CURRENT ARCHAEOLOGICAL RESEARCH IN MONGOLIA
936490_BCAA_Mongolei_06.indd 1
29.04.10 10:01
Bonn Contributions to Asian Archaeology Volume 4
CURRENT ARCHAEOLOGICAL RESEARCH IN MONGOLIA Papers from the First International Conference on “Archaeological Research in Mongolia” held in Ulaanbaatar, August 19th–23rd, 2007
Edited by Jan Bemmann, Hermann Parzinger, Ernst Pohl, Damdinsüren Tseveendorzh Edited by Jan Bemmann
2009 Vor- und Frühgeschichtliche Archäologie Rheinische Friedrich-Wilhelms-Universität Bonn
936490_BCAA_Mongolei_06.indd 2-3
29.04.10 10:01
616 pages, 361 figures, 21 tables
CONTENTS PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 INTRODUCTION Damdinsüren Tseveendorzh Current Archaeological Studies in Mongolia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
The conference and the publication were generously financed by Gerda Henkel Stiftung
STONE AGE Biambaa Günchinsüren Introduction into Stone Age of Mongolia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Damdinsüren Tseveendorzh Mongolanthropus was discovered in Mongolia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Ein Titelsatz ist bei der Deutschen Bibliothek erhältlich (http://www.ddb.de)
Masami Izuho, Batmunkh Tsogtbaatar, Takuya Yamaoka, Kazuhiro Hayashi, Altangerel Enkhtör Preliminary Report of Geoarchaeological Investigation at Khanzat-1, Eastern Mongolia . . . 31 Biambaa Günchinsüren, Dashzeveg Bazargür An Outline of Neolithic Studies in Mongolia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
ROCK ART
Desktop Publishing and Design: Matthias Weis Translations: Authors and Emily Schalk English language editors: Joshua Wright, Susanne Reichert Image editing: Gisela Höhn Final editing: Ute Arents, Güde Bemmann, Ursula Brosseder Printing and binding: AALEXX GmbH – Großburgwedel Printed in Germany Cover illustration: Elisabeth Heinrich
Iadmaa Tserendagva Introduction into the Rock Art of Mongolia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Vladimir D. Kubarev Results and Prospects of Studying Ancient Art Monuments in Mongolia . . . . . . . . . . . . . . . . . 67 Nadsag Batbold, Tserendorzh Bolorbat Petroglyphs of Shaakhar Tolgoi, Mongolia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Iadmaa Tserendagva, Biambaa Günchinsüren Some Images of the Zhavkhlant Khairkhan Petroglyphs, Mongolia . . . . . . . . . . . . . . . . . . . . . . 89
BRONZE AND EARLY IRON AGE ISBN 978-3-936490-31-2 Copyright 2009 by vfgarch.press uni-bonn
936490_BCAA_Mongolei_06.indd 4-5
Bruno Frohlich, Tsend Amgalantögs, Judith Littleton, David Hunt, Janine Hinton, Kellyn Goler Bronze Age Burial Mounds in the Khövsgöl aimag, Mongolia . . . . . . . . . . . . . . . . . . . . . . . . . . 99
29.04.10 10:01
6
7 Jean-Luc Houle, Diimaazhav Erdenebaatar Investigating Mobility, Territoriality and Complexity in the Late Bronze Age: An Initial Perspective from Monuments and Settlements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 Esther Jacobson-Tepfer Archaeology and the Landscape in Mongolia’s High Altai: Inventory and Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
TURKIC AND UIGHUR PERIOD Lucie Šmahelová Kül Tegin Monument and Heritage of Lumír Jisl – The Expedition of 1958 . . . . . . . . . . . . . . 325 Chunag Amartüvshin, Zhamsranzhav Gerelbadrakh Bilgä Kaghan’s Treasures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
Alexei A. Kovalev, Diimaazhav Erdenebaatar Discovery of New Cultures of the Bronze Age in Mongolia according to the Data obtained by the International Central Asian Archaeological Expedition . . . . . . . . . . . . . . . . . 149
Joshua Wright, Chunag Amartüvshin “Unseen by Eye, Unheard by Ear”: The Archaeology of the Early Türks at Baga Gazaryn Chuluu, Mongolia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
Jérôme Magail, Pierre-Yves Milcent, Fabrice Laudrin, Jean François Bussiere, Chimiddorzh Erööl-Erdene, Zham’ian-Ombo Gantulga, Rinchinkhorloo Munkhtulga, Makhbal Tsengel The Joint Monaco-Mongolian “Tsatsyn ereg” Archaeological Expedition . . . . . . . . . . . . . . . 171
Tsagaan Törbat, Dunbüree Batsükh, Jan Bemmann, Thomas O. Höllmann, Peter Zieme A Rock Tomb of the Ancient Turkic Period in the Zhargalant Khairkhan Mountains, Khovd aimag, with the oldest preserved Horse-head Fiddle in Mongolia – A preliminary Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
William W. Fitzhugh The Mongolian Deer Stone-Khirigsuur Complex: Dating and Organization of a Late Bronze Age Menagerie. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
LATE IRON AGE / XIONGNU PERIOD
Ahmet Taşağil Ethno-cultural Distribution of Turkic and Mongolian Tribes between the 6th and 9th Century AD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385 Takashi Osawa The Cultural Relationship between Old Turkic Kingship and Deer Image . . . . . . . . . . . . . . . 401
Hermann Parzinger, Viacheslav I. Molodin, Damdinsüren Tseveendorzh New Discoveries in Mongolian Altai: The Warrior Grave of the Pazyryk Culture at Olon-Güüriin-Gol 10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Kōsetsu Suzuki Revision and Reinterpretation of the Choir Inscription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
Tsagaan Törbat, Pierre-Henri Giscard, Dunbüree Batsükh First Excavation of Pazyryk Kurgans in Mongolian Altai . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
Gleb V. Kubarev, Gilsu So, Damdinsüren Tseveendorzh Research on Ancient Turkic Monuments in the Valley of Khar-Iamaatyn Gol, Mongolian Altai . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
Denis Ramseyer, Nicole Pousaz, Tsagaan Törbat The Xiongnu Settlement of Boroo Gol, Selenge aimag, Mongolia. . . . . . . . . . . . . . . . . . . . . . . 231 Sergei V. Danilov Preliminary Results of the Investigations on a Xiongnu Settlement in Mongolia . . . . . . . . . . 241 Ursula Brosseder Xiongnu Terrace Tombs and their Interpretation as Elite Burials . . . . . . . . . . . . . . . . . . . . . . . 247 Hyeung-Won Yun The Xiongnu Tombs at Khudgiin Tolgoi in Mongolia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281 Bryan K. Miller, Zhamsranzhav Baiarsaikhan, Tseveendorzh Egiimaa, Prokopii B. Konovalov, Judith Logan Elite Xiongnu Burials at the Periphery: Tomb Complexes at Takhiltyn Khotgor, Mongolian Altai . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301 Jean-Paul Desroches, Guilhem André New Perspectives in Xiongnu Archaeology through Studies on the Aristocratic Necropolis of Gol Mod, Mongolia: Activities of the French-Mongolian Archaeological Expedition (MAFM) in the Context of Xiongnu Archaeology . . . . . . . . . . . . . . . . . . . . . . . . . 315
936490_BCAA_Mongolei_06.indd 6-7
Ayudai Ochir, Lkhagvasüren Erdenebold About the Uighur City of Khedun, Mongolia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
KITAN PERIOD AND MONGOLIAN MIDDLE AGE J. Daniel Rogers, Claudio Cioffi-Revilla Expanding Empires and a Theory of Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445 Nikolai N. Kradin, Alexandr L. Ivliev The Downfall of the Bohai State and the Ethnic Structure of the Kitan City of Chintolgoi Balgas, Mongolia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461 Zhamsranzhav Baiarsaikhan A 13th–14th Century Mongolian Grave at Bukhiin Khoshuu . . . . . . . . . . . . . . . . . . . . . . . . . . 477 Ildikó Oka Three Mongolian Coats from the 13th–14th Century Grave at Bukhiin Khoshuu . . . . . . . . . 487
29.04.10 10:01
8 Ernst Pohl Interpretation without Excavation – Topographic Mapping on the Territory of the first Mongolian Capital Karakorum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505 Hans-Georg Hüttel Royal Palace or Buddhist Temple? On Search for the Karakorum Palace. . . . . . . . . . . . . . . . . 535
PREFACE
Noriyuki Shiraishi, Batmunkh Tsogtbaatar A Preliminary Report on the Japanese-Mongolian Joint Archaeological Excavation at Avraga Site: The Great Ordu of Chinggis Khan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
NATURAL SCIENCES Albert Russell Nelson, Chunag Amartüvshin, William Honeychurch A Gobi Mortuary Site through Time: Bioarchaeology at Baga Mongol, Baga Gazaryn Chuluu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565 Christine Lee Who were the Mongols (1100–1400 CE)? An Examination of their Population History . . . . 579 Saran Solongo Luminescence Dating of Fired Bricks from Ancient Mongolian Cities. . . . . . . . . . . . . . . . . . . 593
INDEX OF GEOGRAPHICAL NAMES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597 INDEX OF AUTHORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 609
Since Mongolia’s political opening in the early 1990s, the number of archaeological expeditions under bi- and trinational direction has increased considerably. Scholars from the United States, Japan, Korea, China, Russia, Kazakhstan, Turkey, Hungary, Germany, Italy, Monaco, Switzerland, France as well as from other countries are engaged in cooperative projects with the Archaeological Institute of the Mongolian Academy of Sciences, the National Museum of History or one of the archaeological university institutes in Mongolia. These various research projects (cp. the compilation by Tsogtbaatar and Batbold 20051 and the contribution of Tseveendorzh in this volume) confirm impressively the hospitality and openness of the Mongolian people towards new approaches in research as well as the extraordinarily favourable working conditions in the country. The impetus for organising an international conference in Mongolia was ultimately supplied by the reorientation of our research which so far has focused on Karakorum. This induced us to make a first assessment not only of our own projects (contributions by H.-G. Hüttel and E. Pohl), but furthermore to gain information about the state of research that has been achieved in the meantime. We intended both to document retrospectively the increase of knowledge and, at the same time, to look ahead. In recent years there have been colloquia in Ulaanbaatar or in the respective partner countries, mostly arranged in the context of the numerous cooperation agreements, although these were rather like workshops attended by colleagues who knew each other well. Yet a comprehensive conference that included all teams working in Mongolia still remained a desideratum. Accordingly, our first enquiries and circulars regarding such a conference received positive response. The main concern of the conference was to improve the international network of research groups in Mongolia and to initiate a first exchange of experiences. The aim was to gain an overview of the work being conducted, share experiences in project management and draw Mongolian as well as foreign colleagues together into discussions. Firstly, research deficits would become more apparent, while, secondly, the opportunity would be presented to coordinate future research projects and to cooperate more closely in the field of basic research. Through the various national and international research projects almost the entire time span from the Palaeolithic until the early modern era is covered. Moreover, all source categories are represented, and the expeditions have reached the many diverse regions and natural environments of the country. Thanks to the very generous funding by the Gerda Henkel Stiftung, from 19th to 23rd August, 2007, the results of what had been achieved so far could be presented for the first time after almost two decades of intensive field research in a conference entitled “Archaeological Research in Mongolia”. 1
936490_BCAA_Mongolei_06.indd 8-9
Tsogtbaatar / Batbold 2005: B. Tsogtbaatar / N. Batbold, Archaeological Cooperations in Mongolia.
Bulletin of Japanese Association for Mongolian Studies 35, 2005, 109–126.
29.04.10 10:01
10 Up to 100 colleagues from home and abroad as well as students of archaeological disciplines at different Mongolian universities in Ulaanbaatar attended the 41 presentations, which were organised in seven sections. The conference was opened on the evening of 19 August with a reception given by the Mongolian Academy of Sciences at the National Museum of History, in rooms kindly provided by the museum director, Prof. Dr. A. Ochir. The actual conference began on the following morning at the State and Government “Elite” Centre with a welcome address by the President of the Academy of Sciences, Prof. Dr. B. Chadraa, followed by messages of greeting by Prof. Dr. D. Tseveendorzh (Institute of Archaeology of the Academy of Sciences), Prof. Dr. H. Parzinger (German Archaeological Institute) and Prof. Dr. J. Bemmann (University of Bonn). O. Mironciuc M.A. gave an informative account about the activities of the Gerda Henkel Stiftung, particularly in the context of the special program on Central Asia. The individual sections began with introductory papers on the state of research and current questions. The program on 20 August comprised papers on the subject areas of the “Stone Age”, “Rock Art” and the “Bronze Age and Early Iron Age”. The last mentioned section was continued on the next day, 21 August, followed by papers on the “Late Iron Age / Xiongnu Period”. On 21 August, H. Parzinger together with the German Ambassador P. Fischer inaugurated the newly established research centre of the German Archaeological Institute in Ulaanbaatar. In the evening H. Parzinger gave a lecture on “New Results on the Archaeology of the Scythian Period in South Siberia and Northern Mongolia” at a reception at the Embassy of the Federal Republic of Germany. On 22 August the conference was continued with papers in the section on “Early Historical Periods”. In the afternoon of the same day an excursion to Noyon Uul, the world-renowned cemetery of the Xiongnu period, was on the program. We are very grateful to Dr. N. Polos’mak and her colleague Dr. E. Bogdanov for their elucidative tour of the excavation site, including the impressive findings from tomb 20, which was excavated in 2006 and 2007. Papers in the sections “Medieval Period” and “Natural Sciences” completed the program on 23 August, and the conference concluded in the evening with a reception at the Embassy of the French Republic. Our thanks go to all of the colleagues who described their project at the conference and submitted a manuscript for print. Some authors included a bibliography of their project as a service to the readers. As not all papers were submitted for print, contributions of other colleagues could be accepted. The present volume bears witness to the abundance of archaeological monuments and the diversity of scientific approaches. The contributions present a snapshot, an interim report on research in action, much of which will likely be expanded and complemented, some of it even corrected, in the future. Besides spectacular discoveries and excavation successes (see contributions by Parzinger, Molodin and Tseveendorzh; Desroches and André; Tseveendorzh; Amartüvshin and Gerelbadrakh), it is particularly the long-term and systematic studies in the individual regions (contribution by Jacobson-Tepfer) and comprehensive analyses (contribution by Rogers and Cioffi-Revilla) that add considerably to a greater understanding. Some periods already look back upon a long tradition in research, which provides the results of numerous expeditions and corresponding publications. This applies, for example, to the Stone Age periods, which met with lively interest, particularly of the Russian scholars, and to the Xiongnu epoch, which since the excavations of Kozlov in Noyon Uul in 1925 drew great international attention. The picture that can now be drawn of this archaeological culture is accordingly differentiated. By contrast, investigations on the Old Turkic period are marked rather by linguistic studies. The various groups of monuments likewise reflect a different state of research. Upright stones, some dis-
936490_BCAA_Mongolei_06.indd 10-11
11 playing anthropomorphic features, of the Late Bronze to the early Iron Age, the Old Turkic and Mongolian periods as well as tombs were in the centre of research. Excavations in urban areas and settlements have been carried out on a larger scale only during the past decades. They have changed our perception of the cultures of nomadic pastoralism to a great extent. The antiquarian analysis of entire subject groups, their firm positioning in time and space are in many cases still in the beginning stages, mainly due to the availability of source material and publications. Our present knowledge about the sequence of cultures shows a very static picture: individual, scarcely structured blocks of cultures follow one another, sometimes separated by gaps of several hundred years. The transition from one epoch to the next or from one tribe to the next that is attested by written sources as well as queries as to the reasons for the respective changes in culture have not yet received sufficient attention. Migrations and wars are often taken as seemingly self-explanatory reasons for the disappearance or emergence of cultures, without discussing and questioning the archaeological sources independently. Cooperation with disciplines in the natural sciences has been conducted only to a very limited degree so far. Particularly palaeo-environmental data could be of great relevance for studies on cultural change, economic strategies and the emergence of nomadic polities. If one attempts to assess achievements in the archaeological investigation of Mongolia that have been made thus far, a retrospective is recommendable. Individual stages are evident in the summarising studies by Sergei Kiselev2 and Eleonora Novgorodova3; see also Jettmar 19834. It is worthwhile to compare the articles in the volumes accompanying the exhibitions “Die Mongolen und ihr Weltreich” in Hildesheim and Munich in 19895 and “Dschingis Khan und seine Erben” in Bonn, Munich, Vienna, Istanbul and Budapest in 20056. The apparent increase in knowledge is striking. This should not conceal the fact that in many cases the foundations must still be laid. Building upon past achievements is the task of the younger generation of archaeologists in Mongolia, who are distinguished by their multilingualism and international experience, as well as of their partners abroad. The organisers hope that with this conference an international dialogue of knowledge about archaeology in Mongolia has been initiated, which should be continued during the next years with thematically more narrowly defined activities. In this respect it is gratifying that already in 2008 a second larger international conference took place in Ulaanbaatar, organised by Dr. Ursula Brosseder, University of Bonn, and Dr. Bryan K. Miller, University of Pennsylvania, financed by the Silk Road Foundation. Without the considerable financial resources that were provided by different institutions, such a conference would not have been possible. First and foremost, the Gerda Henkel Stiftung must be mentioned, whose generous funding not only covered a major part of the conference expenses, but which also provided the funds for printing the present publication. We would like to express our gratitude to the representatives of the foundation for their interest in the subject, the generous support and the friendly mentorship of the project. 2 3
S. Kiselev, Mongoliia v drevnosti. Izvestiia Akademii Nauk SSSR, Otdel istorii i filosofii IV, 4 (Moskva 1947). E. Nowgorodowa, Alte Kunst der Mongolei (Leipzig 1980); E. Novgorodova, Drevniaia Mongoliia: nekotorye problemy khronologii i etnokul’turnoi istorii (Moskva 1989).
4
5 6
K. Jettmar, Geschichte der Archäologie in Sibirien und im Asiatischen Steppenraum. Beiträge zur Allgemeinen und Vergleichenden Archäologie 5, 1983, 187–226, esp. 218 et seq. A. Eggebrecht (ed.), Die Mongolen und ihr Weltreich (Mainz 1989). Dschingis Khan und seine Erben. Das Weltreich der Mongolen (München 2005).
29.04.10 10:01
12 The German Academic Exchange Service (Deutscher Akademischer Austauschdienst) has financed the journeys of the German participants by appropriating funds for travel expenses through the scientific exchange program. Hereby our thanks go to Prof. Dr. D. Regdel, Mongolian Academy of Sciences, as well as to Dr. Klaus Birk and Gabriele Buchmann-Schmitz of the DAAD. Fortunately, many participants could cover their travel costs with their own resources, which eased the pressure on our budget considerably. In Ulaanbaatar the diplomatic missions of the French Republic and the Federal Republic of Germany have held receptions in the rooms of the respective embassies, thus emphasising the close cultural ties between both countries and Mongolia. For this, the editors express their sincere gratitude to the Ambassador of the French Republic, His Eminence Monsieur Patrick Chrismant, and to the consul, Monsieur Didier Guilbert, as well as to the Ambassador Extraordinary and Plenipotentiary of the Federal Republic of Germany, His Eminence Mr Pius Fischer, and the Secretary in charge of language and cultural affairs at the embassy, Mr Michael Rossbach. Our colleague of many years, Dr. Kh. Ariunchimeg, contributed significantly through her collaboration in the organisational work to the success of the conference. She was supported by Ts. Egiimaa M.A. and Lk. Mönkhbaiar M.A., members of the staff at the National Museum of Mongolian History respectively at the Archaeological Institute of the Mongolian Academy of Sciences. As English was chosen as conference language, the need for interpreters was limited. We are very grateful to Prof. Tsoros Tse. Jonong and Dr. U. Brosseder for the translation of Mongolian and Russian speeches and lectures. For correcting the English texts penned by second-language speakers, we could engage our friend and colleague, Dr. Joshua Wright, University of Stanford, who edited all texts of nonnative speakers during the past months. A final linguistic revision was carried out by Susanne Reichert, University of Bonn. No less considerable problems were posed by the standardisation of the bibliographical references, as the authors – according to their country of origin and publication traditions – used differing quotation and transliteration systems. Despite a considerable amount of research work, a uniformity with the rules of the Library of Congress was not achieved in all cases. Unification of the bibliography and transliteration, final editing and compilation of an index of geographical names were done by Dr. Ute Arents, Dr. Güde Bemmann, and Dr. Ursula Brosseder. The image edition was carried out by Gisela Höhn, University of Bonn. Sincere thanks are extended to all persons mentioned for their commitment to the preparation of this publication. Typesetting, layout work and printing supervision were in the hands of Weiß-Freiburg GmbH – Graphik & Buchgestaltung. Mongolia lends a great fascination for guests and scholars who are engaged in work there. Almost no one leaves the country unaffected, the vastness and magnificence of the country, its wealth in monuments and the greatness of its tradition inspires respect, awe and affection. One always departs with the wish to return as soon as possible, to continue what was begun and to give an impetus for new undertakings. May this book impart to the reader an impression of the pleasure and enthusiasm felt by the international community of scholars who are working in this country. Jan Bemmann, Hermann Parzinger, Ernst Pohl, Damdinsüren Tseveendorzh
936490_BCAA_Mongolei_06.indd 12-13
CURRENT ARCHAEOLOGICAL STUDIES IN MONGOLIA Damdinsüren Tseveendorzh
The Institute of Archaeology of the Mongolian Academy of Sciences, the leading organization in archaeological studies on the territory of Mongolia, aims at investigating the main problems in the studies of Mongolian prehistory and history through modern archaeology of an international standard. Since the year 1990 the institute has organized over 70 expeditions. Between 2001 and 2007 the institute implemented joint projects in collaboration with universities and scientific organizations of many countries including Russia, The United States, Japan, Germany, France, Hungary, Turkey, Belgium, Italy, Monaco and Kazakhstan. As a result, large amounts of archaeological evidence were gathered that make a significant contribution to the interpretation of key questions in the studies of ancient history and culture of the nomads.
The Stone Age The oldest tools found by research in the south-western and western parts of our country in recent years date as far back as the pre-Acheulian period (approximately 750,000–800,000 years ago), and can possibly be dated earlier once further, more detailed investigations are carried out. In the last three years researchers have focused their efforts mostly on the study of the two to archaeologists relatively unknown regions: Khövsgöl aimag and Bulgan aimag. Excavations were conducted at two important cave sites in Baiankhongor aimag, the Lower Paleolithic Tsagaan Agui (750,000–800,000 years ago) in Baianlig sum and the Mesolithic Chikhen Agui in Baian-Öndör sum in cooperation with the Institute of Archaeology and Ethnology of the Siberian Branch of the Russian Academy of Sciences and the University of Arizona. This cooperation resulted in the identification of dozens of new settlements dating back to the Stone Age in Trans-Altai Gobi regions and other desert regions, the most outstanding of which is Tsakhiurt Valley.
Summer 2009
29.04.10 10:01
EXPANDING EMPIRES AND A THEORY OF CHANGE J. Daniel Rogers, Claudio Cioffi-Revilla
Abstract Several themes consistently play a role in the study of early state formation in eastern Inner Asia, beginning with the Xiongnu state around 200 BC and continuing through the Mongol Empire of the 13th and 14th centuries AD. Because of the nature of these early states scholars have frequently argued that they required the goods, ideas, and cultural knowledge developed by neighbouring agricultural societies. The cultural knowledge, however, came from multiple sources, especially the internal traditions that linked successive steppe polities. Using the Mongol Empire as a prime example, new evidence suggests that the traditional distinction between a pastoralist and agriculturalist economy is insufficient to characterize the complexity of interactions. It is also the case that steppe polities are often described as short-lived entities that succeeded each other in rapid succession. This description de-emphasizes the economic and cultural continuity that transcended the rise and fall of individual political entities. To provide a broad comparative context for analysis of the themes of interaction and continuity, an interdisciplinary effort is underway to develop agent-based simulation models of social complexity. The models currently under development are grounded in the perspective of a canonical theory of social complexity. The repeating and uncertain process of institutional emergence and political development is explained as resulting from a succession of transitions that occur over time, based on a common theme called the “fast” branching process. Macro, or “slow” processes are also modeled at the societal level in an analysis of trajectories of social change. Both processes were usually endogenous; i.e., only partially and often not crucially dependent on neighbouring agrarian societies.
Introduction While scholars working in many disciplines have made great contributions to understanding the dynamic origins, expansion, and organization of states and empires throughout the world, important aspects remain as sources of intense debate or for which little research has been undertaken. The research described here is an interdisciplinary initiative that explores a series of long-standing issues in the history of empire development on the eastern steppes while also
446
J. Daniel Rogers, Claudio Cioffi-Revilla
developing a comparative analysis of social dynamics using computational modeling and a comprehensive theory of social change. In general, the states and empires that emerged from the pastoralist societies of the steppes of Mongolia have been undervalued in the interpretive scope of world history and in social science theories of the emergence of complex societies. In many cases this is the result of the seemingly ephemeral nature of these polities and the broadly held perspective that they are derivative of events in China (e. g. Barfield 2001). Several other recent studies have explored the nature of the relationship with China from other perspectives, but more importantly have considered the broader issues of how these polities may contribute to revisions in prominent social theory (e. g. Di Cosmo 1999; Kradin et al. 2003; Rogers 2007; 2010). This study uses comparative analysis to explore key aspects of these early empires through study of the early documentary record, archeological research at urban centres and other central places in Inner Asia, and through the development of a series of computer simulations that model the emergence of complex social systems. These models are based on a general theory of societal change described here as the canonical theory (Cioffi-Revilla 2005). This theory is the basis for how a series of more specific processes and conditions are evaluated in a broadly interpretive context. This chapter focuses on the dynamic history of early states and empires by first briefly describing the canonical theory and illustrating how it may be applied to specific instances. On this basis, additional organizational aspects are presented that are also fundamental to the emergence of social hierarchies. In particular, we explore the concept of order and the mechanisms for the expansion of social control, once established. This is followed by a discussion of how these concepts are implemented through a series of agent-based computational models of social dynamics. The theory and analysis have implications for the emergence of all complex societies; however, the focus here is on the period of the Bronze Age in Mongolia (ca. 2000 BC) through the end of the Mongol empire (1368 AD), with emphasis on the latter part of this time range.
The Canonical Theory The canonical theory of socio-political complexity (Cioffi-Revilla 2005) provides an explanation of polity formation and evolution based on a mechanism called “canonical variation” on two time scales. Briefly stated, the theory works as follows: On a short-term time scale, canonical variations occur when a given society is subject to situational changes (emergent issues) and the society then succeeds or fails to respond through collective action of some form. Canonical variations occur as this main theme repeats over and over again throughout history, because the exact nature of the situational change, the form of collective action, and the success or failure of action undertaken can vary significantly – albeit within well-specified bounds – while maintaining the same recurring or iterative theme given by situational change, collective action, and resulting outcomes for the polity. As a canonical theory, the iterative and uncertain process of institutional emergence and political development (and occasional decay) is explained as resulting from a succession of transitions, based on variations on a common theme called the “fast” branching process. As each canonical variation of the same “fast” process occurs for a given society, social complex-
Expanding Empires and a Theory of Change
447
Fig. 1. A simplified representation of the canonical theory of social complexity, illustrating the “fast” iterative branching process. In turn, each main event (G, C, N, U, S, and complimentary failure modes) has an associated causal model in conditional logic, such that the branching process with all component sub-models constitutes a composite sequential-conditional model (Cioffi -Revilla 1998, 239–241; 2002; 2005).
ity may accrue, decrease, or remain the same, producing the “slow” accrual process based on emerging experience, statecraft, bonds of trust, norms, institutions and other collective goods, including “social capital” (Coleman 1988; Brown / Lauder 2000). Negative externalities may also be produced, leading to decay in socio-political complexity, as explained below. Thus, the theory explains socio-political evolution by a process of canonical variation. Fig. 1 shows the main events in the “fast” canonical process, denoted by G, C, N, U and S, and their failure modes ~ C, ~ S. The process as a whole generates a political sample space Ω (outcomes on the right). Social complexity emerges (event P, the top outcome in Ω) as a pathdependent phase transition produced by a process of several different albeit specific outcomes. The full model (not shown here) has five other detailed sub-trees (triggering or production rules) for generating each of the five main events in the “fast” branching process from G to S1. The first stage of the fast process begins when an existing group lacks a system of government (i.e. the community is not yet a polity, event G in Fig. 1, left) and may end in a different situation (politically complex phase) when such a community has formed a system of government (the community is a polity after iterations of the event P in Fig. 1, top right). In the early history of Inner Asia this initial phase occurred up to the early Neolithic period, during which time not even chiefdoms are archeologically observable. The process had begun by 2500 BC (terminus ante quem). A situational change makes a group metastable, because a potential for increased (decreased) socio-political complexity is created, but not immediately realized. The realization of a potential for complexity depends on how the rest of the fast process evolves and on how people and environments interact. Xiongnu and Mongol societies succeeded; many others failed. 1
For full details of the fast process see Cioffi-Revilla 2005.
448
J. Daniel Rogers, Claudio Cioffi-Revilla
Given a situational change (C), the group may or may not understand the need for collective action caused by the change (event N in Fig. 1, after C). Causally, N is an information-processing event, involving signal detection, cognition, and other causal events, and is modeled accordingly. If the group does not understand the situational change, it may be destroyed or dispersed without further political development (outcome X* ∈ Ω); hence the critical role of intelligence. If the group grasps the situational change, it may or may not be willing and able to undertake collective action (U), depending on its capacity (asabiya). Collective action occurs in several modes (Lichbach 1996), detailed in Cioffi-Revilla (2005). Sometimes a society fails to undertake collective action (~U), even when it understands a situational change C, by incapacity. If collective action occurs with persistent situational change, then it may succeed (S) or fail (~S), depending on the situation and action. If it fails, the society may be destroyed (outcome Z ∈ Ω). The Mongol federation succeeded; others failed and were absorbed or destroyed as the Mongol imperial polity expanded. Several outcomes in the outcome space (events denoted by X) can all eventually lead to state failure, for example by long-term loss of collective action capacity. Finally, if the society succeeds at time t, then the consequences or societal effects will augment its political complexity (outcome A ∈ Ω) at time t+1, because – even if only on a small scale and temporarily – mobilization of resources, lessons about who to trust, hierarchies of leaders and followers, specialized assignments, division of labor, planning, information sharing, coordination experience, and other elements of governance will have been realized through the experience. Significantly, collective action capacity (CAC) for dealing with the next situational change (threat or opportunity) will increase. The phase transition in the quantum increase in CAC is observable by the formation of multiplex networks on several scales: cognitive, individual, group, and institutional. Such a phase transition has enduring organizational effects on the group, and the next time its situation changes and demands collective action it will draw on more CAC and cope better; it will have more governance experience than before. The phase transition also means the realization of the potential that had been created by the earlier situational change, when the group had become metastable after the initial phase. A single passage through the “fast” canonical process was just described. Over time, a group will experience many such processes, each as a variation on the common theme of challengeresponse. Failure paths lead to political decay or even destruction (events beneath P), so gains in political complexity are not preordained. Importantly, many of the canonical variations of the fast process that can produce successful socio-political development are only partially and often not crucially dependent on developments in other neighbouring societies; many are entirely endogenous or, in other cases, environmentally produced. Situational changes produced by neighbouring societies constitute only a small – albeit sometimes relevant – portion of the total space of possible causal paths to socio-political change.
Constructing Continuity and Change While there are many aspects of the canonical theory to be elaborated, here we will elaborate on only two aspects. One of the key concepts in the formation of any complex organization is the construction of “order” developed through processes of collective action, as presented in the accumulation of social complexity in the canonical theory. How complexity accrues
Expanding Empires and a Theory of Change
449
and order emerges is basic to many forms of research on states and empires, although it is often not made explicit. The term “order” as used here makes reference to long-term cultural constructs that limit cultural variation (entropy reduction) and are usually described as traditions, or even civilizations in a broader sense (Baines / Yoffee 2000, 14–15). The unique aspect of state formation is actually about how these traditions are recombined or otherwise altered to accommodate the innovations necessary to develop new and specialized forms of hierarchy, communication, and collective objectives, at least as defined by the elites. While the emergence or imposition of order can certainly have negative and limiting aspects for individuals, it also produces a kind of simplicity in that relationships and responsibilities become more predictable (Scott 1998). This may, in fact, be one of the most significant aspects of why states and empires became so widespread. There is no question that important forms of tradition and control existed among the agropastoralists of Inner Asia for several thousand years, but of more particular concern is the source of this cultural knowledge (De Marrais et al. 1996; Rogers 2007, 258). In other words, how did these traditions develop? Were they generated primarily from a long agropastoralist history, or were there significant external sources, such as China or the sedentary states further to the west? Importantly, the canonical theory does not require such external sources, precisely because numerous variations of endogenous fast processes can produce successful development of socio-political complexity. In other words, the Deus ex machina of Sino-centric models is neither theoretically needed nor empirically supported as a general causal explanation for understanding the political history of Inner Asia. As with the development of any complex system, the sources of knowledge were varied, however, there is substantial reason to conclude that much of the motivations and mechanisms of control were drawn specifically from a rich and varied steppe tradition and not largely motivated by interaction with China and other agricultural states; i.e., they were often endogenous and only in some cases did they relate to China or to Chinese actors. The construction of order within many societies, but especially among the agropastoral nomads is reviewed briefly here to consider four components, each as part of a single canonical fast process: First, and perhaps most powerful, is the legitimation of leadership, as in the construction of power (Rogers et al. 2005; see also Lichbach 1996, on leadership and CAC), facilitated by the natural rate of occurrence of leaders among the general population (approximately 1:7). Leadership legitimation is a fundamental process that takes place in the formation of chiefdoms and the creation of hereditary or otherwise formalized positions of office. In many instances these offices are also closely linked to kinship, either real or fictive (Rogers 1991, 235–236). The link with kinship is especially evident among the steppe agropastoralist confederations, including the Mongol Empire (Kradin / Skrynnikova 2006). Divine authorization, as the second component, is a basic aspect of legitimation, along with kinship. As in Chinese kingship establishing the right to rule, there is a heavenly mandate (Golden 1982; Lot-Falck 1956), but it is not necessary to interpret the steppe system as a reflection of Chinese practice, since there are also systems of belief based on the Sky God, Tengri, which likewise can serve as a source of divine right. Fundamentally intertwined with the acquisition of legitimate sources of elite authorization are attempts by individuals to acquire prestige (Bender 1978, 213; Friedman / Rowlands 1977, 207), which is another explicitly endogenous canonical variation of the fast process. The ways in which prestige is acquired and leadership is legitimated are endogenous processes designed to construct reliable ways for leaders to control human and therefore, other kinds
450
J. Daniel Rogers, Claudio Cioffi-Revilla
of resources. Control of resources is the third component, the materiality of elite representation (Baines / Yoffee 2000, 14; Earle 1997). This encompasses things from the very small, such as crowns and other symbols of elite control, to major constructions on the landscape as temples, palaces, castles, burial mounds, and monuments that specifically embody a link to the past. The fourth component – particularly relevant in the case of empires – consists of the social mechanisms used to bridge ethnic and linguistic diversity in the construction of order. This is especially important to consider because such diversity is a source of competing ideologies and worldviews, which are potentially disruptive to power relations. In other words, such diversity can represent competition to leadership. In many states, such diversity is considered disruptive and is often dealt with in the harshest terms. However, there are multiple strategies that have been used to bridge diversity, sometimes in the initial consolidation of the polity, but always in the expansion of a polity into an empire. These strategies may be implemented as a situational change, resulting in successful collective action, that is, the actual territorial growth of the polity. The four approaches to societal integration are: elimination of diversity, creation of continuity, overlay mechanisms, and marginal incorporation (Rogers 2007, 263–264). Elimination of diversity is the most draconian approach in that it represents an attempt to exterminate or resettle populations to other regions outside the empire. While the Mongol Empire and other empires did conduct extensive killing in their conquests to force submission (Dawson 1955, xiii), these were not intentional attempts to eliminate diversity. The second approach, creation of continuity, involves purposeful attempts to co-opt local elite symbolism and a wide variety of other cultural ties. This was often done through alliances, treaty, exchange of marriage partners, and especially the creation of new traditions. The third strategy, overlay mechanisms, includes a group of commonly used strategies involving the replacement of key positions in the local population, while leaving many other lower-level functions unchanged. This is a very common, low-investment conquest strategy that was used extensively by the Inner Asian empires to achieve societal integration. It is also a potentially risky strategy, since it makes rebellion relatively easy and does little initially to bridge diversity. The fourth strategy is marginal incorporation. This approach is often seen at the very edges or beyond the margins of expanding empires. In this strategy the expanding empire usually extracts tribute payments. Even when a particular region is incorporated into the empire, the only connection may continue to be through some form of tribute. This minimal approach perpetuates a very fragile connection that does nothing to solidify incorporation of a new region.
Comments on Application to the Xiongnu (Hunnu) and Mongol Empires In Inner Asia the slow process at the societal level eventually generated state-level polities (Xiongnu, Türk, Uighur, Kitan, Mongol Empire, and others) and some failures. The Xiongnu polity formed ca. 200 BC because the Xiongnu society at that time was able to overcome, through collective action, a situational change given primarily by Chinese attacks from the south, which took place while the Xiongnu had a pre-state system of government. In mod-
Expanding Empires and a Theory of Change
451
ern political science terminology, Xiongnu society was collectively affected by the national security issue posed by the Han attacks and, in response, was able to deal with the issue by enacting military and economic policies that eventually significantly mitigated, if not entirely eliminated the Chinese threat. In fact, the Xiongnu turned the Han threat into an opportunity for developing their public finance and avoid taxing their own society. During this process Xiongnu leadership was transformed to accommodate, and in some cases cause the changes through the establishment of systems of order. Had Xiongnu society failed in its collective action it would have transitioned into one of the other forms of lesser social complexity in the outcome-space Ω. For example, the Xiongnu polity may have been destroyed, conquered or dispersed (event X). The latter were failure instances in dealing with invasions, economic, demographic, or environmental changes. For example, power struggles after the 15th century produced political decline and Mongolia was eventually conquered by the Manchus in the 17th and 18th centuries (Crossley 1997). The following is an account of the very beginning formation of the Mongol Empire based on the canonical theory explained above, especially up to and limited to its initial formation, but excluding the subsequent development under Chinggis Khan and his successors. Recall that the slow process, which takes place over the long-term (trans-generational), is composed of numerous fast process iterations, each of which produces positive, neutral, or negative outcomes for the socio-political complexity of the society, corresponding to growth, stasis, or decay in socio-political complexity, respectively. The initial situation – in terms of actors, space, and time – is given by the disunited condition of Mongol and Turkic groups, following the disintegration of the Kitan polity in AD 1125. (Although the neighbouring Jin [Jurchen] polity that governed over the Manchurian groups lasts from 1115 to 1234, it is not part of the future initial Mongol Empire, so we set it aside.) While Mongol society as a whole – i.e. consisting of agropastoralist groups and their associated hierarchies – endured in a general state of disarray that lacked the centralized system of government that would eventually emerge in 1206, it was subject to numerous threats and opportunities (public issues) that presented themselves over time. A major significant problem with previous theoretical models for explaining the formation of the Mongol Empire – or of other polities for that matter – is the insistence on monocausal, linear mechanisms that ignore social complexity. For instance, Kradin and Skrynnikova (2006, 545) summarize in their extensive study eight widely used monocausal explanations for the rise of the Mongol Empire: “(1) various climatic changes; (2) the bellicose and adventurous nature of nomads; (3) overpopulation of the steppe; (4) growth of productive forces and class struggle, weakening of agricultural societies due to feudal disunity (Marxist concepts); (5) the necessity of replenishing an extensive livestock economy by raiding more stable agricultural societies; (6) the reluctance of settled population to trade with nomads (surpluses of livestock that could not be sold anywhere); (7) the personal qualities of steppe society leaders; (8) impulses of ethnic integration.” While Kradin and Skrynnikova note that some of these causal explanations are more viable than others, they make the point that each of these has typically been presented and argued individually. By contrast, the canonical theory allows for multiple causes (causal paths) that can lead to the same outcome of increased socio-political complexity; or also failure. Although multiple, the multiplicity of paths is both finite and specific: not everything goes. Likewise, most significant social changes are likely to have multiple fundamental causes that operate as both fast and slow processes.
452
J. Daniel Rogers, Claudio Cioffi-Revilla
Constructing Artificial Societies States and empires developed and changed in a variety of ways and for many intertwined reasons, as expected by the canonical theory. The development of the canonical theory and other theoretical advances in the study of complex social dynamics presents an excellent foundation for new interpretations. The approach taken here is to build on previous and ongoing efforts in computational historical dynamics (or agent-based “cliodynamics”; paraphrasing Turchin 2003a, 2) – the development of computer simulations. The models being developed are laboratories, in which many ideas about past events, not readily studied archeologically or through documents can be explored. In this laboratory of artificial societies, the basis for events that did happen might have happened, and never happened can all be studied, given the ever-increasing sophistication and power of modern computing guided by viable social theory. This project addresses fundamental questions in theory, methodology, and data concerning human and social dynamic responses to social and environmental challenges in Inner Asia and the Eurasian world system over the past 4,500 years. Inner Asia played an essential catalytic role in the history of human and social dynamics in Eurasia, including the phenomenon of the largest territorial polity to emerge in the evolution of civilizations – the Mongol Empire of the 13th and 14th centuries – as well as the largest scale economy – the Silk Road network – before present-day globalization. The interdisciplinary team combines expertise from political science and international relations, archeology, cultural anthropology, physical anthropology, computer science, economics, and applied mathematics. Computer science expertise includes artificial intelligence, multi-agent systems and evolutionary computation. Over the long run, results from this project will have multiple broader impacts on areas of surviving severe social challenges faced by contemporary human communities, such as global change, terrorist violence, or cultural conflicts – domains where collective action responses and intelligent organizational coordination are essential for human and social survival on several scales. More specifically, there are three synergistic goals for the project: 1. Develop, test, and disseminate a new interdisciplinary theory of long-term societal change and adaptation to complex and evolving social and physical environments, a “generative” theory formalized by a spatial multi-agent computational model. 2. Contribute to the shared understanding of social complexity across the social sciences by integrating concepts and principles within the proposed theoretical framework and research methodology. 3. Produce and disseminate new interdisciplinary data resources created by this project, including a database of archeological sites, chronological information, and a diachronic atlas of Inner Asian polities. While there are multiple products being generated by this research, the principal objective is the development of computational simulation models. The research is moving forward on two main fronts. The computational social science team, led by Claudio Revilla-Cioffi is working to develop the agent-based models of social dynamics, while the Smithsonian archeological teams are assembling new data in conjunction with already available historical information. Three archeological teams led respectively by William Fitzhugh, Bruno Frohlich, and William Honeychurch and Joshua Wright, have conducted fieldwork in Mongolia for two field seasons under the sponsorship of this project. A third field season was conducted in the summer of
Expanding Empires and a Theory of Change
453
2008. The field projects focus on three scales of analysis in three different regions, with each requiring a different set of variables. At the largest scale is the subcontinent, involving the interactions of states and empires. There is also a regional scale that focuses on tribes, confederations, ethnic groups, constituting the sub-segments of states and empires. The third scale is the local and consists of the interactions of households and camps. By working in multiple regions it is possible to test the validity of a comparative approach. For instance, to what extent are the regions different and how can they be used at the larger scale? We build the larger scale model by validating sets of social and environmental interactions within well-controlled data sets. For the local level, there are three intensive survey zones in Mongolia and one in Northern China, derived from existing literature. For each zone, the goal is to develop the following information: – A chronology relevant to the objectives. Basically, the goal is to produce the most detailed chronology possible. – Time series based measures of population changes. This goal involves the use of proxy measures such as site density, site frequency, site size, or ratios of these measures against resource measures such as pasture quality. – A climate model keyed to the same chronological scale. This model is currently under development and will use changes in moisture and temperature over long periods of time. – A measure of resource types including a scale for each – for example, pasture quality that can be related to soil type, slope, and proximity to water. – Social-based strategies for using the resources. What do the agents know and do? – Types of social interactions of individuals within the family and local group. For example, use of ethnographic and historical data on family size and marriage patters (e. g. Christiansen / Altaweel 2006).
The State of the Models Currently there are two models: Household World and Hierarchies World. Each model is currently under development and while there is insufficient space to describe each in detail, it is worthwhile to briefly outline some of the characteristics and potentials. More complete descriptions of the functioning of the models are available in Cioffi-Revilla et al. (2007) and Latek et al. (2007). Both models use the MASON simulation framework (Luke et al. 2005). The Household Model details the behavior and strategies used by individual families as part of camp groups organized as clans in a steppe environment with seasonal fluctuations in resource availability. On a daily basis, each household makes decisions about where to pasture the flocks, while also striving to maintain camp cohesion and avoid forage spaces occupied by other camps. Camps split (expand in population), and merge (contract in population) and herd size expands or decreases based on availability of forage using the Moderate Resolution Imaging Spectroradiometer (MODIS) vegetation indices (Kawamura et al. 2005). The size of household herds is the principal measure of wealth. The complexity of behavior of each household and the environment in which they operate is continually under further development. Fig. 2 and 3 provide examples of visualizations produced by the Household Model. Each step corresponds to one day.
454
J. Daniel Rogers, Claudio Cioffi-Revilla
Fig. 2. Clan wealth graph showing the mean (centre line) and upper and lower ranges of the number of animals owned by each clan after 2,344 steps (days).
Fig. 3. Household wealth histogram comparing the number of households with the average frequency of animals owned by each, after 2,344 steps (days).
Expanding Empires and a Theory of Change
455
Clan wealth is measured by the number of animals owned by each clan, as visualized in Fig. 2. The cyclical trend represented is related to very predictable seasonal changes in herd populations. While this graph does not show it clearly, there is a slow decline in the overall size of herds owned by each clan. The predictable cyclical herd dynamics and the slow decline in herd sizes suggest a trend limited by carrying capacity for this hypothetical 100 x 100 km grassland environment. Household wealth shown in Fig. 3 may not represent actual rates of change, although the overall trends suggest interesting questions. For instance, initially the great majority of households grow in wealth (number of animals), represented by an approximation of a normal curve. That is, there are relatively few poor households, many with moderate wealth, and only a few rich households. In Fig. 3, after only a few years the number of extremely poor households has increased dramatically, also reflected by the lower limits of clan wealth shown in Fig. 2. Over time this trend seems to continue, again probably because of overall population growth and the increasing scarcity of good forage. This evolved distribution, generated by the simulation, is consistent with the Pareto law of wealth inequality (Kleiber / Kotz 2003). With the further development of the Household Model there are a variety of questions of interest that can be explored. For instance: To what extent does the emergence of social entities larger than the camp depend on a particular form of subsistence economy? Can it be herding alone, or does it require a mixed economy of herding, agriculture, hunting, and gathering? The inspiration for this question is the larger issue of how the steppe polities interacted with sedentary agricultural neighbours. Another related question is grounded in a more local perspective: Given available resources and technology, what kinds of buffering mechanisms are most successful at reducing subsistence risk? There are several ways that households might respond to subsistence risk, a few of these include using mobility to find new pasture, diversification of subsistence practices, raiding, resource storage (fodder, dried meat, dried plant foods, etc.), and social alliances through agreements or redundant trade to solidify trading relationships. The second model, the Hierarchies World, applies the canonical theory to explore complex forms of vertical and horizontal social organization, much as might occur in the initial formation of steppe confederations, like the Xiongnu (Honeychurch / Amartuvshin 2006; 2007). In Hierarchies World individual clans become part of social hierarchies with allegiance and control granted to a dominant clan, which is one of the endogenous variations of the fast process within the canonical theory – and somewhat similar to the special case of Carneiro’s (1970) scenario with a strong clan controlling others. These groups of clans represent polities that over time interact with other polities of different sizes and with different resources (see Fig. 4). The size and capabilities of the polities are limited by the amount of resources needed to support a hierarchy and by the managerial capacity of leaders, given the political and environmental constraints of operating in a steppe pastoralist society. The model draws on previous work by Carley and Svoboda (1996), Turchin (2003b), and O’Madadhain et al. (2006) for producing new social structures. The principal work on the Hierarchies World model thus far has implemented only one endogenous variation of the fast canonical process (situational change C produced by a rising dominant clan) and has been carried out by team members Maciej Latek, William Honeychurch, and Maksim Tsvetovat. The initial set of questions posed for this model specifically addressed the degree and nature of interactions, first, between the clans and second, between the polities composed of hierarchies of clans. Among the questions being asked is, to what extent does the number of inter-
456
J. Daniel Rogers, Claudio Cioffi-Revilla
Fig. 4. Clans organized hierarchically within a polity. Clans and polities have differential capabilities based on population and productivity (from Latek et al. 2007, 2).
contingent linkages stabilize or destabilize the emergent network of clan relations? Other questions are concerned with variability in relationships and other conditions that may lead to stability in complex networks (Latek et al. 2007, 2). The networks are crucial to the model, since they represent information and resource flows (social capital). When a polity fails to receive adequate resources from its constituent clans the result is a net decrease in CAC (collective action capacity) for sustaining governance, marking a situational change C for the polity as a whole. Polity leadership may then decide to undertake collective action by expanding (event U in Fig. 1, assuming N occurs) and conquering neighbours to improve resource flows. If this policy is unsuccessful (event ~S), CAC may continue to drop, eventually resulting in the fissioning of the polity (event X). Essentially, a confederation that cannot control enough resources to insure the allegiance of subordinate leaders (i.e., coalition side-payments, in the classical sense of Riker [1962]) will likely suffer the defection of those sub-hierarchies. Before the questions above can be seriously addressed there are several levels of testing that must occur in order to calibrate the simulation with parameters that might be observed in actual polities, based on historical and archeological information. Initial testing has shown that the model responds predictably with positive relationships between oscillations in polity size and CAC. The model also responds predictably to environmental shocks. That is, with a fall in resources there is a decrease in the size of polities. With further testing of the model, the introduction of greater agency by clan elites and the use of a realistic GIS landscape will make it possible to explore an increasing number of fundamental and sophisticated questions about the organization of early states and empires. As each of these models undergoes further development we continually come back to the broader questions by always asking what do we really want the simulations to tell us? What do states and empires do? What do they not do? What factors account for their emergence and success or failure? How can we move away from unilineal causal interpretations and towards dynamic socially relevant explanations?
Expanding Empires and a Theory of Change
457
The Scope of Smithsonian Research in Mongolia Since 2001 the Smithsonian Institution has been working in Mongolia with Mongolian scholars on several different projects. In 2004 this work was formalized through a Memorandum of Understanding, signed by the Smithsonian, the Mongolian National Academy of Sciences and the National Museum of Mongolian History. This agreement has led to a wide variety of projects ranging from the conservation of deer stones to the botany of lichens. Within the last several years our efforts have concentrated on the development of comprehensive archeological survey information in multiple regions on sites of the Bronze Age and later periods, including the development and expansion of early states and empires through the Medieval period. In collaboration with computational social science researchers from George Mason University and other institutions our archeological fieldwork is utilized as part of a larger project to develop computational models (described above) for the development of chiefdoms, states, and empires. These several projects have contributed a variety of results, including new information on Khirigsuurs (Bronze Age mounds), Deer Stone sites, and archeological settlement systems in southern, western, and northern Mongolia. Various aspects of this research are described in other chapters in this volume.
Acknowledgements Major funding for this research was provided by the Human and Social Dynamics initiative of the National Science Foundation (USA), under grant BCS-0527471. The authors wish to especially thank the primary research team, whose work is described here: Sean Luke, Dawn C. Parker, Maksim Tsvetovat, Maciej Latek, William Fitzhugh, William Honeychurch, Bruno Frohlich, Paula DePriest, Randy Latimer, Chunag Amartüvshin, G. C. Balan, Jai Alterman, Randy Casstevens, Joey Harrison, Maction Komwa, and Steve Wilcox. Field research by the Smithsonian teams has been supported in Mongolia by Ch. Dalai, Director of the Institute of History, the Mongolian Academy of Sciences and Damdinsüren Tseveendorzh, Director of the Mongolian Institute of Archaeology. We very much appreciate their collaboration and support.
REFERENCES Baines / Yoffee 2000 J. Baines / N. Yoffee, Order, legitimacy, and wealth: setting the terms. In: J. Richards / M. van Buren (eds.), Order, Legitimacy, and Wealth in Ancient States. New Directions in Archaeology Series (Cambridge 2000) 13–19. Barfield 2001 T. J. Barfield, The Shadow Empires: Imperial State Formation along the Chinese-Nomad frontier. In: S. E. Alcock / T. N. D’Altroy /
K. D. Morrison / C. M. Sinopoli (eds.), Empires: Perspectives from Archaeology and History. Conference “Imperial designs: Comparative Dynamics of Early Empires”, held in Mijas, Spain, in the autumn of 1997 (Cambridge, New York 2001) 10–41. Bender 1978 B. Bender, Gatherer-Hunter to Farmer: A Social Perspective. World Archaeology 10, 1978, 204–222.
458
J. Daniel Rogers, Claudio Cioffi-Revilla
Brown / Lauder 2000 Ph. Brown / H. Lauder, Human Capital, Social Capital, Collective Intelligence. In: S. Baron / J. Field / T. Schuller (eds.), Social Capital: critical perspectives (Oxford, New York 2000) 226–242. Carley / Svoboda 1996 K. M. Carley / D. M. Svoboda, Modeling Organizational Adaptation as a Simulated Annealing Process. Sociological Methods and Research 25, 1996, 138–168. Carneiro 1970 R. Carneiro, A Theory of the Origin of the State. Science 169, 1970, 733–738. Christiansen / Altaweel 2006 J. H. Christiansen / M. Altaweel, Simulation of Natural and Social Process Interactions: An Example from Bronze Age Mesopotamia. Social Science Computer Review 24, No. 2, 2006, 209–226. Cioffi-Revilla 1998 Cl. Cioffi-Revilla, Politics and Uncertainty: Theory, Models and Applications (Cambridge, New York 1998). Cioffi-Revilla 2002 Cl. Cioffi-Revilla, Invariance and universality in social agent-based simulations. In: Proceedings of the National Academy of Science of the U.S.A. 99, Suppl. 3 (Washington 2002) 7314–7316. Cioffi-Revilla 2005 Cl. Cioffi-Revilla, A Canonical Theory of Origins and Development of Social Complexity. Mathematical Sociology 29, 2005, 133–153. Cioffi-Revilla et al. 2007 Cl. Cioffi-Revilla / S. Luke / D. C. Parker / J. D. Rogers / W. Fitzhugh / W. Honeychurch / B. Frohlich / P. DePriest / Ch. Amartuvshin, Agent-based Modeling Simulation of Social Adaptation and Long-Term Change in Inner Asia. In: S. Takahashi / D. Sallach / J. Rouchier (eds.), Advancing Social Simulation: The First World Congress in Social Simulation (Tokyo, New York, Heidelberg, Berlin 2007) 189–200. Coleman 1988 J. S. Coleman, Social Capital in the Creation of Human Capital. American Journal of Sociology 94, Suppl. (Chicago 1988) S95–S120.
Crossley 1997 P. K. Crossley, The Manchus (Oxford 1997) Dawson 1955 Chr. Dawson, The Mongol Mission: Narratives and Letters of the Franciscan Missionaries in Mongolia and China in the Thirteenth and Fourteenth Centuries (New York 1955). De Marrais et al. 1996 E. De Marrais / L. J. Castillo / T. Earle, Ideology, materialization, and power strategies. Current Anthropology 37, 1996, 15–31. Di Cosmo 1999 N. Di Cosmo, State Formation and Periodization in Inner Asian History. Journal of World History 10, 1999, 1–40. Earle 1997 T. Earle, How Chiefs Come to Power: The Political Economy in Prehistory (Stanford 1997). Friedman / Rowlands 1977 J. Friedman / M. Rowlands, Notes towards an Epigenetic Model of the Evolution of ‘Civilization’. In: J. Friedman / M. Rowlands (eds.), The Evolution of Social Systems. Proceedings of a meeting of the Research Seminar in Archaeology and Related Subjects held at the Institute of Archaeology, London University (London 1977) 201–276. Golden 1982 P. B. Golden, Imperial ideology and the sources of unity among the pre-Činggisid nomads of Western Eurasia. Archivum Eurasiae Medii Aevi II, 1982, 37–76. Honeychurch / Amartuvshin 2006 W. Honeychurch / Ch. Amartuvshin, States on Horseback: The Rise of Inner Asian Confederations and Empires. In: M. T. Stark (ed.), Archaeology of Asia. Blackwell studies in global archaeology 7 (Malden 2006) 255–278. Honeychurch / Amartuvshin 2007 W. Honeychurch / Ch. Amartuvshin, Hinterlands, Urban Centers, and Mobile Settings: The “New” Old World Archaeology from the Eurasian Steppe. Asian Perspectives 46, 2007, 36–64.
Expanding Empires and a Theory of Change Kawamura et al. 2005 K. Kawamura / T. Akiyama / H. Yokota / M. Tsutsumi / O. Watanabe / S. Wang, Comparing MODIS Vegetation Indices with AVHRR NDVI for Monitoring the Forage Quantity and Quality in Inner Mongolia Grassland, China. Grassland Science 51, 2005, 33–40. Kleiber / Kotz 2003 Chr. Kleiber / S. Kotz, Statistical Size Distributions in Economics and Actuarial Sciences. Wiley series in probability and statistics (Hoboken 2003). Kradin / Skrynnikova 2006 N. N. Kradin / T. Skrynnikova, Why do we call Chinggis Khan’s Polity “An Empire”? Ab Imperio 5, 2006, 89–118. Kradin et al. 2003 N. N. Kradin / D. M. Bondarenko / T. J. Barfield (eds.), Nomadic Pathways in Social Evolution. The civilization dimension series 5 (Moskva 2003). Latek et al. 2007 M. Latek / W. Honeychurch / M. Tsvetovat, Computational Empires: Hierarchies, Power and Warfare in Ancient Inner Asia. Unpublished paper on file, Center for Social Complexity, George Mason University, Fairfax 2007. Lichbach 1996 M. I. Lichbach, The Cooperator’s Dilemma. Economics, cognition, and society (Ann Arbor 1996). Lot-Falck 1956 E. Lot-Falk, A Propos d’Ätugän, Déese Mongole de la Terre. Revue de l’Histoire des Religions 149, 1956, 145–196. Luke et al. 2005 S. Luke / Cl. Cioffi-Revilla / L. Panait / K. Sullivan, MASON: A Java Multi-Agent Simulation Environment. Simulation: Transactions of the Society for Modeling and Simulation International 81, 2005, 517–527.
459
O’Madadhain et al. 2006 J. O’Madadhain / D. Fisher / S. White / Yan-Biao Boey, The JUNG (Jave Universal Network / Graph) Framework (Irvine 2006). Riker 1962 W. H. Riker, The Theory of Political Coalitions (New Haven 1962). Rogers 1991 J. D. Rogers, Patterns of Change on the Western Margin of the Southeast, A. D. 600–900. In: M. S. Nassaney / Ch. R. Cobb (eds.), Stability, Transformation, and Variation: The Late Woodland Southeast (New York 1991) 221–248. Rogers 2007 J. D. Rogers, The Contingencies of State formation in Eastern Inner Asia. Asian Perspectives 46, 2007, 249–274. Rogers 2010 J. D. Rogers, Inner Asian States and Empires: Theories, Data and Synthesis. Unpublished paper on file, Department of Anthropology, Smithsonian Institution, Washington, DC 2010. Rogers et al. 2005 J. D. Rogers / U. Erdenebat / M. Gallon, Urban Centers and the Emergence of Empires in Eastern Inner Asia. Antiquity 79, 2005, 801–818. Scott 1998 J. C. Scott, Seeing like a State: How Certain Schemes to Improve the Human Condition Have Failed. Yale agrarian studies (New Haven, London 1998). Turchin 2003a P. Turchin, Historical Dynamics: Why States Rise and Fall. Princeton studies in complexity (Princeton, Oxford 2003). Turchin 2003b P. Turchin, Complex population dynamics: A theoretical / empirical synthesis. Monographs in population biology 35 (Princeton 2003).
