form **vrkazana ‘wolf-goat’ has been reconstructed; in the Classical Armenian mythological vocabulary personages of vi·apakÆa¬ ‘dragongoat’ and e¬jerwakÆa¬ ‘stag-goat’ are present, as well as the form *gaylakÆa¬ ‘wolf-goat’ has also been reconstructed (Areshian 1988b, 28). In the 6th century AD an Armenian philosopher-neoplatonic David the Invincible (1980, 29) cited the mythological image of stag-goat e¬jerwakÆa¬ in conjunction with deadresurrecting dogs aralezk’, while criticising beliefs of heathen people. Thus, such personages have strengthened internal continuity in this mytho-ritual sequence. The essential parallelism between the Anatolian and the Slavic traditions as well as the aforementioned intermediary mythological figures are methodologically sufficient to demonstrate the objective existence and the Indo-European origin of the reconstructed metatext. Yet, with a switch back to the code of art history, a third Indo-European, namely the Celtic tradition must be brought into this discussion. The ritual cauldron found at Gundestrup in 1891 (Müller 1892; Olmsted 1979) is prime in importance among the Celtic evidence. The gilded plates covering the cauldron depict several mythological compositions. One of those presents a sitting Stag-god in anthropomorphic hypostase as its central
15. EURASIAN ARCHAELOGY AND CULTURAL SEMIOTICS
293
personage (Fig. 1.7), who has been interpreted as the Gaulish Mercury/*Cernunnos by several scholars. His antlers become a tree in full bud (also interpreted as tendril growing from the antlers), which indicates early spring as the time when the mythic action was supposed to occur. With his left hand the Stag-god (Cernunnos?) strangles a horned snake while with his right hand holding out a torque to a stag. This action takes place in presence of the two other figures of our story: the wolf and the goat. Those are placed according to the metatextual sequence, i.e. the wolf is on the side of the snake, while the goat is depicted beside the stag. Art-historical evidence from other regions supports the presented reconstruction. Among pastoral nomads of Eurasia, speakers of Iranian languages have left an eloquent archaeological testimony of presence of the same metatextual sequence. Sarmatian barrow (kurgan) no. 46 at Ust Labinskaya near the Kuban river contained a headband dated to the 1st–beginning of the 2nd century AD, which was decorated with an appliqué cut out of golden foil (Zasetskaya 1975, 17, fig. 3; Guguev 1992, 117–8). Those have created a horizontal symmetrical composition (Fig. 1.6) with a sacred tree as its centre and two sequences of stags, goats and wolves approaching the tree from both sides. The tree grows from the back of a double-bodied goat and birds are depicted sitting on its branches. It is necessary to mention that the zoomorphic sequence ‘stag-goat-wolfsnake’ signifying the cycle of ‘birth-death-resurrection’ is, at the same time, homologous to the paradigmatic sequence of Proto-Indo-European mythological personages of ‘Man’-‘Twin’-‘Third’-‘Snake’ reconstructed by Lincoln (1991, 12). Thus, the stag as a symbol of celestial sacredness could be associated with ‘Man’, i.e. the priest; the goat with ‘Twin’, i.e. with self-sacrificing function of kingship; the wolf with ‘Third’, i.e. the warrior; and finally the serpent, signifying enemy people supposed to be killed or subjugated, had remained in its zoomorphic hypostase in both sequences. The interdisciplinary reconstruction of the Indo-European mytho-ritual metatext still is in an early stage of development. Further research could lead in several directions. First of all, the abundant evidence scattered through philological, linguistic, archaeological and folkloric records from different parts of Europe, the Near East, Central Asia and India should be combined together according to the rules of semantic, syntactic, and textual interdependence and continuity. Once a satisfactory reconstruction of the whole mytho-ritual metatext would be completed, it may become possible to tackle such problems as: (a) the impact of the Proto-Indo-European socio-cultural environment on the development of
294
GREGORY E. ARESHIAN
the metatext and vice versa (which is a problem of pragmatics, if expressed in terms of cultural semiotics); (b) the possible pre-Indo-European and non-IndoEuropean roots and elements of Proto-Indo-European mythology; (c) the subsequent history of Indo-European myths and rituals, which started since the breakup of the Proto-Indo-European linguistic and cultural unity. In practice the latter will mean attempts to reconstruct processes of emergence and transformation of many specific mytho-ritual systems among different ethnic and social units. Even a cursory glance at the aforementioned belt buckles from Georgia, which presumably belonged to speakers of non-Indo-European languages around the beginning of the Christian era, suggests that the spread of different parts of the initially Proto-Indo-European mytho-ritual metatext has occurred throughout non-Indo-European cultural milieu on many occasions.
Bibliography Abramishvili, R.M. 1978: Tbilisi, Arkheologicheskie pamyatniki (Tbilisi). Anthony, D. 1997: Prehistoric Migration as Social Process. In Chapman, J. and Hamerow, E. (eds.), Migrations and invasions in archaeological explanation (BAR International Series 664) (Oxford), 21–32. Areshian, G.E. 1988a: Indoevropeyskiy syujet v mifologii naseleniya mejdurech’ya Kuri i Araksa II tisyacheletiya do n.e. VDI 4 (187), 84–102. ——. 1988b: Éléments Indoeuropéens dans la mythologie du plateau Arménien et du Caucase du Sud au IIe mil. av. J.-C. In Vavrou·ek, P. and Souœek, V. (eds.), ∞ulmu, Papers on the Ancient Near East (Prague), 19–37. Burney, C. and Lang, D.M. 1971: The Peoples of the Hills: Ancient Ararat and Caucasus (London). David the Invincible (Daiit’ Anya¬t’) 1980: Erkasirut’iink’ p’ilisop’ayakank’. Critical texts combined and introduced by S.S. Arevshatian (Yerevan). De Jesus, P.S. 1980: The Development of Prehistoric Mining and Metallurgy in Anatolia, 1 (BAR International Series 74.1) (Oxford). Filip, J. 1977: Celtic Civilization and its Heritage, 2nd ed. (Wellingborough). Gimbutas, M. 1990: The Collision of Two Ideologies. In When Worlds Collide: The IndoEuropeans and the Pre-Indo-European (Linguistica Extranea 19) (Ann Arbor), 171–8. Guguev, V.K. 1992: Kobyakovskiy kurgan. VDI 4 (203), 116–29. Gummel [Hummel], Y.I. 1940: Arkheologicheskie ocherki (Baku). ——. 1992: Raskopki k yugo-zapadu ot Khanlara v 1941 g. VDI 4 (203), 5–12. Haas, V. 1994: Geschichte der hethitischen Religion (Handbuch der Orientalistik Abt. I, bd. 15) (Leiden/New York/Cologne). Hoffner, H.A. jr 1990: Hittite Myths (Writings from the Ancient World 2) (Atlanta). Ivanov, V.V. 1974: Opit istolkovaniya drevneindiyskikh ritual’nikh i mifologicheskikh terminov, obrazovannikh ot a˜va – ‘kon’’ (jertvoprinosheniye konya i derevo a˜vattha v drevney Indii). In Problemi istorii yazikov i kul’turi narodov Indii, Papers in memoriam of V. S. Vorob’evDesyatovski (Moscow). Jakobson, R. and Ruziœic, G. 1966: The Serbian Zmaj Ognjeni Vuk and the Russian Vseslav epos. In Roman Jakobson, Selected Writings 4: Slavic Epic Studies (The Hague/Paris), 369–79. Jakobson, R. and Szeftel, M. 1966: The Vseslav Epos. In Roman Jakobson, Selected Writings 4: Slavic Epic Studies (The Hague/Paris), 301–68.
15. EURASIAN ARCHAELOGY AND CULTURAL SEMIOTICS
295
K’alant’arian, A.A. and Khachatrian, J.D. 1969: Hnagitakan hazvagyut haytnagorcut’yun Sevana lçi avazanits. Lraber Hasarakan Gitut’yunneri (Journal of Social Sciences) 4, 75–7. Kushnareva, K.K. 1977: Drevneyshie pamyatniki Dvina (Yerevan). ——. 1993: Yuzhnii Kavkaz v IX-II tis. do n.e.: Etapi Kul’turnogo i sotsial’no-ekonomich-eskogo rozvitiya (St Petersburg). Lincoln, B. 1991: Death, War, and Sacrifice: Studies in Ideology and Practice (Chicago/London). Müller, S. 1892: Det store sølvkar fra Gundestrup i Jylland. Nordiske Fortidsminder 1, 35–68. Olmsted, G.S. 1979: The Gundestrup Cauldron (Collection Latomus 162) (Brussels). Otten, H. 1953: Luvische Texte in Umschrift (Deutsche Akademie der Wissenschaften zu Berlin, Institut für Orientforschung Veröffentlichung 17) (Berlin). Özgüç, T. and Akok, M. 1958: Horoztepe. Eski Tunç Devri Mezarli™i ve Ískân Yeri (Türk Tarih Kurumu Yayinlarindan V 18) (Ankara). Piotrovsky, B.B. 1949: Arkheologiya Zakavkaz’ya (Leningrad). Puhvel, J. 1987: Comparative Mythology (Baltimore/London). Sapir, E. 1994: The Collected Works of Edward Sapir 4 (Darnell, R. and Irvine, J. [eds.]) (Berlin/New York). Trigger, B.G. 1989: A History of Archaeological Thought (Cambridge/New York/Port Chester/ Melbourne/Sydney). Yakar, J. 1985: Regional and Local Schools of Metalwork in Early Bronze Age Anatolia, Part 2. Anatolian Studies 35, 25–38. Zasetskaya, I.P. 1975: Zolotie ukrasheniya gunnskoy epokhi (Leningrad).
CHAPTER SIXTEEN WRITING THE LANDSCAPE: PETROGLYPHS OF INNER MONGOLIA AND NINGXIA PROVINCE (CHINA) PAOLA DEMATTÈ
The rock art of China is probably the earliest to have been recorded in historical documents: already by the 6th century AD the geographer Li Daoyuan (472–527) had mentioned in his geographical treatise Shuijing Zhu (Commentary on the Classic of Rivers) the rock engraving he saw in various parts of the Chinese territory as he was surveying the land (Chen Zhaofu 1991, 26–36). Notwithstanding this early start, rock art of China is little known abroad and not a major field of research at home, where archaeologists began its scientific study only in the 1980s. Today there is however considerable growth in rock art research, since these remains offer alternative and suggestive evidence for the understanding of the life and activities of the non-Han populations, which have been active in and around the Chinese world from prehistoric times down to more recent centuries. Rock art sites in China are in fact quite numerous and concentrated in areas which seem to skirt the agricultural civilisations of the Yellow and Yangzi river valleys. These territories were originally inhabited by nomadic or semi-nomadic pastoralists in the north, or fringe agriculturists and horticulturists in the south. The absence of significant concentrations of petroglyphs in areas with the highest density of agricultural settlements may be related either to different lifestyles and/or to its obliteration due to intensive use of the land and overlapping of different cultural activities (especially writing). Among the Chinese rock art concentrations, those found along the northern and western frontiers (Xinjiang, Gansu, Ningxia, Inner Mongolia) play a key role in helping understanding little known aspects of the culture of Northern and Inner Asian populations living in the fluctuating Chinese northern frontier area, and at the same time shed some light on their interactions with the Chinese world.
16. PETROGLYPHS OF INNER MONGOLIA AND NINGXIA PROVINCE
297
Yinshan and Helanshan petroglyphs The rock art sites of south-central Inner Mongolia and Ningxia province are among the richest in China and, in addition to being well preserved and documented, form a coherent stylistic and cultural whole. In these areas petroglyphs are concentrated on the two main mountain ranges: the Yinshan (Yin mountains) of Inner Mongolia; and the Helanshan (Helan mountains) of Ningxia. A huge number of sites are known, many are documented, but many more are still to be discovered and studied. The present study is based on a preliminary research and survey which examined nine petroglyph sites in three general areas: the Yinshan at Urad Rear Banner (Dishui Gou and Bu’erhan Shan; Fig. 1.1–2); the Yinshan at Dengkou county (Ge’er’aobao Gou and Molehetu Gou; Fig. 1.3); and the Helanshan area (Heishimao, Shizuishan city; Daxifeng Gou, Pingluo county; Helankou, Helan county; Siyanjin, Qingtongxia city; Gujingou and Damaidi, Zhongwei county; Fig. 1.4–8). The rock art of the Yinshan and Helanshan consists mainly of engraved signs (petroglyphs) and only minimally of painted images. Petroglyphs were created by engraving, chiselling, incising, or grinding. These different techniques, often paired with stylistic variation, are good time markers, and offer one of the few possibilities for the dating the signs. Scholars have suggested that soon after the introduction of metals in the area, petroglyphs were carved with metal tools; some controlled experiment in the area, as well literature for other areas, shows however that a hard stone was sufficient to obtain similar results (Xu and Wei 1993). The themes with which the petroglyphs of the Yinshan and Helanshan seem to be most concerned relate to the nomad-pastoralist world, and they include a variety of animals (especially ungulates, with a clear preference for cervids and equines; but also felines), as well as elaborate scenes depicting hunting or shepherding activities. Stylistic and comparative archaeological analysis carried out over a number of years by Prof. Gai Shanlin (1986) indicates that petroglyphs production in Inner Mongolia spanned the time from the early Neolithic phases (ca. 9th millennium BC) to the later dynasties (19th century). Gai’s chronology identifies three major periods (in turn divided in phases) of rock art production in the Yinshan: 1) Neolithic – Early Bronze Age (9th millennium BC–11th century BC); 2) Later Bronze Age – Early Iron Age (11th century BC–ca. 3rd/6th centuries
298
PAOLA DEMATTÈ
AD); 3) Historic (6th–19th centuries AD). As for the Helanshan, Xu and Wei (1993) identify three phases which they connect with the Chinese dynastic chronology: 1) Shang – Zhou period (ca. 17th–3rd centuries BC, but mostly Late Zhou – ca. 8th–3rd centuries BC); 2) Qin – Han and Northern and Southern dynasties (3rd century BC–8th century AD); 3) Sui – Tang, Xixia, Mongol Yuan (7th–14th centuries AD). In both areas, style, iconography and associated archaeological evidence indicate that the petroglyphs are to be attributed to various prehistoric and historic northern nationalities whose presence is attested here. Chinese historical sources list among these known nomadic or semi-nomadic populations the Xiongnu, Wuhuan, Xianbei, Tujie Turks, the Xixia Tangut, and eventually the Mongols (Gai Shanlin 1986). Among these, the most productive in terms of rock art production appear to have been the Xiongnu and the Xianbei in the period roughly between the 4th century BC and the 5th century AD.
Landscape and Place Aspects of the north-western Chinese petroglyphic tradition which are relevant to its interpretation can be gleaned by observing the environment within which they are found, their spatial distribution and sign relationships, and the iconography. Petroglyph sites are scattered over large parts of the territory of Inner Mongolia and Ningxia, but they are concentrated in the two main mountain ranges: the Yinshan, running east-west like an arch above the great Yellow River bend; and the Helanshan, running north-south alongside the western side of the Yellow River (Fig. 1). These two mountain chains create a natural barrier protecting from winds and sand the fertile regions to their south and east. The Yinshan and Helanshan are therefore a geographic boundary dividing two lands which offer different kinds of subsistence, but their canyons give multiple possibilities for movements between the two sides. In the early periods (up to 5th or 6th centuries BC), given the generally low population densities, pastoralists under stress or those with semi-sedentary agricultural inclinations could probably enjoy both sides; in the later period increasing pressure from the settled fully agricultural population reduced these movements. Given their role in border setting, it is conceivable that these mountains and their canyons were chosen as petroglyph sites, not only because they provided the stone surfaces necessary for carving, but also because their canyons were
16. PETROGLYPHS OF INNER MONGOLIA AND NINGXIA PROVINCE
299
Fig. 1. A. Area map; B. Clusters of rock art sites in the Yinshan of Inner of Mongolia (1, Bu’erhan Shan; 2, Dishui Gou; 3, Ge-er’aobao Gou and Molehetu Gou) and Helanshan of Ningxia Province (4, Heishimao; 5, Daxifeng Gou; 6, Helankou; 7, Siyanjin; 8, Gujingou and Damaidi).
300
PAOLA DEMATTÈ
on the communication routes which connected the steppe with the more fertile lands further south and ultimately with China. These territories were of the greatest importance to the nomads because they allowed them to maintain a hold on the lands that provided them with resources not available further north (water, lush pastures, some grain) and with the possibility of trading with China. These rugged canyons were therefore crucial to the nomads’ creation and understanding of their own identity and cultural universe, particularly when under an increasing pressure from the expanding agricultural world. Throughout Inner Mongolia and Ningxia, in connection with petroglyph sites there are often ruins of defensive walls of various dynastic periods, beacon towers and garrison cities. These fortifications with their multiple ramifications are, even in our eyes, markers of a defined frontier: symbols of the might of the Chinese empire and of the settled way of life. At the same time it is however clear that they were also places of contact, clash and exchange between people with different mindsets. In Inner Mongolia and in Ningxia the encounter of agricultural settlers with nomadic or semi-nomadic herders has a long history: early agricultural communities are known in the upper Yellow River valley and surrounding areas since the middle to late Neolithic, even though early texts generally portray the western populations as non-agriculturists. Down to the end of the Bronze Age (8th–7th centuries BC), these areas were characterised by a mixed pastoral-agricultural economy; as such the agricultural expansionism of large settled communities cannot have had until then a major conflictive role. However around the 5th century (Late Zhou or Warring States period), with the progressive breakdown of central power in China, the major feudal states fighting for supremacy expanded their territories and their agricultural plans, settled mobile tribes, and put up defensive walls (which eventually coalesced in the creation of the socalled ‘Great Wall’). These structures were not meant just to defend the state from a variety of military threats coming from inside China, but were also there to help northern territorial expansion and control. Nomads were to be driven out of fertile territories and semi-settled people restrained in their movements so that they would not revert to nomadism. Nomadism was seen as damaging to the agricultural economy because it ‘destroyed’ valuable land and it diverted potential agricultural workers from the land. Often, the relation between settled and nomadic populations has been seen as inherently hostile and unbalanced due to the nomads’ ‘envy’ of the wealth of settled people and their lack of understanding of agriculturalism. Some stud-
16. PETROGLYPHS OF INNER MONGOLIA AND NINGXIA PROVINCE
301
ies indicate however that contacts were productive for both sides, and that the northern frontier of China, even after the Great Wall was built, was not a rigid and insurmountable barrier, but a fluctuating and mobile area, which served to both mediate between different cultures and to favour exchanges (Jagchid 1989; Di Cosmo 1999). The contacts between Inner Asian nomads and populations of the Chinese northern frontier, whether in times of war or in peace, contributed to the economic, technical and cultural development of both cultures. Nomads gained a lot from trading with the Chinese world (grain and silk particularly), but also the settled empire benefited enormously from the introduction of new ideas (for example horsemanship, the chariot and the stirrup; but also military techniques). The presence of petroglyphs in these transitional areas, in contact with various portions of defensive walls, the very apparent symbol of the Chinese empire, may indicate that while occupying the area, the nomadic populations were also intentionally creating their own signs and landmarks in the territory. These signs dotting the landscape in significant positions may have had multiple functions, ranging from religious symbols to geographical marks for travelling groups, but at the same time and perhaps most importantly they would have reiterated these peoples’ attachment to a land with which they identified, and which was clearly claimed also by the settled world. Even without adopting geographic determinism, it is pretty clear that any human population is heavily influenced by the environment in which it lives, and that space and landscape (natural or constructed) are of paramount importance for the construction of a personal or group identity (cf. Tuan Yifu 1974). Among mobile peoples, like pastoralists and hunter-gatherers, who by necessity are on the move and have a larger activity area than settled groups, the role of the natural landscape and its landmarks seems to be extremely important. This may be due to the extensive amount of time spent travelling by these populations, but also to the fact that mobile people rarely set up large architectural structures to reinforce group identity and cohesion, and that therefore natural landmarks (recognised as group icons) supply this need. Among Inner Asian nomads, most places where important activities were carried out were selected in relation to the landscape. For the Mongols the entire land (baigal ) was dotted with sacred places which were the focus of ritual and social activities. Rites related to local animistic cults involving natural spirits took place on the slopes of mountains, at rivers or on lake shores, and near prominent trees, even though traditional healing practices were more often held within the ger.
302
PAOLA DEMATTÈ
Within the Helanshan and Yinshan, which corresponded to a crucial locale in the world of the makers, the engravings are located in concentrated spots which have very specific qualities: high position or difficult access, good view, appealing landscape or association with water (springs, waterfalls, rivers). At the same time petroglyphs are carved in positions and ways that allow them to be seen from a certain distance, and on surfaces that are well-exposed to the sun (which may enhance their visibility). In addition, in some cases petroglyph sites are found in association with either small shrines (Ge’er’aobao Gou) or burials (Bu’erhan Shan). These associations indicate that the landscape played a paramount role in the selection of a place in which to carve signs and that in some cases similar choices based on landscape desirability were made also for religious sites. This fact does not automatically make all petroglyph sites into religious sites, but indicates that the two practices were often intertwined. These locations appear then to be desirable places to satisfy both the practical and the spiritual needs of a pastoral and hunting society.
Religious Meanings and Shamanism Studies of prehistoric art often support the idea that rock and cave art are manifestations linked with the spiritual activities of our ancestors. As such religion and ritual have inevitably played a large role in rock art interpretations. While this focus on the spiritual is understandable and from a certain point of view even logical (though limiting), one cannot help but wonder why prehistoric religion has to be viewed as an inherently bizarre enterprise involving mysterious practices. But so it is. In the early part of last century, the standard view was that these manifestations were connected with fertility and hunting magic and ultimately with shamanism. During the Structuralist parenthesis of the 1960s and 1970s these ideas were rejected, but more recently, shamanistic interpretations of petroglyphs have resurfaced, even though the emphasis has shifted away from fertility and hunting magic, towards an obsession for altered states of consciousness and psychotropic drugs (for a summary of interpretive developments, see Bahn and Vertut 1997). According to presently fashionable shamanistic interpretations, petroglyphs are the products of shamans who after the trance (induced by ingestion of hallucinogenic drugs) depict their visions on stone (Lewis-Williams and Dowson
16. PETROGLYPHS OF INNER MONGOLIA AND NINGXIA PROVINCE
303
1983). Even though there are a number of variants, an accepted view holds that the marks of the trance are abstract signs (in the shape of grids, coils, or stars) in proximity of or embedded in the petroglyphs. These are interpreted as phosphenes, i.e. visual by-products of hallucination created by the human brain under duress. Based on the presence of these very simple signs, most rock art is said to have been produced by people who have experienced trance and hallucinations. Other signs of shamanistic trance are reported to be the presence of halfhuman/half-animals creatures (interpreted as the shaman dressed with the skin of animal helpers) or simply of animals (interpreted as animal helpers). From this description, it is clear that given the low threshold for the identification of shamanistic iconography, a large number of petroglyphs can be interpreted as products of shamanistic activities, and that any fanciful story can be attached to these images. This interpretation also short-changes the discussion on the important topic of ‘shamanism’, by reducing this complex phenomenon to a catch-all term referring to the religion of the good savage. The shamanistic interpretive framework is today subject to an increasing critique, which has exposed its theoretical and methodological weaknesses, the lack of scientific data for the neurological claims, as well as its dogmatic approach (Francfort and Hamayon 2001). So even though Inner Mongolia and Ningxia are in a legitimate cultural and geographic context for shamanism, it appears that a shamanistic interpretation of this sort is not useful for understanding the variety of processes which induced these people to create rock art. In fact, even though shamanistic practice may have been part of the religious system of the populations which made the rock carvings, and it is feasible that some carvings were made in a context which had religious implications, this does not mean that the images bore any direct relationship to the practice or representation of ‘shamanism’, or that the ‘shamans’ were the producers. As a matter of fact, in the few cases in which a religious meaning is clearly present, the religion involved is Buddhism. On the other end, it is very likely that alongside or beyond a religious content, there existed other (perhaps somewhat hidden) meanings which have more potential for an understanding of the phenomenon.
Iconography and Communication An analysis of representational themes and internal spatial organisation of signs can help understand what may have been the meaning of these images and how
304
PAOLA DEMATTÈ
they worked within the societies that produced them. The motives represented in the carvings mostly include single animals or hunting and shepherding scenes, but also (though in more limited number) face or mask motives, abstract symbols, hand- or foot-prints and animal tracks (Fig. 2). This focus on hunting and shepherding (activities clearly central to the life, landscape and subsistence of these peoples) indicates that petroglyph-makers paid particular attention to daily life events and mundane activities, even though these may have been representations of events played out in a ritual context. Indeed, in terms of structure and organisation, some of the compositions are arranged as if they had a narrative intention sometimes with an the effect so compelling that the pictures appear as efforts at either communication or long term recording of significant ritual/historical events. This does not mean that rock art was simply a ‘portrayal’ of life on the steppe or a passive recording of myths and legends, but that the subjects and objects of hunting and shepherding, in their different forms and styles, were the symbols by which these people defined themselves. Narrative recording of this kind is visible in a rock panel at Siyanjin. Here, animals and people are composed to create a scene: a camel rider, trotting goats, and a hunter wounding a large beast on whose inside are two goats (Fig. 2b). We cannot be sure (since it is possible that the images were not all carved at the same time as a scene), but perhaps this panel may record a mythical or actual event which involved the killing of a wolf that ate the goats of the herders. The event in any case was of great significance to the makers and their social group. The iconography of the face or mask, which is concentrated in certain areas only (Helankou in the Helanshan and at Molehetu Gou in the Yinshan), could indicate a stronger concern with ritual activities and/or figures associated with worship (Fig. 2c). Still, it seems hard to prove that these faces or masks could be representations of shamans or other religious specialists. Perhaps some of these images may represent deities or spirits associated with nature cults, but given the importance in local religion of ancestor-worship they are more likely to be connected with ancestral representations. The use of the human form or face to identify the ancestor is attested by archaeological and historical evidence as being a long lasting tradition among the pastoral nomads of the Chinese northern frontier. For example the ancient Xiongnu were known for their wooden figurines representing their ancestors, the Tujue Turks for their anthropomorphic tomb stones, and the Mongolians for the ongghot figures (Heissig 1970; Ishjamts 1992). This evidence may then indicate that the faces or masks, while
16. PETROGLYPHS OF INNER MONGOLIA AND NINGXIA PROVINCE
305
Fig. 2. (a) Bu’erhan Shan petroglyphs (Yinshan); (b) Siyanjin rock panel (Helanshan); (c) mask motifs at Helankou (Helanshan).
306
PAOLA DEMATTÈ
different from the more widespread images of animals and hunting scenes, were always concerned with the representation elements that helped define the identity and legacy of the makers. While land, place and iconographic content had a paramount symbolic importance in the process of creation of these images, given the emphasis on sign, record and communication inherent in the petroglyphs, it cannot be overlooked that these pictures functioned also in roles which in literate societies are taken up by writing. Since the producers were mostly non-literate populations, this is not surprising. In times when writing did not exist or literacy was not widespread, the role of pictures and symbols in communicating and recording ritual or social matters was always immense. It is symptomatic that in the later periods, when writing became more widespread among the nomads (post 6th/7th century), writing and engravings mixed, and eventually petroglyph production disappeared and writing took over the same surfaces. At Helankou, inscriptions in Xixia script appended next to earlier petroglyphs describe them as ‘the parents of writing’ or ‘the writing of the spirits of writing’, thus making clear the close connection perceived by literate people between the two sign systems (Fig. 3). At other sites, longer and much later inscriptions in either Chinese or Mongol seem to ignore altogether the petroglyphs and simply compete for space on the same piece of stone. It is also significant to notice that while some of these later inscriptions have a religious content (essentially Buddhist), others (particularly those in Chinese) seem to be more concerned with the dissemination of political or moral contents. While this focus towards the non-religious may be a later development imputable to Chinese imperial influence, it is certainly an indication that not all engravings have to have a religious content or function. If the function of writing and that of petroglyphs may have something in common, there is also a deeper and visual connection between the two, particularly if the single petroglyphs are compared with written signs of the earliest (iconographic and symbolic) types (generally known as pictographs). The stylised and shortened manner used to create early written signs is derivative of many conventions used in rock art. Particularly interesting are figures of horned animals which both in early writing and in rock engravings, while shortened in the body, have very detailed antlers or horns, the very elements which afford sign recognition (Fig. 2a–b). Similarly, in both media, complex structures are broken down and re-assembled in component elements shown in their most characteristic view. For example,
16. PETROGLYPHS OF INNER MONGOLIA AND NINGXIA PROVINCE
307
Fig. 3. Face or mask at Helankou with Xixia inscription.
at Damaidi a chariot pulled by two horses is depicted with this distorted symbolic perspective: the horses are shown sideways, the chariot box is seen from above, while the wheels are depicted as seen frontally; exactly what we find in ancient Chinese writing (Fig. 4). Clearly, there are also remarkable differences between writing and petroglyphs, the most obvious being the arrangement of various signs with respect to each other. A piece of writing, no matter how simple, arranges its signs following a sequential logic which is generally (but not necessarily) linguistic; differently, a pictorial representation places its signs in accordance to a spatial logic, and hierarchies between signs are established by both their size and position. Still, the systematic way of simplifying images transforming them into easily recognisable symbols and the narrative intent, are indications that petroglyphs had potential and were probably used to record and communicate information, perhaps to later generations, neighbouring groups, or even encroaching enemies. The dissemination of such symbols in space makes the larger landscape a sort of pre-modern bulletin board on which every groups gets its message (of ownership, friendship, or antagonism) and the ‘owners’ get specific information or instruction. This sort of communication is also not unlike that of literate cultures (ancient or modern) which have covered their land with writings: edicts, historical inscriptions or simply ‘graffiti’ marking gang territories. Among them, the Chinese have excelled at it, and have succeeded in marking crucial spots
308
PAOLA DEMATTÈ
Fig. 4. Damaidi cart (Helanshan).
of their landscape with texts (and associated images) relating to their cultural identity, from historical records, to religious texts, to political propaganda. The northern neighbours of the Chinese empire did probably the same thing, although in a different way. This means that the uncouth barbarians described in Chinese sources were engaged in similar claiming activities even though they used a semiotic system different from fully developed writing. The carving of images in the landscape, far from being a mysterious religious activity, was more likely an effort to inscribe and take possession of the home territory, and thus define the universe of the social group.
Bibliography Bahn, P. and Vertut, J. 1997: Journey through the Ice Age (Berkeley/Los Angeles). Chen Zhaofu 1991: Zhongguo Yanhua Faxian Shi (History of the Discovery of Chinese Rock Art) (Shanghai). Di Cosmo, N. 1999: The Northern Frontier. In Loewe, M. and Shaughnessy, E.L. (eds.), The Cambridge History of Ancient China: From the Origins of Civilization to 221 BC (Cambridge), 885–966.
16. PETROGLYPHS OF INNER MONGOLIA AND NINGXIA PROVINCE
309
Francfort, H.P. and Hamayon, R.N. (eds.) 2001: The Concept of Shamanism: Uses and Abuses (Budapest). Gai Shanlin 1986: Yinshan Yanhua (Petroglyphs in the Yinshan Mountains) (Beijing). Heissig, W. 1970: The Religions of Mongolia (Berkeley/Los Angeles). Ishjamts, N. 1992: Nomads in Eastern Central Asia. In Masson, V.M. and Dani, A.H. (eds.), History of Civilizations of Central Asia vol. 2 (Paris), 151–69. Jagchid, S., 1989: Peace, War and Trade along the Great Wall (Bloomington). Lewis-Williams, D. and Dowson, T.A. 1983: Signs of All Times. Current Anthropology 29.2, 201–45. Tuan Yi-fu 1974: Topophilia (Englewood Cliffs, NJ). Xu Cheng and Wei Zhong 1993: Helanshan Yanhua (The Rock Art of the Helan Mountains) (Beijing).
CHAPTER SEVENTEEN REFINING THE DEFINITION OF TECHNOLOGY IN THE SOUTHERN ZONE OF THE CIRCUMPONTIC METALLURGICAL PROVINCE: COPPER ALLOYS IN ARMENIA DURING THE EARLY AND MIDDLE BRONZE AGE LAURA A. TEDESCO
Introduction For archaeologists, questions about the origins and the development of metallurgy are of longstanding interest, especially to those of us investigating the various manifestations of emergent complex societies in south-west Asia and Europe. The early writings of V.G. Childe set the stage for current discussions that link metallurgy and social complexity (Childe 1944; 1957; see also Wailes 1996). Archaeologists studying the Bronze Age frequently interpret the presence of metals, especially copper-alloy’s incipient technology, as an indicator of technological and economic complexity and for describing apparent levels of social differentiation. Metal artefacts and debris from metalworking, such as crucibles, slag and moulds, speak to broader questions about the presence of hierarchical and heterarchical social structures, craft specialisation, and social and economic exchange systems. Characterising the mechanics of metal production by identifying the composition and the formation of objects can be a seemingly straightforward analytical task, yet situating the artefacts and the technology into a larger framework of social and economic relationships remains a difficult anthropological question, as is the case for Early and Middle Bronze Age Transcaucasia. I am currently engaged in dissertation research which focuses on the metallurgy of early Transcaucasian cultures from the beginning of the 3rd to the mid2nd millennium BC. The aims of this research are to characterise the technology of copper-base metal making during the Early Bronze and into the beginning of the Middle Bronze Age and to define where metallurgy fits into the region’s larger economic and social context. The research includes several levels of analysis involving the reconstruction of compositional and manufacturing char-
17. COPPER ALLOYS IN ARMENIA
311
acteristics of artefacts in order to document stylistic and technological conventions. These results will contribute to an understanding of the technology of copper-base metal making, the social context of metal production, the role of metallurgy in the expansion of trade networks, and the degree to which metal technology was a specialised craft in Early Bronze Age Transcaucasia. The complete results of these detailed analyses will be forthcoming. In this paper, I provide the background to the project and its placement in a larger archaeological and theoretical framework.
Bronze Age Transcaucasian Chronology and Metalworking Technology Transcaucasia, the territory of the southern Caucasus region, encompasses the modern republics of Georgia, Armenia and Azerbaijan, north-west Iran and eastern Turkey. The region is mountainous with fertile plains and valleys fed by the Kura and Araxes rivers whose headwaters are in eastern Turkey. The rivers join course in eastern Azerbaijan before emptying into the Caspian Sea. The region is ecologically rich and diverse with abundant natural resources, namely polymetallic ores. The earliest evidence for metalworking appears during the Aneolithic, in the final stages of the Shulaveri Shomu Tepe Culture dated to the second half of the 6th millennium BC (Kiguradze 1976; Chataigner 1995). The first experimentations with metallurgy in the region could have been satisfied by the presence of copper on the surface of ore sources. While the methods of ore extraction during the Aneolithic period are unknown, it is likely that the earliest copper-base artefacts needed only to be hammered from nuggets of native copper. The inventory of these earliest copper artefacts is small – about a dozen artefacts; and the finds are isolated. The objects (beads, pins and awls), are similar to those known from Çayönü in south-eastern Anatolia and Yarim Tepe in Iraq. For example, a cylindrical bead discovered at Gargalar Tepesi in Azerbaijan along the central course of the Kura river is very similar to the folded and twisted plate metal beads excavated by Munchaev in the slightly earlier Hassuna levels at Yarim Tepe, dated to the first half of 6th millennium BC (Porada et al. 1992). The chronological transition from the Aneolithic to the Early Bronze Age occurred in the middle of the 4th millennium BC. There are literally hundreds of sites known from across Transcaucasia, except in westernmost Georgia (Sagona
312
LAURA A. TEDESCO
1984). Within the material culture repertoire there are some intra-regional variations but it is generally uniform with distinctive red-black burnished pottery, portable hearth stands, standardised wattle and daub architecture, and simple metal tools and ornaments. Evidence of metalworking in the form of bits of slag, crucible fragments, ladles and traces of metal on the walls of vessels are known from scores of sites (Kavtaradze 1999). Three larger and better-known Early Bronze Age metalworking sites are Amiranis Gora in south-west Georgia, Kül Tepe in Nakhichevan, and Shengavit on the Ararat Plain near Yerevan, Armenia. At Amiranis Gora the excavators have identified a metallurgical workshop from tuyère fragments, a stone-constructed arch-shaped kiln, and an associated large ceramic vessel that contained heavily ground charcoal necessary for producing very high temperatures (Chubanishvili 1971). Carbon samples for dating were taken from the ground charcoal found in the vessel, and this evidence provides clear indication that smelting was conducted as early as the mid-4th millennium BC (Kavtaradze 1999, 74). At Kül Tepe and Shengavit ladles, slag, crucible fragments and casting moulds for axes have been found (Kushnareva and Chubinishvili 1970). The array of associated smelting artefacts from these sites and other contemporary sites such as Khizanaant Gora (Georgia) and Garni (Armenia) strongly supports the idea that the technology of copper-alloy metal production was already reflecting a specialised understanding of complex extractive and smelting processes. The kiln at Amiranis Gora and another of similar construction from Baba Dervish in Azerbaijan are isolated finds; however, the frequency of mould fragments, ladles and crucibles are sufficient evidence that the process of smelting and a developing command of the properties of ores and alloying had been incorporated within the technological tradition across the region by the beginning of the Early Bronze Age. Among the copper-alloyed artefacts themselves excavated from many mid-4th to early 3rd millennium BC sites are daggers, points, spiral rings and fibulae. These have been reported from both burial and settlement contexts (Kushnareva and Markovin 1994). Compositional analyses consistently show that the metal inventory of the Early Bronze Age was made of arsenical copper. Some of the ore sources in the region are known to naturally contain arsenic, but according to Kavtaradze (1999) arsenic contents were high enough at up to 8% ‘that ancient metallurgists in Transcaucasia were intentionally choosing copper with high or low levels of arsenic, [or even directly adding arsenic to the mix, depending upon the technique of manufacture] and functional destination of the artefacts’ (Kavtaradze 1999, 76).
17. COPPER ALLOYS IN ARMENIA
313
A widespread view is that an expansion of Kura-Araxes or Early Transcaucasian culture occurred in a series of migrations around the beginning of the 3rd millennium BC, and that bearers of Transcaucasian cultural traditions, particularly the characteristic red-black burnished pottery were spreading widely. It appears that Transcaucasian metallic ores and metallurgical traditions were being exported, also. Evgeni Chernykh (1992) proposes that the control over ore sources and the specialised technology required in extractive and smelting processes may have created a kind of monopoly driving long-distance trade in radiating directions from Transcaucasia: to the north through the greater Caucasus reaching as far as the Dneiper and Volga rivers, and towards the south and south-west reaching Anatolia, central Iran and Palestine (Chernykh 1992, 54–5). While archaeologists have yet to fully interpret the social and economic relationships between Transcaucasia and its surrounding territories, the recent discovery of a ‘royal’ tomb at Arslantepe in the Malatya Plain of central Anatolia reveals a far more complex picture than previously understood (Frangipane et al. 2002). Arslantepe is recognised as a regional economic and political centre with ties to northern and southern Mesopotamia and to Transcaucasia (Frangipane 1997). Discovered in 1996, the ‘royal’ tomb has been dated to 3000–2800 BC and contains compelling evidence of the significant influence by bearers of Early Transcaucasian culture. Within the elaborate tomb were found numerous red-black burnished vessels; however, the rich assortment of copper-base artefacts affords the most telling evidence of Transcaucasian influence. Among the artefacts, namely a bronze spearhead with silver inlay, a diadem, spiral rings and arm bands made from silver and silver-copper are close parallels in composition and typology to objects found in central Transcaucasia (Frangipane et al. 2002; Kushnareva and Chubinshvili 1970). In addition to the finds in the ‘royal’ tomb, a hoard of nine daggers, twelve spears and a quadruple spiral pendant from Late Chalcolithic/Early Bronze Arslantepe level VIA have been identified on typological and elemental-compositional criteria to suggest a Transcaucasian origin in material and/or manufacture. The artefacts, containing more than 4% arsenic and no nickel, are not made from the more local copper deposits of the Taurus mountains (Palmieri et al. 1993). The quadruple spiral ornament and the leaf-shaped spearheads show striking typological similarities to contemporary artefacts known from Transcaucasia (Palmieri 1981, 109). The finds at Arslantepe, in addition to the widespread appearance of red-black burnished pottery (also known as Khirbet Kerak ware) as far as Palestine, demonstrate a far-reaching
314
LAURA A. TEDESCO
influence by bearers of Early Transcaucasian culture across a vast region during the Early Bronze Age. The command of metallurgical technology, the abundance of particular ores that existed in the Caucasus mountains, along with the movements of semi-nomadic pastoralists may have been the central components stimulating these far-reaching influences. At the end of the Early Bronze with the transition to the Middle Bronze Age at about 2200–2000 BC, the archaeological and metallurgical landscape changes. The settlement pattern in Transcaucasia shifts where many of the Early Bronze sites are abandoned; elaborate and rich kurgan burials are constructed and metallurgy becomes more complex. Tin bronzes are far more prevalent in the Middle Bronze Age nearly replacing arsenical bronzes altogether, not only in Transcaucasia but also across the greater Near East and the northern Caucasus. The number and types of objects (ornaments, weapons, vessels) increases significantly. The artefact forms and composition are a result of increasingly complicated and highly specialised alloying and production. Economic changes are reflected in the expansion and intensification of long-distance trade networks, especially of tin, and in the abandonment and restructuring of many settlements across the entire greater Near East (Dercksen 1996; Kohl 1992, 125–6). While the archaeological record reflects an apparent break in continuity and there appears to be a restructuring of economic relationships across the region, these new metallurgical developments, however, represent an extension of the previous copper-arsenic tradition with a refined, detailed and increasingly specialised knowledge of ores, the properties of metal, and the social and economic value ascribed to the raw materials, particularly tin.
Theoretical Considerations The economic and social relationships between Early Bronze Age Transcaucasia and its surrounding territories are not well understood. World Systems theory has been employed by archaeologists to explain interregional interaction and the development of social complexity (Kohl 1992), yet there are complications in defining the dynamics driving social interactions among cores and peripheries (Stein 1999). A question aptly raised by Phil Kohl (1998) is how do we define the integration and interconnection of larger economic systems, of which surely included copper-alloy metals and metal-making, if what characterises this technological system in a so-called periphery has not yet been adequately defined?
17. COPPER ALLOYS IN ARMENIA
315
Traditional models of craft specialisation are well suited for components of this research (Brumfiel and Earle 1987; Costin 1991; Cobb 1996), though such models have limitations. These studies are well equipped to define how production is organised, yet a craft specialisation model tends to isolate production from other aspects of social complexity (Tringham 1996). Evidence of metal production typically indicates the presence of some form of complicated exchange relationships, what Costin (1991, 4) refers to as ‘extra-household exchange relationships’, and by nature metal technology must involve some degree of specialisation since exceptional practical skills are required, raw materials are not evenly distributed across regions, and there are more consumers than producers. The control of craft production is typically identified as a precursor to more complex political and economic structures, yet Joyce White and Vince Pigott (1996) have demonstrated in Bronze Age Thailand that metal production, while a specialised craft, is not necessarily predicated on a social organisation comprised of full-time specialists working for a controlling, elite social group. Craft specialisation models tend to view the organisation of production operating within the somewhat rigid parameters of political and economic hierarchical structures. However, as White and Pigott (1996, 151)’s Thailand example demonstrates, the community-based mode of copper production in this social context functions within a broader system that is heterarchical and where the control and centralisation of power and production does not follow rigid, hierarchical controls. A similar model to the one proposed by White and Pigott may be applicable to Early Bronze Age Transcaucasia, where there is no evidence of political and economic centralisation and where the archaeological record shows no apparent levels of social differentiation. Typically Early Bronze sites in Transcaucasia are on average 1–2 ha and houses are roughly of equal size and agglutinative. Metal production appears to occur mainly on a small scale, within hamlets, or perhaps even within households. Where potential indications of social differentiation are found in burials, a few documented Transcaucasian sites have individual interments (Elar and Avtoagregat, for example) with few grave goods. The majority of Early Bronze Age sites with excavated cemeteries, however, are collective, (Horom, Keti, Talin, Lanjik, Karchaghbiur, Tsakhkalanj, Berkhaber). These collective burials typically were re-used for centuries as identified by ceramic typologies; they average three to twelve bodies and generally include few grave goods including mostly pottery and occasionally metal artefacts (Badalyan and Avetisyan, personal communication).
316
LAURA A. TEDESCO
Where craft specialisation models may be limited in adequately explaining the relationship of metallurgical technology to other aspects of social and economic complexity, the burgeoning literature on the anthropology of technology attempts to identify a broader socio-technical system comprised of, among other attributes, techniques, technological knowledge and ‘the social coordination of labor’ (Pfaffenberger 1992, 497; Lechtman and Steinberg 1979; Lemonnier 1986; 1992; Ingold 1999). The anthropology of technology is multifaceted but aims to connect different aspects of the relationship of technology and society where technology is not viewed as ‘a means to an end’ (Hegmon 1998, 266), but rather as a form of material expression. One component of an anthropology of technology looks for technological styles where technology itself is seen to play a role in perpetuating and changing social relations. For example, Lechtman (1993) considers how the technology of metalwork is used to mark social distinctions; Dobres and Hoffman (1994) position the working interchange of technology and society as an element of social agency by closely examining small-scale ‘social processes’ within the manufacture of categories of material culture. The concepts of craft specialisation and technological style employed together provide a useful means to begin to define the technology of metalworking in Early Bronze Age Transcaucasia. Assessments of Transcaucasian technological styles, in the sense of that defined by Lechtman (1993) in reference to Andean artisans, are useful in identifying patterns of metallurgical production within Early Transcaucasian culture and assessing the social significance of technological activities. In other contexts, archaeologists have demonstrated that techniques of production bear cultural signatures that reflect gestures repeated in the act of producing an object (Hosler 1994; Lemonnier 1992). They constitute views embedded in the technical practices of artisans about the right and wrong way of doing things, aesthetic preferences for materials, and may even reflect more deeply held world views. An integrally related factor is the social context of production. A critical element in the development of complex societies is the control of resources in the form of manufactured goods and networks of exchange. This research attempts to assess the components relating the social contexts in which production, labour and distribution of copper-alloy metallurgy took place in Transcaucasia.
17. COPPER ALLOYS IN ARMENIA
317
Previous Research Much of the previous research on Transcaucasian metals has concentrated on the identification of the elemental composition of objects in order to sort out the shared characteristics of raw material sources and of production as a means of delineating centres of production and distribution across the region (Abesadze et al. 1958; Abesadze 1969; Avilova et al. 1999; Chernykh 1992; Gevorkian 1980; Selimkhanov 1962; 1978). The work of Chernykh (1992) is significant for integrating compositional data of metals across a broad Eurasian expanse. The results have been interpreted into what Chernykh identifies as metallurgical provinces. These include large, interconnected geographical regions linked by several metallurgically defined factors such as shared access to raw materials, similar technologies identified among centres of production, and shared typological characteristics among corpora of ornaments, tools and weapons. According to Chernykh, within metallurgical provinces exist metallurgical foci. These are centres of production and distribution, interrelating with other metallurgical foci and driving metallurgical traditions of the entire province. Transcaucasia is a metallurgical focus within the southern zone of the Circumpontic Metallurgical Province, which extends from the Adriatic Sea in the west across Eurasia to the Urals in the east and north from the upper Volga basin to the Levant and Mesopotamia in the south. Chernykh’s syntheses clearly emphasise the importance of metals on larger economic and social developments during the Bronze Age. Chronological alignments, however, across the Circumpontic province and within Transcaucasia itself, are problematic. With the exception of Chernykh et al. (2000)’s recently published radiocarbon chronologies and Kavtaradze (1999)’s extremely thorough review and of the central Transcaucasian chronology, radiocarbon dates for the Early Bronze Age do not consistently constitute calibrated chronological series, nor have they been reliably established among other series of dates from adjacent areas such as central and eastern Anatolia, northern Mesopotamia and northwest Iran (di Nocera 2000). The perspective of my research assumes that Chernykh’s general model is correct, and by building on his invaluable work and others, this project’s objective is to provide a detailed description of Transcaucasia’s metal technology, the techniques used to make objects, and the social and economic organisation around metallurgical production in the region. The research will provide detailed descriptions of the artefacts themselves and their archaeological contexts
318
LAURA A. TEDESCO
and information about the techniques of manufacture and composition. These combined lines of inquiry will be used to assess how copper-base metallurgy was organised in a way that has not previously been attempted.
Current Research Programme Lab-based analyses of 48 copper-alloy artefacts recovered from excavations across the Republic of Armenia have been completed. These analyses, including metallography, electron microprobe and x-ray fluorescence, represent a central portion of my current research programme. The artefacts were selected for this study based on excavation from well-provenanced and well-documented contexts. I also relied heavily on the generosity of Armenian archaeologists to make available their unpublished materials and the artefacts themselves for sampling. The artefacts under analysis represent a diverse group: there are weapons (daggers and spears), personal ornaments (spiral rings, a variety of toggle pins, bosses) and a cauldron (from which four samples at the handle, body, rim and a rivet joint were collected and analysed). All of these analyses – metallography, XRF, and microprobe – have been carried out under the direction of Prof. Vincent Pigott and Prof. Thilo Rehren at the Institute of Archaeology, University College London.1 Metallography works through a series of steps of increasingly fine polishing and then chemically etching samples to reveal the metal’s microstructure. The granular microstructure of the metal portrays the history of manufacture of the objects, such as whether heat was applied, or whether hammering and/or forging were part of how an object was formed. X-Ray Fluorescence yields data on elemental composition. Principally we have tested for elements characteristic of Caucasian polymetallic ores like copper, arsenic, tin, nickel, antimony, iron and lead.2 Electron microprobe analysis works on the same principal as
1 Dr Peter Northover at Oxford University, Department of Materials assisted greatly with additional metallographic analyses and with interpreting the samples’ microstructures. Additional analyses and discussions about the samples were also carried out at Lehigh University with Prof. Michael Notis and Dr Aaron Shugar. 2 There are numerous problems with linking archaeological metal objects to their ore sources. The analysis of the ores in Transcaucasia remains problematic since it is often difficult to identify which ores were exploited in which phases of antiquity, and samples collected from copper deposits can never be totally representative of the entire mine complex. Recycling of metal arte-
17. COPPER ALLOYS IN ARMENIA
319
XRF though its detection limits are far greater. It can detect amounts as low as 100 parts per million; it is also capable of rendering extremely precise readings on inclusions entrapped in ancient metal. Microprobe, through the identification of small-scale inclusions, can suggest sources of raw material and complexity in alloying techniques. Overall the results of the compositional analysis show a pattern that we would more or less expect for Transcaucasia in the 3rd to mid-2nd millennium BC.3 The artefacts from the 3rd millennium are arsenical bronzes4 and those artefacts excavated from contexts dated to the first part of the Middle Bronze (end of the 3rd to the middle of the 2nd millennium BC) are more consistently tin-bronze. These data are consistent with the vast body of compositional analytical data that already exists for Bronze Age metals in the Caucasus. My aim though is to achieve a finer focus when combining compositional data with the metallographic and microprobe results. These combined data yield information on multiple levels, allowing for the possibility to describe the technology more thoroughly. If patterns can be identified between manufacturing techniques and elemental compositions, in combination with the archaeological information, then an interpretation can be made about how production was organised, the degree to which it was specialised, and how the exchange networks within the region may have been structured. It also may be possible to make a technological link between production and the intentionality of alloying.5 The 48 copper-alloy artefacts under analyses for this research project represent only a small fraction of the excavated material across Transcaucasia. But by going
facts in antiquity obscures accurate interpretations of elemental compositions, manufacturing techniques and original raw material sources. 3 Final and detailed results of all analyses will be forthcoming. 4 ‘Copper alloys with an arsenic content ranging from 3 to 10% do not seem possible without the direct addition of arsenic minerals. This consideration is supported by both smelting experiments and the analysis of metal objects as well as the copper deposits . . .’ (Palmieri et al. 1993, 574). 5 Palmieri, Sertok and Chernykh (1993) have investigated the ores in eastern Anatolia in an attempt to characterise the nature and composition of various sources of raw materials for copper-base metal production, but the extant data is insufficient to determine the occurrence of copper-arsenic minerals with certainty. The presence of mines in Transcaucasia is known and documented but the arsenic content in them, and the characteristics of alloys when the ores are smelted, remains unknown. Palmieri et al. acknowledge that the technological link needed to understand the production and intentionality of such a high number of Cu-As alloys is missing in current research programmes. (Palmieri et al. 1993, 595).
320
LAURA A. TEDESCO
from more general and broad regional patterns of production and organisation, like those Chernykh identifies within the Circumpontic Metallurgical Province, to multiple levels of finely-honed regional analyses driven by anthropological questions, the research programme described here seeks to contribute to a more complete description of copper metallurgy in Early Bronze Age Transcaucasia where the technology and its social and economic contexts are meaningfully joined.
Bibliography Abesadze, T. 1969: The Production of Metal in Transcaucasia in the 3rd Millennium BC (Tbilisi). Abesadze, T., Bakhtadze, R., Dvali, T. and Dzhaparidze, O. 1958: K istorii medno-bronzovoi metallurgii v Gruzii (Tbilisi). Avilova, L.I., Antonova, Y.V. and Teneyshvili, T.O. 1999: Metallurgical Production in the Southern Zone of the Circumpontic Metallurgical Province during the Early Bronze Age. RA, 51–66. Brumfiel, E. and Earle, T. 1987: Specialization, Exchange and Complex Societies: An Introduction. In Brumfiel, E. and Earle, T. (eds.), Specialization, Exchange, and Complex Societies (Cambridge), 1–9. Chataigner, C. 1995: La Transcaucasie au Neolithique et au Chalcolithique (BAR International Series 624) (Oxford). Chernykh, E.N. 1992: Ancient Metallurgy in the USSR: The Early Metal Age (Cambridge/New York). Chernykh, E.N., Avilova, I.L. and Orlovskaya, L.B. 2000: Metallurgicheskie provintsii i radiouglerodnaya khronologiya/Metallurgical Provinces and Radiocarbon Chronology (Moscow) (in Russian and English). Childe, V.G. 1944: Archaeological Ages as Technological Stages. Journal of The Royal Anthropological Institute of Great Britain and Ireland 74, 7–24. ——. 1957: The Dawn of European Civilization (London). Chubinishvili, T.N. 1971: K drevnii istorii yuznego Kavkaza (Tbilisi.) Cobb, C.R. 1996: Specialization, Exchange, and Power in Small-Scale Societies and Chiefdoms. Research in Economic Anthropology 17, 251–94. Costin, L.C. 1991: Craft Specialization: Issues in Defining, Documenting, and Explaining the Organization of Production. In Schiffer, M.B. (ed.), Archaeological Method and Theory vol. 3 (Tucson), 1–56. Dercksen, J.G. 1996: The Old Assyrian Copper Trade in Anatolia (Istanbul). di Nocera, G.-M. 2000: Radiocarbon Datings from Arslantepe and Norsuntepe. In Marro, C. and Hauptmann, H. (eds.), Chronologies des Pays du Caucase et de l’Euphrate aux IV–III Millenaires (Paris) 73–93. Dobres, M.-A. and Hoffman, C. 1994: Social Agency and the Dynamics of Prehistoric Technology. Journal of Archaeological Method and Theory 1.3, 211–58. Frangipane, M. 1997: A Fourth Millennium Temple/Palace Complex at Arslantepe/Malatya. North-South Relations and the Formation of Early State Societies in the Northern Regions of Greater Mesopotamia. Paléorient 23.1, 45–73. Frangipane, M., di Nocera, G.-M., Hauptmann, A., Morbidelli, P., Palmieri, A., Sadori, L., Schultz, M. and Schmidt-Schultz, T. 2002: New Symbols of a New Power in a ‘Royal’ Tomb from 3000 BC Arslantepe, Malatya (Turkey). Paléorient 27.2, 105–39.
17. COPPER ALLOYS IN ARMENIA
321
Gevorkian, A. 1980: Iz istorii drevneishei metallurgii Armyanskogo Nagorya (Yerevan). Hegmon, M. 1998: Technology, Style and Social Practices: Archaeological Approaches. In Stark, M. (ed.), The Archaeology of Social Boundaries (Washington, DC), 264–279. Hosler, D. 1994: The Sounds and Color of Power: The Sacred Metallurgical Technology of Ancient West Mexico (Cambridge, Mass.). Ingold, T. 1999: Forward. In Dobres, M.-A. and Hoffman, C. (eds.), The Social Dynamics of Technology (Washington, DC), vii–xi. Kavtaradze, G.L. 1999: The Importance of Metallurgical Data for the Formation of a Central Transcaucasian Chronology. In Hauptmann, A., Pernicka, E., Rehren, T. and Yalcin, U. (eds.), The Beginnings of Metallurgy (Bochum), 67–101. Kiguradze, T. 1976: Agmosavlet’ Amierkavkasiis Adresamicat’mok’medo Kulturis Periodizac’ia (Tbilisi) (in Georgian with Russian summary). Kohl, P.L. 1992: The Transcaucasian ‘Periphery’ in the Bronze Age. In Schortman, E. and Urban, P. (eds.), Resources, Power, and Interregional Interaction (New York) 117–37. ——. 1998: Integrated Interaction at the Beginning of the Bronze Age: New Evidence from the Northeastern Caucasus and the Advent of Tin-Bronzes in the 3rd Millennium BC. Paper presented at the American Anthropological Association Meetings (Philadelphia). Kushnareva, K.K. and Chubanishvili, T.N. 1970: Drevnii Kulturii Yuzhnogo Kavkaza (Leningrad). Kushnareva, K.K. and Markovin, V.I. 1994: Epokha Bronzy Kavkaza i Srednei Azii: Ranniaia i Sredniaia Bronza Kavkaza (Moscow). Lechtman, H. 1993: Technologies of Power: The Andean Case. In Henderson, J.S. and Netherly, P.J. (eds.), Configurations of Power: Holistic Anthropology in Theory and Practice (Ithaca), 244–80. Lechtman, H. and Steinberg, A. 1979: The History of Technology: An Anthropological Point of View. In Bugliarello, G. and Doner, D.B. (eds.), The History and Philosophy of Technology (Champaign, IL), 136–60. Lemonnier, P. 1986: The Study of Material Culture Today: Toward and Anthropology of Technical Systems. Journal of Anthropological Archaeology 5, 147–86. —— (ed.) 1992: Elements for an Anthropology of Technology (Ann Arbor). Pfaffenberger, B. 1992: Social Anthropology of Technology. Annual Review of Anthropology 21, 491–516. Palmieri, A. 1981: Excavations at Arslantepe (Malatya). Anatolian Studies 31, 101–19. Palmieri, A., Sertok, K. and Chernykh, E.N. 1993: From Arslantepe Metalwork to Arsenical Copper Technology in Eastern Anatolia. In Frangipane, M., Hauptmann, H., Liverani, M., Matthiae, P. and Mellink, M. (eds.), Between the Rivers and Over the Mountains (Rome), 573–99. Porada, E., Hansen, D.P., Durham, S. and Babcock, S.H. 1992: The Chronology of Mesopotamia, ca. 7000–1600 BC. In Ehrich, R.W. (ed.), Chronologies in Old World Archaeology 3rd ed. (Chicago), vol 1, 77–121; vol. 2, 90–124. Sagona, A.G. 1984: The Caucasian Region in the Early Bronze Age (BAR International Series 214) (Oxford). Selimkhanov, I.R. 1962: Spectral Analysis of Metal Articles from Archaeological Monuments of the Caucasus. Proceedings of the Prehistoric Society 4, 68–79. ——. 1978: Ancient Tin Objects of the Caucasus and the Results of Their Analyses. In Franklin, A., Olin, J. and Wertime, T. (eds.), The Search for Ancient Tin (Washington, DC), 53–58. Stein, G.J. 1999: Rethinking World Systems (Tucson). Tringham, R. 1996: But Gordon, Where Are the People? Some Comments on the Topic of Craft Specialization and Social Evolution. In Wailes (ed.) 1996, 233–39. Wailes, B. (ed.) 1996: Craft Specialization and Social Evolution: In Memory of V. Gordon Childe (Philadelphia). White, J. and Pigott, V. 1996: From Community Craft to Regional Specialization: Intensification of Copper Production in Prestate Thailand. In Wailes 1996, 151–75.
CHAPTER EIGHTEEN THE SAMARA BRONZE AGE METALS PROJECT: INVESTIGATING CHANGING TECHNOLOGIES AND TRANSFORMATIONS OF VALUE IN THE WESTERN EURASIAN STEPPES DAVID L. PETERSON, PAVEL F. KUZNETSOV AND OLEG D. MOCHALOV, WITH A BRIEF APPENDIX BY AUDREY G. BROWN AND EMMETT BROWN
Archaeologists acknowledge an important social dimension to the practice of metallurgy during the Bronze Age in the Eurasian steppes (ca. 3500–1000 BC), but to date there has been little effort to investigate the historical links between the technology of metalwork and its social role. In order to address this question, the Samara Bronze Age Metals Project is investigating the links between the production, circulation and consumption of prehistoric metalwork in the Samara oblast, a Russian administrative district in the Middle Volga and Volga-Ural regions of the western Eurasian steppes (Fig. 1). The project has been divided into three phases. The first phase, completed in 2001, was a collaborative archaeological survey performed by the authors in the Kamyshla district of north-eastern Samara, with the goal of investigating the contribution of local copper production to the consumption of copper-based metalwork during the Bronze Age. The second phase, completed in 2002, was the removal of samples from metal artefacts collected in previous investigations for intensive laboratory analysis. The third phase involves the analysis of the materials, techniques and processes that were associated with the samples collected in the first two phases of the project. This is being used to explore the sociological and technological changes in metal making and metal use in Samara from the Early to Late Bronze Age of the 3rd–2nd millennia BC. This paper presents a summary of the results to date and briefly discusses how the study of the value of metalwork can contribute to broadening the current understanding of prehistoric societies in western Eurasia.
Fig. 1. Location of the Samara oblast on the Eurasian steppes.
18. THE SAMARA BRONZE AGE METALS PROJECT
323
324
D.L. PETERSON, P.F. KUZNETSOV & O.D. MOCHALOV
Previous Approaches to Early Metalwork in the Steppes Metalwork has an important position in the archaeology of the Volga-Ural steppes, a region that follows the Samara river east from the Middle Volga to the Ural river (the traditional border between Europe and Asia) (Fig. 1). Researchers in this region mark the beginning of the Bronze Age with the introduction of copper metallurgy at the turn of the 4th/3rd millennium BC. Many also consider metal production and mobile pastoralism to have been the sole forms of productive economy in the Volga-Ural region throughout the Bronze Age (Vasil’ev et al. 2000, 7–9). The dominant model for the development of metallurgy in the Eurasian steppes posits the establishment ca. 5000–4500 BC of an import economy in which steppe pastoralists became dependent on settled agriculturalists to the south for copper. This informal trade would have been replaced by large-scale mining and copper production as a centralised industry in the southern Urals and Kazakhstan by the mid-2nd millennium BC, through migration, diffusion, social upheavals, and growing demands for copper and bronze (Chernykh 1992, 42–43, 193, 307–8; Kuzmina 1994, 223–30; 2000, 119; Anthony 1998). Research linked to this industrial model has tended to draw a sharp line between metallurgy as the production of metal from ore, and metalworking as the manufacture of metal objects by smithing and casting. This analytical move seems to be tied to an underlying premise that since the existence of metalwork during the Bronze Age depended on the presence of copper, the most significant innovations in working and using it would have originated and diffused from where copper ore was mined and smelted (Chernykh 1992, 7–10). While this division makes sense from a materialist perspective, it underdetermines the role that smiths played in fabricating metalwork and the ways in which metallurgy and metalworking were linked in a broader ‘metal-making’ process (Budd and Taylor 1995). Instead, there is a tendency to approach the technology of metalwork as being synonymous with the metallurgical production of the metal itself, while the potential evidence for metalworking techniques available through the metallographic examination of metal artefacts (Scott 1991) is often ignored. As a consequence, archaeometallurgical research in the Eurasian steppes is sometimes more concerned with tracing the sources of raw materials and, by proxy, paths of cultural influence, migration and the spread of Indo-European languages, than with describing how metalwork was actually made and used, and applying the latter toward broadening the current understanding of prehistoric culture and society.
18. THE SAMARA BRONZE AGE METALS PROJECT
325
During the Bronze Age in Eurasia, the practice of small-scale copper and bronze making became increasingly widespread and, as Budd and Taylor (1995) argue, was oriented toward ritual and symbolic activities rather than the large-scale, specialised extraction and production, and monoplex social roles associated with the industrial model, which are derived from an anachronistic back-projection of modern notions of technological change as driven by rational science. Similarly, research on Bronze Age metalwork in the steppes is often trained on technological origins, which are approached in terms of traits that were passed on genetically from ancestral archaeological cultures to their descendants. However, studies of prehistoric and historic metal making elsewhere suggest that technological origins are less important to understanding technical practice than diffusionist and ethnogenetic approaches typically assume. The true social innovation lies not in the invention of new products and activities but in their widespread adoption, in which they are put to a variety of potentially divergent uses (Renfrew 1986). A technological process like smelting can be accomplished by a variety of techniques, and made to conform to diverse environmental, socio-cultural, and economic circumstances (Gordon and Killick 1993). Techniques and processes are also open to change as they are reproduced over time, and transferred and recreated in different settings (Kohl 1987). Some researchers have recently proposed that the primary output of major Late Bronze Age copper mining localities may have consisted not of copper, but instead of raw or partially reduced copper ore (Chernykh 1994; 1998; Rovira 1999; Díaz del Río et al. this volume). If so it is quite possible that relatively small quantities of ore from various sources reached widely dispersed metalworking sites, where it was smelted for use as needed. In this case, the metal-making process may have been determined more by the activities of small metalworking operations spread throughout the steppes than by hypothetical centres of mining and metallurgy located strictly in ore-bearing regions. Metal making would have been more vulnerable to local variation and innovation than under the standardisation envisioned for centralised production (Chernykh 1992, 7–10). A more satisfactory account of the development of early metal making in the steppes will require greater emphasis on small-scale and dispersed forms of production, the social factors and cultural sensibilities that guided technological practice, and the consumption of its products.
326
D.L. PETERSON, P.F. KUZNETSOV & O.D. MOCHALOV
Value
Technology
Social Practice
Fig. 2. Simplified diagram of the interrelationship of technology, social practice and value.
Technology, Value, and Social Practice This project’s central hypothesis is that the production and use of metal in the prehistoric past were joined in local constructions of value (Peterson 2003). The intensive study of metalwork and metal making can thereby serve as a focus for examining the relationship between technology, value and social practice, in which elements such as the identity, status and rank of social persons is negotiated, defined and contested (Fig. 2).1 This approach combines the concerns of anthropologists (Graeber 1996; 2001) and archaeologists (Lechtman 1994; Bailey 1998; van Wijngaarden 1999) with the constitution of value and the agency of artisans in creating value in technical practice (Dietler and Herbich 1998; Dobres 2000). The study of value in a broad sense (Dumont 1980) may accommodate the symbolic dimensions of metal making that are made marginal or remain unaddressed in the archaeology of the steppes. Consideration of the activities that gave form and meaning to metal artefacts also avoids ‘the arthistory approach to prehistoric metalwork which often focuses on form in the absence of a proper understanding of generative processes’ (Budd and Taylor 1995, 134). Studies elsewhere have found that shifts in prehistoric technology and the widespread adoption of new ways of making and using metal corresponded to changes in cultural value systems (Lechtman 1984; Renfrew 1986; Bradley 1988; Chernykh 1992, 140–69). However, archaeological studies that focus on the technological and social changes surrounding ancient metalwork (for example, Hamilton 1996) are still unusual. 1
Fig. 2 is not meant to delineate the systematic relationship of these elements in a temporal process, rather it illustrates their somewhat ambiguous and mutually influential relationship as a point of discussion.
18. THE SAMARA BRONZE AGE METALS PROJECT
327
In material culture production, technology participates in a process through which meaningful objects are created in acts that are both technical and symbolic (Munn 1986; Dietler and Herbich 1998; Dobres 2000). Artisans create value by manipulating tools and materials to conform to aesthetic and other socio-cultural values in technical practice. The value that becomes attached to objects is subsequently affirmed and redefined in a variety of ways, including the exchange of gifts and commodities, ownership, display, and by their use in mortuary rituals as representative possessions of the deceased or as offerings dedicated by their survivors. However, artisans contribute more to this process than the manipulation of a repertoire of forms, techniques and media that is subsequently appropriated in social practice. They influence prevailing social standards by creating objects of high value through feats of virtuosity, in which they skilfully combine technical mastery with aesthetic sensibilities in the practice of their craft (Boas 1955; Gell 1992). These are among the ways in which continuity and change in material culture occur in the relationship between technology, value and social practice – a relationship that is instantiated by artisans in craft production, in the broader community’s perception of the significance of their acts and products, and by the incorporation of those works and judgments into social practice.
Bronze Age Metalwork in Samara and Adjacent Regions Early metalwork in the steppes was produced through a complex series of operations that included mining, smelting, alloying and the application of metalworking techniques in making finished objects. These techniques and operations are often possible to identify even in prehistoric contexts through laboratory analysis of metal artefacts and debris from archaeological sites (Northover 1989; Scott 1991; Miller 1994). Each of these activities presented an opportunity for the transformation of value: through investments of materials, fuel, knowledge, labour, tools and installations; in the circulation and exchange of raw materials and products (such as ingots); and in redefinitions of the form and significance of materials in the creation finished metalwork. The tension between the durability and mutability of copper-based metalwork – including the potential to recycle it – matches two ways anthropologists have identified in which people perceive material value (Graeber 1996; 2001). First, as the value that becomes attached to things such as heirlooms, ceremonial objects and works of art that have their own ‘cultural biography’ or a singular
328
D.L. PETERSON, P.F. KUZNETSOV & O.D. MOCHALOV
history and significance (Kopytoff 1986). This form of value is frequently linked to adornment of the body. Second, as an attribute of a generic substance or class of objects, like beads and cloth, which is more easily exchanged than things of a singular value. These often serve as currency and the materials for making other things, such as jewellery and clothing in the case of beads and cloth just mentioned. Their value is associated with this latent capacity for action (Graeber 1996; 2001). Copper rings and bracelets from the Khvalynsk cemetery (Eneolithic period, ca. 5000–4500 BC) (Fig. 3) are among the very earliest metalwork from the Eurasian steppes (Agapov et al. 1990, Anthony 1998, 98–9, n. 2). Spectrographic analysis of these and contemporary or slightly later metalwork from the Mariupol-Chapli and Sredni-Stog II style horizons further west has shown that they were made of copper from the Carpathians and Balkans (Chernykh 1992, 46–7). There are strong indications that by the mid-5th millennium BC, these ornaments were already being made in the western steppes with imported or recycled copper. The techniques used to make the copper ornaments from the steppes at this time differ from those identified in roughly contemporary Cucuteni-Tripolye metalwork, and there are a few unique types of metal artefacts known only from Mariupol-Chapli and Sredni-Stog II contexts (Chernykh 1992, 46–7, fig. 13; Ryndina 1998, 151–9). The metalworking techniques identified in the Khavalynsk ornaments were distinctive enough to assign them to their own Middle Volga metalworking focus (Ryndina 1998, 194). Trace element profiles of metalwork from kurgan burials of the Early Bronze Age Yamnaya horizon in the Middle Volga (ca. 3100–2500 BC) suggest that copper ore deposits in the Volga-Ural interfluve were being exploited for metallurgy by the turn of the 4th/3rd millennium BC (Chernykh 1992, 85–6; Anthony 1998, 103). This is supported by finds of local cupric sandstone ore in some Yamnaya burials like one in kurgan 2 at Utëvka I (Vasil’ev 1980, 38–42; Chernykh 1992, 86) (Fig. 3).2 Given the apparent widespread knowledge and regularisation of metalworking – as signalled by the beginning of local metallurgy – by the 3rd millennium BC, a generic value is likely to have developed for copper that was associated with the potential to cast, work and recycle it. Most of the metalwork from the Yamnaya horizon and the Poltavka horizon (Middle Bronze Age,
2
However, the possible use of cupric and ferric sandstone for pigmentation deserves further study, especially in relation to ochre graves.
18. THE SAMARA BRONZE AGE METALS PROJECT
329
Fig. 3. Location of Kamyshla raion and the sites discussed in the text (prepared with the help of Laura Popova).
330
D.L. PETERSON, P.F. KUZNETSOV & O.D. MOCHALOV
Fig. 4. The mace from Kutuluk 1.
ca. 2500–2000 BC) is comprised of types of copper implements common to the circum-Pontic region in the 3rd millennium BC. These include tanged knives and daggers, tetrahedral awls and chisels, adzes and, occasionally, shaft-hole axes. Some were fashioned from arsenic bronze that was probably imported from the Caucasus where large numbers of arsenical copper and arsenic bronze artefacts were being made at this time (Chernykh 1992, 132–3). The distribution of these deceptively utilitarian forms in the steppes is limited and they undoubtedly had special social significance. For example, many are woodworking tools of the type used to build the wagons that first appear in kurgan burials in the steppes and Caucasus ca. 3000 BC (Piggott 1983, 58; Anthony 1998, 103, n. 8; Kuzmina 2000, 119). An early example of metalwork that was clearly of high social significance is a mace from a kurgan burial at Kutuluk I, securely dated to ca. 2930 BC (Kuznetsov 1991; Anthony 1998, 103–4) (Figs. 3–4). This impressive object has no known analogies in the steppes; it is 48 cm long and contains 770 g of copper.3 It joins an implicit threat of force with aesthetic concerns with length, symmetry and the smooth texture and shine of carefully ground and polished copper. It has no visible use-wear, and apparently served as a symbol of the status and/or authority of the deceased. A variety of complex ornaments appeared during the Middle Bronze Age in the steppes in conjunction with the Abashevo style horizon, which began with Late Abeshevo in the Middle Volga forest-steppe zone (ca. 2000–1700 BC). The lost wax casting process became far more widely adopted than previously in making ornaments and figurines like the anthropomorphic ‘idol’ from Galich (Chernykh 1992, pl. 20). Silver is well represented in Abashevo assemblages
3
The weight reported here corrects a previously published figure of 1.5 kg (Anthony 1998, 104).
18. THE SAMARA BRONZE AGE METALS PROJECT
331
and was probably derived from copper-silver deposits in the southern Urals within the distribution of Abashevo sites. In contrast, tin bronze is extremely rare in Abshevo contexts and tin sources are absent within the distribution of Abashevo sites. These patterns suggest that Abashevo metalwork was being made from locally produced metal (Chernykh 1992, 200–2). There are only roughly 500 metal objects known from the Seima-Turbino horizon, which are spread in an unprecedented distribution from the Balkans to the eastern (Asiatic) steppes (Chernykh 1992, 215–33). There is a tendency to distinguish sharply between Seima-Turbino and Abashevo, Sintashta and other contemporary Middle to Late Bronze Age horizons. However, in light of the similarities in the metalwork from each, Seima-Turbino may instead be viewed as having encompassed the most prestigious goods in a broader metals economy, which circulated widely and were widely imitated. For instance, socketed spearheads, which are among the most distinctive Seima-Turbino forms, also occur in Srubnaya-Abashevo burials in the Volga and at Sintashta. But while most are copper and bronze, SeimaTurbino spearheads and axes were also cast in silver (Chernykh 1992, 218, 227, 232) and may be said to represent exceptional versions of more common prototypes. The higher proportion of tin bronzes in Seima-Turbino assemblages as opposed to those of other contemporary horizons (Chernykh 1992, 222) may have had more to do with the selective use of the most prestigious materials for the finest metalwork, rather than simply the location of workshops in orebearing zones as some believe. There is a long standing, though no longer universal, notion in the history of technology that the evolution of society in later prehistory was tied to increases in the efficiency of production resulting from the industrial use of metal, an idea most clearly vocalised in Anglo-American archaeology by V.G. Childe (for instance, Childe 1936). A similar idea seems to underwrite prevailing models of the development of metal making in the steppes, including the overriding emphasis on metallurgy vs metalworking and the increase over time in the gross output of copper and bronze. However, several key developments were as closely or more closely related to aesthetic than utilitarian concerns. Perhaps the most interesting examples of this are the ‘princely’ knives and daggers associated with the Seima-Turbino horizon, with ornamented hilts and pommels adorned with figures of livestock and people (Chernykh 1992, 219, figs. 73–75, 77, pl. 22). In the skilled combination of casting, forging and figural representations, they arguably represent the height of the technology and art of early metalworking in the steppes.
332
D.L. PETERSON, P.F. KUZNETSOV & O.D. MOCHALOV
The Potapovka horizon is named after the Middle to Late Bronze Age component of the Potapovka, Lopatin II and Utëvka VI kurgan cemeteries in Samara (ca. 2200–1700 BC) (Fig. 3), which is closely similar to the better known Sintashta horizon east of the Urals in metalwork, ceramics, horse trappings and mortuary and sacrificial rites (Genning et al. 1992; Vasil’ev et al. 1994; Anthony 1998, 105–6, n. 11). The metalwork from the Potapovka cemetery is to date the most well studied Bronze Age metal assemblage in Samara (Agapov and Kuzminykh 1994). The relatively high proportion of tools and weapons versus ornaments in the Potapovka assemblage bespeaks a ‘military character’ that is striking in comparison to Yamnaya, Poltavka and Abashevo assemblages. According to Agapov and Kuzminykh (1994), the metals present include unalloyed copper derived from ore deposits in the southern Urals and the Volga-Kama and Volga-Ural interfluves including Samara itself, arsenical bronze probably produced from arsenical copper ore deposits at Tash-Kazgan in the eastern Urals, and tin-arsenic and antimony-arsenic bronze. Patterns in the distribution of metalwork in the Srubnaya horizon (ca. 1900–1500 BC) indicate that just as the number of metal objects that were placed in burials declined sharply in the early to mid-2nd millennium BC, new contexts were developing for the consumption and abandonment of metal – hoards and settlements with permanent standing structures. Krasnosamaraskoe is the most well investigated example of the latter in Samara (Fig. 3). Permanent settlements were rare during the earlier Bronze Age in the steppes and their appearance coincided with significant changes in what metal was made and used for. Most of the metal in Srubnaya assemblages from the western steppes has a compositional profile that matches copper ore deposits west of the Urals and indicates the increasing importance of eastern sources in the procurement of metal (Chernykh 1992, 206–207). The majority of Srubnaya metalwork is composed of antimony-arsenic bronze while tin bronze is more common in contemporary assemblages east of the Urals. The addition of tin gives copper a golden colour while arsenic creates a silvery tint. Both are significantly harder and more easily cast than unalloyed copper. The addition of antimony counteracts the brittleness of arsenical bronze making it more functionally equivalent to tin bronze (Hamilton 1996, 14). The functional similarity in these products could mean that differences in the bronze consumed in the eastern and western steppes represent differences in both the technology and taste in the colour of metalwork. Questions of the significance of colour and other characteristics of copper-based metalwork are currently being taken up in ongoing research.
18. THE SAMARA BRONZE AGE METALS PROJECT
333
The Survey An archaeological field survey was conducted in the region of the upper Sok river in the Kamyshla district of north-eastern Samara during the summer of 2001 (Figs. 3 and 5) by a crew of staff and students from Samara State University, Samara State Pedagogical University, the Institute for the History and Archaeology of the Volga and the University of Chicago. The area surveyed was near the villages of Staroe Ermakovo, Chuvashkii Baitugan and Novoe Usmanovo, which geologists include in one of two copper ore bearing regions in Samara. The actual survey area was divided into three parcels totalling 9 km2 that were chosen through the examination of maps, satellite images and historical records. On a brief visit to Kamyshla in May 2001, Russian members of the team identified traces of ancient copper ore mining that are described in accounts of an 18th-century expedition to the area (Pallas 1773; Lepekhin 1795). These were investigated in the survey. We utilised a combination of surface reconnaissance and subsurface shovel testing as a methodology well suited to locating the small, ephemeral sites characteristic of the Bronze Age occupation of the western steppes. Reconnaissance was performed by walking 25 m spaced transects, which yielded a probability of 1.0 of intersecting sites as small as 0.0625 ha (Sundstrom 1993). Areas that were found to contain artefacts or archaeological features were identified as archaeological sites. Small excavation units (1 m2, 2 m2 and 16 m2) were completed in each of these to the level of the clay subsoil (materik), below which evidence of human occupation is no longer found. The excavated soil was screened for artefacts. Laura Popova (University of Chicago) supervised the processing of flotation samples to collect micro- and macro-botanical remains from the cultural strata of each site for further study. The survey identified nine new archaeological sites including three quarries and six habitations (Fig. 5). The latter have been identified by ceramic remains (Fig. 6) as belonging to the Late Bronze Age Srubnaya style horizon (ca. 1700–1500 BC), and represent the first recorded evidence of Bronze Age settlement in the region in which the upper Sok river descends from the beginning of the Ural foothills in Kamyshla. Four sites were invisible at the ground surface and were detected only through our use of subsurface testing. While settlement evidence predating the mid-2nd millennium is rare in the steppes, one of the six settlements (Ermakovo I) also contained a Middle Bronze Age Abashevo habitation component (ca. 2000–1700 BC). One site, the Kibit I settlement, contains the
Fig. 5. The Kamyshla survey area and sites recovered by the survey: 1, Kibit I settlement; 2, Kibit II settlement; 3, Kibit III settlement; 4, Novoe Usmanovo I quarry; 5, Baitugan I quarry; 6, Baitugan II quarry; 7, Baitugan I settlement; 8, Baitugan II settlement; 9, Ermakovo I settlement; 10, artefact scatter.
334 D.L. PETERSON, P.F. KUZNETSOV & O.D. MOCHALOV
18. THE SAMARA BRONZE AGE METALS PROJECT
335
Fig. 6. Srubnaya-type pottery fragments (1–5, 7), an intact ceramic cup (6), and worked stone from Kibit I.
336
D.L. PETERSON, P.F. KUZNETSOV & O.D. MOCHALOV
well-preserved sunken floor of a Srubnaya house, dated by ceramic evidence to approximately the mid-2nd millennium BC (Figs. 6 and 7).4 The results of the survey support an emerging pattern of widespread, smallscale and dispersed production of copper during the Bronze Age period in the western steppes, which challenges the dominant model of Bronze Age metallurgy as a large-scale, centralised industry. Preliminary analysis of faunal remains by Audrey Brown and Emmett Brown indicates a general pattern of mixed cattle and sheep-goat pastoralism, but marked as being relatively sedentary by the high frequency of pig bones in the assemblage (see Appendix below). The Bronze Age pastoral economy would have left as traces short-term or seasonal campsites like those the Samara Valley project found at Peschanyi Dol (see Anthony this volume) (Fig. 3). Using intensive survey techniques, we have shown the existence of as many as five of these in Kamyshia that would have eluded opportunistic identification. Knowledge of them is urgently needed to complement the research on kurgan burials that currently dominates the archaeology of the steppes. In contrast, the house floor at Kibit I represents a permanent settlement. Dating is to the Late Bronze Age period of the mid-2nd millennium BC. Altogether this evidence suggests the combination of the use of a few permanent structures and numerous temporary campsites by small pastoral groups, which may correspond to a time of increasing sedentism. No evidence was recovered from the quarries, such as diagnostic artefacts or carbon samples, which could securely date them to the Bronze Age period. This is a common problem given the nature of mines as ‘landscapes of destruction’ that are notoriously difficult to date. Surprisingly little evidence was recovered for metal making within the settlement sites aside from a small assemblage of slag-like materials. Those most likely to have been associated with Bronze Age metallurgy are of the green, glassy matrix formed by the reduction of copper ore by a non-tapping process, also found at Kargaly (see Díaz del Río et al. this volume). These came from an artefact scatter that included a flaked stone scraper 2 km north of Kibit I (Fig. 5.10). Subsequent testing revealed no site in the area, so that, like the mines, the evidence is ambiguous. Metallic ore is present in the cultural stratigraphy of the sites, but since it is ubiquitous in the region as whole and was not found in association with other clear evidence of
4
This dating was subsequently confirmed by two AMS dates by the NSF-Arizona AMS Laboratory (AA65309 3438 ± 39 BP, AA65310 3392 ± 38 BP). The excavation of the structure in 2004 by Kuznetsov, Mochalov and L. Popova confirmed its association with metal-making activities.
Fig. 7. Location of the 4 × 4 m excavation unit at Kibit I (left), and plan (upper right) and section (lower) of the edge of a house floor encountered in the excavation.
18. THE SAMARA BRONZE AGE METALS PROJECT
337
338
D.L. PETERSON, P.F. KUZNETSOV & O.D. MOCHALOV
production, it does not in itself constitute secure evidence of metallurgical activity. Corroborative evidence is however forthcoming from the 2004 excavation of Kibit I by Kuznetsov, Mochalov and Laura Popova. The small scale and dispersion of metal production – apparently for only small-scale exchange and local consumption – may have been related to the calendar and logistics of mobile pastoralism, a topic for future discussions.
Continuing Research In March and April 2002, Peterson removed samples adequate for metallographic analysis (Scott 1991) from 106 metal artefacts collected through previous investigations, dating to the beginning of the 3rd-middle of the 2nd millennium BC. These represent approximately one-third of the Bronze Age metalwork known from the Samara oblast. The sampling included drafting 1:1 scale drawings of each object, photographing them, and recording their dimensions, weight, appearance, condition and the precise locations from which samples were taken from each artefact. Kuznetsov and Mochalov arranged the cleaning and stabilisation of the sampled objects and the laboratory repair of areas from which samples were taken. Since the time of the conference, Dr Peter Northover (Department of Materials, Oxford University) has mounted, ground and polished these and additional samples for subsequent analysis of elemental composition and manufacturing techniques. This project has chosen value as a focus for understanding the importance of material culture to the people who made and used it. In doing so, the value of objects is being approached as neither a function of any single attribute like the presence of precious metals or scarce alloys, nor of the distance from consumer to source, but as multifaceted, relational, and derived from shifting combinations of meaningful qualities achieved by artisans in technical practice (Peterson 2003). Analysis of the composition and microstructure of these artefacts is being undertaken by wavelength-dispersed spectrometry (WDS) and metallography in order to identify the operations which resulted in slags and other production debris, and the techniques and materials that were combined in creating such characteristics as colour, form, hardness, and durability from which the value of the metalwork was arguably derived. The project thus attempts to move beyond archaeology’s traditional exclusive focus on economics to include an account of the symbolic dimensions of value rooted in aesthetics.
18. THE SAMARA BRONZE AGE METALS PROJECT
339
The results of the analysis will be applied in examining how the technology and value of metalwork was transformed over the course of the Bronze Age period through the changing significance of different uses of metal (adornment, productive and military advantage, etc.), and the relationship of these changes to social practice, as interpreted through the use of metal in burial rites and other social contexts. In the end, the results of the project as a whole will be used to construct a model of the changing technology and value of metal in Samara over the course of the Bronze Age, including the broader relationship between metal making and mobile pastoralism.
Acknowledgments We thank the Wenner-Gren Foundation for its generous funding of this research, and the Russian Academy of Sciences for its financial and institutional support. We also thank the students of the History Department of Samara State Pedagogical University, and Aleksandr Khokhlov, Boris Aguzarov, and Dmitri Kormilitsyn for participating in the field survey. Nikolai Nebritov and his crew personally visited the expedition and gave their expert opinion on the geology of the survey area. Yuri Kol’ev, Irena Vasil’eva, Phil. Kohl, Rabadan Magomedov, and the late Magomed Gadzhiev have generously given additional samples of materials for analysis. We thank David Anthony, Dory Brown, Michael Dietler, Igor Vasil’ev, Adam Smith, Nicholas Kouchoukos, Kathleen Morrison, and K. Aslihan Yener for their advice and assistance. Finally, Peterson gives his special thanks to the Popov family for opening their home to him with the greatest warmth and generosity as we began planning the project in the summer of 2000.
APPENDIX PRELIMINARY FAUNAL REPORT-KAMYSHLA SURVEY 2001 AUDREY G. BROWN AND EMMETT BROWN
The Kamyshla survey was conducted in Samara, southern Russia during the fall of 2001 under the direction of David Peterson of the University of Chicago and of Pavel Kuznetsov and Oleg Mochalov of the Institute for the History and Archaeology of the Volga (Samara, Russia). Five small Bronze Age sites were excavated, four of which, Kibit I–III and Ermakovo I, have produced faunal material. Very few Bronze Age settlement sites from this region have been excavated and even fewer faunal reports have been published. Although the sample size was small and the survey excavations limited in scope, we found it important enough to conduct a faunal analysis by both species and diagnostic element. The project uncovered a total of 349 animal bones (see Fig. 8); 163 of these bones were scrap (specimens identifiable only as bone). The remaining 186 or 53.3% of the bones were identifiable at least to a size category and element (for example, medium mammal rib). Sixty-four of these bones were identifiable to species and element. Cattle (Bos) and sheep/goat (Ovis/Capra) clearly dominate the species present (caprid bones referred to as sheep/goat when species could not be assigned). Five pig (Sus) bones (7.8% of total) were recovered from three of the four sites. Although there were very few measurable specimens present, they were of the size expected for domestic animals in the region. Faunal material from Kibit I was recovered from a Late Bronze Age house floor (see Fig. 7; for species distributions, see Fig. 8). Kibit II and III yielded smaller camps and for our purposes will be treated as a unit. Another possible camp, Ermakovo I was located several kilometres away from these sites (Fig. 5). The bone and species distributions for the camp sites are listed in Fig. 8. The residents of the Kibit and Ermakovo sites appear to have been herders of small and large stock. Cattle may have also been used for traction power. Of further interest is the relatively high frequency of pig bones. Because pigs are not well adapted to pastoralism, their presence indicates relative sedentism.
341
18. THE SAMARA BRONZE AGE METALS PROJECT
Site
Kibit Kibit II Kibit III
TNF Identifiable Identifiable n = 349 n = 186 to species n = 64
Sheep/Goat
Cattle
Pig
Horse
265 17 21
150 10 8
47 5 4
13 2 2
25 3 1
2 0 1
1 0 0
Ermakovo I 46
18
8
1
1
2
1
Fig. 8. Distribution of animal bones from the Kamyshla survey (prepared by Audrey G. Brown and Emmett Brown).
Data on aging, weathering, cut marks and burn marks were collected during processing and will be analysed along with species and carcass part distribution in a future report. Results of the subsequent excavation of Kibit I in 2004 will yield additional evidence concerning how the community provisioned itself, and the nature of its inhabitants’ relationship with animals.
Bibliography Agapov, S. and Kuzminykh, S. 1994: Metall Potapovskogo mogil’nika v sisteme Evraziiskoi Metallurgicheskoi Provintsii (Appendix 1). In Vasil’ev et al. 1994, 167–73. Agapov, S., Vasil’ev, I. and Pestrikova, V.I. 1990: Khvalynskii eneoliticheskii mogil’nik (Saratov). Anthony, D. 1998: The Opening of the Eurasian Steppe at 2000 BCE. In Mair, V. (ed.), The Bronze Age and Early Iron Age Peoples of Eastern Central Asia vol. 1: Archaeology, Migration and Nomadism, Linguistics (Washington, DC) 94–113. Bailey, D. (ed.) 1993: The Archaeology of Value (BAR International Series 730) (Oxford). Boas, F. 1955: Primitive Art (New York). Bradley, R. 1988: Hoarding, Recycling and the Consumption of Prehistoric Metalwork: Technological Change in Prehistoric Europe. World Archaeology 20.2, 249–60. Budd, P. and Taylor, T. 1995: The Faerie Smith Meets the Bronze Industry: Magic versus Science in the Interpretation of Prehistoric Metal-making. World Archaeology 27.1, 133–43. Chernykh, E.N. 1992: Ancient Metallurgy in the USSR: The Early Metal Age (Cambridge/New York). ——. 1994: L’ancienne production minière et métallurgique et les catastrophes écologiques anthropogènes: introduction au problème. Trabajos de Prehistoria 51.2, 55–68. ——. 1997: Kargaly. Zabytii mir (Moscow). ——. 1998: Kargaly: le plus grand ancien complexe minier et de métallurgie à la frontière de l’Europe et de l’Asie. In Frère-Sautot, M.-C. (ed.), Paléometallurgie des cuivres. Actes du colloque de Bourg-en-Bresse et Beaune, 17–18 oct. 1997 (Montignac), 71–6. Childe, V.G. 1936: Man Makes Himself (London). Dietler, M. and. Herbich, I. 1998: Habitus, Techniques, Style: An Integrated Approach to the Social Understanding of Material Culture and Boundaries. In Stark, M. (ed.), The Archaeology of Social Boundaries (Washington, DC), 242–73. Dobres, M.-A. 2000: Technology and Social Agency (Oxford).
342
D.L. PETERSON, P.F. KUZNETSOV & O.D. MOCHALOV
Dumont, L. 1980: On Value. Proceedings of the British Academy LXVI (London), 207–41. Gell, A. 1992: Technology of Enchantment and the Enchantment of Technology. In Coote, J. and Shelton, A. (eds.), Anthropology, Art, and Aesthetics (New York). Gening, V.F., Zdanovich, G.B. and Gening, V.V. 1992: Sintashta (Chelyabinsk). Graeber, D. 1996: Beads and Money: Notes Toward a Theory of Wealth and Power. American Ethnologist 23.1, 4–24. ——. 2001: Toward and Anthropological Theory of Value: the False Coin of Our Own Dreams (New York/Basingstoke). Hamilton, E. 1996: Technology and Social Change in Belgic Gaul: Copper Working at the Titleberg, Luxembourg, 125 BC–AD 300 (MASCA Research Papers in Science and Archaeology 13) (Philadelphia). Kohl, P. 1987: The Ancient Economy, Transferable Technologies and the Bronze Age World System: A View from the Northeastern Frontier of the Ancient Near East. In Rowlands, M., Larsen, M. and Kristiansen, K. (eds.), Centre and Periphery in the Ancient World (Cambridge). Kopytoff, I. 1986: The Cultural Biography of Things: Commoditization as Process. In Appadurai, A. (ed.), The Social Life of Things (Cambridge). Kuzmina, E.E. 1994: Otkuda prishli Indoarii? (Moscow). ——. 2000: The Eurasian Steppes: The Transition from Early Urbanism to Nomadism. In DavisKimball, J., Murphy, E.I., Koryakova, L. and Yablonsky, L.T. (eds.), Kurgans, Ritual Sites, and Settlements: Eurasian Bronze and Iron Age (BAR International Series 890) (Oxford), 118–25. Kuznetsov, P.F. 1991: Unikal’noe pogrebnie epokhi rannei bronzy na r. Kutuluk. In Drevnosti vostochno-ebropeiskoi lesostepi (Samara). Lechtman, H. 1984: Andean Value Systems and the Development of Prehistoric Metallurgy. Technology and Culture 25, 1–36. Lepekhin, I. 1795: Dnevhye zapiski putshestviya po raznym provintsiyam Rossiiskovo gosudarstva (St Petersburg). Miller, H. 1994: Metal Processing at Harappa and Mohenjo-Daro: Information from Non-metal Remains. In Parpola, A. and Koskikallio, P. (eds.), South Asian Archaeology 1993 (Helsinki), 497–510. Munn, N. 1986: The Fame of Gawa (Cambridge). Northover, P. 1989: Non-ferrous Metallurgy in Archaeology. In Henderson, J. (ed.), Scientific Analysis in Archaeology (Los Angeles), 213–36. Pallas, P.S. 1773: Putshestvie po raznym provintsiyam Rossiiskoi imperii (St Petersburg). Peterson, D.L. 2003: Ancient Metallurgy in the Mountain Kingdom: The Technology and Value of Early Bronze Age Metalwork from Velikent, Dagestan. In Smith, A.T. and Rubinson, K.S. (eds.), Archaeology in the Borderlands: Investigations in Caucasia and Beyond (Los Angeles), 22–37. Piggott, S. 1983: The Earliest Wheeled Transport (Ithaca, NY). Renfrew, C. 1986: Varna and the Emergence of Wealth in Prehistoric Europe. In Appadurai, A. (ed.), The Social Life of Things (Cambridge), 141–68. Rovira, S. 1999: Una propuesta metodológica para el estudio de la metallurgia prehistórica: el caso de Gorny en la región de Kargaly (Orenburg, Rusia). Trabajos de Prehistoria 56.2, 85–113. Ryndina, N.V. 1998: Drevneishee metalloobrabatyvaishchee proizvodstvo iugo-vostochnoi Evropy (Moscow). Scott, D. 1991: Metallography and Microstructures of Ancient and Historic Metals (Malibu, CA). Sundstrom, L. 1993: A Simple Mathematical Procedure for Estimating the Adequacy of Site Survey Strategies. Journal of Field Archaeology 20, 91–6. van Wijngaarden, G.-J. 1999: An Archaeological Approach to the Concept of Value. Archaeological Dialogues 6.1, 2–46. Vasil’ev, I.B. 1980: Mogil’nik Yamno-Poltavinskovo vremeni u s. Utëvka v srednem Povolzhe. In Arkheologiya vostochnoevropeishkoi lesostepi (Moscow), 32–58. Vasil’ev, I.B., Kuznetsov, P.F. and Semenova, A.P. 1994: Potaposkii kurgani mogil’nik (Samara). Vasil’ev, I.B., Kuznetsov, P.F. and Turetskii, M.A. 2000: Yamnaya i poltavinskaya kultury. In Kolev, Y.I. (ed.), Istoriya Samarskogo Povolzh’ya c drevneishikh vremen do nashikh dnei (Samara).
CHAPTER NINETEEN UNDERSTANDING THE PRODUCTIVE ECONOMY DURING THE BRONZE AGE THROUGH ARCHAEOMETALLURGICAL AND PALAEO-ENVIRONMENTAL RESEARCH AT KARGALY (SOUTHERN URALS, ORENBURG, RUSSIA) PEDRO DÍAZ DEL RÍO, PILAR LÓPEZ GARCÍA, JOSE ANTONIO LÓPEZ SÁEZ, M. ISABEL MARTINEZ NAVARRETE, ANGEL L. RODRÍGUEZ ALCALDE, SALVADOR ROVIRA-LLORENS, JUAN M. VICENT GARCÍA AND IGNACIO DE ZAVALA MORENCOS
Introduction: Research Design Vast regions of Eurasia have little or no copper ore. Accordingly, Evgenii Chernykh (1992; 1993; Chernij et al. 1990; Chernykh, Avilova et al. 2000; 2002) has posited that metallurgy is the critical factor for understanding the long-distance interactions of eastern European and north-west Asian societies from the Chalcolithic to the Iron Age. He has defined various metallurgical provinces based on the technical and typological characteristics of the centres of metalworking and/or production. The oldest and most important mining and metallurgical centre of the great Eurasian steppes is Kargaly. Production at this centre corresponds to the successive Circumpontic and Euroasiatic Metallurgical Provinces (Chernykh 1996, 87–8). The copper deposits of Kargaly lie in the steppe in Orenburg oblast, about 150 km north-west of its capital city. Kargaly’s 11 principal mining districts cover about 500 km2 (Fig. 1). Since 1990, the Russian Academy of Sciences in Moscow has conducted research in Kargaly under Chernykh’s direction. This research has shown that mining activity began in the Early Bronze Age (the Yamnaya-Poltavka culture) and peaked during the Late Bronze Age (the Srubnaya culture), which was a period of both sedentarisation and intensive metallurgical activity. At about 1400 BC, settlements were abandoned and metallurgy ceased abruptly. There is no evidence for mining until about AD 1745, when Russian industrialists began to exploit the deposits in this region again. These mining operations lasted until 1900, at which point they stopped being profitable (Chernykh, Kuzminykh et al. 1999; Chernykh, Lebedeva et al. 2002, 12, 109).
344
PEDRO DÍAZ DEL RÍO ET AL.
Fig. 1. Regional context of Kargaly Project and Kargaly mining districts (after Chernykh, Lebedeva et al. 2002, 34, fig. 2.13; 49, fig. 3.2).
19. PRODUCTIVE ECONOMY: RESEARCH AT KARGALY
345
Metal analyses of Early Bronze Age artefacts carried out by the Moscow Institute of Archaeology show that smelted copper from Kargaly was distributed as far as the Donets and Oka rivers (an area of nearly 1 million km2) and that ores from Kargaly are found over an area of 80000–100000 km2 (Chernykh 1996, 90, fig. 1; 1998b, 132, 130, fig. 1). The ore body is formed by veins and pockets of malachite in the thick layers of sandstone in the subsoil. There are no slag-heaps because copper was obtained by direct reduction of the ore, in which the resulting slag was crushed to recover entrapped metal prills and droplets (see below). We must rely on other archaeological and historical evidence in order to evaluate Kargaly mining and metallurgical production. Eighteenth-century Russian miners were guided by the remains (pits and galleries) of the prehistoric mining operations. These pits and galleries have been identified and dated by archaeological research, which has also documented domestic and funerary contexts related to them (Chernykh, Kuzminykh et al. 1999; 2000; Chernykh. 2002, 128–39; Zhyrbin 1999). Both phases of exploitation, Bronze Age and modern, correspond to the same archaic technology, one that only used charcoal for fuel and only smelted copper oxide ores (Chernykh 1998b, 131; 1996, 88). This evidence indicates that mining and metallurgical activities reached considerable intensity during the Bronze Age. Chernykh (1994, 63–7; 1998b, 132–3) has suggested that this intensification could well have been the cause of the sudden end of the prehistoric occupation at Kargaly (over-exploitation of the limited forests of the region would have made continued smelting impossible). Since 1993, a multidisciplinary Russian-Spanish team has undertaken a project specifically designed to understand the prehistoric mining and metallurgical activities at Kargaly and the causes of its sudden collapse.1 These investigations have focused on the central mining zone (B) of the Kargaly complex, where Chernykh’s team were conducting excavations of the mining and metallurgical site of Gorny (District V) (Fig. 1) (Rovira 1999; 2004; Vicent et al. 2004). Three parallel and complementary research programmes have been directed towards a comprehensive analysis of the socio-economic, metallurgical and environmental conditions of prehistoric production.
1
Spanish funding for this work comes from the agreement between the Russian Academy of Sciences and the CSIC and from projects PS950031 (1996–99) and PB98–0653 (1999–2002) of the Dirección General de Investigación Científica y Técnica. The principal investigator is M.I. Martínez Navarrete.
346
PEDRO DÍAZ DEL RÍO ET AL.
The first research programme is archaeological in focus. The Russian members of the team have conducted archaeological research to define the chronological, cultural, social and economic character of the settlements during the two periods of mining activity in the region (Bronze Age and 18th–19th centuries AD). This programme has included an extensive systematic survey to locate ancient and modern period settlements, a survey complemented by historical research. This survey and research has led to the discovery of some 20 Bronze Age settlements and burial mounds (kurgans), as well as of a number of later occupations (Chernykh 2002, 59–69). In addition, the Russian members of the team have a) conducted intensive excavations at the Srubnaya-phase site of Gorny, a settlement with rich evidence of metallurgical activity; b) test-pitted other Srubnaya occupations, in particular Gorny 2 and Novenki; and c) conducted excavations at the Yamnaya-Poltavka-, Abashevo- and Srubnaya-phase kurgans of Pershin, Uranbash, and Komisarovo (Chernykh, Lebedeva et al. 2002, 58, 71, 74; for the 21 C14 dates of the settlement of Gorny, see Chernykh 2002, 126–7, 136–7). In order to contribute to recent debates concerning Bronze Age subsistence practices, Spanish members of the team have intensified the recovery of palaeo-economic evidence aimed at analysing the nature of the prehistoric steppe economy, with sub-programmes devoted to archaeozoology, palaeocarpology (including flotation of numerous samples of archaeological sediments), anthracology and palynology (Antipina 1999; Œernych et al. 1998; Morales-Muñiz and Antipina 2003; López et al. 2001; López Sáez et al. 2002a; Uzquiano 2002). In complement to this research, Salvador Rovira (1999; Chernykh and Rovira 1998; Chernykh, Frère-Sautot et al. 1999) has directed archaeometallurgical research aimed at determining the technological characteristics of Kargaly metal production, with particular attention to energy efficiency models. To this end, three sub-programmes have been developed: a) characterisation of Kargaly mineral and metallurgical process through the analysis of copper slag; b) experimental smelting using the technology suggested by archaeological evidence, all of which offer precise indications on energy consumption and other production factors; and c) a comparative analysis of the experimental and archaeological results. These three lines of evidence allow us to draw accurate and reliable conclusions on metallurgical production and energy consumption, a key element when discussing Kargaly’s model of metallurgical production and its historical trajectory. Lastly, our environmental research involved two related programmes: a) systematic recovery of palaeo-environmental data from archaeological and natural
19. PRODUCTIVE ECONOMY: RESEARCH AT KARGALY
347
deposits and b) modelling the present-day landscape, with particular attention to the formation processes of the palynological record (Khotinsky 1984; Chibilev 1996; Gaillard et al. 1994). As part of our palaeo-environmental work, we developed six palaeopalynological sequences supported by radiocarbon dates (Fig. 2.1). Four sequences came from Srubnaya-phase archaeological sites under excavation, two from natural deposits. We also analysed charcoal and other botanical remains recovered by flotation from archaeological deposits at Gorny (López et al. 2001; 2003; López Sáez et al. 2002a; 2002b; Uzquiano 2002). These data permit us to define five bioclimatic phases. Our modelling of the present-day landscape combined observations in the field with analysis of multiband images from the Landsat 5 satellite’s Thematic Mapper (TM) sensor in order to develop analytical regional maps of the vegetation, soils, distribution of humidity, etc. Relief morphology was mapped using a digital terrain model (D’Antoni and Spanner 1993; Vicent et al. 2000). Finally our research design involved collection of data on the present-day pollen rain (Fig. 3). Two transects were defined: one (10 km × 5 km) was placed around the Srubnaya site of Gorny, in an area of intense mining activity in both prehistoric and modern times. The other transect (5 km × 5 km) was placed some 10km south of the former around the Srubnaya site of Novenki, in an area with no mining activity. These two areas were sampled independently. Using a 500 m grid we randomly selected 55 sampling units. These represented about 10% of the two transects. In each of those unit we took pollen samples in the upper 10 cm of the topsoil, established a detailed inventory of the flora, and described other relevant variables (such as the incidence of farming and stock-raising) (Chibilev 1996; López-Sáez 2002). These observations established the effect of geographic factors on vegetational distribution in the present-day landscape and thus help us understand representations of that vegetation in the palynological record. This enables us to interpret palaeopalynological sequences in terms of past vegetation distributions. At the same time, differences in mining and metallurgy between the sampling units demonstrate the impact of these activities on the landscape (Fig. 2.2).
Results of the Archaeometallurgical and Palaeo-Environmental Research at Kargly The archaeometallurgical survey at Gorny benefits from the large collection of smelting debris and metal objects uncovered during archaeological field work
348
PEDRO DÍAZ DEL RÍO ET AL.
Fig. 2. 1. Bioclimatic phases: estimated chronology BP and palaeoclimate (after Khotinsky 1984) and archaeology (after Chernykh, Avilova et al. 2000; 2002). Arrows up and down (1–4) suggest the evolution of arboreal pollen (López et al. 2001; López-Sáez et al. 2002a; 2002b); 2. Representation of selected ecological groups. Each pollynomorph has been assigned to a single ecological group. The correlation is based on the floral catalogue of the Kargaly region and on the species identified during our field work. These pollynomorphs might correspond to several other ecological groups in other regions; 3. Cubic regression model: distance to forests explains .27646 of Betula pollen variance.
19. PRODUCTIVE ECONOMY: RESEARCH AT KARGALY
349
Fig. 3. Sampling model on digital terrain model: location of palaeopalynological sequences and pollen rain samples at Kargaly (South Urals, Russia).
(Chernykh, Kuzminykh et al. 1999; Chernykh 2002, 24–5, 54, 72, 85, 88, 92, 105, 111, 119). The results of their analysis by scanning electron microscopy, X-ray fluorescence spectroscopy and metallography show that metallurgical technique was a primitive one that worked copper oxide ores by a non-intentional slag smelting process. That is to say, without adding fluxes to get a low viscosity and lower melting point slag. A glassy matrix containing silica compounds forms the copper slag. Some samples exhibit well-formed lathes of pyroxene and akermanite crystals. Fayalite is usually absent. Thus, they must be slags obtained from a direct ore reduction process. As the slag viscosity is very high, most of the copper formed during that process remains trapped within the matrix. Cast objects were finished by cold hammering and, occasionally, annealing. The experimental replication of copper smelting started with copper ore selection. After this, the ore was crushed into small pieces and charged in the furnace
350
PEDRO DÍAZ DEL RÍO ET AL.
mixed with charcoal. After four hours, some big lumps of slag were obtained. Analysis using scanning electron microscope facilities allowed us to determine the chemical composition of the crystallographic phases. The identification was completed using an optical microscope. Remarkably, both archaeological and experimental slags are very similar in composition and phase structure. Copper was recovered by crushing the slag into stone mortars, taking apart the prills visible to the naked eye, and washing the slag dust onto a dish to remove the smaller portions by difference of density. The copper obtained in this way was heated into a crucible, melted and poured into a small ingot-mould carved in wood. As we know the charcoal consumption in each step of the process, we are in a position to calculate the amount of fuel needed to obtain copper in situ. Based on the experimental results, to obtain a 1 kg ingot of copper one would have to burn 65 kg of charcoal (obtained from about 500 kg of dry wood). Thus, our experimental smelting leads us to estimate that over the 300-year occupation of Gorny the copper production would have been 21.4 mt (requiring 10700 mt of wood as fuel) (Rovira 1999, 109–10; Horne 1982). We will now focus on what our research has discovered about the amount, composition and distribution of the region’s woodlands and the implications of this in evaluating the scale of metallurgical production at Kargaly. Based on archaeological, archaeometallurgical and archival evidence, Chernykh (1994, 63, 65; 1998b, 130, 132) considers that the entire cycle of mining, smelting and ingot-casting was practised intensively at Kargaly during the occupational spans of the permanent settlements of the Srubnaya culture, with most of the products of this process being exported from the region. Chernykh (1998a, 72) calculates that a 150,000 mt of copper were produced in the Kargaly region during the Bronze Age, 100,000 mt of which were created during the Srubnaya period. Given the large amounts of fuel required in ore smelting, Rovira (1999, 111) estimates that 75 million mt of wood would have been consumed in the Bronze Age. Based on the simplest assumption that production was constant over time, this would amount to 37500 mt of wood per year. Following Chernykh’s estimate of forest productivity in the Orenburg region, this implies the annual felling of 150 ha of woodland (Chernykh 1994, 60). These estimates stand in contrast to the limited amount of woodlands in the region today. If they are correct, either the availability of forest resources has diminished drastically since the Bronze Age or one must propose an alternative model of how prehistoric metallurgy operated. Palaeobotanical evidence sug-
19. PRODUCTIVE ECONOMY: RESEARCH AT KARGALY
351
gests that birch (Betula pendula) and oak (Quercus robur) were the main fuels (Uzquiano 2002; López-Sáez et al. 2002). It is unlikely that dung was used as a permanent fuel-source (Lebedeva 2004, 244–7). We must, therefore, evaluate the local and regional energetic potential by analysing the distribution of forests of these species over time. Based on our cartographic analysis at Landsat TM images we estimate that today forest covers about 2.6% of the Kargaly region. For the most part this is gallery forest located along the sides and headwaters of the numerous seasonal watercourses that make up the region’s drainage network. This is constituted predominantly of birch and poplar (Populus tremula). The only forest composed of willow (Salix) and poplar occurs along the banks of the river Usolka, the only permanent watercourse in the study area. If we assume a fully sustainable exploitation of forests with a cycle of 60 years for full recovery of initial productive capacity, Chernykh’s estimation of metallurgical production would demand 9000 ha of woodland.2 Given our estimate of a current forest cover of 2.6%, this would require a territory of 3500 km2, an area seven times the size of the whole mining complex of Kargaly. If we reduce the exploitation cycle to 30 years, the forested area would be more than 1700 km2, more than three times the size of the mining district. These estimates establish a threshold beyond which forest regeneration is impossible. In short, the present amount of woodland is incompatible with a sustainable metallurgical model over the long term. It should be noted that the current situation is the result of the massive impact of agriculture and stock-raising (particularly connected with Soviet colonisation) added to the effects of the long recent phase of mining (from 1745 to 1900). We must suppose, therefore, that the available forests observed in the present constitute only a fraction of what existed in the Bronze Age. According to Chernykh (1994, 63–7; 1998b, 132–3), the mining complex collapsed at the end of the Bronze Age mainly because forests were exploited above the level of sustainability. To evaluate this hypothesis we examined the variability of the representation of different species of trees at Kargaly over the course of the palaeopalynological sequences obtained in our research (Fig. 2.2). Our results suggest that the process of change is fairly close to the expectations of Chernykh’s model. The
2
The estimation of the length of the cycle is based on historical evidence (Chernykh 1994, 60). Exploitation may have been based on pruning trees, combined or not with cutting them down.
352
PEDRO DÍAZ DEL RÍO ET AL.
proportion of the ‘autochthonous arboreal’ pollen category is from bottom to top as follows: a) a relative minimum very close to present values during the Srubnaya occupation (pollen phase 3); b) a certain recovery during pollen phases 4 and 5; and finally c) an absolute minimum in the present (pollen phase 6). Nevertheless, the palynological evidence alone cannot corroborate any of these hypotheses because it does not permit a quantitative evaluation of the areas covered by pollen-producing species. We can, however, establish the significance of these changes for the shape of the landscape on the basis of our research on the formation processes of the pollen record. Results indicate that the representation of ‘autochthonous arboreals’ (and especially of its most important component, birch) depends on the distance between the point where pollen is obtained and the location of groups of trees (Fig. 2.3, Fig. 4). Analysis of the present-day pollen rain permits us, therefore, to model the significance of quantitative variations in arboreal pollen in terms of the local distribution of forests. To do this we compare the values obtained in the different palaeopalynological samples with the variability observed in present-day samples in relation to their location in the map of regional forests. This research demonstrates that values in Srubnaya times correspond with the situation today in contrast with what occurs in pre- and post-Srubnaya phases. We must conclude, then, that forest distribution, at least around Gorny, was similar to that seen now, that is, it was limited to gallery forest. Before and after Srubnaya times, values suggest sampling points were closer to forests, and, thus, that these forests were denser with respect to the points of observation. In order to confirm and generalise these conclusions we carried out a numerical classification of all the pollen spectra obtained in the sampling areas in terms of the proportion of the different ecological groups represented in them. In a first trial, a hierarchical classification using k-mean method identified five groups of spectra. Then using Landsat TM images of the areas within 250 m of the sampling points of present-day pollen rain, we measured the present-day distribution of the Normalised Difference Vegetation Index (NDVI) (Fig. 5.1). As is to be expected, these values proved to be relatively homogenous within each group of the classification. This permits us to conclude that NDVI values for palaeopalynological samples classified within a group would fall within the observable distribution for that group today. Seventy-five percent of the samples attributable to the Srubnaya period (Biozone Karg-3) from Gorny are classified, interestingly, within group 2, which presents a mean NDVI value of about 66 (Fig. 2.1). When the Geographical Information System displays the points
19. PRODUCTIVE ECONOMY: RESEARCH AT KARGALY
353
Fig. 4. Map of distances between sampling points and forests at Kargaly (South Urals, Russia).
whose NDVI value is within one standard deviation around the mean of this distribution, we see that all of these correspond to a natural formation of herbaceous steppe modified by mining activities (Fig. 5.2). We may tentatively conclude, therefore, that the distribution of vegetation around the mining settlement of Gorny was almost identical to what we can observe today, and that the fuel productive potential of forests in the Gorny mining district during Srubnaya phase was similar to that of the present. Taking into account the extensive distribution of Kargaly copper throughout the Srubnaya or Srubnaya-Abashevskaya cultural area, we must then conclude that most of
354
PEDRO DÍAZ DEL RÍO ET AL.
Fig. 5. 1. Normalised Difference Vegetation Index (NDVI) and sampling model; 2. Predictive model at Kargaly (South Urals, Russia).
19. PRODUCTIVE ECONOMY: RESEARCH AT KARGALY
355
it was exported, not as metal, but as copper ore or lumps of slag-like material, a hypothesis already proposed by Chernykh (1998a, 132–3; 1994, 65; Rovira 1999, 112) as an alternative to that of intensive metallurgical production at Kargaly. Deciding between these two scenarios has broad implications for all aspects of the historical interpretation of Kargaly metallurgy since they involve opposing models of production, circulation and the social division of labour. Therefore, we must rethink current assumptions about the socio-economic structure of the Late Bronze Age and the nature of regional and trans-regional exchange networks during this period. Because of their temporal and spatial scale and because of the importance of Eurasian metallurgical technology, the metallurgical provinces defined by Chernykh and his collaborators constitute the essential framework within which this reinterpretation must take place.
Acknowledgments We are grateful to E.N. Chernykh and his team, and especially to Tamara O. Teneishvili for helping us feel that we all spoke the same language. We thank Antonio Gilman for his critical and interactive translations and María Cruz Berrocal and Elías López Romero for helping prepare the illustrations.
Bibliography Antipina, E.E. 1999: Kostnye ostatki zhivotnykh s poseleniya Gornyi (biologicheskie i arjeologischeskie aspekty issledovanya). RosA 1, 103–16. Œernych [Chernykh], E.N., Antipina, E.E. and Lebedeva, E.J. 1998: Produktionsformen der Urgesellschaft in den Steppen Osteuropas (Ackerbau, Viehzucht, Erzgewinnung und Verhüttung). In Hänsel, B. and Machnik, J. (eds.), Das Karptenbecken und die Osteuropäische Steppe. Nomadenbewegungen und Kulturaustausch in der vorchristlichen Metallzeit (4000–500 v.Chr.) (Rahden, Westf.), 233–52. Chernij [Chernykh], E.N., Avilova, L.I., Bartseva, T.B., Orlovskaia, L.B. and Teneishvili, T. 1990: El Sistema de la Provincia Metalúrgica Circumpóntica. Trabajos de Prehistoria 47, 63–101. Chernykh, E.N. 1992: Ancient Metallurgy in the USSR. The Early Metal Age (Cambridge). ——. 1993: Rythm and Models of the Fundamental Cultural and Technological Destructions after Metal Discovery. In Martínez Navarrete, M.I. (ed.), Theory and Practice of Prehistory: Views from the Edges of Europe (Santander), 275–300. ——. 1994: L’ancienne production minière et métallurgique et les catastrophes écologiques anthropogènes: introduction au problème. Trabajos de Prehistoria 51.2, 55–68. ——. 1996: The Dawn of Mining and Metallurgy in Eastern Europe: The New Discoveries. In Bagolini, B. and Lo Schiavo, F. (eds.), The Colloquia of the XIII International Congress of Prehistoric and Protohistoric Sciences, Forli, 8–14 September 1996 vol. 10: The Copper Age in the Near East and Europe (Forlì), 85–93.
356
PEDRO DÍAZ DEL RÍO ET AL.
——. 1998a: Kargaly: le plus grand ancien complexe minier et de métallurgie à la frontière de l´Europe et de l’Asie. In Frère-Sautot, M.-C. (ed.), Paléometallurgie des cuivres. Actes du colloque de Bourg-en-Bresse et Beaune, 17–18 oct. 1997 (Montignac), 71–6. ——. 1998b: Ancient Mining and Metallurgy in Eastern Europe: Ecological Problems. In Hänsel, B. (ed.), Man and Environment in European Bronze Age. The Bronze Age: the First Golden Age of Europe (Kiel), 129–33. —— (ed.) 2002: Kargaly vol. II: Gornyi – poselenie epokhi pozgnei bronzy: Topografiya, litologiya, stratigrafiya: Proizvodstvenno-bytovye i sackral’nye sooruzheniya: Otnositel’naya i absolyutnaya khronologiya (Moscow). Chernykh, E.N., Avilova, L.I. and Orlovskaya, L.B. 2000: Metallurgical Provinces and Radiocarbon Chronology (Moscow). Chernykh, E.N., Avilova, L.I., Orlovskaya, L.B. and Kuzminykh, S.V. 2002: Metallurgiya v Tsirkumpontiiskom areale: ot edinstva k raspadu. RosA 1, 5–23. Chernykh, E., Frère-Sautot, M.-C., Happ, J. and Rovira, S. 1999: Experimentations de fonderie dans le site de minerai de cuivre de Kargali (Oural – Russie). CU+. Bulletin du Groupe de Travail International sur la paléométallurgie des cuivres et des minerais associés. Association pour la Promotion de l´Archéologie de Bourgogne (A.P.A.B.) 1, 2–4. Chernykh, E.N., Kuzminykh, S.V., Lebedeva, L.Y., Agapov, S.A., Lunkov, V.Y., Orlovskaya, L.B., Teneishvili, T.O. and Valkov, D.V. 1999: Arkheologicheskie pamyatniki epokhi bronzy na Kargalakh (poselenie Gornyi i drugie). RosA 1, 77–101. Chernykh, E.N., Kuzminykh, S.V., Lebedeva, L.Y. and Lunkov, V.Y. 2000: Issledovanie kurgannogo mogilnika u s. Pershin. Arkheologicheskie pamyatniki Orenburzhya IV, 63–84. Chernykh, E.N., Lebedeva, L.Y., Kuzminykh, S.V., Lunkov, V.Y., Gorozhanin, V.M., Gorozhanina, E.N., Ovchinnikov, V.V. and Puchkov, V.H. 2002: Kargaly vol. I: Geologogeograficheskie kharakteristiki—Istoriya otkrytii, ekspluatatsii i issledovanii: Arkheologicheskie pamyatniki (Moscow). Chernykh, E.N. and Rovira, S. 1998: La metalurgia antigua del cobre en Kargaly (Orenburg, Rusia): informe preliminar. In Frère-Sautot, M.-C. (ed.), Paléometallurgie des cuivres. Actes du colloque de Bourg-en-Bresse et Beaune, 17–18 oct. 1997 (Montignac), 77–83. Chibilev, A.A. 1996: Prirodnoie naslediie Orenburgskoi oblasti (Orenburg). D’Antoni, H. and Spanner, M.A. 1993: Remote sensing and modern pollen dispersal in Southern Patagonia and Tierra Del Fuego (Argentina): Models for Palaeoecology. Grana 32, 29–39. Gaillard, M.J., Birks, H.J.B., Emanuelsson, U., Karlsson, S., Lageras, P. and Olausson, D. 1994: Application of modern pollen/land-use relationships to the interpretation of pollen diagrams – reconstructions of land-use history in south Sweden, 3000–0 BP. Review of Palaeobotany and Palynology 82, 47–73. Horne, L. 1982: Fuel for the metal worker. Expedition 25.1, 6–13. Khotinsky, N.A. 1984: Holocene vegetation history. In Velichko, A.A. (ed.), Late Quaternary environments of the Soviet Union (London), 179–200. Lebedeva, L.Y. 2004: Glava 8. Arkheobotanischeskoe issledovaniya. In Chernykh, E.N. (ed.), Kargaly vol. III: Selishe Gornyi—Arkheologicheskie materialy; Tekhnologiya gorno-metallurgicheskogo proizvodstva; Arkheobiologicheskie issledovaniya (Moscow), 240–8. López, P., Chernykh, E.N. and López Sáez, J.A. 2001: Palynological analysis at the Gorny site (Kargaly region): The Earliest Metallurgical Centre in Northern Eurasia (Russia). In Goodman, D.K. and Clarke, R.T. (eds.), Proceedings of the IX International Palynological Congress, Houston, Texas, 1996 (Dallas), 347–55. ——. 2002: Glossarii: sovremennaya flora Kargalov. In Chernykh 2002, 170–4. López, P., López-Sáez, J.A., Chernykh, E.N. and Tarasov, P. 2003: Late Holocene vegetation history and human activity shown by pollen analysis of Novienki bog (Kargaly Region, Orenburg Oblast, Russia), Vegetation History and Archaeobotany 12, 75–82. López Sáez, J.A., López García, P. and Martinez Navarrete, M.I. 2002: Glava 10. Palinologicheskie issledovaniya na kholme Gornogo. In Chernykh 2002, 153–65. Morales-Muñiz, A. and Antipina, E. 2003: Srubnaya Faunas and Beyond: A Critical Assessment of the Archaeozoological Information from the East European Steppe. In Levine, M., Renfrew, C. and Boyle, K. (eds.), Prehistoric Steppe Adaptation and the Horse (Cambridge), 329–51.
19. PRODUCTIVE ECONOMY: RESEARCH AT KARGALY
357
Rovira, S. 1999: Una propuesta metodológica para el estudio de la metalurgia prehistórica: el caso de Gorny en la región de Kargaly (Orenburg, Rusia). Trabajos de Prehistoria 56.2, 85–113. ——. 2004: Glava 4: Metallurgiya medi: izuchenie tekhnologii. In Chernykh, E.N. (ed.), Kargaly vol. III: Selishe Gornyi—Arkheologicheskie materialy; Tekhnologiya gorno-metallurgicheskogo proizvodstva; Arkheobiologicheskie issledovaniya (Moscow), 106–33. Uzquiano, P. 2002: Prilozhenie 1. Opredelenue drevesnykh ostatkov s Gornogo. In Chernykh 2002, 166–69. Vicent [García], J.M., López Garcia, P., Martinez Navarrete, M.I., López-Sáez, J.A., de Zavala Morencos, I. and Rovira Llorens, S. 2004: Teledeteccíon, Arqueologia e Historia de la Vegetacíon: el Proyecto Kargaly. International Journal of Archaeology Suppl. 1 (abstracts from the Astrobiology Science Conference 2004, NASA, Ames, 28 March–1 April), 9. Vicent García, J.M., Rodríguez Alcalde, A.L., López Sáez, J.A., Zavala Morencos, I. de, López García, P. and Martínez Navarrete, M.I. 2000: ¿Catástrofes ecológicas en la estepa? Arqueología del Paisaje en el complejo minero-metalúrgico de Kargaly (Región de Orenburg, Rusia). Trabajos de Prehistoria 57.1, 29–74 Zhyrbin, I.V. 1999: Elektrometricheskie issledovaniya na poselenii Gornyi. RosA 1, 117–24.
CHAPTER TWENTY WHY HAVE SIBERIAN ARTEFACTS BEEN EXCAVATED WITHIN ANCIENT CHINESE DYNASTIC BORDERS? KATHERYN M. LINDUFF
Introduction As a beginning graduate student in the study of Chinese art history and archaeology, I was surprised to learn that Bronze Age archaeological contexts in China dating from the 2nd and 1st millennia BC contained materials that were not Chinese. At that time, these materials were identified simply as ‘Siberian’. The presence of such artefacts in early Chinese dynastic contexts had fuelled debates in the late 1940s, for instance, between the Swedish philologist Bernhard Karlgren (1946) and the German art historian Max Loehr (1949a–b) about the origins of bronze production and the primacy of Siberia or China in that invention. Moreover, speculations about the relationship between the peoples of these regions, the early Chinese and the Siberians, essentialised their ways of life: the sedentary, agricultural and politically centralised peoples in China, while those in southern Siberia where thought to be mobile herders who roamed across great tracts of open steppe. That model, commonly called the ‘steppe and the sown’, also alleged that these lifestyles were self-generating, exclusive and incompatible (Lattimore 1940; Whittaker 1994). In the second half of the 20th century this bi-polar model was feverishly defended, evidently sustained by the ethnic tension along the borders between China and Russia as well as the politics of the Cold War. Likewise, the types and styles of artefacts commissioned by each set of patrons were looked upon as stemming from opposing traditions. For instance, the animal imagery on the Chinese ritual bronzes is a composite made up of parts of wild animals or is fantastic altogether. By contrast, the representations of animals, either singly or in combat, that adorned portable bronze items such as belt plaques, ornaments, knives and daggers of the peoples of the steppe were more naturalistically rendered (Karlgren 1945; Loehr 1949a–b; Bunker et al. 1970).
20. SIBERIAN ARTEFACTS IN ANCIENT CHINA?
359
Although there are significant differences between the lifestyles and artefacts produced by each, this past research has concentrated primarily on the extremes of those specialisations. In doing so, the possibility of understanding mixtures of adaptive subsistence strategies and styles and types of artefacts is precluded. From ancient Chinese written sources, we are told that these peoples were in contact with each other. The residue of that interaction in the form of bronze artefacts resides in many archaeological settings in the Northern Zone1 of China as well as in the Yellow River basin over the past several decades. More precise archaeological understanding of the lifestyles practised in this area has only very recently been systematically studied (Linduff 2002–04; Liu Li 1996; Underhill et al. 1998). Preliminary reports show, for instance, that there is wide variation in subsistence strategies ranging from fully sedentary agriculturally based economies to semi-mobile agro-pastoral societies, and more rarely, fully nomadic adaptation. So, too is there regional variation among styles of artefacts (Bunker et al. 1997). What this variation says both about the steppe vs the sown model as well as interaction among these peoples of southern Siberia and northern East Asia is what interests me here. Both ancient as well as modern authors have found convenient ways of framing the relationship between peoples at the frontiers of early Dynastic China and those at dynastic centres. The most common explanation for the large numbers of objects of Siberian prototype in Chinese settings was that they were collected as war booty, such as those uncovered in the royal cemetery at Anyang. There they have been explained by most to provide a show of power and authority of the Shang over a ‘foreign’ enemy (Li 1980; K.C. Chang 1986). Although the ways of life and artefacts were viewed as exclusive, recent archaeological investigation in northern China suggests that the supposed cultural, and even ecological, borders were far less rigid and hostile and far more fluid than ancient and even modern authors have proposed. This paper assumes that communication among diverse peoples at the frontiers and even in the dynastic centre was not only frequent, but also suggests that it was opportunistic and diverse in purpose. These relationships during the 2nd and early 1st millennia BC depended especially on local economic, political and probably personal factors. Here I will discuss the internal structure and dynamics of four sites, three in the northern frontier area during early dynastic Chinese history (ca. 1900–5th 1
This includes the provinces of present-day northern and western China: Liaoning and Jilin to the north-east, Inner Mongolia and Gansu to the north, and Xinjiang in the west.
360
KATHERYN M. LINDUFF
century BC) and one at the capital of the Shang dynasty itself (ca. 1250–1050 BC), in response to the question: ‘Why are Siberian artefacts found within the borders of early dynastic China?’ (Map 1). Each of the four cases documents a distinctive type of interaction, and none suggests that these artefacts were collected only as war booty. Moreover, I suggest that selection of certain types of artefacts and iconography afforded some benefit, either political or social, to their owners. I have found that the presence of ‘Siberian’ artefacts in these Bronze Age settings represents at least the following four types of interface: • transmission of ideas, in this case metallurgical technology (at Huoshaogou, Gansu province, ca. 1900–1600 BC); • the presence of surplus extraction and trade, in this case of raw materials (at Zhukaigou, Inner Mongolia, ca. 1500 BC); • the establishment of a marriage alliance (at Anyang, Henan province, ca. 1250–1050 BC); • the assertion of military control over hostile peoples within dynastic political limits and perhaps, of ethnogenesis (at Yuhuangmiao, district of Bejing, ca. 8th–5th centuries BC).
Siberian Artefacts in Chinese Contexts Transmission of Ideas/Technology. Recently recovered evidence of the transference of ideas and technology comes from the site called Huoshaogou, in Gansu province in present-day western China (Map 1). The site is far to the west of the supposed dynastic centre of the Xia and is dated from ca. 1900 to ca. 1600 BC. Recent excavations have documented knowledge of both metallurgy and horse breeding before comparable activities can be found in the Central Plain (Fitzgerald-Huber 1995; Linduff 2000; 2002). At Huoshaogou, habitations and burials of these agricultural peoples include local style pottery as well as metal artefacts with Andronovo (Federovo) and Seima-Turbino analogies, or from cultures in eastern Kazakhstan and other parts of southern Siberia dated from the early 2nd millennium BC (Barnard 1987; 1993; Bunker 1997; Linduff 1997; 2000; Li 2000) (Fig. 1).
20. SIBERIAN ARTEFACTS IN ANCIENT CHINA?
Fig. 1. Map showing Neolithic, Xia and Shang archaeological sites.
361
362
KATHERYN M. LINDUFF
The most conspicuous are the production of metal artefacts using metalworking techniques including cold and hot forging, as well as bi-valve casting and alloying of metals including arsenic-copper and copper-tin-lead that represent their earliest and easternmost use in East Asia at about 1900 BC. Although metallurgy was not at the vanguard of exchange systems in this region at this time, the technology could easily diffuse from one area to another at a level appropriate to each local context and must have developed in areas close to natural resources. These sites were not predictably stabile and critically important technologies such as metalworking or animal breeding were easily transferable to less developed areas situated close to natural source deposits or breeding plains for live resources. Economic development and dependency seem not to be linked in north Asia at this time and these early industries were relatively self-sufficient (Linduff 1995, 142). Recently, more such artefacts have been identified in dated contexts in Xinjiang province west of Houshaogou (Mei 2001) and are being used to trace further the route of transference of metallurgical knowledge into western China from southern Siberia. These technologies were apparently transferred at a somewhat later date around 1600 BC into the existing dynastic centre (as at Erlitou) in the Central Plain where casting technology was adjusted to conform to a local practice. Surplus Extraction and Trade. The site at Zhukaigou, Inner Mongolia, was excavated in 1972 and 1973 (Xu 1989; anon. 1990; Linduff 1995), and a site report has very recently been published (NZWKYOB 2000). Zhukaigou stands in an important location where the steppe cultures meet the settled farming civilisation of the Chinese. The site itself is a village occupied by settled farmers located far north and west of the mid-Shang dynastic centres, with which it is contemporary. Although it is located far outside the sphere of influence of the dynastic centre, mid-Shang style bronze vessels and ge daggers were excavated in the Levels 4 and 5, dating from between 1500 and 1250 BC, along with local ceramic and metal objects. In addition, bronze knives with analogies among Siberian artefacts of the Karasuk archaeological culture were also found in elite burials at Zhukaigou (Fig. 2). Zhukaigou is strategically located between great lodes of copper to its north and the capital of the middle Shang to their south. Extraction of this raw material would have provided local Zhukaigou elites the opportunity to control a commodity central to maintenance of power and authority of the Shang. Among the thousands of ritual objects commissioned by the political elite of the Shang,
20. SIBERIAN ARTEFACTS IN ANCIENT CHINA?
363
Fig. 2. Bronze spatulas (bishou) from Huoshaogou, Yumena, Gansu (after Linduff 2000, fig. 10).
vessels and weapons made of bronze afforded them symbols of political, social, as well as spiritual power in rituals dedicated to commemorating their ancestors and to declaring social inequality. The locals at Zhukaigou must have channelled surplus metal ores to the ore-poor production centres at the dynastic core (Wallerstein 1976; Cooter 1977; McGuire and Paynter 1991), and by maintaining close ties, their own status at Zhukaigou was enhanced and economic independence was provided. At Zhukaigou, social hierarchies were marked both by numbers and types of bronze objects in burial (Linduff 1995), and the mid-Shang (ca. 1550–1250 BC) as well as Karasuk items found in the burials of elites there suggest that this was a prestige-good economy in which these rare bronze items marked the
364
KATHERYN M. LINDUFF
special status of their owners (Schortman and Urban 1987; Kristianson 1991; Helms 1992). This was a site where exchange took place, where goods of nonlocal prototypes, whether of Shang or steppe manufacture, were new and special. In a place such as Zhukaigou where power of the leadership was probably measured through the possession of rare goods, stability would have been fragile since status was tied to successful relations with the dynastic centre and/or steppe dwellers or traders. Zhukaigou was abandoned by 1250 BC, for instance, when the Shang must have secured other sources of ores when they moved the capital to Anyang or when hostile groups may have occupied the territory between them and Zhukaigou, making transport of ores impossible. Marriage Alliance. The archaeological site at Anyang, Henan, in the Central Plain (Anyang 1976; 1977; 1986; 1987) was the location of the capital of the late Shang period, 1250–1050 BC (anon. 1994) (Fig. 1). Many artefacts with Siberian prototypes, including bow-shaped objects, curved knives with animal pommels, horse trappings of all sorts, and chariots were discovered there (Li 1980). From inscriptional evidence found at the site, we know that during this period the Shang were frequently in contact with hostile, outsider groups to their north and west and they often boasted of success in war including the capture of chariots, personnel and objects (Keightley 1978). For these reasons, most of the Siberian-type objects discovered there have been explained as war booty collected from these defeated enemies. Analysis of the intact Tomb of Lady Hao excavated in 1984, however suggests to me that other types of relationships also existed between the Shang and their non-Shang neighbours. Fu Hao was a consort of the third king at Anyang and her tomb, dated about 1200 BC, yielded all the emblems of a Shang elite. It was a tomb systematically filled in layers with bronze ritual vessels, weapons, and carved jades, bone and ivory (Chang C.-L. 1986; Chou 1970; K.C. Chang 1980; 1986). In the spaces closest to her body were found scores of frontier style items including curved knives, mirrors and horse trappings, all items with southern Siberian (especially Karasuk) analogies (Fig. 3). These identify her with ‘outsider’ groups described in Shang period inscriptions as their sometimes-hostile northern neighbours. Her marriage to the king suggests that union with her provided a political alliance of strategic importance to the stability of the dynastic frontiers (Linduff 1996; Keightley 1999). Lady Hao’s affiliation with the Shang cannot be doubted for she was buried with full ‘royal’ paraphernalia. However, while living and dying within Shang
20. SIBERIAN ARTEFACTS IN ANCIENT CHINA?
365
Fig. 3. Zhukaigou, Inner Mongolia, Phase V, bronzes (after Kaogu xuebao 1988.3, 325).
political limits, she must have retained identity with her parent culture, one of those outsider groups documented in ancient Chinese histories. Late Neolithic practice among the Yangshao people at Shijia, for instance, in the Central Plain shows that women married out and were returned to their natal kinship group upon death when they were buried with their brothers (Gao et al. 1993; Keightley 2001). In the case of Lady Hao during the Shang, however, increased male dominance in social and ritual spheres did not allow the return of the physical body to her parent culture, but appears to have endorsed identification of the female lineage in other ways at burial. The association is well documented in Lady Hao’s tomb where frontier style markers such as mirrors, curved blade knives and jingle-headed tools and weapons were placed directly next to her body. I have proposed elsewhere that she may have been affiliated with horsebreeding groups whose presence in Shang elite life was so essential to the demonstration of power of the court (Linduff 2002). Other tombs at Anyang that include horse trappings and chariots also document this practice. And even though those buried in large stepped tombs at Anyang have been thought to be Shang elite, the Fu Hao tomb with its ‘Siberian’ style artefacts suggest to me that the notion of cultural homogeneity proposed by ancient historians even in the ‘core’ area of the Shang must be re-evaluated. Regional Alliance Through Military Control and Ethnogenesis. The cemetery at Yuhuangmiao, is in Yanqing county, in the district of Beijing (anon. 1990; 1989;
366
KATHERYN M. LINDUFF
Map 1). This site is well within the borders of the Zhou dynastic state of Yan and can be dated between ca. the 8th and 5th centuries BC. The area was described in ancient Chinese historical accounts as the home of the Shanrong people, thought to be nomadic, and ethnically not Zhou. Excavation of the cemetery in Yanqing county yielded over 280 niche-style tombs with evidence of sacrificed goats, cattle and canines, and large numbers of belt plaques, bronze daggers, and short swords of Tagar type, from southern Siberia, using animal decor. Both the shape and contents of these tombs were unlike their counterparts in Zhou political centres in the Central Plain. These daggers and belt plaques are common to mounted warriors of the steppe, but not the Zhou armies. This cemetery was located well within the political and cultural limits of the victorious Zhou, so why would these tombs so consistently display objects that would identify them with Zhou enemies at their borders? In this frontier zone, as the Chinese histories suggest, local groups may have been employed to bring ‘nuisance’ groups under control in order to shield the core and to stabilise the confederacy. Even though these opponents were not always powerful or threatening, their control was especially important for Zhou efforts to centralise their leadership, control, and power. In most settings in this region, such as in the tombs of the elite in the Yan capital, Zhou style objects dominated funerary display and thereby confirmed local allegiance to the central government (Sun 2001). In the Yuhuangmiao tombs, however, the absence of objects of Zhou style, implies that the deceased saw themselves as a coherent unit of ‘outsiders’ with an established, or perhaps even created, identity that could confirm their separateness and through which they made a place for themselves in the new dynastic organisation. Their residence and burial inside the Yan state of the Zhou confederation confirms neither a mobile nor pastoral lifestyle but suggests that the men may have been mercenaries, a mounted cavalry in the service of the Zhou military, probably hired and pitted against hostile non-Zhou groups in or near this strategically located border state. Such clearly delineated accord exhibited in burial practice at the cemetery at Yanqing, suggests ethnogenesis and if so, it was structured by contemporary political interaction (Brumfiel and Fox 1994).
Discussion From the movement of peoples and technology in the early 2nd millennium into western China exemplified by the excavation of the cemetery and habitation
20. SIBERIAN ARTEFACTS IN ANCIENT CHINA?
367
Fig. 4. Bronzes of the northern type unearthed from the Fu Hao Tomb and their northern counterparts: 1–6, from Fu Hao Tomb; 7, from Yantoucun, Suide, Shaanxi; 8, Inner Mongolia; 9, Naimatai, Guinanxian, Qinghai; 10, Linzheyu, Baode, Shaanxi; 11, Transbaikalia, Russia; 12, Chirnokovo, Krasnoiarsk, Russia (after K.C. Chang 1986, fig. 51).
Fig. 5. Yuhuangmiao, tomb 86. Remains of wooden coffin and animal sacrifice (after Beifang kaogu sishininian 1990, 81, fig. 35).
368
KATHERYN M. LINDUFF
site at Huoshaogou; to the establishment of ‘frontier’ towns such as Zhukaigou in strategic locations in relation to raw materials; to the marriage practice of the dynastic leaders for political purposes as evidenced at the capital of the Shang at Anyang; to the employment of warriors of non-Zhou backgrounds to provide border control of other, potentially hostile groups, the presence of ‘outsider’ artefacts provided visual markers and kept clear the identity of non-dynastic peoples who occupied the frontiers as well as the dynastic core in the 2nd and 1st millennia BC. Their presence was of vital importance to the economic, social and political well being of the dynastic system and a diverse set of arrangements can be distinguished. Likewise, the use of these ‘alien’ artefacts in each context would seem to underscore choices made by their owners to differentiate themselves deliberately as elites or simply as outsiders. That separateness provided the owners with a strategic political, occupational, economic or personal advantage within the context of their society. This is especially clear in the case of the atmosphere of the Zhou since they were consciously and selectively trying to construct and project theirs as a coherent, integrated culture. It is clear from ancient and modern texts that the limits of Chinese culture were drawn according to lifestyle that could be demonstrated by display of certain artefacts. But in the four examples just discussed where such artefacts were displayed, all but Yuhuangmiao were clearly agricultural, or agro-pastoral communities. In the case of Yuhuangmiao, we can only assume that because they buried their men and women inside of Yan state territory over a period of up to 300 years, that even they lived permanently somewhere nearby. Thanks to excavations in China over the past few years we now know that the peoples and their artefacts identified by Teploukov, Griaznov, Rudenko and Kiselev, among others, in southern Siberia were not limited to that area or lifestyle and extended into the heartland of China. The relationship of the peoples who bore such artefacts with the dynastic centres varied according to the specific circumstances of time and place. Some such as the villagers at Huoshaogou adopted new technologies with origins far from their own communities and probably unwittingly became transmitters of very special knowledge about metallurgy that eventually transformed life at the dynastic centres in the Yellow River basin. Others such as the leaders at Zhukaigou used these artefacts to promote and maintain high status for themselves. And, either by their own design and/or that of the Chinese who saw them as ‘outsiders,’ many such as Fu Hao as well as the warriors at Yuhuangmiao did not forsake their actual or
20. SIBERIAN ARTEFACTS IN ANCIENT CHINA?
369
adopted roots and carried these signal artefacts to death even within a ‘foreign’ land. Technology exchange, surplus extraction, marriage alliance and ethnogenesis are, then, some of the reasons I think that there are Siberian artefacts inside of ancient Chinese dynastic borders.
Bibliography Anon. 1989: Beijingshi wenwu yanjiushuo Shanrong wenhua kaogudui. Beijing Yanqing Jundushan Dongzhou Shanrong buluo mudi fajue jilue. Wenwu 8, 17–35. Anon. 1990: Wenwu Kaogu gongzuo sanshinian (Beijing), 142–3. Anon. 1994: Yinxu de faxian yu yanjiu (Archaeological Excavation and Researches of the Yin Ruins) (Beijing). Anyang City Museum 1986: Anyang Tiexi Liujiazhuang nan Yindai muzang fajue jianbao (Excavation of Yin Dynasty Tomb in South Liujiazhuang, Tiexi at Anyang). Zhongyuan wenwu 3, 14–23. Anyang Excavation Team, Institute of Archaeology, CASS 1976: 1975 nian Anyang Yinxu de xinfaxian (New Discoveries at Yinxu, Anyang). Kaogu 4, 264. Anyang Work Team, Institute of Archaeology, CASS 1977: Anyang Yinxu nuli jisikeng de fajue (Excavation of the Sacrificial Pits of Slaves at Yin Ruins in Anyang). Kaogu 1, 20–36. ——. 1987: Yinxu 259, 260 haomu fajue baogao (The Excavation of Tombs M259 and M260 at the Yin Ruins). Kaogu xuebao 1, 99–118. Barnard, N. 1987: Bronze Casting Technology in the Peripheral ‘Barbarian’ Regions. Bulletin of the Metals Museum 12, 3–37. ——. 1993: Thoughts on the Emergence of Metallurgy in Pre-Shang and Early Shang China, and a Technical Appraisal of Relevant Bronze Artefacts of the Time. Bulletin of the Metals Museum 19, 3–48. Brumfiel, E.M. and Fox, J. (eds.) 1994: Factional Competition and Political Development in the New World (Cambridge). Bunker, E.C., Chatwin, C.B. and Farkas, A.R. 1970: ‘Animal Style Art’ from East to West (New York). Bunker, E.C., Kawami, T., Linduff, K.M. and Wu En 1997: Ancient Bronzes of the Eastern Eurasian Steppes from the Arthur M. Sackler Collections (New York). Chang, K.C. 1980: Shang Civilization (New Haven). —— (ed.) 1986: Studies in Shang Archaeology (New Haven). Chang Cheng-lang 1986: A Brief Discussion of Fu Tzu. In Chang, K.C. 1986, 103–19. Chou Hung-hsiang 1970: Fu-X Ladies of the Shang Dynasty. Monumenta Serica 29.1, 346–90. Cooter, S.W. 1977: Preindustrial Frontiers and Interaction Spheres: Prolegomenon to a Study of Roman Frontier Regions. In Miller, D. and Stefan, J. (eds.), The Frontier, Comparative Studies (Norman), 81–107. Fitzgerald-Huber, L. 1995: Qijia and Erlitou, The Question of Contacts with Distant Cultures (John King Fairbank Center for East Asian Research 6) (Cambridge, Mass.). Gao Qiang and Yun Kuen Lee 1993: A Biological Perspective on Yangshao Kinship. Journal of Anthropological Archaeology 12, 266–89. Griaznov, M. 1969: Ancient Civilization of Southern Siberia (Geneva). Helms, M. 1992: Long Distance Contacts, Elite Aspirations and the Age of Discovery in a Cosmological Context. In Schortman, E.M. and Urban, P.A. (eds.), Resources, Power, and Interregional Interaction (New York), 157–74. Institute of Archaeology, CASS 1980: Yinxu Fuhaomu (Tomb of Lady Hao at Yinxu in Anyang) (Beijing). Karlgren, B. 1945: Some Weapons and Tools of the Yin Dynasty. Bulletin of the Museum of Far Eastern Antiquities XCII, 101–44.
370
KATHERYN M. LINDUFF
Keightley, D.N. 1978: Sources of Shang History (Berkeley/Los Angeles). ——. 1999: Women in the late Neolithic and Shang Period in China. Nan nü 1.1, 1–63. Kristianson, K. 1991: Chiefdom, State, and System of Social Evolution. In Earle, T.K. (ed.), Chiefdoms: Power, Economy and Ideology (Cambridge), 16–43. Lattimore, O. 1942: Inner Asian Frontiers of China (Oxford). Li, J.J. 2000: Mapping Artefacts of the Frontier: An Approach to the Study of the Yan Mountainous Area in the Eastern Zhou Period (8th–3rd Century BC) (PhD Dissertation, University of Pittsburgh). Linduff, K.M. 1995: Zhukaigou, Steppe Culture and the Rise of Chinese Civilization. Antiquity 69, 133–45. ——. 1996: Art and Identity: The Chinese and Their ‘Significant Others’ in the Shang. In Gervers, M. and Schlepp, W. (eds.), Cultural Contact, History and Ethnicity in Inner Asia (Toronto Studies in Central and Inner Asia No. 2) (Toronto), 12–48. ——. 1998: The Emergence and Demise of Bronze-using Cultures Outside the Central Plain in Ancient China. In Mair, V. (ed.), The Bronze Age and Early Iron Age Peoples of Eastern Central Asia (Washington, DC), 619–43. ——. 2000: The Emergence of Metallurgy in China (Lewiston/Queenston/Lampeter). ——. 2002: A Walk on the Wild Side: Horses and Horse Gear in Shang China. In Renfrew, C. (ed.), Late Prehistoric Exploitation of the Eurasian Steppe (Cambridge). ——. 2002–04: Early Complex Societies in NE China: The Chifeng International Collaborative Archaeological Research Project. Journal of Field Archaeology 29.1–2, 45–73. Liu Li 1996: Settlement Patterns, Chiefdom Variability, and the Development of Early States in North China. Journal of Anthropological Archaeology 15, 237–88. Loehr, M. 1949a: Weapons and Tools from Anyang and Siberian Analogies. American Journal of Archaeology LIII, 126–44. ——. 1949b: Ordos Daggers and Knives: New Material Classification and Chronology. Artibus Asiae XII, 23–83. McGuire, R.H. and Paynter, R. 1991: The Archaeology of Inequality (Cambridge, Mass.). Mei, J.J. 1999: Copper and Bronze Metallurgy in Late Prehistoric Xinjiang: Its cultural context and relationship with neighboring regions (BAR International Series 865) (Oxford). NZWKYOB 2000: Zhukaigou: Qingtong Shidai Zaoqi Fajue Baogao (Zhukaigou: Excavation Report of the Early Bronze Age Site) (Beijing). Schortman, E.M. and Urban, P.A. 1987: Modeling Interregional Interaction in Prehistory. In Schiffer, M.B. (ed.), Advances in Archaeological Method and Theory vol. 11 (San Diego), 37–95. Sun Yan 2001: A Study of Classification of Material Culture and Mortuary Practices of the Yan in the Early Western Zhou Period (PhD Dissertation, University of Pittsburgh). Underhill, A.P., Feinman, G.M., Nicholas, L., Bennett, G., Fengshu Cai, Haihuang Yu, Fengshi Luan and Hui Fang 1998: Systematic, Regional Survey in SE Shandong Province, China. Journal of Field Archaeology 25, 453–74. Wallerstein, I. 1976: A World-System Perspective on the Social Sciences. British Journal of Sociology 27.3, 343–52. Whittaker, C.R. 1994: Roman Empire: A Social and Economic Study (Baltimore). Xu Yongjie 1989: The Siba Culture Site at Huoshaogou, Donghuishan, Minle County. In The Almanac of Chinese Archaeology (1988) (Beijing).
CHAPTER TWENTY-ONE SENSING THE SHADOW EMPIRE: NEW APPROACHES TO THE PROBLEM OF THE KHAZAR STATE IRINA R. SHINGIRAY
In this paper I would like to suggest some new approaches to the conceptualisation of the origins and political organisation of the Khazar empire, viewed primarily from an anthropological perspective. While an anthropological approach is rarely applied to Khazar archaeology, certain anthropological models may greatly aid the archaeological process and help us reach a better understanding of the Khazar state, which was largely influenced by the outside world (Khazanov 1994). The approach I intend to use is based on the work of Thomas Barfield (1989; 1990; 2001), which focuses on nomadic shadow empires and their relations with sedentary states. Using this approach, I emphasise the importance of political, military and economic relations between the nomadic Khazar empire and the Islamic caliphate from the 7th to the 10th centuries AD. These relations had a profound impact on the political organisation and economic stability of the Khazar empire, and deserve greater attention. In addition, I would like to discuss the implications of this approach for future archaeological investigations relevant to the Khazar problem. Locating, mapping and studying the Khazar sites in the nomadic core territory of this empire is necessary in order to understand the nature of this polity, which was primarily a military state, not a commercial one, as some have suggested (see, for example, Golden 1980; Di Cosmo 1999). The analysis of the material culture of the Khazar nomads, who represented the core of the political and military power of this empire, will afford a better understanding of the internal social relations among the Khazar confederate tribes, as well as their external political and economic relations with the Islamic caliphate. Finally, I will suggest several methods for locating nomadic Khazar sites in the core territory of their empire based on applications of satellite remote sensing techniques. The nomadic Khazar empire is not well known. Therefore, a brief historical sketch regarding the development and demise of this empire is necessary in order to ground the discussion of the anthropological and archaeological approaches that can be applied to it.
372
IRINA R. SHINGIRAY
Historical Background The Khazar state was a confederacy of nomadic Turkic tribes, which emerged in the north-western Caspian region in the 7th century AD after the disintegration of the extensive Western Turkish empire.1 After the collapse of the latter, the geopolitical situation of the western Eurasian steppe was complicated by the emergence of two Turkish successor powers within the Pontic-Caspian region. One was the Khazar state, which occupied the arid steppe of the north-western Caspian region, and was therefore heavily dependent upon resources from the fertile lands of the Transcaucasus. The second power was Magna Bulgaria, which was formed in the prosperous north-eastern Pontic region and had lucrative trade and political alliances with Byzantium. Initially, Magna Bulgaria overshadowed the Khazars politically. However, in the middle of the 7th century AD the Khazars began their expansion by conquering Bulgaria and subjugating most of its tribes, thus becoming the predominant political power in the region. Simultaneously, the Khazars entered into a state of war with the army of the expanding Umayyad caliphate, a conflict that lasted until the middle of the 8th century AD. As a result, the Transcaucasian ‘breadbasket’ of the Khazars became a part of the caliphate and ceased to be in the Khazar sphere of influence. Despite losing the Transcaucasus, the Khazars not only managed to preserve their hegemony within the Pontic-Caspian steppe, but were able to strengthen and expand their political and military power throughout most of the western Eurasian steppe and beyond. The Khazar imperial army, consisting of a large number of heavily armoured horsemen, represented a formidable military force, which enabled the Khazars to manipulate various sedentary groups politically. According to some contemporary sources, the Khazars controlled as many as 25 different nations (Artamonov 1962, 191). During its existence, the Khazar empire had a major influence on multiple sedentary and semi-sedentary populations of Eastern Europe, the Caucasus and those on the frontier with the Islamic lands, transforming the political, economic and cultural make-up of these regions. 1
The Western Turkish (Kök) empire, which controlled the Pontic-Caspian region prior to the Khazars, was itself an offshoot of the larger (previously unified) Turkish empire. The Turkish empire controlled most of the Eurasian Steppe from the Yellow to the Black Seas, and promoted the silk trade from China to Sassanian Iran and Byzantium during the 6th and the beginning of the 7th century AD.
21. NEW APPROACHES TO THE PROBLEM OF THE KHAZAR STATE
373
When the Abbasid dynasty replaced that of the Umayyads around AD 750, the caliphate’s policy toward the Khazars changed dramatically. Instead of conflict, the Khazars were presented with new possibilities of trade with the caliphate, which grew economically powerful during the early Abbasid period.2 The Khazars were engaged in the lucrative trade of slaves, furs and other luxury goods, which were acquired in the steppe or in the forest region to the north and re-exported to the Abbasid caliphate. As a result of this trade, large quantities of Abbasid silver dirhams found their way into the steppe, north-western Eastern Europe and Scandinavia.3 The Khazars controlled an important trade network, at the nexus of which they founded their capital Itil on the Lower Volga, in a region where no previous urban settlements had existed (Fig. 1). Itil became one of the largest trade emporiums on the Caspian Sea during that time (Golden 1980, 15). The prosperity of the Khazar empire lasted until the middle of the 10th century AD, after which time the empire rapidly disintegrated. The reasons for the collapse of the Khazar empire are disputed, but they are attributed either to the expansion of Kievan Rus; to assaults by Pechenegs or Ghuz nomads; or to economic crises. A realignment of military and political powers on the steppe followed the collapse of the Khazars, and subsequently the Seljuks assumed political hegemony in that part of the Eurasian steppe in the 11th century AD.
Archaeological Background and the Khazar Problem In the course of the 20th century, Russian and Soviet archaeologists uncovered a large amount of archaeological data from the Khazar period in the western Eurasian steppe. Most of these excavations were the result of salvage work from settlements in the Don and Donets river regions, the Crimea and the northern Caucasus (see Pletneva 1999). Typological studies of these remains suggested a relatively homogenous material culture, which coincided spatially with the extent of the Khazar empire, and was dated chronologically from the end of the 8th to the middle of the 10th century AD. This material culture was 2
As the Islamic world-economy expanded during the so-called Golden Age of Islam, the Khazar empire became linked to that economic system (see Lombard 1975). 3 Recent studies in historical numismatics by Thomas Noonan (1998) present strong evidence to support these Khazar-Islamic caliphate trade relations and exchange as a main source of the Khazars’ economic power.
Fig. 1. The Khazar empire and the Islamic caliphate (ca. 900).
374 IRINA R. SHINGIRAY
21. NEW APPROACHES TO THE PROBLEM OF THE KHAZAR STATE
375
termed the Saltovo-Maiatskaia culture (the SMC), and ascribed to the Khazar empire. Svetlana Pletneva (1999) divided the SMC into five ‘geo-ethnic’ variants, corresponding to various sedentary, semi-sedentary and nomadic tribes (such as Alans, Burtas, Bulgars and Pechenegs) occupying different geographic regions. She also suggested a sixth hypothetical variant corresponding to the geographic core territory of the empire as known from the historical sources, namely the region between the Lower Volga and northern Daghestan. The material culture of the latter variant, according to the author, most likely belonged to the Khazars proper, but this region was (and is) the least investigated, and the Khazar material culture from the core of that empire is virtually unknown. In fact, the material culture from the western margins of the Khazar empire (which was better studied and served as a basis of knowledge about the Khazars) did not necessarily belong to the members of the Khazar confederacy, but instead to their subject tribes and tributaries. Pletneva’s theory of ‘geo-ethnic regionalisation’ spurred contentious debates among archaeologists, who tried to equate her ‘geo-ethnic’ variants with specific ethnic groups. Although some strongly criticised the ethnocentric approach to the interpretation of the SMC material culture,4 the focus of the Khazar problem in archaeology continues to lie largely in the domain of ethnicity. Most archaeologists and historians engaged in this debate, however, do not attempt to approach the problem of ethnicity in the Khazar society from the perspective of the Khazar political organisation, even though such an approach seems to be more appropriate for the Khazar case. These archaeologists are still reluctant to deviate from the scheme of unilinear evolutionary development of the nomadic societies, as articulated by Pletneva (1967), and therefore, the problems concerning the ethnicity and political organisation of the imperial Khazar state are far from being settled. Pletneva’s scheme of the internal socio-economic development of Khazar society, termed ‘From the Pastures to the Cities’, presupposes a gradual evolution from a nomadic society, to a semi-nomadic one, to a sedentary one, to the emergence of urbanism and a ‘feudal’ society. In short, archaeologists tend to approach the development of nomads from the same perspective as they approach sedentary societies. 4
For instance, Werbart (1996, 217) suggested that the Khazar empire should be perceived as a multi-ethnic, multi-cultural, multi-lingual conglomerate with a ‘pluralism of the social structures and economy’, something that, according to her, cannot be inferred from the archaeological material culture of the SMC.
376
IRINA R. SHINGIRAY
Anatoly Khazanov (1994) has remarked on the difficulties of conceptualising a nomadic empire, such as that of the Khazars, within the framework of unilinear socio-economic evolution. It is difficult to explain the emergence of nomadic empires on the basis of their socio-economic prerequisites, namely a rudimentary pastoral economy and virtually classless society with a minimal labour division and simple relations of production. Instead, Khazanov (1994) argued that there are external influences that brought economic, socio-political and cultural changes into nomadic societies. In fact, the nature of the nomads’ relations with the outside world determined the level of socio-political development, political integration and centralisation of power in their societies. Unfortunately, many useful concepts and implications based on anthropological research are often ignored or underestimated by Khazar scholars.5
Anthropological Approach: The Khazar State as a Shadow Empire Anthropological research shows that it is strong external political influences – such as imperial expansion by large sedentary states – that often compelled nomads to aggregate into large tribal confederations for military purposes. And since their non-autarkic economy could not sustain those large supra-tribal units and the state apparatus, the longevity and size of these nomadic polities depended on their success in linking themselves economically to primary sedentary states via trade, tribute, extortion, or raids. Barfield (1989) has demonstrated this on the basis of multiple historical cases from the Chinese/ nomadic frontier in eastern Eurasia. He found that nomadic empires often emerged, interacted and collapsed together with the primary sedentary empires. He termed them ‘shadow empires’, as they were secondary phenomena in response to the influences of the sedentary imperial power, and were economically and politically synchronous with the latter. In my view, the nomadic Khazar empire is a good example of such a shadow empire, which emerged, developed and collapsed synchronously with the formation, expansion and sub5
Some historians and archaeologists, including Pletneva, realise the importance of external influences on the Khazar empire. There is, after all, a great deal of historical and archaeological evidence for Khazar ‘international’ warfare or the important long-distance trade in luxury goods, such as slaves and furs, which came from the lands of the Khazars. However, most of these authorities consider external relations and trade to be of secondary importance in the development of the Khazar empire, thus underestimating the significance of these factors.
21. NEW APPROACHES TO THE PROBLEM OF THE KHAZAR STATE
377
sequent regionalisation (namely the loss of political centralisation) of the Islamic caliphate, on whose resources the Khazars depended. Nomadic shadow empires possessed a centralised military force and a threetier state apparatus, which was structurally linked to the sedentary state (see Barfield 2001). In Barfield’s words (1990, 167), the state structure ‘had its roots in foreign relations’, namely the manipulation of the sedentary powers, ‘not in the evolution of social organisation on the steppe itself’. The political system of such an imperial nomadic confederation was based on a mechanism of the redistribution of the prestige goods and benefits among the confederate tribes. State-organised nomads, however, were much more efficient at extorting wealth than producing it – exploitation and taxation of scattered nomadic households did not make sense.6 Therefore, the wealth for maintaining a nomadic empire came almost exclusively from the outside world, and in the case of the Khazars, mainly from the lands of the caliphate. In order to extract that wealth, the Khazars’ strategies included military raids and political manipulations, such as extortion, tribute, marriage alliances, trade agreements and taxation. These imperial economic activities that related to the state’s ‘foreign affairs’ were a prerogative of the ruling stratum of the nomadic society, and were central to their power (Barfield 2001, 14; 1981). War and trade were complementary components of the Khazar imperial policy directed at acquisition of resources, not at conquering the lands of the sedentary societies. Other Khazar economic activities included border trade; the extraction and re-export of luxury goods; and the taxation of foreign merchants and various nonnomadic tribes subjugated by them. After a century of warfare with the caliphate, the Khazars were presented with possibilities for trade in luxuries (slaves, pelts and exotics for precious metals and textiles) by the Abbasids – we then witness the impact of Khazar imperialism. It resulted in the economic and political transformation of various tribal people in the Khazar sphere of influence, and the mobilisation of resources under Khazar control. It becomes apparent that the Khazar state was not a commercial state that mainly benefited from transit trade and economic demands of primary sedentary powers. Instead, it was formed as a military state, where the military source of power was instrumental in the formation of this empire, its politics of trade and war, and its gain of authority among other political powers of the region. The Khazar
6
It is known that the Khazar nomads were not taxed by their state; instead, they had to enter into imperial military service when called upon to do so (Dunlop 1954, 232).
378
IRINA R. SHINGIRAY
empire had a great demand for resources and it used its military authority to acquire them. Because the Khazars depended on luxury goods and silver from the caliphate, the political and economic stability of the Khazar empire was largely determined by the political-economic conditions in the caliphate. Consequently, the major economic and political crisis in the Abbasid caliphate in the 10th century AD (the so-called ‘silver crisis’) most likely brought about the demise of the Khazar imperial confederation – the shadow empire of the caliphate.7
Archaeological Implications Chronologically, the above scenario is reflected in the development of the SMC material culture. Already toward the end of the 8th century AD, the SMC acquires certain homogeneity of its material culture across the western Eurasian steppe. This convergence becomes more and more apparent toward the 10th century AD, attesting to the period of increased centralisation and imperialism, after which it suddenly vanishes in the middle of the 10th century. More implications, however, can be made from the proposed approach of conceptualising the Khazar empire as a shadow empire of the caliphate. At the political core of the Khazar empire, the imperial nomads did not settle down and adopt crafts and urbanism, as used to be believed (Pletneva 1967; 1999). Instead, they continued their seasonal movements and practice of war and trade, which they monopolised at the higher level of political organisation. In fact, the seasonal movements of the imperial nomads were combined with the functions of military and political surveillance of imperial trade routes and territories, pacification of local inhabitants, and distribution of gifts and luxury goods to the tribal leaders. The Khazars, for instance, moved annually from their base in Itil to northern Daghestan, to the Don river, and back to Itil (Artamonov 1962, 387). Nomads also benefited from pillage and raids into the territories of sedentary states, border trade, and the accommodation of long-distance trade, while some of them became merchants themselves. The nomads were attracted to the routes, either for war or for trade, in order to participate in the
7
The increased attacks on the Khazars by the Rus, Ghuz, and Pechenegs toward the middle of the 10th century AD probably were provoked by the economic and, thus, political weakness of the Khazar empire, which was primarily induced by the silver crisis.
21. NEW APPROACHES TO THE PROBLEM OF THE KHAZAR STATE
379
exchange of goods, to tax and safeguard foreign merchants, provide them with transport, water and food supplies, and to maintain rest facilities. As a priority, the archaeological material culture in the core territory of the Khazar empire should be studied. Most likely, only in the core area we will be able to differentiate between the two main types of material culture: that belonging to the members of the Khazar imperial confederacy, perhaps even of the ruling elite; and that belonging to the tributaries and commoners based on their social status. Studying the geographic distribution and the nature of these different types of material culture should also provide us with more clues regarding the organisation of the Khazar society based on the location of areas of activity and other sites in the geographic and political landscape of that region. In order to be able to implement the above proposals, the range of archaeological investigations has to include Khazar nomadic sites, and not only urban sites favoured by researchers. In other words, regional surveys should be carried out in the core territory, which can produce data relevant to the diverse types and distribution of the Khazar sites. In order to assess the impact of the Khazar-Islamic caliphate relations on the Khazar empire and society, one must locate and study the sites related to directional trade and warfare between the two empires. Archaeological evidence along the trade routes, especially in the core territory of the Khazar empire, should be most informative for such a study.8 Special attention should be paid to the distribution of imports from the caliphate within the Khazar empire. An analysis based on these data can point to specific trends of trade and indicate areas composed of different social groups in Khazar society, and their activities. If possible, Khazar exports to the caliphate and the influence of the Khazar material culture on the latter should be similarly investigated.
Remote Sensing Approach In the context of the archaeological tasks suggested above, it makes sense to address efficient and cost-effective ways to initiate a survey of the Khazar sites within their core territory. This core territory in the north-western Caspian region is topographically flat, arid and devoid of any tall vegetation, which makes a remote sensing approach highly effective (Tsutskin 1987). Analysis of remotely 8
For a more detailed discussion concerning the trade routes in that territory, see Prokopenko 1999.
380
IRINA R. SHINGIRAY
sensed imagery to determine areas with the highest potential to contain archaeological sites has been widely used in the field of archaeology. A methodology for applying this approach to the region under discussion is currently being developed. The areas of special attention include: 1) sand dunes (especially where they are intersected by the routes), since they serve as fresh water indicators, or dune lakes; 2) extra-vegetated and/or salinised areas around artesian wells; 3) reed-growths, which provided shelter for pastoralists and their animals in the open steppe, and were often utilised as building material and fuel; and 4) crossroads and burial mounds. Reeds, dunes, artesian wells, mounds and crossroads are easily identifiable on remotely sensed optical and radar imagery. Their locations, size and distributions provide important information to a field archaeologist engaged in a surface survey of archaeological sites in this type of terrain. Moreover, a remote sensing approach aids navigation in such a largely featureless terrain (Holcomb 2001).
Conclusions In conclusion, I would like to stress once again my thesis that the imperial confederation of the Khazars can be best conceived of as a shadow empire, which emerged in response to the formation and expansion of the Islamic caliphate, and then collapsed once the caliphate’s central control declined and the economic link between the two empires via the Caucasus was destabilised. The political organisation of the Khazar empire had its roots in Khazar/caliphate relations, rather than in the internal socio-economic evolution of the Khazar society. Lastly, in order to gain a better understanding of the operating social networks within the Khazar empire, a regional survey is needed in its core territory to locate nomadic Khazar sites and study their material culture. This is the only way to test the proposed approach archaeologically and evaluate the hypothesis of the Khazar state as a shadow empire of the Islamic caliphate.
Bibliography Artamonov, M. 1962: Istoriia Khazar (Leningrad). Barfield, T.J. 1981: The Hsiung-nu Imperial Confederacy: Organization and Foreign Policy. Journal of Asian Studies 41.1, 45–61. ——. 1989: The Perilous Frontier: Nomadic Empires and China (Cambridge). ——. 1990: Tribe and State Relations: The Inner Asian Perspective. In Khoury, P. and Kostiner, J. (eds.), Tribes and State Formation in the Middle East (Berkeley), 153–82.
21. NEW APPROACHES TO THE PROBLEM OF THE KHAZAR STATE
381
——. 2001: The Shadow Empires: Imperial State Formation along the Chinese-Nomad Frontier. In Alcock, S.E., D’Altroy, T.N., Morrison, K.D. and Sinopoli, C.M. (eds.), Empires: Perspectives from Archaeology and History (Cambridge), 10–41. Di Cosmo, N. 1999: State Formation and Periodization in Inner Asian History. Journal of World History 10, 1–40. Dunlop, D.M. 1954: The History of the Jewish Khazars (Princeton). Golden, P.B. 1980: Khazar Studies: An Historico-Philological Inquiry into the Origins of the Khazars (Budapest). Holcomb, D.W. 2001: Imaging Radar and Archaeological Survey: An Example from the Gobi Desert of Southern Mongolia. Journal of Field Archaeology 28.1–2, 131–41. Khazanov, A.M. 1994: Nomads and the Outside World (Madison, WI). Lombard, M. 1975: The Golden Age of Islam (Amsterdam). Noonan, T.S. 1998: The Islamic World, Rusia, and the Vikings, 750–900: The Numismatic Evidence (Aldershot). Pletneva, S.A. 1967: Ot kochevii k gorodam: Saltovo-Maiatskaia kul’tura (Moscow). ——. 1999: Ocherki Khazarskoi arkheologii (Moscow). Prokopenko, I.A. 1999: Istoriia Severokavkazskikh Torgovykh Putey IV v. do n.e.–XI v. n.e. (Stavropol). Tsutskin, E.V. 1987: Nekotorye napravleniia kosmicheskoi arkheologii. In Tsutskin, E.V. (ed.), Arkheologicheskie issledovaniia Kalmykii (Elista), 114–33. Werbart, B. 1996: Khazars or ‘Saltovo-Majaki Culture’?: Prejudices about Archaeology and Ethnicity. Current Swedish Archaeology 4, 199–221.
CHAPTER TWENTY-TWO RECENT XERORADIOGRAPHIC ANALYSIS OF KURA-ARAXES CERAMICS MARYFRAN HEINSCH AND PAMELA VANDIVER
The Kura-Araxes horizon (3500–2500 BC) is the largest archaeological horizon for either the Near East or Europe, encompassing a large geographic region that includes portions of northern Iran, eastern Turkey and much of the Caucasus (Fig. 1). This horizon is predominantly defined by a single ceramic type that is based on morphological and decorative criteria. While morphological and decorative characteristics are no doubt important indicators of socially conditioned practices of production, so too are the forming processes for raising the walls of these vessels, though these processes have yet to be examined for KuraAraxes type wares. The results of a small pilot project involving the xeroradiography of Kura-Araxes type wares and other contemporary ceramic types found together at sites in Daghestan and Azerbaijan indicate that there is greater variability and diversity in the production practices underlying these wares than previously was described. The observed contrasts in manufacturing techniques revealed in xeroradiographic images raise new sets of research questions with interesting implications for the development of ceramic technologies in the Caucasus as well as for social interactions related to ceramic production and consumption during the Kura-Araxes period. Previous interpretations of Kura-Araxes horizon sites have frequently attempted to make sense of distinct regional variations in decoration and morphology within the Kura-Araxes ceramic type (Khanzadian 1967; Dzhaparidze 1980; Kushnareva and Chubinishvilli 1971). Many of these interpretations relied on competing migration theories for peoples to whom the manufacture of these wares was attributed. Less sophisticated variants of Kura-Araxes pottery were alternatively taken to indicate the earliest point of cultural origin (Kushnareva 1997; Burney 1977) or the degraded remnants of a pottery style that had migrated a great distance from its homeland (Sagona 1984). More recently, both Bochkarev (1982) and Ismailov (1983) have indicated, based on these distinct regional stylistic and morphological variations, that the
Fig. 1. Map of the Kura-Araxes horizon (modified after Kushnareva 1997).
22. XERORADIOGRAPHIC ANALYSIS OF KURA-ARAXES CERAMICS
383
384
MARYFRAN HEINSCH AND PAMELA VANDIVER
Kura-Araxes horizon may best be described as a community of multiple cultures or, ‘block of cultures’ that are closely related in some unspecified manner. The organisation of these cultures and the relations between them is ambiguous. Further, the relationships between Kura-Araxes type wares and other contemporary wares at the same sites have not been investigated thoroughly and generally have not been taken into account when describing regional variation in ceramic assemblages. This interpretive perspective raises two questions. First, what is the nature and extent of the regional variation in Kura-Araxes horizon pottery technologies and how can this best be described? Second, how does this variation relate to patterns of social relationships within and between communities within this hypothesised ‘block of cultures’? A useful framework for an investigation into these questions may be derived from perspectives within the anthropology of technology that are concerned with the implications of interaction between people and the objects they create (cf. Dobres 2000; Ingold 1993; Pfaffenberger 1992). In particular, a socio-technical approach examines these objects in terms of the social practices that both produce them and are in part shaped by them. As such, this approach relies predominantly on chaîne operatoire (Leroi-Gourhan 1946; Cresswell 1976, 6; Lemonnier 1992, 6; 1986, 149) analysis in combination with theories of practice (see Ortner 1984) to describe changing patterns of technological activity within fluctuating social contexts (Dobres and Hoffman 1994). Chaîne operatoire analysis focuses on reconstructing sequences of techniques involved in both the production and consumption of objects. Each technique is examined as a ‘total social fact’ (Mauss 1979 [1936]), reflecting the integration of socially conditioned behaviours with material resources. The interpretation of this data through theories of practice (Bourdieu 1977; Giddens 1984) describe mechanisms for generating variation in techniques of ceramic production and consumption by situating individual agency within a broader scope of community practice. This combination provides a link between variability in Kura-Araxes material culture and variability in social context and organisation, though it requires a more detailed examination of social practices underlying, in this case, ceramic technologies, than previous analyses and interpretations have offered. The first question regarding the nature and extent of variation in Kura-Araxes horizon ceramics may thus be addressed through an examination of their chaînes operatoire to identify variability in the techniques employed in their manufacture and consumption beyond the already apparent morphological and decorative
22. XERORADIOGRAPHIC ANALYSIS OF KURA-ARAXES CERAMICS
385
divergences. The second question regarding the relationships between ceramic variability and social relationships may then be addressed through the articulation of techniques with theories of practice to describe patterns of social relationships relating to technological activity within and between communities. To begin with then, it is essential to describe as fully as possible, the techniques required in producing and using Kura-Araxes wares. This requires multiple forms of analysis, though one form of analysis, xeroradiography, offers a convenient starting point to address the production aspects of pottery. Xeroradiography provides information regarding the assembly of vessel walls for most types of pottery. It involves the exposure of an electrostatically charged selenium plate to produce images that have far greater edge enhancement than those images produced by conventional radiography. This edge enhancement makes visible variations in the amount, shape and alignment of pores in the ceramic fabric that relate to specific differences in production techniques (Vandiver 1988, 145). The amount of these pores can indicate the extent of the forming pressure applied to the clay, a measure of which corresponds to forming technique (Vandiver 1988, 142). The shape and orientation of these pores further indicate the forming methods used. For instance, a coiling technique would produce horizontal rows of pores at varying intervals throughout the ceramic body (Vandiver 1988, 143). A pot thrown on a wheel would exhibit a pattern of elongated pores oriented at about a 30- to 45-degree angle as a result of upward pressure applied to clay moving across a horizontal centripetal axis (Vandiver 1988, 143) (Fig. 2). Xeroradiography also reveals variation in the relative size and amount of temper. Qualitative sorting and coding of these variations provides a method for selecting a smaller sample from large collections of sherds for more labour-intensive analyses such as petrography of inclusions, or instrumental neutron activation analysis, or determination of firing temperature ranges. A sample of 20 specimens from sites in Daghestan and Azerbaijan was xeroradiographed to refine future research methodology. The specimens selected represent two roughly coeval ceramic types that are represented at Kura-Araxes horizon sites in both regions. Relationships between these two types regarding their production and use have never been examined. Additionally, the KuraAraxes type wares selected for this study represent two distinct regional variants of this type. Again, the possible relationships between these two variants have not been closely examined. By selecting specimens across types and typological variants we hoped to observe differences in their chaînes operatoires that contribute to these distinctions.
386
MARYFRAN HEINSCH AND PAMELA VANDIVER
ORIENTATION OF PORES: PARALLEL TO THE WALL rows of horizontal pores – coiling fairly even distribution of pores in a consistent diagonal direction – throwing pores distributed around discrete blocks of clayey material – building in slabs PERPENDICULAR TO WALL IN VERTICAL OF HORIZONTAL DIRECTION SECTION even distribution – throwing or paddle and anvil
butt join – coiling or slab construction
bevel join – coiling or slab construction
Fig. 2. Orientation of pores (modified after Vandiver 1988, fig. 3).
The 23 samples of Kura-Araxes type wares in the sample, like all KuraAraxes type wares, are burnished and are hand made, as opposed to having been thrown on a wheel (Munchaev 1975). While the Kura-Araxes type includes specimens that are decorated with incised or modelled motifs (Munchaev 1975), those included in the sample lacked ornamentation. Also, while some Kura-Araxes wares are black or red in colour (Kushnareva 1997), those included in the sample ranged from grey to brown to orange. Evidence of a ‘lug’ handle often considered diagnostic of this ceramic type was present on two of the samples for this study. The other ware type included in the sample is known as ‘High Quality Ware’ (HQW) (Gadzhiev et al. 2000). Like Kura-Araxes ware, HQW is similarly burnished, and typically displays patterns of tight rocker bar impressions or incised zigzags around the shoulder of the vessel body. HQW is found associated with Kura-Araxes type ceramics in domestic features at Velikent in Daghestan (Gadzhiev et al. 1998, 198). In contrast to the widespread distribution of Kura-Araxes type wares, the distribution of HQW appears limited to the Caspian littoral and adjacent foothills from Azerbaijan through Daghestan and into Chechnya (Munchaev 1975; Gadzhiev et al. 2000).
22. XERORADIOGRAPHIC ANALYSIS OF KURA-ARAXES CERAMICS
387
Inclusion size range
Pore size Range
Ceramic body thickness range
Kura-Araxes Textile-Formed
N/A
