The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads 178969406X, 9781789694062

Citation preview

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Edited by

Alison V.G. Betts, Marika Vicziany, Peter Jia and Angelo Andrea Di Castro

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Edited by

Alison V.G. Betts, Marika Vicziany, Peter Jia and Angelo Andrea Di Castro

Archaeopress Archaeology

Archaeopress Publishing Ltd Summertown Pavilion 18-24 Middle Way Summertown Oxford OX2 7LG www.archaeopress.com

ISBN 978-1-78969-406-2 ISBN 978-1-78969-407-9 (e-Pdf)

© Authors and Archaeopress 2019 Cover images come from the following chapters in this book: Chapter 3, Figure 79. Xiaohe Cemetery: Painted wooden face (photo Xinjiang Institute of Archaeology); Chapter 4, Figure 5. Adunqiaolu: House Site F1 (photo Chinese Academy of Social Sciences); Chapter 7, Figure 5. Berel (Altay, Kazakhstan): horse tack ornament in gilded wood in the shape of griffin and mountain sheep with curling mane of Achaemenid inspiration (photo Mission Archéologique Française en Asie centrale – MAFAC).

All rights reserved. No part of this book may be reproduced, or transmitted, in any form or by any means, electronic, mechanical, photocopying or otherwise, without the prior written permission of the copyright owners. The figures showing the location of the archaeological sites in this book remain the copyright of the editors unless written permission is given to reuse or republish them. Permission is to be sought by writing to [email protected] Printed in England by Severn, Gloucester This book is available direct from Archaeopress or from our website www.archaeopress.com

Contents List of Figures������������������������������������������������������������������������������������������������������������������������������������������������������������������������������� ii Preface����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������� vii Contributors������������������������������������������������������������������������������������������������������������������������������������������������������������������������������ viii Chapter 1 Ancient Xinjiang at the International Crossroads������������������������������������������������������������������������������������������������������������1 Marika Vicziany Chapter 2 Xinjiang in Prehistory���������������������������������������������������������������������������������������������������������������������������������������������������������������9 Alison Betts Chapter 3 The Xiaohe (Small River) Cemetery and the Xiaohe Culture���������������������������������������������������������������������������������������19 Yidilisi Abuduresule, Wenying Li, Xingjun Hu Chapter 4 Tianshan as a Bridge: New Studies of Bronze Age Archaeology in the Western Tianshan, Xinjiang, China��������������������������������������������������������������������������������������������������������������������������������������������������������������������52 Dexin Cong Chapter 5 A Report on Archaeological Findings in the Upper Yili Valley������������������������������������������������������������������������������������64 Xuetang Liu Chapter 6 Painted Pottery of the Late Bronze Age – Early Iron Age in Central Asia: New Data from Southern Uzbekistan����������������������������������������������������������������������������������������������������������������������������������������������������������������������������������84 Nikolaus G.O. Boroffka and Leonid M. Sverchkov Chapter 7 Shifting Exchange Patterns During the Bronze and Iron Ages Between China and the West in Eurasia��������97 Henri-Paul Francfort Chapter 8 The Kashgar Oasis: Reassessing the Historical Record������������������������������������������������������������������������������������������������113 Marika Vicziany and Angelo Andrea Di Castro Chapter 9 The Kashgar Oasis in Buddhist and Pre Buddhist Times: the Archaeological and Environmental Record����������������������������������������������������������������������������������������������������������������������������������������������������������143 Angelo Andrea Di Castro, Marika Vicziany and Xuan Zhu Chapter 10 Unravelling Farming and Metallurgy in Ancient China with Nuclear Science�����������������������������������������������������171 John Dodson, Pia Atahan and Xiaoqiang Li Chapter 11 Early Wheat Cultivation and Plant Use in Xinjiang Prehistory: New Evidence Based on Starch Analysis����181 Peter Jia and Florence Chau Index�������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������200 i

List of Figures Chapter 1 M. Vicziany: Ancient Xinjiang at the International Crossroads

Figure 1. Diagram showing some of the key archaeological sites and locations referred to in this chapter........................................ 1

Chapter 2 A. Betts: Xinjiang in Prehistory

Figure 1. Diagram showing the key Xinjiang sites discussed in this chapter............................................................................................. 9 Figure 2a and b. View of the pre-Bronze Age site of Goukou in the north-eastern Tianshan; stone tools from Goukou................. 10 Figure 3. Qiemu’erqieke stone stela, Xinjiang Altai...................................................................................................................................... 10 Figure 4. Ceramic vessels from the Adunqiaolu cemetery, Bortala Valley............................................................................................... 12 Figure 5. Deer Stone Khirigsuur Complex, Xinjiang Altai........................................................................................................................... 13 Figure 6. Chawuhugou cemetery, painted jugs............................................................................................................................................. 14

Chapter 3 Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture

Figure 1. Diagram showing the distribution of the Xiaohe Culture in Xinjiang...................................................................................... 20 Figure 2. Diagram showing the location of the Xiaohe Culture Cemeteries, Regions A and B.............................................................. 20 Figure 3. Cemetery 09LE50 located on the top of a yardang........................................................................................................................ 21 Figure 4. Detail of Cemetery 09LE50............................................................................................................................................................... 21 Figure 5. Plan of Cemetery 09LE4 showing the location of the graves...................................................................................................... 22 Figure 6. Cemetery 09LE4: General view looking south............................................................................................................................... 22 Figure 7. Cemetery 09LE4: Burials 3-15, looking west.................................................................................................................................. 22 Figure 8. Cemetery 09LE4: Burial 1, looking south-west.............................................................................................................................. 23 Figure 9. View of Cemetery 09LE8 on top of the yardang............................................................................................................................ 23 Figure 10. Cemetery 09LE8: Burial 1................................................................................................................................................................ 23 Figure 11. Cemetery 09LE8: Mummified body. Burial 1............................................................................................................................... 24 Figure 12. Cemetery 09LE8: Burial 2................................................................................................................................................................ 24 Figure 13. Artefacts collected from Kongque River Delta Cemetery Burial 36........................................................................................ 25 Figure 14. Mummified body found at Kongque River Delta Cemetery Burial 36.............................................................................25 Figure 15. Basket collected from Kongque River Delta Cemetery 37......................................................................................................... 26 Figure 16. Tiebanhe Cemetery......................................................................................................................................................................... 26 Figure 17. Female mummy found at Tiebanhe Cemetery 2......................................................................................................................... 26 Figure 18. General view of the Gumuguo Cemetery..................................................................................................................................... 27 Figure 19. Radial burial patterns in the Gumuguo Cemetery..................................................................................................................... 27 Figure 20. Excavated burials at the Gumuguo Cemetery............................................................................................................................. 27 Figure 21. Evidence of looting at the Gumuguo Cemetery.......................................................................................................................... 28 Figure 22. A timber enclosed burial chamber at LT Cemetery.................................................................................................................... 28 Figure 23. Plan showing the location of graves at Xianshuiquan Cemetery No 10................................................................................. 29 Figure 24. View of Xianshuiquan Cemetery No 10....................................................................................................................................... 29 Figure 25. Timbers found at Xianshuiquan Cemetery No 10...................................................................................................................... 29 Figure 26. Carved timber found at Xianshuiquan Cemetery No 10........................................................................................................... 30 Figure 27. Basket found at Xianshuiquan Cemetery No 10......................................................................................................................... 30 Figure 28. The Xiaohe Cemetery before the excavations of 2002-2005..................................................................................................... 30 Figure 29. The Xiaohe Cemetery: view from southeast to northwest....................................................................................................... 30 Figure 30. The northern section of the Xiaohe Cemetery before excavation.......................................................................................... 31 Figure 31. A general view of the Xiaohe Cemetery before excavation...................................................................................................... 31 Figure 32. Plan of Xiaohe mound before excavation.................................................................................................................................... 32 Figure 33. Xiaohe Cemetery: timber coffins.................................................................................................................................................. 33 Figure 34. Xiaohe Cemetery: timber poles with coffins............................................................................................................................... 33 Figure 35. Xiaohe Cemetery: details of the timber poles............................................................................................................................. 33 Figure 36. Xiaohe Cemetery: Burial 75. Coffin with clay shell covering surrounded by poles.............................................................. 34 Figure 37. Xiaohe Cemetery: Burial 75. Detail of the clay shell covering of the coffin........................................................................... 34 Figure 38. Xiaohe Cemetery: The horns of cattle (Bos taurus) from Burial 13.......................................................................................... 34 Figure 39. Xiaohe Cemetery: Plan of burials and timber poles in layers 1-3............................................................................................ 35 Figure 40. Xiaohe Cemetery: Southern section. Burials in layer 2............................................................................................................. 36 Figure 41. Xiaohe Cemetery: Southern section. Burials in layers 4 and 5................................................................................................. 36 Figure 42. Survey route for the Northern Cemetery.................................................................................................................................... 37 Figure 43. Coffins in Northern Cemetery....................................................................................................................................................... 37 Figure 44. Coffin in the Northern Cemetery.................................................................................................................................................. 38 Figure 45. Rescue work in the Northern Cemetery...................................................................................................................................... 38

ii

Figure 46. Textile found in the Northern Cemetery..................................................................................................................................... 38 Figure 47. Female symbol on wooden pole in the Northern Cemetery..................................................................................................... 38 Figure 48. Basket found in the Northern cemetery...................................................................................................................................... 39 Figure 49. Phallic symbol found in the Northern Cemetery....................................................................................................................... 39 Figure 50. Xiaohe Cemetery: Burial 13 covered in cattle hide.................................................................................................................... 39 Figure 51. Xiaohe Cemetery: Burial 13. Female body................................................................................................................................... 39 Figure 52. Xiaohe Cemetery: Pole with female symbol ............................................................................................................................... 40 Figure 53. Xiaohe Cemetery: Burial 24. Poles with female and male symbols......................................................................................... 40 Figure 54. Xiaohe Cemetery: Pole with male symbol................................................................................................................................... 40 Figure 55. Xiaohe Cemetery: hafted stone axe.............................................................................................................................................. 40 Figure 56. Xiaohe Cemetery: Burials 24 and 28. Wooden arrows with feathered fletching................................................................... 40 Figure 57. Xiaohe Cemetery: Wooden arrow with bone tip........................................................................................................................ 41 Figure 58. Xiaohe Cemetery: stone mace head.............................................................................................................................................. 41 Figure 59. Xiaohe Cemetery: bronze arrowhead........................................................................................................................................... 41 Figure 60. Xiaohe Cemetery: Burial 13. Slivers cut from animal ears........................................................................................................ 41 Figure 61. Xiaohe Cemetery: Burial 75. Basket in a wooden bucket.......................................................................................................... 42 Figure 62. Xiaohe Cemetery: Woven basket................................................................................................................................................... 42 Figure 63. Xiaohe Cemetery: Wooden bucket................................................................................................................................................ 42 Figure 64. Xiaohe Cemetery: Horn and leather container.......................................................................................................................... 42 Figure 65. Xiaohe Cemetery: Burial 13. Leather bag.................................................................................................................................... 42 Figure 66. Xiaohe Cemetery: Burial 13. String and bead bracelet.............................................................................................................. 43 Figure 67. Xiaohe Cemetery: String and bead necklace............................................................................................................................... 43 Figure 68. Xiaohe Cemetery: Golden earring on a bronze mirror.............................................................................................................. 43 Figure 69. Xiaohe Cemetery: Felt hat with string, fur and feather ornament......................................................................................... 43 Figure 70. Xiaohe Cemetery: Burial 13. Leather boots................................................................................................................................. 43 Figure 71. Xiaohe Cemetery: Decorated feather........................................................................................................................................... 43 Figure 72. Xiaohe Cemetery: Phallic symbol in wood and string............................................................................................................... 44 Figure 73. Xiaohe Cemetery: Burial 24. Basket and twigs on top of coffin............................................................................................... 44 Figure 74. Xiaohe Cemetery: Short strings made of animal tendons........................................................................................................ 44 Figure 75. Xiaohe Cemetery: Burial 11. Woven textile with tassels........................................................................................................... 45 Figure 76. Xiaohe Cemetery: Burial 11. Woven textile................................................................................................................................. 45 Figure 77. Xiaohe Cemetery: Burial 13. Woven and embroidered textile................................................................................................. 45 Figure 78. Xiaohe Cemetery: Burial 13. Human face carved from wood................................................................................................... 45 Figure 79. Xiaohe Cemetery: Painted wooden face...................................................................................................................................... 45 Figure 80. Xiaohe Cemetery: Burial 2. Wooden replica of human body.................................................................................................... 45 Figure 81. Xiaohe Cemetery: Wooden figurine with clothing and earrings............................................................................................. 46 Figure 82. Xiaohe Cemetery: Wooden sticks carved with human faces and string decoration............................................................. 46 Figure 83. Table showing radiocarbon dates from the Xiaohe Cemetery: Southern Section................................................................ 47 Figure 84. Table showing radiocarbon dates from the Xiaohe Cemetery: Northern section................................................................ 48

Chapter 4 D. Cong: Tianshan as a Bridge: New Studies of Bronze Age Archaeology in the Western Tianshan, Xinjiang, China

Figure 1. Diagram showing the Bortala River area and adjacent regions................................................................................................. 52 Figure 2. General view of the site of Turigen................................................................................................................................................ 54 Figure 3. View of the site of Hu’ertuoleha from above................................................................................................................................ 54 Figure 4. Table showing radiocarbon (AMS 14C) dates for the Adunqiaolu site...................................................................................... 56 Figure 5. Adunqiaolu: House Site F1................................................................................................................................................................ 57 Figure 6. Adunqiaolu: Pottery sherds found inside House F1..................................................................................................................... 57 Figure 7. Adunqiaolu: House plan of F1, F2 and F3....................................................................................................................................... 58 Figure 8a and b. A stone cist burial at the Adunqiaolu Cemetery.............................................................................................................. 59 Figure 9. Adunqiaolu: Stone enclosure of burial SM9.................................................................................................................................. 60 Figure 10. Adunqiaolu: The northern region of the cemetery (SM4 top centre)..................................................................................... 60 Figure 11. Adunqiaolu: Burial SM4 from above............................................................................................................................................. 60 Figure 12. Adunqiaolu: Burial SM4-1.............................................................................................................................................................. 61 Figure 13. Adunqiaolu: Earring found in burial SM4-1................................................................................................................................ 61 Figure 14. Adunqiaolu: Pottery vessel found in burial SM4-1..................................................................................................................... 61 Figure 15. Adunqiaolu: Stone enclosure burial of SM50-1 and SM50-2..................................................................................................... 62

Chapter 5 X. Liu: A Report on Archaeological Findings in the Upper Yili Valley

Figure 1. Diagram showing prehistoric sites in the Upper Yili River Valley............................................................................................ 65 Figure 2. Qiongkeke: Ceramics from the settlement site............................................................................................................................. 65 Figure 3. Qiongkeke: Ceramics from the settlement site............................................................................................................................. 66 Figure 4. Qiongkeke: Microlithic cores from the settlement site............................................................................................................... 66 Figure 5. Qiongkeke: Burial in Cemetery No. 1.............................................................................................................................................. 67 Figure 6. Qiongkeke: Ceramics in Cemetery No. 1........................................................................................................................................ 68

iii

Figure 7. Qiongkeke: Ceramics in Cemetery No. 2........................................................................................................................................ 69 Figure 8. Tangbalesayi: Types of burials......................................................................................................................................................... 70 Figure 9. Tangbalesayi: Ceramics and personal ornaments from the burials.......................................................................................... 71 Figure 10. Kukesuhexi: Types of burials......................................................................................................................................................... 72 Figure 11. Kukesuhexi: Ceramics from the burials....................................................................................................................................... 72 Figure 12. Kalasu: Plan of house F1................................................................................................................................................................. 73 Figure 13. Kalasu: Stone artefacts from house F1......................................................................................................................................... 73 Figure 14. Kalasu: Ceramics from house F1.................................................................................................................................................... 74 Figure 15. Gongliu Shankou: Ceramics from the cemetery......................................................................................................................... 75 Figure 16. Wu’ertangou: Plan and cross-section of stone mound burial.................................................................................................. 75 Figure 17. Jialegesiha’yinte: Ceramics from the burials............................................................................................................................... 76 Figure 18. Bietebasitao: Ceramics and stone artefacts from the burials................................................................................................... 77 Figure 19. Jirentuohai: Ceramic and metal artefacts from the burials...................................................................................................... 78 Figure 20. Sa’erbulakegou: Ceramics from the burials................................................................................................................................. 79 Figure 21. Tiemulike-goukou: Ceramics from the burials........................................................................................................................... 79 Figure 22. Examples of the animals depicted on the rock art of the Yili Valley...................................................................................... 80 Figure 23. Examples of the animals depicted on the rock art of the Yili Valley...................................................................................... 81

Chapter 6 N. Boroffka and L. Sverchkov: Painted Pottery of the Late Bronze Age – Early Iron Age in Central Asia: New Data from Southern Uzbekistan

Figure 1. Diagram showing the distribution of Central Asia Yaz I type hand-made painted pottery.................................................. 85 Figure 2. Diagram showing the location of Maydatepa in the Bandykhan microregion........................................................................ 85 Figure 3. Excavation view (from the east) of Yaz I architecture in the upper layers of Maydatepa..................................................... 86 Figure 4. Detail of a complex fireplace in the Yaz I settlement of Maydatepa......................................................................................... 86 Figure 5. Partial view of the southern profile at Maydatepa...................................................................................................................... 87 Figure 6. Maydatepa: Hand-made Yaz I type pottery................................................................................................................................... 88 Figure 7. Maydatepa: Hand-made Yaz I type pottery................................................................................................................................... 89 Figure 8. Maydatepa: Wheel-made pottery from Yaz I contexts................................................................................................................ 90 Figure 9. Maydatepa: Various finds from Yaz I contexts............................................................................................................................. 91

Chapter 7 H.-P. Francfort: Shifting Exchange Patterns During the Bronze and Iron Ages Between China and the West in Eurasia

Figure 1. Diagram showing the Eurasian cultural areas and sites of the Bronze Age discussed in this chapter................................ 97 Figure 2. Sarazm (Tajikistan): Stone implements of steppic (staff) or ‘oasis’ (with handles) types..................................................... 98 Figure 3. Sarazm (Tajikistan): Painted pottery of ‘Balochi’ type................................................................................................................ 98 Figure 4. Diagram showing some of the Eurasian sites of the Iron Age discussed in this chapter..................................................... 101 Figure 5. Berel (Altay, Kazakhstan): Kurgan No. 11 horse tack ornament.............................................................................................. 101 Figure 6. Berel (Altay, Kazakhstan): Kurgan No. 11 horse wooden ornament........................................................................................ 101 Figure 7. Diagram showing main branches of the Silk Roads................................................................................................................... 103 Figure 8. Gandharan toilet tray with royal wine filtering and drinking ceremony seen frontally.................................................... 104 Figure 9. Noin-Ula (Mongolia) tapestry with royal wine filtering and drinking ceremony seen in profile...................................... 105

Chapter 8 M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record

Figure 1. Diagram showing the location of the eight main archaeological sites in the Kashgar oasis.............................................. 115 Figure 2. Former location of Pigeon House near Khan-oi......................................................................................................................... 116 Figure 3. Road sign placed near the original location of the Kizil Debe Stupa in contemporary Kashgar........................................ 116 Figure 4. Satellite view of the area of Khan-oi in 2006............................................................................................................................... 117 Figure 5. Agricultural encroachment on Khan-oi in 2012......................................................................................................................... 117 Figure 6 a and b. Satellite images of the site of Topa Tim in 2005 (a) and 2013 (b)............................................................................... 118 Figure 7. The stupa of Topa Tim in the background, 2007........................................................................................................................ 118 Figure 8. Satellite image showing the line of karez adjacent to the Mori Tim sacred area.................................................................. 119 Figure 9. One of the 48 karez wells in the vicinity of the sacred area of Mori Tim................................................................................ 119 Figure 10. Bulldozers removing part of the old wall of Kashgar in 2005................................................................................................. 120 Figure 11. Tomb of Satuq Boghra Khan in Artush...................................................................................................................................... 132 Figure 12. Exterior of the tomb of Bibi Maryam with the female caretaker.......................................................................................... 132 Figure 13. Interior of the tomb of Bibi Maryam showing the tomb with ritual rods............................................................................ 133 Figure 14. Desert Shrine of Ordam Padshah in 1928. This is the grave (mazar) of Ali Arslan Khan.................................................... 134 Figure 15. Remains of animal sacrifices outside the tomb of Satuq Boghra Khan................................................................................ 134 Figure 16. Exterior of the Tomb of Apa Khoja in Kashgar......................................................................................................................... 135 Figure 17. Interior of the Tomb of Apa Khoja.............................................................................................................................................. 136 Figure 18. Lithograph c. 1890 showing the types of costumes.................................................................................................................. 137 Figure 19. Detail of the inscription on the cauldron outside Tomb of Satuq Bogra Khan in Artush.................................................. 137

iv

Chapter 9 A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times: the Archaeological and Environmental Record

Figure 1. Diagram showing some of the key archaeological sites discussed in this chapter............................................................... 144 Figure 2a and b. Aerial view of the site of Yawaluk/Tegurman in 2002 (a) and in 2012 (b)................................................................. 146 Figure 3. Plan of the ruins of Yawaluk/Tegurman surveyed by Pelliot in 1906..................................................................................... 146 Figure 4. Mori Tim stupa A on the right and stupa B on the left............................................................................................................. 147 Figure 5. Fragment of a figurative stucco from the debris of the monastic site of Topa Tim............................................................. 147 Figure 6. Tractor marks on the old structures of Khan-oi in 2013........................................................................................................... 148 Figure 7. Encroachment of irrigation canals on the old structures of Khan-oi in 2013....................................................................... 149 Figure 8. Sand dunes covering up part of the cemetery at Yopurga....................................................................................................... 149 Figure 9. Plan of the ruins of Eskishahar surveyed by Pelliot in 1906..................................................................................................... 150 Figure 10. Image of the ruins of Eskishahar as it appeared during the Pelliot expedition in 1906.................................................... 151 Figure 11. Remains of the walls of Eskishahar in 2013............................................................................................................................... 151 Figure 12. The caves of the Three Immortal Buddhas, on the cliffs above the Chakmak river........................................................... 154 Figure 13. Plan and cross section of the caves of the Three Immortal Buddhas surveyed by Pelliot in 1906................................... 154 Figure 14. Fragments of a Buddhist manuscript discovered in Yawaluk................................................................................................ 155 Figure 15. Water well in old Kashgar in 2014............................................................................................................................................... 158 Figure 16. Water logged fields around Topa Tim in 2005.......................................................................................................................... 159 Figure 17. Ruins of a Buddhist stupa and monastery (Toukguzkaznakh)............................................................................................... 162

Chapter 10 J. Dodson, P. Atahan and X. Li: Unravelling Farming and Metallurgy in Ancient China with Nuclear Science

Figure 1. Diagrammatic representation of the three main isotopes of Carbon..................................................................................... 171 Figure 2. General rules of isotopic fractionation........................................................................................................................................ 172 Figure 3. Representation of the basis of the radiocarbon dating method.............................................................................................. 172 Figure 4. Diagram showing Eurasia and Gansu Province........................................................................................................................... 174 Figure 5. Photograph showing surface scatter at Huoshiliang site......................................................................................................... 175 Figure 6. Cation profiles in Huoshiliang sediments.................................................................................................................................... 175 Figure 7. Diagram showing the distribution of early wheat sites with robust AMS radiocarbon dating.......................................... 176 Figure 8. Table showing AMS 14C dates for samples taken from the surface scatters at Huoshiliang and Gangangwa sites........ 178 Figure 9. Diagram showing plot of δ13C and δ15N data................................................................................................................................ 178

Chapter 11 P. Jia and F. Chau: Early Wheat Cultivation and Plant Use in Xinjiang Prehistory: New Evidence Based on Starch Analysis

Figure 1. Diagram showing sites discussed in this chapter....................................................................................................................... 182 Figure 2. Table showing maximum length (mm) of starch granules....................................................................................................... 183 Figure 3. Table showing boxplots of starch granule maximum length of the comparative reference collection data................... 185 Figure 4. Starch granules of Leguminosae family....................................................................................................................................... 186 Figure 5. Gramineae starch............................................................................................................................................................................. 186 Figure 6. Other species of starch plants....................................................................................................................................................... 187 Figure 7. Stone and specimen collection site.............................................................................................................................................. 187 Figure 8. Specimen collection and stone tools............................................................................................................................................ 188 Figure 9. Table showing archaeological samples. Boxplots of starch grain maximum length............................................................ 188 Figure 10. Wheat starch recovered from stone mortar.............................................................................................................................. 189 Figure 11. Wheat starch recovered from grinding stone........................................................................................................................... 189 Figure 12. Wheat starch recovered from grinding stone at Kalasu site.................................................................................................. 189 Figure 13. Modern reference granules of wheat and barley..................................................................................................................... 190 Figure 14. Modern reference starch granules: left – oats, right – broomcorn millet............................................................................ 191 Figure 15. Starch granules with multilateral/polyhedral shape found at Kalasu site.......................................................................... 191 Figure 16. Modern reference starch granules – sorghum.......................................................................................................................... 192 Figure 17. Modern reference starch granules – buckwheat...................................................................................................................... 192 Figure 18. Ancient starch granules................................................................................................................................................................ 193 Figure 19. Possible suo-yang (Songaria Cynomorium) starch granules...................................................................................................... 194 Figure 20. Starch granules of da-yun (Cistanche deserticola) extracted from modern reference plant............................................... 194 Figure 21. Da-yun (Cistanche deserticola) starch granules........................................................................................................................... 195 Figure 22. Ancient da-yun (Cistanche deserticola) starch granules found from stone tools unearthed at Luanzagangzi site.......... 196 Figure 23. Table showing plant starches identified from the archaeological sites mentioned in this chapter................................ 196

v

Preface This book had its origins in a collaborative project between the University of Sydney and the Archaeological Institute of the Chinese Academy of Social Sciences in Beijing. In 2006 Alison Betts and Peter Jia began, with Chinese colleagues from CASS, a study of the impact of the cultures of the Eurasian steppe on the rise of the state in central China through exploration of the prehistory of Xinjiang. In 2014 the team was enhanced by additional collaborators from Monash University, Marika Vicziany and Angelo Andrea Di Castro, as well as John Dodson, then Head of the Institute for Environmental Research, Australian Nuclear Science and Technology Organisation (ANSTO). Funding for collaborative fieldwork from the Australian side has been provided by the Australian Research Council through two major Discovery grants, East meets West: an archaeological study of early contact between China and Eurasia (DP0770997: 2007-2010) and The Silk Roads in the Bronze Age: critical links between Eurasia and China (DP150100121: 2015-2018). Additional funding with smaller grants was provided by The University of Sydney (through the China Studies Centre and the Bridging Grants Scheme), Monash University’s School of Philosophical, Historical and International Studies (Faculty of Arts) and ANSTO.

The papers included in this volume are based on a selection of lectures presented at a workshop, East and West: Past and Future, sponsored by the China Studies Centre at the University of Sydney in May 2012. The volume presents a fresh framework for the prehistory and early history of Xinjiang that will be of value and interest to all readers concerned with the role of Inner Asia in world history. In addition to thanking our European and Australian collaborators for their unique contributions, we are grateful to our Chinese colleagues for generously sharing with us their ground-breaking original chapters about new discoveries based on recent fieldwork in Xinjiang. The editors are grateful to the Chinese Academy of Social Sciences (Beijing), the China Studies Centre at Sydney University and the Arts Faculty at Monash University for their continued support of research on the early history and cultures of Xinjiang. We are also grateful to Dr Uri Gilad for his support in preparing the 16 figures showing the location of the archaeological sites referred to in this volume. Funding for Dr Gilad’s work came from the ARC project on The Silk Roads in the Bronze Age (DP150100121: 2015-2018). And finally, we thank the editorial team at Archaeopress for their assistance in preparing these papers for publication. We are especially grateful for the support and patience showed by Dr David Davison. Alison Betts, Marika Vicziany, Peter Jia, Angelo Andrea Di Castro

vi

Contributors

Yidilisi Abuduresule is the former Director of the Xinjiang Bureau of Relics and Director of Excavations at the world famous Bronze Age cemetery of Xiaohe in the Taklamakan Desert.

John Dodson is the former head of Environmental Sciences at the Australian Nuclear Science and Technology Organisation (ANSTO). He has undertaken extensive environmental field research in China and is currently based in Xian as a research professor with the Institute of Earth Environment, CAS.

Pia Atahan has a PhD in environmental change from the University of Western Australia and until recently worked as a scientist at the Australian Nuclear Science and Technology Organisation in Sydney. Her research focused on early agriculture in China and the use of isotopes for unravelling animal and human diets from skeletal remains in China.

Henri-Paul Francfort is a Member of the Institute of the Académie des Inscriptions et Belles-Lettres. He is Emeritus Director of Research for the CNRS – UMR 7041 ‘ArScAn’, The Archaeology of Central Asia. He has had a long and distinguished career in the archaeology of Central Asia, particularly Afghanistan and more recently Tadjikistan.

Alison Betts is Professor of Silk Road Studies at the University of Sydney. She has been working on the archaeology of Central Asia for more than two decades. Over the past decade she has been working with Jia on a collaborative field programme with the Institute of Archaeology, Chinese Academy of Social Sciences to study the Bronze Age of Xinjiang and the influence of the Eurasian Bronze Age on early Chinese cultural development.

Xingjun Hu is a Professor at the Xinjiang Institute of Cultural Relics and Archaeology. As a Director of Fieldwork, he has worked in Xinjiang for many years. Most recently he has directed the excavation of a cemetery with Afanasievo burial features. Peter Weiming Jia is a Senior ARC Research Fellow at the University of Sydney. He has worked on starch residue analysis and other archaeobotanical studies in Xinjiang. He has been involved in archaeological fieldwork on the Bronze Age in Xinjiang for more than a decade.

Nikolaus G.O. Boroffka is a senior member of the Eurasia-Department, German Archaeological Institute/ Deutsches Archäologisches Institut in Berlin. He has undertaken extensive archaeological fieldwork across Central Asia and Eastern Europe. He has a wide range of research interests including mining, metallurgy and environment-human interactions.

Wenying Li is Professor and the Director of Xinjiang Institute of Cultural Relics and Archaeology. She has worked extensively on Xinjiang’s late prehistoric archaeology. Amongst her other work she has been involved in the excavation of the Xiaohe cemetery and is the only female archaeologist on the expedition team.

Florence Chau studied a Bachelor of Science and Bachelor of Arts (Honours) at the University of Sydney, majoring in Botany, Archaeology and English Literature. She is currently working at the University of Sydney and the Australian National Maritime Museum.

Xiaoqiang Li is the Deputy Director of the Key Laboratory of Vertebrate Evolution and Human Origins, Institute of Vertebrate Palaeontology and Palaeoanthropology of Chinese Academy of Science. He is also a Professor at the Graduate University of Chinese Academy of Science. He has worked in the field of palaeo-environmental and environmental archaeology in China for many years. One of his teams is working on the Xinjiang and Central Asia area.

Dexin Cong is the Director of the Frontier Archaeological Department, Archaeological Institute, Chinese Academy of Social Sciences and co-Director of the Adunqiaolu Project at Bortala Valley, a major longterm study of late prehistoric archaeology in Xinjiang, China. Angelo Andrea Di Castro is Adjunct Research Fellow of the Centre for Ancient Cultures of Monash University. He has worked extensively on archaeological sites in Italy, Nepal, Australia and China and published various books and academic papers. His recent research projects include the cultural exchanges along the Silk Road, the archaeology of the Kashgar oasis and the goddess cults of Gandhara.

Xuetang Liu is Professor and the Head of the School of Ethnology and Sociology, Xinjiang Normal University. He was the Director of a large scale archaeological excavation in the Yili River Valley at the Xinjiang Institute of Cultural relics and Archaeology before moving to the university. vii

Leonid M. Sverchkov is a senior member of the Fine Arts Scientific Research Institute of the Academy of Sciences of Uzbekistan/San’atshunoslik Ilmiy-Tadqiqot Instituti, O’zbekistan Badiiy Akademiyasi in Tashkent/ Toshkent. He has undertaken extensive archaeological fieldwork in Central Asia. His many publications range from the history of the Neolithic to the Early Middle Ages.

has worked on Xinjiang since a consultancy with the Asian Development Bank and the State Council of China in 2000. Her two major projects have been about the Buddhist culture and monuments of Kashgar and the Bronze Age in the Bortala Valley. In 2008 she published a monograph with George Michell, Yen Hu Tsui and John Gollings: Kashgar-Oasis City on China’s Old Silk Road (London: Frances Lincoln).

Marika Vicziany is Professor Emerita and Director of the National Centre for South Asian Studies (Monash Asia Initiative) in the Arts Faculty at Monash University and a specialist on long term economic development in Asia. In addition to her work on South Asia, Vicziany

Xuan Zhu is a Senior Lecturer in the School of Earth, Atmosphere and Environment at Monash University and specialises in environmental decision analysis and modelling, geographical information systems and remote sensing.

viii

Chapter 1

Ancient Xinjiang at the International Crossroads Marika Vicziany that the mummies stood as an isolated phenomenon, lacking a broader context to explain their culture’s unique features. In the intervening years, many studies on specific aspects of the mummies have been published. We now know that their DNA makeup signalled their largely Eurasian ancestry (Li et al. 2010, 2015) rather than the earlier view that the mummies were of European origins (Romgard 2008: 21, 36). Moreover, in the Li 2010 study, the mixing of Asian and European features has been traced to southern Siberia. DNA analysis, in other words, suggests far longer migration routes and not only in one direction. The foods of the mummies have also been studied (Yang 2014; Yang et al. 2014), as have their cattle (see Chapter 3 in this volume; Mai et al. 2016), showing that their domesticated antecedents came from both Western Asia and China. So far, however, core data on the archaeology of key sites such as Xiaohe have not been published and without this the other studies are essentially disembodied specialist reports, without a broad and meaningful context. Our

This book unveils the ancient secrets of Xinjiang, western China, one of the least known but culturally rich and complex regions of the heart of Asia. Xinjiang is perhaps most famous for its lethal desert, the Taklamakan (Figure 1), its name in the local Uyghur language translating as ‘You can go in, but you will never come out’. For those who have a good grasp of geography it may also be known as the core part of the routes along the ancient Silk Roads, hazardous paths from one small patch of green oasis to another, days apart across perilous sandy wastes. In recent years Xinjiang has also become known for its so-called ‘mummies’, remarkably preserved bodies protected from decay by heat and saline sands. Some two decades have passed since news of the remarkable Tarim Basin mummies was first publicised outside China (Barber 1999; Mallory and Mair 2000), but as the world was introduced to this extraordinary Bronze Age marvel, the lack of further archaeological information on Xinjiang and surrounding lands meant

Figure 1. Diagram showing some of the key archaeological sites and locations referred to in this chapter.

1

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads reports, without a broad and meaningful context. Our volume contains the long awaited first authoritative review of the archaeological framework for Xiaohe and related sites that constitute the Xiaohe culture, by its chief scholars Yidilis Abuduresule, Wenying Li and Xingjun Hu. The importance of Xiaohe lies in its age, for together with Qiemu’erqieke and Tianshanbeilu it is one of the three earliest Bronze Age cultures of Xinjiang (see next chapter by Betts).

by sand storms, the structure might well have been a tomb when it was first built. One can suggest that the presence of dozens of goat and oxen horns and dried snakes and eagles around the cabin/tomb (Abuduresule 2008: slides 97 and 101) points to a totemic culture involving a person of power (Abuduresule 2008: slide 98 showing an ‘Unearthed Marble-head of a Staff of Authority’). Abduresule’s hypothesis assumed that Australian indigenous people lived only in temporary grasswooden shelters, a view that was encouraged by other similarities he observed between Aboriginal and Xiaohe peoples in their animal skin clothing and weapons. If the Xiaohe people lived in temporary dwellings, it was hardly surprising that no evidence of this remained. On the other hand, if Hedin’s account of living conditions in Lop Country are to be believed, the dwellings might have been substantial and permanent, but made of materials that have not survived into the present day. When the Swedish explorer Hedin mapped the Tarim River at the turn of the 20th century he reported on the reeds used by the local people (‘Lopniks’) along the river banks of Lop country as construction material and fibre for clothing. These reed structures were substantial and towered over the local residents. Hedin’s photographs do not suggest that these were temporary structures.1 Perhaps reeds were also the primary building materials used by the people of the Xiaohe culture at a time when the water resources of the Xiaohe areas were plentiful?

Chapter 3 reveals the unexpected geographical spread of the Xiaohe culture in the eastern and southern parts of the Taklamakan desert, a culture which despite its many common features also incorporated unique characteristics at particular sites such as the Northern Cemetery in the lower Keriya River area. The Xiaohe culture in its various iterations remains a puzzle for Abduduresule and his colleagues, as explained to us during his first visit to Australia in 2008. He was attracted to the indigenous Australian collections at the Museum of Victoria because he hoped to solve the problem of why the houses and other buildings used by the Xiaohe people have not survived, even in ruined form. The Xiaohe culture is understood entirely from coffins, the mummies and their grave goods, although recently work has also been undertaken to study the changing environment of the region (Li et al. 2013). Until his visit to Melbourne, Abuduresule did not know that research on Australian indigenous cultures had in recent years moved away from notions about nomadic, wandering Australian Aborigines and paid greater attention to permanent structures including the fish dams and basalt stone houses used by indigenous Gunditjmara people in the Condah Lake areas of southwestern Victoria (McNiven 2015; Mirtschin 2013). As a result of new research, indigenous Australian subsistence economies have increasingly come to be defined as societies of ‘hunter-gatherer-cultivators’ (Keen 2004: 95-96). Recognition of the ingenuity of Aboriginal engineering and environmental management has accelerated since the earliest eel dams of Lake Condah were dated back to some 6,600 years (Hinchliffe 2018; McNiven 2015).

Thanks to new research reported in the present collection, we know that the Bronze Age people used a variety of construction materials determined by the nature of locally available materials. In Chapter 4 by Cong, we have documentation about the enormous size of the Andronovo-type houses which were built by the Bronze Age residents of Adunquiaolu. These houses used the plentiful supply of rocks in the upper Bortala Valley, in contrast to the environmental conditions of the Taklamakan desert-based sites of Xiaohe. Irrespective of the size of houses or the building materials used, the permanence of such habitations does not mean that the Bronze Age peoples resided there all year around.

The lack of any comparable discoveries about the nature and patterns of human settlements in the Xiaohe culture has only served to increase the puzzles about the lives of the Bronze Age peoples of the Taklamakan desert. The largest structure found so far in the case of the Xiaohe sites is a sizeable wooden ‘hut’. In 2008, Abuduresule at a public lecture in Melbourne speculated that this ‘mysterious cabin’ might have been constructed for the use of a local shaman who may have officiated at the Xiaohe burials (Abuduresule 2008: slides 94, 95, 96, 98). The illustration, however, describes this large cabin as a ‘special tomb,’ although no human remains were found in it. Given the history of looting in the area, together with physical degradation caused

Chapter 3 by Abuduresule et al. leads us into the first part of this collection which includes the work of other Chinese archaeologists in Chapters 4 and 5. These three chapters are based on excavation reports that have never before been published in English except in summary Hedin’s report of his journey down the Tarim River into Lop Country speaks of the extensive use of the river reeds which were so high and prolific that on at least one occasion his crew had to burn off a considerable area in order to clear a channel through the waterways. His account is accompanied by numerous photographs of his workmen and the local villagers standing in front of their substantial reed huts towering over the heads of the residents (Hedin 1903: 321, 325, 420-423, 437,439, 443, 449).

1 

2

M. Vicziany: Ancient Xinjiang at the International Crossroads form for a few international conferences. Publication of this work in this volume was a time consuming process and had to await, in the first instance, completion of the individual reports to the Chinese Bureau of Cultural Relics. Once that had happened, the difficult task of translation, checking and cross checking began in order to ensure that these unique chapters are as accurate as possible and unambiguous in their statements. It was essential to do this because other non-Chinese writings have sometimes misunderstood and then misrepresented the nature of the evidence. Chapter 3, therefore, represents the first, much awaited, full official summary of the excavations and cultural context of the sites by the primary excavators involved in elucidating the evidence thrown up by the remarkable excavations of the Bronze Age cemeteries of the Xiaohe culture with their extraordinary levels of organic preservation. The arduous task of collecting and analysing the scattered evidence has taken more than 30 years. The people of Xiaohe emerge from this analysis as remarkable humans whose daily habits, judging by their grave goods, bear an uncanny resemblance to modern day living. Their animal skin boots, as protection against the extreme cold of the winters, are little different from modern day ugg boots, while the ephedra twigs carried in their grass-woven pouches remind us of the medicinal properties of the same plant which remains an essential component of the modern nasal and sinus decongestant called Sudafed. Even the name Sudafed is based on the word ephedra, a herb that has many properties including the ability to act as a dramatic stimulant. Perhaps the Xiaohe peoples followed a form of shamanism in which ephedera’s medicinal properties helped them to achieve heightened levels of awareness?

strong relationships between the protection provided in the deep winters to the settlers and the care of their valuable animal herds in this shared accommodation. Liu’s chapter complements that by Cong in focussing away from the Bortala River Valley to archaeological sites in the upper Yili River Valley. Together, the broadening perspective provided by Liu and Cong are dramatically enhancing our knowledge of the prehistory of the western Chinese regions in going beyond the previous focus of research on the desert areas of Xinjiang. Not only are these scholars providing geographical and topographical benchmarks for documenting the Bronze Age, they are also starting to map out the unique connection of these settlement sites, graves and rock paintings to the western, southern and northern regions beyond China – a theme also elaborated by Francfort in his contribution to this volume. More specifically, Liu presents a broad chronology for the known sites of the Bronze and Iron Ages based on excavations at Qiongkeke. Qiongkeke, he suggests, provides a base reference point by which to assess other archaeological sites in the Yili Valley that appear to be interrelated. It is significant that in these first ever investigations into the pre-history of the far western part of Xinjiang both Cong and Liu provide not only detailed reports about their findings but also benchmarks for human settlement patterns stretching beyond China into eastern Kazakhstan. This painstaking, comparative documentation is helping us to understand how the Bortala and Yili river valleys served as gateways from Eurasia into western Xinjiang and to the far east. Exactly when and how these Bronze Age settlements spread across such a wide area is still, however, to be revealed by new research. Chapter 2 by Betts in this collection addresses some of this by providing an overview of the important cultural connections that linked up the human settlements of the vast spaces of Xinjiang during the Bronze Age.

The next two chapters by Cong and Liu shift our attention away from the eastern, desert regions of Xinjiang to the even less studied western, river valleys defined by the glacier-covered Tianshan Mountains that separate China from Kazakhstan. Cong focuses on one of the most important new archaeological excavations in western China, namely the site of Adunqiaolu, some 17 kms from the border of China and Kazakhstan, in the Bortala River Valley.2 He provides the first properly documented, absolute dates for the Bronze Age in this part of Xinjiang. The characteristics of the excavated mounds, houses and cemeteries bear a strong relationship to excavated sites further to the west, beyond the borders of China rather than to the eastern parts of Xinjiang, specifically with those of the Andronovo cultural style (see also Ruan 2013). The huge houses and the width of the walls suggest

Bronze Age settlers, it seems, protected their precious animals and other assets not only from the weather as described by Cong but also other threats to their security. A paper published by Jia et al. in 2018 describes the ruins of three fortified hills found only a few years ago in the Bortala River Valley, dating from the Late Bronze Age. They are located at strategic positions that provide a look-out over expansive, green fields, one of the ‘richest areas of seasonal pastures in the upper Bortala Valley’ (Jia 2018: 70). The authors suggest that the inhabitants of these fortified hill stations were defending their interests against competitor pastoral groups. The question yet to be answered, however, is ‘who was being protected from whom’? These fortifications also reinforce the nature of Cong’s findings in the Bortala Valley, for they indicate that the large houses built by Bronze Age people were not the only construction projects of that period. Again, this

In 2015-2018 the editors of this book were involved in an Australian Research Council funded project, DP150100121, that collaborated with Professor Cong and his team in analysing the evidence uncovered by the excavations in Adunqiaolu and bringing the findings to international attention.

2 

3

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads new evidence for the far western Bronze Age sites at the foot of the Tianshan Mountains underscores the puzzle that remains with the lack of any settlement data for the Xiaohe culture described in Chapter 3.

horse harnessing and mounted pastoral nomadism’, he argues that the limited focus of previous research on west to east exchanges needs to be expanded and take into account the exchanges between Central Asia and South-Afghanistan and the Indus-Baluchistan regions. The need for this extended perspective agrees with arguments put by other authors in this collection (see the two chapters on Kashgar). The special features of the Late Bronze Age and Iron Age emerge from the enormous diversity and complexity of the goods and artefacts that Francfort has studied over such a wide region. He suggests this reflects the growing elaboration of the technologies, social structures and religions that supported such production and exchange. Moreover, the exchange of goods and ideas accelerated. Both production and trade appeared to feed a greatly expanded demand by diverse regional elites for more raw materials, especially metal, stone and clay/plaster casting. The value of speedy communications has, for millennia, been recognized by scholars to be important in the survival of societies. But carefully documented cases such as those by Francfort are not easy to find. In his chapter, horsemanship replaced Bronze Age elite chariots and played a transformative role in ‘permitting ….speed of movement in long distance treks and raids, and speed in delivering showers of arrows by noblemen or laymen riding their horses’. Francfort argues that the nomadic populations thrived in mountainous and semi-desert areas, without having to invest the laborious effort required by the sophisticated but fragile agrarian empires of the Greeks and others. But when nomadic armies did successfully conquer the oases of South and Central Asia and Xinjiang, they acculturated and adopted the lifestyles of the agrarian centres like Bactra. Such cultural adoptions form a long running theme in the history of the regions mapped out by Francfort. His unique contribution is to give equal weight to the exchanges that occurred along the north-south axis (i.e. between the steppes of central Asia and the southern Bactrian-Hindukush regions) and the west-east exchange routes which are more typically discussed by other scholars of the Silk Road.

Liu’s account of the Iron Age in the Yili Valley is complemented by the chapter by Boroffka and Sverchkov. Their study helps us to understand the transition between the Bronze and Iron Ages through an analysis of the painted pottery that spanned both eras in southern Uzbekistan. Their unique GermanUzbek collaborative excavations of Bandykhan, especially the site of Maydatepa, began in 2005 and continue to challenge the idea that distinctive cultural forms emerged in particular regions and reflected self-contained technological and artistic formations. Instead they point to the convergence of cultural styles from southern Uzbekistan to the Ferghana Valley and Xinjiang. Even the unfired bricks of the Xintala site in Xinjiang closely resemble those of Maydatepa, not to mention the overlapping chronologies and the shared characteristics of the painted pottery. This evidence reflects not only close cultural contacts between such dispersed sites but also, possibly, influences transferred by human migrations. The authors plead for the notion of human migrations to be taken seriously, and draw on information about changing climates to propose that the settlers from Xintala moved to the upper Bandykhansay delta in response to the desertification of north-western Xinjiang. This is an important hypothesis because, in contrast to the majority of scholars working on the cultural links between Eurasia and Xinjiang, Boroffka and Sverchkov propose the movement of people and ideas from Xinjiang into the Eurasian regions to the west of the Tianshan Mountains rather than from the west to the east. A similar east-west movement of ideas and technologies is documented by Dodson, Atahan and Xiaoqiang in their chapter. In searching for explanations about the cultural affinities between Xinjiang and southern Afghanistan Boroffka and Sverchkov make an observation that needs to be taken seriously:

Francfort’s knowledge of this extensive terrain and time scale is based on in-depth excavations that he has been involved with for many decades – for example, the Sarazm site in the upper Zaravshan river valley of modern day Tadjikistan has been studied since 1984. The artefacts found on this site alone speak to long distance trade and the exchange of ideas with Iran, Baluchistan, the Indus Valley and the Siberian steppes. The images on the petroglyphs of the KarakoramHimalayan regions correspond to designs on engraved wood and woven textiles in the sandy parts of Xinjiang that lack stone, while in the Bortala and Upper Yili River Valleys the rock art corresponds to designs found on the rocks of the Altai Mountains (see Cong and Liu

It seems that the separation of regions with highly similar archaeological complexes is due rather to the history of research, caused by the order in which phenomena were studied by different research teams located within different contemporary administrative borders. Francfort’s chapter develops this point further by drawing on an even wider canvas for interregional comparisons ranging from the Mediterranean to the eastern part of Xinjiang and from South Asia to the Altai Mountains and Russia. While acknowledging that a critical feature of the transition from the Late Bronze Age to the Iron Age in Eurasia ‘was the emergence of 4

M. Vicziany: Ancient Xinjiang at the International Crossroads in this book). In contrast to Boroffka and Sverchkov, Francfort places less importance on human migrations to explain the appearance of parallel artefacts, designs and raw materials across a vast expanse. Human migrations assume ‘too many migrations from too many places to too many other places’. International trade networks seem to hold the key but there remain many gaps in the evidential record for complete certainty. The third period discussed by Francfort begins c. 250 BC and the cultural influences of the Greeks. Whether the Hellenistic styled artefacts discussed by Francfort came from Greece or were fashioned by local artists in the Grecian tradition remains unknown but both cases testify to the long distance trade and cultural links between northern Xinjiang and the Mediterranean, mainly as a result of the demands by elite nomadic societies. Elite tastes, it seems, were not restricted to goods of military value but also a wide variety of personal items. The Chinese chronicles portrayed the nomadic tribes of Eurasia as ‘barbarians’ but the archaeological record speaks to their cultural sophistication, much of it based on adopting and adapting technologies, ideas and styles from agrarian empires.

the result that local people benefitted from the end of empires. Unlike the other chapters in this collection that draw on evidence from many excavation sites, Kashgar’s archaeological record has suffered from long term neglect. In Chapter 9, Di Castro, Vicziany and Zhu argue for the urgent need to conserve and study the ancient monuments of Kashgar for they are rapidly disappearing as a result of the combined pressures of agricultural and urban development. Using photographs taken since 2005, observations from on-site visits, satellite images and interviews with local curators such as Mr. Qadir (the former Director of the Kashgar Museum) they suggest that some of the Buddhist sites might be amongst the earliest in western China. In particular, they have identified eight key archaeological sites (Mori Tim, Topa Tim, Khan-oi, Yawaluk, the Three Immortal Buddha Caves, Upal, Ancient Shule and Eskishahar). The Mori Tim site also appears to have the remnants of an ancient irrigation system – but its date and the corresponding dates for the stupas is yet to be determined. Their arguments for new research on Buddhist Kashgar are based on their study of the Buddhist economies of Sanchi in India, Anuradhapura in Sri Lanka and Angkor Wat in Cambodia. Not only were these three great centres of Buddhist learning and practice but they were also outstanding examples of how Buddhist monasteries managed the land and water systems that sustained these cultures. In contrast to Kashgar, where the absence of comprehensive excavations means that we have no dates for any of the eight sites, in these other centres the parameters of knowledge are constantly expanding thanks to the application of new technologies such as Lidar. For this reason, the authors have identified a range of methodologies that are needed for the study of Kashgar’s Buddhist heritage and fragile oasis environment, if the cultural achievements of this significant Buddhist centre are to be recognised.

The importance of preserving the archaeological record of Xinjiang is taken up in the next two chapters about the Buddhist monuments in the oasis town of Kashgar which sits at the meeting point between the Tianshan and Pamir mountains. These chapters go well beyond the Bronze and Iron Ages into the period c. 200 CE up to the Islamic era starting in c. 1000 CE. In Chapter 8, Vicziany and Di Castro argue that in contrast to imperial narratives that focus on the rise and fall of kings and warriors, the history of Kashgar needs to be understood as an integral part of international trade routes that benefited the local populations, irrespective of the identity of the overlords. In this sense, this chapter speaks to the themes raised by Francfort’s chapter. By reviewing examples of the literature in Chinese, Persian and European sources, Vicziany and Di Castro suggest that the rise of Kashgar as a centre for Buddhist culture was only possible, in the first instance, because of the location of the Kashgar oasis on crossroads flowing not only west-east but also south-north. The authors also question the usefulness of writing the history of western China by being overly dependent on the Chinese chronicles which focus on imperial histories. Drawing on the new scholarship about Indian empires such as the Mughals, for example, they argue that the fall of empires does not typically disadvantage the local elites or people because the rise of smaller successor states kept the fabric of society together in a familiar way, without ordinary people having to bear the burdens of an exploitative imperial superstructure. Small states were less demanding of the economic surplus than larger political formations with

Dodson, Atahan and Li are also concerned with demonstrating the value of particular technologies in documenting the history of Xinjiang. Using nuclear techniques they have provided absolute dates and, therefore, unique insights into the mineral and agricultural industries of the Bronze Age in the Hexi Corridor of Gansu. Their hypothesis is that the Corridor was an important conduit for the transfer of ideas from Mongolia into the Hexi region and further on. The excavated mineral remains show that bronze making in Gansu is at least some 4000 years old. While not as old as bronze manufacture in Mesopotamia, the mineral richness of the area is demonstrated by the continued mining for copper over the very long run. In this as in the other chapters, a research priority has been to establish reliable dates for the archaeological sites and the material evidence that they have 5

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads revealed. In addition to providing absolute dates for the manufacture of alloys they have also estimated reliable dates for the presence of wheat in Gansu. The earliest wheat seeds in this area were collected from sediments at Xishanping (Dadiwan) near the Xi River. The wheat and barley seeds are dated from about 4650 cal. years B.P. The sediment section shows that millet and rice cultivation are even older – some 5100 years. The presence and dates for wheat are especially important, given the role that wheat has played in diversifying the agricultural base of the emerging Chinese empires. Dodson et al. maintain (see Figure 7 and related discussion in Chapter 10) that robust AMS radiocarbon dating shows that wheat entered Gansu and further north-west some 4000 years ago and from there moved to the central, southern and western parts of China including Xinjiang. For many scholars, the role of Mongolia as a transition zone remains controversial (Spengler et al. 2016; Stevens et al. 2016). Without engaging in this debate, the importance of this chapter is that it compels us to reassess the relationship between Xinjiang, Gansu and Mongolia – something that has not attracted much attention until now.

and Frachetti (2004: 206-207, 242-243, 368) in the case of eastern Kazakhstan and southern Russia. Most significantly, this chapter addresses the controversies about the nature of prehistoric pastoralism. The authors show that Bronze Age people did not depend on their animal herds alone. This evidence agrees with the latest insights we have into the lifestyles and diets of the people of the Xiaohe culture. Using human hair/ dentin and animal bones Qu and colleagues concluded that while the ‘initial settlers of the Xiaohe culture primarily engaged in animal husbandry’ their diets became more complex over time and included, in some cases, surprisingly large quantities of wheat (Qu et al. 2018: 2010-2012). Conclusion The main conclusions of this collection can be summarised as follows. First, this monograph brings to public attention the first detailed and systematic account by Chinese archaeologists of the new, recently discovered Bronze Age sites of Xinjiang stretching from the Tianshan Mountains to the far eastern part of the Taklamakan desert. This contrasts with the previous literature that was focussed on the eastern desert region. Three chapters in this volume provide an important corrective to previous understandings of the Bronze Age by assessing the recent archaeological findings in the mountains bordering China and Kazakhstan. This refocus draws attention to western China’s river valleys and mountain passes as conduits for the transmission of cultures, technologies and peoples from further west, in Central Asia, to western China and eastwards. None of this diminishes the importance of Chapter 3 by Abuduresule et al., for this is the most comprehensive published discussion in English about the complexities of the Xiaohe culture located in the Taklamakan desert. Chapter 3 is critical for providing the essential context for ongoing work by this team and their colleagues in addressing the ever growing puzzles about the Xiaohe culture whether it is about its desert-based oasis environment (Li et al. 2013) or the world’s oldest glue (Rao et al. 2015) and paint brushes made of cattle heart muscle (Mai et al. 2016) identified amongst the grave goods of the mummies.

The last case study offered by Dodson et al. uses nuclear isotopes to analyse bone collagen from humans and animals found at sites in the Hexi Corridor. Their analysis shows that, despite the presence of some wheat in the palaeobotanical remains, millet was the main component of the diet of local Bronze Age people in the period from the late 3rd millennium to the early 2nd millennium cal. BC. By contrast, they did not find much evidence for the consumption of protein in the diet of the Bronze Age settlers. They do not suggest that this ‘intriguing’ diet was typical of Xinjiang or Gansu but it certainly justifies their call for further analysis of bone collagen from other sites. In the final chapter Jia and Chau also analyse the consumption habits of the Bronze Age peoples of Xinjiang, but this time their focus is on four archaeological sites in the far west of Xinjiang, near the Tianshan Mountains. Their methodology involves an analysis of the ancient starch granules found on eight grinding stones. This is the first work for Xinjiang using the identification of ancient starches to study the plants used by Bronze Age peoples who lived here c. 2300-1000 BC. The starches found by Jia and Chau came from cultivated and wild plants and food and non-food plants, including medicinal species. In showing the multiple character of the starches found on grinding stones, we can conclude that some mixed farming was practiced at that time but that the grinding stones by themselves do not prove that farming dominated. This supports the work of other scholars showing that wild plants were also processed (Liu et al. 2010; Liu et al. 2013; Liu et al. 2011; Yang et al. 2012). Moreover, cereals were also traded as noted by Francfort in this volume

Secondly, our findings stress the powerful and enduring cultural links between western China and the regions to the west, north and south. There were many Silk Roads along many routes, not only those along a west-east direction. Moreover, the direction in which innovative cultural and technical ideas flowed remains complex and multi-layered. The archaeological history of Xinjiang is still too young and scattered to allow us to resolve debates about which entry points mattered more than others. Rather we prefer to consider multiple entry points for the same or similar ideas and technologies. The porous nature of Xinjiang’s borders then and now, 6

M. Vicziany: Ancient Xinjiang at the International Crossroads together with the incomplete archaeological record, makes it hard to assume that one particular corridor was more important than another.

600 kms, distances were no obstacle to the spread of Bronze Age or other cultures. The surprising mobility of the Bronze Age peoples contributed to the growing complexity of these cultures over time as successive waves of new migrants arrived bringing with them wheat from the west and millet from the east (Qu et al. 2018: 2012). The next chapter by Betts takes up some of these themes by mapping the range of pre-historic cultural sites in Xinjiang and identifying features that relate to other parts of Eurasia and China.

Thirdly, our understanding of pastoralism in the Bronze Age has been expanded by these chapters demonstrating that Bronze Age herders, like their east Kazakhstan counterparts, were brilliant environmental managers and multi-skilled in combining herding with farming, artisanal, trading and security-oriented activities as a way of minimising the risks of living in difficult, extremely cold, semi-arid and desert environments. Fragile ecologies needed special management skills for survival.

References Abuduresule, Y. 2008. Treasures in the Tarim Basin, Xinjiang, China: Looking for the Ancient Civilisations. Public Lecture for the Monash Asia Institute and the National Gallery of Victoria. Melbourne: NGV, May. Barber, E. 1999. The Mummies of Ürümchi. London: Pan Books. Fletcher, R. 2011. Low-Density, Agrarian-based Urbanism: Scale, Power and Ecology in Michael E. Smith (ed.) The Comparative Archaeology of Complex Societies. Cambridge: Cambridge University Press: 289-381. Frachetti, M. 2004. Bronze Age Pastoral Landscapes of Eurasia and the Nature of Social Interaction in the Mountain Steppe Zone of Eastern Kazakhstan. Unpublished PhD dissertation, University of Pennsylvania. Hedin, S. 1903. Central Asia and Tibet: Towards the Holy City of Lassa, Vol. 1. London: Hurst and Blackett. Hinchliffe, J. 2018. How smoked eel is at the heart of Australia’s next World Heritage bid. The Age, 9 December, viewed on 8 March 2019, . Keen, I. 2004. Aboriginal Economy and Society: Australia at the Threshold of Colonisation. South Melbourne: Oxford University Press. Li, C. et al. 2015. Analysis of ancient human mitochondrial DNA from the Xiaohe cemetery: insights into prehistoric population movements in the Tarim Basin, China. BMC Genetics 16:78, viewed on 29 April 2019, . Li, J. et al. 2013. Buried in Sands: Environmental Analysis at the Archaeological Site of Xiaohe Cemetery, Xinjiang, China. PLoS ONE 8: 7, July 22. https://doi. org/10.1371/journal.pone.0068957. Li, C. et al. 2010. Evidence that a West-East admixed population lived in the Tarim Basin as early as the early Bronze Age. BMC Biology 8: 15, viewed on 29 April 2019, . Liu, X. 2001. Migration and Settlement of the YuezhiKushan: Interaction and Interdependence of

Fourthly, different chapters have emphasised different mechanisms of cultural-technological transfers. We feel that instead of focussing on the debate about whether trade was more important than human migration, we need to recognise that both were possible and could serve as substitutes for each other depending on what was opportune at particular points in time. Fifthly, many chapters emphasised the need to solve the remaining puzzles of Xinjiang’s long historical connections with bordering regions by expanding the research technologies that investigators use. Archaeological findings have made major contributions to increasing the parameters of our understanding of the processes of long-term interchanges of people, goods, technologies and ideas but they need to go hand in hand with a re-examination of documents which can now be approached with fresh eyes. The case of Kashgar, however, is exceptional, for here we urgently need systematic archaeological investigations before the pre-Islamic heritage is lost for ever. In summary, while the role of Xinjiang as a transition zone has long been appreciated, at this stage, we simply do not yet know enough to assert that any one route or mechanism was more important than another. We cannot even generate a chronology suggesting which routes preceded others, for the archaeological and historical records remain imprecise or unknown. Nor can we yet reconstruct the sequences by which these transmissions happened once they had crossed into western China. What this collection has drawn attention to is the misleading view that transmissions along the Silk Road occurred mainly from west to east. This needs to be revised by reinstating the notion of many ‘Silk Roads’ and zones of transmission from south to north and also from east to west, involving not only the mountainous-valley-steppe routes from Kazakhstan, Russia and Central Asia but also paths via Afghanistan, Pakistan, India, Mongolia and Central China. And it may well be that all of these routes were equally significant. What we can say on the basis of Chapter 3 is that if the Xiaohe culture replicated itself across a ‘gap’ of some 7

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Rao, H et al. 2015. Proteomic identification of adhesive on a bone sculpture-inlaid wooden artifact from the Xiaohe Cemetery, Xinjiang, China. Journal of Archaeological Science 53: 148-155. http://dx.doi. org/10.1016/j.jas.2014.10.010. Romgard, J. 2008. Ancient Human Settlements in Xinjiang and the Early Silk Road Trade. Sino-Platonic Papers 185: 1-126. Ruan, Q. 2013. Studies on the discoveries of Andronovo affiliation found in Xinjiang, China. Western Archaeology 7: 125-154 [阮秋荣:《新疆发现的安 德罗诺沃文化遗存研究》，《 西部考古 》2013 年，7 期, 125-154 页] Spengler, R. et al. 2016. The spread of agriculture into northern Central Asia: Timing, pathways, and environmental feedbacks. The Holocene 26/10: 15271540. DOI: 10.1177/0959683616641739. Stevens, C. et al. 2016. Between China and South Asia: A Middle Asian corridor of crop dispersal and agricultural innovation in the Bronze Age. The Holocene 26/10: 1541-1555. DOI:10.1177/0959683616650268. Yang, R. et al. 2014. Investigation of cereal remains at the Xiaohe Cemetery in Xinjiang, China. Journal of Archaeological Science 49: 42-47. Yang, Y. 2014. Proteomics evidence for kefir dairy in Early Bronze Age China. Journal of Archaeological Science 45:178-186.

Nomadic and Sedentary Societies. Journal of World History 12: 262-292. Mai, H. et al. 2016. Characterization of cosmetic sticks at Xiaohe Cemetery in early Bronze Age Xinjiang, China. Nature Scientific Reports 6:18939. DOI: 10.1038/ srep18939. Mallory, J.P. and V.H. Mair 2000. The Tarim Mummies: ancient China and the mystery of the earliest peoples from the West. London: Thames & Hudson. McNiven, I. et al. 2015. Phased redevelopment of an ancient Gunditjmara fish trap over the past 800 years: Muldoons trap complex, Lake Condah, southwestern Victoria. Australian Archaeology 81: 4458. Mirtschin, A. 2013. The Aboriginal Stone Huts of Lake Condah. YouTube video, 16 Dec, viewed on 8 February 2019, . Qu, Y. et al. 2018 Diverse lifestyles and populations in the Xiaohe culture of the Lop Nur region, Xinjiang, China. Archaeological and Anthropological Science. 10: 2005-2014. DOI 10.1007/s12520-017-0520-7.

8

Chapter 2

Xinjiang in Prehistory Alison Betts From the Palaeolithic to the start of the Bronze Age small groups of hunter-forager-fishers roamed throughout Xinjiang, in a manner similar to those whose traces have been found across Eurasia and southern Siberia. This changed in the Bronze Age when, by the 3rd millennium BC, Xinjiang began to assume its role as the crossroads of Inner Asia. The story of Xinjiang in prehistory is still a fragmentary one, but a recent spate of fieldwork has begun to build a general framework that can be used to understand not only the early cultural development of the region but also its relationships to the lands surrounding it. Using the most recent research, this chapter presents a summary of the prehistory of Xinjiang in order to place the varied papers in this volume into a broader context.

influenced the nature of human activity within and across it (Figure 1). Xinjiang covers some 1665 million square kilometres of Inner Asia, ringed by mountains and filled by harsh and often impenetrable deserts. Passage through the region is commonly around the southern rim, where the vast Taklamakan desert meets the northern edge of the Tibetan plateau. Here rivers flowing off the mountains form a scattered chain of small oases, separated by long and hazardous stretches of dune fields. Running east-west across the centre is the Tianshan, a narrow, steep chain of young fold mountains. The peaks are capped with permanent glaciers from which the summer melt flows down through deeply incised valleys to the deserts below. The southern flanks are dry and generally barren, but the northern slopes and highland basins support rich mountain grasslands. Along the foothills of the northern Tianshan runs a narrow belt of land no

China’s most westerly territory, Xinjiang, has a distinctive and complex landscape that has strongly

Figure 1. Diagram showing the key Xinjiang sites discussed in this chapter.

9

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 2a and b. View of the pre-Bronze Age site of Goukou in the north-eastern Tianshan; stone tools from Goukou (photo Betts).

more than a kilometre or so wide, suitable for dry farming (Jia et al. 2011), while the Yili River cuts the western Tianshan range in two with a grassy valley broadening out into the Semirech’ye (Seven Rivers) steppe in eastern Kazakhstan, a natural route to and from the Eurasian steppe. The eastern end of Xinjiang sits between the Altai Mountains, the Tianshan and the Tibetan Plateau, offering only a narrow desert corridor into China, flanked to the north by the stony wastes of the Gobi Desert. This route, known as the Hexi Corridor, leads into Gansu and eventually down to the middle Yellow River Valley.

large stone cists within a stone enclosure. Geometric designs are occasionally found painted on the interior of the stone slabs of the cists. The entrances to the enclosures are marked by large, often highly detailed, anthropomorphic stelae representing both men and women (Jia and Betts 2010; Lu et al. 2001; XIA ASS 1981, 1985) (Figure 3). The Qiemu’erqieke people carved clusters of stylised human faces onto rocks near the areas where they buried their dead, and probably also where they lived. Little is known of the Qiemu’erqieke economic basis as no settlement sites have yet been identified. Excavations at a Qiemu’erqieke site in the Chinese Altai have provided dates around 2400 BC (XICRA 2014), while another recently excavated cemetery, Sa’ensayi, near Urumqi, in the central region of the northern Tianshan, contained a grave of

Little is known of the early prehistory of Xinjiang. Prior to the Bronze Age, the region supported a small population of hunter-foragers (Debaine-Francfort 1988: 7 ff.; Jia et al. 2011: 171), but their presence has largely been recorded only through surface scatters of stone tools (Figure 2). Recent excavations at the cave site of Tongtiandong in the southern Altai have revealed evidence for human activity at around 40 k BP (Yu and He 2017), and one other stratified but undated preBronze Age context is known, from the site of Jilintai in the Yili Valley (Ruan 2013). By at least the mid-3rd millennium BC, a new group of people were moving in from the north-west. Archaeologically referred to as the Qiemu’erqieke tradition (Jia and Betts 2010; Kovalev 2008; Kovalev (ed.) 2015; Kovalev et al. 2008; XIA ASS 1981), this distinctive culture is distributed across the northern part of Xinjiang, in the Altai, but also in an expansion south-west along the Tarbagatay range to the western Tianshan where traces of them have been found in the Bortala Valley (Jia et al. 2017), and south-east along the Altai and across the Gurbantunggut desert to the eastern Tianshan at Jimusar and neighbouring regions. The Qiemu’erqieke cultural signature includes distinctive incised ovoid round-bottomed jars, stone vessels of the same design but with few or no incisions, and well-made flaked stone arrowheads. Burials are in

Figure 3. Qiemu’erqieke stone stela, Xinjiang Altai (photo Betts).

10

A. Betts: Xinjiang in Prehistory Qiemu’erqieke type, with similar ceramics and stone bowls, dated to 2470 cal. BC (XIA 2013). It is possible that the tradition continued in similar form into the early to mid-2nd millennium, c. 1700 BC.

domesticates, were introduced into Xinjiang from Eurasia at some time in the 3rd millennium BC (Betts et al. 2014; Dodson et al. 2013; Flad et al. 2010; Liu et al. 2017; Spengler et al. 2016; Stevens et al. 2016). Isotopic studies of diet at Tianshanbeilu show that the population generally consumed both wheat (C3) and millet (C4), with the exception of the earliest dated individual (1940-1765 cal. BC) who had an exclusive diet of millet (Wang et al. 2017).

Qiemu’erqieke cultural traits are strongly developed and highly distinctive, but their roots can be seen in the Early to Middle Bronze Age cultures of the Altai and southern Siberia, the Afanasievo and the Okunevo. These include cist burials, ceramics, some roundbottomed, with incised patterns, stone arrowheads and anthropomorphic stelae. It is most likely that the Qiemu’erqieke people were predominantly mobile pastoralists, as were the Afanasievo and Okunevo peoples (Frachetti 2008: 39-40), probably with small scale supplementary agriculture (Chapter 11 in this volume). Cemeteries are in the mountain grassland areas used by pastoralist groups up to recent times. Since the greatest concentration of sites occurs along the Altai, with the spread into northern Xinjiang and into the Tianshan a more limited expansion, it seems likely that the Qiemu’erqieke culture arrived in Xinjiang through a gradual search for new pastures and other resources, but it is certainly possible that it assimilated indigenous pre-Bronze Age populations as it expanded.

The third early tradition in Xinjiang, in the south, is that of Xiaohe, a largely aceramic culture based in the rivers and oases of the Tarim Basin (Abuduresule et al. in Chapter 3 in this volume). It is internationally famous for the highly elaborate nature and quite remarkable organic preservation of its burials. The Xiaohe culture is known from two key sites, the Xiaohe Cemetery itself (XAT 2004, 2005, 2007) and the cemetery of Gumuguo (Han 1986; Wang 2014) in the same region. More recently a similar cemetery, the Northern Cemetery, was found some 600 km to the south-west on the lower reaches of the Keriya River, showing that the culture was once widespread along the rivers that cut through the Taklamakan desert. The Xiaohe people cultivated wheat and barley (Yang et al. 2014), although probably on a small scale, and herded cattle (Abuduresule et al. in Chapter 3 in this volume; Cai et al. 2014; Mai et al. 2016). The earliest phase for the Xiaohe culture has been dated to 2200-1880 cal. BC when it began to develop in the centre and eastern region of the Tarim Basin (Chapter 3 in this volume). The type site lies in the now barren, saline depression of Lop Nur, once a massive lake fed by the Tarim River. By the early to mid-2nd millennium BC, regional variations in the material remains indicate that the Xiaohe culture began to split into two groups. The two groups of the Xiaohe culture are represented by the variants of Keriya–Xiaohe and Gumugou-Xianshuiquan respectively (Abuduresule et al. in Chapter 3 in this volume).

As Eurasian pastoralists began to move into the northwest of Xinjiang, small-scale oasis farmers from the Hexi Corridor and Gansu were beginning to work their way westwards to settle on the south-eastern end of the Tianshan. In the oasis of Hami, archaeologists have identified the Tianshanbeilu (also known as the Linya) culture, associated with a rich corpus of elaborate and colourful painted ceramics. The Tianshanbeilu culture is known only from the Tianshanbeilu Cemetery in Hami City (Li 2009; Lu et al. 2001). Hami lies towards the south-eastern end of the Tianshan, at the north-west end of the Hexi corridor connecting central China and Eurasia. The nature of the cultural influences is clear in the appearance of hand-made, painted redware pottery undoubtedly derived from eastern styles. The designs are mainly executed in black with some use of a purplish-red paint. Globular jars with double handles are among the most common forms. There are also taller cylindrical jars, again with double handles and painted decoration. The Tianshanbeilu culture is still poorly dated, with some disputed C14 dates and several opinions on relative chronology. Some scholars have argued that the earliest parallels are with the 4th to 3rd millennia BC Machang culture of Gansu (Han 2005: 81; Han 2012) but more recent research suggests that the site should be dated from c. 2000-1300 BC, based both on new typological analyses and C14 dates on some of the skeletons (Wang et al. 2017). Some ceramic vessels from the cemetery find strong parallels with those of the early 2nd millennium BC Siba culture of Gansu. The oasis farmers in Gansu cultivated millet, a Chinese domesticate. Wheat and barley, West Asian

The Xiaohe people exploited a lacustrine and riverine environment bordering on an extreme desert (Li et al. 2013; Zhang et al. 2017). Their cattle were clearly central to their lives. Many of their coffins were covered in cattle hides, and painted bovine skulls with massive horns still attached were placed in the burials. Cattle hearts were used as the basis of cosmetics, cattle ears were found in graves and the Xiaohe people wore boots made of cattle hide (Mai et al. 2016). The rivers and lakes, too, were integral to their survival. Apart from providing water to survive, the Xiaohe material culture was almost totally derived from the plants that the rivers and lakes sustained. Timber was heavily exploited, notably the stands of massive Populus euphratica that line the rivers of the Tarim Basin. Burials were in wooden coffins, some formed in the shape of canoes. The Xiaohe people seem to have 11

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads made very little pottery but stored their possessions in elaborately woven baskets.

Qiemu’erqieke peoples. They spread into the western hills and mountains of Xinjiang, along the Tianshan towards the east and extended south into the Pamirs, close to the western end of the Tibetan Plateau. This group shared broad affinity with the loosely defined Andronovo complex that appeared widely across Eurasia in the later Bronze Age (Koryakova and Epimakhov 2007; Kuz’mina 2007; 2008), more specifically the eastern Federovo variant (Jia et al. 2017). Ceramics are characterised by handmade, flat-based pottery vessels, most commonly shouldered jars with a limited array of incised and stamped decoration (Figure 4). Graves vary in form but are sometimes covered with stone mounds (kurgan), while many contain stone built cists formed by large rough-cut slabs of rock. Grave goods include pots as well as some personal items including gold earrings. Houses are typically quite large rectangular semisubterranean structures with double rows of stones defining the outer walls. Inside there are a variety of features including partition walls, hearths, postholes and pits (Jia et al. 2017; XICRA et al. 2018). A hoard of bronzes, the Aga’ersen Hoard (Mei 2000:14), was found in Gongliu County near Yili Township in western Xinjiang, containing three shaft-hole axes, three sickles, three gouges, two chisels and one hammer. The tools find strong parallels in Andronovo types found in the Semirech’ye region of eastern Kazakhstan (Kuz’mina 1998).

DNA studies have shown that the Xiaohe culture had its roots in Eurasia, but like the Qiemu’erqieke culture, it may also have absorbed elements of a pre-Bronze Age indigenous population (Betts et al. 2018). Genetic studies of the Xiaohe population (Li et al. 2010; Li et al. 2015) indicate that the original Xiaohe people comprised an admixture of both East and West Eurasian ancestral populations, with the likelihood that the admixed population might have had some relationship with South Siberian Bronze Age populations. As noted by Abuduresule et al. (Chapter 3 in this volume), as time went by the population added some traces of Indian lineages and probably some from the Central Asian oases. Overall, however, the south/west Asian components are relatively minor in the Xiaohe population, while the Eurasian ancestry remained dominant. Qiemu’erqieke, Tianshanbeilu and Xiaohe/Gumuguo are the earliest Bronze Age traditions to appear in Xinjiang, initially occupying distinct physical and environmental niches. The Qiemu’erqieke people favoured mountain valleys suited to transhumant pastoralism. The Tianshanbeilu culture appears in the Hami oasis. While their occupation sites are lost, it seems likely that the Tianshanbeilu people followed the traditions of their eastern forebears, living a fairly sedentary life farming cereal crops. Although the Xiaohe/Gumuguo peoples may have had their origins among Eurasian transhumant pastoralists (Betts et al. 2018), they adapted themselves to a lacustrine/riverine environment around the Lop Nur lakes and along the Tarim River, herding cattle and growing some cereals.

Sites of this period have been excavated along the Kazakh border southwest of Tacheng, in the Bortala Valley (Jia et al. 2017; Chapter 4 in this volume) at the occupation site of Adunqiaolu and others, as well as numerous cemeteries. Similar cultural traditions have also been identified at cemeteries in the Yili Valley (Chapter 4 in this volume), and at Jirentai Goukou (Jartai Pass), also in the Yili Valley, an occupation site with evidence for one of the earliest known examples of the use of coal as fuel (XICRA et al. 2018). The sites and cemeteries of this type in the Bortala Valley date

By the early 2nd millennium BC, new Eurasian agropastoralists moved in from the west, occupying some of the same lands as the southerly expansion of

Figure 4. Ceramic vessels from the Adunqiaolu cemetery, Bortala Valley (photo Doumani Dupuy).

12

A. Betts: Xinjiang in Prehistory

Figure 5. Deer Stone Khirigsuur Complex, Xinjiang Altai (photo Betts).

generally from 1800 – 1300 cal. BC (Chapter 4 in this volume; Jia et al. 2017). The most southerly outlier, the site of Xiabandi at Tashkurgan in the Pamirs, has very simple, crudely made, flat bottomed grey ware vessels that are broadly paralleled in Andronovo forms, and is dated from c. 1900 – 1500 cal. BC (Jia et al. 2017: 633). While the earlier Bronze Age sites are associated with the east-west dispersal of crops, the Federovo affiliated tradition can be linked to the broader spread of metallurgy from Eurasia into China (Mei 2000, 2003, 2009; Mei et al. 2017). As with the cereals, new ideas and techniques were brought into Xinjiang by Eurasian peoples and then transmitted further down the Hexi Corridor and perhaps Mongolia through technological diffusion by various means.

into the Iron Age (c. 1300-700 BC) (Fitzhugh 2009) also extended down into the Xinjiang Altai (Figure 5). The culture is marked by sometimes very large stone tumuli surrounded by spoked stone circles. The deer stones are tall slender stelae carved with images of deer with exaggerated antlers in the steppic ‘animal style’, and less frequently horses. To the south, there is evidence that the mountain regions from the Pamirs eastwards along the northern rim of the Tibetan plateau were also exploited by pastoralist groups. Little fieldwork has been carried out south of the Tarim Basin, but one key exception is the Liushui Cemetery in the Kunlunshan range. This remarkable cemetery was marked by a large field of cobble-stone rectangles, alternately of black and white rocks, creating an extraordinary chequerboard design. Ceramics from the graves are coarse unpainted handmade grey ware jars with incised zigzag decoration. Some vessels have handles. The Luishui tradition seems to find its influences from the northwest in derived Karasuk forms and decorative styles, suggesting a continuing spread of cultural contact among transhumant pastoralists along the mountains of western Xinjiang. The Federovo affiliated site of Xiabandi in its later phases develops a local tradition of plain round bottomed jars and bowls, a ceramic assemblage found also in the Xiangbaobao cemetery in the same area. This localised tradition represented the Iron Age in the Pamirs, with signs of some wider influence, in for example the round bottomed vessels, from other sites along the western mountain rim of Xinjiang, such as the Yili Valley.

Following these fairly clear cultural clusters, it becomes a little harder to define the Late Bronze and Bronze-Iron Age transition in Xinjiang. A number of sites are known but there is limited publication, very little in the way of absolutely dating and a notable mixture of styles within the material cultural remains. It appears though that broad patterns of settlement type and economies established in the earlier Bronze Age still persist. In the hill and mountain country along the western rim of Xinjiang from the Altai to the Pamirs transhumant (agro) pastoralists continued to exploit the grasslands, their camp sites and burials identifiable through handmade grey ware pottery vessels with incised decoration. In the Xinjiang Altai, coarse grey ware vessels broadly derived from Karaksuk traditions, known as the Kuxi variant, have been found following looting of graves (AIX 2015: 61-106). A new nomadic culture centred in Mongolia, the ‘Deer Stone Khirigsuur Complex’, dating from the late Bronze Age

Along the southern rim of the Tianshan, the influence of the farming tradition of the Tianshanbeilu culture in the Hami oasis spread westwards, creating isolated 13

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 6. Chawuhugou cemetery, painted jugs (photo Xinjiang Uygur Autonomous Region International Cultural Exchange Association).

pockets of settlement with distinctive localised painted pottery traditions. In the oasis of Hami itself, painted ceramics at the cemetery of Yanbulake derive from earlier Tianshanbeilu forms but with cultural influences from sites to the east in Gansu and Qinghai, indicating continued contact between these regions and Xinjiang (Han 2017). In the greater Hami district, there are rich areas of upland grazing at Balikun (Barkol) in the Tianshan. Here sites have been found in the summer herding pastures at c. 2300 asl. The occupation site of Hongshankou and burials at Dongheigou have produced locally distinct ceramics yet continuing to show influences from earlier Tianshanbeilu shapes with double handles and some painted designs. The occupation sites are stone built with irregular layouts (Xibei et al. 2014; XICRA 2009).

on the north-eastern slopes of the Tianshan, and in the lush grasslands of the Yili where, as is the case today, farming and pastoralism may have achieved a more even balance in the lifestyles of the Late Bronze and early Iron Age people there (Chapter 3 in this volume). Towards the eastern end of the narrow dry farming belt on the northern slopes of the mid-Tianshan an exploratory sounding was carried out at the farming settlement of Luanzagangzi (Jia et al. 2011). The early stage at the site is dated from 1300 to 1000 cal. BC. This stage at Luanzagangzi represents a local Late Bronze/ Early Iron Age variant, influenced by the painted ceramics of the Hami region. Ceramic assemblages in sites of the northern Tianshan slopes at this time are distinctive for their double handed jars and painted net and pendant triangle designs typified in ceramics from the cemetery of Banjiegou and other sites in the region. There are parallels in some of the Chawuhugou ceramics with those seen in the Yili Valley, while double-handled vessels show similarities with the sites in the northern Tianshan.

West of Hami on the southern slopes of the Tianshan, the Subeixi cemetery and occupation site (XIAA et al. 2003) is an example of an oasis settlement derived from the earlier Tianshanbeilu tradition, also maintaining contacts with the east. The settlement site contained mudbrick houses and the ceramics were painted vessels with shapes and designs influenced by the Late Bronze/ early Iron Age Xindian culture of Gansu-Qinghai (c. 1000 BC onwards). Radio carbon dates for the Subeixi settlement are c. 593-173 cal. BC and for the cemetery from 791-360 cal. BC.

The Yanqi Basin lies at the mouth of a long valley running north-westwards into the Tianshan, to a watershed with the Yili Valley, clearly a route linking the Tarim Basin with the Eurasian steppe back into the Bronze Age. Near Chawuhugou, the cemetery of Xintala (XICRA 1988）also has painted pottery but appears to have an earlier chronological stratum associated with handmade unpainted wares with incised zigzags that have some parallels with the western Xinjiang Andronovo affiliated cultures. This suggestion is reinforced by the recovery of metal artefacts comprising a copper awl and a knife as well as a stone mould, and surface finds that included a socketed celt, an arrowhead and an awl of Andronovo type (Kuz’mina 1998). The apparently later painted pottery horizon has designs seemingly echoing some of the earlier incised patterns, as well as some shapes and forms also found on sites of Qiongkeke Stage II type in the Yili Valley. Some shapes may also echo derived Karasuk influence.

Further west again along the southern slopes of the Tianshan in the Yanqi Basin are two excavated cemeteries of this Late Bronze/Early Iron Age period. Chawuhugou represents an example of a regional variant of the painted pottery oasis cultures (XIA 1999), although this far to the west there is no strong evidence for continued contact with Gansu cultures. The ceramic assemblage is dominated by distinctive spouted vessels (Figure 6). Painted designs comprise a variety of geometric shapes including net patterns, chequered squares and triangles. Some agricultural settlements also start to appear in dry-farming areas, 14

A. Betts: Xinjiang in Prehistory In the Yili valley (Chapter 4 in this volume: Figs 6, 7, 15, 17-21) ceramics from cemeteries falling broadly within the Qiongkeke Stage II tradition are a mixture of painted and plain wares with a range of forms. Painted designs and some shapes echo painted pottery known from the Chust culture (Kuz’mina 2007: 438), a sedentary late Bronze to early Iron Age tradition with advanced metallurgy in the rich, fertile Ferghana Valley running westwards from Kirghizstan into Uzbekistan. It is likely though that influences on the Yili cultures run to the east as well as to the west as both shapes and painted designs find parallels in contemporaneous ceramic assemblages in the mid-Tianshan and Hami regions.

is the transmission of West Asian domesticated food plants, most notably wheat, with the concomitant east to west transmission of cultivated millets, taking place at some time in the 3rd millennium BC. Next was the eastward spread of Eurasian metallurgical techniques (Mei 2000), predominantly in the 2nd millennium BC, and particularly in association with the sites of Andronovo affinity. The third is the horse, with a precocious period of domestication in the Botai culture as early as the 4th millennium BC in the Eurasian steppe, but possibly with a later second phase of domestication in Eurasia (Gaunitz et al. 2018). The horse was not widely adopted as a dominant herd animal in Xinjiang until well into the Iron Age, but there is evidence for horses and horse riding at the Chawuhugou and Subeixi cemeteries (XIA 1999; XIAA et al. 2003). An associated technological import from Eurasia was the spokewheeled chariot. This inspirational development out of the solid wheeled ox-drawn wagons of earlier Eurasian prehistory spread rapidly out across the ancient world into both the Middle East and China. In the Altaic Deer Stone Khirigsuur Complex the kurgans contained ritually interred horses. Palaeopathological analyses show that the horses were probably bridled and used for transport, and Taylor et al. (2015) suggest that they were used as both riding animals and to draw chariots. Domestic horses first reached China, along with chariots, during the late Shang Dynasty at c. 1300-1200 BC (Kelekna 2013; Yuan and Flad 2006).

At the western end of the Tarim Basin, in the vicinity of the Kashgar Oasis, there are hints of connections to the west, but these are ill-defined and based primarily on surface finds. Surveys of the site of Aketala recovered a copper knife, some coarse sandy pottery and a number of ground stone implements including knives, sickles, mortars, pestles, spindle whorls and grinding stones. Similar crescent-shaped stone knives are known from the Chust culture of the Ferghana Valley (Askarov 1992: 447-451). While Aketala’s place in the prehistoric chronology of Xinjiang is uncertain, it is perhaps also likely to belong to this period, dating broadly from the late second to the early first millennium BC (Mei 2000: 9), but the possibility of earlier levels cannot be dismissed. The later prehistory of Xinjiang does not conform to what might be considered more traditional patterns of development from hunter-gather-foragers to agricultural villages. There is no Neolithic and no process of indigenous domestication. Domesticated crops and animals were introduced from outside, not to settled villages, but into the hands of transhumant pastoralists. Yet there is a distinct pattern of development. In the earlier Bronze Age, the region is influenced from both west and east. The impact of Eurasian cultures is greater, appearing with the spread of the Qiemu’erqieke agro-pastoralists across the northern half of Xinjiang, possibly also forming an ancestral population for the Xiaohe/Gumuguo peoples in the south. In this early stage eastern influence is limited largely to the Hami oasis with the people of Tianshanbeilu. In the next stage the Andronovo/ Federovo affiliated agro-pastoralists extend across the mountain regions of the western half of Xinjiang from the Altai to the Pamirs. This is followed by a blending of influences as oasis settlements and painted pottery spread westwards and north across the Tianshan, while transhumant pastoralists continue to graze their herds in the mountain grasslands.

Xinjiang has always had its own highly individual pattern of cultural development despite its position at the ‘Crossroads of Asia’. Much of this has been dictated by its landscape of mountains and deserts. Its north has been roamed mostly by nomadic groups, while the oases have nourished individual pockets of settled life, often quite distinct from one to another, for the most part because of great distances separating them. Despite this, cultural influences, ideas and technologies have been regularly interchanged. The ebb and flow of the ancient cultures of Xinjiang has created a land of rich complexity and remarkable social diversity. References AIX. Archaeological Institute of Xinjiang 2015. Collection of Archaeological and Historical Research in Xinjiang Altai. Beijing: Relics Press. [新疆考古研究所:《新 疆阿勒泰地区考古与历史文集》2015 年, 北京: 文 物出版社]. Askarov, A. 1992. The Beginning of the Iron Age in Transoxiana, in A. Dani and V. Masson (eds) History of Civilisations of Central Asia: 441-458. Paris: UNESCO. Betts, A. et al. 2014. The Origins of Wheat in China: potential pathways for its introduction. Quaternary International 348/20: 158-168. Betts, A. et al. 2019. A New Hypothesis for Early Bronze Age Cultural Diversity in Xinjiang, China.

These prehistoric populations of Xinjiang acted as the vehicles by which several key practices and technologies were transmitted from west to east. The earliest of these 15

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Archaeological Research in Asia 17: 204-213. https:// doi.org/10.1016/j.ara.2018.04.001. Cai, D. et al. 2014. The origins of Chinese domestic cattle as revealed by ancient DNA analysis. Journal of Archaeological Science 41: 423-434. http://dx.doi. org/10.1016/j.jas.2013.09.003. Debaine-Francfort, C. 1988. Archéologie du Xinjiang des Origins aux Han. Paléorient 14(1), 5-29. Dodson, J. et al. 2013. Origin and spread of wheat in China. Quaternary Science Reviews 15: 108-111. http:// dx.doi.org/10.1016/j.quascirev.2013.04.021. Fitzhugh, W. 2009. The Mongolian Deer Stone Khirigsuur Complex: Dating and Organization of a Late Bronze Age Menagerie, in J. Bemmann, H. Parzinger, E. Pohl and D. Tseveendorj (eds) Current Archaeological Research in Mongolia: Papers from the 1st International Conference on Archaeological Research in Mongolia: 183199. Bonn: University of Bonn. Flad, R. et al. 2010. Early wheat in China: results from new studies at Donhuishan in the Hexi Corridor. The Holocene 20/6:955-965. Frachetti, M. 2012. Multiregional Emergence of Mobile Pastoralism and Nonuniform Institutional Complexity across Eurasia. Current Anthropology 53/1: 2-38. Frachetti, M. 2008. Pastoralist Landscapes and Social Interaction in Bronze Age Eurasia. Berkeley: University of California Press. Gaunitz, C. et al. 2018. Ancient genomes revisit the ancestry of domestic and Przewalski’s horses. Science 10.1126/science.aao3297 (2018). Han, J. 2005. Chronological and typological analysis of Xinjiang Bronze and early Iron Age. Xinjiang Relics 2005 (3), 57-91. [韩建业:《新疆青铜时代 – 早期铁 器时代文化的分期和谱系》,《 新疆文物 》2005 年, 3期 , 57-91 页]. Han, J. 2012. The ‘Painted Pottery Road’ and early Sino-Western cultural exchanges. Studia Classica et Orientalia Anabasis 3: 25-42. Han, J. 2017. Analysis on the rounded bottom pot found in Central Asia in the mid-late second millennium BC. Archaeology 2017(9): 82-93. [韩建业:《公元前 2千 纪中后叶亚洲中部地区的圜底陶罐》 ,《 考 古》2017 年, 第 9 期, 82-93 页]. Han, K. 1986. Study on the skeletons found at Gumugou cemetery in Xinjiang Konghe, Journal of Archaeological Studies 3:361-384. [韩康信:《新疆孔雀河古墓沟墓 地人骨研究》,《考古学报》1986 年 3 期, 361-384 页]. Jia, P. and A. Betts 2010. A re-analysis of the Qiemu’erqieke (Shamirshak) cemeteries, Xinjiang, China. Journal of Indo-European Studies 38/3-4: 1-43. Jia, P. et al. 2011. New Evidence for Bronze Age Agricultural Settlements in the Zhunge’er (Junggar) Basin, China. Journal of Field Archaeology 36/4: 269280. Jia, P. et al. 2017. Adunqiaolu: new evidence for the Andronovo in Xinjiang, China. Antiquity 91/357: 621639.

Kelekna, P. 2013. The Horse in Human History. Cambridge: Cambridge University Press. Koryakova, L. and A. Epimakhov 2007. The Urals and western Siberia in the Bronze and Iron Ages. Cambridge: Cambridge University Press. Kovalev, A. (ed.) 2015. Earliest Europeans in the Heart of Asia: The Chemurchek Cultural Phenomenon. St. Petersburg: St. Petersburg State Museum. Kovalev, A. 2008. Discovery of new cultures of the Bronze Age in Mongolia (According to the data obtained by the International Central Asiatic Archaeological Expedition), in Proceedings of Third International Conference of Turpan Study in the Origins and Migration of Eurasian Nomads: 343-370, Turpan, Xinjiang: Institute of Turpan Study. Kovalev, A. et al. 2008. Radiometric dating of Kurgans in the Mongolian Altai, in Ancient and Medieval Nomads of Central Asia: 172-186. Russian Federation: Department of Archaeology, Ethnography and History, Ministry of Education and Science. Ковалев А.А., Эрдэнэбаатар Д., Зайцева Г.И., Бурова Н.Д. Радиоуглеродное датирование курганов Монгольского Алтая, исследованных Международной Центрально-азиатской археологической экспедицией, и его значение для хронологического и типологического упорядочивания бронзового века Центральной Азии // Древние и средневековые кочевники Центральной Азии. Барнаул: Азбука, 2008. С. 172186. Kuz’mina, E. 1998. Cultural connections of the Tarim Basin people and pastoralists of the Asian steppes in the Bronze Age, in V. Mair (ed.). Journal of IndoEuropean Studies, Monograph Series No. 26, Vol. 1; Archeology, migration and nomadism, linguistics: 63-93. Washington D.C.: Institute for the Study of Man, Philadelphia. Kuz’mina, E. 2007. The origins of the Indo-Iranians. Leiden: Brill. Kuz’mina, E. 2008. The prehistory of the Silk Road. Philadelphia: University of Pennsylvania Press. Li, C. et al. 2010. Evidence that a West-East admixed population lived in the Tarim Basin as early as the early Bronze Age. BMC Biology 2010 8:15. https://doi. org/10.1186/1741-7007-8-15. Li, S. 2009. Study on the first phase in Tianshanbeilu cemetery. Turpan Studies 1:1-3. [李水城:《天山北路 墓地一期遺存分析》,《吐魯番學研究》2009 年 第 1 期, 第 1-3 頁]. Li, C. et al. 2015. Analysis of ancient human mitochondrial DNA from the Xiaohe cemetery: insights into prehistoric population movements in the Tarim Basin, China. BMC Genetics 2015 16/78 1-11. https:// doi.org/10.1186/s12863-015-0237-5. Li, J. et al. 2013. Buried in sands: environmental analysis at the archaeological site of Xiaohe Cemetery, Xinjiang, China. PLoS ONE 8/7 July 22, 2013. https:// doi.org/10.1371/journal.pone.0068957. 16

A. Betts: Xinjiang in Prehistory Liu, X. et al. 2017. Journey to the east: Diverse routes and variable flowering times for wheat and barley en route to prehistoric China. PLoS ONE 12/11: 0187405. Lu, E. et al. 2001. A preliminary analysis of Bronze Age archaeology in Xinjiang. In: Su Bingqi and Contemporary Chinese Archaeology. Science Press, Beijing, pp. 179-184. [吕恩国、常喜恩、 王 炳 华:《新疆青铜时代考古文化浅论》, 《苏秉琦与当代中国考古学》, 北京: 科学出版社 2001 年]. Mai, H. et al. 2016. Characterization of cosmetic sticks at Xiaohe Cemetery in early Bronze Age Xinjiang, China. National. Science Reports 6, 18939. http:// dx.doi.org/10.1038/srep18939. Mei, J. 2000. Copper and Bronze Metallurgy in Late Prehistoric Xinjiang. BAR International Series 865. Oxford: Archaeopress. Mei, J. 2003. Cultural interaction between China and Central Asia during the Bronze Age. Proceedings of the British Academy 121: 1-39. Mei, J. 2009. Early Metallurgy and Socio-Cultural Complexity: Archaeological Discoveries in Northwest China, in B. Hanks and K. Linduff (eds) Social Complexity in Prehistoric Eurasia: 215-32. Cambridge: Cambridge University Press. Mei, J. et al. 2017. The appropriation of early bronze technology in China, in J. Maran and P. Stockhammer (eds) Appropriating Innovations: Entangled Knowledge in Eurasia, 5000-1500 BCE: 232-241. Oxford: Oxbow Books. Ruan, Q. 2013. Studies on the discoveries of Andronovo affiliation found in Xinjiang, China. Western Archaeology 7: 125-154. [阮秋荣:《新疆发现的安德 罗诺沃文化遗存研究》,《西部考古》2013 年, 7 期, 125-154 页]. Spengler, R. et al. 2016. The spread of agriculture into northern Central Asia: Timing, pathways, and environmental feedbacks. The Holocene 26/10: 15271540. DOI: 10.1177/0959683616641739. Stevens, C. et al. 2016. Between China and South Asia: A Middle Asian corridor of crop dispersal and agricultural innovation in the Bronze Age. The Holocene 26/10: 1541-1555. DOI: 10.1177/0959683616650268. Taylor, W. et al. 2015. Equine Cranial Morphology and the Archaeological Identification of Riding and Chariotry in Late Bronze Age Mongolia. Antiquity 89: 854-871. Wang, B. 2014. Gumugou. Xinjiang Peoples Publishing House, Wulumuqi. [王炳华:《古墓沟》乌鲁木齐: 新疆人民出版社 2014 年]. Wang, T. et al. 2017. Tianshanbeilu and the Isotopic Millet Road: reviewing the late Neolithic/Bronze Age radiation of human millet consumption from north China to Europe. National Science Review https://doi.org/10.1093/nsr/nwx015. XAT 2004. Xiaohe archaeological team 2004. Survey and excavation report at Xiaohe cemetery in 2002. Frontier Archaeology 3:338-398. [新疆文物考古研究

所小河考古队:《2002 年小河墓地考古调查与发 掘报告》,《边疆考古研究》第 3 辑，2003 年， 第 338. ～ 398 页]. XAT 2005. Xiaohe Archaeological Team 2005. New achievement at Xiaohe Cemetery in Lop Nur. Turfan Studies 1:114-125. [新疆文物考古研究所小河考 古队:《罗布泊小河墓地考古发掘的重要收获》, 《吐鲁番学研究》2005 年 1 期, 114-125 页]. XAT 2007. Xiaohe archaeological team 2007. Brief report on the excavation at Xinjiang Lop Nur Xiaohe cemetery in 2003. Cultural Relics 10: 4-42. [新疆文物 考古研究所小河考古隊:《新疆罗布泊小河墓地 2003年发掘简报》,《文物》2007 年, 10 期, 4-42 页]. XIA 1999. Xinjiang Institute of Archaeology 1999: Xinjiang Chawuhu – A large-scale clan cemetery excavation report. Beijing: Oriental Press, 431-415. [新疆文物考古研究所 1999: 《新疆察乌呼-大型 氏族墓地发掘报告》。北京: 东方出版社, 431-415 页]. XIA 2013. Xinjiang Institute of Archaeology 2013. Xinjiang Saensayi Cemetery. Beijing: Relics Press. [新 疆文物考古研究所: 《新疆萨恩萨伊墓地》, 2013 年, 北京: 文物出版社]. XIA ASS. Xinjiang Institute of Archaeology, Academy of Social Science, 1985. Ancient Antiquities of Xinjiang. Beijing: Wenwu Press. [新疆社会科学院考古研究 所: 《新疆古代文物》, 北京: 文物出版社]. XIA ASS. Xinjiang Institute of Archaeology, Academy of Social Science, 1981. Summary of excavation at Qiemu’erqieke cemetery. Cultural relics 1981/1: 2332. [新疆社会科学院考古研究所:《克疆木齐墓地 发掘简报》,《文物》1981 年, 1 期, 23-32 页]. XIAA et al. Xinjiang Institute of Antiquity and Archaeology and Turfan Prefectural Museum 2003. The Subeixi Site and Cemeteries in Shanshan County, Xinjiang. Chinese Archaeology, 3:135-142. XICRA 1988. Xinjiang Institute of Cultural Relics and Archaeology 1988. The excavation of the Xintala site, Heshuo, Xinjiang. Archaeology1988/5: 399-407. [新疆文物考古研究所:《新疆和硕新塔拉遗址发 掘简报》,《考古》1988 年 5 期, 399-401 页]. XICRA 2009. Xinjiang Institute of Cultural Relics and Archaeology 2009. Brief report on the excavation of the Dongheigou site during 2006-2007 in Barkul, Xinjiang. Archaeology 2009/1:3-27. [新疆文物考古 研究所, 西北大学文化遗产与考古学研究中心: 《(新疆巴里坤县东黑沟遗址 2006 ～ 2007 年发掘 简报》,《考古》2009 年 1 期, 3-27 页]. XICRA 2013. Xinjiang Institute of Cultural Relics and Archaeology 2013. The Sa’ensayi Cemetery of Xinjiang. Beijing: Relics Press. [新疆考古研究所:《新疆萨恩 萨伊墓地》, 北京: 文物出版社 2013 年]. XICRA 2014. Xinjiang Institute of Cultural Relics and Archaeology, 2014. Tuganba’er cemetery of Habahe, Xinjiang. Cultural Relics 2014 /12: 18-28. [新疆文物 考古研究所:《新疆哈巴河托干拜2号墓地发掘简 报》,《文物》2014 年 12 月, 18-28 页]. 17

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads XICRA et al. 2018. Xinjiang Institute of Cultural Relics and Archaeology, Culture Bureau of the Ili Kazakh Autonomous Prefecture and Culture Bureau of Nilke County 2018. The Jartai Pass Site in Nilka County, Xinjiang. Chinese Archaeology 18 2018/18: 38-51. Xibei et al. 2014. Centre of Archaeology and Heritage Conservation for the Silk Road at Xibei University, Hami Bureau of Relics, Barkol Bureau of Relics 2014 Brief report of on the 2008 survey at Barkol Hongshankou site. Cultural relics 2014/7: 17-30. [西 北大学丝绸之路文化遗产保护与考古学研究中 心, 哈密地区文物局, 巴里坤县文物局:《新疆巴里 坤红山口遗址 2008 年调查简报》,《文物 》2014 年, 7 期, 17-30 页].

Yang, R. et al. 2014. Investigation of cereal remains at the Xiaohe Cemetery in Xinjiang, China. Journal of Archaeological Science 49 (September 2014): 42-47. Yu, J. and J. He 2017. Significant discoveries from the excavation of Jimunai Tongtiandong site, Xinjiang. Relics News Page 8, December 1, 2017. [于建军、 何嘉宁:《新疆吉木乃通天洞遗址发掘获重要收 获》,《文物报》2017 年 12 月 1 日第 8 页]. Yuan, J. and R. Flad 2006. Research on Early Horse Domestication in China, in M. Mashkour (ed.) Equids in Time and Space: 124-131. Oxford: Oxbow. Zhang, Y. et al. 2017. Holocene environmental changes around Xiaohe Cemetery and its effects on human occupation, Xinjiang, China. Journal of Geographical Science 27/6: 752-768. http://dx.doi. org/10.1007/ s11442-017-1404-6.

18

Chapter 3

The Xiaohe (Small River) Cemetery and the Xiaohe Culture

1

Yidilisi Abuduresule, Wenying Li, Xingjun Hu The Xiaohe Culture is a unique Bronze Age complex which began to develop in the central and eastern regions of the Tarim Basin in southern Xinjiang in the 3rd millennium BC. By the 2nd millennium BC it began to split into two groups. The two groups of the Xiaohe culture are represented by the variants of Keriya– Xiaohe and Gumugou-Xianshuiquan respectively and are different in terms of cultural contexts. The appearance of variations between the two groups may be the result of uneven development in social, cultural and economic development within different regions. One of the major features of the Xiaohe Culture is its close relationship to the cultures of the Eurasian steppe, leading to the hypothesis that it had its origins in the steppe (Lin 2002). The Xiaohe Culture also displays features that reflect connections with other cultural traditions in the Tarim Basin and the Hami region to the north-east. There is evidence of cultural interaction and influence between the Xiaohe and Hami Tianshanbeilu complexes.1

Taklamakan desert near the lower Keriya River. The cultural content of the Northern Cemetery is almost identical to that of the Xiaohe Cemetery. This new site has greatly improved our knowledge of the extent and nature of the Xiaohe Culture.

Expeditions to the Taklamakan Desert in the early 20th century revealed a number of ancient cemeteries. Aurel Stein from Britain (1928), Sven Hedin and Folke Bergman from Sweden (Bergman 1939: Hedin 19435), and Huang Wenbi from China (1948) all conducted surveys around the Lop Nur region where they recorded well preserved burials. After World War II, archaeological fieldwork in the Taklamakan Desert was mainly carried out by Chinese scholars. The discovery of the Gumugou Cemetery on the banks of the Kongque River in 1979 opened a new page in the archaeology of the Taklamakan (Han 1986; Wang 1983a, b, 2014), with full scale excavation carried out at the site. Until very recently, the detailed nature of the research was greater than any subsequent excavation in the region. The next major field project in the Taklamakan was the Xiaohe Cemetery where, from 2002 to 2005, archaeologists from the Xinjiang Institute of Relics and Archaeology conducted a complete excavation (XAT 2004, 2005, 2007). Based on the discoveries there, people began to name the ancient remains that appeared along the Talimu-Kongque River Delta the Xiaohe Culture (Xiaohe Archaeological Team 2005). Then a new site called the Northern Cemetery was found in the centre of the

Geographically, the Xiaohe Culture can be categorised into three groups (or variants) in different regions: the northern Lop Nur and Kongque River area (Region A); the Southern Lop Nur area (Region B); and the central Taklamakan area (Region C) (Figures 1, 2). Each cultural region contains a number of cemeteries, sharing common features but also slightly different.

1 

Over the past decade, a number of international publications have discussed the Xiaohe Culture (e.g. Barber 1999; Mallory and Mair 2000; Thornton and Schurr 2004; Yang et al. 2014; Mai et al. 2016; Li 2010; Li et al. 2010, 2015; Zhang et al. 2017). This broad perspective has helped us to understand this unique desert phenomenon, but some discussions may be based on incomplete information which can sometimes lead to inaccurate conclusions. To help resolve this issue, the following paper, based on multidisciplinary studies, gives the reader a relatively complete picture of the Xiaohe Culture. The regional distribution of the Xiaohe Culture

Region A: Northern Lop Nur and Kongque River Region 1. LF Cemetery LF Cemetery (Stein 1988 [1928]: 263-73)2 is located on a steep yardang3 terrace 4 kilometres north-east of Stein’s LE fortress. It was first recorded by Stein in 1914 when eight burials were excavated. The cemetery is surrounded by a rectangular timber enclosure. Each Based on recent fieldwork, the authors have updated the evidence contained in Stein’s report. A yardang is a small streamlined hill or raised terrace carved from bedrock or any consolidated or semi-consolidated material by the dual action of wind abrasion, dust and sand, and deflation. The resulting pattern is a combination of the original rock distribution, and the fluid mechanics of the air flow and resulting pattern of erosion. The term was first brought from the Turkic (steep bank) into English usage by Hedin (1903) when he noted these features in Xinjiang. 2  3 

Also known as Ordek’s Necropolis.

19

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 1. Diagram showing the distribution of the Xiaohe Culture in Xinjiang.

Figure 2. Diagram showing the location of the Xiaohe Culture Cemeteries, Regions A and B.

20

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture burial was placed in a timber coffin made from beams of Populus euphratica, a tree commonly found in the desert river valleys. The tops of the coffins were covered by cattle skins. Each contained a single male body with a woollen hat, linen or woollen cloak and leather boots. The clothing was well made and decorated with complex ornaments. One hat was decorated with feathers and the skin of a small mammal, while another was embroidered with ten rows of red thread. A small bag tied to the edge of a cloak contained wheat kernels. One coffin also contained two woven grass baskets.

three to five large pieces of timber were found at the cemetery, which is likely to contain intact burials. A timber coffin was exposed on another part of the terrace. This was shaped like a boat and constructed with two pieces of smooth curved timber boards. Such boat-shaped coffins are normally without a floor or a cover. Similar coffins, timber poles and two fragmentary mummies, apparently from damaged burials, were sporadically distributed across the surface of the terrace. A male mummy with a woollen hat and clothing was collected from the surface of the terrace. Baskets, wooden pins, bows and leather shoes were also found during the survey.

2. 09LE50 Cemetery This cemetery is located in Ruoqiang County, also on top of a small yardang terrace, 9.4 km north-east of LE fortress (Figures 3-4). It was discovered by local heritage administration officers during a survey in 2003 (Abuduresule and Li 2007). Two enclosures marked by

3. 09LE4 Cemetery This cemetery was found on the top of a yardang terrace during the Third National Relics Survey in 2009, 4.65 kilometres north of LE fortress (Figures 5-8).

Figure 3. Cemetery 09LE50 located on the top of a yardang (photo Xinjiang Institute of Archaeology).

Figure 4. Detail of Cemetery 09LE50 (photo Xinjiang Institute of Archaeology).

21

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 5. Plan of Cemetery 09LE4 showing the location of the graves (illustration Xinjiang Institute of Archaeology).

It is likely to be the LQ cemetery recorded by Stein (1988 [1928]: 263-73). Many timber poles 1030 cm long were found lying on the surface in the northeast area, some with traces of burning. The poles showed some evidence of having been grouped into some thirteen rectangular enclosures. One enclosure measured 1.9 x 0.8 m, with the distance between each pole of around 0.35-0.4 m. Another enclosure measuring 1 x 0.4 m made from timber boards was also found. The upper parts of the boards had been burnt. Some bronze objects were also found near this cemetery; one was a round plate with two holes similar to bronze pieces found in Hami district. Stein also recorded a wooden human figure (L.Q.ii.01) 60 cm long next to the feet of one of the bodies.

Figure 6. Cemetery 09LE4: General view looking south (photo Xinjiang Institute of Archaeology).

Figure 7. Cemetery 09LE4: Burials 3-15, looking west (photo Xinjiang Institute of Archaeology).

22

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture 4. 09LE8 Cemetery This cemetery was found in 2009 during the Third National Relics Survey. It is located on the top of a yardang terrace, 4.7 kilometres north of LE fortress (Figures 9-12). The survey team found two pit-burials which had been looted sometime in the past. The original shape of burial No. 1 (M1) is unclear due to the damage caused by the looting. The shape of burial No. 2 (M2) is relatively clear; it is a rectangular pit measuring 2.1 x 0.9 m, and 1 m deep. In this burial, a body was found inside a boat-coffin with the two curved timber boards measuring 1.9 x 0.25 x 0.25 m. Other discoveries around this grave included 12 pieces of wooden board and a few fragments of cow hide near the body, that of a middle-aged male with dark brown hair and brown facial hair. A woollen netting cloth was tied around his waist. Figure 8. Cemetery 09LE4: Burial 1, looking south-west (photo Xinjiang Institute of Archaeology).

Figure 9. View of Cemetery 09LE8 on top of the yardang (photo Xinjiang Institute of Archaeology).

Figure 10. Cemetery 09LE8: Burial 1 (photo Xinjiang Institute of Archaeology).

23

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads draped in cow hide. A basket with a net pattern was found on the top of the coffin. Inside the coffin was the mummified body of an old female wrapped in brown woollen cloth, with two woollen hats. She was wearing a pair of leather shoes and a number of small bells were found at her waist. In the cloth there were three small pockets, one purposely filled with a bunch of ephedra branches. The clothes were fastened with seven pins, six of wood and one of bone. A cloth doll and a wooden comb were also found near her body. 6. Burial No. 37, Kongque River Delta Cemetery This burial is located inside a ruined house on high land southwest of burial No. 36. Sven Hedin excavated it in 1934 and gave the house the number II. Whether the house and burial are associated cannot be ascertained. The burial comprised a wooden coffin and an incomplete human skeleton associated with a pair of leather shoes, a basket and tools for making the basket (Figure 15). 7. A cemetery found on the northern bank of the Kongque River by Huang Wenbi In 1930, the Chinese archaeologist Huang Wenbi (1948: 98-101) found this cemetery and excavated two burials. At the end of one of the burials was an upright wooden pole. There was a wooden coffin covered with black cow hide containing a body wearing a woollen hat and wrapped up in woollen clothes. Two baskets were placed

Figure 11. Cemetery 09LE8: Mummified body. Burial 1 (photo Xinjiang Institute of Archaeology).

Figure 12. Cemetery 09LE8: Burial 2 (photo Xinjiang Institute of Archaeology).

5. Burial No. 36, Kongque River Delta Cemetery

next to the head. The mummified body had a tattoo with three red and green strokes on the forehead. Another burial marked by a timber pole on the surface of the ground was disturbed sometime before the excavation. A rectangular rock about 1 m long was erected inside the burial pit. A basket, a painted pot, six bone drills

This burial was found on the top of a terrace located northwest of LQ (Bergman 1939: 136-139) and excavated by Sven Hedin and Chen Zongqi in 1936 (Figures 13-14). The coffin was made of a hollowed out half tree trunk 24

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture

Figure 13. Artefacts collected from Kongque River Delta Cemetery Burial 36 (after Bergman 1939: Pl. 26).

Figure 14. Mummified body found at Kongque River Delta Cemetery Burial 36

(after Bergman 1936: 136, Fig. 32).

25

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads chest. Her whole body was covered with tree branches and reeds. She wore a woollen hat and was wrapped in a woollen blanket and sheepskin. A basket and a wooden comb were found next to her left shoulder. 9. Gumugou Cemetery Gumugou Cemetery (Wang 1983a, b, 2014) is situated on the north bank of the dry Kongque River bed, 85 km west of the Loulan fortress (Figures 18-21). Stein visited the site, recording it as LS Cemetery, and excavated several graves (Stein 1988 [1928]: 263-73). In 1979, 42 burials were excavated by the former Archaeological Institute of the Xinjiang Academy of Social Science. Two groups of burials were identified by the excavators. In the first group 36 burials were almost identical to those from Xiaohe Cemetery. The burials in this group usually had two timber poles at the eastern and western ends of the grave. Between the timber poles was a wooden coffin. Most coffins contained a body with a woollen hat and leather shoes, wrapped in a woollen blanket with some jade or bone objects, and with a small bag filled with ephedra leaves lying on the chest. Artefacts were typically found next to the body; these included baskets containing wheat kernels and other food, a wooden basin (or bowl, cup), and cups made of animal horn. Some burials in this cemetery contained stone figures.

Figure 15. Basket collected from Kongque River Delta Cemetery 37 (after Bergman 1939: Pl. 25.5).

and some jade, stone and bone artefacts were found in the grave. 8. Tiebanhe Cemetery This cemetery is located near Lop Nur, close to the mouth of the Kongque River (Figures 16-17). Two burials in this cemetery were excavated in 1979 by staff from the Archaeological Institute of Xinjiang Academy of Social Science (Mu 1995: 387). During the excavation, they found one well preserved pit burial with two large tree trunks erected on either end of the grave. The body of a woman was found with baskets laid on her head and

Figure 16. Tiebanhe Cemetery (photo Liu Yusheng).

Figure 17. Female mummy found at Tiebanhe Cemetery 2 (photo Xinjiang Institute of Archaeology).

26

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture

Figure 18. General view of the Gumuguo Cemetery (photo Xinjiang Institute of Archaeology).

Figure 19. Radial burial patterns in the Gumuguo Cemetery (photo Xinjiang Institute of Archaeology).

Figure 20. Excavated burials at the Gumuguo Cemetery (photo Xinjiang Institute of Archaeology).

where timber boards were used to make an enclosure attached to the walls of the burial pit. The tops of the boards rose above ground level to form a rectangular enclosure on the surface.

The second group comprised six graves, each with seven circles made of small timber poles, forming a radial pattern around the burial. Based on the report written by Stein (1988 [1928]), this cemetery had a third group, 27

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 21. Evidence of looting at the Gumuguo Cemetery (photo Xinjiang Institute of Archaeology).

10. LT Cemetery

local officers during the Third National Archaeological Survey in 2009. The burials in this cemetery are distributed within an area of about 700 sq m; most have been seriously looted. The description presented here is based on seven damaged burial pits.

This cemetery was discovered by Ellsworth Huntington in 1906 and excavated by Stein in 1914 (1988: 263-73) (Figure 22). The cemetery, consisting of 22 burials, lies on the top of a pebble mound, 8 km east of the Gumugou Cemetery. The burial practices in this cemetery are close to those in Gumugou, specifically the timber board enclosure. Stein excavated six burials and found only a few incomplete human skeletons but no coffins.

These seven graves each lie within rectangular pits 2.5-3 m long, 1.2-1.6 m wide and 1-2 m deep. Timber coffins formed with the curved planks of Populus euphratica are similar to those found in the Xiaohe Cemetery. The coffins here are also surrounded by enclosures of timber planking. On the exterior the enclosure is usually surrounded by a larger circle built of small timber poles. Some artefacts collected during the survey included wooden anthropomorphic carvings, poles, sheep skins, woollen cloth, baskets, small timber boards and human bones.

11. Xianshuiquan Cemetery No 10 The cemetery is located on a pebble mound along the foothills of Kuluketage Mountain about 6 km north of the dry Kongque River bed, 2.5 km northwest of the Loulan Fortress (Figures 23-27). It was discovered by

Figure 22. A timber enclosed burial chamber at LT Cemetery (photo Xinjiang Institute of Archaeology).

28

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture Region B: The Southern Lop Nur Region: Xiaohe Xiaohe cemetery is the only site discovered in this area. It is situated south-west of Lop Nur, around 4 km east of Xiaohe, near the small dry branch of the Kongque River (XAT 2004, 2005, 2007). The cemetery forms a small sand mound 7 m high, 74 m long and 35 m wide (Figures 28-32). The site was first discovered by the Swedish explorer Folke Bergman in 1934 when he excavated 12 burials. Sixty-six years after Bergman, in 2000, Chinese archaeologists visited the site for the first time. From 2002-2005 the Xiaohe site was completely excavated by staff from the Archaeological Institute of Xinjiang Uygur Autonomous Region. The excavation uncovered the entire cemetery and recorded a total of 167 burials (Figures 33-41).

Figure 23. Plan showing the location of graves at Xianshuiquan Cemetery No 10 (illustration Xinjiang Institute of Archaeology).

Figure 24. View of Xianshuiquan Cemetery No 10 (photo Xinjiang Institute of Archaeology).

Figure 25. Timbers found at Xianshuiquan Cemetery No 10 (photo Xinjiang Institute of Archaeology).

29

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 26. Carved timber found at Xianshuiquan Cemetery No 10 (photo Xinjiang Institute of Archaeology).

Figure 27. Basket found at Xianshuiquan Cemetery No 10 (photo Xinjiang Institute of Archaeology).

Figure 28. The Xiaohe Cemetery before the excavations of 2002-2005 (photo Xinjiang Institute of Archaeology).

Figure 29. The Xiaohe Cemetery: view from southeast to northwest (photo Xinjiang Institute of Archaeology).

30

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture

Figure 30. The northern section of the Xiaohe Cemetery before excavation (photo Xinjiang Institute of Archaeology).

Figure 31. A general view of the Xiaohe Cemetery before excavation showing the middle wall dividing the northern and southern sections (photo Xinjiang Institute of Archaeology).

Two timber fences were found at the Xiaohe Cemetery. One fence stood across the middle, separating the cemetery into south and north sections, while another fence was built at the southern end. It is probable that there was once a third fence as traces of timber were recovered at the northern end during the excavations. One hundred and thirty-nine burials in five different layers were found in the southern section, but most tombs in the north were damaged by strong winds, particularly the burials on the top layer. There are only 28 burials preserved in the upper layers due to wind erosion.

Types of burials in the Xiaohe Cemetery All the burials can be categorized into three types based on construction details. The first is most common. A coffin was laid in a pit cut into the sand, with a timber pole erected at the head of the grave, to the east or northeast. The form of pole is believed to be linked to sexual symbolism. Each grave contained one long, narrow wooden coffin without a base (Figures 3335). The coffin forms changed in shape through time, straight in the early stages and slightly curved later. 31

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 32. Plan of Xiaohe mound before excavation (illustration Xinjiang Institute of Archaeology).

32

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture

Figure 33. Xiaohe Cemetery: timber coffins (photo Xinjiang Institute of Archaeology).

Figure 34. Xiaohe Cemetery: timber poles with coffins (photo Xinjiang Institute of Archaeology).

Figure 35. Xiaohe Cemetery: details of the timber poles (photo Xinjiang Institute of Archaeology).

33

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Almost all burials are of single individuals; double burials are rare. In most cases the bodies have woollen hats on their heads and are wrapped in woollen cloaks. A basket made of grass is often associated with the dead. There are very few burials with multiple bodies while some tombs have carved wooden anthropomorphic figures inside the coffin. The second type of burials used a clay shell to cover the entire wooden coffin (Figures 36-37). Four such burials have been excavated. Two were found in the southern section in layers 4 and 5, and two more were uncovered in the northern section. All four contained females placed in boat-shape coffins. The clay-covered coffins were surrounded by rectangular chambers made of timber boards attached to the clay shell. Some artefacts were found inside the chambers. Each timber chamber was surrounded by a circle of timber poles roughly 2 m in diameter. The poles were around 2 m high, painted with splashes of red colour and cattle skulls were placed on some of them (Figure 36). Two other clay shell burials found in the northern section were damaged by wind. Based on the number of cattle heads collected in the north section, it is likely that there were at least seven clay shell burials there.

Figure 37. Xiaohe Cemetery: Burial 75. Detail of the clay shell covering of the coffin (photo Xinjiang Institute of Archaeology).

Figure 38. Xiaohe Cemetery: The horns of cattle (Bos taurus) from Burial 13 (photo Xinjiang Institute of Archaeology).

The elaborate nature of this type of grave and the large number of cattle and sheep heads discovered suggests that the people buried in them might have been of higher social status. The artefacts and animal heads may well be sacrifices associated with ceremonies of mourning. There is only one example of the third type of burial (BM28), found at the northern end of the north section. It stands out from the others on the basis of its form and associated artefacts even though it was seriously damaged. It is also the largest burial in the entire

Figure 36. Xiaohe Cemetery: Burial 75. Coffin with clay shell covering surrounded by poles (photo Xinjiang Institute of Archaeology).

34

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture cemetery. The structures marking the grave comprised a rectangular chamber 7 sq m in extent and 1.5 m deep. It had a trapezoidal entrance built with wooden polygonal poles and planks. A timber wall divided the chamber into two rooms; the interiors of the rooms were painted with red or black ‘S’ designs combined with vertical stripes and net patterns. Several layers of cattle skin covered this wooden structure. Clay was piled up on either side of the chamber, and on top of the clay were seven layers of cattle skulls. The entrance to the chamber was simply constructed using timber planks and poles. A timber pole 3 m tall, decorated with eight incised grooves was erected in front of the entrance. No body was found inside the chamber but a few bones from a female were collected from the

nearby sand outside. A stone club head with a timber handle, a bone carved with a human face, a circular bronze mirror and a bronze bell were found inside the chamber. A painted wooden plate, a wooden carving of a human face and over a hundred severed animal heads were found on the top of the grave or collected from other places nearby. The Phases of the Xiaohe Cemetery Xiaohe burials can be divided into two phases based on stratigraphical differences and burial contexts. In the southern section, the latest phase includes the burials found in the upper layers 1, 2 and 3 which share some common features (Figures 39-40). The deeper burials

Figure 39. Xiaohe Cemetery: Plan of burials and timber poles in layers 1-3 (illustration Xinjiang Institute of Archaeology).

35

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 40. Xiaohe Cemetery: Southern section. Burials in layer 2 (photo Xinjiang Institute of Archaeology).

Figure 41. Xiaohe Cemetery: Southern section. Burials in layers 4 and 5 (photo Xinjiang Institute of Archaeology).

recorded from layers 4 and 5 belong to the earlier phase (Figure 41). The major features of the burials in the northern section are similar to those in layers 4 and 5 of the southern section, so they can be assigned to the earlier phase. The differences in cultural features between the two phases are sufficient to suggest that the Xiaohe Cemetery was used over a long period of time.

section and 70 in the northern. In other words, the 167 burials that we excavated comprised less than half of the estimated total original number of some 357 graves. Region C: The Central Taklamakan Region: The Northern Cemetery Only one cemetery, the Northern Cemetery, has been found in this region so far. It is located on the bank of the lower Keriya River in the heart of the Taklamakan Desert. It was discovered by chance when a ChineseFrench archaeological team carried out a survey around this area in 20084 (Figure 42). The cemetery lies

In addition to the 167 preserved graves that we excavated, many timber boards from damaged coffins were scattered around the cemetery. A count of these boards makes it possible to estimate that at least 190 tombs were partially destroyed either by natural erosion or by human disturbance. Of these former 190 graves, about 120 burials have been located in the southern

Documents kept in the Xinjiang Archaeological Institute; DebaineFrancfort and I. Abuduresule 2000.

4 

36

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture

Figure 42. Survey route for the Northern Cemetery (illustration Xinjiang Institute of Archaeology).

Figure 43. Coffins in Northern Cemetery (photo Xinjiang Institute of Archaeology).

of the Tarim-Kongque River is likely to be the major region of its distribution. Along with the discovery of the Xiaohe Cemetery (Figures 50-82), recent research into this culture has greatly improved our understanding of its cultural contexts, chronology, distribution and the relationship between the cemeteries. Studies have expanded into physical anthropology, DNA, metallurgy and environmental analyses (Mei 2009; Yang et al. 2014; Qiu et al. 2014; Li et al. 2010; Li et al. 2015).

on an oval sandy mound 48.5 m long (south to north), 34 m wide (east to west), and 4 m high, covering about 1650 sq m. Wooden poles, carved posts representing sexual symbols, timber coffin planks, human bones, and artefacts such as clothes, boots, hats, pins, baskets, cattle skins, horns, stone slabs, jade beads, painted wood and bone pipes have been found on the surface of the cemetery. Pot sherds, a bronze knife, a jade axe, a grinding stone, stone arrowheads and microlithic blades were also found during the survey. The cemetery is estimated to have contained over one hundred tombs. The burial rites and artefacts found here are similar to those found in the Xiaohe Cemetery (Figures 43-49).

Chronology The basic chronology for each phase of the Xiaohe Tradition is derived from C14 dates which have been obtained from three key sites. Gumuguo is the earliest with dates ranging from 3092-1531 cal. BC (Wang 1983a, 2014; Flad et al. 2010: Table 1), but the single 4th millennium BC date is on a wooden artefact. The rest, including dates on a woollen blanket and a sheepskin,

Preliminary research results about the Xiaohe Culture The material remains of the Xiaohe culture are spread over the central and eastern Tarim Basin, but the delta 37

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 46. Textile found in the Northern Cemetery (photo Xinjiang Institute of Archaeology).

Figure 44. Coffin in the Northern Cemetery (photo Xinjiang Institute of Archaeology).

Figure 47. Female symbol on wooden pole in the Northern Cemetery (photo Xinjiang Institute of Archaeology).

Figure 45. Rescue work in the Northern Cemetery (photo Xinjiang Institute of Archaeology).

C14 AMS dates from the Xiaohe Cemetery (Tabs 1, 2; Flad et al. 2010, Tab. 1). Those from the southern section range from c. 1800-1400 cal. BC; those from the northern section from c. 1950-1500 cal. BC. Five stratified layers of burials were excavated in the southern section while

fall in the range c. 2500-1500 cal. BC. Radiocarbon dates obtained for the Northern Cemetery indicate that it was in use around 1880-1700 cal. BC.5 There are twenty-eight 5 

C14 dates provided by the laboratory of Beijing University.

38

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture

Figure 49. Phallic symbol found in the Northern Cemetery (photo Xinjiang Institute of Archaeology).

Figure 48. Basket found in the Northern cemetery (photo Xinjiang Institute of Archaeology).

Figure 50. Xiaohe Cemetery: Burial 13 covered in cattle hide (photo Xinjiang Institute of Archaeology).

Figure 51. Xiaohe Cemetery: Burial 13. Female body (photo Xinjiang Institute of Archaeology).

39

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 52. Xiaohe Cemetery: Pole with female symbol (photo Xinjiang Institute of Archaeology).

Figure 53. Xiaohe Cemetery: Burial 24. Poles with female and male symbols (photo Xinjiang Institute of Archaeology).

Figure 55. Xiaohe Cemetery: hafted stone axe (photo Xinjiang Institute of Archaeology).

Figure 56. Xiaohe Cemetery: Burials 24 and 28. Wooden arrows with feathered fletching (photo Xinjiang Institute of Archaeology).

40

Figure 54. Xiaohe Cemetery: Pole with male symbol (photo Xinjiang Institute of Archaeology).

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture

Figure 57. Xiaohe Cemetery: Wooden arrow with bone tip

(photo Xinjiang Institute of Archaeology).

Figure 58. Xiaohe Cemetery: stone mace head (photo Xinjiang Institute of Archaeology).

Figure 59. Xiaohe Cemetery: bronze arrowhead (photo Xinjiang

Institute of Archaeology).

Figure 60. Xiaohe Cemetery: Burial 13. Slivers cut from animal ears (photo Xinjiang Institute of Archaeology).

41

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 61. Xiaohe Cemetery: Burial 75. Basket in a wooden bucket (photo Xinjiang Institute of Archaeology).

Figure 63. Xiaohe Cemetery: Wooden bucket (photo Xinjiang Institute of Archaeology).

Figure 64. Xiaohe Cemetery: Horn and leather container (photo Xinjiang Institute of Archaeology).

Figure 62. Xiaohe Cemetery: Woven basket (photo Xinjiang Institute of Archaeology).

the northern section was heavily eroded with only the lowest layers preserved. Xiaohe Phase I, comprising layers 4 and 5 in the southern section of the Xiaohe Cemetery and the preserved graves from the northern section, may be dated at c. 2000 BC to 1700 BC, and Phase II, layers 1 to 3 in the southern section and all other burials destroyed prior to excavation, at c. 1700 BC to 1450 BC or even later. Phase III has no C14 date

Figure 65. Xiaohe Cemetery: Burial 13. Leather bag (photo Xinjiang Institute of Archaeology).

available but on the basis of the bronze dagger found near Lop Nur we would suggest the date of around 1000 BC. 42

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture

Figure 66. Xiaohe Cemetery: Burial 13. String and bead bracelet (photo Xinjiang Institute of Archaeology).

Figure 67. Xiaohe Cemetery: String and bead necklace (photo Xinjiang Institute of Archaeology).

Figure 69. Xiaohe Cemetery: Felt hat with string, fur and feather ornament (photo Xinjiang Institute of Archaeology).

Figure 70. Xiaohe Cemetery: Burial 13. Leather boots (photo Xinjiang Institute of Archaeology).

Figure 68. Xiaohe Cemetery: Golden earring on a bronze mirror (photo Xinjiang Institute of Archaeology).

Figure 71. Xiaohe Cemetery: Decorated feather (photo Xinjiang Institute of Archaeology).

43

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 72. Xiaohe Cemetery: Phallic symbol in wood and string (photo Xinjiang Institute of Archaeology).

Figure 73. Xiaohe Cemetery: Burial 24. Basket and twigs on top of coffin (photo Xinjiang Institute of Archaeology).

Figure 74. Xiaohe Cemetery: Short strings made of animal tendons (photo Xinjiang Institute of Archaeology).

44

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture

Figure 77. Xiaohe Cemetery: Burial 13. Woven and embroidered textile (photo Xinjiang Institute of Archaeology).

Figure 75. Xiaohe Cemetery: Burial 11. Woven textile with tassels (photo Xinjiang Institute of Archaeology).

Figure 78. Xiaohe Cemetery: Burial 13. Human face carved from wood (photo Xinjiang Institute of Archaeology).

Figure 76. Xiaohe Cemetery: Burial 11. Woven textile (photo Xinjiang Institute of Archaeology).

Figure 79. Xiaohe Cemetery: Painted wooden face (photo Xinjiang Institute of Archaeology).

Figure 80. Xiaohe Cemetery: Burial 2. Wooden replica of human body (photo Xinjiang Institute of Archaeology).

45

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 82. Xiaohe Cemetery: Wooden sticks carved with human faces and string decoration (photo Xinjiang Institute of Archaeology).

Figure 81. Xiaohe Cemetery: Wooden figurine with clothing and earrings (photo Xinjiang Institute of Archaeology).

Xiaohe Phase I – the earliest burials

there is the ‘wooden house’ with a female body, and in Phase II the male body in a clay covered wooden coffin encircled by large timber poles. Each seems to hold special significance relative to the rest of the graves.

Remains associated with Xiaohe Phase I include layers 4 and 5 in the southern section of the Xiaohe Cemetery, twenty-eight burials in the northern section of the Xiaohe Cemetery, and the Keriya River Northern Cemetery. They also include the earliest group at Gumugou Cemetery and the burials in Tiebanhe Cemetery. The Keriya Northern Cemetery may have had some burials earlier than Xiaohe Phase I, but due to the damage caused by looting it is impossible to be certain about this. The typical pattern for Xiaohe Phase I is a female burial in a clay-sealed wooden coffin at the centre with a number of single burials surrounding it. Each burial has an upright wooden pole shaped to symbolise a sexual organ. Straight poles with a curved red painted tip represent male symbols while straight poles with a black painted oval end represent female symbols (Figures 52-54). In some cases at Xiaohe and the Northern Cemetery these are reversed so that female symbols occur on male graves and vice versa. A 3-5 m tall timber pole painted in red is also erected at the head of the grave. In both the early and late periods of use at the site, there seems to have been an area of ritual or cult focus in the northern section. In Phase I

Xiaohe Phase II – the later burials Xiaohe Phase II is represented by the burials found in layers 1, 2 and 3 in the southern section of the Xiaohe Cemetery, and all the other burials damaged before the excavation at the Xiaohe Cemetery. It also includes the later group at the Gumugou Cemetery. There are changes in burial rites between Phase I and Phase II. For instance, the number of large red timber poles drops dramatically in Phase II and disappears completely in the latest period (layer 1) of Phase II. There are also fewer burials with sexually symbolic wooden poles. Less than half of the 180 burials have symbolic sexual organs, and less than 30 burials out of 180 have red poles, whereas almost all of the burials in Phase I had both of these attributes. Based on the stratigraphic evidence at Gumugou Cemetery, the burials typical of Xiaohe Phase I are the first at the site. Burials of males with graves surrounded 46

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture

Figure 83. Table showing radiocarbon dates from the Xiaohe Cemetery: Southern Section (Xinjiang Institute of Archaeology).

by a radiating ‘sunburst’ pattern of hundreds of short timber poles appeared in Phase II at the Gumugou Cemetery (Wang 1983a, b, 2014). The large number of short timber poles found associated with damaged burials at Xiaohe suggests that this typically male

‘sunburst’ burial should also exist in Phase II at the Xiaohe Cemetery.6 It has been estimated that over 6  If the ‘sunburst’ burial is a special form, then the Gumugou Cemetery should also have some ordinary burials of the same period.

47

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 84. Table showing radiocarbon dates from the Xiaohe Cemetery: Northern section (Xinjiang Institute of Archaeology).

300 short timber poles were used to construct each individual ‘sunburst’ burial and around 900 poles have been found in Xiaohe cemetery. This suggests that there are at least three of these ‘sunburst’ male burials at Xiaohe.7 Two burials recorded by early explorers might be later than Phase II. Burial 36 found by Sven Hedin and the grave excavated by Huang Wenbi at Kongque River Delta had coffins made from one single log like a canoe. Another burial found near Lop Nur had a mummified body with a costume similar to that found at the Xiaohe Cemetery and should be considered to be part of the Xiaohe Tradition. However, a bronze dagger with a late design is associated with this body.8 This suggests that the Xiaohe Culture may have extended into the late Bronze Age around 1000 BC allowing us to consider the possibility of a Xiaohe Phase III.

the delta of the lower Kongque River and the Tarim River banks (Regions A and B) and the lower Keriya River (Region C). Although there are three distinct regions in which the Xiaohe cultural sites can be found, our study has shown that the Xiaohe Cemetery north-west of Lop Nur (Region B) has more in common with the Northern Cemetery on the lower Keriya River (Region C) than others, although the two cemeteries are 600 km apart. By contrast, the other sites found in the northern area of Lop Nur in the delta and banks of the Kongque River have some common features, but can be differentiated from the Xiaohe and the Keriya Northern Cemeteries. The differences between these two branches of the Xiaohe culture arise mainly in small details such as the patterns decorated on artefacts, the ways in which woollen clothes are made, and the style of costumes and baskets. As a result we speak of the two groups as distinct variants of the Xiaohe Culture: the Keriya-Xiaohe and the GumugouXianshuiquan branches. Why and how these differences between the two branches developed remains unclear. They could have been caused by different living traditions, or by variations in social and economic development within Xiaohe society.

The two main branches of the Xiaohe Culture The Xiaohe Culture is largely located in the central and eastern regions of the Taklamakan Desert, along 70 timber coffins were collected from the surface in the northern sector. 8  This information was provided by Mr. Niu Geng, Bureau of Relics of Ba Zhou. The bronze dagger is stored in the Bazhou Museum. 7 

48

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture Relationships between the Xiaohe Culture and its neighbours

in daily life. Analysis of the small metal items found at Xiaohe indicates that bronze alloy metallurgy was well developed in Xiaohe society. A number of marks left by sharp metal tools used for woodworking have been recorded on the surface of the timber found in burials, suggesting common use of bronze tools in the Xiaohe community. Archaeologists have collected a number of pot sherds from cylindrical vessels with impressed decoration in the Taklamakan Desert over the years, indicating that the Xiaohe people might have used pottery for daily life but not for burials. In addition, a pit house found on a mound on the northern shore of Lop Nur is likely to be related to the Xiaohe culture. Finds associated with the structure include baskets, but also some pottery sherds with painted or impressed designs9. This find has provided an important clue for tracing a more complete picture of the Xiaohe tradition.

To analyse the Xiaohe Culture in relation to neighbouring cultures it is necessary to consider a wide range of archaeological discoveries both temporally and spatially, and to examine the available data for the Neolithic and Bronze Ages within a broad area of Eurasia. Across the eastern Eurasian steppe in the late Neolithic there are some common traits within ceramic assemblages. Vessels are typically cylindrical in form with rounded or flat bases, decorated with complex wave patterns incised or impressed on the surface (Lin 2008). Comparisons could be made between these forms found widely across eastern Eurasia and southern Siberia from the Neolithic to the Bronze Age and the baskets and wooden bucket found at Xiaohe. The wooden human carvings found at Xiaohe are also found within the Eurasian tradition (e.g., Savinov and Podol’skii 1997: 279, Tab. I, 5-7). It appears then that the origins of the Xiaohe culture should be sought in the Eurasian steppe, from where the rich traditions associated with it developed through long distance migration and an extensive period of cultural interaction.

There are some close parallels between artefacts from the Xiaohe Culture and that of Linya/Tianshanbeilu. The baskets and wooden buckets found at Xiaohe are similar to the form of the third pottery group at Linya/Tianshanbeilu (Li 2002). Large numbers of bronze artefacts were discovered in Tianshanbeilu Cemetery and there is a question as to where the technology came from originally. In fact, the bronzes found in both cultures share common features in their shapes and designs. The early technology of bronze alloy casting at Linya/Tianshanbeilu is likely to have been brought into Hami Basin by cultural interaction between Xiaohe and Linya. A bronze mirror and a leaf shaped arrowhead from Xiaohe parallels finds from Linya in Hami (Li 2002). It seems most likely that in the Hami Basin the Xiaohe cultural tradition blended with elements of the Machang and Siba cultures (in the Gansu region: see Figure 1 in this chapter), originally from the Upper Yellow River in East Asia, creating a new culture, the Linya Culture, particularly notable for its bronzes (Liu and Li 2007). Such bronze items and pottery are also found in other cultures in the East Tianshan region. This suggests that after Xiaohe blended into Linya in Hami, the cultural interaction extended throughout a large region of the eastern Tianshan area.

Other regional evidence should also be considered. In the east of Xinjiang, one of the most important Bronze Age discoveries is the Tianshanbeilu Cemetery in Hami City, associated with the so-called Linya Culture (see Figure 1 in this chapter). Li (2009) has identified one among several classes of ceramics which is believed to be the result of acculturation between the Gansu prehistoric tradition in Eastern Asia and the Eurasian style of vessels with pointed bases. However, so far there is no evidence for direct connections between the Eurasian cultures in the west and the Linya Culture of Tianshanbeilu in the east. It could be reasonable to assume that the influence from the steppe cultures comes into the Linya culture from the southwest. The Hami region which the Linya Culture occupied is just to the north-east of Lop Nur and the area of the Xiaohe Culture. Xiaohe burials usually have neither pottery nor bronzes. It is therefore easy to gain the impression that the Xiaohe Culture was aceramic, with little knowledge of bronze working. Yet study of the artefacts found in Xiaohe burials strongly suggests that a large number of items from the graves may have been for ceremonial purposes only. Xiaohe people might not have chosen pottery as appropriate items for grave goods. Moreover, although some grave goods may be items in everyday use, in general burial goods are selected or produced following the dictates of ceremonial needs, strongly influenced by religion, tradition or even by individual preference. In other words, Xiaohe burial goods cannot be representative of the full range of material culture, particularly items in common use

Physical anthropology and DNA analysis of Xiaohe people Although there is little knowledge of the genetic origins of the Neolithic population of the Tarim Basin, large numbers of well-preserved, naturally mummified human bodies and skeletons have provided precious data for the study of physical anthropology and DNA analysis. Some preliminary results are very promising, even though this ongoing study is far from complete. 9 

49

Archaeological records in the Xinjiang Archaeological Institute.

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Based on his physical anthropological study, Han Kangxin (1986) suggests that the people buried at Gumugou can be identified as being of ancient European ancestry, while human bodies found at Xiaohe also appear to have some features pointing to ancient European origins. However, the results of DNA analysis show a much more complex picture than had been previously imagined. Individuals from Xiaohe Phase I (layer 5, Xiaohe Cemetery) carry the gene C4 representing east Eurasian ancestry, but they also carry genes H and K which relate to west Eurasian populations (Li et al. 2010). The first settlers had European paternal lineages, and maternal lineages of European and southern Siberian origin. This suggests that hybridization between East and West Eurasian populations occurred as early as Xiaohe Phase I. The haplotype genes of C4, H and K emerged in the Eurasian steppe more than 10,000 years ago. Presumably, these people met in the Eurasian steppe and hybridized long before migrating to Xinjiang. The genes identified from individuals in Xiaohe layer 4 show more complex variation than those of layer 5, a pattern that continued into Xiaohe Phase II, in layers 3, 2 and 1, where the genes show greater mixing still (Li et al. 2015). The origin of the mitochondrial lineages is more widespread, with the presence of west Eurasian, east Eurasian and Indian lineages. This indicates that the Xiaohe people increasingly exchanged genes with near neighbours soon after moving into the Tarim River valley. This may account for the marked genetic change over time in the Xiaohe population (Li 2010; Li et al. 2010; Li et al. 2015). The later Xiaohe people carried diverse east Eurasian maternal lineages, including a dominance of C4 and C5, generally linked to southern Siberia. It is possible that this mix also includes contact with the settled oasis dwellers of Central Asia, while indirect contact with East Asians is not proven but cannot be excluded.

led to its fall and what role did the environment play in this decline? These are all difficult questions to which we continue to seek answers. It is possible that environmental change accelerated the movement of Xiaohe peoples into peripheral regions. The Xiaohe tradition may have been rapidly acculturated by the painted pottery cultures from the East, leading to the complete disappearance of the Xiaohe tradition in the Tarim Basin. Future research should focus on these painted pottery traditions in the late Bronze Age to discover what traces might be left of the Xiaohe people. Further ethnological analysis is needed on the human remains associated with these painted pottery traditions in peripheral regions such as Hami, Turpan and the Yanqi Basin. In this way, it might be possible to trace the descendants of the Xiaohe people. References Abuduresule, I. and W. Li 2007. Investigation of the dyeing textiles collected from the robbed Tomb near Loulan LE, in L. Damo (ed.) Study on Central Taklamakan Silk Dress Culture. Shanghai: Donghua University Press. [伊弟利斯、李文瑛《楼兰LE 附近被盗墓及其染织服饰的调查》, 载《大漠 联珠－环塔克拉玛干丝绸之路服饰文化考察报 告》, 东华大学出版社 2007 年，59～75 页]. Barber, E. 1999. The Mummies of Ürümchi. London: Macmillan. Bergman, F. 1939. Archaeological Researches in Sinkiang: Especially the Lop-Nor Region. Stockholm: Bökforlags Aktiebolaget Thule. Debaine-Francfort, C. and I. Abuduresule 2000. Keriya, mémoires d’un fleuve. Archéologie et civilisation des oasis du Taklamakan. Mission archéologique franco-chinoise au Xinjiang. Paris: Findakly. Flad, R et al. 2010. Early wheat in China: results from new studies at Donghuishan in the Hexi Corridor. Holocene 20/6:955-965. Han, K. 1986. Study on the skeletons found at Gumugou cemetery in Xinjiang Konghe, Journal of Archaeological Studies 3:361-384. [韩康信：《新疆孔 雀河古墓沟墓地人骨研究》，《考古学报》1986 年 3 期, 361-384 页]. Hedin, S. 1943-45. History of the Expedition in Asia 19271935 (with Folke Bergman). Stockholm: Elanders Boktryckeri. Huang, W. 1948. Archaeology in Lop Nur, Collections of Northwest Scientific Observation Book 1: 98-101. [黄文 弼：《罗布淖尔考古记》，《中国西北科学考察 团丛刊之一》，1948 年，第 98～101 页]. Li, C. 2010. The molecular genetics studies on Xiaohe ancient remains. Unpublished PhD dissertation, Jilin University. [李春香:《小河墓地古代生物遗骸 的分子遗传学研究》，吉林大学博士学位论文, 2010 年 6 月]. Li, C. et al. 2010. Evidence that a West-East admixed population lived in the Tarim Basin as early as the

Conclusion The rise of the Xiaohe Culture occurred around the second half of the 3rd millennium BC in the middle and eastern regions of the Tarim Basin. After the 2nd millennium BC, as the result of cultural development, it formed two variants, the Keriya-Xiaohe and Gumugou-Xianshuihe branches. The origins of the Xiaohe tradition are rather complex; they include some local aspects, some from the northern Eurasian steppe, some from the eastern Eurasian steppe and some from elsewhere. Among these traditions, the Linya/Tianshanbeilu from Hami seems to be closely related culturally, indicating the possibility of cultural interaction with Xiaohe. Many questions raised by the Xiaohe tradition may be addressed by further and more detailed studies of, for instance, the genetic structure of the Xiaohe people. Moreover, where did the Xiaohe people go after the demise of the Xiaohe culture? What 50

Y. Abuduresule, W. Li and X. Hu: The Xiaohe (Small River) Cemetery and the Xiaohe Culture early Bronze Age. BMC Biology 2010 8:15. https://doi. org/10.1186/1741-7007-8-15. Li, C. et al. 2015. Analysis of ancient human mitochondrial DNA from the Xiaohe cemetery: insights into prehistoric population movements in the Tarim Basin, China. BMC Genetics 16/78: 1-11. https://doi. org/10.1186/s12863-015-0237-5. Lin, M. 2002. The origins of Tocharian and its migration. Xinjiang Wenwu (2002)3/4: 69-82. [林梅村：《吐火 罗人的起源与迁徙》，《新疆文物》2002 年，第 3-4 期，69-82 页]. Li, S. 2002. The interaction between northeast China and Central Asia during the second millennium BC: an archaeological perspective, in K. Boyle, C. Renfrew and M. Levine (eds) Ancient interactions: east and west in Eurasia: 171-181. Cambridge: McDonald Institute for Archaeological Research. Li, S. 2009. Study on the first phase in Tianshanbeilu cemetery, Collection in memorial of Mr. Yu Weichao:193-202. Beijing: Relics Press. [李水城： 《天山北路墓地一期遗存分析》，《俞伟超先 生纪念文集》第 193-202 页，文物出版社，2009 年]. Lin, Y. 2008. Two phenomena, one hypothesis, in Collection of Liyun: 246-250. Beijing: China Encyclopedia Press. [林沄：《两个现象，一个假 设》，载《林沄学术文集》，北京，中国大百科 全书出版社，2008年，246-250 页]. Liu, X. and W. Li 2007. A new observation on the origins of Chinese early bronze cultures and related issues. Tibetan Studies 3:1-63. [刘学堂、李文瑛： 《中国 早期青铜文化的起源及其相关问题新探》,《藏 学学刊》第三辑，1-63 页]. Mallory, J. and V. Mair 2000. The Tarim Mummies. New York: Thames and Hudson. Mei, J. 2009. Early Metallurgy and Socio-Cultural Complexity: Archaeological Discoveries in Northwest China, in B. Hanks and K. Linduff (eds) Social Complexity in Prehistoric Eurasia: 215-32. Cambridge: Cambridge University Press. Mu, S. 1995. Discovery and Research on Loulan ancient mummy. Loulan Cultural Studies 384. Urmuqi: Xinjiang People’s Press. [穆舜英：《楼兰古尸的 发现及其研究》，《穆舜英、张平：《楼兰文化 研究论集》384 页，乌鲁木齐：新疆人民出版社 1995 年]. Qiu, Z. et al. 2014. Paleo-environment and paleodiet inferred from Early Bronze Age cow dung at

Xiaohe Cemetery, Xinjiang, NW China. Quaternary International 349: 167-177. Savinov, D. and M. Podol’skii (eds) 1997. Okunevskiy sbornik. St Petersburg: Institute of the History of Material Culture, Russian Academy of Sciences. Окуневский Сборник. Санкт-Петербург: Институт истории материальной культуры Российской Академии наук. Stein, A. 1988 [1928]. Innermost Asia, Detailed Report of Explorations in Central Asia, Kan-Su and Eastern Iran. Oxford: The Clarendon Press. 1988 reprint New Delhi: COSMO. Thornton, C. and G. Schurr 2004. Genes, language, and culture: an example from the Tarim Basin. Oxford Journal of Archaeology 23/1 February 2004: 83-106. Wang, B. 1983a. The primary Research of Archaeological Excavation at Kongque River Gumugou Cemetery. Xinjiang Social Science 1983/1. [王炳华：《孔雀河 古墓沟发掘及其初步研究》，《新疆社会科学研 究》1983 年第 1 期]. Wang, B. 1983b. Several issues in the social life of Gumugou people. Journal of Xinjiang University Philosophy and Social Science 1983/2. [王炳华：《古 墓沟人社会文化生活中几个问题》，《新疆大学 学报 (哲学社会科学版)》1983 年第 02 期]. Wang, B. 2014. Gumugou. Xinjiang Peoples Publishing House, Wulumuqi. [王炳华：《古墓沟》乌鲁木 齐：新疆人民出版社 2014 年]. XAT. Xiaohe Archaeological Team 2004. Survey and excavation report at Xiaohe cemetery in 2002. Frontier Archaeology 3:338-398. [新疆文物考古研究 所小河考古队:《2002 年小河墓地考古调查与发 掘报告》，《边疆考古研究》第 3 辑，2003 年， 第 338. ～ 398 页]. XAT. Xiaohe Archaeological Team 2005. New achievement at Xiaohe Cemetery in Lop Nur. Turfan Studies 1:114-125. [ 新疆文物考古研究所小河考古 队：《罗布泊小河墓地考古发掘的重要收获》， 《吐鲁番学研究》2005 年 1 期，114-125 页]. XAT. Xiaohe Archaeological Team 2007. Brief report on the excavation at Xinjiang Lop Nur Xiaohe cemetery in 2003. Cultural Relics 10:4-42. [新疆文物考古研究 所小河考古隊:《新疆罗布泊小河墓地2003年发掘 简报》，《文物》2007 年，10 期，4-42 页]. Yang, R. et al. 2014. Investigation of cereal remains at the Xiaohe Cemetery in Xinjiang, China. Journal of Archaeological Science 49 (2014) 42-47.

51

Chapter 4

Tianshan as a Bridge: New Studies of Bronze Age Archaeology in the Western Tianshan, Xinjiang, China Dexin Cong Over the past few years, fieldwork has been carried out in the western Tianshan area encompassing the Tianshan mountain ranges of western Xinjiang, China and its adjacent areas (Figure 1). Recent work has focused primarily along the banks of the Bortala River where archaeological sites have been found ranging in date from c. 2000 to 1000 BC (Jia et al. 2017). In addition to the study of a number of cemeteries, major excavation has been carried out at the occupation site of Adunqiaolu in Wenquan County. As the largest body of archaeological research undertaken in the Bortala River area, the excavation and survey of sites around Adunqiaolu are key to understanding the wider area, which is one of the objectives of this chapter. The Adunqiaolu house site and cemeteries are the most important remains discovered in Xinjiang in recent years, yielding dates

around the 19th to the 15th century BC, belonging to the Middle Bronze Age. The prehistoric archaeology of the eastern Eurasian steppes has been studied in considerable detail by Russian scholars (e.g. Chernykh 2009; Koryakova & Epimakhov 2007; Kuz’mina 2007, 2008; Shishlina 2008), and to a lesser extent by Europeans and Americans (e.g. Frachetti 2008, 2012; Hanks and Linduff (eds) 2009; Hanks et al. 2007; Doumani et al. 2015). Chinese scholarship on neighbouring Xinjiang began more recently and is still limited in scope. Given China’s rapid economic expansion, the current focus is on rescue excavations urgently conducted in response to rapid economic development and construction of infrastructure. The bulk of this new research has been focused on the most

Figure 1. Diagram showing the Bortala River area and adjacent regions.

52

D. Cong: Tianshan as a Bridge: New Studies of Bronze Age Archaeology in the Western Tianshan visible sites, typically cemeteries from various periods and historical occupation sites. Prehistoric occupation sites, such as Adunqiaolu, have received little attention.

the Etuokesai River. The total length of the river is over 252 km, draining a river basin area of about 15,928 sq km (Figure 1).

Xinjiang, originally known in ancient China as the ‘Western Region’, is located in the hinterlands of the Eurasian continent, a key bridging zone between East and West. Archaeological data shows that no later than 2000 BC, or even earlier in the Xinjiang region, cultural interactions took place between Xinjiang and the western parts of Central Asia, and between Xinjiang and the East – that is mainland China (Xiao 2004; Han 2012, Ruan 2013; Betts et al. 2014; Spengler et al. 2014; Qu et al. 2017). In recent years, the pace of fieldwork in the western Tianshan area has increased, particularly in the Bortala Mongolian Autonomous Prefecture. At the beginning of the 21st century, archaeological surveys of Bronze Age sites along the Bortala River, ending with the Third National Relics Survey in 2009, provided a solid foundation for future research (BRXUAR 2011).

The Upper and Middle Bortala Rivers are located in a west-east valley formed by the Alatao and Biezhentao Mountains, which consist of small ranges in the western Tianshan. The river bed lies close to the foothills of the Biezhentao Mountains in the south and is separated from the Alatao Mountains in the north by a wide alluvial fan. The topography of the valley is characterized by several Alatao alluvial fans leading south towards the Bortala River at an elevation of about 2000 m, with low hills and small mountains above. The southern alluvial fans form important areas of open pasture in the semi-arid steppe, drained by a dense network of seasonal streams. Vegetation consists mainly of stipa grass (Stipa sp.), bindweed (Convolvulus arvensis) and other herbaceous plants. With an annual average temperature of 3.6°C the climate is cool, with a temperate, semi-arid, continental climate dominating, despite long hours of sunshine. The area is frost-free between 141 to 166 days a year, with an average annual rainfall of 204 mm.

Based on those surveys, in 2010, the Institute of Archaeology of the Chinese Academy of Social Science (CASS) carried out a systematic archaeological program in Wenquan County, in the Bortala River area, making a number of important finds (AIC et al. 2013). The Bortala River region, the entrance to the West Tianshan, is rich in archaeological remains, but is as yet very much understudied. The basic characteristics and chronology of Bronze Age culture in this region are still unclear and represent a critical gap in our knowledge. Ongoing excavations and research at Adunqiaolu, and the wider Bortala River region, are now leading to a comprehensive and more accurate understanding of the context of Bronze Age culture in the western Tianshan and the Northern Steppes, helping to explore the nature of relationships between Chinese and Eurasian cultures in prehistory.

Because of its special geographical location and mountainous steppe environment, the Bortala River Basin formed an important ancient connection between the western Tianshan of Xinjiang and the Semirech’ye region further west in Kazakhstan. As a key bridge between East and West, the region has provided a platform for various aspects of archaeological and anthropological research, advancing understanding of the distribution of nomads along the Eurasian Steppes, their migration, the interaction between nomadic and agricultural peoples, nomadic cultural influences (especially interactions with the Yellow River Shang-Zhou cultures), and the investigation of cross-influences between prehistoric nomadic and agricultural economies.

The Bortala River Basin The Bortala Mongolian Autonomous Prefecture is located in the north-western Xinjiang Uygur Autonomous Region, connecting Tacheng in the northeast and Yili Kazak Autonomous Prefecture in the south – a total area of 24,934 sq km. To the northwest is the border separating China and the Republic of Kazakhstan. The Bortala River, situated in the middle of the Bortala Mongolian Autonomous Prefecture, is an internal river in western Xinjiang. The upper Bortala River, also known as the Wenquan (Hot-Springs) River, originates between the Alatao and Biezhentao Mountains in the western Tianshan, flowing eastward through Wenquan (Hot Springs) County. After Bole, the river joins the Daheyanzi River in Jinghe County, and from there turns north-east into Aibi (Ebinur) Lake. In Wenquan County, the two main tributaries of the Bortala River are the Wenquan or Hot Springs River and

Archaeological sites and cemeteries in the Bortala River region Ancient Bronze Age sites in the Bortala River region have four main characteristics. First, the sites are generally found on alluvial fans in front of the Alatao Mountains, at elevations of between 2400 to 1600 m, near small seasonal or perennial rivers. Second, the sites are concentrated on the north bank of the upper and middle reaches of the Bortala River, including Etuokesai, Turigen, Hongqi, Bulake, Kazhan, Qionglusitai, Adunqiaolu, Kuketawu, Shaleibiliuke, and Arixiate. Third, sites such as Adunqiaolu and Hu’ertuoleha are often found in the form of clusters of settlements, with sections of the ruins clearly having been built to preconceived plans. Fourth, there appears to be little obvious relationship between the occupation 53

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads sites and cemeteries. Only a few occupation sites and burials have been identified as having been constructed simultaneously. Before I turn to a discussion about Adunqiaolu and cemeteries in the vicinity of that occupation site, I begin below with a more general summary of other occupation sites and burials in the Bortala River area, for these provide a general context for locating and understanding Adunqiaolu.

square enclosures. Well preserved L-shaped, doublestone enclosures have been found on the surface at Turigen and Hu’ertuoleha (Figures 2, 3). Etuokesai Etuokesai is a site within the Turigen occupation cluster. It is located on a piedmont alluvial fan at a location in the Dongdubulage village summer pasture in Wenquan County, close to Turigen Gulley. Its southern aspect faces the Etuokesai Mountains, the other three sides being open, facing the lowlands. The total area of the site is approximately 1.8 sq km. Preliminary investigations have identified typical double-sided stone rectangles (house remains) visible on the surface. One of the stone rectangles is approximately 18 x 20 m, with a doorway on the north side of the rectangle. The height of the stones is about 50 cm. Inside the stone rectangle are crossed, doublestone walls forming four segments. The other rectangle is 25 m long, 22.7 m wide, and the width of the doublestone outer walls is approximately 1.3 m. A burial with a circular shape was set in the centre of this rectangle. A separate, small burial was also found in the corner.

Occupation Sites Occupation sites are situated in three or four concentrated areas. From the Upper to Lower Bortala River are the Kazan, Adunqiaolu, Turigen (Shabibuliuke) and Hu’ertuoleha occupation clusters. These were usually constructed on the flat top of a small hill or foothill, with large stones forming double-circular or

Turigen

Figure 2. General view of the site of Turigen (photo Chinese Academy of Social Sciences).

Figure 3. View of the site of Hu’ertuoleha from above (photo Chinese Academy of Social Sciences).

54

Turigen is located some 22 km northeast of Wenquan Township, on the eastern bank of the Zhuerken-Turigen River, between the Chagantunge Xiang and Kundelun Pastures, about seven km north-west of Shabibuliuke. The site is oriented north to south, covering approximately 50,000 sq m on the lower slopes of the Alatao Mountains. The site combines a settlement and cemeteries. Burial chambers found here are composed mainly of large stone slabs arranged in two parallel rows from north to south. Most house remains are almost square, surrounded by double rows of stone with a doorway on the east side. At the top of the hill is what appears to be a square altar, constructed of three layers of stones. The outer edge of the altar measures approximately 13 m, with an overall height of about 1.5 m. Near the altar is a large isolated upright boulder with a shallow curving depression formed by wind erosion on which patterns of human faces have been engraved. These carved faces are

D. Cong: Tianshan as a Bridge: New Studies of Bronze Age Archaeology in the Western Tianshan associated with the early Bronze Age Qiemu’erqieke culture centred in the Altai, but also spread down the mountains on either side of the Gurbantunggut desert into the northern Tianshan (Jia and Betts 2010). They occur quite separately from other more common rock art clusters depicting animals and stylized human figures (see Rock Art below).

mountain valley, is more than 3,000 m long and 2,000 m wide. Hundreds of the burials are well preserved, either within stone or soil mounds or surrounded by stone enclosures. One unusual stone-enclosure burial is 17 m in diameter and 0.65 m high. The burial is surrounded by an external, double-row, stone circle approximately 30 m in diameter.

Hu’ertuoleha

The Arixiate Cemeteries

Hu’ertuoleha is located in the piedmont valleys in the Alatao Mountains, approximately 10 km northeast of Turigen, in the Huji’ertu zone of the Kundelun Pastures. ‘Hu’ertuoleha’ means ‘twin’ in the Mongolian language, referring to two adjacent hills, one on the east and another on the west, about 400 m apart, separated by a river. On one of the hilltops more than 10 rectangular house sites have been found, each measuring approximately 16 to 18 m north-south and 14 to 16 m east-west. The entire area encompasses about 50,000 sq m.

The Arixiate cemeteries are located on the Arixiate Grassland, northwest of Haribuhuzhen Township, Wenquan County. The tombs are distributed across an Alatao open alluvial fan, surrounded by flat terrain. The entire burial area consists of three large cemeteries, plus a number of sporadically distributed burials outside the cemeteries, totalling 1,479 tombs. The surface forms of the tombs are diverse, including stone mounds, stone circles and sarcophagus tombs. One stone-mound tomb, located on the west bank of the Arixiate River, and the largest and most complex burial construction so far found, is surrounded by a stone circle 70 m in diameter. The stone mound in the centre of the tomb is 2.5 m high. Surrounding the outside of the tomb is a doublerow stone circle, divided internally into 17 similar fanshaped sections. The stone enclosure tombs found at Qiongkusitai and Arixiate cemeteries are noticeably different from those found at Adunqiaolu (see below).

Cemeteries There are many burials at Hu’ertuoleha and more than 200 cemeteries in the Wenquan County area alone. Most of the burials are stone structures, including stonemound tombs, stone-enclosure tombs, stone-mound tombs surrounded by stone enclosures, stone-slab tombs, soil-mound tombs, and human stone-figured tombs.

Rock Art Rock engravings have been found in many locations along the Bortala River. One particularly important cluster was found near the Benbutu Village of Angelige Xiang, Wenquan County, on a small hill on the northern bank of the Etuokesai River tributary.

The Aoluogasheli Cemetery The Aoluogasheli cemetery is located on a foothill of the Alatao Mountains about 60 km west of Wenquan County, in the Hotuerha Pastures of Zhalaimute Xiang. The cemetery covers an area measuring approximately 400 m from east to west, and 800 m north to south. About 40 burials have been found in this cemetery, including simple stone mounds and stone mounds surrounded by stone enclosures. One of the burials is circular, with double stone circles around a stonemound tomb. The diameter of the external edge of the stone circle is about 40 m, with a 1.5 m high circular stone mound in the centre, 18 m in diameter. Another burial had a black pole-like stone erected on its eastern side. The Aoluogasheli cemetery typifies the most common type of burials found in the Bortala River area, which also includes the Wusitebiezhen and Xialebaketu cemeteries. The dates of the burials in the cemeteries extend over a wide range, from the Bronze Age to the Iron Age, up to about the first century BC.

This small hill, overlooking the surrounding flat grasslands, appears dark-brown, a colour distinguishing it from other surrounding areas. More than 1000 petroglyphs, mostly line engravings, have been found on the surface of its rocks. Subjects vary from livestock and wild animals to humans hunting or dancing, with various symbolic objects. Between the Upper and Lower Bortala River, petroglyphs have been found in more than 40 locations. While it is notoriously difficult to date rock art, these petroglyphs have been included here because they form part of the distinguishing characteristics of the region that we are describing. The Adunqiaolu Site Work at the major, ongoing excavation of the Bronze Age site of Adunqiaolu has helped to elucidate the relationship between it and the sites identified above. Adunqiaolu is located 41 km north-west of Wenquan Township, near the Turigen village of Chagantunge Xiang (Jia et al. 2017). The site sits on an open slope

The Qiongkusitai Cemetery The large Qiongkusitai cemetery, located adjacent to the Husita Village of Chagantunge Xiang in the Alatao 55

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads of the Alatao foothills, bounded by gullies and creeks on its eastern and western sides. The Adunqiaolu cemetery lies about a kilometre from the occupation site on a lower-level slope of the foothills. Figure 4 gives absolute dates for the Adunqiaolu occupation sites and

cemeteries. The dates range from the 19th to 15th centuries BC, namely the Middle Bronze Age. Eleven groups of ruins were recorded during the preliminary investigations. Centred on low hills, these

Figure 4. Table showing radiocarbon (AMS 14C) dates for the Adunqiaolu site (Chinese Academy of Social Sciences).

56

D. Cong: Tianshan as a Bridge: New Studies of Bronze Age Archaeology in the Western Tianshan old settlement sites stretch out in all directions to the east, south, west and north. Granite stones shaped by ancient glaciers surround the ruins between the hills and valleys, as far as the north bank of the Bortala River. ‘Adunqiaolu’ means ‘stones like horses’ in the Mongolian language. The Institute of Archaeology of CASS excavated ruins F1, F2 and F3 in 2011 and 2012. The artefacts found include pottery sherds, bronze chisels, animal bones and stone tools. The animal remains have been radiocarbon dated to the 18th and 17th centuries BC.

slabs, projects outward to the south. The exterior of the house measures 22 m in length (18 m inside) by 18 m in width (14.6 m inside). The stone slabs of the F1 wall are large, up to 3 x 1.5 m in size. The house is designed symmetrically, divided into four units, indicating different functional areas. The F1 house has fan-shaped sections in both the southeast and south-west corners. The fan in the south-east corner is built with double rows of stones, while the one in the south-west corner is more circular in shape. Bones of livestock and pottery sherds, as well as layers of clean soil, were unearthed alternately in the two corners. In the middle of F1, double rows of stone walls divide the house into two major south and north sections. Several subsections can also be identified from remains of small stone piles or walls. Several layers of stone piles deposited on the north-east corner subsection inside F1 align in rows with north-south orientations. A large circular stone pile is in the centre of the north-

On the southern slope of a hill on the western side, in a line from its base to the top of the hill (i.e., from south to north), four house sites built with double rows of stone slabs are distributed along the slope. Among these four structures, the site designated F1 is the largest (Figures 5 and 7). F1 is rectangular in shape, built with large stone slabs forming two parallel outer walls, 0.98 to 1.33 m apart. The doorway, also constructed of large stone

Figure 5. Adunqiaolu: House Site F1 (photo Chinese Academy of Social Sciences).

Figure 6. Adunqiaolu: Pottery sherds found inside House F1 (photo Chinese Academy of Social Sciences).

57

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 7. Adunqiaolu: House plan of F1, F2 and F3 (illustration Chinese Academy of Social Sciences).

The Adunqiaolu Cemeteries

east subsection. In the main part of the north-west subsection is another rounded stone mound a single layer high. Two square-shaped subsections form the north-east and north-west corners of F1. Several pits, probably used for storage, were also found inside the house, as were a number of pottery sherds with incised decoration on their surface. (Figure 6).

The burial sites begin about one kilometre from the settlement site. Most of the burials are square stone enclosures, but a few small, stone-mounded tombs are also found on the edges of the primary burial area. Sixty-four square, stone-slab enclosure burials have been recorded in the Adunqiaolu cemetery (Li and Lu 2003, Jia et al. 2017) (Figure 8). Cemeteries similar to Adunqiaolu include the Southwest Cemetery at the Etuokesai Checkpoint, the Kuketawu stone-enclosure cemetery, a cemetery located 76 km from Kazan, and the Gulijiaba Cemetery.

Ruins F2 and F3 have walls built of double rows of stones, and appear to be extensions of house F1, which they adjoin. The shapes of the F2 and F3 structures are more irregular and polygonal than F1. F2 measures more than 17 m in length and more than 14 m at its widest point. The longest diameter of F3 is about 17.8 m on the eastern side of the stone wall. A gap on the west side of F3 is presumably a doorway (Figure 7).

One cemetery is 1800 m south of the small hill where houses F1-F2-F3 are located. The burials are randomly 58

D. Cong: Tianshan as a Bridge: New Studies of Bronze Age Archaeology in the Western Tianshan

Figure 8a and b. A stone cist burial at the Adunqiaolu Cemetery (photo and illustration Chinese Academy of Social Sciences).

to three burials within a single stone-slab enclosure (Figure 10).

placed in a space of about 500 m. Based on the density of the burial distribution, the cemetery can be divided into three zones: north, central and south. More than 60 tombs have been identified in the three zones. All are stone cist burials, surrounded by rectangular or square stone-slab enclosures (Figure 9). Some of the better-preserved tombs also contained wooden coffins. Burial body postures varied and included sideways flexed burials. Some burials involved cremations and dismemberment of the bodies; some tombs were empty, possibly indicating that the bodies had been removed after burial. The largest stone-slab enclosure is square, approximately 9.9-10 m in length. In 2011 and 2012, 11 tombs were excavated, with some containing two

Tomb SM4 is a stone-slab enclosure, located in the northern part of the cemetery. There is a gap in the middle of the north side of the stone enclosure. Within the enclosure are two stone cist burials, SM4-1 and SM4-2. The burial pit is rectangular, with an east-west orientation. Burial SM4-1 lies to the north of SM4-2. Four large slabs form the sarcophagus, which appears to have been modified at some point in the past (Figure 11). The gap between the sarcophagus and the wall of the burial pit was filled with fine gravelly soil. The SM4-1 59

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 9. Adunqiaolu: Stone enclosure of burial SM9 (photo Chinese Academy of Social Sciences).

Figure 10. Adunqiaolu: The northern region of the cemetery (SM4 top centre) (photo Chinese Academy of Social Sciences).

Figure 11. Adunqiaolu: Burial SM4 from above (photo Chinese Academy of Social Sciences).

60

D. Cong: Tianshan as a Bridge: New Studies of Bronze Age Archaeology in the Western Tianshan sarcophagus was well-sealed with solid slabs on each side of the cist. The cover stones consisted of four large, thin stone slabs covered by a 3 cm thick layer of mud plaster. Some burned human bone fragments were found on the bottom of the cist. A space about 30 cm wide separated the two graves. The bottom of the cist was also formed of flat stone slabs. Inside the sarcophagus was a preserved, but highly decayed, timber coffin made of 13-15 cm diameter wood sticks. The north side of the coffin was damaged by the collapse of the cover stone and was almost bare of any trace of wood. The better-preserved side of the coffin was composed of five layers of timber sticks with tenon joints. The well-preserved human skeleton found inside the cist was male, about 30 years old, lying on his side, legs drawn up, body aligned to the north, but head turned to the west (Figure 12). Next to the body were gold-inlaid bronze earrings (Figure 13), pottery (Figure 14) and a sheep talus.

Figure 13. Adunqiaolu: Earring found in burial SM4-1 (photo Chinese Academy of Social Sciences).

Burial SM50, located in the southern region of the cemetery, is a tomb inside a rectangular stone enclosure about 7.1 m long and 2.8 to 3 m wide. The stone slabs for SM50 are smaller and more inclined inward than SM4. There are two burial pits in the SM50 enclosure, with one pit, SM50-1, containing two cists (Figure 15). Cremated human bones were found on the bottom of both cists in SM50-1. A small broken pot was unearthed

Figure 14. Adunqiaolu: Pottery vessel found in burial SM4-1 (photo Chinese Academy of Social Sciences).

near the bottom at the western end of the cists. SM502 contained an almost complete human skeleton which bore no signs of cremation. The body has been identified as a female adult (25-30 years old) lying sideways with flexed legs inside the cist but missing a skull and right lower limb. The head should have been facing toward the west according to the position of the body. Within the sarcophagus to the north, next to the female adult skeleton, was a baby’s skull, scapula, ribs and other remains, implying that SM50-2 was probably a burial for a mother and her infant. Conclusion Given the absolute dates that we have established, the excavation and field survey at Adunqiaolu not only provide the basic chronology and cultural context for the Bronze Age in the area, but also give direction for future research (Shao 2009; Xiao 2004). A large number

Figure 12. Adunqiaolu: Burial SM4-1 (photo Chinese Academy of Social Sciences).

61

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 15. Adunqiaolu: Stone enclosure burial of SM50-1 and SM50-2 (photo Chinese Academy of Social Sciences).

of sites similar to Adunqiaolu have been recorded through field surveys in the past two years, and the number is expected to increase in future survey seasons. The growing body of information demonstrates that during the early to mid-2nd millennium BC, a society represented by the Adunqiaolu site occupied the Bortala River Valley and adjacent areas. Fieldwork at Adunqiaolu has also revealed new information about Bronze Age societies in the western Tianshan Mountains. It illustrates clearly that the cultural influence of Andronovo (Fedorovo) Eurasian peoples spread into western Xinjiang and became well established there (Jia et al. 2017). The specific location of occupation sites F1-F3, on a platform half-way to the top of a hill, with magnificent glacier-clad mountains on one side and unobscured views overlooking a cemetery in the near distance, implies it had special meaning to the people who once lived there. The complex plans and scale of the architecture and the location and landscape of the F1-F3 group of ruins suggest that the site was likely an important communal place frequently used by the local Bronze Age societies located in the Bortala River Basin.

There are many questions yet to be answered in future studies: why and how this society migrated to this area; how did these people adapt to the local environment; how was society was organised – what kind of social system did they have; how did the settlers interact with the local indigenous people and people in neighbouring regions? References AIC et al. 2013. Archaeological Institute of CASS, Bortala Mongolian Autonomous Prefecture Museum and Bureau of Relics of Wenquan County. Adunqiaolu Site and Cemetery, Wenquan County of Xinjiang. Kaogu No. 7: 25-32. [中国社会科学院考古研究所、博尔塔拉 蒙古自治州博物馆、温泉县文物局:《新疆温泉 县阿敦乔鲁遗址与墓地》，《考古》2013年第7 期 25-32 页]. Betts, A. et al. 2014. The Origins of Wheat in China: potential pathways for its introduction. Quaternary International 348/20: 158-168. BRXUAR 2011. Bureau of Relics of Xinjiang Uyghur Autonomous Region 2011. The Fixed Relics of Wenquan County. Document Collection of Third National Relics Survey of Xinjiang Uyghur Autonomous Region. [新疆维吾尔自治区文物局: 《新疆维吾尔自治区第三次全国文物普查资料汇 编·温泉县不可移动文物》，2011 年]. Chernykh, E. 2009. Formation of the Eurasian Steppe Belt Cultures, in B. Hanks and K. Linduff (ed.) Social Complexity in Prehistoric Eurasia: 115-145. Cambridge: Cambridge University Press. Doumani, P. et al. 2015. Burial ritual, agriculture, and craft production among Bronze Age pastoralists at Tasbas (Kazakhstan). Archaeological Research in Asia 1-2, January–April 2015: 17-32, https://doi. org/10.1016/j.ara.2015.01.001.

The Adunqiaolu cemetery offers unique features differentiating it from other cemeteries found in Xinjiang in the past. The types of graves and their contents find their best parallels not further east into Xinjiang but to the west. North of the Alatao Mountains similar burials are recorded at Begazy in the Semirech’ye region in southeast Kazakhstan (Margulan 1979). Basikan River, one of the seven rivers in the Semirech’ye area, is only about 30 kilometres from Adunqiaolu. The excavation of the Adunqiaolu cemetery and the artefacts unearthed from burials such as pottery, gold inlaid bronze earrings and anthropomorphic stone figures record the cultural exchanges that took place between ancient societies in the western Tianshan and the people in the Semirech’ye area of Central Asia. 62

D. Cong: Tianshan as a Bridge: New Studies of Bronze Age Archaeology in the Western Tianshan Frachetti, M. 2008. Pastoralist Landscapes. Berkley: University of California Press. Frachetti, M. 2012. Multiregional Emergence of Mobile Pastoralism and Nonuniform Institutional Complexity across Eurasia. Current Anthropology 53/1 (February 2012): 2-38. Han, J. 2012. The ‘Painted Pottery Road’ and Early Sino-Western Cultural Exchanges. Anabasis 3, Studia Classica et Orientalia: 25-42. Hanks, B. and K. Linduff (eds) 2009. Social complexity in prehistoric Eurasia. Cambridge: Cambridge University Press. Hanks, B. et al. 2007. Towards a refined chronology for the Bronze Age of the southern Urals, Russia. Antiquity 81: 353-67. Jia, P. and A. Betts 2010. A re-analysis of the Qiemu’erqieke (Shamirshak) cemeteries, Xinjiang, China. Journal of Indo-European Studies 38/3-4, Fall/ Winter 2010: 1-43. Jia, P. et al. 2017. Adunqiaolu: new evidence for the Andronovo in Xinjiang, China. Antiquity 91/357: 621639. Jia, P. et al. 2017. Bronze Age Hill Forts: new evidence for defensive sites in the western Tian Shan, China. Archaeological Research in Asia. https://doi. org/10.1016/j.ara.2017.10.005. Koryakova, L. and A. Epimakhov 2007. The Urals and Western Siberia in the Bronze and Iron Ages. Cambridge: Cambridge University Press. Kuz’mina, E. 2007. The Origins of the Indo-Iranians. Leiden: Brill. Kuz’mina, E. 2008. The Prehistory of the Silk Road. Philadelphia: University of Pennsylvania Press. Li, J. and E. Lu 2003. Archaeological Survey at Adunqiaolu site, Wenquan County. Xinjiang Wenwu No. 1：20-27. [李金国、吕恩国:《温泉县阿敦乔鲁遗存的考

古调查和研究》，《新疆文物》2003 年第１期 20-27页]. Margulan A. 1979. Begazy-Dandybaevskaya Culture of Central Kazakhstan. Alma-Ata: Academy of Sciences of Kazakhstan SSR. Маргулан А.Х. 1979. БегазыДандыбаевская культура Центрального Казахстана, Алма-Ата: Академия Наук Казахской ССР. Ruan, Q. 2013. Studies on the discoveries of Andronovo affiliation found in Xinjiang, China. Western Archaeology 7:125-154. [阮秋荣:《新疆发现的安德 罗诺沃文化遗存研究》，《西部考古》第 7 期, 125-154 页]. Shao, H. 2009. Investigation of the Andronovo and Related Relics found in Xinjiang. Frontier Archaeology 8: 81-97. [邵会秋:《新疆地区安德罗诺沃文化相关 遗存探析》,《边疆考古研究所》2009 年第 8 辑 81-97页]. Shishlina, N. 2008. Reconstruction of the Bronze Age of the Caspian Steppes. Life Styles and Life Ways of Pastoral Nomads. BAR International Series 1876. Oxford: Archaeopress. Spengler, R. et al. 2014. Early agriculture and crop transmission among Bronze Age mobile pastoralists of central Eurasia. Proceedings of the Royal Society B 281: 20133382, http://dx.doi.org/10.1098/ rspb.2013.3382. Qu, Y. et al. 2017. Diverse lifestyles and populations in the Xiaohe culture of the Lop Nur region, Xinjiang, China. Archaeological and Anthropological Sciences. DOI 10.1007/s12520-017-0520-7. Xiao X. 2004. Discussion about Xinjiang Prehistory. Study of Northwest, No. 2: 74-83 [肖小勇, 2004.《关 于新疆史前研究的讨论》,《西域研究》2004年第 2 期 74-83 页].

63

Chapter 5

A Report on Archaeological Findings in the Upper Yili Valley Xuetang Liu by large scale rescue excavation associated with major construction work on national infrastructure projects. Fieldwork has been focused mainly on two tributary valleys of the Yili River, the Kashi Valley and the Tekes Valley, but some cemeteries in other areas have also been partially excavated. In all, more than 1000 burials have been excavated. The activity areas of the living, occupation sites, are less easily identifiable and only a few have been explored, although those that have been studied have been excavated extensively, exposing over 1000 sq m in total. Field surveys have also resulted in the recording of an additional 1000 or more burials and 40 rock art sites. The most significant of these are described in this chapter (Figure 1).

The Yili, one of the most important rivers in Central Asia, rises in China in the central Tianshan and runs 1500 km westward, finally flowing into Lake Balkhash in Kazakhstan. Its fertile valley acts as a corridor, connecting east and west, serving as a critical stimulant to cultural interaction. This paper provides the first ever summary of archaeological discoveries in the area of the Upper Yili Valley, from the source of the river to the Huoerguosi delta. As our understanding of the unique prehistory of Xinjiang is still at a very early stage, this preliminary report serves to outline the rich and complex story of the region, offering a basic background accessible to scholars working in the wider region. It presents a broad chronology for the known sites of the Bronze and Iron Ages based on excavations at the occupation site and cemeteries of Qiongkeke. The Yili River Valley served as a gateway from the Eurasian steppe into Xinjiang and the archaeological data recovered there to date reflects a mix of Eurasian influences and locally developed traditions.

Archaeological investigation and excavation of burials are mainly associated with three periods: the Bronze Age, the early Iron Age and the historical period. The Bronze Age and the early Iron Age remains date from the mid-2nd millennium to the late 1st millennium BC. There are more than 1000 cemeteries and 50,000 graves ascribed to these two periods. The cemeteries are distributed on both sides of the Yili Valley and on the banks of its tributaries, including many small secondary tributaries running into tributary deltas. Rock engravings have also been found in the Yili Valley. The key sites have been described below, with most carvings probably dated to the later prehistoric periods.

In 1958, Huang Wenbi was the first to study the prehistory of the Yili Valley, identifying a number of ancient graves in the Chabuchaer County of Zhaosu (Huang 1983: 19-20). Following his work, others began the excavation of a large number of burials in cemeteries including Saerhuobu, Stud Horse Farm (AIXASS 1962), Xiatai1 and Poma (Mu and Wang (eds) 1985) in Zhaosu County (AIXASS 1962); Halatubai in Nileke (XUARM 1988); Yimuchang in Tekesi (Zhang 1989); Stud Sheep Farm in Gongnaisi (AIXASS 1985); Tiemulike (XICRA 1988), Heishanto (Mu and Wang (eds) 1985) in Xinyuan, and Suodunbulake in Chabuchaer, as well as excavation at the Fish Pond occupation site in Xinyuan (XUARM 1991). To date, 266 burials have been excavated in these cemeteries. Survey work has also been conducted at the important copper mining site of Nulasai in Nileke (Wang 1984, Mei and Li 1998), and a substantial number of bronze artefacts have been collected from sites in the Yili Valley (Wang 1962; Wang and Cheng 1989).

Occupation Sites and Cemeteries in the Upper Yili Valley While many cemeteries have been recorded in Xinjiang, very few prehistoric occupation sites have been found. In recent years, however, excavations at the Qiongkeke and Kalasu occupation sites have generated important new evidence for this early era. The prehistory of Xinjiang up until the Bronze Age is represented only by surface scatters of microliths, including cores and blades, that have been found at more than forty locations in Xinjiang (Wang 1985; Xing 1993; Yidilis 1993; Yang 1987). The only example of stratified levels associated with microlithic artefacts is from the basal layers of Qiongkeke. For the subsequent Bronze and Iron Ages, Qiongkeke

In the 21st century, archaeological research in this region has made significant progress, stimulated 1  The excavation report has not been published. For short notes see: AIXASS 1979; Mu and Wang 1985.

64

X. Liu: A Report on Archaeological Findings in the Upper Yili Valley

Figure 1. Diagram showing prehistoric sites in the Upper Yili River Valley.

represents the type site for a basic chronology of the Upper Yili Valley. The Qiongkeke Sites The Qiongkeke Occupation Site Qiongkeke (Figures 2-4), first noted in 2001, lies in the middle reaches of the Kashi River of Nileke County. In the years prior to excavation, the occupation site had been badly damaged, leaving only c. 450 sq m undisturbed. Excavations in 2002 completely uncovered this part of the site, identifying three main stages of usage by humans. The lowest levels were occupation levels of the early prehistoric period, marked by microlithic artefacts, the middle layer, again occupation levels, dated to the Bronze Age. Following abandonment of the occupation site, it was covered by an Iron Age cemetery (XICRA 2003). The lowest levels, those associated with the first use of the site, produced two fragments of microlithic cores (Figure 4). These two cores are the first microlithic remains to be found in Xinjiang in a well-stratified context, in this case below a Bronze Age deposit. The lowest levels also contained three bone arrowheads and a pierced sheep knuckle bone.

Figure 2. Qiongkeke: Ceramics from the settlement site (illustration Liu).

65

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads attached to the surface of the vessel. The stone artefacts included a whetstone and a circular grindstone. The Qiongkeke Cemeteries Qiongkeke Cemetery 1 Qiongkeke Cemetery 1, dated on the basis of Iron Age ceramics, is located on the south bank of the Kashi River, Nileke County, immediately adjacent to the Qiongkeke occupation site. In 2001 the Xinjiang Institute of Cultural Relics and Archaeology excavated  the cemetery, uncovering 50 burials and two altars. The burials in this cemetery could be easily identified by small mounds above the ground (Figure 5). Some graves were surrounded by single or double lines of stone circles. Most were primary burials with only a few secondary burials. A significant number of the skeletons were missing fingers or toes. In some graves the skeletons were missing all the palm bones and in some cases palm bones were missing from both hands. Some hand bones, and foot bones, were broken or scattered. The skeleton in burial 51 looked unusual as the body was not inside the stone coffin. Instead the position of the skeleton seemed to indicate that this person was buried alive and died attempting to climb out of the grave.

Figure 3. Qiongkeke: Ceramics from the settlement site (illustration Liu).

About half the graves contained funerary objects, generally one or two items. Only a few had up to four artefacts. The grave goods included ceramics, iron, wood, stone and bronze items. Most of the iron objects were badly decayed; in only a few cases was it possible to distinguish the shapes of small knives or swords. There were several wooden containers which could be identified as bowls or jars, even though they were poorly preserved. The pottery vessels included jars, cups and bowls (Figure 6). Sheep knuckle bones were found in several burials. Sheep bones were also found inside some of the ceramic vessels, suggesting that meat was placed inside these before burial (Zhang 1990).

Figure 4. Qiongkeke: Microlithic cores from the settlement site (photo Liu).

In the Bronze Age layers, a well-preserved stone hut circle was uncovered, constructed with pebbles forming an enclosed oval 6.2 m long and 6 m wide. Hard clay covered the floor, and pottery sherds and animal bones were found in the overlying occupation deposit. West of this was a hearth pit 50 cm in diameter and 20 cm deep. A burial found outside the stone circle contained a single male body. The skull was represented only by some fragments which seemed to have been cut into pieces and then polished.

The pottery vessels were well made and manufactured by hand. The fabric was reddish, uniform in thickness, with a smooth surface. The shapes were fairly simple, mostly round-bottomed jars without handles. Fourteen examples were decorated with geometrical patterns in triangular or net motifs, or simply parallel lines.

Large amounts of pottery sherds were recovered from the occupation deposits (Figures 2-3), together with a few stone implements, bone objects, and a fragment of bronze. The pottery was hand-made, and the vessels were flat bottomed jars with straight or slightly splayed sides. A few pieces were decorated with incised geometric patterns, impressed parallel lines, or plastic decoration in the form of small clay dots or a thin band

Qiongkeke Cemetery 2 Qiongkeke Cemetery 2 is also located on the south bank of the Kashi River near the Qiongkeke occupation site and again, based on the ceramics, is dated to the Iron Age. The cemetery has been looted. Some graves 66

X. Liu: A Report on Archaeological Findings in the Upper Yili Valley

Figure 5. Qiongkeke: Burial in Cemetery No. 1 (illustration Liu).

bases and occasionally ring bases. These vessels are similar in form to those of the late Andronovo culture of the Eurasian steppe (Jia et al. 2017). Qiongkeke Stage I might be seen as the Andronovo expansion to the east, where it formed localised variants in the upper reaches of the Yili Valley (Zhang 1990). Several cemeteries found in the upper Kashi Valley such as Wutulandi, Caiqiaomen, Kujiergou, Kalasu, Tangbalesayi (Figures 8-9), Kukesuhexi (Figures 10-11), and Shenbukeqi cemeteries can be classified as typical of Qiongkeke Stage I. The Tangbalesayi Cemetery found near the Township of Nileke County contained 24 burials (Figure 8) (Ruan 2013) of which some could be categorised as of Stage I type. In these, the burials were placed in simple rectangular pits, with the body laid on its side and the knees flexed. The funerary objects were different from those found in the other burials in the cemetery. The ceramic vessels were wide-mouthed jars with a slightly curved shoulder and a ring base (Figure 9: 1-12). The earrings (Figure 9: 17-19) were similar to those found in the Eurasian Andronovo cultural complex of the western steppe. The influence of Qiongkeke Stage I spread out into the surrounding area. The Tacheng Weixiao Site, located in the northwest corner of Tacheng city (Li 1991), belongs to this cultural complex, as does the Sazicun Site (XICRA 1996). To the south, the cemetery of Xiabandi in the highland Pamirs also shows connections to Qiongkeke Stage I (XICRA and YHAO 2005).

had sarcophagi built using 30-60 cm long stone slabs, while others were simple earthen pits. Grave goods were rare, usually only one or two pieces in a burial chamber. A small number of graves had four or five pieces. The objects included pottery, iron items, bronzes, sheep bones and golden earrings (Figure 7). Chronology Chronologically, apart from evidence of microlithic industries, the earliest cultural stage in the Upper Yili Valley region, based on stratigraphic evidence, is Stage I, the Bronze Age levels in the Qiongkeke occupation site (Lower Qiongkeke). This is followed by Stage II of Qiongkeke (Upper Qiongkeke), represented by Iron Age graves in the Qiongkeke cemeteries. In the following section other sites in the Yili Valley are discussed in relation to the type sites of Qiongkeke Stages I and II. Qiongkeke Stage I Apart from the Bronze Age occupation site of Qiongkeke, cemeteries of Stage I type are mainly distributed in the upper reaches of the Yili region, especially in the three major tributaries of the Yili River. The pottery vessels found from the burials are simple in design. They are open jars, mainly with flat 67

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 6. Qiongkeke: Ceramics in Cemetery No. 1 (illustration Liu).

Over the years, many bronze artefacts have been collected in the Yili Valley. These date both to the Bronze Age and to the early Iron Age. The Bronze Age collections contain mostly weapons and these should also be classified with Qiongkeke Phase I.

The stone items included natural cobbles, and stone processed to make a variety of objects including vessels, hoes, hammers and grinding stones (Figure 13). The pottery was hand-made and grey in colour. Only a small number of sherds were ornamented with incised lines, nail impressions, triangular incised patterns, impressed broken lines and raised plastic decoration (Figure 14) (AIX, YBCR and YHAO 2008). The vessels are jars, some with shoulders and slightly closed mouths. The forms of the vessels and Dwelling No. 1 conform closely to ceramics and architecture of Andronovo (Federovo) affinity in the Bortala Valley dated from the 19th to the 16th centuries cal. BC, while dates from burials elsewhere in the western Tianshan range from around the mid eighteenth to the mid fourteenth centuries cal. BC, (Jia et al. 2017; Cong in this volume).

Kalasu Occupation Site Not only does Kalasu in Nileke County have an important Bronze Age cemetery (see above), it also has one of the few occupation sites that has been excavated in the region. The site can be dated on typological grounds to the Bronze Age and is affiliated with Qiongkeke Stage I. Kalasu was first found in 2006 and then excavated in 2007. An area of over 475 sq m was opened up, showing the site to be covered by nearly two meters of loess. Below this was a structure, Dwelling No 1, sub-rectangular in shape, 15.2 m long and 9.5-11 m wide (Figure 12). Inside the structure, the floor surface was of hard packed brown earth, with one hearth and several ash pits, as well as a number of postholes. Stone artefacts and pottery sherds were recovered from the occupation deposits.

Qiongkeke Stage II Qiongkeke Stage II (see Chronology above) is represented by the two Qiongkeke Cemeteries. Other cemeteries of Stage II type are densely distributed in the 68

X. Liu: A Report on Archaeological Findings in the Upper Yili Valley

Figure 7. Qiongkeke: Ceramics in cemetery No. 2 (illustration Liu).

between Bronze Age, Andronovo affiliated Qiongkeke Stage I and the less clearly defined, but nonetheless distinct, Iron Age Stage II, the name Yili River Basin Culture is too broad to precisely define the later prehistoric culture of the Upper Yili region. The finds from Qiongkeke provide a more robust typology on which to base a definition and therefore, it is suggested that the Upper Qiongkeke or Qoingkeke Stage II is a more appropriate name for this cultural complex. The basic characteristics of this culture can be summed up in several points:

upper reaches of the Yili Valley, mainly along the Kashi River, in Nileke, Tekesi and Tokkuztara Counties. Most cemeteries that have been found there and excavated can be classified as Qiongkeke Stage II. No occupation sites have yet been found. Sites of this type have previously been given other names. Chen and others have called Stage II sites the Yili River Basin Culture (Chen 2000; Zhang 2002). However, the later prehistoric material cultures identified so far in the Yili Valley are poorly classified, so it is necessary to clarify the cultural context and regional distribution of this complex. A clear definition will not only help our understanding of the later prehistory of the Yili Valley but is also important for the archaeology of Central Asia as a whole. New archaeological discoveries suggest that the prehistoric culture of the Yili Valley is multifaceted, with a variety of different influences. Based on the stratigraphic evidence from the excavation at Nileke Qiongkeke, specifically the clear distinction

a.

b.

69

The cemeteries are usually located on the banks along the upper reaches of the Yili River, where the rainfall is higher and the vegetation more lush. The size of the cemeteries varies. The earlier cemeteries are generally larger. The number of graves ranges from around ten to 100.

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 8. Tangbalesayi: Types of burials (illustration Liu).

c.

d.

e.

Most graves are covered by an earthen mound, but some earlier burials have almost no mound. Instead, these graves are clearly indicated with a ring of stones on the surface of the ground. Later burials are sometimes covered with a huge earthen mound as well as several large stone circles surrounding it. The grave pits are rectangular in shape and some have chambers hollowed out of the side walls. A few graves are lined at the base with pebble paving.

f.

70

The most common burial is a single body in a supine position with the head to the north or north-east. Some burials are secondary, or there is evidence for the burial of several bodies at different times in the same pit. Missing fingers or toes are quite common, which may imply a special burial rite. Some skulls were deliberately perforated. Grave goods are few; usually only one or two items or occasionally three or four. Very few tombs have more funerary objects and it is very

X. Liu: A Report on Archaeological Findings in the Upper Yili Valley

Figure 9. Tangbalesayi: Ceramics and personal ornaments from the burials (illustration Liu).

rare to see burials with rich items. The grave goods can be categorized into two groups. The first group consists of objects for daily living, such as ceramic, wooden and metal artefacts. These items are usually found next to the skull; sheep tail bones are found inside the wooden containers. The second group of funerary objects includes tools and weapons such as iron knives with a long handle, arrowheads and swords. A very few golden objects are found in some burials.

BC. Its early phase falls in the first half of the millennium, the late phase in the second half. The early phase is represented by the first group of Stage II burials at Qiongkeke Cemetery. The mounds are low or absent, the graves are closer together and there is more painted pottery. The late phase can be seen in the second group of Stage II burials at Qiongkeke Cemetery where the graves are more widely dispersed, the mounds are higher and chained along the ridge, while painted pottery becomes less frequent or is absent.

The culture of Qiongkeke Stage II existed in the Upper Yili Valley for the whole of the 1st millennium

It is premature to come to any conclusions about the origins of the Qiongkeke Stage II Culture. The 71

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 10. Kukesuhexi: Types of burials (illustration Liu).

Figure 11. Kukesuhexi: Ceramics from the burials (illustration Liu).

72

X. Liu: A Report on Archaeological Findings in the Upper Yili Valley

Figure 12. Kalasu: Plan of house F1 (illustration Liu).

Figure 13. Kalasu: Stone artefacts from house F1 (scale 1 cm) (illustration Liu).

73

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Xigelike Mountain, south of Gongliu County. It has been completely excavated, revealing c. 100 graves and two altars. Based on the form of construction, there are four different burial types. There are simple earthen pits, pits with a cave-like side-chamber, sarcophagi built with large slabs of stone, and rock-cut graves. Associated grave goods mainly comprise pottery, with lesser numbers of iron, bronze, stone, and gold artefacts. The pottery is hand-made, red in colour, and dominated by jars and round bottomed vessels (Figure 15) (XICRA 2006). The Touhu and Wu’ertangou Cemeteries These two cemeteries, dated to Qiongkeke Stage II, are located in Chabuchaer Xibo Autonomous County. Touhu cemetery comprised 30 burials. Some graves were surrounded by stone circles up to 10 m in diameter. The Wu’ertangou cemetery contained c. 500 burials, most covered by stone mounds (Figure 16). The Jialegesiha’yinte Cemetery The Jialegesiha’yinte Cemetery is located on the south bank of the Kashi River. In 2003 archaeologists excavated it and found 209 burials. Most of these can be compared with Qiongkeke Stage II; only a few are thought to date to the late Han and Jin Dynasties (c. 200 BC – AD 200). There were clear differences in structure between burials. Some were ringed with stone circles while others had mounds of stones covering the grave. A sacrificial pit was found inside one burial mound containing 40 sheep legs, and a number of sheep and cattle heads (XICRA, BRYKAP and BRNC 2006)

Figure 14. Kalasu: Ceramics from house F1 (illustration Liu).

evidence indicates that around 1000 BC, the Stage II Qiongkeke Culture suddenly replaced the Andronovo complex completely in the Upper Yili Valley. It is difficult to make any connections between Qiongkeke Stage I and Stage II. If viewing this culture broadly, including the eastern Tianshan and Upper Yellow River region, it seems that painted pottery has been continually extending westwards since it was first developed in the Upper Yellow River area. Through complex interactions with local cultures, this expansion finally reached the Upper Yili Valley, establishing a unique material culture and replacing the Andronovo culture.

Most graves had a single body as the primary burial, with the head to the north-west and the feet to the south-east. There was only a small number of secondary burials, some with missing skulls. In several cases the lower body was clearly in situ but the upper body seemed to have been disturbed somehow in the past. Some graves contained two bodies, usually male and female adults, perhaps man and wife. One burial with double chambers contained two adults, male and female, plus one female child, possibly a family tomb. In a considerable number of cases the skeletons were missing finger or toe bones, perhaps associated with some special burial customs. This was found to be a very common phenomenon. It was initially believed that it was due to rodent activity but there was no indication of burrows or other evidence of animal disturbance. Only the bones of the hands and feet were affected but there was no sign of gnawing or removal of other smaller bones. It seems therefore more likely that the removal of these bones was linked to some form of burial ritual.

Qiongkeke Stage II also extends into the Lower Yili Valley, entering the Semirech’ye region east of Lake Balkhash. Russian scholars believe this culture was associated with the Saka-Wusun tribes but new research has presented a more nuanced discussion of the Iron Age in this region, highlighting the problems of matching historical records to archaeological realities (Chang et al. 2003). The Gongliu Shankou Reservoir Cemetery This cemetery has finds and grave forms typical of Qiongkeke Stage II. It is located on the foothills of 74

X. Liu: A Report on Archaeological Findings in the Upper Yili Valley large burial mounds were 25 to 60 m in diameter and several metres high. Under the burial mounds there were double burial pits or one pit with multiple chambers. A ditch filled with a large amount of ash was found surrounding the burial pits, possibly evidence of ritual activities related to the funerary ceremony. Based on the associated artefacts, this kind of large burial can probably be dated to around the 3rd to 4th centuries AD.

Figure 15. Gongliu Shankou: Ceramics from the cemetery (illustration Liu).

Unlike the large tumuli, the medium and small mounds are probably prehistoric in date, related to Qiongkeke Stage II. The medium sized mounds were between 15 to 25 m in diameter, and about one m high. They often had a single or double stone circle around the mound. The small burials had mounds less than one m in diameter and were below one m in height. The small burials were rarely looted. They usually had a single burial pit or a pit with a side cave, but only rarely had double pits or chambers.

Most of the burials had few funerary objects. The grave goods, such as pottery vessels (Figure 18), were generally placed near the skull. Small ornaments were found

There were few grave goods; some burials had no associated artefacts at all. The items represented included iron, pottery, wood, stone, bone and gold fragments (Figure 17). Most burials also contained sheep tail bones or vertebrae, or occasionally bones from other animals. The iron objects comprised small knives, swords, rings or tubular pieces. The bronzes were mainly ornaments: hairpins, bracelets and similar items. There were a few bone arrowheads, while gold earrings and gold pellets were found in some graves. The Bietebasitao Cemetery The Bietebasitao Cemetery is located on the south bank of the Kashi River (Liu et al. 2004). The Xinjiang Institute of Cultural Relics and Archaeology excavated c. 100 burials here in 2003. Two main periods of use could be identified, a late one in the first few centuries AD and an earlier prehistoric one broadly of Qiongkeke Stage II type. The graves were distributed from east to west and could be divided into eight sub-sections. The burials were covered with mounds of various sizes; large, medium and small types. Most of the graves had small mounds, about 10 had medium sized mounds and only a very few were categorized as large. The

Figure 16. Wu’ertangou: Plan and cross-section of stone mound burial (illustration Liu).

75

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 17. Jialegesiha’yinte: Ceramics from the burials (illustration Liu).

around and on top of the bodies, suggesting that they were attached to clothing. Apart from pottery, funerary objects included iron items, bronzes, stone, bone, wood and gold. Iron was used for hairpins, and bronzes for knives, hairpins, buttons, bracelets, and larger items such as mirrors. The stone objects were mostly circular grinding stones.

Valley, both dated to Qiongkeke Stage II. In 2001 two altars were found at Qiongkeke cemetery (XICRA 2003). Forty to sixty cm of gravel was laid out as a pavement. On top was placed a rock mound 12.75 m in diameter and 0.8 m high. Another altar was found at Bietebasitao cemetery in 2003. Here the structure looked like a large burial mound, but inside the mound was a roughly square pit 1 m deep, filled with burned ashes. The sides of the pit were reddened and baked hard by repeated exposure to fire. A smaller pit was found next to the large pit, apparently in association. Inside this pit was a stone phallus (Figure 18, No. 9).

The Jirentuohai, Sa’erbulakegou and Tiemulike-goukou Cemeteries The cemeteries of Jirentuohai (Figure 19), Sa’erbulakegou (Figure 20) and Tiemulike-goukou (Figure 21) are dated around the 7-8th centuries BC (XICRA 2004, 1988; Zhou 2004). The graves and the grave goods from the burials share common similarities and they too can be broadly linked to Qiongkeke Stage II.

In 2004, a ritual site was found in south Tiemulikegou, a small tributary of the Kashi River. The altar consisted of an earthen mound two m high and 35 m in diameter. Under the mound on the original ground surface was a stone circle 34 m in diameter. Several layers of pebble paving were found inside the soil mound, with a horse skeleton placed on one of the layers (Zhou 2004).

Ritual Sites in the Upper Yili Valley There are number of sites believed to be ritual altars associated with cemeteries found in the Upper Yili 76

X. Liu: A Report on Archaeological Findings in the Upper Yili Valley

Figure 18. Bietebasitao: Ceramics and stone artefacts from the burials (illustration Liu).

Based on the very limited evidence available for the nature of cult practice in the Yili Valley in this period, it is difficult for now to speculate about the nature and use of these structures.

from the Bronze and Iron Ages, and as such forms an essential part of the cultural history of the Upper Yili Valley. The rock art displays, techniques and styles are similar to those found in the Altai Mountains. The purpose of the rock art is not clear. Other than some human figures, the main subjects are animals such as sheep, deer, horses, camels, dogs and cattle. The engravings show dynamic images of movement: the animals are shown running or jumping, in both realistic and exaggerated styles. There are some thematic scenes, such as herders on horseback or hunters holding a bow and arrow or stick. Images of wolves attacking animals may have deeper meanings. Techniques used were mainly pecking, incising and grinding.

Rock Art from the Upper Yili Valley Other evidence for prehistoric activity in the Yili Valley includes more than 40 rock art sites along the Upper Yili Valley, including the Kashi, Tekesi and Gongnaisi Valleys (Figures 22-23). Rock art is difficult to date, and these may belong to several different periods. Compared to the Altai Mountains in northern Xinjiang and the eastern Tianshan area, the amount of rock art found in the Yili Valley is relatively small but at least some, if not most, very probably dates 77

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 19. Jirentuohai: Ceramic and metal artefacts from the burials (illustration Liu).

Qiongkeke Stages I and II, representing respectively the Bronze and Iron Ages. Sites and graves of Andronovo affinity disappear towards the end of the 2nd millennium BC, to be replaced by traces of people using painted pottery of quite new forms. Painted ceramics represent a tradition that appears at the eastern end of the Tianshan Mountains in the Hami oasis in the early Bronze Age, brought by early oasis agriculturalists from the Hexi corridor and Gansu (Han 2005; Han 2012; Wang et al. 2017). The appearance of painted pottery as a cultural trait in the first millennium BC in the Yili Valley suggests the gradual drift of people and/or cultural traditions along the Tianshan Mountains from the east to the west. Qiongkeke Stage II most probably represents a complex mix of influences from the east, west and

Conclusion The Yili Valley is the gateway from the Eurasian steppe into Xinjiang, acting as a fertile well-watered corridor attracting the agro-pastoral populations of the western grasslands into the heart of the Tianshan Mountains. Apart from water and pasture, the valley also offered copper and coal (Mei 2000; Wang and Ruan 2016; XICRA, BRYKAP and BRNC 2017), key resources that were exploited from the Bronze Age, Qiongkeke Stage I, onwards. The Bronze Age sites and cemeteries of Qiongkeke Stage I are part of the easternmost limits of the greater Andronovo ‘cultural complex’ into the mountain grasslands of the Altai, Tianshan and Pamir regions (Jia et al. 2017). There is no clear relationship between 78

X. Liu: A Report on Archaeological Findings in the Upper Yili Valley

Figure 20. Sa’erbulakegou: Ceramics from the burials (illustration Liu).

Figure 21. Tiemulike-goukou: Ceramics from the burials (illustration Liu).

79

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads [新疆文物考古研究所，伊犁州文物局, 尼勒克 县文物管理所:《尼勒克县喀拉苏遗址考古发 掘简报》,《新疆文物》2008 年 3-4 期, 33-43 页]. Chang, C. et al. 2003. Iron Age society and chronology in South-east Kazakhstan. Antiquity 77/296: 298312. Chen, G. 2000. Preliminary study on the culture of the Xinjiang Yili River Basin. Journal of Eurasian Studies 2: 1-35. Han, J. 2005. Chronological and typological analysis of Xinjiang Bronze and early Iron Age. Xinjiang Wenwu 2005/3: 57-99. [《新疆青铜时代-早期铁器时代文 化的分期和谱系》,《新疆文物》2005 年 3 期, 5799 页]. Han, J. 2012. The ‘Painted Pottery Road’ and Early Sino-Western Cultural Exchanges. Studia Classica et Orientalia Anabasis 3: 25-42. Huang, W. 1983. The Chapter II Yili Survey, in Xinjiang Archaeological Excavation Report: 19-20. Beijing: Cultural Relics Press. Jia, P. et al. 2017. Adunqiaolu: new evidence for the Andronovo in Xinjiang, China. Antiquity 91/ 357: 621-639. Li, X. 1991. New archaeological discoveries in Xinjiang Tacheng. Study of the western regions 1:104. [李肖 《新疆塔城市考古的新发现》,《西域研究》1991 年 1 期, 104 页]. Liu, X. et al. 2004. Major discoveries on the Bietebasitao cemeteries of Nilelke County, Xinjiang. Study of the western regions 2004/1. [刘学堂; 托呼提; 阿里甫: 《新疆尼勒克县别特巴斯陶墓群全面发掘获重要 成果》,《西与研究》2004 年 1 期]. Mei, J. 2000. Copper and Bronze Metallurgy in Late Prehistoric Xinjiang: its cultural context and relationship with neighbouring regions. BAR Int. Ser. 865. Oxford: Archaeopress. Mei, J. and Y. Li. 1998. Preliminary study of ancient copper smelting technology. Study of Natural Sciences 17/3: 289-295. [梅建军, 李延祥;新疆奴拉赛古铜 矿冶遗址冶炼技术初步研究[J]; 自然科学史研究; 1998 年03 期，289-295 页]. Mu, S. and M. Wang (eds) 1985. The ancient peoples of Xinjiang and Cultural Relics. Beijing: Relics Press. [穆 舜英、王明哲:《论新疆古代民族考古文化》, 载 新疆维吾尔自治区考古研究所编:《新疆古代民 族文物》, 文物出版社, 1985 年]. Ruan, Q. 2013. Studies on the discoveries of Andronovo affiliation found in Xinjiang, China. Western Archaeology 7: 125-154. [阮秋荣:《新疆发现的安 德罗诺沃文化遗存研究》, 载西北大学丝绸之 路文化遗产保护与考古学研究中心等编《西北 考古》第七辑, 三秦出版社, 2013 年, 第 125-154 页]. Wang, B. 1962. Ancient bronzes unearthed from Teksi County. Cultural Relics 7-8: 116-118. [王炳华 1962 《特克斯县出土的古代铜器》,《文物》1962 年 7-8 期, 116-118 页].

Figure 22. Examples of the animals depicted on the rock art of the Yili Valley (illustration Liu).

also local developments. Study of the prehistory of the Yili Valley is still at an early stage and the Qiongkeke chronology offers a basic framework for further refinement and study. While there is broad uniformity in the material cultural remains of Stage I, there is visible variety in the ceramic sequences of Stage II. As research progresses, it may become possible to explain the apparent gap between Qiongkeke Stages I and II and explore more closely the variations in the sites that belong to Qiongkeke Stage II. References AIXASS 1962. Archaeological Institute of Xinjiang Academy of Social Sciences. Excavation at ancient tombs in Zhaosu County. The Heritage 7-8:98-111. [中国社会科学院新疆分院民族研究所考古组: 《昭苏县古代墓葬试掘简报》,《文物》1962 年 7-8 期，98-111 页]. AIXASS 1979. Archaeological Institute of Xinjiang Academy of Social Sciences 1979. Archaeological Work Over 30 Years (1949-1979 year). Beijing: Relics Press. [新 疆维吾尔自治区博物馆、新疆社会科学院考古研 究所:《建国以来新疆考古的主要收获》, 载《文 物考古工作三十年 (1949-1979 年) 》，北京: 文物 出版社，1979 年]. AIXASS 1985. Archaeological Institute of Xinjiang Academy of Social Sciences. The Zhongyangchang sarcophagus burials in Gongnaisi of Xinyuan, Xinjiang. Archaeology and Cultural Relics 2. [新疆社会 科学院考古研究所1985《新疆新源巩乃斯种羊场 石棺墓》, 《考古与文物》1985 年 2 期]. AIX, YBCR and YHAO 2008. Archaeological Institute of Xinjiang, Yili Bureau of Cultural Relics and Yili Heritage Administrative Office. Excavation at Kalasu site. Xinjiang Cultural Relics 3-4: 33-43. 80

X. Liu: A Report on Archaeological Findings in the Upper Yili Valley

Figure 23. Examples of the animals depicted on the rock art of the Yili Valley (illustration Liu).

81

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Wang, B. 1985. Preliminary study of microliths in Xinjiang. The proceedings of the Quaternary Research in the Arid Area of Xinjiang. Urumqi: Xinjiang People’s Press. [王炳华:《新疆细石器遗存的初步研究》, 《干旱区新疆第四纪研究论文集》, 新疆人民出 版社, 1985 年]. Wang, B. and Z. Cheng 1989. The bronzes unearthed in Xinjiang Gongliu. Cultural Relics 1989/8: 99-100. [王博、成振国:《新疆巩留出土一批铜器》,《文 物》1989 年 8 期, 99-10 页]. Wang, M. 1984. Ancient copper mine site in Nileke. Yearbook of Chinese Archaeology: 176-177. Beijing: Heritage Press. [王明哲:《尼勒克古铜矿遗址调 查》,《中国考古学年鉴》, 文物出版社, 1984 年, 176-177 页]. Wang, T. et al. 2017. Tianshanbeilu and the Isotopic Millet Road: reviewing the late Neolithic/Bronze Age radiation of human millet consumption from north China to Europe. National Science Review. https://doi.org/10.1093/nsr/nwx015. Wang, Y. and Q. Ruan, 2016. New results from 2015 field work at Jirentai-Goukou site in Nileke county, Xinjiang. The Western Regions Studies 2016/1:132-134. [王永强, & 阮求荣. (2016). 《2015 年新疆尼勒克县 吉仁台沟口考古工作的新收获》. 《西域研究》 (1), 132-134 页]. XICRA 1988. Xinjiang Institute of Cultural Relics and Archaeology. Tiemulike Cemeteries found in Xinyuan. The Heritage 8:63-70. [新疆文物考古研究 所:《新疆新源铁木里克古墓群》,《文物》1988 年 8 期, 63-70 页]. XICRA 1996. Xinjiang Institute of Cultural Relics and Archaeology. Ancient tombs found in Sazicun of Tacheng County. Xinjiang Cultural Relics 1996/2: 14-22. [新疆文物考古研究所等:《托里县萨孜 村古墓葬》,《新疆文物》1996 年 2 期。14-22 页]. XICRA 2003. Xinjiang Institute of Cultural Relics and Archaeology. The excavation of Qiukeke No. 1 Cemetery. Xinjiang Cultural Relics 2-3: 13-53, [新疆文物考古研究所:《尼勒克县穷科克一号墓 地发掘报告》,《新疆文物》2003 年 2,3 期, 13-53 页]. XICRA 2004. Xinjiang Institute of Cultural Relics and Archaeology. Brief report of excavation on the Jirentuohai cemetery of Nileke County, Xinjiang. Xinjiang Relics 3: 60-87. [新疆文物考古研究所:《伊 犁州尼勒克县奇仁托海墓地发掘简报》,《新疆 文物》2004 年 3 期, 60-87 页]. XICRA 2006. Xinjiang Institute of Cultural Relics and Archaeology. The excavation of Shankou Shuiku Cemetery in 2005 of Yili Gongliu. Xinjiang Cultural Relics 1: 1-40. [新疆文物考古所:《2005 年度伊犁州 巩留县山口水库墓地考古发掘报告》,《新疆文 物》, 2006 年 1 期, 1-40 页]. XICRA, BRYKAP and BRNC 2017. Xinjiang Institute of Relics and Archaeology, Bureau of Relics of Yili Kazaks Autonomous Prefecture, Bureau of

Relics of Nileke County. Jirentai – Goukou site of Nileke County in Xinjiang. Kaogu 2017/7: 57-70. [新疆文物考古研究所, 伊犁哈萨克自治州文物 局, & 尼勒克县文物局. (2017). 《新疆尼勒克 县吉仁台沟口遗址》. 《考古》, 2017(7), 57-70 页]. XICRA, BRYKAP and BRNC 2006. Xinjiang Institute of Relics and Archaeology, Bureau of Relics of Yili Kazaks Autonomous Prefecture, Bureau of Relics of Nilke County. Brief report on the excavation at Jialesika’yinte cemetery in Nileke County. Xinjiang Relics 2006/3-4: 1-28. [新疆文物考古研究所、伊犁 哈萨克自治州文物局:《尼勒克县加勒克斯卡茵 特山北麓墓葬发掘简报》,《新疆文物》2006 年 3-4 期, 1-28 页]. XICRA and YHAO 2005. Xinjiang Institute of Cultural Relics and Archaeology, Yili Heritage Administrative Office. The excavation in 2004 at Xiabandi cemetery in Tashikuergan. Xinjiang Cultural Relics 1: 3-12. [新 疆文物考古研究所:《塔什库尔干县下坂地墓地 2004年度考古发掘简报》,《新疆文物》2005 年 1 期, 3-12 页]. Xing, K. 1993. Preliminary study on Xinjiang microlithics. Xinjiang Relics 4:60-71. [邢开鼎:《新疆 细石器初探》,《新疆文物》, 1993 年 4 期，60-71 页]. XUARM 1988. Xinjiang Uygur Autonomous Region Museum. Excavation of the Wu Sun tomb in Halatubai. Xinjiang Cultural Relics 2: 2-18. [新疆维 吾尔自治区博物馆:《尼勒克县哈拉图拜乌孙 墓的发掘》,《新疆文物》1988 年第 2 期, 2-18 页]. XUARM 1991. Xinjiang Uygur Autonomous Region Museum. The excavation at the site of Yuchi in Xinyuan County, Xinjiang. Archaeological and Cultural Relics 3. [新疆维吾尔自治区博物馆文物队:《新疆 新源县七十一团一连渔塘遗址发掘简报》,《考 古与文物》1991 年3 期]. Yang, Y. 1987. Microlithic Sites in Xinjiang. Prehistory Studies 4:17-25. [羊毅勇:《试论新疆的细石器》, 《史前研究》1987 年 4 期, 17-25 页]. Yidilis, A. 1993. Xinjiang Microlithic. Xinjiang Relics 4:1959 [伊弟利斯:《新疆地区细石器遗存》,《新疆文 物》1993 年 4 期, 19-59 页]. Zhang, Y. 1989. Excavation and research on Yili valley mound burials. Xinjiang Cultural Relics 3: 11-22. [张玉 忠:《伊犁河谷土墩墓的发掘和研究》,《新疆文 物》1989 年第 3 期, 11-22 页]. Zhang, Y. 1990. Ancient fortress found in Yili River Valley. Relics and Museum 2:12-19. [张玉忠:《伊犁 河谷新发现的古城堡及相关遗迹》,《文博》1990 年 2 期, 12-19 页]. Zhang, Y. 2002. Archaeological review of Xinjiang. Archaeology 6: 4. [张玉忠:《新疆考古述略》,《考 古》2002 年 6 期, 4 页]. Zhou, X. 2004. Brief report of excavation on the cemeteries of Jialegesiha’yinte and Tiemuli Goukou of Nileke County, Xinjiang. Study of the western 82

X. Liu: A Report on Archaeological Findings in the Upper Yili Valley regions 2004/4. [周小明:《新疆尼勒克县加勒格 斯哈音特和铁木里克沟口墓地考古发掘成果简 述》,《西与研究》2004 年 4 期].

83

Chapter 6

Painted Pottery of the Late Bronze Age – Early Iron Age in Central Asia: New Data from Southern Uzbekistan Nikolaus G.O. Boroffka and Leonid M. Sverchkov The archaeological complex discussed in this chapter, including hand-made painted pottery,1 was first discovered in the upper layers of the settlement mound at Anau in the northern foothills of the Kopet Dagh by the expedition of R. Pumpelly in 1904 (Figure 1). H. Schmidt considered it a ‘period of barbarian occupation’ (Pumpelly 1908). After the excavations at Yazdepe and the periodisation of the stratified material found there, this ‘barbarian’ culture, dated to the early 1st millennium BC, became known in scientific publications as Yaz I, sometimes also as the Early Iron Age Oxus culture (Кuz’mina 1972; Kuz’mina 2008: 281321). In the early stages of research, attention was primarily focussed on the southern and south-western connections of the material from Yazdepe, period I, since the Early Iron Age monuments of northern Iran appeared much better studied (Маsson 1959: 44-48). In the former, painted ceramics disappear around the middle of the 2nd millennium BC but occur again in the first half of the 1st millennium BC (Pogrebova 1977: 160, 162). Perhaps for this reason the period Yaz I was dated 900-650 BC, while the more archaic Chust sites further east were placed later in the second half of the 2nd millennium BC (Маsson 1959: 48, 115).

debate lost its relevance because the synchronisation of the Yaz I culture with the Chust culture of Fergana was confirmed. Those scientists who had supported a neutral position in this debate were right to reject the existence of a common source for the Chust and the Yaz I cultures and instead explained the similarities by a convergence process, based predominantly on previously autochthonic cultures (Маsson 1959: 116; Khlopina and Khlopin 1976; Sagdullaev 1985, 1989; Francfort 2001). Although in recent years the distribution of Yaz I type sites has substantially grown, what Sagdullaev noted 20 years ago is still true – that the problem of the origin of the archaeological, not simply the ‘ceramic’, complexes of the Yaz I type is still far from being resolved (Sagdullaev 1989: 63). Without any pretence that the results may be valid for all the Yaz I type sites, important new data have been recovered in southern Uzbekistan. Research at Bandykhan (Figures 1-2) during the 1970s by E.V. Rtveladze and A.S. Sagdullaev (Rtveladze 1976; Rtveladze 2007) identified several mounds covering periods ranging from the Late Bronze Age to the early Medieval period. They include the site of Maydatepa (then named Bandykhan I), which yielded a ‘clean’ Yaz I related site (i.e. not much disturbed by later settlement) (Figures 2-3). Since 2005 a joint Uzbek-German expedition renewed work around Bandykhan, including the site of Maydatepa (Sverchkov and Boroffka 2007; Sverchkov and Boroffka 2009; Boroffka and Sverchkov 2013).

The analogies between the materials from Yazdepe and Chust are generally accepted, and as a consequence of the discovery of more and more sites of Yaz I type the concept was formed that this culture spread from the Fergana valley into the oases regions of the Amu-darya, the Murgab and even further south (Casal 1961; Zadneprovskiy 1962; Аskarov 1979; Аskarov 1981). Simultaneously, an opposing idea was proposed, with a cultural spread from Iranian Khorasan towards the east and north: to Mundigak and Tillyatepe in Afghanistan and from there to Kuchuktepa and Bandykhan in southern Uzbekistan (Sarianidi 1977, 1981). When the first calibrated radiocarbon dates were published for Yaz I type sites, which placed the phenomenon approximately at 1500-1000 BC, this

The settlement consists of a large mound, with a horseshoe-shaped citadel in the east and a lower surrounding settlement area, mainly to the west. The ‘citadel’ bears a modern cemetery and cannot be studied because it would disturb contemporary graves, but a trench of 20 m x 10 m was excavated in the lower settlement to the west. Architecture, with rectangular rooms containing complex fireplaces (Figure 4) was uncovered. The building material is varied, including clay blocks, loaf-shaped loess bricks and unfired mould-pressed mud bricks sized 42 cm x 21 cm x 11 cm. At least five building phases can be separated in the stratigraphic sequence, which reaches more than 3 m in depth. Between the ‘citadel’ and the excavation in the open settlement, part of an ancient canal was

This denomination is used by us here with reservation, mainly because it is already firmly established in scientific literature. In fact, it is not a good descriptor, since in Central Asia the quantity of vessels with painting usually makes up only 1-5% of the entire pottery from corresponding features. This picture is usually distorted (as here too) by the preferential publication of the painted material.

1 

84

N. Boroffka and L. Sverchkov: Painted Pottery of the Late Bronze Age – Early Iron Age in Central Asia

Figure 1. Diagram showing the distribution of Central Asia Yaz I type hand-made painted pottery, the location of Bandykhan and other sites mentioned in this chapter.

Figure 2. Diagram showing the location of Maydatepa in the Bandykhan microregion.

85

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 3. Excavation view (from the east) of Yaz I architecture in the upper layers of Maydatepa. The round pits in the foreground are later (photo Boroffka, DAI Eurasien-Abteilung).

Figure 4. Detail of a complex fireplace in the Yaz I settlement of Maydatepa. The fireplace with a central round hollow lies on a platform raised above the beaten earth floor. An integrated ash-pit is visible in the west (right in the image) and a plastered basin to the east of the fireplace (left in the image) has not yet been cleaned in this excavation stage. The scale is 2 m long (photo Boroffka, DAI Eurasien-Abteilung).

identified, which functioned during the Yaz I period and had been re-dug several times (Figure 5). This, together with analyses of botanical macro-remains from flotation samples and a study of the animal bones, indicates an agricultural economy including stock-breeding and irrigation farming. The mud brick architecture, usual in Central Asia and of a developed

economy, contradict the old idea of a ‘barbarian’ period and ‘decline’ relative to the preceding Bronze Age, even though the monumental architecture of the latter appears to be lacking. It must, however, be kept in mind that the ‘citadel’ at Bandykhan, where monumental architecture might be expected, could not be excavated due to its modern use as a cemetery. 86

N. Boroffka and L. Sverchkov: Painted Pottery of the Late Bronze Age – Early Iron Age in Central Asia

Figure 5. Partial view of the southern profile at Maydatepa. The brown clay visible at the bottom and to the right is the sterile ground, into which a canal was cut. The canal was filled by blackish and greenish humus layers and re-cut at least twice, as visible in the left part of this section. The scale is 2 m long (photo Boroffka, DAI Eurasien-Abteilung).

Throughout the five building phases the material hardly changed in its principal features. The architecture, as well as the main characteristics of the finds, testifies to an unusually conservative and unbroken continuity. Most of the pottery is typical for the Yaz I type groups: hand-made, usually of light colours or unevenly fired (but in any case oxidizing), while black or grey ware is absent. Of the hand-made ceramics, around 9-10 per cent were painted in red, brown or black, very rarely polychrome, mostly with geometric motifs. The two very common forms, with simpler rim-shapes (Figures 6-7) are open hemispherical bowls of various sizes and round-bodied pots, generally with short rims turned outwards. The hand-made ceramics also include footed bowls (Figure 6: 8, 9), of which especially the miniature versions are reminiscent of the stemmed vases of the Bactro-Margiana Archaeological Complex (BMAC). Other shapes made only by hand are vessels with an oval plan (Figure 6: 1, 11, 15), small rectangular boxes (Figure 6: 18), large flat lids with central strap-handles (the only handles applied to any vessel type) (Figure 6: 2), cauldrons with lateral spouts (Figure 7: 8), rough cylindrical vessels (Figure 7: 10), and convex vessels open at both the wider and the narrower end (probably baking hoods) (Figure 7: 9). In addition, roughly 10 per cent of the total material was wheel-made from, to all appearance, the same clay as the hand-made pottery. The wheel-made vessel shapes (Figure 8) correspond to the hand-made ones, but include only pots and open bowls of various sizes.

(Figure 8: 2-3) and rare decoration with incised wavy lines (Figure 8: 7-9) were well known. Some of the Maydatepa sherds belonged to footed vessels.2 The paste of these BMAC sherds is visually quite different from the Yaz I ware, although ongoing mineralogical analyses will yield more precise information. Although there was evidently some contact between the bearers of the BMAC and those of the Yaz I traditions, the latter are clearly quite alien and stand apart from the local BMAC technology, which was revived again later in the Yaz II period, in which the potter’s wheel and conical formed lower vessel parts are characteristic (Sverchkov and Boroffka 2013; Boroffka and Sverchkov 2013).3 Long-distance trade also appears to have declined, although ornaments made from sea-shells (not freshwater) found at Maydatepa (Figure 9: 1-2) show that, to a degree, far-reaching contacts did still exist. Small abstract anthropomorphic figurines discovered at Maydatepa are the first of their kind known in the Yaz I period. A number of metal objects, especially the sickles and straight knives (Figure 9: 6) were already known from other sites of this period, especially Chust (Sprishevskiy 1974: fig. 26: 1; 27: 10.21). Casting moulds for such objects prove the production of such tools at the site (Figure 9: 7), although no ore sources 2  These wheel-made ceramics are very similar to those known from the site of Mollali in the Mirshade oasis, some 40 km to the northeast of Bandykhan. 3  It is possible that the evolution of culture from BMAC to Yaz II was not so sharply or completely interrupted on the eastern periphery of the former distribution area (for example, in the region of Denau or in north-eastern Afghanistan). Thus, in the Denau region, so far not a single site of Yaz I type is known yet and only two fragments of painted pottery are known from the entire area. Supposedly at Buyrachi there is no interruption of the BMAC tradition (see Sagdullaev 1989: 62).

Amongst the pottery there are a few ceramics of the older tradition of the Bactro-Margiana Archaeological Complex, where wheel-made vessels, often with separately formed lower parts, some footed vessels 87

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 6. Maydatepa: Hand-made Yaz I type pottery (illustration R. Boroffka).

88

N. Boroffka and L. Sverchkov: Painted Pottery of the Late Bronze Age – Early Iron Age in Central Asia

Figure 7. Maydatepa: Hand-made Yaz I type pottery (illustration R. Boroffka).

89

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 8. Maydatepa: Wheel-made pottery from Yaz I contexts (illustration R. Boroffka).

90

N. Boroffka and L. Sverchkov: Painted Pottery of the Late Bronze Age – Early Iron Age in Central Asia

Figure 9. Maydatepa: Various finds from Yaz I contexts. 1 Ocean shell ornaments, 2 abstract anthropomorphous figurines, 3 grooved stone hammer, 4 stone pestle, 5 pedestalled stone mortar, 6 copper-alloy finds, 7 stone casting mould (complete figure and illustrations by R. Boroffka, photo Boroffka, DAI Eurasien-Abteilung).

are known in the nearby area. Future analyses might show whether the metal was possibly obtained from the prehistoric mining regions in Tajikistan, where the deposits at Mushiston (Parzinger and Boroffka 2003: 234-252) could have been accessed following the valleys to Denau in the northeast and then north into the Hissar

mountains. Fragmentary grooved stone hammers (Figure 9: 3), usually associated with mining activity, have been discovered at Maydatepa, but may also have been used in the exploitation of rock salt, which occurs in the wider surroundings (Sverčkov 2009: 151). A copper-alloy omega-shaped pendant (Figure 9: 6, upper 91

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads right) has a good analogy with similar pendants in the Late Bronze Age settlement of Parkhai-Depe of the Sumbar region far to the west (Khlopin 1975: 116, 117, fig. 1:1), again confirming long distance contacts. Small stone pedestalled mortars and pestles (Figure 9: 4-5) are well known from Yaz I type contexts from a nameless site near Mirshade (best images in Pugachenkova et al. 1991: figs 7-8) and from Chust (Sprishevskiy 1974: figs 16: 1; 17: 1-2). The entire Yaz I sequence at Maydatepa has been radiocarbon-dated to cal. BC 1400-1000 (Görsdorf 2007: 132; Boroffka and Sverchkov 2013). The site was then deserted, possibly only for a short period, as some of the pits of the Yaz II time indicate. After this the citadel area has only been used as a cemetery in the modern period.

Fergana valley and the Aketala cultural group, at least four sites of the Chust culture (all in the east of the Fergana valley) yielded, besides the painted pottery, grey-black ceramics, which could be compared to the Aketala group (Zadneprovskiy 1962: 28; Zadneprovskiy 1997: 51). Would it then not be more appropriate to consider the Chust culture of Fergana, together with the Burgulyuk culture of the Tashkent region, simply as the local western version of one cultural phenomenon that extended into the Tarim Basin up to the Late Bronze and Early Iron Age? It seems that the separation of regions with highly similar archaeological complexes is due rather to the history of research, caused by the order in which phenomena were studied by different research teams located within different contemporary administrative borders. On the other hand, the interconnection of the materials of Yaz I type with the cultures of Xinjiang, Fergana and the Tashkent area appears paradoxical since sometimes the analogies between Yaz I and Burgulyuk and similar evidence from Xinjiang are closer than to Chust. As archaeological research advances in the Central Asian valleys between the Amu-darya and Syr-darya rivers, the increasing number of new sites and new layers related to the groups with hand-made painted ceramics means that the Early Iron Age4 is becoming increasingly better understood in this area. Over time it will be more difficult to consider this Central Asian region as self-contained or independent of the material evidence found in the north, east or south (Sverchkov and Boroffka 2009).

Many archaeologists who have dealt with comparative materials for the Yaz I, Chust and Burgulyuk cultures have focused attention on numerous and diverse analogies to the so-called cultures of the painted ceramics of Xinjiang, in particular the monuments from the Tarim Basin (Kiselev 1960: 253; Zadneprovskiy 1962: 107; Masson and Sarianidi 1972: 164-165; Litvinskiy 1981: 159-160; Antonova 1984: 57-58; Kohl 1984: 189191; Zadneprovskiy 1997: 96-98; Lhuillier 2007; Lhuillier 2013). This connection has also been observed by Chinese researchers, especially in the last decade after new research in the Xinjiang Uyghur Autonomous Region of the People’s Republic of China. It is even assumed that the important technological innovation of iron metallurgy was introduced into western China from Eastern Turkestan, i.e. from Fergana (Mei 2000: 6769; Mei and Shell 2002: 229-230). However, H.P. Francfort still rejects such connections between Fergana and the Tarim Basin, based on the fact that between the Chust culture and the southern foothills of the Tianshan lies the Aketala cultural group, in which there is grey pottery, but no painted pottery (Francfort 2001: 228229, 232). Although Lhuillier (2007) notes the striking similarity in the combination objects and especially of decorative motifs on the pottery, in her conclusions she again separates the various characteristics. Thus, for example, she considers the analogy of stone knives as given by the similar economies, while the geometric styles of the painting are thought to be due to simple coincidence (Lhuillier 2013: 189-196). Her argument about the long distances between the sites in Xinjiang and Central Asia can hardly be accepted, if we consider the contacts that did exist. Even Lhuillier (2007: 6) recognizes ‘contacts bilatéraux, plutôt sporadiques’, which can hardly have been only sporadic given the general long-distance connections documented at Maydatepa, both to the east and to the west. In our opinion it is necessary to consider not only the pottery or each artefact and tool individually, but the entire archaeological complex in its totality (Sverchkov and Boroffka 2009). Moreover, besides the complete identity of the stone tools in the painted pottery culture of the

A situation very similar to Maydatepa has been discovered at the settlement and cemetery of Xintala, and in the burials at Quhui, located in the north-eastern part of the Tarim Basin, near the city of Karashar. The Xintala mound occupies an area of roughly 4 hectares and reaches a height of 5 m. The size of the cemetery has not yet been established and it has not been excavated. In Quhui several destroyed graves were studied, but their funerary construction remains indeterminate (Debaine-Francfort 1988: 16-18; Chen and Hiebert 1995: 265, 267; Zadneprovskiy 1997: 96-98; Mei 2000: 10). The Xintala architecture was of unfired bricks, among the earliest evidence for this building material in Xinjiang, with brick sizes of 42 cm x 21 cm x 11 cm. In the pottery production from the settlement two traditions were observed. One is expressed by ceramics with red or brown paint on a light background,5 showing geometric designs – triangles, hatching, zigzags and wavy lines. The other ceramic tradition is represented by black – The distribution of Yaz I type material actually appears to be lacking only where it has not been sought. 5  It is important to stress that the painting was executed on a light background, not on a red one (Zadneprovskiy 1997, 97). 4 

92

N. Boroffka and L. Sverchkov: Painted Pottery of the Late Bronze Age – Early Iron Age in Central Asia brown pottery with the admixture of coarse temper, and is less frequent than the first. The ware, fired under reducing conditions, was decorated with stamps and incisions and included designs in the form of filled triangles and fir trees or fish-bone, which are very characteristic of the Andronovo culture of the steppe. It is not quite clear whether these steppe ceramics were found together with the painted pottery in the same layers or preceded them, but the latter is more likely from the Central Asian point of view. Curved stone knives or sickles ‘of Chust type’, as well as grinding stones, mortars and pestles, hoes and a polished axe of imported nephrite were recovered as well. The finds of a stone mould, a copper-alloy knife and fragments of a copper-alloy vessel, similar to finds at Maydatepa, indicate knowledge and practice of metalworking and a socketed axe of Sejma-Turbino type, unfortunately without context, underlines the connections with the cultures of the steppes. During the excavations of the settlement some grains of millet (Setaria italica) were recovered, whose radiocarbon analysis gave an age of cal. BC 1680-1490, while a second analysis of carbon yielded a date of BC 1700-1470 (Chen and Hiebert 1995: 265, 267). The lifetime of the Xintala settlement thus appears to have been approximately from the early 17th century BC to the mid-15th century BC.

north-eastern China occurred around 1400 BC (Tarasov et al. 2006: 298). If we accept that the climate of northeastern China during this period was steppe-like, then the environment in the Tarim Basin around BC 1400 was probably quite similar to what it is today, namely desert or semi-desert. One region the migrants from Xintala clearly chose as a new habitat was the upper delta of the Bandykhansay (Baysunsay), which must have been similar to the ecosystem familiar to them before the onset of desertification. About BC 1100-1000 the settlement of Maydatepa was deserted, without any obvious signs being visible archaeologically: there are no indications of fire or destruction due to military events or natural catastrophes. Possibly the inhabitants returned towards the north (in a very broad sense), to somewhere between the Kashkadarya valley and Xinjiang. Unfortunately, we do not have radiocarbon dates for the numerous monuments of the same culture with painted ceramics from Kashkadarya or for the Burgulyuk culture of the Tashkent area. For the entire Zaravshan valley there is only one date from the lower layers of Koktepa, which is given as BC 650 ± 50 (Isamiddinov 2002: 64). Although we must, therefore, remain sceptical about this date, we are indeed dealing with an amazingly conservative culture, as we already saw at Maydatepa, where the architecture and the finds hardly show any evolution for around 400 years. Thus, at present we have no other means of constructing a comparative chronology for the Yaz I phenomenon. We must also warn against accepting any unqualified belief in the modern methods of absolute dating, the problems of which are not always completely understood yet (e.g. Kuz’min et al. 2007; Heußner and Boroffka 2013) and which may lead to unjustified high dates in the absolute chronology of monuments, sometimes more correctly dated by classical archaeological methods. Based on the new evidence from Maydatepa, the sites with hand-made painted pottery, previously dated to the 1st millennium BC, now need to be placed in the second half of the 2nd millennium BC. However, this does not necessarily also give a date for their end everywhere, as even the uncertain date from Koktepa may indicate. Painted pottery is sometimes encountered in late Achaemenid contexts (e.g. Sagdullaev 1985: fig. 4: 2126, 28). In the Fergana valley, outside the limits of the Achaemenid Empire, painted pottery remains typical for the inheritors of the Chust culture – the Eylatan and Shurabashat cultures, dated from the middle of the 1st millennium BC to the beginning of the 1st millennium AD. In some cases, painting is still encountered on vessels of the Early Middle Age, especially in the northern Fergana valley, and, as far as we know, the situation is similar in Xinjiang.

The archaeological sites of Xintala and Maydatepa thus appear extremely similar, even the size of the unfired bricks resembling each other. Despite the distance between them it is difficult to assume that this correspondence is by chance. Given that the date for the end use of the Xintala site at around BC 1450 coincides with the beginning of human settlement at Maydatepa, it is hard not to compare the evidence for these two dispersed sites. The material similarities of the painted pottery from Xinjiang and southern Central Asia, together with the chronological sequences, lead us to do more than merely assume that there were close contacts between these two regions. Instead this evidence supports our plea for the idea of actual migrations to be taken seriously. The only difference in the pottery complexes of the two monuments (representing their respective regions) is the presence of other ceramic traditions, which in Xintala are those of the Andronovo culture and at Maydatepa those of the BMAC. This merely demonstrates the influence of the immediately adjacent cultures: in Xinjiang this was Andronovo, in Surkhandar’ya the BMAC tradition. In each case the contacts to the neighbours were most probably of peaceful nature. Thinking about the reasons for such distant migrations, in the first place, we consider the ecological crisis which occurred in the steppe zone of Eurasia during the last third of the 2nd millennium BC and lasted to the beginning of the 1st millennium BC (Yablonskiy 2005: 778). One of the climatic oscillation peaks in

The date for the appearance of hand-made painted ceramics in the Tashkent oasis and the length of their 93

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads persistence, can only be stated approximately, as was the case already 30 years ago. As with the Maydatepa site, the Burgulyuk culture does not have any local predecessors in the Tashkent region. Nor does it have any inheritors. In other words, it appears ‘out of the blue’ and departs to somewhere unknown. The archaeologist Kh. Duke, who mainly studied the Burgulyuk culture, dated it to the 9th – 7th centuries BC: ‘… which does not exclude reservations relative to the conditionality and preliminary character of this dating, since the archaeological studies on the settlements of the Burgulyuk culture are not yet completed’ (Duke 1982: 71-72). It is, therefore, quite possible, as M.I. Filanovich assumed for Shashtepa, that the Burgulyuk culture continued to exist into the 6th – 4th centuries BC or even up to the appearance of the Kaunchi culture with its Sarmatian character, in the 3rd to 2nd centuries BC (Filanovich 1982: 121). Equally, it could happen that in the Tashkent oasis Burgulyuk sites will soon be identified for periods before the 1st millennium BC. We here propose a more or less direct connection between the hand-made painted pottery groups of Maydatepa and Xinjiang. Again, we wish to stress that the evidence for this connection includes the entire complex in both cases: namely, the architecture, pottery, stone and metal tools, as well as the economic and social system. However, we are not suggesting that this is valid for all sites with painted pottery. Instead, we insist on the importance of establishing a separate chronology for each site and micro-environment, including, of course those with hand-made painted pottery. That each individual phenomenon needs to be seen in its entirety is well demonstrated if we remember the well-known recurrence of the so-called ‘pseudo-Tripolie’ pottery in the period from the transition of the 10th-11th century to the 12th century AD, which included all of Maverannakhr at the time and is preserved even today in remote mountain regions.

проблемам истории Центральной Азии в древности (II тысячелетие до н.э.), Душанбе, 17-22 октября 1977 г.: 167-179. Москва: Наука. Boroffka, N. and L. Sverchkov 2013. The Jaz II and III period pottery. Classification and chronology viewed from Bandikhan, Southern Uzbekistan, in M. Wagner (ed.) Pottery chronology of the Early Iron Age in Central Asia: 49-74. Warszawa: Kazimierz Michałowski Foundation, Institute of Archaeology, University of Warsaw. Casal, J.-M. 1961. Fouilles de Mundigak. Mémoires de la Délégation Archéologique Française en Afghanistan 17, Paris: C. Klincksieck. Chen, K. and F. Hiebert 1995. The Late Prehistory of Xinjiang in relation to its Neighbors. Journal of World Prehistory 9/2: 243-300. Debaine-Francfort, C. 1988. Archaéologie du Xinjiang des origines aux Han. Premiẻre partie. Paléorient 14.1: 5-29. Duke 1982. Дуке, Х. 1982. Туябугузские поселения бургулюкской культуры. Ташкент: Фан. Filanovich 1982. Филанович, М.И. 1982. Шаштепа – древнейшее поселение оседлых земледельцев на территории Ташкент, in Г.В. Шишкина (ed.) У истоков древней культуры Ташкента: 91-124. Ташкент: Фан. Francfort, H.-P. 2001. The cultures with painted ceramics of South Central Asia and their relations with the north-eastern steppe zone (late 2nd – early 1st millennium BC), in R. Eichmann and H. Parzinger (eds, Migration und Kulturtransfer. Der Wandel vorderund zentralasiatischer Kulturen im Umbruch vom 2. zum 1. vorchristlichen Jahrtausend. Akten des Internationalen Kolloquiums Berlin, 23. bis 26. November 1999: 221-235. Kolloquien zur Vor- und Frühgeschichte 6, Bonn: Dr. Rudolf Habelt GmbH. Görsdorf, J. 2007. Information zu den 14C-Datierungsergebnissen von Majdatepa (Bandixon I). Труды Байсунской научной экспедиции 3: 132. Heußner, K.-U. and N. Boroffka 2013. Dendrochronological and Radiocarbon dating of samples from Kurganzol (Uzbekistan) in M. Леонид Сверчков (ed.) Курганзол – крепость Александра на юге Узбекистана: 182-184. Ташкент: SMIASIA. Isamiddinov 2002. Исамиддинов, М.Х. 2002. Истоки городской культуры Самаркандского Согда. Ташкент: Издательство народного наследия имени А. Кадыри. Kiselev 1960. Киселев, С.В. 1960. Неолит и бронзовый век Китая (по материалам научной командировки в КНР). Советская археология № 4: 244-266. Khlopin 1975. Хлопин, И.Н. 1975. Поселение эпохи бронзы Пархай-Депе. Краткие сообщения Ордена Трудового Красного Знамени Институт Археологии 142: 116-121.

References Antonova 1984. Антонова, Е.В. 1984. Несколько заметок о первобытной археологии Синьцзяна, in Б.А. Литвинский (ed.) Восточный Туркестан и Средняя Азия. История. Культура. Связи: 55-60. Москва: Наука. Askarov 1979. Аскаров, А.А. 1979. К вопросу о происхождении культуры племен с расписной керамикой эпохи поздней бронзы и раннего железа, in А.В. Виноградов, М.Г. Воробьева, М.А. Итина, Л.М. Левина, Е.Е. Неразик, Ю.А. Рапопорт (eds) Этнография и археология Средней Азии: 34-37. Москва: Наука. Askarov 1981. Аскаров, А.А. 1981. Южный Узбекистан во II тысячелетии до н.э., in М.С. Асимов (ed.) Этнические проблемы истории Центральной Азии в древности (II тысячелетие до н. э.). Труды Международного симпозиума по этническим 94

N. Boroffka and L. Sverchkov: Painted Pottery of the Late Bronze Age – Early Iron Age in Central Asia Khlopina and Khlopin 1976. Хлопина, Л.И. and И.Н. Хлопин 1976. К происхождению комплекса Яз-Тепе I Южного Туркменистана. Советская археология № 4: 200-203. Kohl, P. 1984. Central Asia. Palaeolithic beginnings to the Iron Age. Recherche sur les Civilisations, Synthèse 14. Paris: Éditions Recherche sur les Civilisations. Кuz’mina 1972. Кузьмина, Е.Е. 1972. К вопросу о формировании культуры Северной Бактрии (‘Бактрийский мираж’ и археологическая действительность). Вестник древней истории № 1: 131-147. Кuz’mina 2008. Кузьмина, Е.Е. 2008. Арии – путь на юг. Москва, Санкт-Петербург: Летний сад. Kuzmin, Y., L. Nevesskaya, S. Krivonogov and G. Burr 2007. Apparent 14C ages of the ‘pre-bomb’ shells and correction values (R, ΔR) for Caspian and Aral Seas (Central Asia). Nuclear Instruments and Methods in Physics B 259: 463-466. Lhuillier, J. 2007. Les relations des cultures à céramique modelée peine d’Ouzbekistan et du Xinjiang, in 3ème Congrès du Réseau Asie – IMASIE, 26-27-28 sept. 2007, Paris, France, Maison de la Chimie, École des Hautes Études en Sciences Sociales, Fondation Maison des Sciences de l’Homme. Thématique 6: Espaces, rituels, sociétés, Atelier 33: Transformations des milieux et des peuplements en Asie centrale, 2e-1er millénaire av. J.-C.: nouvelles recherches, viewed 24 Feb. 2012, . Lhuillier, J. 2013. Les cultures à céramique modelée peinte en Asie Centrale Méridionale. Dynamiques socio-culturelles à l’âge du fer ancient (1500-1000 av. n.è.). Mémoires de la Mission Archéologique Française en Asie Centrale 13. Paris: Editions De Boccard. Litvinskiy 1981. Литвинский, Б.А. 1981. Проблемы этнической истории Средней Азии во II тысячелетии до н. э. (Среднеазиатский аспект арийской проблемы), in М.С. Асимов (ed.) Этнические проблемы истории Центральной Азии в древности (II тысячелетие до н. ẹ): Труды Международного симпозиума по этническим проблемам истории Центральной Азии в древности (II тысячелетие до н. э.), Душанбе, 17-22 октября 1977 г.: 154-166. Москва: Наука. Masson 1959. Массон, В.М. 1959. Древнеземледельческая культура Маргианы. Материалы и исследования по археологии СССР 73. Москва, Ленинград: Издательство Академии наук СССР. Masson, V. and V. Sarianidi 1972. Central Asia: Turkmenia Before the Achaemenids, London: Thames & Hudson. Mei, J. 2000. Copper and Bronze Metallurgy in Late Prehistoric Xinjiang. Its cultural context and relationship

with neighbouring regions. British Archaeological Reports (BAR), International Series 865. Oxford: Archaeopress. Mei, J. and C. Shell 2002. The Iron Age Cultures in Xinjiang and their Steppe Connections, in K. Boyle, C. Renfrew and M. Levine (eds) Ancient interactions: east and west in Eurasia: 213-234. McDonald Institute Monographs. Cambridge: McDonald Institute for Archaeological Research. Parzinger, H. and N. Boroffka 2003. Das Zinn der Bronzezeit in Mittelasien I. Die siedlungsarchäologischen Forschungen im Umfeld der Zinnlagerstätten. Archäologie in Iran und Turan 5. Mainz am Rhein: Verlag Philipp von Zabern. Pogrebova 1977. Погребова, М. Н. 1977. Иран и Закавказье в раннем железном веке. Москва: Наука. Pugachenkova, G., E. Rtveladze, K. Kato, B. Turgunov and J. Il’yasov (eds) 1991. Antiquities of southern Uzbekistan / Пугаченкова, Г.А., Ртвеладзе, Э.В., Като, К., Тургунов, Б.А. and Ильясов, Дж.Я. (eds), Древности южного Узбекистана, Hachioji: Soka University Press. Pumpelly, R. (ed.) 1908. Explorations in Turkestan. Expedition of 1904. Prehistoric civilizations of Anau. Origins, Growth, and Influence on Environment 1-2. Carnegie Institution of Washington Publication 73. Washington D.C.: Carnegie Institution of Washington. Rtveladze 1976. Ртвеладзе, Э.В. 1976. Новые древнебактрийские памятники на юге Узбекистана, in В.М. Массон (ed.), Бактрийские древности. Предварительные сообщения об археологических работах на юге Узбекистана: 93-103. Ленинград: Наука. Rtveladze 2007. Ртвеладзе, Э.В. 2007. Археологические исследования в Бандыхане в 1974-1975 гг. Труды Байсунской научной экспедиции 3: 67-95. Sagdullaev 1985. Сагдуллаев, А.С. 1985. О соотношении древнеземледельческих комплексов Ферганы и Бактрии. Советская археология № 4: 21-32. Sarianidi 1977. Сарианиди, В.И. 1977. Древние земледельцы Афганистана. Материалы Советско-Афганской экспедиции 1969-1974 гг. Москва: Наука. Sarianidi 1981. Сарианиди, В.И. 1981. Древняя Бактрия: новые аспекты старой проблемы, in М.С. Асимов (ed.), Этнические проблемы истории Центральной Азии в древности (II тысячелетие до н. ẹ): Труды Международного симпозиума по этническим проблемам истории Центральной Азии в древности (II тысячелетие до н. э.), Душанбе, 17-22 октября 1977 г.: 180-191 Москва: Наука. Sagdullaev 1989. Сагдуллаев, А.С. 1989. Некоторые аспекты проблемы происхождения среднеазиатских комплексов типа Яз I. Советская археология № 2: 49-65. 95

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Sprishevski 1974. Спришевский, В.И. 1974. Каталог археологических материалов эпохи камня и бронзы. Академия Наук Узбекской ССР, Музей Истории Народов Узбекистана им. М.Т. Айбека. Ташкент: Фан. Sverchkov, L. 2009. A history of research on ancient mining in Uzbekistan. Archäologische Mitteilungen aus Iran und Turan 41: 141-164. Sverchkov and Boroffka 2007. Сверчков, Л.М. and Н. Бороффка 2007. Археологические исследования в Бандыхане в 2005 г. Труды Байсунской Научной Экспедиции. Археология, История и Этнография 3: 97-141. Sverchkov and Boroffka 2009. Сверчков, Л.М. and Н. Бороффка 2009. О некоторых проблемах археологического исследования культур расписной керамики эпохи поздней бронзы – раннего железа, in Д.А. Алимова, Ш.Р. Пидаев (eds) Столице Узбекистана Ташкенту 2200 лет. Материалы международной научной конференции, посвященной 2200-летнему юбилею города Ташкента: 33-39. Ташкент: Фан. Sverchkov and Boroffka 2013. Сверчков, Л.М. and Н. Бороффка 2013. Период Яз-II: этапы и хронология. Записки Восточного Отделения Российского Археологического Обществы (ЗВОРАО), Новая серия 2013.

Tarasov, P., G. Jin and M. Wagner 2006. Mid-Holocene environmental and human dynamics in northeastern China reconstructed from pollen and archaeological data. Palaeogeography, Palaeoclimatology, Palaeoecology 241: 284-300. Yablonskiy 2005. Яблонский, Л.Т. 2005. Археологоантропологическая гипотеза к проблеме формирования культур сакского типа, in Центральная Азия. Источники, история, культура. Материалы Международной научной конференции, посвященной 80-летию доктора исторических наук Е.А. Давидович и действительного члена Академии наук Таджикистана, академиu РАЕН, доктора исторических наук Б.А. Литвинского. Москва, 3-5 апреля 2003 г.: 776-791. Москва: Восточная Литература. Zadneprovskiy 1962. Заднепровский, Ю. А. 1962. Древнеземледельческая культура Ферганы. Материалы и исследования по археологии СССР 118. Москва, Ленинград: Издательство Академии наук СССР. Zadneprovskiy 1997. Заднепровский, Ю.А. 1997. Ошское поселение к истории Ферганы в эпоху поздней бронзы. Бишкек: Мурас.

96

Chapter 7

Shifting Exchange Patterns During the Bronze and Iron Ages Between China and The West in Eurasia Henri-Paul Francfort The most important historical phenomenon occurring from the Bronze to the Iron Age in the civilisation of Eurasia was the emergence of horse harnessing and mounted pastoral nomadism (Anthony 2007; Boyle et al. 2002; Hanks and Linduff 2009; Kuz’mina 2007; Levine et al. 2003; Menghin and Parzinger 2007). From the early Eneolithic and Bronze Age populations of agro-pastoralists and stock breeders to the settled ‘urban’ agriculturalists in the Middle Bronze Age (mid3rd millennium BC) (Gening et al. 1992; Kohl 2007; Zdanovich 2002) and the mounted pastoral nomadic tribes in the 1st millennium BC (Koryakova and Epimakhov 2007; Stark et al. 2012), the relationships between the ‘sown’ and the steppe altered and changed considerably the picture of East-West relationships and boosted the pattern of exchange between the 3rd to the 1st millennium BC. However, this vision of an East-West exchange axis is certainly an oversimplification. This chapter tries to show that we also need to consider the North-South relationships by examining the exchanges between Central Asia and the regions South of the

Hindu Kush, notably the South-Afghanistan and the Indus-Baluchistan regions (Figure 1). We consider three chronological phases in this chapter: (1) the 3rd -2nd millennia BC; (2) the 1st millennium BC and (3) the turn and the beginning of our era (c. 2nd century BC – 2nd century AD). It is not possible in this limited space to examine all the manifestations of these complex relationships. We propose to select only some aspects of the archaeological and cultural complexes to illustrate the importance of these exchange phenomena. From the 3rd to the 2nd millennium BC During this period from the 3rd to the 2nd millennium BC, the agro-pastoralist cultures of Afanasevo and Okunevo were characterized by cattle herding and the beginnings of metallurgy in a vast part of the steppe zone (Alekshin et al. 2010; Avanesova and Dzhurakulova 2008; Frachetti 2012, 2017; Fribus 1998; Gösdorf et al.

Figure 1. Diagram showing the Eurasian cultural areas and sites of the Bronze Age discussed in this chapter.

97

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 2. Sarazm (Tajikistan): Stone implements of steppic (staff) or ‘oasis’ (with handles) types. Penjikent Museum, ca. 3rd millennium BC (photo Mission Archéologique Française en Asie centrale – MAFAC).

1998; Krasnienko 2002; Kubarev et al. 1999; Leont’ev and Kapel’ko 2002; Mallory 1997; Nagler 1999; Savinov and Podol’skij 1997; Jia and Betts 2010). We shall highlight their relations with the proto-urban settlements in the southern zone, an area influenced by the Middle Eastern and Indus-Baluchistan cultures. We shall use as examples a selection of various sets of archaeological data: Sarazm excavation data; Afanasevo and Okunevo cultural sets; rock art. The excavations at Sarazm (Tajikistan, Zaravshan valley) brought to light not only the history of the proto-urban evolution far away from the great fluvial basins but also the links between cultures that would have been commonly considered as remote, separate and without contacts. But surprisingly Sarazm showed contact with various archaeological cultures and areas, and a very large network of relations and exchanges (Figure 2).

Figure 3. Sarazm (Tajikistan): Painted pottery of ‘Balochi’ type. Excavation VII, ca. mid-3rd millennium BC. TajikFrench excavations (photo Mission Archéologique Française en Asie centrale – MAFAC).

of the site. Cross-regional interconnections are also suggested by various Balochi, Mundigak (III-IV) and Iranian (grey Gorgan type) styles of potteries found at several places and within stratigraphic levels at Sarazm (Figure 3) (Lyonnet 1996). The pottery is undergoing archaeometric analysis to enable us1 to differentiate between imported and locally made vases. The imported potteries seem to appear only in small quantities compared to the locally made ceramics ‘imitating’ exogenous shapes and decorations.

The Sarazm site is located in the upper valley of the river Zaravshan, 40 km upstream from Samarkand, in the Republic of Tadjikistan, near the city of Penjikent. The excavations, in cooperation with our team of the French Archaeological Mission in Central Asia since 1984, exposed the architectural and material remains of a proto-urban settlement connected with the world on the probable basis of minerals trade involving stone and metals (Besenval 1987, 1988; Isakov 1986, 1991, 1996; Isakov and Lyonnet 1988; Mutin et al. 2017; Razzokov 2008; Razzokov et al. 2006). Sarazm was included in the UNESCO list of world heritage monuments in 2010.

All these artefacts belong to normal proto-urban assemblages and point towards the Middle East (Iranian Plateau) and southwards to the Indus-Baluchistan region as sites of origin. This last southern connection is especially important with Mundigak (Mutin and Razzokov 2014) and the Indus-Baluchistan region since not only Sarazm but also the remains of Taloqan in

The complexity of the connections (Afanasevo steppe cum proto-urban oasis) are well exemplified by the burial No. 5 in excavation No. IV: a young lady was buried with precious goods, jewels etc. notably copper, silver, gold beads and beads of lapis (from Afghanistan) and a shell bangle (from the Indian Ocean). This provides evidence of wide-ranging exchange networks. The same remark can be made a propos a gold rosette and a unique stone cylinder seal of Proto-Elamite style from another part

1  The project is part of the Franco-German ANR-DFG programme ROXIANA directed by N. Boroffka (from DAI, the German Archaeological Institute) and H. P. Francfort (from CNRS, the French National Centre for Scientific Research).

98

H.-P. Francfort: Shifting Exchange Patterns During the Bronze and Iron Ages Takhar province of Eastern Afghanistan (Lyonnet 1977, 1981, 1988) display Baluchi type pottery. At a later period (mid-3rd millennium) Shortughaï (NE Afghanistan) is clearly a settlement outpost of the Harappan civilization in Central Asia (Francfort 1984a, 1985a, b, 1988b; Francfort et al. 1989). Towards the north of Baluchistan, in Pakistan, at Mehrgarh and Sibri, materials that definitely belong to the Oxus Civilization have been found (BMAC) (Jarrige 1988, 2017; Santoni 1984, 1988). At this time, the Indus Civilization was in contact with the Oxus Civilization which was also connected to the Middle and Near Eastern empires of Akkad, Ur III and Isin-Larsa to the West (Amiet 1986, 2004; Francfort 2016; Francfort and Tremblay 2010; Sarianidi 2005, 2007; Sarianidi and Dubova 2011). In short, we are sure that not only did a West-East connection function during the proto-historic period, but also North-South links were active, in a complex network of relations between agrarian cultures. In this respect, along with numerous artefacts, the introduction into Central Asia of the zebu and water buffalo and perhaps of some varieties of plants from India during the Harappan period is highly significant.2 Similarly, the introduction of millet in the late periods of Shortughaï, as well as some typical stone tools, indicate relations with the eastern regions, namely China, perhaps Xinjiang (Debaine-Francfort 1988). But this is not all.

stones, metals and gold can be considered (Mallory 1991). But Sarazm, as prestigious as it is, is just a part of a bigger picture. Recent discoveries at Farkhor in Southern Tajikistan, precisely opposite Shortughai on the right bank of the Panj river (Bobomullaev et al. 2015; Vinogradova and Bobomullaev 2016), allows us to consider the attractiveness of the ‘Oxus gold’ of this region already in the Bronze Age (Francfort 2016), possibly at some moment in connection with the Ur ‘royal tombs’ (Jansen et al. 2016). Moreover, if we bring into this scenario some rock art representations engraved in Siberia, Kazakhstan, the Himalayas and the Upper Indus valley, we see a consistent and continuous pattern of exchange across a wide space of steppes and mountains. One marker of such relations is the widespread Okunevo design known for its image of the human face or mask, the so called mascoid pattern, seen from the Yenissei valley to Upper Indus, to Zanskar and Ladakh. Identified long ago by Jettmar they can be traced to the Znamenka stele in the Minusinsk basin, to the Yenissei canyon boulders and to Indus and Ladakh rocks (Bruneau 2012; Bruneau and Bellezza 2013; Francfort et al. 1990, 1992; Jettmar 1982; Jettmar and Thewalt 1987). In passing we note the problem of shamanist interpretations of this sort of image: the horned masks or heads are sometimes supposed to represent shamanic ceremonial masks and thus to indicate shamanic religion (Devlet 1997; Rozwadowski 2004). But this unproven theory does not answer the question of how and why these designs spread so widely: were migration or exchange the driving force? In any case, the mascoids could be better interpreted as headgear worn by hunters. This does not exclude their ritual function, but such a purpose is not necessarily connected to any practice of trance or shamanism (Francfort et al. (eds) 2001; Lorblanchet et al. 2006). Moreover, the engravings of mascoids do not necessarily indicate a firm ethnic or linguistic marker. Instead, are we not perhaps facing a shared fuzzy set of technical and ideological functions related to the imaginary world of the hunters and herders of the steppe?

There is another remark that should be made about the Sarazm burial that we describe here. The lady’s (‘princess’) tomb is located with four others inside a large stone circle recalling only the burials of the Afanasevo steppe culture, and no other. The Afanasevo culture is known in South Siberia, Altai and Kazakhstan and more recently Uzbekistan (Avanesova and Dzhurakulova 2008). Connected with the same steppe Siberian culture, some grooved big stones without recognizable functional use have been found. These are unusual in the proto-urban context and recall the Afanasevo culture (Boroffka and Sava 1998). This ancient eastern steppe connection in the Zaravshan valley has been recently confirmed by the pottery and other material from the site of Zhukov, in Uzbekistan, lower down the river. It is unanimously considered to be a pure Afanasevo settlement (Avanesova 2013; Avanesova and Dzhurakulova 2008). In the Sarazm assemblage, some Kel’teminar types of pottery exhibit links with the Kyzyl Kum desert and the lower reaches of the Zaravshan River (Lyonnet 1996).

Another example of rock images relevant to our argument is the bullock cart represented in profile. The profile view is the ancient Near Eastern manner of drawing a cart (Littauer and Crouwel 1979) and this style was spread as far as the Yenissei (Afanasevo/ Okunevo) and Bronze Age Bactria. Shall we, using such evidence, trace a migration from the Near East to Siberia or even from South Russia (Yamnaya culture) to Siberia (Afanasevo)? We don’t think so because there are many reasons for the spread of images, goods, peoples and ideas, but the chronologies are not yet sufficiently precise today to support such eastwards migration theories of Indo-European peoples and languages (the literature is enormous but we cannot

What are the possible explanations for such exogenous materials at Sarazm? Migrations (the traditional Soviet historical explanation) or exchanges and relations based on the temporary residence of peoples from various origins for specific purposes such as collecting See J. Desse and G. Willcox contributions to the Shortughaï volume (Francfort et al. 1989).

2 

99

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads tackle this question in this article). Later, during the 2nd millennium, the steppe is peopled by the Andronovo culture with its many variants. In this time, the chariot representation changed; the view is no longer in profile, but it is now as if seen from above, in a sort of ‘split representation’. This new way of depicting chariots dominates the entire steppe world, except in Ferghana (Chust culture) where the profile image remains dominant. Here agriculture is practiced on a large scale, and there are many parallel designs that are reminiscent of the Eneolithic and Bronze Age pottery paintings of the proto-urban type (Novozhenov 2012). The petroglyphs of China and the pictogram ‘chariot’ of the Shang period belong to the same group of representations as the Andronovo chariot images (Francfort 1992, 2002). When thinking about profile and split chariot depictions, what sort of relationships are we looking at, between the northern steppes and the proto-urban oasis cultures?

from various regions and cultural origins could have settled in the same neighbourhood possibly at the same time for specific reasons. We saw these situations with Sarazm and Zhukov, Shortughaï and Farkhor, Nausharo and Sibri. During the third and second millennia, a huge network of relationships connected regions and settlements north of the Hindu Kush-Himalaya crest line, but also the regions north and south of this notional line. Distance, rivers and mountains never constituted obstacles or ‘natural’ boundaries. At the same time, we will probably never obtain the kind of reliable statistical information about international exchange that we are accustomed to in the modern era. For the proto-history of Central-South-Eastern Asia we shall have to remain satisfied with understanding the trends and evolution of exchange by using the available material evidence and constructing holistic but reasonable models about how production, consumption and technological factors supported inter-regional trade.

Summary

Steppe riders and empires: the early nomadic cultures between the Achaemenids and the Eastern Zhou-Warring States (1st millennium BC)

Focusing systematically on mainly human migration as an explanation for the parallel designs that appear in widely dispersed archaeological sites leads, we think, to assuming too many migrations from too many places to too many other places. We prefer to think of this evidence as reflecting the existence of widespread exchange networks, but await more evidence to substantiate the argument that some migrations were more important. Eventually such data may need to come from other disciplines. In a similar manner, the appearance of cattle, sheep, goat, wheat, barley, and millet have to be interpreted carefully. Animal husbandry and the cultivation of crops comes from somewhere and moves, ‘travels’, to be adopted or implemented in some other place, but this does not prove that groups of people or tribes migrated long distances. Peoples, technologies, ideas and ways of looking at the world all move or ‘travel’. Eurasia is an open system. Thinking about these movements as networks, and using graphs or hypergraphs to analyze the data, provides us with an overview of disparate data to see what patterns emerge. This is a better approach than using vague descriptors such as ‘relations’, ‘exchanges’ or ‘clouds’. Having said that, we acknowledge that it is still too early to produce holistic and quantified models for reconstructing the ‘silk roads’ (which were earlier really ‘lapis-lazuli roads’) on the basis of small quantities of data equivalent to one donkey load of traded goods. Such a limited analysis is not a serious exercise. On the other hand, using the unique Afghan Badakhshan deposit and the many discoveries in situ from Mesopotamia to the Levant provides us with the possibility of tracing a network of many different and nuanced itineraries for lapis. It is now also possible to extend the roads and itineraries to tin, gold, perhaps chlorite. A new and important conclusion is that peoples

In this second part of this chapter we consider relations between examples of the widely dispersed images of chariots and composite beings or monsters and how these images share some stylistic elements. In the previous section we saw that the split chariot image dominated the steppe and China (the Shang dynasty) and this is manifest on some Mashan silks. However, the Chinese images then reverted to profile representation, at Mashan again for instance, so that with the Zhou pictorial bronzes we see mixed images of chariots in both the split and profile style. Apparently, this change occurred under Achaemenid influence. A close inspection of the scenes depicting hunter bowmen on chariots shows that these obviously have been inspired by images of the Persian royal hunt: the composition with charioteer, the king archer and the feline image are similar to the well-known ‘Darius cylinder’. After this change, the profile view continued into the Han and subsequent periods (Thote 1999; Francfort 2002). The story of composite beings, monsters, in the art of the steppe is also very revealing. Until the 4th-3rd centuries BC, no Near Eastern influence was visible in the steppe art as shown by the discoveries in the kurgans (burial mounds) of Arjan 1 and Arjan 2 in Tuva (Chugunov et al. 2006; Chugunov et al. 2010; Francfort 2009; Grjaznov 1984). There we find images of animals or parts of animals, in the recognizable steppe style, but no composite animals or monsters. And the same appears also to apply to the early animal representations of Chilikty-2 in Eastern Kazakhstan (Figure 4). That means that the griffin was not originally present in the steppe in any form. The horned lion, a specific form 100

H.-P. Francfort: Shifting Exchange Patterns During the Bronze and Iron Ages

Figure 4. Diagram showing some of the Eurasian sites of the Iron Age discussed in this chapter.

of ‘griffin’, comes directly from Persia and the satrapies of Central Asia, from the courtly art and silverware of Persepolis, for instance. We can follow the griffin image up to the Altai, to Siberia and to Xinjiang (Figures 5, 6). The horned lion, palmette (decorative palm leaf), eagle head and other Achaemenid motifs appear in the Altai

(Berel 11; Ukok: Polos’mak 2001; Stark et al. (eds) 2012) but also from Persia to Ningxia (Debaine-Francfort 2016; Francfort 2007; 2009; Francfort et al. 1998, 2000). Achaemenid arts in the steppe, in the elite burials, whether imported or imitated, are an important chapter in the relationships we are seeking to trace. The south Ural areas in Russia and Kazakhstan are very important. At Filippovka the Persian influence on style is present in many artefacts: amphora, bracelet, pendant, torque, rhyton, goblet, throne foot, etc. (Treister and

Figure 5. Berel (Altay, Kazakhstan): Kurgan No. 11 horse tack ornament in gilded wood in the shape of griffin and mountain sheep with curling mane of Achaemenid inspiration, first quarter of 3rd century BC. Kazakh-FrenchItalian excavation (photo Mission Archéologique Française en Asie centrale – MAFAC).

Figure 6. Berel (Altay, Kazakhstan): Kurgan No. 11 horse wooden ornament in the shape of Persepolis griffin, of Achaemenid inspiration, first quarter of 3rd century BC. Kazakh-French-Italian excavation (photo Mission Archéologique Française en Asie centrale – MAFAC).

101

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Jablonskij 2012; Jablonskij 2014; Kurinskikh et al. 2013). The kurgan Issyk (Semirech’ye, Kazakhstan) displays Achaemenid imports (Akishev 1978). At Lebedevka, Achaemenid and Graeco-Roman imports can be seen in the rhyton-jug. At Isakovka, Hellenistic silverware is inscribed (in Khorasmian and Parthian), and an Achaemenid phiale is reused as a phalera on horse tack, while a four lobed dagger sheath in Scytho-Sarmatian style has also been discovered. At Sidorovka a curling feline image, a typical steppe theme, is represented in another style. At Prokhorovka, a 4th century bowl from the Black Sea has been discovered and Achaemenid godroned omphalos phialae are reused as phalerae. At Volodarka, a phalera depicts Bellerophon fighting Chimera, a composition very close to the Olynthus mosaics in 4th century BC Macedonia (Balakhvancev and Jablonskij 2007; Balakhvantsev and Yablonskij 2009a, b; Gutsalov 2012; Koryakova 2002, 2007; Koryakova and Epimakhov 2007; Livshits 2003; Mochkova 1994; Treister 1997, 2004, 2009a, b, 2010, 2012).

of Heidelberg are replete with evidence of Upper Indus rock art images related to Central Asian steppe and Achaemenid art (Bandini-König 2003, 2005, 2007; Francfort 2013b, 2015). The absence of rocks, and therefore of rock art, in the sandy Taklamakan desert of Xinjiang (north of the Himalayas and Tibet) is not mysterious but more importantly is compensated by discoveries made in the cemeteries and towns of the Keriya valley (Djumboulaq-Qum), in the Turfan basin (Zhaghunluk, Wupu, Yanghai) and in the piedmonts of Tianshan and Kunlun (Debaine-Francfort and Abduressul (eds) 2001; Francfort 1998, 2001). In these parts of Xinjiang, we can find artistic motifs and styles similar to the steppe art engraved on wood artefacts and woven into textiles. Imports from India have also been discovered there, for example, etched carnelian beads. Summary During the Early Iron Age, the steppe network and its connections with the ‘upper satrapies’ of Central Asia starts to borrow artefacts and art forms from the Achaemenid Empire -- this occurs even outside the borders of the Empire and continues after the fall of the empire until the mid-third century kurgans of Ak-Alakha, Berel 11 and Pazyryk 5, for instance. Eventually these designs and artefacts were transmitted to China. These borrowings could only have occurred because the recipients desired to change the form or the style of a pre-existing image of an already extant representation (e.g. a monster) or material element (phalera). These composite beings or monsters, influenced by Achaemenid tastes, fulfilled the desire for the more elaborated shapes of a convincing ‘monster’, a role previously played by wolves or felines. This sort of transfer was not imposed by the culture in which these ideas and styles originally arose; rather, it is our opinion, that the adoption of such designs reflects a deliberate choice made by the recipient culture exercising a conscious selection process. Motivated by psychological and economic desires, the acts of giving gifts, taking booty, or exchanging goods and ideas probably played a role, in addition to the tendencies amongst elites to imitate the styles and tastes of foreign satrapies and royal courts with which they were in contact. The Greek successors of the Achaemenids inherited these exchange networks, both in the steppe and in the agrarian world, and populated distant regions with their own tastes for Hellenized artefacts and shapes, parallel to the ongoing and age-old practice of the movement of people, goods and ideas. This was the prelude, in the steppe, to the Silk Road. In addition to the earlier material links, a number of new roads also connected Central Asia with the southern parts of the Indus-Baluchistan region, a link supported by the evidence in rock art styles and archaeological finds.

Towards the southern areas, it is remarkable that, as in the previous period, the rock engravings of Tibet (Rutog), Ladakh-Zanskar and the Upper Indus areas provide a rich corpus of images related to steppe art. Notably, these Iron Age petroglyphs are not unique to a single style or period but belong to the whole millennium from the early 1st millennium to what is conventionally called Antiquity. We can recognize the Arzhan-Maiemir style but also later styles (Francfort et al. 1990, 1992). The same applies to Xinjiang where, without rock surfaces, this style appears engraved on wood or worked into textiles (Debaine-Francfort 2016; Francfort 1998). Similarly, it is possible to identify artefacts belonging to the steppe art corpus that were discovered long ago (the Pattan find related to early times) (Chugunov et al. 2010) or more recently. On the other hand, Early Iron Age pottery similar to the Yaz-I complex of Central Asia has been discovered and excavated in the Peshawar area at Akra (Magee et al. 2005). This demonstrates that the southern extension of the network, therefore, predates the creation of the Indian and Gandharan satrapies by the Achaemenid administration. The routes connecting the Central Asian regions with the Indus-Baluchistan area have been more or less well known since the Achaemenid period, when the satrapies of Areia, Margiana and Bactria were linked by roads and mountain passes to Arachosia and Drangiana but also with the Gandharan and Indus provinces (Briant 1996, 2012; Vogelsang 1992). However, no sources apart from archaeological finds, especially petroglyphs, can be found on the Karakorum-Himalaya roads and passes (Holdich 1910; Zelinskij 1964a, b). The remarkable series of beautiful volumes edited and published by H. Hauptmann et al., D. Bandini-König and collaborators at the Academy 102

H.-P. Francfort: Shifting Exchange Patterns During the Bronze and Iron Ages Graeco-Bactrian, Saka-Yuezhi, Xiongnu and Han: the connection between the great steppe-empires from Bactria to Mongolia at the time of the Silk Road (c. 2nd century BC to 2nd century AD)

by a Greek artist or a local artist trained in the Greek style. However, if we consider this piece along with the Achaemenid objects scattered along the steppe road and with the Achaemenid and Greek finds from the frozen Altai Kurgans that we discussed above, we can appreciate the intensity of the long-distance exchanges that took place in the world of the nomadic elites, from the Mediterranean to China.

After the conquest of Central Asia by Alexander the Great, an important Seleucid regime controlled the ‘upper satrapies’ where Antiochos the First (son of Seleukos the First, general of Alexander the Great, and the Bactrian princess Apamè) was installed as the Viceroy by his father. In c. 250 BC independence was proclaimed by the satrap Diodotos, the first of a succession of Graeco-Bactrian sovereigns lasting until c. 130 BC. Then, in brief, the Saka-Yuezhi nomadic tribes, coming from the steppes, captured control of Bactria, mixing with the preexisting populations and cultures. This was the prelude to the building of the Kushan Empire, the great contemporary to the powers of China, Parthia and Rome. During these periods, the east-west and north-south communication networks were never seriously curtailed. In this part of the chapter my illustrations are from the post GraecoBactrian period. For the Seleucid and Graeco-Bactrian period in Central Asia the bibliography is enormous, a great part of it being on and around the site of Aï Khanum in northern Afghanistan that was excavated under the direction of Paul Bernard from 1964 to 1978 (Figure 7).

Later, during the Saka-Yuezhi period in Bactria, the case of the Tillya Tepe burial provides an exemplary indication of the nature of these exchanges (Boardman 2003; Francfort 2011b, c; Pfrommer 1996; Rubinson 2008; Sarianidi 1985; Sarianidi and Koshelenko 1982). The burial, located in western Bactria, NW Afghanistan, near Shiberghan, belongs to one interment: a single male warrior’s tomb is surrounded by five or six female tombs reliably dated to the mid-1st century AD. At this time in the 1st century, the coins and artefacts of Bactria take their iconographic inspirations from a wide range of cultures: Roman, Greek, Graeco-Bactrian, Parthian, Saka, Yuezhi and Xiongnu. The tombs coincide with the transformation of the exchange routes from the steppe roads to the Silk Roads. The four lobed dagger sheath, for example, spread among Saka, Parthian and Sarmatian weaponry from the Altai to Nimrud Dag, and from Palmyra to Noin-Ula (Francfort 2011b; Schiltz 2002). The reverted hindquarters seen on images of animals, on the other hand, is a good example of a stylistic trait that we can follow from Tillya Tepe to Mawangdui in China (Pirazzoli-t’Serstevens 2007; Debaine-Francfort 2016), when its first appearance occurred in the steppe world, at Arjan 1.

A Greek helmeted hoplite in bronze, about 50 cm high, found in a kurgan burial in Xinjiang, dates from the late 4th century BC (Francfort 1988a: 182, Fig. 8). We don’t know how this piece came to northern Xinjiang or whether it had been manufactured there

Figure 7. Diagram showing main branches of the Silk Roads and Central Asian archaeological sites discussed in this chapter.

103

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 8. Gandharan toilet tray with royal wine filtering and drinking ceremony seen frontally (photo Kurita, courtesy Mr and Mrs Hirose, and Ms J. Pons).

The paintings and reliefs of the Khalchayan monument in southern Uzbekistan belong to the same period, possibly the reign of Heraios, one of the first Kushan rulers (Abdullaev 2007; Bernar and Abdullaev 1997; Bernard 1987; Pugachenkova 1966, 1971). The Shampula textiles found in Xinjiang near Khotan (Bunker 2001; Keller 2001; Museum and Archaeology 2001; Wang and Xiao 2001) display interesting similarities with designs found in the site of Khalchayan: wing shapes, flower shapes, warrior postures, dress, spear and ribbon in the style of the Yuezhi. They all clearly belong to the universe of the Yuezhi (pre- and early Kushan) kings and princes. Our observation of images from Sampula and Khalchayan also suggests to us that the Miran wall paintings (Nikonorov 2000; Reisinger 2010) discovered long ago by Stein in Xinjiang date from an earlier date (1st or early 2nd c. AD) than generally thought (3rd c. AD) (Francfort 2014b).

imports are spread along the steppes, as are Chinese mirrors and textiles. Such artefacts were fashionable among well-to-do ladies, and it is possible to trace a Gandharan extension, thanks to the images engraved on the so-called ‘toilet trays’ (Figure 8). A good example is the embroidery recently found in NoinUla (overview: Polos’mak 2010, 2011, 2012). These items were certainly woven in Bactria as noticed long ago by many researchers, for their design is similar to tapestries from other kurgans at the same site. There sits the king on a folding chair (sella curulis3) drinking wine filtered with a special device (tripod and hanging strainer) and collected in a crater of typical GraecoBactrian and Yuezhi Hellenic form (Figure 9). The pictorial composition is reminiscent of some Gandharan toilet trays and recalls the designs of Graeco-Buddhist reliefs and Sassanian silverware (Francfort 2014a). The same can be said of the dragons that appear in the ornamental frieze (Francfort 1979, 2012b).

During the Kushan period, but farther east, in Noin-Ula (Noyon uul), Mongolia, the remarkable elite Xiongnu tombs have revealed textiles that are linked to the pictorial traditions of the Yuezhi: the decorative faces closely resemble the Khalchayan ‘portraits’ while the local ornaments have integrated Graeco-Roman design elements. These artefacts were most probably manufactured in Bactria (Francfort 2011c; Polos’mak et al. 2011; Polos’mak et al. 2008; Rudenko 1969; Yatsenko 2012). More information has recently been published about Noin-Ula based on these recent discoveries made by the Mongol-Russian expedition of textiles displaying battle scenes, rituals and courtly compositions (Polos’mak 2015; Polos’mak et al. 2011).

One of the most striking examples of the stylistic similarities between Gandharan and Bactria-Sogdiana designs around the turn of the era can be seen in a comparison between the battle engraved on the bone Orlat plates of the Samarkand region and the coins of the Indo-Scythians Azes I and Azes II and Azilizes from south of the Hindu Kush (Bernar and Abdullaev 1997; Ilyasov 2003; Ilyasov and Rusanov 1997/98; Litvinsky 2001). The image of the king with armour, spear or whip and standard are the same, including some details of the horse’s tack, specifically the bridles and phalerae. However, in the Orlat burial No. 2, a long sword with 3  This motif also appears on some Kujula Kadphises coins imitated from Augustus, and actually found in sites at Tillya Tepe and at Sirkap in Taxila.

Even in this Mongolian site, the network of exchange and relations extended southwards. Graeco-Roman 104

H.-P. Francfort: Shifting Exchange Patterns During the Bronze and Iron Ages

Figure 9. Noin-Ula (Mongolia) tapestry with royal wine filtering and drinking ceremony seen in profile. Note the dragons in the ornamental frieze, first century AD. MongolRussian excavation (from N. Polos’mak, Nauka, 2 (38), 2011: 116-117).

scabbard slide in nephrite points towards the eastern steppes and China as does the dog buried in the dromos of the tomb. The other side of the Orlat plaque, with a very dynamic representation of a hunt, can easily be compared with the Takht-i Sangin ivory plate with hunting scene, and with a famous gold plaque from the Siberian collection of the Hermitage Museum depicting a boar hunt (Francfort 2014a).

Bactrian kingdom followed the Seleucid example and expanded the network of inter-regional relations to an unprecedented scale from Bactria to Mongolia and China. The mechanisms are diversified but we can compare for instance the taste of the Xiongnu elites for Graeco-Bactrian luxury goods to the symmetrical desire from the Saka-Yuezhi elites for Chinese lacquers and mirrors (Francfort 2011a, 2014a; Rapin 2007; Rapin et al. 2001). This consumer demand and its supporting network amplified the exchanges from India (Gandhara) to the Mediterranean, under the Kushans, the Parthians and the Romans, functioning in both directions.4 The elites shared a number of regalia and paraphernalia, as

Summary After Alexander’s conquest and the establishment of the Graeco-Oriental monarchies a great wave of Hellenic style and design developed and dispersed beyond its original home territories. After the establishment of the breakaway Seleucid regime, the Graeco-

See for instance the coinage of Kujuka Kadphises the first great Kushan king and the use and representations of the Roman sella curulis in the steppe world (Rtveladze 2012).

4 

105

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads well as some specific military techniques such as the use of cataphractarii and mounted archers. This period represented a tremendous extension of the previous networks, by whatever possible mechanisms. Exactly how some of this happened may remain unknown, for example the movement of Bactrian-made Saka-Yuezhi textiles to Khotanian Sampula in Xinjiang or to Xiongnu Noin-Ula (Noyon uul) in Mongolia. The Chinese chronicles mention the defeat of the Yuezhi by the Xiongnu at an earlier date but what the archaeological evidence presented in this chapter shows is the large scale complex political, military and cultural history that has not been narrated in any chronicle.

system, the city fortification and the land-tenure structure of the agrarian empires, of the Achaemenids or the Greeks, was much slower, more labor-intensive and perhaps to some degree fragile. Therefore, the network we consider here can be taken as some sort of complementary and rival binary system, best described by the historians studying the relationships of the Chinese empires with the steppe tribes, confederations, states and empires from the Chinese and Classical sources to Grousset, Lattimore and Di Cosmo (Di Cosmo 2002). The proof of it is given by the systematic acculturation and complementary economies of the nomads: when they reach and conquer Bactria or the oasis world in general, they become Bactrians or generally settlers; and the same process was reiterated by the Huns, the Hephtalites, the Turks, the Mongols, and the Timurids as it was done before by the Iranians, Medes and perhaps the ‘Andronovians’ if not the ‘Afanasevoids’. This kind of replication in history is not completely new or original, but it is clearly a part of the mechanism of Eurasian history, subject to attempts of comprehensive overviews, explanations and theorising (for instance: Aruz 2003; Beckwith 2009; Cunliffe 2015; Grousset 1970; Hansen 2017; Turchin 2009). The network we sketchily described in this chapter had not just chronological and cultural dimensions, but also spatial ones; it was not only the East-West popularised ‘Great Khorasan Road’ or ‘Silk Road’, but also the interconnected North-South ‘Vieille Route de l’Inde’ (Foucher 1942-1947; Neelis 2006; Rtveladze 2012). The ‘Vieille Route de l’Inde’ was not only used at least twice in history for momentous occurrences such as the Aryan invasion and the Saka coming via the ‘Hanging Passes’ in the 2nd century BC, as we would believe if we only rely upon texts, but it was a permanent link with the steppe world via either Bactria or the Pamir-Karakorum-Western Himalayas. In this respect, there is some balance or some alternate domination, or some complementarity, between the steppe world’s populations and the agrarian empires, and this begins in the Bronze Age.

Conclusion For many logical and historical reasons, we cannot assert that the network of Eurasian exchanges was broader under the Roman dynasty of the Julio-Claudians than during the time of the Afanasevo culture, in terms of extension and even in terms of intensity. What then on earth changed during all the centuries I have summarily reflected on in this chapter? First the technologies became more diversified and the social structures and religions more elaborate and subjected to control within larger socio-political units. Second, the speed and the quantity of all the agents and items involved in the inter-regional exchange of goods and ideas increased greatly. This became obvious when I considered the diversity of goods that were transformed, notably the metals, but the techniques of production improved too, allowing reproduction in great numbers of elite goods. In my earlier work I noted the changing techniques used for speeding up production of stone structures and clay and plaster castings for the reproduction of artefacts (Francfort, 1984b; 2012a). Irrigation techniques for improving the quality and quantity of the crops have also been recorded (Francfort and Lecomte, 2002). Similarly, as we have noticed above, horsemanship was the main transformation during these centuries in Eurasia, permitting speed and multiplication: speed of movement in long distance treks and raids, and speed in delivering showers of arrows by noblemen or laymen riding their horses. This contrasts greatly with the chariot elites of the Bronze Age whose advantage was only to arrive at the fighting place not too exhausted by long runs on foot (see, for example, accounts by Homer). Moreover, the nomadic populations could maintain a great number of people in mountainous and semi-desert areas, and for them there was no need to go back to a permanent home for ploughing, irrigating, or harvesting; the food was always mobile and thus at hand. It only slightly diminished in spring during the birthing and milking season and complementary diets were available from neighbouring agriculturalists (Francfort 2013a). On the other side of the socio-economic divide, the sophisticated administered postal angaron and fodder

References Abdullaev, K. 2007. Nomad Migration in Central Asia, in J. Cribb and G. Herrmann (eds) After Alexander. Central Asia before Islam: 73-98. Oxford: Oxford University Press. Akishev, K.A. 1978. Kurgan Issyk. Moscow: Iskusstvo. Alekshin, V. et al. (eds) 2010. Drevnie kul’tury Evrazii. Materialy mezhdunarodnoj nauchnoj konferencii, posvjashchennoj 100-letiju so dnja rozhdenija A. N. Bernshtama. St Petersburg: Rossijskaja Akademija Nauk, Institut Istorii Material’noj Kul’tury. Amiet, P. 1986. L’âge des échanges inter-iraniens. 35001700 avant J.-C.. Paris: Editions de la réunion des Musées Nationaux. 106

H.-P. Francfort: Shifting Exchange Patterns During the Bronze and Iron Ages Amiet, P. 2004. L’élargissement des échanges interiraniens. Essai de mise à jour. Akkadica 125: 5-25. Anthony, D. 2007. The Horse, the Wheel and Language. How Bronze Age Riders from the Eurasian Steppes Shaped the Modern World. Princeton, Oxford: Princeton University Press. Aruz, J. (ed.) 2003. Art of the First cities. The Third Millennium B.C. from the Mediterranean to the Indus. New York: The Metropolitan Museum of Art. Avanesova, N. 2013. Zhukov, un ‘sanctuaire’ énéolithique d’anciens nomades dans la vallée du Zaravshan (Ouzbékistan). Paléorient: 85-108. Avanesova, N. and D. Dzhurakulova 2008. Drevnejshie nomady Zaravshana, in Sh. Pidaev (ed.) Cultures of Nomadic peoples of Central Asia: 13-33. Samarkand: International Institute for Central Asian Studies Samarkand. Balakhvancev, A. and L. Jablonskij 2007. Akhemenidskaja emal’ iz Filippovki (problema khronologii pamjatnika). Rossiiskaya arkheologiya 1: 143-149. Balakhvantsev, A. and L. Yablonskij 2009a. Once Again on the Question of the Dating of Inscriptions from Prokhorovka. Ancient Civilizations 15: 137-165. Balakhvantsev, A. and L. Yablonskij 2009b. A Silver Bowl from the New Excavations of the Early Sarmatian Burial-Ground Near the Village of Prokhorovka. Ancient Civilizations 15: 167-182. Bandini-König, D. 2003. Die Felsbildstation Thalpan. I Kataloge Chilas-Brücke und Thalpan (Steine 1-30). Materialen zur Archäologie der Nordgebiete Pakistans 6. Mayence: Verlag Philipp von Zabern. Bandini-König, D. 2005: Die Felsbildstation Thalpan. II Katalog Thalpan (Steine 31-195). Materialen zur Archäologie der Nordgebiete Pakistans 7. Mayence: Verlag Philipp von Zabern. Bandini-König, D. 2007. Die Felsbildstation Thalpan. III Katalog Thalpan (Steine 196-450). Materialen zur Archäologie der Nordgebiete Pakistans 8. Mayence: Verlag Philipp von Zabern. Beckwith, C. 2009. Empires of the Silk Road. A History of Central Eurasia from the Bronze Age to the Present. Princeton/Oxford: Princeton University Press. Bernar, P. and K. Abdullaev 1997. Nomady na granice Baktrii (k voprosu etnicheskoj i kul’turnoj identifikacii. Rossiiskaya arkheologiya 1: 68-86. Bernard, P. 1987. Les nomades conquérants de l’empire gréco-bactrien. Réflexions sur leur identité ethnique and culturelle. Comptes-rendus de l’Académie des Inscriptions and Belles-Lettres 131/4: 758-768. Besenval, R. 1987. Découvertes récentes à Sarazm (R.S.S. du Tadjikistan): attestation des relations au IIIe millénaire entre l’Asie centrale, l’Iran du Nord-Est and le Baluchistan. Comptes-rendus de l’Académie des Inscriptions et Belles-Lettres 131/2: 441-456. Besenval, R. 1988. L’étain dans l’Asie centrale protohistorique: une source possible pour les métallurgies moyen-orientales. Examen de l’hypothèse, in J.-C. Gardin (ed.) L’Asie centrale and

ses rapports avec les civilisations orientales des origines à l’âge du fer. actes du colloque franco-soviétique, Paris, 19-26 Novembre 1985: 229-235. Mémoires de la M.A.F.A.C. I. Paris: Diffusion de Boccard. Boardman, J. 2003. Three Monsters at Tillya Tepe. Ancient Civilizations from Scythia to Siberia 9/1-2: 133146. Bobomullaev, S. et al. 2015. Resul’taty issledovanij mogil’nika Farkhor – Pamjatnika epokhi srednej bronzy na juge Tadzhikistana, vesnoj 2014 goda. Muarrikh Istorik 4: 73-98. Boroffka, N. and E. Sava 1998. Zu den steinernen ‘Zeptern/Stössel-Zeptern’, ‘Miniatursäulen’ und ‘Phalli’ der Bronzezeit Eurasiens. AMIT 30: 17-111. Boyle, K. et al. (eds) 2002. Ancient interactions: east and west in Eurasia. McDonald Institute Monographs. Cambridge: McDonald Institute for Archaeological Research. Briant, P. 1996. Histoire de l’empire perse de Cyrus à Alexandre. Paris: Fayard. Briant, P. 2012. From the Indus to the Mediterranean: The Administrative Organization and Logistics of the Great Roads of the Persian Empire, in S. Alcock et al. (eds) Highways,Byways, and Road Systems in the Pre-Modern World: 185-201. New York: John Wiley & Sons, Inc. Bruneau, L. 2012. Étude thématique et stylistique des pétroglyphes du Ladakh (Jammu et Cachemire, Inde). Eurasia Antiqua 18: 67-87. Bruneau, L. and J. Bellezza 2013. The Rock Art of Upper Tibet and Ladakh — Inner Asian cultural adaptation, regional differentiation and the Western Tibetan Plateau Style. Revue des Études Tibétaines 28: 5-161. Bunker, E. 2001. The Cemetery at Shanpula, Xinjiang. Simple Burials, Complex Textiles, in D. Keller and R. Schorta (eds) Fabulous Creatures from the Desert Sands. Central Asian Woolen Textiles from the Second Century BC to the Second Century AD: 15-45. Riggisberger Berichte 10. Riggisberg: Abegg-Stiftung. Chugunov, K. et al. 2006. Der Goldschatz von Arzhan. Ein Fürstengrab der Skythenzeit in der südsibirischen Steppe. Munich: Schirmer/Mosel. Chugunov, K. et al. 2010. Der skythenzeitliche Fürstenkurgan Arzan 2 in Tuva. Archäologie in Eurasien Band 26, Steppenvölker Eurasiens Band 3. Mayence: Verlag Philipp von Zabern. Cunliffe, B. 2015. By Steppe, Desert, and Ocean. The Birth of Eurasia. Oxford: Oxford University Press. Debaine-Francfort, C. 1988. Etude comparative de matériels lithiques protohistoriques chinois (Chine métropolitaine and Asie centrale), in J.-C. Gardin (ed.) L’Asie centrale and ses rapports avec les civilisations orientales des origines à l’Age du Fer. Actes du colloque franco-soviétique, Paris 19-26 Novembre 1985: 197-206. Mém. de la Mission Archéologique Française en Asie Centrale I. Paris: Diffusion de Boccard. Debaine-Francfort, C. 2016. D’un monde à l’autre. Hybridations et randonnées célestes dans le monde 107

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads chinois, de l’âge du bronze au début de notre ère, in P. Linant de Bellefonds et al. (eds) L’homme-animal dans les arts visuels Image et créatures hybrides dans le temps et dans l’espace: 238-251. Paris: Les Belles Lettres/ Presses Universitaires de Paris Nanterre. Debaine-Francfort, C. and I. Abduressul (eds) 2001. Keriya, mémoires d’un fleuve. Archéologie and civilisation des oasis du Taklamakan. Paris: Editions Findakly. Devlet, M. 1997. Okunevskie antropomorfnye lichiny v rjadu naskal’nykh izobrazhenij Severnoj i Central’noj Azii, in D. G. Savinov and M. L. Podol’skij (eds), Okunevskij Sbornik. Kul’tura, Iskusstvo, Antropologija: 240-250. St Petersburg: Khakasavtodor. Di Cosmo, N. 2002. Ancient China and its enemies: the rise of nomadic power in East Asian history. Cambridge: Cambridge University Press. Foucher, A. 1942-1947. La vieille route de l’Inde de Bactres à Taxila. MDAFA I. Paris: Diffusion de Boccard. Frachetti, M. 2012. Multiregional Emergence of Mobile Pastoralism and Nonuniform Institutional Complexity across Eurasia. Current Anthropology 53/1: 2-38. Frachetti, M. et al., 2017. Nomadic ecology shaped the highland geography of Asia’s Silk Roads. Nature 543/7644:193-198. Francfort, H.-P. 1979. Les palettes du Gandhâra. MDAFA XXIII. Paris: Diffusion de Boccard. Francfort, H.-P. 1984a. The early periods of Shortughaï (Harappan) and the western Bactrian culture of Dashly, in B. Allchin (ed.) South Asian Archaeology 1981: 70-175. Cambridge: Cambridge University Press. Francfort, H.-P. 1984b. Fouilles d’Aï Khanoum III Le sanctuaire du temple à niches indentées 2. Les trouvailles. MDAFA XXVII. Paris: Diffusion de Boccard. Francfort, H.-P. 1985a. The relationship between urban lowlands and mountaneous areas in protohistory as seen from Shortughaï. Journal of Central Asia VIII/2: 125-131. Francfort, H.-P. 1985b. Tradition harappéenne and innovation bactrienne à Shortughaï, in J.-C. Gardin (ed.) L’archéologie de la Bactriane ancienne. Actes du colloque franco-soviétique de Dushanbe: 95-104. Paris: CNRS. Francfort, H.-P. 1988a. Central Asia and Eastern Iran, in J. Boardman et al., (eds) The Cambridge Ancient History: 165-193. Second Edition, Vol. IV. Cambridge/New York/New Rochelle/Melbourne/Sydney: Cambridge University Press. Francfort, H.-P. 1988b. A propos de l’urbanisation du site de Shortughaï (Afghanistan). Une approche archéologique des transformations de l’économie de production. Bulletin du Centre Genevois d’Anthropologie 1: 15-34. Francfort, H.-P. 1992. New Data Illustrating the Early Contacts Between Central Asia and the North-West of the Subcontinent, in C. Jarrige (ed.) South Asian Archaeology 1989: 97-102. Monographs in World

Archaeology 14. Madison Wisconsin: Prehistory Press. Francfort, H.-P. 1998. De l’art des steppes au sud du Taklamakan. Bulletin of the Asia Institute 12: 45-58. Francfort, H.-P. 2001. Des cousins des Scythes?, in C. Debaine-Francfort (ed.) Keriya, mémoires d’un fleuve. Archéologie and civilisation des oasis du Taklamakan. Paris: Editions Findakly: 156-158. Francfort, H.-P. 2002. Images du char en Eurasie orientale des origines à la fin du 1er millénaire av. J.-C., in V. V. Bobrov (ed.) Pervobytnaja arkheologija. Chelovek i iskusstvo. Sbornik nauchnykh trudov, posvjashchennyj 70-letju po dnja rozhdenija Jakova Abramovicha Shera: 80-89. Novosibirsk: Institut Arkheologii i Etnografii R.A.N. Francfort, H.-P. 2007. Choix des nomades and choix des sédentaires en Asie Centrale dans l’adaptation de thèmes and de motifs des arts de la Perse achéménide, in P. Rouillard, C. Perlès, and E. Grimaud (eds) Mobilités, immobilismes. L’emprunt et son refus: 267-282. Paris: Diffusion de Boccard. Francfort, H.-P. 2009. La mort chez les Scythes de l’Europe à l’Altaï: la culture de Pazyryk et ses prédécesseurs à la lumière de quelques découvertes récentes, in J. Guilaine (ed.) Sépultures et sociétés. Du Néolithique à l’Histoire. Séminaire du Collège de France: 153-192. Collection des Hespérides. Paris: Errance. Francfort, H.-P. 2011a. Les objets romains and chinois découverts dans la steppe d’Asie centrale (env. 2e siècle av. J.-C. -2e siècle ap. J.-C). Archéothéma 19: 5057. Francfort, H.-P. 2011b. Tillya Tépa (Afghanistan) la sépulture d’un roi anonyme de la Bactriane du Ier siècle P.C. Topoi 17/1: 277-347. Francfort, H.-P. 2011c. Tillya Tepe and connections with the Eurasian steppes, in J. Aruz and E. Valtz Fino (eds) Afghanistan: Forging Civilizations along the Silk Road: 88-101. New Haven/London: The Metropolitan Museum of Art/Yale University Press. Francfort, H.-P. 2012a. L’art oublié des lapidaires de la Bactriane aux époques achéménide et hellénistique. Persika 17. Paris: Diffusion de Boccard. Francfort, H.-P. 2012b. Royal wine drinking ceremonies from Bactria and Gandhâra: about Saka/Yuezhi and Indo-Scythian Representations, in A. Petrova (ed.) The last Ecyclopaedist. Festschrift for B. Litvinsky. 467-481. Moscow: Institute for Oriental Studies, Department of Ancient Orient. Francfort, H.-P. 2013a. Pratiques guerrières dans les steppes et les oasis de l’Asie centrale au premier millénaire av. J.-C., in O. Buchsenschutz et al. (eds), Archéologie de la violence et de la guerre dans les sociétés pré et protohistoriques: 81-98 Paris: Éd. du CTHS (Actes des congrès nationaux des sociétés historiques et scientifiques). Francfort, H.-P. 2013b. L’art oublié des lapidaires de la Bactriane aux époques achéménide et hellénistique. Persika 17. Paris: Diffusion de Boccard. 108

H.-P. Francfort: Shifting Exchange Patterns During the Bronze and Iron Ages Francfort, H.-P. 2014a. Les nomades installés dans la Bactriane (2nd c. BC – 1st c. AD): Nouvelles découvertes. Comptes-Rendus des séances de l’Académie des Incriptions et belles-Lettres, 2013, NovembreDécembre 157/4: 1543-1576. Francfort, H.-P. 2014b. Les peintures de Miran et l’art de l’Asie centrale. Monuments et Mémoires de la Fondation Eugène Piot 93: 23-64. Francfort, H.-P. 2015. Une idole protohistorique d’Aï Khanoum hellénistique (Bactriane, Afghanistan), in V. Schiltz (ed.) De Samarcande à Istanbul: étapes orientales II: 41-51. Paris: CNRS éditions. Francfort, H.-P. 2016. How the twins met: Indus and Oxus Bronze Age Civilizations in Eastern Bactria. Shortughaï revisited forty years later, in N.A. Dubova et al. (eds) To the memory of Professor Victor Sarianidi: 461-475. Transactions of Margiana Archaeological Expedition. Moscow: N.N. Miklukho-Maklay Institute of Ethnology and Anthropology of Russian Academy of Sciences/Margiana Archaeological Expedition, Altay State University. Francfort, H.-P., with contributions by Ch. Boisset et al. 1989. Fouilles de Shortughaï: recherches sur l’Asie centrale protohistorique. Mémoires de la Mission Archéologique Française en Asie centrale, vol. II. Paris: Diffusion de Boccard. Francfort, H.-P. et al. (eds) 2001. Shamanism. Uses and abuses of a concept. Bibliotheca Shamanistica 10. Budapest: Akadémiai Kiado. Francfort, H.-P. et al. 1990. Pétroglyphes archaïques du Ladakh and du Zanskar. Arts Asiatiques XLV: 5-27. Francfort, H.-P. et al. 1992. Archaic Petroglyphs of Ladakh and Zanskar, in M. Lorblanchet (ed.) Rock Art in the Old World. Papers presented in Symposium A of the AURA Congress Darwin (Australia) 1988: 147-192. New Delhi: Indira Gandhi National Center for the Arts. Francfort, H.-P. et al. 1998. Découverte d’une tombe gelée d’époque scythe (IVe s. av. J.-C.) à Berel’ dans l’Altaï (Kazakhstan oriental). Comptes-rendus de l’Académie des Inscriptions and Belles-Lettres novembre-décembre 142/4: 1165-1175. Francfort, H.-P. et al. 2000. La fouille d’un kourgane scythe gelé du IVe siècle av. notre ère à Berel’ dans l’Altaï (Kazakhstan). Comptes-rendus de l’Académie des Inscriptions and Belles-Lettres 144/2: 775-806. Francfort, H.-P. and Lecomte, O. 2002. Irrigation and société en Asie centrale des origines à l’époque achéménide. Annales Histoire, Sciences Sociales 57/3: 625-663. Francfort, H.-P. and X. Tremblay 2010. Marhaši and la Civilisation de l’Oxus. Iranica Antiqua XLV: 51-224. Fribus, A. 1998. Proiskhozhdenie afanas’evskoj kul’tury. Kemerovo: Kemerovo State University. Gening, V. et al. 1992. Sintashta. Arkheologicheskie pamjatniki arijskikh plemen Uralo-Kazakhstanskikh stepej 1. Cheljabinsk: Juzhno-Ural’skoe knizhnoe izdatel’stvo.

Gösdorf, J. et al. 1998. Neue 14C-Datierung für die Sibirische Steppe und ihre Konsequenzen für dir regionale Bronzezeitchronologie. Eurasia Antiqua 4: 73-80. Grjaznov, M. 1984. Der Grosskurgan von Arzan in Tuva, Südsibirien. Mat. zur Allgemeinen und Vergleichenden Archäologie 23. Munich: Verlag C.H. Beck. Grousset, R. 1970. The empire of the steppes: a history of Central Asia. Rutgers: University of Princeton. Gutsalov, S. 2012. On links between the Ancient Nomads in the Southern Foothills of the Urals and Central Asia. Ancient Civilizations from Scythia to Siberia 18/1: 29-50. Hanks, B. and K. Linduff (eds) 2009. Social Complexity in Prehistoric Eurasia. Monuments, Metals and Mobility. Cambridge: Cambridge University Press. Hansen, V. 2017. The Silk Road: A New History with Documents with coverage of the Mongols and Marco Polo. New York/Oxford: Oxford University Press. Holdich, Th. 1910. The Gates of India, being an historical narrative. London: Macmillan and Company. Ilyasov, J. 2003. Covered tail and ‘flying’ tassels. Iranica Antiqua XXVIII: 259-325. Ilyasov, J. and D. Rusanov 1997/98. A Study on the Bone Plates from Orlat. Silk Road Art and Archaeology 5: 107-159. Isakov, A. 1986. Sarazm – novyj rannezemledel’cheskij pamjatnik Srednej Azii. Sovetskaya Arkheologiya 1: 152-167. Isakov, A. 1991. Sarazm. Dushambe: Donish. Isakov, A. 1996. Sarazm: An Agricultural Center of Ancient Sogdiana. Bulletin of the Asia Institute 8/1994: 1-12. Isakov, A. and B. Lyonnet 1988. Céramique de Sarazm (Tadjikistan, URSS): Problèmes d’échanges and de peuplement à la fin du Chalcolithique and au début de l’Age du Bronze. Paléorient 14/1: 31-47. Jablonskij, L. 2014. Ekipirovka rannesarmatskogo elitnogo voina (po materialam filippovskikh mogil’nikov). Rossiiskaya arkheologiya 2: 27-38. Jansen, M. et al. 2016. Platinum group placer minerals in ancient gold artifacts – Geochemistry and osmium isotopes of inclusions in Early Bronze Age gold from Ur/Mesopotamia. Journal of Archaeological Science 68: 12-23. Jarrige, J.-F. 1988. Le complexe culturel de Mehrgarh (Période VIII) and de Sibri. Le ‘Trésor’ de Quetta, in J.-F. Jarrige (ed.) Les citées oubliées de l’Indus. Archéologie du Pakistan (catalogue de l’exposition du Musée National des Arts Asiatiques Guimet): 111128. Paris: Association Française d’Action Artistique. Jarrige, J.-F. 2017. Indus-Oxus Civilizations: More Thoughts, in V. Lefèvre et al., (eds) South Asian Archaeology and Art 2012. Volume 1. Man and Environment in Prehistoric and Protohistoric South Asia: New Perspectives: 19-33. Turnhout: Brepols. 109

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Jettmar, K. 1982. Petroglyphs and Early History of the Upper Indus Valley. Zentralasiatische Studien 16: 293308. Jettmar, K. and V. Thewalt 1987. Between Gandhara and the Silk roads. Rock-carvings along the Karakorum Highway. Mayence: Philipp von Zabern. Jia, P. and A. Betts 2010. A re-analysis of the Qiemu’erqieke (Shamirshak) cemeteries, Xinjiang, China. Journal of Indo European Studies 38/3-4: 275-317. Keller, D. 2001. Fabulous Creatures from the Desert Sands, in D. Keller and R. Schorta (eds) Fabulous Creatures from the Desert Sands. Central Asian Woolen Textiles from the Second Century BC to the Second Century AD: 9-13. Riggisberger Berichte 10. Riggisberg: Abegg-Stiftung. Kohl, P. 2007. The Making of Bronze Age Eurasia. Cambridge World Archaeology. Cambridge: Cambridge University Press. Koryakova, L. 2002. Between Steppe and Forest: Iron Age Societies of the Urals, in K. Boyle et al. (eds) Ancient interactions: east and west in Eurasia: 265-292. McDonald Institute Monographs. Cambridge: McDonald Institute for Archaeological Research. Koryakova, L. 2007. On the Northern Periphery of the Nomadic World: Research in the Trans-Ural Region, in J. Aruz et al. (eds) The Golden Deer of Eurasia. Perspectives on the Steppe Nomads of the Ancient World: 102-113. The Metropolitan Museum of Art Symposia. New York/New Haven/London: The Metropolitan Museum of Art/Yale University Press. Koryakova, L. and A.V. Epimakhov 2007, The Urals and Western Siberia in the Bronze and Iron Age, Cambridge World Archaeology. Cambridge: Cambridge University Press. Krasnienko, S. 2002. Pamjatniki afanas’evskoj kul’tury na jugo-zapade krasnojarskogo kraja, in J. Piotrovskij (ed.): Stepi Evrazii v drevnosti i srednevekov’e. K 100-letiju so dnja rozhdenija M. P. Grjaznova: 171-175. SaintPetersburg: State Hermitage Museum. Institute for the History of Material Culture, Russian Academy of Sciences. Kubarev, V. et al. 1999. Neue Grabfunde der Afanas’evoKultur am mittleren Katun. Eurasia Antiqua 5: 71-94. Kurinskikh, O. et al. 2013. Detali Paradnogo Trona Akhemenidskogo Tipa iz Filippovki. Vestnik Drevnej Istorii 2: 180-203. Kuz’mina, E. 2007. The Origin of the Indo-Iranian., Leiden Indo-European Etymological Dictionary Series 3. Leiden/Boston: Brill. Leont’ev, N. and V. Kapel’ko 2002. Steinstelen der OkunevKultur. Archäologie in Eurasien. Mayence: Verlag Philipp von Zabern. Levine, M. et al. (eds) 2003. Prehistoric steppe adaptation and the horse. McDonald Institute Monographs. Cambridge: McDonald Institute. Littauer, M. and J. Crouwel 1979. Wheeled vehicles and ridden animals in the ancient Near East. Handbook of

Oriental Studies. Section 7 Art and Archaeology, Volume: 1/2B/1. The Ancient Near East 2B. Leiden: Brill. Litvinsky, B. 2001. The Bactrian Ivory Plate with a Hunting Scene from the Temple of the Oxus. Silk Road Art and Archaeology VII: 137-166. Livshits, V. 2003. Three Silver Bowls from the Isakovka Burial Ground N°1 with Khwarezmian and Parthian Inscriptions. Ancient Civilizations from Scythia to Siberia 9/1-2: 147-172. Lorblanchet, M. et al. (eds) 2006. Chamanismes and arts préhistoriques. Vision critique. Hespérides. Paris: Errance. Lyonnet, B. 1977. Découverte de sites de l’âge du bronze dans le N.E. de l’Afghanistan: leurs rapports avec la civilisation de l’Indus. Annali 37: 19-35. Lyonnet, B. 1981. Etablissements chalcolithiques dans le Nord-Est de l’Afghanistan: leurs rapports avec les civilisations du bassin de l’Indus. Paléorient 7/2: 5774. Lyonnet, B. 1988. Les relations de la bactriane orientale avec le monde indo-baluche à partir des données céramologiques, du 3e au 2e millénaires avant notre ère, in J.-C. Gardin (ed.) L’Asie centrale and ses rapports avec les civilisations orientales des origines à l’âge du fer. actes du colloque franco-soviétique, Paris, 19-26 Novembre 1985: 143-151. Mémoires de la M.A.F.A.C. I: 143-151. Paris: Diffusion de Boccard.  Lyonnet, B. 1996. Sarazm (Tadjikistan) céramiques (Chalcolithique and Bronze ancien). Mémoires de la Mission Archéologique Française en Asie Centrale VII. Paris: Diffusion de Boccard. Magee, P. et al. 2005. The Achaemenid Empire in South Asia and Recent Excavations in Akra in Northwest Pakistan. American Journal of Archaeology 109: 711741. Mallory, J. 1991. In Search of the Indo-Europeans. Language, Archaeology and Myth. London: Thames & Hudson. Mallory, J. 1997. A la recherche des Indo-Européens. Langue, archéologie, mythe. Paris: Seuil. Menghin, W. and Parzinger, H. (eds) 2007. Im Zeichen des Goldenen Greifen. Königs Gräber der Skythen. Munich, Berlin, London, New York: Prestel. Mochkova, M. 1994. Le cimetière sarmate de Lebedevka dans le Sud de l’Oural. Les dossiers d’archéologie 194: 84-87. Museum, Xinjiang Uygur Autonomous Region, Xinjiang Institute of Archaeology 2001, Sampula in Xinjiang of China. Revelation and Study of Ancient Khotan Civilization, Urumqi: Xinjiang People’s Publishing House. Mutin, B. and A. Razzokov 2014. Contacts across the Hindu Kush in the Bronze Age. Additional Insights from Sarazm – Soundings 11-11A (Tajikistan). Archäologische Mitteilungen aus Iran und Turan 46: 123147. Mutin, B. et al. 2017. Resuming Tajik-French Fieldwork at Sarazm, Tajikistan. Preliminary Activity Report 110

H.-P. Francfort: Shifting Exchange Patterns During the Bronze and Iron Ages on the 2011-2012 Field Seasons, in V. Lefèvre et al., (eds) South Asian Archaeology and Art 2012. Volume 1. Man and Environment in Prehistoric and Protohistoric South Asia: New Perspectives: 197-210. Turnhout: Brepols. Nagler, A. 1999. Waren die Träger der Okunev-Kultur Nomaden? Eurasia Antiqua 5: 1-28. Neelis, J. 2006. La Vieille Route Reconsidered: Alternative Paths for Early Transmission of Buddhism Beyond the Borderlands of South Asia. Bulletin of the Asia Institute 16/2002: 143-164. Nikonorov, V. 2000. Unikal’nyj greben’ iz Kampyrtepa. Materialy Tokharistanskoj Ekspededicii. Arkheologicheskie issledovanija Kampyrtepa 1: 131137. Novozhenov, V. 2012. Chudo kommunikacii i drevnejshij kolësnyj transport Evrazii. Moscow: Izdatel’stvo Taus. Pfrommer, M. 1996. Die Gräber von Tillya-Tepe – Eine späthellenistisch-frühkaiserzeitlische Nekropole zwischen achämenidischen Erbe, griechischer Tradition und nomadisch-zentralasiatischer Gegenwart, in B. Funck (ed.) Hellenismus: Beiträge zur Erforschung von Akkulturation und politischer Ordnung in den Staaten des hellenistischen Zeitalters: Akten des Internationalen Hellenismus-Kolloquiums, 9.-14. März 1994 in Berlin: 91-120. Tübingen: J.C.B. Mohr (Paul Siebeck). Pirazzoli-t’Serstevens, M. 2007. Inner Asia and Han China: Borrowings and Representations, in T. Lawton (ed.) New Frontiers in Global Archaeology: Defining China’s Ancient Traditions: 1-20. Washington, AMS Foundation for the Arts, Sciences and Humanities. Polos’mak, N. 2001. Vsadniki Ukoka. Novosibirsk: Infolio Press. Polos’mak, N. 2010. My vypili Somu, my stali bessmertnymi. Nauka iz pervykh ruk 33/3: 50-59. Polos’mak, N. 2011. Svet dalekoj Ellady. Nauka iz pervykh ruk 37/1: 94-107. Polos’mak, N. 2012. Embroideries on Garments from Kurgan 20 of the Noin-Ula Burial Ground. Anabasis. Studia Classica et Orientalia 3: 267-288. Polos’mak, N. 2015. Nouvelles découvertes de tentures polychromes brodées du début de notre ère dans les tumuli n°20 et n°31 de Noin-Ula (République de Mongolie). Arts Asiatiques 70: 3-32. Polos’mak, N. et al. 2011. Dvadcatij noin-ulinskij kurgan. Novosibirsk: Infolio éditions. Polos’mak, N. et al. 2008. The Burial Construction of Noin Ula Mound 20, Mongolia. Archaeology, Ethnology & Anthropology of Eurasia 34/2: 77-87. Pugachenkova, G. 1966. Khalchajan. Tashkent: FAN. Pugachenkova, G. 1971. Skulptura Khalchajana. Moscow: Iskusstvo. Rapin, C. 2007. Nomads and the Shaping of Central Asia: from the Early Iron Age to the Kushan Period, in J. Cribb and G. Herrmann (eds) After Alexander. Central Asia before Islam: 29-72. Oxford: Oxford University Press.

Rapin, C. et al. 2001. La tombe d’une princesse nomade à Koktepe près de Samarkand. Comptes-rendus de l’Académie des Inscriptions and Belles-Lettres JanvierMars 145/1: 35-92. Razzokov, A. 2008. Sarazm (orudija truda i khozjajstvo po eksperimental’no-trasologicheskim dannym). Dushambe: Edjod. Razzokov, A. et al. 2006. Sarazm 5500. Dushambe: ECHOD. Reisinger, M. 2010. New Evidence about Composite Bows and Their Arrows in Inner Asia. The Silk Road 8: 42-22. Rozwadowski, A. 2004. Symbols through Time. Interpreting the Rock Art of Central Asia. Poznan: Institute of Eastern Studies Adam Mickiewicz University. Rtveladze, E. 2012. Velikij Indijskij Put’. Iz istorii vazhnejshikh torgovykh dorog Evrazii. SaintPetersburg: Nestor-Istorija. Rubinson, K. 2008. Tillya Tepe: Aspects of Gender and Cultural Identity, in K. Linduff and K. Rubinson (eds) Are all Warriors Male? Gender Roles on the Ancient Eurasian Steppe: 51-63. Lanham, New York, Toronto, Plymouth UK: Altamira Press. Rudenko, S. 1969. Die Kultur der Hsiung-Nu und die Hügelgräber von Noin Ula. Bonn: Rudolf Habelt Verlag. Santoni, M. 1984. Sibri and the South Cemetery of Mehrgarh: Third Millenum Connections between the Northern Kachi Plain (Pakistan) and Central Asia, in B. Allchin (ed.) South Asian Archaeology 1981:52-60. Cambridge: CUP. Santoni, M. 1988. Aspects matériels des cultures de Sibri et de Mehrgarh VIII (plaine de Kachi, Baluchistan, Pakistan) à la fin du troisième and au début du deuxième millénaires, in J.-C. Gardin (ed.) L’Asie centrale et ses rapports avec les civilisations orientales des origines à l’Age du Fe: 135-141. Mémoires de la Mission Archéologique Française en Asie Centrale I. Paris: Diffusion de Boccard. Sarianidi, V. 1985. The golden hoard of Bactria: from the Tillya-tepe excavations in northern Afghanistan. Leningrad: Aurora Art Publishers. Sarianidi, V. 2005. Gonurdepe. City of Kings and Gods. Ashgabat: Miras. Sarianidi, V. 2007. Necropolis of Gonur. Athens: Kapon Editions. Sarianidi, V. and N. Dubova 2011. Nouvelles sépultures sur le territoire de la « nécropole royale » de Gonur Dépé. Arts Asiatiques 65: 5-26. Sarianidi, V. and G. Koshelenko 1982. Monety Tilljatepa, in G. Bongard-Levin (ed.) Drevnjaja Indija. Istorikokul’turnye svjazy: 307-318. Moscow: Nauka. Savinov, D.G. and M. Podol’skij (eds) 1997. Okunevskij Sbornik. Kul’tura, Iskusstvo, Antropologija. StPetersburg: Khaksavtodor. Schiltz, V. 2002. Les Sarmates entre Rome and la Chine. Nouvelles perspectives. Comptes-rendus de l’Académie des Inscriptions and Belles-Lettres. 146/3: 845-887. Stark, S. et al. (eds) 2012. Nomads and Networks. The Ancient Art and Culture of Kazakhstan. New York: 111

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Institute for the Study of the Ancient World/ Princeton University Press. Thote, A. 1999. De quelques conventions picturales. Le char et ses représentations aux Ve-IVe siècles avant notre ère. Études chinoises. Mélanges de sinologie offerts à Monsieur Jean-Pierre Diény XVIII/1-2: 179-220. Treister, M. 1997. New discoveries of Sarmatian complexes of the 1st century A.D. A survey of publications in VDI. Ancient Civilizations 4/1: 36100. Treister, M. 2004. Cloisonné and champlevé decoration in the goldwork of the late Hellenistic-early Imperial periods. Acta Archaeologica 75: 189-219. Treister, M. 2009a. Silver-gilt Bowl from Burial-mound B at Prokhorovka. Ancient Civilizations 15: 183-189. Treister, M. 2009b. Silver Phialai from the Prokhorovka Burial-mound N°.1. Ancient Civilizations 15: 94-135. Treister, M. 2010. ‘Achaemenid’ and ‘Achaemenidinspired’ Goldware and Silverware, Jewellery and Arms and their Imitations to the North of the Achaemenid Empire, in J. Nieling and E. Rehm (eds) Achaemenid Impact in the Black Sea Communication of Power: 223-279. Aarhus: Aarhus University Press. Treister, M. 2012. Silver Phalerae with a Depiction of Bellerophon and the Chimaira from a Sarmatian Burial in Volodarka (Western Kazakhstan). A Reappraisal of the Question of the So-Called GraecoBactrian Style in Hellenistic Toreutics. Ancient Civilizations from Scythia to Siberia 18/1: 51-109. Treister, M. and L. Jablonskij (eds) 2012. Vlijanija akhemenidskoj kul’tury v juzhnom Priural’e (V-III vv. do n.e.). Drevnjaja torevtika i juvelirnoe delo v vostochnoj Evrope. Moscow: TAUS.

Turchin, P. 2009. A theory for formation of large empires. Journal of Global History 4/02: 191-217. Vinogradova, N. and B. Bobomullaev 2016. Predvaritel’nye resultaty issledovanij moil’nika Farkhor – pamjatnika epokhi srednej bronzy na juge Tadzhikistana vesnoj 2014 goda, in N. Dubova, E. Antonova et al. (eds) To the memory of Professor Victor Sarianidi: 534-552. Transactions of Margiana Archaeological Expedition. Moscow: N.N. MiklukhoMaklay Institute of Ethnology and Anthropology of Russian Academy of Sciences, Margiana Archaeological Expedition, Altay State University. Vogelsang, W. 1992. The Rise and Organisation of the Achaemenid Empire. The Eastern Iranian Evidence. Studies in the History of the Ancient Near East III. Leiden/New York/Cologne: E.J. Brill. Wang, B, and X. Xiao 2001. A General Introduction to the Ancient Tombs at Shanpula, Xinjiang, China, in D. Keller and R. Schorta (eds) Fabulous Creatures from the Desert Sands. Central Asian Woolen Textiles from the Second Century BC to the Second Century AD: 47-78. Riggisberger Berichte 10. Riggisberg: AbeggStiftung. Yatsenko, S. 2012. Yuezhi on Bactrian Embroidery from Textiles Found at Noyon uul, Mongolia. The Silk Road 10: 39-48. Zdanovich, D. 2002. Arkaim: Nekropol’ (po materialam kurgana 25 Bol’shekaraganskogo mogil’nika) kniga 1. Cheljabinsk: Juzhno-Ural’skoe knizhnoe izdatel’stvo. Zelinskij, A. 1964a. Ancient routes through the Pamirs. Central Asian Review 13/1: 44-54. Zelinskij A. 1964b. Drevnie puti Pamira. Strani i Narody Vostoka III: 99-119.

112

Chapter 8

The Kashgar Oasis: Reassessing the Historical Record Marika Vicziany and Angelo Andrea Di Castro1 This is the first of two chapters that address the importance of Kashgar for a better understanding of the historical and cultural importance of Xinjiang. Research on Kashgar has been seriously neglected by Western scholars. Even the impressive two volume work on Xinjiang edited by Mair, after a conference held in Philadelphia in 1998, included only one reference to Kashgar in the index compared with 14 to Kucha, 10 to Turfan, five to Urumqi and three to Khotan (Mair 1998, Vol. 2, Index). We believe that Kashgar merits equal attention. We begin this chapter with a discussion of the neglect Kashgar has suffered in Western scholarship and the consequence of this, in particular for the preservation of China’s cultural heritage. The rest of this chapter reviews the literature about Kashgar as presented in various historical documents and secondary sources. Our main argument is that the history of Kashgar has been dominated by chronologies about the marauding activities of various nomadic tribes and their settlement in the oases of the Tarim Basin, the efforts by successive Chinese emperors to contain nomadic incursions and expansions, and the rise and fall of kings who were either local personalities, or foreign imposters or puppets placed into power by whatever dominant power ruled southern Xinjiang. 1

Asia) and the east (China and East Asia) and second, the persistence of trade indicates the long term prosperity of the oasis towns despite periods of political decline due to conflicts caused by competing powers. In other words, we suggest that the human settlement pattern in the Kashgar oasis in the very long run can best be understood by studying trade patterns rather than the rise and fall of empires. Indeed the fall of emperors and their empires, can herald periods of prosperity for ordinary peoples as the new histories on India show. In the future, we also hope that scholars will closely study ‘everyday aspects of social relations in a specific historical setting’, an innovative approach that BellerHann has started in her study of Xinjiang for the period 1880-1949 (2008: 26).

There is a parallel narrative to this conventional history that weaves itself into and out of the history books but rarely receives the kind of regular attention it deserves, namely stories of the merchants who crossed the Tarim Basin and the Tianshan Mountains and carried many commodities for use by local and distant consumers. We believe that this narrative needs to be the focus of future research for two reasons: first, the Kashgar oasis was a major transition zone for long distance trade between the west (Europe and Central

In this chapter we simplify the historical facts by using the name Kashgar to describe the town also known as Shule before the 8th century AD. The modern, sinicised name for Kashgar is Kashi, but we use the term Kashgar because this is what appears in the historical accounts and generally in scholarship. The etymology of the name ‘Kashgar’ has exercised many scholars, in particular Sir Aurel Stein whose explanation remains the most detailed and comprehensive discussion of the problem (Stein 1907: Vol. I, 47-51; see also Pelliot 1959: 196-214). As Stein (1907 Vol. I: 48, note 7) argues after a meticulous examination of all sources, the name Kashgar did not come into general usage until the raids of the town by the Arab Governor of Khorasan Qutayba in 715-716 AD. Gibb (1922: 473, 474) has argued that Qutayba never reached Kashgar and that a later raid by a group of Turkish soldiers in 717 AD has become conflated with the movements of Qutayba which were apparently restricted to the Ferghana Valley. However in a more recent work, Beckwith (1987: 82, note 149), looking at medieval Arabic sources (Ibn A’tham alKūfī) does not discard the possible historicity of the

The environmental history of Kashgar is an equally important tool for understanding the evolution of this oasis town. Hence our second chapter in this book considers the inter-relationships between human settlement patterns and the environment together with human interventions to moderate the natural conditions of the Kashgar oasis. The Neglect of Kashgar’s Heritage Monuments

The authors of this chapter are members of Monash University’s Kashgar Research Project, a multidisciplinary, international research collaboration established in 2004 with the Xinjiang Normal University and the support of the Department of Tourism and the Department of External Affairs, Xinjiang Government. Our first major book was published in 2008 (Michell et al. with photographs by Gollings). In this chapter we have used some of the original photographs taken by John Gollings during our visit to Kasghar in 2005. All 7,000 original photographs taken by Gollings can be found on a digital archive at Monash University under the heading Kashgar Project: Gollings Collection (https://www.monash.edu/library/collections/special/ kashgar). We are grateful to John Gollings for his contribution and ongoing involvement in the Kashgar Project. We also take this opportunity to thank the British Library for permission to reproduce Figure 14 from the Frank Ludlow Collection.

1 

113

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads raid of Qutayba. Despite this, Stein is probably right in suggesting that the name Kashgar came into general use much later than the name Shule. Stein (1907: Vol. I, 48, 52) did not find the many etymological disputes about the name of Kashgar surprising, given the importance of the town and general region as a conduit between China and the west and the south and its role as an emporium for pre-historical and pre-modern long distance trade. Official Chinese economic policy today is rejuvenating the old historical importance of Kashgar as an international trading hub to promote the PRC’s ‘Belt and Road policy’.

the site of some of the oldest Buddhist stupas5 in China (see Di Castro, Vicziany and Zhu in the next chapter); and the conversion of local peoples by Satuq Boghra Khan in about 934 AD was the beginning of local people adopting Islam (see below about the ‘Islamic Records on Kashgar’).6 Many peoples, tribes, religious leaders, warriors, bureaucrats, soldiers and kings have fought over the Kashgar oasis to achieve regional supremacy. According to contemporary farmers in Kashgar, the soil surrounding the ancient city of Khan-oi (this may have been another old capital of the oasis, some 10 kms east of modern Kashgar) is prized for its richness, which according to local tradition, is the result of the vast quantities of blood that were shed in the religious wars between Islamic warriors and the residents who resisted conversion. Despite these legends, the religious beliefs of the local people of Khan-oi at the time of the early Islamic conversions remain uncertain. The legend about how Satuq Boghra Khan became a Muslim (see below section on ‘Islamic Records on Kashgar’) only tells us that the King of Kashgar during Satuq’s youth was a patron of temples that housed ‘idols’. Whether or not there were religious wars of great ferocity is unknown. An analysis of the soils of Khan-oi for phosphate might throw light on why peasants believe the soil to be so rich (see Di Castro, Vicziany and Zhu in the next chapter).

The Kashgar oasis (Vicziany and Di Castro 2014) historically consisted of the town of Kashgar surrounded by orchards, cultivated fields of wheat,2 millet and possibly barley, and eventually the fringes of the Taklamakan desert. According to Stein, the Kashgar that he visited in the early 20th century was the same town as the Kashgar of the early 16th century. He based this on a careful comparison of his own observations and those by Mirza Haider who lived in Kashgar c. 15141530 AD and wrote about it as an older man by which time he was the dominant political figure in Kashmir (Stein 1907: Vol. I, 73; Hambis 1964: 5). Stein’s Kashgar is known today as ‘old Kashgar.’3 However, this is probably a misnomer because the site of Upal (Wupaer), near the gardens of Mahmud al-Kashgari’s tomb (1005-1102 AD), might turn out to be an older centre of human settlement in the oasis (see below and Di Castro, Vicziany and Zhu in the next chapter). Indeed, as we argue below and in the next chapter, there have been many ‘Kashgars’.

Much remains to be done to protect the cultural heritage of Kashgar, despite recent urban renewal projects and increasing investment by the governments and private companies in the local economy. The new economic prosperity of Kashgar need not be achieved to the detriment of the city’s significant cultural monuments, especially the ones from the Buddhist era. Indeed, the cultural monuments of Kashgar could well augment the new wealth that might be generated by the enormous trade potential that has become a priority for the central and provincial governments. In 2010, Kashgar was declared to be a Special Economic Zone (SEZ). This was bound to happen given Beijing’s stress since 2000 on cross border trade and the role of the nearby international border crossing at the Torugart Pass (ADB 2002: Fig. 6-2, 133). However, it was the first new SEZ in 15 years and it was accompanied by the central government ordering the Guangdong government to transfer some US$1.5 billion to Kashgar over five years, as a way of increasing the impact of the

The Kashgar oasis is a land of contradictions: on the one hand there is little question that the general area has been for thousands of years one of the most important entry points into China from Central Asia, the Middle East, Europe and South Asia. Despite this, there has been little effort or investment in research on its premodern history. The importance of Kashgar is based on many considerations: wheat may have entered into central China via the southern parts of Xinjiang given the early dates for wheat at various sites in Afghanistan and the Ferghana Valley;4 the Kashgar oasis remains The spiritual significance of wheat in Kashgar is captured by one of the legends of Bibi Maryam (see main text below) in which she sought to escape her attackers by hiding under a store of wheat provided by a local farmer (Hayes 2012: 191). 3  The old city centre of Kashgar used to be distinct from the new, modern part which lies outside the old city walls. During our visits to Kashgar in 2013, 2014 and 2018 we noticed that much of the old city had been demolished. The dramatic wall of the old city along the ring road near the former British and Russian consulates had been retained together with a section of the high point of the old city to the width of some 6 houses and up to the original height we had documented in our photographic essay published in London (Michell et al. 2008). 4  The current archaeological record favours the arrival of wheat in China via the northern route, but that is partly because little archaeological research has taken place in the Kashgar region. 2 

Many observers wrongly describe ‘stupa’ as a Buddhist temple; however, the stupa is not a temple. The world ‘stupa’ came into use by British scholars and administrators in the early 19th century to describe a dome shaped structure originally built to house the relics of the Buddha or other sacred items associated with him and his followers. 6  See Grenard 1900. According to Roux (1990: 47), despite its Islamic appearance, the legend of Satuq Boghra Khan’s conversion should be appreciated in the broader Turkic-Mongolian religious context. 5 

114

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record central government’s allocation of development funds to Kashgar (Fish 2010). In 2005 an annual trade fair was inaugurated which in June 2012 was also expected to increase local investment by some 20 per cent (Shasha 2012). In April 2013 we drove along the road from Kasghar city to Upal (Wupaer) and noticed that the first half of the 45 kilometres to Upal had been developed into an industrial zone with many construction sites, factories, warehouses, artisanal workshops, wholesale and retail outlets. In April 2015 China and Pakistan signed one of the biggest ever aid grants received by Pakistan. US$46 billion will be spent on the development of ‘an economic corridor’ between Kashgar and Gwadar, the Chinese built port in the southern province of Baluchistan in Pakistan (Houreld 2015). This is one of the largest initiatives taken by China under President Xi Jinping’s new strategy to upgrade and expand the country’s infrastructure links with its neighbours under the ‘Belt Road’ initiative. As these goals come to fruition, Kashgar is likely to benefit but measures are needed to ensure that the new prosperity protects the cultural heritage of the city. The next chapter addresses the concerns we have about that (see also Vicziany and Di Castro 2016).

had already vanished. Figure 1 shows the location of the eight main archaeological sites that we have investigated since 2005. Many of these were probably Buddhist, but their identity awaits verification with detailed research and dating, as we argue in the next chapter. Amongst the buildings that have disappeared is the ‘Pigeon House’ or Kaptar Khana that Sir Aurel Stein (1907: Vol. I, 85) documented. On the ground before us in 2005 we saw a large concrete sign that had previously been positioned in front of the ‘Pigeon House’ – the writing on the sign said that this was an important cultural relic and protected by the government (Figure 2). As we spoke to the local peasants who showed us the site, we learnt that the local government had recently given permission to reclaim the land for agricultural purposes. It is unlikely that the peasants deliberately destroyed the building: rather the building may have been subjected to continued erosion and decline until there was little left worth preserving from the viewpoint of the local community that is poorly educated about the cultural and historical importance of Kashgar. A photograph in Stein’s book (1907: Vol. I, fig. 17 facing page 112) shows that the structure was already in ruins and probably prone to ongoing decline. Yet in his time it was still 16.5 feet high, 23 feet long and 22 feet wide.

By the time of our first period of fieldwork in Kashgar in 2005, many of the archaeological sites of the oasis

Figure 1. Diagram showing the location of the eight main archaeological sites in the Kashgar oasis discussed in this chapter.

115

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Satellite image, when compared to Google Earth imagery, shows the encroachment of new farming lands (Figures 4 and 5). This comes on top of decades during which farmers have dug up the historic soils of Khan-oi and removed them to nearby fields because of their alleged fertility. A few kilometers along the road beyond Khan-oi is the site of another stupa locally known as Topa Tim. Mr Qadir, the former Director of the Kashgar Museum, located and identified it from a careful study of old historical documents. A visit to the area during April 2013 confirmed Figure 2. Former location of Pigeon House near Khan-oi, the images from Google Earth that showing the concrete sign erected by the Kashgar Municipality the site is threatened by a large allfor the protection of this site (photo Gollings). weather road that divides the main mound (possibly the stupa) from This considerable structure was characterised by many associated monastic structures (Figure 6). The main small niches and a central void that contained hundreds Topa Tim mound (Figure 7) remains intact, but other of bones which led Stein to comment that the building parts of the site have been heavily disturbed by the was not typically Buddhist, Islamic or Zoroastrian but construction of the new road (see Di Castro, Vicziany probably Christian (Stein 1907: Vol. I, 85-86). In Stein’s and Zhu in this volume). day the surrounding environment was a ‘desolate area on barren land’ in contrast to the situation today where Those cultural sites that do have some ‘protection’ are the site is the middle of farmed fields. only minimally guarded. For example, the site of the This was not our only experience of how the cultural sites of the Kashgar oasis continue to decline and disappear. In 2005, Abduyraehim Qadir, the then Director of the Kashgar Museum agreed to show us the site of a ‘watch tower’ within the limits of the city.7 We were interested in this as many Chinese sources describe the old Buddhist stupas as ‘watch towers’. He offered to check the site before our visit the next day. On our return we found a tearful man who informed us that the watch tower had recently been demolished to make way for the expansion of the road which now bore the name of the destroyed structure. The ‘watch tower’ in question was the Kizil Debe, one of Kashgar’s stupas. Stein described it and Rhie also referred to it in her analysis of the structural features of Kashgar’s Buddhist stupas. Stein (1907: Vol. I, 78) reported considerable siltation around ‘Kizil Debe’ or the ‘red mound’. Rhie (1999: Vol. 1, 250) noted the rectangular base of the structure and dated the ‘stupa’ to the 2nd century AD. Today the only indication of its existence is the place name of the neighborhood and road signs reading ‘Kizil Debe’ (Figure 3). The old city of Khan-oi is another cultural site that appears to be under threat from development- this time from farmers rather than urban expansion. A Quickbird

Figure 3. Road sign placed near the original location of the Kizil Debe Stupa in contemporary Kashgar (photo Di Castro).

We are sorry to report that the highly knowledgeable Mr Qadir, died in early May 2018.

7 

116

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record

Figure 4. Satellite view of the area of Khan-oi in 2006 (acquired from Quickbird, March 2, 2006).

Figure 5. Agricultural encroachment on Khan-oi in 2012 (from Google Earth).

at a distance of about 547 m from the main stupa as estimated from Google Earth images. Without further investigations, we cannot be sure that the land surrounding Mori Tim has remained stable – what we can see is that the immediate environment has suffered successive flooding and erosion. There is a risk that the vibrations of the passing trains have further destabilized the area around this very important monument.

Mori Tim stupa is surrounded by a wire fence and a guard attends each day. He opens and closes the gate to visitors. The enclosed area extends well beyond the stupa and includes ancient soils, old agricultural fields demarcated by small earthen boundaries and irrigation channels that linked the fields to the old Chakmak River bed. Beyond the wire fence towards the orchard of Beshkerem and the city of Kashgar, there is much evidence of recently abandoned farm houses and fields. In contrast to the green fields of the orchards, this area is characterised by salt encrusted, arid earth that glistens in the sun. Salination is an index of the fragile environment of this area. Another man-made source of fragility arises from the main railway line that connected Kashgar to Urumqi in 1999. It runs

The site around the Mori Tim stupas is more significant than we initially realised: in 2007 we discovered via the study of Google Earth images a long line of irrigation wells or karez (Figure 8). On the ground inspections in April 2013 confirmed that there were 48 irrigation wells which could date from the Buddhist era in Kashgar. 117

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 6 a and b. Satellite images of the site of Topa Tim in 2005 (a) and 2013 (b) showing new road dividing the stupa on the right from the monastery on the left (from Google Earth).

Proper dating, however, needs to be done. Until then, all possible measures should be taken to protect the stupa and the surrounding, unique environment. Within the city of Kashgar itself, in 2005 we witnessed the destruction of part of the wall that surrounded the limits of the old city of Kashgar, dated to the 16th century AD (Stein 1907: Vol. I, 73, note 1) (Figure 10). The bulldozers were acting in response to the demands of urban residents who lived close to the old city walls – they were fearful of the wall collapsing on top of them (Vicziany: 2005, 15). The removal of the wall also facilitated the expansion of the municipality by providing room for new roads and other infrastructure. Today, the images on Google Earth show that this part of Kashgar has been intensively developed with many new, densely packed houses. Unfortunately, we cannot verify that the occupants of these houses are the former residents. The unprecedented pace of urban development throughout China has been a history of richer residents buying land close to the city centres and then pushing the old residents to the periphery of China’s towns or into less desirable locations. In 2011

Figure 7. The stupa of Topa Tim in the background, 2007 (photo Di Castro).

118

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record

Figure 8. Satellite image showing the line of karez adjacent to the Mori Tim sacred area (acquired from Quickbird, March 2, 2006).

On a more dramatic scale, fieldwork in April 2013 revealed that the integrity of the city centre of old Kashgar has now been compromised by urban renewal projects that have seen the removal of some 80 per cent of the buildings that we visited in 2005. A good slice of the original circular wall and its adjacent houses have been retained in an area that slopes down to the Tumen River. At night, the wall is illuminated and the whole area looks very attractive. The demolitions are not regarded by local residents as especially controversial from the viewpoint of cultural conservation; rather, controversy is usually focused on questions of financial compensation and the provision of alternative housing. Residents had complained about the poor quality of the old housing while local authorities were worried about possible collapse in this earthquake prone area (see discussion on earthquakes in the next chapter).8 On the other hand, the urban renovations that have been undertaken in this area appear to have been implemented without any assessment of the relative importance of the old historical buildings, artisanal workshops and dozens of mosques that used to define the old town. An exception to this generalization is the effort that has been made to retain much of the original character of the old artisan shops adjacent to the Idka Khan mosque. Many of the small mosques in this area are managed and funded by particular artisan guilds.

these ‘forced demolitions’ were banned by the State Council of China because of the escalating violence and civil disorder that they have generated (Xinhua 2011). Another view is that even if there were safety concerns about the old wall of Kashgar city, there are ways of conserving the structure while protecting residents.

Figure 9. One of the 48 karez wells in the vicinity of the sacred area of Mori Tim (photo Di Castro).

8  For example, in 1948 the dome of the Apa Khoja Mausoleum was damaged by an earthquake (Hayes 2012: 33, fig. 2).

119

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 10. Bulldozers removing part of the old wall of Kashgar in 2005 (photo Vicziany).

based his knowledge of the names and distances on the reports of Marinos who in turn had collected information from the trade expedition organised by the Syrian entrepreneur Maes Titianus in about 20-1 BC. It was an enormously ambitious expedition, expected to cover some 2000 miles with the sponsors hoping to eliminate the exploitative middlemen who extracted high profits from the Chinese silk and other goods that travelled from central China to the ‘Stone Tower’ on the western borderlands of the Tarim Basin and beyond.

The Historical Significance of Kashgar Kashgar’s historical monuments and sites are worthy of careful preservation because the town has great historical significance, especially because of its role as a cultural centre in the evolution of Buddhism and Islam in China. In this part of this chapter we consider the European, Chinese and Persian records that speak of Kashgar’s importance, followed by an examination of the emergence of Kashgar as a cultural centre. The earliest Western documents

However, Sir Aurel Stein disputed this on the basis of a careful comparison of different records and his own travels through the border regions. He (1907: Vol. I, 54, note 17) rejected his own earlier misidentification of the Stone Tower. Stein’s quest to find the direct route between Bactria and Kashgar as followed by the European and Chinese caravans explains the enormous number of photographs of the passes and mountains of western China held in the Stein photographic archives in the British Library. Stein (1907: Vol. I, 55) finally agreed with Baron Richthofen that the Stone Tower was none other than Irkeshtam which at the time of Stein’s visit was a ‘Russian customs station and fortified frontier post on the road to Kashgar’. Irkeshtam provided fodder and fuel enabling the caravans from the west to replenish their supplies before continuing down the mountains towards Kashgar. Rapin (2001: 218-219) referring to the Stone Tower of Ptolemy (VI. 13), says that this marked the crossroad of the main trade routes between the land of the Saka tribes (giving access to Bactria and Sogdiana) and the Tarim oases. This was one terminal of the itinerary of the merchant Maes Titianus. From here it was possible to reach Kasia khora, the Kasia of Ptolemy (VI, 15). Rapin argued that the Stone Tower rather than being an actual watchtower made of stone (which has caused it to be identified by various scholars as Tashkurgan/Taxkorgan in Xinjiang), should be recognised as a natural formation called the Throne of Suleiman (Takht-i Suleiman), near Osh in Kyrgyzstan, in the Ferghana Valley. The Saka controlled the crossroads of the Lithinos Pyrgos (Stone Tower), from where they eventually entered Bactria in the 2nd century BC when they attacked Ai Khanum/Eucratideia. At the time of Maes Titianus, the Saka, who were occupying Bactria, were in competition with the Yuezhi/ Tocharoi for the control of the trade routes. Despite these local rivalries the trade along the Silk Road continued to prosper (Lerner 1998).

The earliest Western document to speak of the Kashgar region was the Geography by Ptolemy: he provided accompanying maps all compiled during the early 2nd century AD. Ptolemy (1966: VI, 15-16) referring to the mountain range, or simply the region around Kashgar, uses one of the old names for the city, namely ‘Kasia’. The main objective of the Geography was to produce ‘an imitation through drawing of the entire known part of the world together with the things that are, broadly speaking, connected with it’ (Berggren and Jones 2000: 3). Many of the 8,000 localities mapped by Ptolemy in his grid of longitudes and latitudes were first used by his contemporary Marinos of Tyre, but copies of the latter’s work have been lost for centuries thereby making us dependent on Ptolemy. In his Geography Ptolemy includes Serike or China which was approached overland from Bactria into Xinjiang via the ‘Stone Tower’ located in the land of the nomadic Saka (Berggren and Jones 2000: Appendix C, The Trade Route Across Central Asia, 151; also 140 Map 8a based on Map 8 Ptolemy).9 Ptolemy 9  The Stone Tower referred to by Ptolemy has been widely identified by many scholars as the Xinjiang town of Tashkurgan/Taxkorgan.

120

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record It is significant to the central theme of this chapter that Maes Titianus has hardly attracted any scholarship, Cary’s work being one of the few exceptions. More importantly, Cary himself suffers from the tendency of many historians to overplay the role of political centralisation as a precondition for inter-regional trade. Cary’s (1956: 132) date for Maes Titianus’s trading expedition is based on his assumption that it must have occurred before the Kushan ‘invasions’ of south western China. Cary asks why Maes failed to establish a permanent trade link between China and the West. His (1956: 133-134) answer was that the Kushan invasion ‘had the effect of cutting communications between China and the western world for more than half a century’. McLauchlin, writing some five decades after Cary, dates the Maes’ expedition to about AD 100 which is about a century later than Cary’s date of c.20-1 BC. But the two authors agree on one thing – that political control by the Chinese emperor was critical to the success of Maes’ trade expedition. McLauchlin (2010: 108) assumes that because the Han Emperor at this time (i.e. AD 100) ‘controlled’ the Tarim Basin, conditions were right for trade because the ‘unique conditions’ of Han control over the Tarim Basin only lasted for ten years.

and economic decline. Those views were overturned by Bayly who pointed out that the decline of the Mughal elite at the centre of the empire did not automatically lead to the decline of other groups, other cities or other regions. His study showed how the end of the Mughal Empire saw the Mughal soldiers, merchants and administrators adjusting to their new circumstances by working for successor states and local lords. Moreover, after the Mughals, Gangetic India thrived thanks to the freedom and ingenuity of the merchants and service gentry (Bayly 1983: 8). It is this kind of history that has yet to be written for the Tarim Basin. Some early Western accounts of Xinjiang do, in fact, focus on matters of commerce even if these have been overwhelmed by the political narratives that dominate contemporary historical explanations of the evolution of the Tarim Basin. After Ptolemy, another example of Western interest in the Kashgar region occurs in about 353 AD in the work of the Roman historian Ammianus Marcellinus. Marcellinus (1989: XXIII, 6.60) wrote a detailed account of the Roman Empire and the lands beyond it, including China or Seres. He tells us of the tireless merchants who moved towards China via the border village of Lithinos Pyrgos (the above mentioned ‘Stone Tower’), an unambiguous reference to the same border crossing named by Ptolemy some 200 years earlier (Macpherson 1805: Vol. 1, 209). From this we can surmise that the ‘Stone Tower’ was not a building but probably a prominent hill that marked the border crossing (see above note 9).

Our problem with this kind of explanation of early ‘western’ traders to China is that they give a special agency to the Chinese emperor and his representatives in the Tarim Basin. Why should we assume that Chinese control over the Tarim Basin was more conducive to trade than control by others? Why should Chinese imperial conquests be equated with social and economic stability? More importantly, what kind of ‘control’ was possible given the vast distances that had to be covered between the imperial city of Chang’an (modern day Xian) and the western edge of the Tarim Basin and the poor state of communications? Would it not be better to think of the control of the Chinese emperor as a kind of tributary relationship with western China? And was this tributary relationship not also typical of other political states that contested the claims of supremacy by the Chinese emperor? If we think of loose tributary models of ‘political control’ then the persistence of long distance trade becomes a stronger possibility irrespective of whether the Chinese emperor or some other king was dominant at any particular time. And if trade continued, even when the rule of the Chinese and others had collapsed, what kind of related activities did the merchants, other intermediaries and peasants engage in?

The third important reference to Kashgar occurs in The Travels by Marco Polo. In the late 13th century, Marco Polo with his father and uncle entered China via Kashgar as had thousands of other travelers and traders before them. At that time Kashgar was part of the Mongol Empire founded by Chinggis Khan. The Polos were Venetian merchants who already had extensive experience of living and trading in the Middle East and Turkey. Yet they were impressed by what they saw in western China: It has villages and towns a plenty. The biggest city and the most splendid, is Kashgar. The inhabitants live by trade and industry. They have very fine orchards and vineyards and flourishing estates. Cotton grows here in plenty, besides flax and hemp. The soil is fruitful and productive of all means of life. This country is the starting point from which many merchants set out to market their wares all over the world. (Polo 2016: 80).

These questions have arisen as a result of our exposure to debates in South Asian history about the impact of the decline of the Mughal Empire. Initially, the decline of the Mughals (one of the largest empires in Indian history c. 1526-1858) was seen as giving way to political decentralisation, followed by social disorder

Our examples of early western references to the Kashgar region all have one thing in common – they stress the importance of Kashgar’s role in international overland trade. In the next section we consider the Chinese records. The oldest of these predated the European by 121

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads two hundred years and they confirm the strategic and economic significance of the Kashgar oasis. In contrast to European accounts the Chinese ones are far more concerned with political issues largely because the compilers of the Chinese reports were appointed by the Emperors to write chronicles for the imperial courts.

documents in the latter compilation are Chapters 96 and 21 in the Hanshu respectively called ‘The Memoir on the Western Regions’ (hereafter MWR) and ‘The Memoir on Zhang Qian and Li Guangli’ (hereafter MZL). Our main source has been Hulsewé’s translation in 1979 of the Hanshu.

The first Chinese records of Kashgar

General Zhang Qian had been sent on a mission by the Emperor of China, Wudi (149-86 BC),15 to find the leader of the Yuezhi. The Yuezhi were the old enemies of the Xiongnu but by the time General Zhang Qian found them they were well settled into a good life west of Ferghana and not interested in helping the Chinese emperor fight the Xiongnu. Zhang Qian’s report provided the Chinese Emperor with the first reliable account of the socio-economic and political situation in China’s western region and the ‘extreme west’ i.e. beyond the western mountains. In the process he discovered the famous ‘Heavenly Horses’ of the Ferghana Valley (Hulsewé MWR 1979: 132-4). The Chinese Emperor became convinced that these powerful, sturdy and fast-moving blood-sweating horses could save China from the deprecations of the Xiongnu.16 Yu (1998: 81), however, points to deeper economic and social motives for the campaigns that followed General Zhang Qian’s initial visit, including Emperor Wudi’s determination to secure the western regions of China by subduing the extreme west on the other side of the mountains.

The first Chinese records that mention Kashgar arise from General Zhang Qian (d. 114 BC) passing through the Kashgar oasis on his way from the capital of the Chinese empire in Chang’an to the Ferghana valley in search of allies to fight the marauding Xiongnu (or Hun).10 The Xiongnu were seen by the Chinese chroniclers as a savage, nomadic,11 warlike people who extracted tribute in return for not attacking the Chinese. The Xiongnu confederation developed an imperial structure that inhabited the steppes to the north of China. Drompp (2005: 14-19) speaks of the ‘Xiongnu model’ by which he meant that as the principle enemy of the Han, they were fashioned into a stereotype of barbarism which was subsequently applied by the Chinese courts to other nomads.12 On General Zhang Qian’s return to the Chinese capital in 125 BC he reported his experiences and observations to the Emperor13 and these have been incorporated into the two earliest histories of China: the Shiji by Sima Qian (c. 90 BC) and the Hanshu or The History of the Former Han Dynasty by the Ban family. The former covers the period from China’s antiquity to c. 90 BC and the latter the period 202 BC to c. 23 AD.14 To understand Kashgar’s history, the two critical

The Yuezhi were not keen on trading their horses for gold, so Emperor Wudi appointed Li Guangli as the Ershi General (104-101 BC) to lead military campaigns to establish a horse supply for China (Hulsewé MWR 1979: 135-6; Loewe 1979: 42-43). The four year-long campaigns were very costly and allegedly involved some 100,000 Chinese troops (Hulsewé MWR 1979: 135; Dubs 1955: vol. 2, Appendix V, 132-135). The costs of the campaigns were recouped by bringing the whole of the western side of the Tarim Basin and Ferghana Valley into a tributary relationship with China (Yu 1998:

de la Vaissiere (2005: 43) notes that Ancient Letter No. 2 from the Sogdian cache found by Stein in Niya, called the Xiongnu ‘Huns’. This letter is dated to 311 AD some 60 years before the Xiongnu ‘swept across the borders of the Roman Empire’ (de la Vaissiere 2005: 46). Kim, using Indian and Tibetan sources, has also confirmed the identity of the Xiongnu as Huns (Kim 2013: 28). 11  Recent work on nomads and pastoralism has rejected the idea that nomads/pastoralists were primitive, narrow specialists who had no knowledge of other forms of production. New understandings argue that nomadism and pastoralism typically co-exists with farming, craftsmanship, trade and the provision of a wide range of other goods and services: for a review of the literature see Vicziany et al. (forthcoming). Modelling Pastoralism’s Varied Characteristics. 12  For a meticulous critique that rejects the Xiongnu was a ‘barbaric’ people and culture see Kim 2013; see also Brosseder and Miller (eds) 2011 for a comprehensive review of the key debates and evidence about the Xiongnu and their empire. 13  The report was probably verbal rather than written down (Loewe 1979, 31). 14  This has led many scholars to assume that the Shiji is older than the Hanshu. The original texts were probably written on wooden slats that have not survived. What has survived are the oldest printed paper editions of the Shiji from the 12th century and the Hanshu from the 11th century (Loewe 1979: 11-12). A careful comparison of the two texts, together with a consideration of the conditions in which they were compiled and then transmitted over 1,000 years, led Loewe to conclude that the Hanshu was the more reliable of the two (Loewe 1979: 11-25). However, even the Hanshu cannot be taken literally given that its original compilation on wood occurred some ‘150 years after the events it describes’ i.e. between about 36 AD and 131 AD by the Ban family that was well connected to the Chinese court (Lowe 1979: 25-26). The above historiographic sketch is important because for all these reasons Hulsewé was compelled to re-translate the critical 10 

chapters in the Hanshu in order to promote a better understanding of the evolution of the relationship between the Chinese empire and the western regions beyond the Gansu corridor and beyond the mountains that divide western China from Central Asia in the west and South Asia to the south. The critical chapters are: chapter 96 in the Hanshu (the famous ‘Monograph on the Western Regions’) and chapter 61 which provides a biography of General Zhang Qian (d. 114 BC) and General Li Guangli (d. 88 BC). In the Shiji chapter 123 is virtually identical to chapter 61 in the Hanshu which, through the biography of General Zhang Qian, tells us a great deal about the western region. The puzzle about the relationship between chapters 96 (HS) and 123 (SS) is solved by Loewe and Hulsewé who argue that ‘… the Hanshu ch.61 is primary, and that Shih-chi ch.123 was practically lost, to be reconstructed out of Hanshu materials- chiefly Hanshu ch.61’ (Loewe 1979: 25). That reconstruction occurred sometime in the 3rd-4th centuries AD (Loewe 1979: 26). 15  As Hulsewé (MZL 1979: 209, note 774) notes, the dates for Zhang Qian’s travels are not certain but we can be more confident of the date on which he was given a special title by the Emperor when he returned from his mission in 123 BC. 16  In fact, the horses did not sweat blood but rather suffered from parasites that caused bleeding through the skin.

122

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record 82-83). Despite the preoccupation of these histories with the Chinese Emperor’s military campaigns, the overriding impression created of the Tarim Basin and further west is of lands of wealth and opportunity – the Memoir of the Western Region speaks of lands of the grapes and lucerne (alfalfa) of Da Yuan (Ta Yuan) in the Ferghana valley:

Despite these chronicles and the opinions of various scholars, we need not jump to the conclusion that Chinese power was centralised, predictable or consistent. The larger any pre-modern empire became, the greater the problem of ‘control’ and the more likely were periods of strong influence interspersed with weak influence. Closer to the Chinese capital, more centralisation of power was possible although the conditions of the Tarim Basin ensured that there was until modern times a considerable gap between Chinese imperial designs and what actually happened on the ground. By 76 AD the entire Tarim Basin is said to have come under Chinese imperial control as a result of General Ban Chao’s campaigns against local rulers, in particular the King of Khotan (Stein 1907: Vol. I, 53). The General eventually established his headquarters in Kucha and received the title of Protector. After that he became the Marquis of Western China in 91 AD. He may have represented the Chinese emperor but the reality is likely to have been one in which General Ban Chao acted pretty much as an independent operator out of Kucha. Evidence of the instability of the Chinese presence can be seen in the reassertion of imperial power by General Ban Chao’s youngest son, Ban Yong, in about 123-127 AD, when Kashgar together with other towns in the Tarim basin were again placed under imperial control (Hitch 1988: 171). Ban Chao also had to contend with the expanding influence of the Kushan Empire further west and south.

The general custom is to enjoy wine and the horses enjoy lucerne (Hulsewé MWR 1979: 135). They also grew wheat and rice (Hulsewé MZL 1979: 210 note 776). The capital of Da Yuan (Ta Yuan) had a population of 300,000 people of whom 60,000 were ‘able to bear arms’, while the countryside had some 70 towns (Hulsewé MWR 1979: 131-2). Interspersed with the towns of the west of China and the extreme west beyond the mountains were barren lands, pastoral communities and remnants of various tribes. Kashgar is specifically mentioned as the state of Shule with ‘1,500 households, 18,647 individuals, 2,000 persons ‘able to bear arms’ and ‘markets with stalls’ (Hulsewé MWR 1979: 141-2). The frequent references to armed men, courtly intrigues in the various capitals, and Han forces contesting local powers may give the impression of these regions in a constant state of war. In fact, we cannot assume this. Typically, the Chinese annals exaggerate the size of armies and the nature of conflict was not of the modern sort in which civilian casualties have been enormous. For this reason, it becomes possible to imagine the long term growth of oasis towns such as Kashgar despite the ongoing contests to control them.

The Chronicle on the Western Regions in the Hou Hanshu includes a section on Kashgar (Hill 2009: 4344). It is reported as ‘controlling’ 21,000 households (i.e. a population of more than 100,000 assuming 5 family members per household) and 30,000 soldiers. This suggests that Kasghar had grown greatly since the reports in the Hanshu when it only had ‘1,500 households, 18,647 individuals, 2,000 persons ‘able to bear arms’ (Hulsewé MWR 1979: 141-2). According to the Hou Hanshu Kashgar was, in other words, a good sized oasis town by the early part of the 2nd century AD but still only about a tenth the size of Imperial Rome which in the early 1st century AD was possibly the world’s largest city with an estimated population of about one million (Morley 1996: 35).

The next Chinese document that mentions Kashgar is the Hou Hanshu or History of the Later Han Dynasty, written by Fan Ye (d. 445 AD). It includes a section called ‘Chronicle of the Western Regions’ based on the expeditions and military campaigns of General Ban Chao (d. 102 AD). The General’s youngest son Ban Yong (d. 128 AD) wrote a report about his father’s expeditions in c. 125 AD (de la Vaissiere 2005: 38, note 73). The Western Chronicle covers the period from about 25 AD to c. 125 AD (Hill 2009: vi) and refers to Kashgar as one of the 48 districts of the ‘Western Regions’ (Hill 2009: 43-44). The chronicle describes the social, economic and political history of the countries and kingdoms adjacent to Kashgar on its western, northern and eastern sides. From these accounts, we can envisage a regional settlement pattern showing town and farming communities interspersed with nomadic and seminomadic peoples and how, in the case of the Yuezhi, the nomadic people quickly settled into sedentary ways of living and even developed independent and powerful kingdoms such as the Kushan (Hill 2009: 29). General Ban Chao’s extraordinary military campaigns saw the expansion of Chinese ‘power’ far beyond western China right up to the borders of the Parthian Empire (Mallory and Mair 2000: 69).

By the time of General Ban Chao’s campaigns the Yuezhi had created the Kushan Empire, c. 50-268 AD (Neelis 2011: 144), one of the most sophisticated political and cultural formations of its era. The Kushans recognised the importance of Kashgar for its gatekeeper role in accessing China and its highly desirable silks. Now it was their turn to seek alliances with the Chinese emperor and they hoped to do this through marriage. When the proposed alliance failed, 70,000 Kushan soldiers attacked Kashgar in 87-91 AD (Hitch 1988: 123

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads 183)17 and began a systematic campaign against General Ban Chao throughout the Tarim basin. Some 20 years later, between 114 AD and 119 AD the Kushans placed their preferred man into Kashgar and declared him to be king (Stein 1907: Vol. I, 55; Brough 1965: 589). This happened when the King of Kashgar, An-Guo, died and his mother placed his nephew on the throne. But the Kushans preferred the former King’s uncle, Chenpan, who had been sent earlier on by An-Guo as a hostage to the Kushan King. The hostage then became the new King of Kashgar with Kushan military support.18 This account in the Hou Hanshu tells us that both the old and new kings of Kashgar were local people, not Kushan functionaries. Such evidence supports Brough’s argument that the Kushans did not exercise any direct power in the Tarim Basin at this time (Brough 1965: 589), preferring in this instance to rule via a local proxy or puppet. On the other hand, by bringing a massive army into the Tarim to push out the Chinese, the Kushans demonstrated their high regard for the lucrative trade that crisscrossed this region (Hitch 1988: 183). We also suggest, in the next chapter, that Kushan patronage of Buddhism might have promoted this new religion in the Kashgar region via the installation of Chenpan as the King of Kashgar.

Sogdian monks, such as Kang and An, who belonged to émigré families and used the Sogdian community trade networks that were to be found all along the southern and northern trade routes of the Tarim Basin up to Gansu (de la Vaissiere 2005: 78-79 and Map 3, 59). According to Stein (1907: Vol. I, 56-57), the Theravāda Buddhist tradition dominated in Kashgar as it did in Bactria while further to the east, in Khotan, the Mahāyāna tradition was practiced. As we suggest in Di Castro, Vicziany and Zhu (in this volume), the prosperity of Kashgar and more generally the Tarim Basin, resulted in a generous flow of donations and labour services for the construction of important religious monuments from local people, passing merchants and the entrepreneurial activities of the monks and nuns. While the importance of trade is not disputed, some scholars argue that the Kushans did ultimately assert direct control of the southern parts of the Tarim Basin. Brough, for example, has analysed the ‘Kharoshti documents’ found by Stein at Niya. Kharoshti was the language of the Kushans but appears to have been adopted in the kingdom of Shan-shan that stretched from Niya to the town of Loulan. However, a careful study of the names of the Kings of the Shan-shan dynasty revealed that they were neither Indian nor Iranian. On the other hand, an analysis of the titles that the kings used were shown to be imperial designations typical of Kushan administration. However, we believe that Brough (1965: 596-598) goes too far in suggesting from this evidence that the Kushans incorporated parts of the southern Tarim in their empire from about the mid-2nd to 3rd centuries. After that, Brough (1965: 600601) suggests that Shan-shan ‘submitted to China’.

The numismatic records for the southern Tarim Basin region also support the idea that Kushan influence was important (Hitch 1988: 185-187); Kushan technology may also have been introduced into Niya and elsewhere (see Di Castro, Vicziany and Zhu in this volume). Despite this evidence, Kushan control remained indirect and required, as was typical of the kingdoms of the Tarim, forging alliances with any potential threat (Neelis 2011: 144 and note 242). More importantly, Kushan power depended on encouraging luxury trade and it was the success of this trade that then generated the economic surpluses that were probably used to patronise the Buddhist orders that were building stupas and managing monasteries. As Neelis (2011: 144) argues, the Kushan kings have been praised for their role in giving donations for these construction projects, but in reality local communities and governments probably played a more important role in financing the early Buddhist monasteries (see Di Castro, Vicziany and Zhu in this volume).

What does it mean when we say that Shan-shan was reabsorbed into the Chinese Empire after 265 AD (Brough 1965: 602)? As Brough (1965: 603) himself notes, the Chinese, like the Kushans, wanted the benefits of trade and for the local people ‘there is no indication that a change of dynasty made any difference to [them]’. In like fashion, the assertion of Kushan power probably had only a marginal influence on local communities. Hitch’s argument about the fluctuating influences of the Kushan and the Han supports Brough’s perception that regime change had little direct impact on the local populations. In other words, the so called ‘conquests’ of emperors have typically been exaggerated: for example, when General Ban Chao ‘conquered’ the Tarim Basin he probably did so not with a vast military force but with only 30 men. The General, moreover, acted more like an envoy. Similarly, the Kushans would not have required many generals and vast armies (Hitch 1988: 180). These interpretations of the nature of political power and control over the Tarim Basin emphasise the small movements that were needed to tip the balance. In 106 AD the Chinese Protector General Ren Zhang was driven from Kashgar by the local populace influenced, Hitch

Nevertheless, it was during the time of the Kushans (1st – 3rd century AD) that monastic Buddhism established itself in Kashgar. The agents of transmission included Hitch notes (1988: 183) that records say that this army represented about 70 per cent of the total Kushan military forces. He too puzzles about how this vast army, its families and retainers, could support themselves in the arid Tarim Basin but he does not reject the figures. We suggest that the number is greatly exaggerated and that Hitch was concerned for good reasons. 18  On the royal hostage from Kashgar see also Zürcher 1968: 353, 369; Narain 1968: 216, 218. On the monastery where the royal hostage was detained see Meunié 1942; Meunié 1943-45: 151-162; Bussagli 1968: 42. 17 

124

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record argues, by the spreading presence of the Kushans up to Kucha. No dramatic, violent rebellion, as chronicled by Chinese historians in their stories of the 105 AD revolt, was needed to make this happen: instead the revolt was a ‘gradual turning away from Chinese authority orchestrated by the Kushans’ (Hitch 1988: 181). We find Hitch’s explanation to be a plausible one.

to collect taxes while ‘ruling’ indirectly through the existing courts (Stein 1907: Vol. I, 58) and a system of vassalage and suzerainty. The complexity of Kashgar’s political economy emerges more clearly when we get information about the activities of the Persian speaking Sogdian merchants who not merely travelled between China and Central Asia but also established themselves as trading communities on the numerous branches of the Silk Road along which commodity exchange occurred. Sogdians followed many religions, including Zoroastrianism, Manichaeism and Buddhism, and as such they blended in well with the syncretic cultural mix they would have found in the oasis towns of the Tarim. Millward (2007: 29) writes that by the early 4th century AD the Sogdians had already set up a ‘network […] across the Tarim Basin and Gansu corridor’. Etienne de la Vaissiere’s analysis of Sogdian and other sources provides the evidence for this conclusion. His chapters on the Sogdian ‘networks’ document how long distance trade was regular and systematic (de la Vaissiere 2005, chapters 1 and 2). This network clearly had a much longer gestation, and presents us with a parallel narrative about the importance of the commercial interactions that we suggested at the start of this chapter. We can surmise that the Sogdians were already very active at the time of the Kushans. They continued to provide a critical social and economic network for international commerce until about 755 AD. According to de la Vaissiere (2005: 5) by the 10th century AD the characteristic social and economic features of Sogdian commerce had disappeared. In other words, local kingdoms rose and fell and great powers arrived and sought to extract their share of wealth by controlling the Tarim Basin, but for over 500 years the lives of ordinary people were probably more influenced by the Sogdian traders than any other factor. In its early days, the emerging Sogdian trade network overlapped with the flourishing Buddhist culture of Kashgar, contributing to the wealth required to support large monastic and agrarian communities (see Di Castro, Vicziany and Zhu in the next chapter). Inevitably Sogdians also entered into the service of the Chinese emperor as soldiers and administrators. Perhaps the most famous of these was the SogdianTurk General An Lushan who rebelled against the Chinese emperor in the heart of the Tang dynasty’s territory (i.e. north eastern China) between 755 and 763 AD (Millward 2007: 36-37). de la Vaissiere insists that An Lushan’s revolt was a ‘Sogdian movement, supported by Sogdian traders’ (quote in Millward 2007: 36) 21 and although the Tang dynasty survived, it no longer had the power to claim control of the

Throughout this time, Kashgar’s size and position made it into a highly prized possession as the many struggles over its control show: the Chronicle on the Western Regions speaks of the Yuezhi making persistent attempts to control the King of Kashgar (Hill 2009: 43). When Chinese power was dominant (c. 126 AD) the people of Yarkand abandoned their support of Khotan and switched their allegiances to Kashgar with the result that the Kashgar oasis ‘became powerful and a rival to Qiuci (Kucha) and Yutian (Khotan)’ (Hill 2009: 43). But when the threat of the Kushans receded, the royal families of Kashgar fell to fighting each other, and despite the attempt by the Chinese Emperor to subdue the town with 30,000 troops in c. 170 AD, the Kashgar oasis remained in turmoil from the viewpoint of the ambitious Chinese Emperor (Hill 2009: 45). The renowned stele of Cao Quan documents how the king of Kashgar, Hede, was punished (in 169-170 AD) for failing to pay tribute to the Chinese emperor. Confronted by Cao Quan’s troops ‘Hede’s face turned white, he retreated and died’ (Watt et al. 2004: 121). This historical narrative of the political turmoil in the Kashgar oasis needs to be treated with great caution. For one thing, there is no record suggesting that the population of Kashgar collapsed as a result of conflict. The population of this oasis town in c. 200 AD appears to have remained at about 21,000 families and 20,000 soldiers (Mallory and Mair 2000: 70). Second, we have evidence to suggest that even during periods of Kushan hegemony, Kashgar continued to be ruled by local Kings. In the mid-5th century the King of Kashgar chose to send gifts to the Emperor of the Northern Wei, Wencheng Di (452-466 AD),19 including an incombustible cloth allegedly worn by Buddha (Stein 1907: Vol. I, 58; Puri 1987: 47). After the Hephthalite conquest (late 5th – early 6th century AD), the Kingdom of Kashgar regained, to a certain extent, independent sovereignty under the western Turks.20 The latter invaded western China in the middle of the 6th century AD but preferred This occurred during the period of the so-called ‘Southern and Northern dynasties’ (Nanbei chao) (420-589 AD) when Chinese power was characterized by political fragmentation up to the time of the Sui Dynasy in the late 6th century AD. 20  During the 5th century the Hephthalites migrated to the western regions and settled in the Tarim, Bactria and Sogdia: see for instance: Litvinsky 1996: 135-144; Beckwith 2009: 102, 116 and 406, note 56. On the Hephthalites in the Kashgar oasis see: Sinor 1990: 299; Litvinsky 1996: 141. On the Turkic Empire see for instance: Sinor 1990; Sinor and Klyashtorny 1996; Beckwith 2009: 112-139. Relevant information and data can also be found in the older works of Chavannes (1903) and Grousset (1941: 103-161). 19 

In a personal communication (February 2013) Professor Emeritus Colin Mackerras disagreed with this interpretation of the An Lushan rebellion, the scale of which goes well beyond any conceivable support that the Sogdian traders may have provided.

21 

125

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Tarim Basin. The Sogdian revolt was put down with the help of the Uyghur empire which, Drompp notes (2005: 25), changed the dynamics between the Chinese and Uyghur states in favour of the Uyghur extracting greater quantities of silk and more Chinese princesses.

… as populated by ‘loose-rein’ peoples who maintained much of their traditional way of life and traditional leadership despite being part of the Tang administrative structure… In other words, we imagine the frontier zones as being far less under central control and far less destructively rebellious than the official records and some contemporary historians assume. We also support Drompp’s (2005: 13) warnings against too literal an interpretation of Chinese imperial sources that tend to portray the nomadic populations on the frontiers as scheming barbarians focussed only on raids as a way of acquiring wealth.

After the Sogdian revolt many different powers wrestled with each other to ‘control’ or establish the right to extract tribute from the oasis towns of the Tarim, including the Tibetans. But this did not mean the inevitable collapse of the oasis towns or the destruction of trade. As Millward (2007: 39) notes, the Sogdians cooperated with a wide range of contesting interests, thereby ensuring their own prosperity and the ongoing flow of goods, ideas, and people ‘from Ferghana to Manchuria’. We return to the question of international trade in the second part of this chapter when we consider the kind of economic circumstances needed to sustain the emergence of Kashgar as a Buddhist and then Islamic cultural hub.

The records of the Tang dynasty also tell us about the emergence of a Uyghur Empire in central Mongolia in about 744 AD. The Uyghurs were skilled traders and exchanged their horses for vast quantities of Chinese silk. As Drompp (2005: 24) notes, in this regard the Uyghurs inherited a tradition of exchange that began with Xiongnu-Chinese relations. When rebels challenged the power of the Chinese emperor, it was the Uyghurs who came to the rescue. They were the critical factor in putting down the An Lush an rebellion, a paradoxical situation because the Uyghur Empire was modelled on Sogdian ideas as much as the An Lushan revolt was a revolt of Sogdian soldiers, administrators and traders (Millward 2007: 45). The Uyghur reward for assisting the Chinese emperor was more silk in exchange for horses (Drompp 2005: 79). The story of the Uyghur Empire is important to the history of the Kashgar oasis because when the Uyghur Empire started to fall apart about a hundred years after its rise (c. 850 AD), the Uyghurs were compelled to scatter far and wide. Many of them migrated towards the south and western most parts of the Tarim Basin. On their arrival in these oases, including Kashgar, the Uyghurs were still mainly Manichaeans, for Manichaeism had been the state religion of the Uyghur Empire since their conversion in 763 AD (Mackerras 1990: 329-335).23

The Chinese records of the Tang dynasty The Chinese records of the Tang dynasty, as translated and analysed by the renowned French scholar Édouard Chavannes (1903), reveal the manner in which the Chinese emperor re-established control of western China and beyond, up to Kabul and the borders of Persia, by about 658 AD (Stein 1907: Vol. I, 59). Kucha reemerged as the headquarters of the Chinese presence in western China, a region now known as the ‘Four Garrisons’ – it included Kashgar. As Stein stresses, the naming of Kashgar in the administrative reorganisation of western China under the Tang dynasty shows the persistent value of this border settlement to the Chinese (Stein 1907: Vol. I, 60). The resistance of the Western Turks to Chinese power also persisted the longest in Kashgar and Khotan, taking Imperial China another two years to put down local rebellions. Direct rule by China of the Kashgar region was a recipe for ongoing rebellions, in contrast to the preceding years of indirect rule when the King of Kashgar sent at least two envoys to the Chinese Emperor bearing tribute to Chinese power as long as it remained remote.22 From the mid-7th century until the early 8th century AD (with interludes) the Tang dynasty’s imperial armies ruled western China by successfully fending off local rebellions and incursions by Tibetan forces. The above narrative is typical of many accounts of the Tang dynasty’s relationship with the periphery i.e. the borderlands including western China. The preoccupation with rebellions seems to inevitably stress the assumed centralised nature of Chinese imperial power. However, we incline to the more subtle interpretation of Drompp (2005: 161) who described the frontier areas

The Chinese accounts about the events in the Tarim Basin from about 200 BC to 900 AD confirm western observations about Kashgar, but in much greater detail because the stakes were higher: he who controlled Kashgar controlled one of the critical gateways between the ‘east’ and the ‘west’, a transition zone in which the lucrative nature of international trade greatly influenced political, diplomatic and strategic interests in this borderland. We also learn that irrespective of the dynasties and tribes that ruled Kashgar, a kingship system had been in place for a long time. The hierarchical nature of Kashgar society is suggestive of 23  Today, many Uyghurs express surprise on learning about their community’s belief systems in the days before they converted to Islam.

The Kashgar embassies of 635 AD and 639 AD are described by Stein: 1907, Vol. I, 61. 22 

126

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record a social inequality that had important implications for the control of land and water resources in this oasis.

the archaeological evidence suggests (see Di Castro, Vicziany and Zhu in the next chapter). The arrival of Islam built on this cultural tradition, with Kashgar emerging by the early 11th century as a major centre of Muslim learning.

The Persian records The first Persian reference to Kashgar appears on the famous inscription of the Sasanian Emperor Shahpur I on the Ka’bah of Zoroaster at Naqsh-i Rustam. Amongst other things the inscription lists all the lands that fell under the control of the Persian Empire in 260 AD and the borders of those territories. Kashgar is named in this inscription although as Frye states, there is no evidence to suggest that the Sasanian armies ever ventured that far east. Rather, as Shahpur I conquered more and more lands, the Kushan Empire that was already weak gave up claims to its former territories. Hence Frye (n.d.) concludes that the reference to Kashgar in this inscription gives ‘… the actual or the pretended extent of the Kushan empire up to the borders of the state of Kashgar which was more or less restricted to eastern Turkestan’.

Kashgar as an Important Cultural Centre Kashgar emerged as a vibrant cultural centre during the Buddhist era and remained so until the arrival of Islam via the overland route in the 10th century, some three hundred years after Islam reached eastern and central China.24 The cultural vitality of Kashgar depended on the wealth generated by regional and international trade. When the Sogdian traders eventually disappeared by the 10th century, they were replaced by other actors. We disagree with Lattimore’s view that international trade was all about tribute and luxury goods exchanged between embassies and powerful rulers and ‘had very little to do with the subject peoples of the oases’ (Lattimore 1940: 175). Those ‘subject peoples’ produced the food, clothing, slaves and military conscripts that supported commerce. City states like Kashgar were the very basis of sustainable, long distance regional and international trade.

The Sasanians rekindled the religion of the Achaemenids, namely Zoroastrianism, and the social hierarchies that it preferred, giving priests and the extended Iranian family special authority in all civilian matters (Frye, n.d.). Under the leadership of Kartir, a religious purist, all religions other than Zoroastrianism were declared to be heresies. The new puritanical tone can be seen in the inscriptions on the Ka’bah of Zoroaster. Despite this, other religions (especially Christianity) persisted in Sasanian dominions; Kashgar was less affected by the new fanaticism because it was not directly ruled by the Sasanians. The long history of religious syncretism in Kashgar continued, with many religions coming together in the one location. Indeed, this era most likely coincided with the evolution of a strong Buddhist presence in Kashgar, although we await the dating of the Buddhist monuments to establish this with certainty.

Buddhist evidence Kumārajīva (344-413 AD) arrived in Kashgar on his return voyage from Jibin (probably in the region of Gandhara-Kashmir)25 to Kucha in c. 355 AD. He had left Kucha at the age of nine to study Buddhism in Kashmir and after three years in Jibin he reached Kashgar aged 12 and stayed for a year with his mother who had accompanied him on his travels, after adopting the life of a Buddhist nun. By origin, she was a Kuchean princess who had married a Kashmiri Brahmin serving in the court of the King of Kucha. At the time of Kumārajīva’s visit to Kashgar, the town must have been seething with religious controversy for it was here that Kumārajīva learnt about the Mahāyāna stream of Buddhist thought thanks to the teachings of Sūryasoma. Sūryasoma was from Yarkand but living in Kashgar during Kumārajīva’s visit. Under his influence Kumārajīva abandoned the Theravāda Buddhist tradition that dominated the religious life of Kashgar at this time and began following the Mahāyāna tradition (Rhie 2002: Vol. 2, 387-402; Dutt 1988: 303-4).26 The religious fervor supports the view

The old western, Chinese and Persian references to Kashgar, spanning the last two hundred years before the common era up to the early 9th century AD all have three things in common: they stress the strategic importance of Kashgar as an entry point into and out of China to the West; they describe the unstable political history of the region to the north, east and west of Kashgar; and they speak of the economic significance of the Kashgar oasis as a trading emporium on the Silk Road. As a long term garrison town, irrespective of the rulers that were imposed by the authorities that dominated the locality, Kashgar had a civilian population accompanied by a considerable retinue of soldiers. It is not until the 4th century AD that we start to hear about Kashgar as a cultural centre thanks to the travels of Kumārajīva and other Chinese Buddhist pilgrims. The flowering of this cultural life, however, had taken longer to develop as

Gernet (1962: 77, 82) notes that before the rise of Hangchow as China’s main international and domestic commercial/trading centre during the southern Sung dynasty, Canton played that role in the early 9th century with perhaps half of the merchant community being Muslims from the Gulf, Middle East and Central Asia. By the time of the Mongol invasion of 1276, Hangchow also had a considerable community of foreign merchants including Muslim traders from Central Asia. 25  On Jibin see Petech 1950: 63-80. 26  On his return to Kucha, Kumārajīva was ordained and slowly matured into a major Buddhist scholar. His fame resulted in him being captured by the Chinese who took control of Kucha in 384 AD. 24 

127

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads that monastic Buddhism had firmly established itself in Kashgar by the early 4th century AD, with religious orders replacing merchants as the purveyors of new ideas.

a low opinion of the residents of Kashgar whom they continue to regard as crude despite their attachment to Islam. The quality of Buddhist learning was also primitive according to Xuan Zang. It occurred in ‘several hundreds of Sangharamas, with some 10,000 followers… [but] without understanding the principles, they recite many religious chants’ (Beal 1983: II, 307). Despite these negative judgments Xuan Zang came to depend on Kashgar’s depository of Buddhist manuscripts when he lost some vital documents and ordered copies to be made for him (Stein 1907: Vol. I, 71). Xuan Zang’s report of such a large number of Buddhist believers and his reference to the oasis’s specialist records affirms the vibrancy of Kashgar as a religious centre. That reputation rested on the capacity of the oasis to support so many monks and nuns in addition to kings, nobles, soldiers and landed gentry.

About 50 years after Kumārajīva, between 399 and 414 AD, Faxian and his companions were travelling to India from Khotan when they decided to detour to Kashgar in 402 AD.27 At this time, Kashgar was probably at the height of its importance to international Buddhism (Stein 1907: Vol. I, 67-68), as suggested by the various Buddhist relics that attracted the comments of pilgrims. Faxian described Kashgar as a land of wheat, high quality woollen and silk fabrics and Buddhism. Each day the monks prayed for the wheat to ripen. Faxian was shown a widely venerated relic – Buddha’s stone spittoon. He did not mention Buddha’s alms bowl which Kumārajīva referred to but as Stein (1907: Vol. I, 68) notes, that is because Faxian had seen a competing alms bowl in Peshawar and obviously decided that the one in Kashgar was not authentic.28 Another relic kept in Kashgar was one of Buddha’s teeth around which a stupa had been built. It was the focus of the lives of ‘more than a thousand monks and their disciples, all students of the Hinayana’ (Faxian cited by Stein 1907: Vol. I, 69). The cult of relics in Kashgar suggests that the oasis had expanded its reputation as an international centre for Buddhist thought and practice and attracted pilgrims from distant lands by the late 4th century AD.

The beginning of the decline of Chinese domination over Kashgar and western China more generally started in 715 AD when the Muslim General Qutayba allegedly sacked Kashgar. The following period was one of renewed political contests to control the Tarim Basin and the surrounding areas between the Chinese Emperor, the Tibetans, various Turkic raiders and local rulers who were constantly clamouring to maintain or reassert their independence (Stein 1907: Vol. I, 62-65). Kucha remained the main symbol of Chinese power in the Tarim Basin, even though it was cut off from central China for many decades. The Chinese Emperor also recognised that the king of Kashgar deserved more attention than most local potentates and so in 728 AD conferred a royal title on him (Stein 1907: Vol. I, 66). Despite this, Chinese power was receding (and nonexistent after the mid-8th century AD) and in 941 AD the ruler of Kashgar, Satuq Boghra Khan, converted to Islam and in so doing brought the history of Kashgar into the Islamic period. The arrival of Islam in Kashgar roughly coincided with the final disappearance of the Sogdian commercial networks that had propelled long distance, overland international trade for so many centuries. What replaced the Sogdians? And how did the new commercial arrangements interact with the new Islamic societies and polities that emerged? It is this difficult question that we consider next, before examining the Islamic conversions.

Another Chinese pilgrim who left an account of Kashgar in c.644-645 AD was Xuan Zang. Kashgar was reported by him to be a typical oasis town with plenty of sand interspersed with orchards. But it was uncivilized because the residents were ‘fierce and impetuous […] mostly false and deceitful’ (Beal 1983: II, 306).29 Xuan Zang may well have started the negative stereotype of Kashgar’s inhabitants, as this judgment was repeated by others.30 In 2003 during fieldwork in Xinjiang we learned that many Uyghurs of Urumqi and Ili still have For the rest of his life he lived in the Chinese capital of Chang’an where he was given a vast establishment to translate Buddhist texts from Sanskrit and also correct old, erroneous translations. 27  The controversy regarding the visit of Faxian to Kashgar is still unsettled and goes back to the 19th century. For instance Legge (1886: 18, 22) was apparently following the interpretations of Klaproth and Watters who believed that Faxien went to Kashmir or Ladakh. On the other hand Stein (1907: Vol. I, 67-69) seems to agree with Beal (1869: 15-17; 1983, xxviii), for both said Faxien detoured to Kashgar because an important religious festival was going to be held there. 28  This is an odd explanation as most religions, including Buddhism, are replete with competing ‘authentic relics’. 29  Beal (1983: II, 306) transcribed Kashgar as Kie-Sha. 30  Perhaps he was repeating Chinese stereotypes about the Hephthalites perceived as barbaric nomads. At the time of his visit, the Hephthalites dominated the region. Xuan Zang says that the people of Kie-Sha (Kashgar) compressed the head of new born babies with a wooden board (Beal 1983: II, 306). This practice of deforming the skull was common amongst the Hephthalites and other Turkic populations. Some Uyghur families still do this to their babies today (see also Macartney 1985: 124, who witnessed this practice in the first half of the 20th century).

The great gap in Western knowledge about Kashgar’s history Why is our knowledge of Kashgar’s international trade after the arrival of Islam so limited, especially during Ming times (1368-1644)? Is it that the overland trade route declined because of Central Asian political instability and the emergence of cheaper and safer sea trade as suggested by Rossabi (1990: 351, 367, 369-370)? Lattimore also speaks of that decline when he writes about the ‘new age of maritime power’ which came 128

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record with revolutionary social changes associated with the ‘origins, rise and triumph of capitalism’ (Lattimore 1940: 5). But there are alternative views to these. We tend to side with Kwangmin Kim who stresses the historical continuities between the pre-Islamic and Islamic period suggesting that neither of these factors prevented overland commerce (Kim 2008: 22-23). Did the Ming abandon overland trade for seaborne commerce as suggested by the Imperial voyages of Admiral Zheng He (1371-1433)? Dardess suggested not, for these expeditions, despite their enormous size, were not commercial ventures but imperial displays of power (Dardess 2012). On the other hand, Dardess argues that Ming indifference to Central Asia was the result of an obsession with the northern regions and Japan (Dardess 2012: 1-24). Kwangmin Kim disputes this, arguing convincingly that the Ming began to build up a tribute system with western China and Central Asia – the Qing inherited this legacy and perfected it. Kim’s perspective opens up the possibility that the international trade of Kashgar and other towns in Xinjiang became increasingly oriented towards Muslim commercial centres in Central Asia and the Middle East during the Islamic period. This is an important view of the long term sustainability of the western trading systems of China and Xinjiang.

Suzhou with any international, overland trade that may have involved Kashgar. The great gap in our knowledge remains, therefore, for the long years of the Ming dynasty (1368-1644) and only gradually reduces as we approach the Qing era (1644-1912). There is an even greater gap for the five hundred years between 900 AD and 1368 AD. These three scholars, nevertheless, provide invaluable insights into the commerce of western China and help us to understand what kind of networks probably survived and promoted ongoing regional and international trade after the Sogdian decline. The portrait of trade by Kwangmin Kim (2008) complements that by Millward (1998). Kim focuses on the way in which Muslim merchant-saints collaborated with the Qing dynasty and Millward stresses the role of Chinese migrant merchants who were encouraged by the Chinese emperor to settle in western China. Kim’s vivid account of trade in the far south west of the Tarim under Khoja Emin stresses the wealth of the caravan merchants who sided with the Qing dynasty despite their earlier opposition to it. Khoja Emin was a Muslim merchant from the Turfan basin who had been appointed by the Qing emperor as the Governor of Altishah, the region between Kashgar and Khotan, in I759. He was one of the many Muslim noblemen who collaborated with the Chinese in establishing a new commercial empire in the far west of China and beyond in Central Asia. He became the highest ranking Muslim in the Chinese imperial system and was put in charge of local laws, tax collection and keeping his eye on rebellious Muslims such as the émigré Khojas (Kim 2008: 212-214). In promoting these many objectives, Khoja Emin depended on the loyalty of the caravan merchants called Bo-de-er-ge. They were Nashqibandi sufis from Central Asia, Bokhara and India, eager to continue to prosper from inter-regional trade, which included providing supplies to Chinese troops in western China (Kim 2008: 227). Local merchants called them Andijanians and they were the richest group of traders. Yarkand had some 200 and Kashgar some 170 Andijanian merchant households in 1764 (Kim 2008: 231-232). Not surprisingly, we learn from Kim that Khoja Emin behaved as if he was the actual local ruler of Yarkand-Kashgar and people believed him to be that, rather than the mere intermediary appointed by the Chinese Emperor. By acting in this superior manner, Khoja Emin was able to expand the number and range of Muslim collaborators with the Qing dynasty in western China and Central Asia, something that the Chinese were keen to do not only for reasons of power but as a way of protecting the international trade that yielded such lucrative customs duties.

Certainly accessing the archives of Central Asia, the Middle East and Russia, together with applying multidisciplinary and multi-lingual skills are very important as Subrahmanyam and Lewis have argued (Grillot and Etter 2012; Kramer 1999) but the Chinese records remain surprisingly rich when studied very closely as Kwangmin Kim (2008) and James Millward (2007) have done for the Qing dynasty period for the years 1696-1814 and 1759-1864 respectively. The work of Marmé (2005; 2007) also reminds us that the centre of economic growth in Ming and Qing times was already on the eastern seaboard rather than in the western regions. This was probably always so because western China’s fragile oasis economies could hardly support the intensive agricultural and industrial production that defined central and eastern China. While eastern Chinese cities were production centres, this does not rule out the role that the western, oasis based cities played in the long distance trade of those industrial goods. By piecing together the insights that Kwangmin Kim, Millward and Marmé have about the nature of Chinese commerce, we might be able to say something about what replaced Sogdian commercial networks after 900 AD and specifically how the commerce of Kashgar may have fared after the arrival of Islam. Unfortunately, all three studies still leave an enormous historical gap in our knowledge because Kwangmin Kim and Millward focus on the Qing period while Marmé’s work relates mainly to Ming times but does not link the growth of

Kwangmin Kim’s (2008) discussion of local and international trade stresses how even the worst political instability during late Ming/early Qing times on China’s 129

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads western and northern borders promoted trade in grain and horses (Kim 2008: 23-24). Trade was part of the Ming tribute system and it provided the foundations for the more commercially oriented economy that the Qing built. In developing that new mercantile empire, the Chinese emperor depended critically on a wide range of Muslim collaborators who had earlier opposed the ‘infidel’ Chinese. The Qing determination to protect traders and implement Islamic law to support the Muslim merchants was one indication of how important the Muslim traders remained.31 The focus of the Ming trading system on western horses for Chinese silk was largely replaced by the jade-silk trade (Kim 2008: 267). But hundreds of other items were also bought and sold over short and very long distances as China gradually became integrated into a newly emerging world economy based on silver exchanges.

The merchant activities that Millward and Kim describe represent the far western end of this extraordinary industrial-commercial activity on the eastern seaboard of China.33 In summary, we believe that there is enough evidence to suggest that Kashgar’s international trade orientation towards Central Asia and South Asia has been an enduring characteristic of that oasis town’s identity for some 2,000 years or more. The specific contours of that trade, the traded items and the traders may have changed but the importance of private enterprise never dulled. With few exceptions, the Emperors of China could not and did not ignore the trade of western China and beyond. Some Emperors were pro trade and others supported state monopolies but in both cases private trade in the very long run persisted with few exceptions even if that trade had to sit alongside state monopolies. That persistence guaranteed the long term economic, political and cultural importance of Kashgar. Similar statements can probably be made about the kings, sultans and emperors who ruled in Central Asia and South Asia. For these kingdoms on China’s borders, trade with the Tarim towns ultimately meant trading with one of the most productive economies in the world, namely central and eastern China. However, until the Islamic archives mentioned at the start of this section are studied more closely, we have only a sketchy knowledge of the inner commercial life of Kashgar. For example, no authors speak of the Kashgari merchants as international traders; the assumption is that they confined themselves to local trade alone. However, this contradicts the report by Shaw in 1871 of ‘Toorkees’ based in Leh, a sizeable town in the Ladakh region. The ‘Toorkees’ we can safely assume were Uyghurs although the British called them ‘Mughals’ (Shaw 1871: 11, 101 and note 1).34 Shaw (1871: 11) thought these traders to be escapees from the Qing conquest of Kashgar and was very impressed by their dignity:

Kim stresses the wealth of the international caravan traders while Millward speaks of the rising prosperity of the Chinese merchant residents of Xinjiang. Both are right in the sense that the greatest trading profits were not made by local Turkestan merchants. This reminds one of the Sogdian trading system where the greatest profits also went to the Sogdians who, after all, took the greatest risks in trading over very long distances and across many borders. Millward’s account also agrees with Kim’s in stressing the commercial benefits of the Islamic cultural networks that the Naqshbandiyya merchants used to promote commodity exchange between Central Asia and China (Millward 1998: 155). Both authors show how the international trade that passed through the Tarim oasis towns, including Kashgar, brought trading profits to a wide range of merchants even if they shared unequally in that prosperity. Millward (1998: 155-175) writes of rich Kokandi merchants with trade links to the far west, local ‘Turkistani’ merchants or ‘Bukharans’, newly arrived Chinese merchants, Tungan merchants and various Turkestan noblemen or ‘Begs’ who used their audiences with the Chinese emperor as trading opportunities. Driving much of the international overland trade was the rise of new industrial centres during Ming and Qing times, best exemplified by Marmé’s study of Suzhou.32

stronger than ever before as an industrial town and transformed in its commercial character to such a degree that Marmé stresses the discontinuities between the old and new Suzhou. The story of the rise of Suzhou is a story of commercial modernisation in the precolonial era, before European dominance takes over discourses about China’s history. Suzhou emerged largely thanks to the combined energetic output of sojourning merchant craftsmen who moved in from Fujian and Guangdong, women textile workers who belonged to thousands of rural and periurban, patri-corporations and hundreds of pawnbrokers who gave credit to farmers so they were not compelled to sell their food crops immediately after harvests when prices were low (Marmé 2007: 156-164, Table 5 and note 67). Suzhou became a vast storehouse of industrial goods which thousands of merchants then carried to international markets in Japan, southeast Asia, America and Europe (Marmé 2007: 164-5). 33  Marmé does not speak of that westward trade flow, preferring instead to concentrate on the new shipbuilding industry that developed in Suzhou and resulted in hundreds of Chinese made boats plying international trading routes even when the emperor banned private trade. 34  Shaw was the first Englishman to enter Kashgar. In his account of that city, he speaks of the local ‘Toorkees’ (Shaw 1871: 159-160).

Kwangmin Kim’s (2008: 282) portrait of Governor Odui, one of the successors to Khoja Emin, provides further insights into trade in the southern Tarim. Odui was in charge of the jade mines of the Chinese emperor, a monopoly systematically and routinely evaded by Chinese, local and international merchants who all worked together to maximize the profits for themselves in trading the jade produced in the state-controlled mines. 32  By 1684 the Chinese emperor himself noted that the people of Suzhou were mainly working in industry and trade and very few in agriculture (Marmé 2007: 144). Suzhou grew in response to local and inter-regional commercial demands, especially for cotton and silk textiles. In late Ming times it had half a million people and by about 1800 some 1.4 million, making it the largest industrial-commercialnon capital city in the world renowned for its high quality production and ostentatious lifestyles. The dynamism of Suzhou was irrepressible – not even the extreme destruction of much of its infrastructure during the Qing conquest held it back. It re-emerged relatively quickly, 31 

130

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record Their large white turbans, their beards, their long and ample outer robes, reaching nearly to the ground, and open in front showing a shorter undercoat girt at the waist, their heavy riding boots of black leather, all gave them an imposing aire; while their dignified manners… made them seem like men amongst monkeys compared with the people around them.

Qara-Khanids ‘ruled from the Oxus to the Tarim basin’ (Biran 2005), a regime that persisted until 1200 despite the many internal disputes and external threats that emerged from the decentralised nature of the dynasty.35 Decentralised rule facilitated an expansion of international trade along the Silk Road and promoted the growth of new and old towns and expanded industries including crafts, mines, irrigation and agriculture (Biran 2005). The usual exchange of Chinese products like silk for Central Asian goods such as glass and slaves continued right up to the lands controlled by Chinggis Khan in addition to the trade with nomads (Biran 2005). The Qara-Khanids were also great patrons of literature, art and architecture throughout Central Asia and Kashgar, promoting a Turkic interpretation of Islam.

With these words of caution, we will now focus on some of the key cultural developments in Kashgar during the Islamic period. Much of the existing knowledge is based on legends and anecdotes. These stories should not, however, be dismissed. Instead they can inform future researchers about the issues and sites worthy of deeper analysis if we are to establish the cultural significance of the Kashgar oasis through its Islamic prism. The Islamic period gave rise to many important public monuments, a new cultural vibrancy and a recognition of the underlying problems of survival in an oasis environment, namely water management. In this way, the Islamic era also represents strong continuities with the preoccupations of local people who lived during the Buddhist period.

The most powerful story in Kashgar tells the legend of how Satuq Boghra Khan converted to Islam. He was the first Kashgari and Turkic ruler to do this, despite his uncle’s opposition. The full story of that conversion was recorded by Jamal Qarshi in his Mulhaqat al-surah, written more than 600 years before Jarring’s visit to Kashgar. Qarshi was a resident of Kashgar and claims that his account is based on a lost book about the history of Kashgar called the Tarikh-i-Kashghar (Jamal Qarshi in Levi and Sela 2010: 73-74). One of Satuq’s uncles was the Governor of Artush (Artux), a small town north of Kashgar. He was a prominent merchant who looked after camel caravans entering Artush from Bokhara, an activity which compelled Satuq to watch his uncle’s interactions with the traders. His uncle was also a clandestine Muslim, and it was his influence that persuaded Satuq to convert to Islam, a process that required him to kill his other uncle who was the King of Kashgar and an opponent of Islam (Jamal Qarshi in Levi and Sela 2010: 75-76). Satuq then took control of Kashgar at which point – allegedly – some 200,000 families joined him in a mass conversion process (Grenard 1900: 43). He died in 955/956 AD and his tomb today can be found in Artush, north-west of Kashgar (Figure 11).

The Islamic records on Kashgar Western knowledge of the Islamic period begins with the religious legends as told to the European explorers. Jarring summarized the Islamisation of Kashgar in his account of the religious cult surrounding the Ordam Mazar between Kashgar and Yengisar when he followed the route that links up the religious shrines of the Kashgar oasis in 1929. Before Jarring, many Muslim sources also recorded these legends or ‘historical accounts’ of the Islamic conversion of the people of the Tarim Basin, but most of them are not known or accessible to western scholars. Thanks to the diligence of translators, an extract from the account by Jamal Qarshi in 1303 in his Mulhaqat al-surah can now be read. But we remain very much at the dawn of any serious study of Islamic texts that speak of Kashgar. The same applies to the Russian sources and archives that remain unexplored by both western and Chinese scholars. New work by Hayes (2012) on the legends of the Tarim Basin, however, heralds a much needed shift in western scholarship.

Satuq’s youngest daughter36 was unmarried and deeply involved in his conversion movements, especially as resistance to Islam mounted. Ala-nur Khanim, famously known throughout the Kashgar region as Bibi Maryam,

Satuq Boghra Khan (d. c. 955 AD), the iconic symbol of Islam in Xinjiang today, belonged to the Qara-Khanid dynasty, started by a confederation of Central Asian tribes whose history remains ‘extremely sketchy’ and ‘highly debated’ (Biran 2005). The frequently told legend of his conversion (see below) claims that he was the first Qara-Khanid to become a Muslim who then left the responsibility of mass conversions throughout Central Asia to his ancestors. Satuq’s legitimacy derived from the Abbasid Caliph in Baghdad but by 1010 the

35  For most of the 12th century the Qara-Khanids were in the orbit of the Qara-Kitai Empire (originally the Liao dynasty 907-1125 AD) that ruled Kashgar indirectly and left intact the essential features of local Qara-Khanid government. The Qara-Kitai, a nomadic and culturally tolerant regime, were greatly influenced by Chinese religious values as well as the Christian Nestorian ones and never assimilated into the urban Islamic societies or belief systems that they ruled (Biran 2005). The Liao are the only non-Chinese group to establish a dynasty regarded as ‘Chinese’ by official Han historians (Biran 2005: 1). On the Qara-Khanids (Karakhanids) and Qara-Kitai (Karakhitay) see also Davidovich 1997 and Sinor 1997. 36  The identity of Bibi Maryam is confused and many traditions coexist (Hayes 2012: 182-183).

131

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 12. Exterior of the tomb of Bibi Maryam with the female caretaker (photo Di Castro).

Figure 11. Tomb of Satuq Boghra Khan in Artush with a bronze cooking cauldron dedicated by Yakub Beg in the foreground (the cauldron was used to distribute food to pilgrims (photo Di Castro).

entered the battle and was also killed (Landsdell 1894: 75). The place where she died became a sacred Islamic site on which the shrine of Bibi Maryam sits today, rebuilt and enlarged by Yakub Beg who undertook extensive renovations in the 19th century (Bellew 1875: 324) (Figures 12, 13). The renovation process has been described by Hayes as a revived ‘ancestralism’ that helped Yakub Beg to legitimate his regime (Zarcone 1999; Hayes 2012: 190). Today the grave of Bibi Maryam is popular with women seeking remedies ranging from migraines to childlessness.

had a son called Ali Arslan Khan who became the first martyr to Islam. Bibi Maryam was a virgin when her son was born, the father allegedly being St Gabriel. The conception was inspired by the apparition of a tiger, so the son was named ‘The Lord Exalted Tiger’ (Bellew 1875: 334, Grenard 1900: 10-11). Observers have noted the parallels between this birth legend and various Christian and Turkic-Mongol traditions (Roux 1990: 221; Bellew 1875: 335-6). Arslan tried to convert the other nobles in Kashgar to Islam but when they resisted he took up arms and forced them to flee towards Yengishar and ultimately Khotan. At the head of the resistance were two Buddhist princes called Jukti Rashid and Nukti Rashid (Jarring 1935: 349; Shaw 1878: 109; Hayes 2012: 141, 168, 248).

The site of Arslan’s grave is the Ordam Mazar (graveyard) which remains one of the important Islamic sites in the Kashgar oasis attracting thousands of pilgrims who drown themselves in ecstasies of grief as they recite the legend of Arslan.38 Jarring was the first European scholar to describe the various beliefs and rituals surrounding the oases and desert shrines that form part of the Ordam Padshah system. In recent years, pilgrimage to this site has been banned by the authorities fearful of public disorder. When active, the site is well known for having the highest stack of wooden branches or rods (reaching 10 metres into the air in Jarring’s time) sitting on top of

Arslan died in this conflict but his surviving kin continued the war to assert the supremacy of Islam over Buddhism, fighting not only in the Kashgar region but also in neighbouring Bactria and Ferghana. Arslan was beheaded by his enemies and his head thrown to the dogs near the walls of Kashgar (Bellew1875: 335; Lansdell 1894: 74).37 His mother, Bibi Maryam, then

bodiless head continued to sit on the horse and slaughtered hundreds of Buddhists with the sword it clasped in its teeth (Hayes, 2012, 162163). 38  His body is buried here; his head is buried beneath the shrine of Hazrat Padshah (Bellew 1875: 310).

The special significance of the headless martyr has given rise to many different versions of Ali Arslan’s beheading: (a) his head was hung on the fort of old Kashgar, (b) on losing his head, Arslan’s horse killed hundreds of infidels by opening and shutting his eyes, (c) the 37 

132

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record were essential, of course, to refresh pilgrims who had travelled long distances. Water is equally important for ritual cleansing. And most fundamentally of all, water is a prerequisite for agriculture and horticulture. For generations, the Kashgar oasis was famous for its orchards (Polo 2016: 80). The need for good water management in the Kashgar oasis is something that we discuss elsewhere in this volume (see the next chapter by Di Castro, Vicziany and Zhu). Many different stories are told in Kashgar about the role that Bibi Maryam played in improving water availability: one version told to us by a local resident says that after being wounded by a spear she dragged the spear along the ground and caused a new channel to flow with water. Having done that she then died. Another version of this story is that on her death Bibi Maryam requested the construction of a canal linking the Beshkerem orchard to Artush. That canal, named after her, was built and later renovated by Yakub Beg. A related water story is that when Bibi Maryam visited her father’s grave in Artush she met an old man who caused the waters of the shrine to follow her when she returned to Beshkerem. This stream is said to be the origin of the Chakmak (Lightening) River, one of the life sources of the Kashgar oasis through its many incarnations. When the water level is low, locals say that is because she disobeyed the old man’s instructions by looking behind her (Hayes 2012: 190-191).40 To ensure that the spring waters needed for irrigation flowed into the canal, sacrifices of sheep and offerings of grains and oils used to be made at the tomb of Satuq Boghra Khan (Hayes 2012: 191; Figure 15).

Figure 13. Interior of the tomb of Bibi Maryam showing the tomb with ritual rods (photo Di Castro).

the sacred grave of Arslan. The rods are festooned with innumerable fluttering cloths designed to drive away evil spirits (Figure 14).39 It is significant that the festival celebrating the martyrdom of Arslan occurs during the Shia month of Muharram – providing further evidence of the incorporation of sectarian variety into the religious festivals of the Muslims of Kashgar.

The early history of Islam in Kashgar matches the long experience of that town with religious controversies. We believe that the stories about religious wars are probably based on historical facts, but that the level of violence has been exaggerated as part of the tradition of venerating cultural heroes. What is most interesting about these stories is that they verify the nature of the historical circumstances that have defined the life of Kashgar in the long run – the international caravan trade, desert battles and graveyards and a preoccupation with water. The stories of the conversions to Islam are especially powerful because a second conversion process occurred after the first. According to the legends, Hasan, the son of Satuq Boghra Khan, spent some ten years campaigning for Islam in Central Asia and Iraq and then returned to Kashgar to find that the local people had abandoned Islam and reinstated the Buddhist princes of Khotan as their rulers (Hayes 2012: 146).41 Hasan was eventually

Another important shrine that belonged to the Ordam Padshah system at the time of Jarring’s visit was Dost Bulaqim. Its special feature was a sacred pond or well said to have suddenly appeared when a disciple of Arslan solved the terrible drought that afflicted the local village. He threw his dagger into the ground causing water to rush out and form a pond (Jarring 1935: 352). Jarring noted that the other shrines in this system also had wells and madrassas (religious schools) and that the Ordam mazar in particular had three wells in addition to the madrassa and population of 400 living in the adjoining village (Jarring 1935: 353). Water wells Information about the tomb of Arslan Khan can be found in Xinjiang Kaogu Yanjiusuo 1995: 584-587. In 2005, 2007 and 2013 visits to the Ordam mazar were prohibited by the Xinjiang Government. But the site of the Laddered Tombs near Yopurga has the same rituals and had an active holy grave culture when we visited in 2005. The ladders standing before the graves are made from the trunks and branches of poplar which are used to transport the deceased to the grave site. Sometimes the wooden poles are first taken to the shrine of Bibi Maryam to be blessed and then carried in procession to other mazars. For images of the Yopurga cemetery see Michell et al. 2008: 35-41.

39 

There is another legend about the seven maids who responded to Bibi Maryam’s orders and ended up as martyrs for Islam. Shrines to them are located upstream along the Kushan River in Yengisar (Kashgar Team fieldwork April 2005; Hayes 2012: 192). 41  Mention of the conflicts between Khotan and Kashgar are also recorded in Chinese annals and in the Dunhuang documents. These 40 

133

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 14. Desert Shrine of Ordam Padshah in 1928. This is the grave (mazar) of Ali Arslan Khan (Frank Ludlow Collection, reproduced with the permission of the British Library).

With the arrival of Islam, Kashgar’s status as a cultural centre was elevated into that of a leading town in western China, as testified by Arabic histories of Kashgar dating from the 11th century. At about the same time,

Mahmud al-Kashgari was travelling between Kashgar and Baghdad, both major centres of learning, in order to write the first Turkish lexicon, the Dīwān Lughāt itTurk (al-Kashgari 1032 [1982]) which remains till today an incomparable source about the pre-Islamic culture of the Turkic traditions. The discovery of Mahmud alKashgari’s tomb in 1981-1982 was accompanied by the discovery of documents and old books about the region (Barat 1994: 77-81). Another early Islamic scholar of Kashgar, was Yusuf Khass Hajib (Yūsuf Khāṣṣ Hājib), author of the Ḱutadğu biliğ or The Wisdom of Royal Glory, written in 1069-1070 (Barthold and Spuler 2012; Hayes 2012: 132). This ‘masterpiece of the Turkic-language people’ (Nazarov 1994: 82) takes the form of a ‘didactic poem’ in Qara-Khanid Turkish giving advice to the king’s family about managing the affairs of the state and the private royal household (Marlow 2012). Hajib’s book gives us an insight into the hierarchical structure of society in Kashgar: it speaks of the king, the vizier, nobles, princes, servants, descendants of the prophet, the scholars and ulama, physicians, dream interpreters, astrologers, wives and commoners (Yusuf Hass Hajib in Levi and Sela 2010: 76-81). The work has also been praised as a superb essay on the virtues of beauty over ugliness and justice over oppression (Nazarov 1994: 8288).

refer to victorious campaigns led by the King of Khotan against the Muslim ruler of Kashgar. In 971 AD a dancing elephant taken in Kashgar was sent by the King of Khotan to the Chinese emperor (Emmerick 1986: 271; Skjaervo 2004: 41). 42  The legends of these martyrdoms are embedded in the calendar for Muharram the most important event in the religious year, thereby proving that in the oral traditions of Kashgar (and Central Asia too) the sectarian differences between Shia and Sunni have been ignored (Hayes 2012: 248, 251; Beller-Hann 2008: 310).

The first two hundred or so years of Qara-Khanid rule represents the first period of Islamisation in the Kashgar region. The great cultural efflorescence that is reflected in the literature, poetry, architecture and cult of the shrines established a permanent base for legitimising Islamic cultural hegemony in the centuries that followed. It was also this era that gave

Figure 15. Remains of animal sacrifices outside the tomb of Satuq Boghra Khan (photo Di Castro).

killed by Jukti Rashid, as was his brother Husayn,42 but ultimately Islam asserted itself as the dominant belief system in this region.

134

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record

Figure 16. Exterior of the Tomb of Apa Khoja in Kashgar (photo Gollings).

rise to some of the most enduring Islamic monuments in Kashgar today. The period of Qara-Khanid dominance gave way to the second wave of Islamic evolution under Chinggis Khan and his descendants from about 1218 onwards. At first Mongol dominance involved a symbiosis between Chinggis Khan’s tribal economy and animistic beliefs on the one hand with, on the other hand, the sedentary economies and complex beliefs of the people he conquered, including those of Kashgar. Like other powerful rulers who dominated the Silk Road, the Mongols appreciated how the wealth generated by international trade was beneficial to their own economic needs. This, together with the extraordinary size and spread of the Mongol Empire, brought disparate people, cultures, products and traders into a unique interaction with each other (Biran 2004: 350; 2007: 87-89).43 Prominent in these overland trading networks were the Muslim, Uyghur and Italian merchants (Biran 2004: 351) who, like Marco Polo and Ibn Battuta, themselves travelled between Europe and China rather than depending on the old relay system of using multiple intermediaries (Biran 2007: 91-92).

monuments from this period remain in Kashgar today (Hayes 2012: 244), probably because when extensive renovations of Islamic sites were undertaken in the 19th century they focused on those legends and sites associated with the heroic battles of the initial conversions during the time of Satuq Boghra Khan. The compilation of a comprehensive inventory of the Islamic monuments of Kasghar in historical times and today would help to determine how well the Mongol period is represented in this city and whether there are more sites worthy of preservation given the growing appreciation of Chinggis Khan in China. By the late 16th century the vibrancy of Mongol control started to disappear under pressure from what Hayes (2012: 32) describes as the third wave of Islamisation in the Tarim Basin driven by competing Sufi religious factions who vied for political power. This eventually resulted in the establishment of a powerful new state characterized by the old familiar feature that it was established by a foreign invading force originating from the Ferghana Valley. The Naqshbandiyya Khojas, however, were also defined by the highly innovative feature of being a theocratic state that underscored its legitimacy by claiming direct descent from the Prophet Mohammad (Hayes 2012: 43-47). Kashgar was very much the geographical centre of their power, as the 72 tombs of the Khoja dynasty in the mausoleum of the Apa Khoja suggests (Figures 16, 17). Of course, well before assuming the mantle of kingly power, the Naqshbandiyyas had already distinguished themselves as highly competent administrators and successful merchants.

Most important of all, Mongol dominance did not weaken Islam or the cultural urbanism of Kashgar and the surrounding area. The descendants of Chinggis Khan’s son Chagatai gradually abandoned their tribal animism and became Islamicised. However, unlike the period of the first conversions, few shrines and The Mongols were great patrons of trade, as shown by their construction of commercial infrastructure (Biran 2007: 66). Even Samarkand and Bukhara, towns totally devastated by the Mongol invasion, recovered quickly to emerge into a new era of economic prosperity and religious liberalism (Biran 2007: 67-68). The Mongol nobility themselves contributed to this commercialisation by their apparently insatiable demand for luxury goods and the more general demands created by building a new state (Biran 2004: 350-351).

43 

The sufi era, in terms of political authority, ended with the reign of Yakub Beg from 1864 to 1877 (Kim 2004: xiv). It 135

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 17. Interior of the Tomb of Apa Khoja (photo Gollings).

was the period that coincided with the beginning of the ‘Great Game’ of competition between Russia, Europe and China for influence over this part of Central Asia. Commercial rivalry was unrelenting between Russia and Britain. British assessments of Yakub Beg were, perhaps for this reason, very positive, as the writings of Bellew, Rawlinson and Shaw demonstrate.44 Boulger, by contrast, was a great critic of Yakub Beg whom he described as a ‘military dictator’ running a police state (Boulger 1878: 167).45

embroiled in local affairs and eventually declared himself to be the ruler or Beg of Kashgar (Rossabi 1975: 174-191). He also represented the sufi cultural wave that had engulfed much of Central Asia at that time. Yakub Beg’s preoccupation was with political legitimacy. As a result he balanced the extensive taxation needed to maintain his court and army (Figure 18)46 with cultural programs designed to link his regime to the founders of Islam in Kashgar. He renovated many shrines and monuments (Apa Khoja, Bibi Maryam, etc.), undertook pilgrimages to significant sites such as the Ordam Mazar, personally cooked and distributed food to the poor (Figures 11 and 19)47 and introduced Shariah (Islamic Law) with an unprecedented puritanism in this region. Shaw (1871: 392, 401, 402) gave graphic descriptions of the executions and amputations he saw and heard about during his visit to Yakub Beg’s court, including the display of a human head near one of Kashgar city’s gates as a warning to criminals.48 By the time of

Yakub Beg, like many other rulers of Kashgar, was a foreigner. He served as a soldier in the Khokand court of eastern Ferghana and came to Kashgar as the escort for the last Khoja king who had been living in exile in Khokand. Instead of returning, he found himself For example, on 15 June 1874 Major-General Sir H.C. Rawlinson (1874: 414-415) read out to a meeting of the Royal Geographical Society of London reports from Forsyth’s mission to Kashgar and prefaced this with the following comment: ‘… Yacub Beg has proved himself a very wise and able ruler. He has repressed brigandage, encouraged trade, raised a very decent military force and generally strengthened and improved his country’. Bellew (1875: 302) was also impressed by the high state of law and order in Kashgar. Shaw’s visit to Kashgar in 1868-1869 gives one a very different impression of what constituted order. His account speaks of the strict Shariah (Islamic Law) discipline that Yakub Beg implemented. 45  Unlike Rawlinson, he describes the increased trade of Kashgar as ‘pure fiction’ and Yakub Beg as a ‘pecuniarily embarrassed ruler’ (Boulger 1878: 166). Kashgar’s external trade was indeed oriented towards the south and west and mainly involved Badakshan and the towns of Khokand, Bokhara, Tashkent and Leh (Boulger 1878: 164-165). Trade with Russia was a very small affair, largely because Russian trade focused on Ili, in the north of Xinjiang (Boulger 1878: 174). This description verifies the need for a detailed scrutiny of the Islamic archives that we spoke of earlier in this chapter. 44 

Another reference dated this illustration to 1885. Yakub Beg died in 1877 but his soldiers would have been dressed and armed in similar fashion. 47  The cauldron in Figures 11 and 19 sits in the compound of the mausoleum of Satuq Boghra Khan in Artush; it was a gift given by Yakub Beg to the Artush community in 1869. It has the following dedication written on its rim: ‘In Hijri 1286, on the 27th day of Rejeb month, this dash qazan [community bowl] was given to the mazar (mausoleum) of the Honorable Padishah Sultan Satuq Boghra Khan, by the friend of Alims [the knowledgeable, learned ones], the leader of the miserable, the Ghazi who bears the name of Muhammad Yaqub Beg’. 27 Rejeb was the date of Muhammad’s Ascension to Heaven known as Mi’raaj Nabi. Our thanks to Dr Ayxem Eli and Dr Farid Bezhan for their help with this translation. 48  However, Shaw did comment that these measures were not particularly brutal for their time. 46 

136

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record

Figure 18. Lithograph c. 1890 showing the types of costumes probably worn by the military during Yakub Beg’s reign (Vicziany archive).

argues (2008: 308-310) many other beliefs survived the Islamic surge. Visitors to Kashgar today can see plenty of evidence of the liberal, syncretic Islam that is still practiced and tolerates animal sacrifices, Shamanism, magic and many other pre-Islamic beliefs and practices (Michell et al. 2008: 145). Our account of the Islamic records on Kashgar concludes as we approach the end of the Qing dynasty in the early twentieth century. After 1900 China was increasingly characterised by domestic turmoil, exacerbated by different foreign powers seeking to maximise their influence in east Asia.

Figure 19. Detail of the inscription on the cauldron outside Tomb of Satuq Bogra Khan in Artush dedicated by Yakub Beg (photo Di Castro).

Conclusion

Yakub Beg’s death in 1877, ordinary people felt under tremendous economic and social pressure despite his policies which saw the expansion of agriculture, mining and cross border trade especially with the countries to the east and south of Kashgar (Kim 2004: 137).49

Kashgar’s long term historical significance as a major conduit between China and the countries on its far western border is undisputed. However, in-depth historical analysis of the history of the city and the evolution of its oasis has been lacking. Research, surveys, excavations and dating are all urgently needed, for without detailed information about the history of the city of Kashgar it will be difficult to protect its cultural heritage. That heritage is threatened, as noted in the opening section of this chapter. As we explain in Di Castro, Vicziany and Zhu (in this volume), the cultural heritage of Kashgar includes remnants of traditional technologies for managing the land and water resources. These resource questions are ever more pressing now that Kashgar is the focus of development projects that form part of China’s strategy of developing cross border trade. The underlying environmental conditions of the oasis have not changed in the

On balance, Yakub Beg’s cultural programs were intended to give Islam a privileged position in the local belief systems. In this he failed, for as Beller-Hann 49  Despite heavy taxation, southern Xinjiang, especially Kashgar, was still very prosperous when the Qing dynasty re-asserted control in 1877. Tax records show that the grain tax collected from the city of Kashgar was the highest after Yarkand and all the southern cities provided more grain to the Chinese Emperor than the cities in north Xinjiang (Kim 2004: 123). South western Xinjiang had not been devastated to the same extent as the north in the decades of fighting before the Qing reasserted control because there were fewer Chinese soldiers and civil servants in the south and less infighting amongst the Muslim noble families and factions (Kim 2004: 123). The relative prosperity of Kashgar allowed Yakub Beg to import labour from Urumqi to accelerate his public works projects (Kim 2004:124).

137

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads References

long run – in particular, the dependence on the meltwaters of the local mountain systems and the mismatch between river waters and the needs of agriculture (see the next chapter). Old technologies that addressed these problems, therefore, need to be taken seriously. China today has already become an importer of food and fuel, partly as a strategy for conserving its own scarce water resources. During fieldwork in 2013 we were again reminded of these pressures when on the outskirts of the Kashgar we saw hundreds of acres devoted to desalination by evaporation techniques. On the newly reclaimed land nearby, we also saw tomatoes and other crops growing under plastic covers and wondered about the sustainability of these approaches to increasing agricultural output.

ADB (Asian Development Bank) 2002. The 2020 Project: Policy Support in the People’s Republic of China. Manila: ADB. al-Kashgari, Mahmud 1032. [Edition of 1982], Compendium of the Turkic Dialects, in R. Dankoff and J. Kelly (eds) Sources of Oriental Languages and Literatures, Turkic Sources: 7, Parts 1-3. Cambridge, Mass.: Harvard University Printing Office. Barat, K. 1994. Discovery of History: The Burial Site of Kashgarli Mahmud, in H.B. Paksoy (ed.) Central Asia Reader: The Rediscovery of History: 77-81. London: M.E. Sharpe. Barthold, W. and B. Spuler. 2012. Kās̲h̲g̲h̲ar in Encyclopaedia of Islam, Second Edition. Leiden: Brill, viewed on 15 Sept. 2012, < http://www.brill.com/ publications/online-resources/encyclopaediaislam-online>. Bayly, C.A. 1983. Rulers, Townsmen and Bazaars: North Indian Society in the Age of British Expansion, 17701870. Cambridge: Cambridge University Press. Beal, S. 1869. Travels of Fah-Hian and Sung-Yun, Buddhist Pilgrims, from China to India (400 A.D. and 518 A.D.). Translated from the Chinese by Samuel Beal. London: Trübner and Co. Beal, S. 1983. Si-Yu-Ki: Buddhist Records of the Western World. Translated from the Chinese of Hiuen Tsiang A.D. 629. New Delhi: Munshiram Manoharlal. [Reprint of 1st edition of 1884]. Beckwith, C.I. 1987. The Tibetan Empire in Central Asia. Princeton: Princeton University Press. Beckwith, C.I. 2009. Empires of the Silk Road. Princeton: Princeton University Press. Beller-Hann, I. 2008. Community Matters in Xinjiang 1880-1949: Towards a Historical Anthropology of the Uygur. Leiden/Boston: Brill, Bellew, H.W 1875. Kashmir and Kashgar: A Narrative of the Journey of the Embassy to Kashgar in 1873-74. London: Trübner. Berggren, J.L. and A. Jones. 2000. Ptolemy’s Geography: An Annotated Translation of the Theoretical Chapters; Princeton: Princeton University Press. Biran, M. 2004. The Mongol Transformation: From the Steppe to the Eurasian Continent, in Johan Arnason and Björn Wittrock (eds) Eurasian Transformations Tenth to Thirteenth Centuries: Crystallizations, Divergences, Renaissances: 339-361. Leiden/Boston: Brill. Biran, M. 2005. The Ilak-Khanids (or Qarakhanids), in Encyclopedia Iranica, vol. XII, Columbia University Press, New York, 621-628. Biran, M. 2007. Chinggis Khan, Makers of the Muslim World. Oxford: Oneworld Publications. Boulger, D.C. 1878. The Life of Yakoob Beg; Athalik Ghazi & Badaulet; Ameer of Kashgar. London: Wm. H. Allen & Co.

This chapter’s main objective was to reassess the historical record of Kashgar. We argue that we need to turn away from the history of kings and elites and refocus on systematic research about what sustained daily life for ordinary people including local and long distance trade. If we wish to challenge the simplistic chronologies that focus on the rise and fall of empires as sources of daily misery, we need microhistories of specific locales to show how people lived and benefited from the lack of centralised political systems when these collapsed. If we wish to understand how the history of Kashgar linked up with the tumultuous economic development of eastern cities such as Suzhou, we need to study the micro-details of trade and human exchanges between the eastern and western parts of China and beyond into Central and South Asia. If we believe that Kashgar was a significant centre for Buddhist and Islamic learning, and the available records point to this, we need to invest effort in understanding how that cultural life worked and what kind of economic and social activities made that possible. Here the connections between religious orders and the profits of regional and international trade are likely to prove a fruitful area of analysis. The tools needed for this reassessment are interdisciplinary and multilingual – these need to be applied to intellectual excavations in the hitherto inaccessible archives of Central Asia, Russia, the Middle East and even China. Archaeological excavations and surveys are likely to prove equally fruitful, as we suggest in the next chapter. Acknowledgements We thank Professor Marylin Martin Rhie (Smith College, Mass. USA), Professor Emeritus Colin Mackerras (Griffith University, Australia) and Dr Ayxem Eli (University of New South Wales, Australia) for their comments on an earlier draft of this chapter. 138

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record Brosseder, U. and B.K. Miller (eds) 2011. Xiongnu Archaeology – Multidisciplinary Perspectives of the first Steppe Empire in inner Asia; Series on Bonn Contributions to Asian Archaeology 5, Jan Bremmann (ed.) Freiburg: Feiburger Graphishe Betriebe. Brough, J. 1965. Comments on Third-Century ShanShan and the History of Buddhism. Bulletin of the School of Oriental and African Studies 28: No. 3, 582612. Bussagli, M. 1968. The Problem of Kaniṣka as seen by the Art Historian, in A.L. Basham (ed.) Papers on the Date of Kaniṣka: 39-56. Leiden: Brill. Cary, M. 1956. Maës, Qui et Titianus. The Classical Quarterly, New Series 6: Nos. 3/4 (Jul. – Oct.), 130134. Chavannes, É. 1903. Documents sur les Tou-Kiue (Turcs) Occidentaux. Paris: Librairie d’Amérique et d’Orient Adrien Maisonneuve. Dardess, J.W. 2012. Ming China, 1368-1644: A Concise History of a Resilient Empire. Maryland: Rowman and Littlefield Publishers Inc. de la Vaissiere, E. 2005. Sogdian Traders. A History. Trans. by James Ward. Leiden: Brill Academic. Drompp, M.R. 2005. Tang China and the Collapse of the Uighur Empire. Leiden/Boston: Brill. Dubs, H.H. et al. 1955. The History of the Former Han Dynasty by Pan Ku. Translation Vol. 2, Sponsored by the American Council of Learned Societies. Baltimore: Waverley Press Inc. [Reprint of the First edition of 1938]. Dutt, S. 1988. Buddhist Monks and Monasteries of India: Their History and Their Contributions to Indian Culture. New Delhi: Motilal Banarsidass [Reprint of the First edition of 1962]. Emmerick, R.E. 1986. The Iranian Settlements to the East of the Pamirs, in E. Yarshater (ed.) The Cambridge History of Iran vol. 3, The Seleucid, Parthian and Sasanid Periods, Part 1: 263-275. Cambridge: Cambridge University Press. Fish, I.S. 2010. ‘China’s Hottest Cities and Kashgar’, Newsweek, 25 September, viewed on 13 Nov. 2019, http://europe.newsweek.com/chinas-hottestcities-and-kashgar-72333. Frye, R.N. (n.d.), The Heirs of the Achaemenids: Emperor Ardeshir and the Cycle of History, The Circle of Ancient Iranian Studies, 1998-2015, SOAS, viewed on 21 April 2019, . Gernet, J. 1962. Daily Life in China on the Eve of the Mongol Invasion 1250-1276. London: George Allen & Unwin. Gibb, H.A.R. 1922. The Arab Invasion of Kashgar in A.D. 715. Bulletin of the School of Oriental Studies 2: No. 3, 467-474. Grenard, M.F. 1900. La légende de Satok Boghra Khân et l’histoire. Journal Asiatique Sér. 9, Tome 15: Janvier-Février, 5-79.

Grillot, T. and A-J. Etter. 2012. History Speaks Many Languages: Interview with Sanjay Subrahmanyam, viewed on 6 Jan. 2013, . Grousset, R. 1941. L’empire des steppes. Paris: Payot. Hambis, L. 1964. Sites et monuments de la région de Kachgar et de Toumchouq, in Hambis, L. et al. (eds) Mission Paul Pelliot, Vol. II. Toumchouq. 1-17. Paris: Documents archéologiques publiés sous les auspices de l’Académie des Inscriptions et BellesLettres. Hayes, T. 2012. Enshrined Memory: Landscape, History and Identity among the Uyghur Muslims of China’s Tarim Basin. Unpublished PhD dissertation, University of Queensland. Hill, J.E. 2009. Through the Jade Gate to Rome: A Study of the Silk Routes during the Later Han Dynasty 1st to 2nd Centuries CE. South Carolina: Booksurge. Hitch, D.A. 1988. Kushan Tarim Domination. Central Asiatic Journal 32: Nos. 3-4, 170-192. Houreld, K. 2015. China and Pakistan Launch Economic Corridor Plan Worth US $46 billion. Reuters: 20 April, viewed on 10 May 2019, . Hulsewé, A.F.P. 1979. China in Central Asia: The Early stage 125 BC-AD 23: An Annotated Translation of Chapters 61 and 96 of the History of the Former Han Dynasty. Leiden: Brill. Jamal, Q. in S.C. Levi and R. Sela. 2010. Islamic Central Asia: An Anthology of Historical Sources. Bloomington: Indiana University Press, 73-76. Jarring, G. 1935. The Ordam-padishah system of Eastern Turkistan Shrines. Geografiska Annaler, Vol. 17, Supplement: Hyllningsskrift Tillagnad Sven Hedin, 348-354. Kim, H. 2004. Holy War in China: the Muslim Rebellion and State in Chinese Central Asia 1864-1877. Stanford: Stanford University Press. Kim, H.J. 2013. The Huns, Rome and the Birth of Europe. Cambridge: Cambridge University Press, Chapter 3 ‘The Huns in Central Asia’: 17-43. Kim, K. 2008. Saintly Brokers: Uyghur Muslims, Trade, and the Making of Qing Central Asia, 1696-1814. Unpublished PhD dissertation, University of California, Berkeley. Kramer, M. 1999. Bernard Lewis, in Encyclopedia of Historians and Historical Writing. 719-20. London: Fitzroy Dearborn, Vol. 1, 719-720, viewed on 6 Jan. 2013, . Lansdell, H. 1894. Chinese Central Asia. Vol. 2. A Ride to Little Tibet Volume. New York: Charles Scribner’s Sons. 139

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Lattimore, O. 1940. Inner Asian Frontiers of China. New York: American Geographical Society. Legge, J. 1886. A Record of Buddhistic Kingdoms. Being an account by the Chinese Monk Fâ-Hien of travels in India and Ceylon (A.D. 399-414) in search of the Buddhist books of discipline. Oxford: Clarendon Press. Lerner, J. D. 1998. Ptolemy and the Silk Road: from Baktra Basileion to Sera Metropolis. East and West 48: Nos.1/2, 9-25. Litvinsky, B.A. 1996. ‘The Hephthalite Empire’, in B.A. Litvinsky et al. (eds) History of Civilizations of Central Asia, vol. III, The Crossroads of Civilizations: A.D. 250-to 750: 135-162. Paris: UNESCO. Loewe, M.A.N. 1979. Introduction to A.F.P. Hulsewé, (1979), China in Central Asia: The Early stage 125 BCAD 23: An Annotated Translation of Chapters 61 and 96 of the History of the Former Han Dynasty. Leiden: Brill, 1-71. Macartney, Lady 1985. An English Lady in Chinese Turkestan, with an Introduction by Peter Hopkrik, Oxford/Hong Kong: Oxford University Press. [Reprint of the First edition of 1931]. Mackerras, C. 1990. The Uighurs, in D. Sinor (ed.) The Cambridge History of Early Inner Asia: 317-343. Cambridge: Cambridge University Press. Macpherson, D. 1805. Annals of Commerce, Manufactures, Fisheries and Navigation, With Brief Notices of the Arts and Sciences Connected With Them. Containing the Commercial Transactions of the British Empire and Other Countries, From the Earliest Accounts to the Meeting of the Union Parliament in January, 1801; and Comprehending the Most Valuable Part of the Late Mr Anderson’s History of Commerce with a Large Appendix, London: Nicholas and Son. [Digitised by Google, Vol. 1]. Mair, V.H. (ed.) 1998. The Bronze Age and Early Iron Age Peoples of Eastern Central Asia, 2 vols. Washington/ Philadelphia: The Institute for the Study of Man and the University of Pennsylvania Museum Publications. Mallory, J.P. and Mair, V.H. 2000. The Tarim Mummies. London: Thames and Hudson. Marlow, L. 2012. Advice and advice literature, in Encyclopaedia of Islam, Vol. 3, E J Brill Online, viewed on 15 Sept. 2012, . Marmé, M. 2005. Suzhou: Where The Goods Of All The Provinces Converge. California: Stanford University Press. Marmé, M. 2007. Survival through transformation: how China’s Suzhou centered world economy weathered the general crisis of the seventeenth century. Social History 32: No. 2, 144-165.

Marcellinus, A. 1989 in Caltabiano, Matilde (ed.) Ammiano Marcellino Storie. Milano: Rusconi [in Italian]. McLauchlin, R. 2010. Rome and the Distant East: Trade Routes to the Ancient Lands of Arabia, India and China. UK/USA: Continuum. Meunié, J. 1942. Shotorak, Memoires de la Delegation Francaise en Afghanistan (MDAFA), Vol. X, Paris: Éditions d’art et d’histoire. Meunié, J. 1943-45. Le couvent des otages chinois de Kanishka au Kāpiça. Journal Asiatique 234: 151162. Michell, G. et al. with photographs by Gollings, J. 2008. Kashgar, Oasis city on China’s Old Silk Road. London: Frances Lincoln Publishers. Millward, J.A. 1998. Beyond the Pass: Economy, Ethnicity, and Empire in Qing Central Asia, 1759-1864. California: Stanford University Press. Millward, J.A. 2007. Eurasian Crossroads: A History of Xinjiang. New York: Columbia University Press. Morley, N. 1996. Metropolis and Hinterland. The City of Rome and the Italian Economy 200 BC-AD 200, Cambridge: Cambridge University Press. Nazarov, B. 1994. Kutadgu Bilig: One of the First Written Monuments of the Aesthetic Thought of the Turkic People in H.B. Paksoy (ed.) Central Asia Reader: The Rediscovery of History: 82-88. London: M.E. Sharpe. Narain, A.K. 1968. The Date of Kaniṣka, in A.L. Basham (ed.) Papers on the Date of Kaniṣka: 206-239. Leiden: Brill. Neelis, J.E. 2011. Early Buddhist Transmission and Trade Networks Mobility and Exchange Within and Beyond the Northwestern Borderlands of South Asia (Dynamics in the History of Religions 2). Leiden: Brill. Pelliot, P. 1959. Notes on Marco Polo. I. Ouvrage Posthume publié sous les auspices de l’Académie des Inscriptions et Belles-Lettres et avec le concours du Centre national de la Recherche scientifique. Paris: Imprimerie Nationale, Librairie A. Maisonneuve. Personal Communications from Professor Emeritus Colin Mackerras, February 2013. Petech, L. 1950. Northern India According to the Shui-Ching-Chu. Serie Orientale Roma, II. Rome: Istituto Italiano per il Medio ed Estremo Oriente. Polo, M. 2016. The Travels. London: Penguin Books. Ptolemy 1966. Claudii Ptolemaei Geographia, in C.F.A. Nobbe (ed.) Hildesheim: Georg Olms Verlangsbuchhandlung [In Greek]. Puri, B.N. 1987. Buddhism in Central Asia. Delhi: Motilal Banarsidass. Rapin, C. 2001 [1998]. L’incompréhensible Asie centrale de la carte de Ptolémée. Propositions pour un décodage, in O. Bopearachchi et al. (eds) 140

M. Vicziany and A.A. Di Castro: The Kashgar Oasis: Reassessing the Historical Record ‘Alexander’s Legacy in the East. Studies in Honor of Paul Bernard’, Bulletin of the Asia Institute 12: 201225. Rawlinson, H.C. et al. 1873-1874. Extracts of Letters from Members of Mr Forsyth’s Mission to Kashgar Relating to the Geographical Results of the Mission. Proceedings of the Royal Geographical Society of London 18: 414-444. Rhie, M.M. 1999. Early Buddhist Art of China and Central Asia, Vol. 1, Later Han, Three Kingdoms and Western Chin in China and Bactria to Shan-shan in Central Asia. Leiden: Brill. Rhie, M.M. 2002. Early Buddhist Art of China and Central Asia, Vol. 2, The Eastern China and Sixteen Kingdoms Period in China and Tumshuk, Kucha and Karashahr in Central Asia – Text. Leiden: Brill. Rossabi, M. 1970. The Tea and Horse Trade with Inner Asia during the Ming. Journal of Asian History 4: 136168. Rossabi, M. 1975. China and Inner Asia. From 1368 to the Present Day. London: Thames and Hudson. Rossabi, M. 1990. The ‘decline’ of the central Asian caravan trade, in James D. Tracy (ed.) The Rise of Merchant Empires: Long Distance Trade in the Early Modern World, 1350-1750: 351-370. Cambridge: Cambridge University Press. Roux, Jean-Paul 1990. La religione dei Turchi e dei Mongoli. Genova: ECIG [Italian translation of the 1st French edition of 1984]. Shasha, D. (ed.) 2012. International trade fair opens in Xinjiang. Xinhua, 28 June, viewed on 21 June 2014, . Shaw, R.B. 1871. Visits to High Tartary, Yarkand and Kashgar, (formerly Chinese Tartary), and return journey over the Karakoram Pass. London: John Murray. Shaw, R.B. 1878. A Sketch of the Turki Language as Spoken in Eastern Turkistan (Kashgar and Yarkand). Calcutta: Baptist Mission Press. Sinor, D. 1990. The establishment and dissolution of the Türk empire, in D. Sinor (ed.) The Cambridge History of Early Inner Asia: 285-316. Cambridge: Cambridge University Press. Sinor, D. and S.G. Klyashtorny. 1996. The Türk Empire in B.A. Litvinsky et al. (eds) History of Civilizations of Central Asia, vol. III, The Crossroads of Civilizations AD 250 – to 750: 327-347. Paris: UNESCO. Sinor, D. 1997. The Kitan and the Kara Khitay, in M.S. Asimov and C.E. Bosworth (eds) History of Civilisations of Central Asia, Vol. IV, The Age of Achievement: A.D.750 to the End of the Fifteenth Century, Pt 1: 227-242. Paris: UNESCO. Skjaervo, P.O. 2004. Iranians, Indians, Chinese and Tibetans: The Rulers and Ruled of Khotan in the

First Millennium, in S. Whitfield and U. SimsWilliams (eds) The Silk Road. Trade, Travel, War and Faith: 34-41. Chicago: Serindia Publications, Chicago. Stein, M.A. 1907. Ancient Khotan: Detailed Report of Archaeological Explorations in Chinese Turkestan. Oxford: Clarendon Press. Types and costumes from Chinese Turkestan c. 1890. The Universal Geography with Illustrations and Maps, Division XIII. London: Virtue & Co Limited. Yu, T. 1998. A Study of Saka History. Sino-Platonic Papers 80: July. Yusuf Hass, Hajib in S.C. Levi and R. Sela. 2010. Islamic Central Asia: An Anthology of Historical Sources: 76-81. Bloomington: Indiana University Press. Vicziany, M. 2005. Ethics, Archaeology and the Engineering Profession in Developing Countries. 4th Asia-Pacific Forum on Engineering and Technology Education, Conference Proceedings, Bangkok, UICEE, September: 15. Vicziany, M. and Di Castro, A.A. 2014. ‘The Buddhist Cultures of the Chakmak River System’, Lecture to the Department of History and Ethnography, Xinjiang Normal University, Urumqi, 14 October. Vicziany, M. and Di Castro, A.A. 2016. The Hidden Treasures of Buddhist Kashgar. TAASA Review, The Journal of the Asian Art Society of Australia 24: 1417. Vicziany, M. et al. (forthcoming) Modelling Pastoralism’s Varied Characteristics. Watt., C.Y. et al. (eds) 2004. China. Dawn of a Golden Age 200-750 AD, New York/NewHaven/London: Metropolitan Museum of Art and Yale University Press. Xinhua 2011. China issues new rules to ease tensions over forced demolition. People’s Daily January 23, viewed on 21 June 2014, . Xinjiang Statistical Year Book 2011, Section 5-11, viewed on 6 Jan. 2013, . Xinjiang Institute of Relics and Archaeology 1995. ‘Report on the Relics of Artux, Kashgar, Hotan areas during the Qarakhanid dynasty’, in Xinjiang Institute of Relics and Archaeology ed. New Achievments in Archaeological Research in Xinjiang during the Span (1979-1989)], 582-588. Urumuqi: Xinjiang Renmin Chubanshe. [新疆文物 考古研究所 1995. 阿图什喀什和田地区卡拉汗朝 遗迹调查, 载于 新疆文物考古研究所编著 新疆文 物考古新收获 (1979-1989): 582-588页. 乌鲁木 齐: 新疆人民出版社]. 141

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Zarcone, T. 1999. Quand le saint légitime le politique: le mausolé de Afaq Kwaja à Kashgar. Central Asian Survey 18: 225-241.

Zürcher, E. 1968. The Yüeh-chih and Kaniṣka in Chinese Sources, in A.L. Basham (ed.) Papers on the Date of Kaniṣka: 346-390. London: Brill.

142

Chapter 9

The Kashgar Oasis in Buddhist and Pre Buddhist Times: the Archaeological and Environmental Record Angelo Andrea Di Castro, Marika Vicziany and Xuan Zhu In the previous chapter, Vicziany and Di Castro analysed a selection of the historical records about Kashgar and the kind of biases that need to be addressed. As we saw, enormous gaps remain in our knowledge, especially about the daily lives of ordinary people who lived under the shadow of kings, courts and the religious authorities of the oasis. Archaeological studies have the potential to answer many questions that historical documents cannot, especially when those documents are concerned largely with the affairs of the elites. The scattered documentary records for Kashgar stressed the rise and fall of many small city states in western China, their continuous feuding, the jostling of families to grab the title of king from their competitors and the many incursions of outsiders from the east, the west, the north and the south. This struggle to control the Kashgar oasis was driven by the resources of the oasis and its strategic significance at the crossroads between the Himalayan-Pamir mountains and the Tianshan, all of which suggests that Kashgar had the economic and social prerequisites to support a large, sedentary, monastic Buddhist presence and later the madrassas and mosques that made Kashgar into an important Islamic cultural centre. As Neelis has put it (2011: 292), Kashgar’s importance to China was assured by its role as the meeting point between the northern and southern parts of the Silk Road that skirted the Taklamakan desert in order to link China with Central Asia and Europe.1

been vital for the support of the city as an Islamic hub. Most important of all, the Buddhist sites and monuments are at greater risk today than the Islamic ones partly because the Muslim culture of Kashgar remains a vibrant part of China’s social life and this ensures that some of the key Islamic monuments receive community and government support and protection. Already in the 19th century, Yakub Beg began extensive renovations on many of these. By comparison, there has been no social life relevant to the Buddhist sites for many centuries. But our greatest problem for Kashgar is that we still have no clarity on exactly when Buddhism reached this oasis and the dates for the buildings that survive from that era. Nevertheless, the modest remains which we see today suggest that the Buddhist period represented not only a cultural flowering but also a great leap forward in production methods to sustain a large clerical community and many international pilgrims and traders. This religious life also promoted an expanded commerce within the region and internationally. As we saw in the previous chapter, Xuan Zang speaks of Kashgar having some 10,000 monks in the mid-7th century AD while studies of the Buddhist sites of India tell us that international trade played a major role in sustaining Buddhist cultures wherever monasteries and stupas were built (see previous chapter). Using a broad timeframe, this chapter maps out the nature of the environment in which the Buddhist culture of Kashgar was able to thrive. In advocating the reconstruction of the landscape, human settlement patterns, and the management of water, soil and other resources, we seek to take the archaeology of Kashgar beyond the initial interest that the European explorers had in document collections and the focus on the study of buildings and relics which has typified post-revolutionary Chinese scholarship. We begin this chapter with an account of the archaeological remains of the Buddhist sites, stressing the need to establish reliable dates. We are dealing with possibly some of the oldest Buddhist remains in western China yet, compared to Dunhuang, Kashgar remains the Cinderella of the Silk Road. It has attracted no significant international projects or collaborations between western and

Much of this chapter concentrates on the Buddhist monuments and land usage patterns of ancient Kashgar, partly because the archaeological remains are so distinctive and partly because we can begin to understand the nature of the agricultural economy that supported the Buddhist culture from comparative studies of the Buddhist monasteries in India and Sri Lanka. The environmental and economic underpinnings of Buddhist Kashgar would also have The authors of this chapter are members of Monash University’s Kashgar Research Project, a multidisciplinary, international research collaboration established in 2004 with the Xinjiang Normal University and the support of the Department of Tourism and the Department of External Affairs, Xinjiang Government. Our first major book was published in 2008 (Michell et al. with photographs by Gollings).

1 

143

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 1. Diagram showing some of the key archaeological sites discussed in this chapter.

Chinese scholars. In documenting the nature of the eight major Buddhist sites, we also discuss the kind of economy that was needed to support such large populations of religious, military and political elites, local and international traders and a significant number of permanent residents in the first half of the 1st millennium AD. This is followed by an exploration of the limited evidence about the pre-historical settlement patterns in and near the Kashgar oasis. It could well be that the area around the village of Upal (Wupaer), for instance, located at a distance of some 45 kms south-west of the modern city of Kashgar, was the site of the early human settlements in western China during the Neolithic period and before.2 Finally, we speculate about the question of whether climate change models can be used to improve our understanding of the environmental and physical constraints of human settlement and cultural achievements. There is clearly a direct relationship between the history of Kashgar and its fragile oasis condition that depends on the meltwaters of the surrounding mountains and human technological ingenuity. The factors determining the

flow of snowmelts is an important consideration in mapping out the general landscape. The Buddhist Sites of the Kashgar Oasis The first systematic description of the stupas3 of Kasghar and their environment was given by Sir Aurel Stein (Stein 1907: Vol. I, 76-86). He did not, however, spend much time in Kashgar and carried out only limited excavations, preferring to rush on to Khotan and Niya. As a result, some of Stein’s bold statements about the Kashgar stupas do not stand up to scrutiny today. For example, he asserted that all the stupas in Kashgar, Niya, Rawak and Endere had square bases which dated them later than the round bases of early stupas in the Indian north-western regions, which he declared to be ‘totally absent in Eastern Turkistan’ (Stein 1907: Vol. I, 83). Rhie has disputed this and writes that the Topa Tim stupa is perhaps the oldest of the five remaining structures on the outskirts of Kashgar city, although her argument appears to be based on the reports of Stein (1907) and Pelliot (Hambis 1964), rather than on new fieldwork. According to Rhie (1999: 249), Topa Tim, like other early Gandharan structures, had a circular base which could

Personal communication from Professor Yidilisi Abuduresule, 2013, Urumqi. If this hypothesis is correct, then such early agriculture in the Kashgar area contrasts with some Bronze Age sites like Adunqiaolu in north-western Xinjiang that have not yet yielded any evidence of Neolithic settlement.

2 

On stupa archaeology, architecture and symbolism see for instance Dallapiccola and Lallemant (eds) 1980; Snodgrass 1985.

3 

144

A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times suggests its construction around the 1st century AD. On the other hand, the proportions suggest that the stupa was built later than this, perhaps between 100-200 AD. The problem is that there has been no archaeological fieldwork to provide accurate dates for any of Kashgar’s stupas or associated monasteries. We are not aware of any systematic analysis of their structure, dates or evolution. In the case of Topa Tim, knowledge about it and its location had totally disappeared until it was rediscovered by the Director of the Kashgar Museum in 2003.4 Despite this highly significant local discovery, no excavations, sampling or investigation has been conducted even though Rhie suggests that this might represent the oldest stupa in this part of western China. Perhaps Chinese scholars have been working on these stupas, but if so their work has not come to our attention yet.5 Thus the Buddhist monuments of Kashgar provide us with a dramatic example of the urgent work that needs to be done – conservation, new excavations if necessary, surveys and laboratory analysis in addition to a proper inventory of Chinese scholarship about these important structures. Modern technologies can help us begin to understand some of these issues, and we will return to this question below. In this chapter, for example, we suggest that a reconstruction of Kashgar’s environmental evolution during the pre-Islamic period might also help us achieve a better appreciation of its long-term, post-Buddhist historical evolution.

We have some questions about these statements, partly because Rhie was extrapolating from the drawings of Stein rather than basing her comments on fieldwork. In the case of Kizil Debe and Tegurman, we still need more evidence to show what kind of stupas these structures were. The word ‘Tim’ in Turkish means ‘mound’ and it has been used for a very long time by local people to describe all the stupas of Kashgar. However, some of these structures are so poorly preserved that it is difficult to distinguish them visually from the watch towers that were constructed by Chinese imperial troops during their campaigns in western China. Without further fieldwork, we cannot be sure that the ‘Tims’ identified by Stein and Pelliot (both followed by Rhie) were stupas at all. With every passing year it also becomes more difficult to verify what these structures were – for example the site of Tegurman Tim (Yawaluk/Dakiyanus)7 has been heavily disturbed by the construction of a highway and power and telephone cables (Figure 2). The stupa vanished some time ago. When we visited in 2005, the site was occupied by an extensive but abandoned brick factory with multiple kilns. The disused factory remains visible on Google Earth. The stupa described by Pelliot in 1906 was also marked by him on a diagram (Figure 3). Qadir, the late Director of the Kashgar Museum, was convinced that there was a significant stupa at the Tegurman Tim site – and his views have to be taken seriously because he had an intimate knowledge of the area based on his own explorations and excavations there. One of his important findings was a 6th-7th century AD ‘Jar with Three Handles’ in Sogdian style (Watt 2004: 190191; Di Castro 2013: 9-10; Vicziany and Di Castro 2015: 14-15). According to Qadir, the Tegurman Tim site stretched across a large area that is now divided by a major highway. The last excavation in 2000 ran for only 18 days but in addition to the significant vase many fragments of pottery as well as birch bark manuscripts with Brahmi script were also found.

The other four stupas, suggests Rhie (1999: 250-252), were built after Topa Tim: 1. 2. 3.

4.

Kurghan Tim, with a square base, dated to the 1st to 2nd centuries AD; Kizil Debe, with a rectangular base, dated to the 2nd century AD; Khākān-ning-shahri, (the ‘Grand Khan’s town’), also called Tegurman/Tigharman, (the ‘windmill’)6 of intermediate age between Topa Tim and the other two; and Mori Tim, with a square base, dated around the 2nd to 3rd centuries AD influenced by sophisticated Bactrian styles.

Kizil Debe has also vanished as we noted in the introduction to our previous chapter in this volume. The identification of Kurghan Tim is also problematic – we have been unable to locate it.8 So in effect, today we only have two remaining stupa sites in Kasghar, the one called Mori Tim and the other, Topa Tim. Moreover, if the second structure situated on the Mori Tim site turns out to be a stupa as Di Castro has suggested9 then

4  The Director of the Kashgar Museum, Abduyraehim Qadir (d. early 2018) gave us the benefit of his enormous knowledge of the region and Kashgar’s Buddhist monuments. 5  We know that Khan-oi was excavated in the 1950s as the signs displayed on the site in Chinese and Uyghur clearly state this. In 1957 the site was declared protected for its cultural heritage importance. We were told that in the 1980s excavations had been carried out again, and that a report in Uyghur language by the city of Kashgar was published in 1983. Unfortunately we haven’t been successful in finding this publication. The only report available to us refers to Tang and Qara-Khanid period material found at Khan-oi (Xinjiang Institute of Archaeology 1995a). 6  See also Pelliot 1906: 263-268; Hambis 1964: 18-21; Hambis ed, 1961: pl. CXXIII, figures 335, 336; Stein 1980 [1921]: Vol.1, 80-82. This ‘stupa’ in Stein’s time was still 32 feet high and part of an urban settlement with extensive fortifications – see Stein 1980 [1921]: Vol. 1, opposite 82 figures 33 and 34. Strangely Stein never refers to Pelliot (1906) who published a plan of the site including the ruined position of the stupa.

The Europeans called this site Tegurman Tim but it is more commonly known as Yawaluk/Dakiyanus (Personal communication by Abduyraehim Qadir, October 2013). See also report written by Qadir and his team in Relics Team of Cultural Relics Office in Kashi Prefecture 2002: 54-57; Xinjiang Cultural Relics  Bureau et al. (eds), 1999: 267. 8  Stein (1907: Vol. I, 72-74) indicates that it was located to the NorthWest of the Old City, between the Chini Bagh (the residence of McCartney, the English Consul) and the old cemeteries, including the Apa Khoja mausoleum. 9  Di Castro calls it ‘Mori Tim Stupa B’ and Stein a ‘ruined mound’ or ‘B’ or an ‘oblong mound’. Stein (1907: Vol. I, 84 and Vol. II, pl. XXII) 7 

145

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 2a and b. Aerial view of the site of Yawaluk/Tegurman in 2002 (a) and in 2012 (b) (from Google Earth).

Figure 3. Plan of the ruins of Yawaluk/Tegurman surveyed by Pelliot in 1906 (from Pelliot 1906).

stupas survive in contemporary Kashgar; and another three may have survived in the greater Kashgar area,10 although we have not visited these.

in fact, Kashgar had at least six stupas, not five, at the time of Aurel Stein’s visit. In that case three of the six compared this monumental structure to a stepped pyramid with niches on all sides or ‘Qocho’. Franz (1980: 41-42) defined this type of monument in Gaochang, as a ‘temple-stupa’ or ‘Nischenpagode’; see also Di Castro 2008: 263-266.

Just outside Kashgar there were remains of more stupas south-east of Khan-oi. Pelliot visited these and a brief notice about  them can be found in Hambis 1964: 32-33 (for  Dongkul and Aqqach [called

10 

146

A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times

Figure 4. Mori Tim stupa A on the right and stupa B on the left (photo Di Castro).

In the case of the surviving stupas, we have the usual problems of dating their origins. Di Castro gives a later date for construction than Rhie, but we have no certainty. Rhie also suggests that these stupas were located within monastic residential complexes or vihāra/sanghārāma. We have Xuan Zang’s eye witness account of the ‘several hundreds of Sangharamas, with some ten thousand followers’ (Beal 1983: 307) in Kashgar during the mid-7th century AD to back up Rhie. The most elaborate stupa was Mori Tim, which Rhie (1999: 254) does not find surprising because in her opinion it was probably built later than the other four and reflects the inspiration of Mahāyāna rather than Theravāda ideas.11

of Topa Tim being the oldest stupa in Kashgar and perhaps western China, proper study of this site is urgent. High resolution satellite images of the Topa Tim area, for example, show that a new, wide road now passes between the stupa and the nearby mound (probably the remains of the related monastery). Fragments of a stucco figurative frieze were observed during our visit in October 2013 lying on the disturbed ground (Figure 5). The stucco fragments had been exposed by the long and heavy rains in the area a week or so before our visit. However, most of the churned earth was the work of road construction and the removal of earth by local farmers and builders. Similar risks to the integrity of the Khan-oi site also exist (see below).

The only recent work on the stupas of Kashgar is that by Di Castro (2008), who made a careful analysis of the Mori Tim stupa (Figure 4) and compared it with accounts of more than a century ago by Stein and Pelliot. Given the constraints on foreign scholars undertaking scientific measurements in China, Di Castro’s observations are based on an extrapolation of the dimensions of the Mori Tim Stupa from rectified remote sensing imagery.

Despite the constraints, Di Castro has drawn our attention to the unsatisfactory state of our knowledge of the Kashgar stupas by showing the discrepancies between the existing stupa remains and the now badly dated accounts of Kashgar from the early 20th century.

Following the advice of Abduraehim Qadir, we also found the Topa Tim stupa, located some kilometres east of the ruins of the old city of Khan-oi on the left hand side of a road that starts in Kashgar city. Our inspection cannot rule out Rhie’s suggestion that the base of the stupa is possibly round, but the debris around the stupa from the decay of its structure and materials prevents us from any certainty on this subject. Given that the roundness of the base may provide strong confirmation Akekashixiang on Google Map]) and 33-35 (for Tope Chahr and other stupas). These stupas were probably connected to the ones we have discussed in this chapter although Rhie, for example, does not mention them. 11  Why Rhie links Mori Tim with the Mahāyāna tradition is not fully explained by her. See also Stein 1907: Vol. I, 81-85; Hambis 1964: 27-30, Di Castro 2008: 278.

Figure 5. Fragment of a figurative stucco from the debris of the monastic site of Topa Tim (photo Di Castro).

147

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Nevertheless, the visual information revealed by the sites is highly suggestive. Between the Mori Tim ‘Stupa A’ and the so-called ‘Stupa B’ there are secondary structures suggestive of monastic residences or rooms for ritual usage (Di Castro 2008: 275-276). Excavating these minor structures, after Stein had carried out his investigations, Pelliot found fragments of painted stucco (not clay as in most sites in the Tarim Basin) of draperies and one ear, indicating that a large statue of Buddha (or a Bodhisattva) had been installed either in one of the ‘small rooms’ or in one of the niches of the larger Stupa B (Hambis 1964: 30; Di Castro 2008: 276).

that they worshipped mainly at the nearby site of Topa Tim and/or they had small domestic Buddhist shrines within their living quarters. The remains of Khan-oi suggest a considerable urban settlement. The ‘main’ area today is protected by a wall that was 7 m wide, and extends for about 850 square metres (Xinjiang Cultural Relics Bureau et al. (eds) 1999: 268). However, observations on the ground match satellite images suggesting that the site of Khanoi was much larger than this, and probably covered an area of about 3 kms x 1 km. The area identified by Xinjiang Cultural Relics Bureau et al. (1999: 268) was the residential zone, or the ‘fort’, of Khan-oi, historically known by the name of Hasa Tam (Stein 1907: Vol. I, 79, Hambis 1964: 23-24).

Our visit to the ancient city site of Khan-oi also suggested to us a separation between royal and religious spaces at the time when Buddhism was a prominent religion in Kashgar. The urban area of ancient Khan-oi itself does not appear to have included any major stupa, but further archaeological fieldwork could correct this view and perhaps reveal the presence of shrines, or smaller stupas inside the royal centre. Till now, however, we have no evidence to conclude that Khan-oi was a Buddhist centre, but it certainly was a royal residence, as its name suggests (‘the Khan’s abode’). We also have no conclusive dates for Khan-oi’s establishment. Nor do we know anything about the religious beliefs of the royalty and commoners who lived in Khan-oi. What we do know is that even if Buddhism was important it was not to the exclusion of other religious ideas. According to the Tang Annals the Iranian cult of the heavenly god Ahura Mazda was also practiced in Kashgar during the period of the Turkic dominion (Chavannes 1903: 121; Stein 1907: Vol. I, 71, note 48). The only explanation we have so far about Khan-oi is given by local oral traditions that speak of bloody religious wars in this area (see below). Another explanation that we need to consider is that the Khan-oi elite might have been Buddhists but

Most of Khan-oi has crumbled now but the mud walls are still 5-6 m high in parts. Other details reported for Khan-oi include the palace and coins from the Tang dynasty (Xinjiang Institute of Archaeology, Social Science Academy 1983: 154) and Qara-khanid eras but the excavations did not yield sufficient evidence to prove that Khan-oi was the capital of the Qara-khanid during the 10th to 13th centuries AD (Xinjiang Institute of Archaeology 1995a: 587-588). However, old Kashgar did have a royal centre because Mahmud al-Kashgari referred to Kashgar as ordu kand, meaning the town of the King’s residence (al-Kashgari 1032: Vol. 1, 148, 270). Despite its alleged importance during the Tang dynasty (618-907 AD) we have no maps of the structure of this urban centre. Today the integrity of the site is being compromised by the expansion of agriculture, irrigation canals and road works (Figures 6, 7). Khan-oi was not the only urban settlement in the Kashgar oasis or on its desert fringe. The keeper of

Figure 6. Tractor marks on the old structures of Khan-oi in 2013 (photo Di Castro).

148

A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times

Figure 7. Encroachment of irrigation canals on the old structures of Khan-oi in 2013, visible in the middle of the image (photo Di Castro).

cultural relics at Yopurga, who accompanied us in 2005 during our first visit to some of the desert mazars (shrines/graves) on the edge of the Kashgar prefecture, indicated that in the recent past the remains of ancient ‘towns’ have been exposed in the area when severe storms blew away large quantities of sand to reveal extensive urban structures that were normally buried by the dunes. When the sands then blew in the opposite direction, these structures were again swallowed up by the desert (Figure 8). He told us that perhaps up to six old urban sites from the ‘Buddhist period’ may well lie buried in the sands around Mori Tim, Topa Tim, Khan-oi and Yopurga. This oral tradition cannot be discounted, but nor does it prove that these other towns existed or that they were Buddhist. They may, for example, have been large shrines. Only a reliable archaeological ground or belowground survey could establish the truth or otherwise of these local legends. New remote sensing technologies make aerial surveys possible now. Excavation work can also be made more precise using ground penetrating radar (GPR). The farmers working the lands adjacent to the Khan-oi site tell stories about great religious battles between Buddhist and Muslim Kings, fought to convert the former to Islam and thereby turn the rest of the population away from Buddhism and other religions that were followed by the Khan-oi residents. Despite these stories, Iranian cults, for example, appear to have been important even after the arrival of Buddhism and Islam.

Figure 8. Sand dunes covering up part of the cemetery at Yopurga (photo Gollings).

previous chapter in this volume). By themselves, however, folk legends are no substitute for scientific inquiry. Modern research techniques such as Optical Stimulated Luminescence (OSL), can be used to date ancient natural sediments and anthropic strata, 12

During our visits to the Khan-oi site we noticed some farmers collecting soil from Khan-oi – they told us that the earth was very fertile in this area because so much blood had been shed during the religious wars of hundreds of years ago. The belief in raging religious wars has been kept alive by the cult surrounding the Ordam Mazar (see Vicziany and Di Castro in the

On OSL see Wintle 1996: 131-132. For applications of OSL see Bailiff et al. 2013 and Coningham and Gunarwadhana 2013: 26-27. The biochemistry of phosphorus can offer good opportunities for minimally invasive methods of investigation. For instance, aluminium

12 

149

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 9. Plan of the ruins of Eskishahar surveyed by Pelliot in 1906 (from Hambis 1961).

whereas buried organic remains can be examined by means of phosphate and nitrogen analysis. GPR can help in reconstructing plans of buried structures and the extent of the settlements,13 as can remote imaging (Luo et al. 2017: 79-88).

fortress protected by 7 m thick mud walls. In Pelliot’s time, two sections of the city walls remained: one was about 200 metres long and the other 287 metres. In 1990 when the Kashgar Cultural Relics Team surveyed the site only 10 metres of the wall remained and most of the site had been taken over by a construction company.

Another ancient site in the Kashgar prefecture that has been partially investigated and surveyed but apparently not fully excavated is that of Eskishahar (Aisikeshaer / Aisike Sa’ercun meaning ‘old city’) on the south-eastern fringe of Kashgar. Parts of the original wall remain to a height of more than three metres although the old city has vanished with the construction of new dwellings and a memorial park celebrating the conquests of Ban Chao who died in c. 102 AD (see the previous chapter). Various Chinese archaeologists have identified Eskishahar with Wuji during the Eastern Han Dynasty, or Shule in the Tang Dynasty or a town ruled by Ruohetuorixiti in the 9th-10th centuries AD (Xinjiang Annals Editorial Board (eds) 2007: 168-169). With the arrival of Islam, the city elder of Eskishahar (by the name of Nuoguterebute), led the resistance to the ‘Arabs’ but was defeated and escaped to Khotan (Xinjiang Institute of Archaeology, Social Science Academy 1983: 153). The layout of the city was explored and mapped by Paul Pelliot in 1906 (Figures 9, 10, 12, 13).14 As with Khan-oi, the city was a

Neither the date nor historical evolution of Eskishahar/ Wuji has ever been established although some pottery shards were dated to c. 220-589 AD. Sections of the wall remain as the photographs below show (Figure 11) – these provide clues about where future research might begin. Such analysis is important not only for an intrinsic understanding of the nature of Eskishahar but also its relationship with Khan-oi and what today is called the ‘old’ Kashgar town on the mound in the city centre of Kashgar (see Vicziany and Di Castro in the previous chapter of this volume for an account of the recent demolitions). How were these three towns related chronologically, historically and culturally?15 339. Unfortunately, because of the new housing developments in this part of Kashgar city it will be difficult to evaluate the exact extent of the old site of Eskishahar. Nevertheless, we need to investigate the claim that this was the ‘capital’ after the abandonment of Khan-oi and before the construction of the ‘old’ Kashgar town that Mirza Haidar referred to. De Filippi (1932: 492), for example, writes as follows about Eskishahar: ‘… in 1513 […] Aba Bakr destroyed Eski Shahr, the ancient Kashgar, turning the River Kisil Su from its course, it is said, so that it might undermine the fortifications (built of dried mud-brick, as usual in Turkestan). Then, in seven days, by the labours of 10,000 men, he built the present Kashgar, a little distance above the ancient one.’ 15  Zürcher (1968: 369) quotes the Hou Hanshu (77.6b) where it says that the King of Kashgar, Chung, ‘revolts and entrenches himself at the city of Wu-chi’ [Wuji]. Other Chinese archaeologists have

and iron phosphate can give good data about plants and crops (Rapp and Hill 1998: 195). On phosphate analysis see Walker 1992. Nitrogen analysis can also offer remarkable data about dietary habits, see for instance Hedges and Reynard 2007. 13  On the various applications of GPR see for instance Conyers, 2012. 14  Hambis 1964: 17-18; Hambis (ed.) 1961: pls. CXXIV.337, CXXV.338-

150

A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times

Figure 10. Image of the ruins of Eskishahar as it appeared during the Pelliot expedition in 1906 (from Hambis 1961).

Figure 11. Remains of the walls of Eskishahar in 2013 (photo Di Castro).

these questions would enable us to begin to reconstruct the urban history of this significant oasis town. The remaining walls of Kashgar and Khan-oi also require analysis.17

In October 2013, on a trip into the north-eastern part of the treeless desert some 10 to 15 kms east of Mori Tim we came across an extensive site marked by a large sign declaring this to be ‘Ancient Shule’.16 We were surprised to see this sign because other papers on Kashgar’s history had pointed to the location of ancient Shule to the southeast of Kashgar town. Finding answers to

The documentary records for the Kashgar oasis, together with the observations by various foreign visitors in the 20th century are too thin and dispersed

identified Eskishahar with the Shule of the Tang period or with the city of the late 9th to the early 10th century as mentioned above (Xinjiang Annals Editorial Boards (eds) 2007: 168). 16  Professor Chen (Xinjiang Normal University, Geography School), Professor Vicziany, and Dr Zhu undertook this fieldtrip in April 2013.

17  We have been reminded of the value of archaeological investigations that study the walls of Chinese towns by the recent discovery of the walls of one of the oldest and largest Neolithic urban sites at Shimao, in northern China (Xinhua 2012).

151

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads to allow us to draw strong conclusions about the history and settlement pattern in this oasis. Archaeological fieldwork and laboratory analysis are critical if we are to give Kashgar its proper place in Central Asian and Chinese history and protect its historically significant Buddhist monuments. If the Kashgar stupas are amongst the earliest Buddhist structures in western China then detailed and systematic research is urgently needed. There are now a number of important studies for other Asian civilisations that provide models for the kind of strategies that could be developed if we wish to answer some of the questions we have raised in this and the previous chapter.

formed only a small part of the large Khmer civilization that covered various urban, peri-urban and rural areas across a broad landscape of more than 1000 square miles (Fletcher and Evans 2012). The urban core was populated by the royal families, aristocrats, other elite elements and traders. The ordinary people who supported the massive superstructures of Buddhist Angkor lived outside the urban core in the rural hinterland. Coningham and Gunawardhana have applied Fletcher’s model of low density urban settlement to suggest that the Buddhist monasteries of the Anuradhapura hinterland functioned as major administrative centres, not merely as sacred sites for religious contemplation (Coningham and Gunawardhana 2013: 461-464). Their meticulous analysis of 706 inscriptions in the Anuradhapura hinterland (Coningham and Gunawardhana 2013: 445) recorded land grants and immunity from taxation. This led them to define power as a decentralised system of urban-rural controls in which the ‘King’ was not omnipotent. Paradoxically, the initial generosity of the King led to a situation in which land grants and tax exemptions to monastic orders ultimately undermined secular authority in favour of religious power (Coningham and Gunawardhana 2013: 450-451, 462-3).

Comparisons with the Buddhist Sites of South and Southeast Asia Ambitious studies have been made of the three ancient Buddhist sites of Anuradhapura, Sanchi and Angkor. These, together with observations about the Buddhist sites of western India, provide models for developing a research agenda on Buddhist Kashgar. The multi-volumed reports about Anuradhapura, the Lankan capital for some 1500 years from c.400 BC to 1100 AD (Coningham and Gunawardhana 2013) use methodologies that incorporate many non-invasive and invasive techniques that are equally relevant to study of the known and unknown parts of Buddhist Kashgar (see for example, Chapter 3 of Coningham and Gunawardhana Vol. 3, 2013: 21-31). In particular, the chapter on ‘Environment and Water Management’ (Coningham and Gunawardhana 2013: 191-228) informs us about the kind of variables we need to consider when connecting the patterns of urban development to the land management systems of the hinterland that sustained the urban centres and religious sites of Kashgar. Shaw’s earlier work on the irrigation systems of Sanchi also discussed the technologies of water management and linked these to pre-Buddhist cultural practices. Most important of all, she documented how the urban demands of Buddhist Sanchi compelled the community to change its cropping patterns to support a greatly expanded population (Shaw, 2016, Chapter 14). These studies provide frameworks for analysing how Buddhist Kashgar managed its own natural environment to deal with the influx of Buddhist monks, teachers and pilgrims, as well as the traders and royal patrons they attracted.

Shaw’s analysis of the Buddhist monuments of Sanchi in northern India provides another example of the way in which archaeological research is reordering our knowledge of pre-modern political economy (Shaw 2000 and 2009; Shaw and Sutcliffe 2001; Shaw et al. 2016). She has noted, for example, that religious buildings were not the only monumental structures in the Sanchi landscape. Equally important were irrigation works which could be larger than ritual buildings and demanded great investments of materials, human labour and managerial skills. She writes that ‘[a] preliminary comparison between the Sanchi dam and Stupa 1 on the hill above shows that the volume of the dam is approximately nine times that of the latter’ (Shaw 2016: 252). Supporting the growing Buddhist population and culture, in other words, put great pressures on both secular and religious authorities, as well as on the local residents and the environment. Nevertheless, such organisational efforts facilitated the conversion of the normally arid agriculture of the Sanchi region into a wet rice agrarian system. The higher yields of rice crops, compared with the yield of rain-fed traditional wheat crops made it possible to deal with the increasing population density associated with the elaboration of the Buddhist culture of Sanchi. When Buddhism and Sanchi then declined, the region reverted to rain-fed wheat cultivation that continues to define the region today (Shaw: 2016, 249). Shaw’s analysis of pre- and post-Buddhist Sanchi, provides us with a measure of the innovations that were urgently needed to support the expanding population that came with Buddhism. In contast to Coningham and Gunawardana,

Another relevant study is that by Roland Fletcher and his team working on Greater Angkor. Since a seminal paper published in 2009, Fletcher’s notion of a low density urban settlement pattern for civilisations based in tropical forests has compelled scholars to look well beyond the familiar boundaries of ancient cities (Fletcher 2009; Fletcher 2012; Fletcher and Evans 2012). For example, the Angkor Wat site is now known to have 152

A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times Shaw placed less confidence in inscriptions which could only be afforded by the elites. Large donations were needed for capital works, but the daily maintenance of the sangha required regular, local and sustainable forms of support, especially food (Shaw: 2016, 51). In explaining how monastic orders were supported by the ordinary populace she emphasises the extent to which pre-Buddhist non-elite religious cults were absorbed into Buddhist practices. This gave the pre-Buddhist beliefs greater legitimacy and reduced the social gap between the monks and the local people (see also Bailey and Mabbett: 1993, 155-157). In addition to analysing popular, pre-Buddhist rituals, Shaw has reconstructed the settlement patterns of the general population to understand the possible interactions between the monastic and labouring communities (Shaw: 2016, Chapter 13).

related to the Mori Tim stupas? These karez appear to be different from the underground brick channels or qanat of northern Xinjiang: in Kashgar they seem to take the form of wells that are fed with underground water rather than the built, covered underground channels that define the oasis of Turfan. As in Central Asia, we surmise that the water in the karez used to rise to the surface via capillary action. The mismatch between the abundance of meltwaters in the local rivers and the needs of agriculture that we discussed above, suggests that these karez were a response to the fluctuating availability of water from river sources in Kashgar. A second possibility is that the course of the local river might have shifted. There is some evidence for the shifting contours of the Chakmak River in Kashgar.18 But again, what we do not know is whether these karez belonged to the Buddhist era or not.

Might these models of low density urban settlements apply to the Kashgar region? Perhaps, but the environmental conditions of the Kashgar oasis were radically different from the monsoon tropical forest areas that Fletcher, Coningham, and Gunawardhana have described. The oasis was also different from the arid region of Sanchi studied by Shaw. The mountain meltwaters that feed Kashgar’s rivers may appear, at first sight, to provide a more reliable source of water than temperamental monsoons. Commenting on the water resources of ‘Chinese Turkistan’ and its ‘extreme oases’, Lattimore noted that ‘The water must come from a river and it so happens that the water is most plentiful at precisely the time when it is most needed … At the same time the crops are almost completely free of the danger of damage by unseasonable, unwanted rain’ (Lattimore 1940: 156-157). Contemporary research disputes this easy picture for farming in the Kashgar oasis. According to one report some 93 per cent of Kashgar’s water supply today is provided by the meltwaters in the river system (Huang et al. 2005: 209 Table 2). However, Huang et al. also show that water availability is not optimum for crop production for there is a mismatch between the arrival of meltwaters in June-August and the needs of crops which are planted and harvested in March to June and again in September. Today, agriculture in Kashgar depends on the water resources not only in the rivers but also in reservoirs in March, soil-water in April to May and groundwater from March to May (Huang 2005: 213215). Our question is how did Buddhist Kashgar respond to these environmental challenges?

In the next section we speculate about the socioeconomic base of Buddhist Kashgar and the changing needs of the Buddhist communities as they shifted from mendicant Buddhist practices to monastic landlordism.

Di Castro and Vicziany found evidence of the remnants of what appear to be an extensive traditional Central Asian irrigation system (karez) lying near the Buddhist site of Mori Tim (Vicziany and Di Castro 2016). When these karez were built is the critical question and for this proper dating is needed. After that we can ask what role did these karez play and how, if at all, are they

The caves of the Three Immortal Buddhas, on the Chakmak river, opposite Yuqi Moriwan (Pelliot’s Utch Meravan), are the only remaining, visible religious

Conjectures about the Socio-Economic Base of Buddhist Kashgar Early mendicant Buddhism in Kashgar The political economy of the first wave of itinerant, mendicant Buddhism into Kashgar probably made only minimal demands on the society and environment. The long history of Kashgar’s trade meant that local people were most likely accustomed to providing goods and services for visitors and itinerant merchants. At a later date, it is possible that Buddhism was officially promoted in Kashgar by King Chenpan in the first quarter of the 2nd century AD. As noted by Vicziany and Di Castro in the previous chapter, the Kushans kept a royal hostage from Kashgar in a Gandharan Buddhist monastery (possibly Shotorak, not far from modern Kabul; see chapter 8 note 18). Given this, it is reasonable to conjecture that the Kushans were also attempting to influence the local politics of Kashgar through their backing of Buddhist institutions. Kushan patronage of the King of Kashgar and Buddhism more generally in the oasis is not incompatible with other evidence suggesting that the earliest Buddhist monks might have arrived in Kashgar in the company of traders of different ethnic backgrounds from the eastern parts of Central Asia (Sen, 2012: 15-16).

There may well have been three different sites for the core of the Kashgar settlement in the long run, for there is evidence of the Chakmak River having shifted at least three times, perhaps in response to critical events.

18 

153

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 12. The caves of the Three Immortal Buddhas, on the cliffs above the Chakmak river (photo Gollings).

purposes of serious research. The presumed age of these caves alone justifies a closer analysis of what they represented.19 The caves have heavily damaged wall paintings which are certainly worthy of further study if we assume that Buddhism could have entered China via the west or south-west.20 Other kinds of evidence are equally important, including the study of how these caves were used and how they were reached given their great height. An examination of photographs of the caves show regular square cavities (mortises) on the rock cliff that probably accommodated the beams of the scaffoldings that were used as stairs to reach the Figure 13. Plan and cross section of the caves of the Three Immortal entrance of the caves. A careful Buddhas surveyed by Pelliot in 1906 (from Pelliot 1908). subsurface analysis of the debris accumulated at the foot of the cliff where the caves are located might yield useful data caves of Kashgar (Figures 12, 13). Could these have about the history of the caves’ occupation. served as the early residences of the first wave of monks who took temporary shelter in caves and 19  Recently, Marylin Rhie has hypothesised that the caves might date used them as their base for meditation. Pelliot (1906) to the 2nd-3rd centuries AD: ‘…the shape of the three doorways published a detailed description of these caves after follows the post and lintel form best known from Gandharan relief systematically surveying them on the basis of previous sculpture, such as seen in Figure 5.34c in Vol. I (Rhie 1999). This would tend to suggest an early rather than a later date for these caves, observations by Petrovsky, the Russian Consul of perhaps c. 2nd-3rd centuries AD’ (Rhie 2018). Kashgar (1882-1902). Today, the caves are virtually 20  Pelliot (1906 Figure 2) published a plan of the caves. The ignored by the tour guides because it is said that there conventional dating of the caves ranges from the 3rd-5th centuries AD to the Tang period (see also Di Castro 2013: 9). For illustrations is nothing much to look at. They are also very difficult of the interior of the caves see also Xinjiang Cultural Relics Bureau to reach. However, the demand by tourists for dramatic et al. (eds) 1999: 269; Xinjiang Annals Editorial Board (eds) 2007: 19 art and architectural forms should not determine the figure 89. 154

A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times The political economy of monastic Buddhism Over time monastic institutions began to emerge in the oasis alongside these modest cave-based residences. The caves themselves may have continued to attract more ascetic Buddhist sects that wished to remain relatively isolated from the general population. The arrival of monastic Buddhism created a new political economy in the region. According to Zürcher (2012), for example, after c. 300 AD monastic Buddhism began to be transmitted into China by Indian and Kashmiri monks. Zürcher provides us with a relative date to set alongside the absolute dates we hope to establish for the Buddhist monuments of Kashgar. The construction of large stupas and monasteries needed an agricultural and industrial support base. Capital investment for construction then had to be supplemented with capital for the maintenance of the infrastructure and the Buddhist sangha (priestly community). Donations of land, equipment, food, clothes and labour services were needed in addition to luxury items which played prominent roles in Buddhist rituals – silk banners, for instance (Liu 1988: 92-102). In return for these donations the monasteries reciprocated by providing a wide range of services that were valued by local elites and ordinary people. Borrowing from Shaw, we can assume that in Kashgar, as in Sanchi, over time the growth of the monasteries meant that the sangha became more integrated with local communities and beyond. Providing medicines and funeral rites, the monasteries became ‘agents for development and the alleviation of everyday suffering (dukkha)’ (Shaw 2016: 262). As in Sanchi, it was probably the growing interdependency between the sangha and the people that promoted Buddhist ideology rather than any religious conversion processes.

Figure 14. Fragments of a Buddhist manuscript discovered in Yawaluk (from Xinjiang Cultural Relics Bureau et al. 1999).

of Sanskrit manuscript leaves and impressions left on clay lumps were recovered in 1997 from a pottery container in Dakiyanus/Yawaluk. These are on display at the Kashgar Museum22 and they have also appeared in official Chinese publications about important discoveries in Xinjiang. Figure 14 shows a bark leaf probably inscribed with the Prajñāpāramitā sūtra (from the Mahāyāna school). According to Professor Francesco Sferra, these texts may be related to a Gilgit manuscript studied by Tucci and dated to the 7th and 8th centuries AD. If so, they probably reached Yawaluk from Gilgit or Kashmir.23 Similar manuscript fragments also exist in the Petrovsky collection in St Petersburg, for Petrovsky (the Russian Consul in Kashgar in the late 19th Century) was an avid collector. These also await systematic study because we do not know whether they originated in the oasis of Kashgar or elsewhere (Oldenburg 1983). Monastic – lay relations One of the key points that needs explaining are the labour systems that enabled the maintenance of the monasteries and their considerable residents. Perhaps some monks worked as labourers in either the fields or artisanal workshops, for we have evidence from Dunhuang suggesting this possibility (Gernet 1995: 96; Chen 1973: 137, 146). At the minimum, the monks must have been routinely involved in supervising agricultural and manufacturing operations, recruiting and training labourers and novices from various sources. Perhaps different kinds of labour worked side by side – monks, novices, peasants, pilgrims, penitents, contracted and volunteer labour. In other words, the prohibition against manual labour in some Buddhist texts should not be taken literally (Gernet 1995: 70, 95, 100, 103, 119; Chen 1973: 141, 146, 148 ff.; Liu 1988: 152153; Shaw 2009: 32 ff.).

In South Asia, many Buddhist buildings and the sangha developed thanks to generous royal land grants. Later on, grants by wealthy families increasingly replaced royal patronage as the Kings declined in importance or turned away from Buddhism. No doubt similar processes were at work in Kashgar but we do not know who the donors were – in the case of India and Sri Lanka these are revealed in many inscriptions.21 Hitch (2009: 5) has gone so far as to say that Kashgar and Yarkand ‘…are exceptional in the basin in that they have had significant populations in the historical period yet have offered up no documents of any kind from the pre-Turkic period, not even Sanskrit or Chinese ones’. However, this is not strictly the case as fragments of Buddhist manuscripts were discovered in the monastic site of Yawaluk/Dakiyanus (Tegurman), to the north of Kashgar city. It is also worth noting that fragments

22  Visit by Di Castro and Vicziany to the Kashgar Museum in October 2013. 23  Personal communication from Professor Sferra, 5 October 2015. The fragments of this manuscript (see Figure 14) were originally published in Xinjiang Cultural Relics Bureau et al. (eds) 1999: 267, figure 0729.

Among the many publications that deal with the inscriptions in the Buddhist chaityas of India see, for instance, Kosambi 2009 and Nagaraju 1981, and in the case of Anuradhapura in Sri Lanka, Coningham and Gunawardhana 2013: Vol. 3, 452-457.

21 

155

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Food and non-food goods may have been supplied by local farms or villages. Alternatively, the monasteries may have controlled farming land. Our suggestion at this stage of our research is that the area in the immediate vicinity of the Mori Tim stupa may have been cultivated by, or on behalf of, the Mori Tim monastery. Observation by one geographer from the Xinjiang Normal University noted that the soils were very fine, a characteristic of old cultivation.24 Again careful study of these propositions is needed through scientific geochemical analysis, excavation and the dating of the karez and canals.

Khano-oi) but without dating we cannot say which of these, if any, existed during the Buddhist era. Solving these questions will be greatly aided by dating the Buddhist structures, the karez, the canals and the agricultural soils. At this stage we can only report that on the Mori Tim site, there are the remnants of three karez lines: the original line of 48 wells that Di Castro discovered and then another two lines that Di Castro and Vicziany visited in October 2013: one line had three remaining karez and the other, six. These last two lines lie outside the protective fencing that surrounds the southern part of the Mori Tim site and in contrast to the first line of karez they had been greatly disturbed by various kinds of excavation projects. But we have enough evidence so far to dispute the view of Hitch (2009: 3) that the karez technology is unique to Turfan.

Studies of the Anuradhapura site provide one model of how Buddhist institutions developed. A quantitative analysis of the epigraphic record from the early historic (c. 340 BC to 200 AD) to the medieval period (c. 600 AD – 1200 AD) led to a careful analysis of the distribution of power (Coningham and Gunawardha 2013: 445-457). At first, secular kings granted land to the monasteries but over time relations of production in the hinterland increasingly fell into the hands of the monasteries. The monasteries may even have been the authorities under whose auspices the agricultural frontier of Anurdahapura continued to expand. Some aspects of this model may well have applied to Kashgar although the relative scarcity of land (defined by the boundaries of the oasis environment) may have made land grants less dynamic than in the case of the Anuradhapura hinterland.

Most likely the karez that we can still see today adjacent to the Mori Tim site were part of a much larger irrigation system. Our hypothesis is based on new discoveries on the other side of the Chinese border. Recent geo-archaeological and aerial research methods in Kazakhstan have revealed an extensive karez system of 9,000 wells organized in 261 karez lines for a length of 124 kms in the Sauran desert region. New mapping techniques have generated a massive increase in the original three lines that up to 2004 were thought to define the traditional irrigation system in that area (Sala and Deom 2008).

The oasis environment

The Mori Tim site raises another issue: during our visits in 2005 and 2013 the ground glistened with salt. When the local farmers tried some decades ago to extend their farms and orchards towards the Mori Tim site, they discovered that farming there was not sustainable. This compels us to ask: what was the nature of the soil during Buddhist times? What kind of food and non-food crops were grown? If we find evidence of cotton cultivation, this might have led to the soil exhaustion cotton is known to cause. Might the combined environmental pressure of intensive food production and cotton cultivation have depleted the land and caused critical levels of soil salinity?

Whatever the labour and tax arrangements, the Kashgar oasis provided an ideal base for Buddhist establishments because of its environmental capacities and international trading connections – the next two factors that we consider. The settlement pattern of the oasis shows that in the centre of the Kashgar oasis there are even today extensive orchards and farming lands that fan outwards from the periphery of the town into the surrounding lands which then lead to the remaining Buddhist monuments at the edge of the cultivable area. The distance today between the food supply and the stupas and their surrounding structures is relatively short, not exceeding 15 kms (corresponding to a three to four hour walk). For the pre-Islamic era, however, the location of Kashgar’s urban core relative to the monasteries and cultivated lands remains to be ascertained. As we noted in in the first parts of this chapter, old Kashgar had a number of urban centres (Kashgar, Shule, Eshkishahar and

One of our research priorities is to reconstruct the ancient landscape and hydrology of Kashgar, including the land and water management practices. We should not assume that the decline of the Buddhist monuments and cultural centres was the result of religious wars between the Buddhist princes and the newly converted Muslim ones. Perhaps an environmental crisis also contributed to the decline of the Buddhist establishments? Soil analysis based on core sampling could reveal the nature of the agrarian evolution in this region over the long run. Optical Stimulated Luminescence (OSL) can be used to date the sediments

24  Professor Chen from the Xinjiang Normal University accompanied Professor Vicziany to the Mori Tim site in April 2013 to verify that the 48 wells (karez) discovered by Di Castro (who had studied satellite images) still existed on the ground. Qadir, the former Director of the Kashgar Museum, also told us in October 2013 that he was aware of the karez on the Mori Tim site.

156

A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times in the older river beds and artificial channels. New microwave remote imaging together with pollen analysis would help us to construct a comprehensive picture of agriculture at the time when the Mori Tim and Topa Tim sites were international attractions for a multitude of regular, foreign pilgrims, scholars and traders.

We have assumed that wheat production dominated, but this needs to be demonstrated against archaeological and paleobotanical evidence. What if wet rice cultivation became the norm in response to the population pressure that built up as Mori Tim, Topa Tim and other religious sites grew in importance? Alternatively, if wheat did remain the dominant cereal how extensive did agricultural lands become? Many other possibilities also need to be considered: did agriculture in Kashgar involve mixed farming techniques and crops? In building our model for the Kasghar oasis we should avoid any deterministic focus on monocultures. For example, in the Arundahapura and Angkor Wat hinterlands swidden cultivation preceded and continued alongside other farming techniques (Gilliland 2013: 1019, 1026; Fletcher 2012: 307).

This brings us to a further critical question: what technology was used to produce food in sufficient quantities to sell to passing merchants in addition to supporting the hundreds and perhaps, as Xuan Zang reported, thousands of monks? We believe that wheat was probably the main staple followed by some rice and millet. This assumption is based on deductions derived from current information about cereal output in the Kashgar prefecture today. In 2011, total grain production was 2.575 million tons almost equally divided between wheat (1.232 million tons) and corn (1.211 million tons). No official figures have been given for rice and sorghum production but deducting wheat and corn output from total production suggests that rice and sorghum yielded only some 132,000 tons of food in 2011 (Kashi Prefecture Statistics Bureau 2011).25 Driving through the orchard of Beshkerem in April 2013, we noted that the predominant food crop was wheat interspersed with grape vines and fruit trees. On the other hand, we have to seriously consider the possibility that rice was either grown in Kashgar or imported in considerable quantities, given the importance of rice in Buddhist rituals (Shaw: 2016, 250-252).

Other crops may also have been important for humans and animals. Even earlier forms of wheat persisted and were important as animal feed. We also need to consider the possibility that millet and barley were cultivated during the Buddhist period. Certainly millet is a very attractive crop because of its risk averse properties: it is fast growing, able to handle uncertain water supplies and is salt resistant. As such it seems ideally suited to the salinity of the soils that we saw. But as noted earlier, how far back the problem of salinity stretches remains unknown. Similarly, we can postulate that perhaps some barley may have been cultivated in the Kashgar area (Personal communication from Dr Zhao, December 2013), especially on higher lands where less water is available and less irrigation possible.

Early historical sources (such as the travels of Faxien in 414 AD) confirm that Kashgar was already a wheat producing oasis that also yielded fruits. Wheat, in contrast to wet rice cultivation, requires less water so we can surmise that the irrigation system of Kashgar was probably canal based rather than tank, reservoir or dam based. The area surrounding the Mori Tim stupas has many irrigation canals that run from the now dry bed of the Chakmak River to the old fields. On the other hand, the evidence of the existence of karez technology suggests that there were problems of accessing river water if only uncovered canals were used. Wheat was probably the dominant crop, but some small amounts of rice appear to have been grown in the Kashgar oasis for a long time.26 During the Buddhist period, minimal quantities of rice would have been needed to satisfy the demands of the Chinese garrison, some traders and Buddhist rituals.

Irrigation systems The water management systems of western China consist of five kinds of technology: tanks, wells, uncovered canals, underground and covered canals or karez/qanat and the underground water wells (also called karez locally) that we have discovered. As noted in the previous chapter, managing the water supplies in the oasis is recorded in the many legends about Bibi Maryam during the Islamic period.27 There are other stories too about old urban wells in the towns of the Kashgar prefecture. In old Kashgar city there is a well with special properties not far from the main road. It was still used as a source of drinking water and for washing clothes in 2005. The well survived the recent demolitions to this part of old Kashgar when we visited the area in 2014 (Figure 15). Although the well was probably no older than the 16th century AD (the likely date of ‘old’ Kashgar’s foundation) its survival provides evidence of the

We thank Dr Guibin Zhang (Monash Alumnus) for locating and translating the relevant parts of this source. 26  For example, a confidential Military Report on Kashgaria prepared by the British Intelligence Branch in Simla (India) contains a map by Captain A.R.B. Shuttleworth showing rice fields along the Tumen River in 1908 (Tamm 2012). The map was found by Tamm in the Political and Secret Department files of the Indian Office Records in the British Library, London. 25 

See the previous chapter in this volume in the section on The Islamic records on Kashgar.

27 

157

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads (dombel) were used to ensure more reliable water supplies for irrigation systems. Bertrand believes that tank irrigation systems also started in Central Asia especially in the time of the Kushan Empire.30 Then this technology could have migrated into northwestern India and from there into China.31 Moreover, many tanks were built in Gandhara and the Tarim Basin close to monasteries (Bertrand 2012: 41-42).32 The sites where tanks have been found in the Tarim Basin are Khotan, Charkilik, Rawak, Loulan, Niya, and Turfan (Bertrand 2012: 43-46). Despite his innovative work, a lot of suppositions stand behind Bertrand’s speculations and many of them cannot be verified. No reports of ancient tanks in the Kashgar prefecture have yet been found by us although there are many tanks scattered throughout the countryside, as well as in some urban areas, including the large tank in front of police headquarters in the ‘old town’ of Kashgar and two tanks outside the monumental complex of the Apa Khoja mausoleum. Certainly, the image of the Kashgar oasis as a major Buddhist cultural centre with large numbers of stupas, monasteries and Buddhist followers suggests to us the possibility that water tanks did exist for irrigation and for personal use, although the ritual and cultic practices that Bertrand surmises cannot be proven.33 Nevertheless, if the technology of water tanks is closely associated with monasteries in Taxila, Khotan and elsewhere as Bertrand (2012: 51) notes, then it is likely that they also existed in Kashgar for ‘[s]uch structures are very similar from one oasis to another’. In addition to tanks and wells, we need to

Figure 15. Water well in old Kashgar in 2014 (photo Di Castro).

general value of wells even inside the urban centres of the oasis. Despite the importance of water sources, we have not found many studies about them. The exception is the work of Bertrand (2012) who investigated the tanks of the Jingjue Kingdom at Loulan on the far eastern side of the Tarim Basin. Bertrand’s work is brimming with ideas but remains essentially speculative and even contradictory. He suggests that the irrigation systems of Jingjue (perhaps more generally in western China) were indigenous responses to environmental challenges rather than imports of Chinese technology. Chinese technology could have entered western China with the military campaigns waged on behalf of the Chinese emperor (e.g. c. 128 BC). Sustained campaigns needed reliable food supplies best provided by military farms (tuntian).28 Bertrand (2012: 35-37) considers this transmission route but discounts it as the source of wider dissemination. Rather he (2012: 47-48) suggests that indigenous technologies were developed perhaps with some admixture of imported ideas from Gandhara in Afghanistan and Pakistan.29 However, Bertrand wavers on the relationship between indigenous and imported technologies. Exactly how local technologies were invented is not addressed; instead he (2012: 38-39) stresses that in Central Asian cultures tanks and special devices for measuring water levels

On this point see also Mukhamedjanov 1994: 265-272. Bertrand’s explanation of migration is somewhat confusing: he does not make a clear distinction between the migration of technology and the migration of people. On the one hand he (2012: 3, 5-6, 35, 44, 46-51, 53-54) claims that there was mass migration from Gandhara in the 3rd century AD following the Sasanid conquest and on the other he (2012: 51) also admits that this is a supposition. Bertrand (2012: 46) is also vague in his use of the notion of ‘Indo-European development’ and the ‘Gandharan administration system’ and gives no particular linguistic, cultural or historical references. By contrast, Harmatta (1994: 434-435) more simply admitted that the dissemination of Gandhari script in the Tarim region could have happened when members of the Kushan administration originally based at Balkh fled into western China after the Sasanian invasion of 245 AD. In addition to this, there is also the possibility that a ‘diaspora’ of merchants from the western regions carried specific technologies and ‘know how’ into the southern Tarim Basin. 32  Bertrand (2012: 42-43) refers to Xuan Zang’s account of ten tanks ‘secretly connected together’ near a large stone stupa in Taxila in 630 and 645 but unfortunately there is no archaeological evidence for this. 33  Religious rituals connected to water tanks are well documented in the Indian subcontinent. For instance at Sanchi, Shaw and Sutcliffe (2001: 68-71) identified a conspicuous number of dams and tanks associated with local Naga cults. The depiction of serpents not only connected Buddhism to the pre-Buddhist beliefs of local people but also gave the Naga cults new roles as protectors of Buddhist relics (Shaw 2016: 138-139). However, in the Tarim basin, apart from a unique representation of a Nagini emerging from a square tank depicted on a wall painting in Dandan-Uiliq (a reference to a narrative connected with a local legend (Stein 1907: Vol. I, 253, Vol. II, pl. II; Bertrand 2012: 480)), we are not aware of any such a cultic association with the tanks in western China. 30  31 

The best study on the tuntian system is by Luo et al. 2017: 79-88. See also Ma and Sun 1994: 240-243. 29  On the possibility of locally developed technologies see also Gilliland et al. 2013: 1013, 1026. 28 

158

A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times

Figure 16. Water logged fields around Topa Tim in 2005, the stupa is hardly visible on the horizon line in the middle of the image (photo Di Castro).

map the old uncovered canals of the Kashgar oasis – we see evidence of these on most of the stupa sites. Only a systematic survey (both aerial and ground recognition with soil sampling for analysis) of the Kashgar oasis might reveal what lies beneath the surface of the land.

stupa had resulted in the ground around the stupa rising by some 15 feet (c. 4.6 m). He (1907: Vol. I, 78) also commented on siltation in the case of Kizil Debe. Could that siltation have contributed to the decline of these structures even before they were overtaken by the urban expansion of Kashgar?

All the above named water technologies for irrigation and personal use most likely co-existed and overlapped. The exact nature of Kashgar’s old irrigation system would have be determined by a number of factors including the needs of the food and non-food crops, the availability and sources of water, the distance between the water sources and the crops, the distance between the water and the stupas and monasteries, the topography, the contours of the runoff and many other variables. The volume and flow of runoff from the Tianshan Mountains was also a critical variable, as it was these meltwaters that fed the local rivers and determined the nature of the groundwater supplies and the moisture content of the soil.

We also need to abandon the assumption that land was a stable factor. Kashgar is prone to earth movements because it lies at the meeting point between the Tianshan and Pamir Mountains. Instability is also caused by the Kashgar-Yecheng fault system. The earliest local earthquake recorded by modern equipment in the Kashgar area occurred in 1902 and measured c. 7.7 in magnitude (Kulikova 2016: 68 and Appendix E1). Earthquakes are quite common in the Kashgar area, although they do not necessarily cause deaths if the epicentre is in a remote area. On 11 August 2012, however, an earthquake with a magnitude of 5.2 destroyed some 30 ‘housing units’ and injured about 26 people (AFP 2011).What impact did events of this kind have, over the long run, on destabilizing large historical monuments such as the stupas? What has been the long-term history of earthquakes in Kashgar? Natural disasters are as much a factor of destruction as manmade wars. Recent work on the decline of Angkor Wat, for example, has suggested that the environmental burden of that civilization was so great that it proved unsustainable and collapsed under its own weight rather than as a result of attacks by the Thai (Fletcher and Evans, 2012: 61). The international significance of these findings are considerable: in getting our history straight, we can avoid blaming people for events that were beyond human control (Vicziany 2012: 5, 9).

We also note that in parts of the oasis there is sometimes a lot of surface water. For example, during our visit to Topa Tim in April 2005 we found it surrounded by water logged fields (Figure 16). Stein (1907: Vol. I, 80) also commented on the excessive surface waters surrounding the Topa Tim site but the causes for it then and now are not known. As noted in our previous comments on Lattimore, the hydrology of the Kashgar oasis is complicated. Our research questions are not about the scarcity or abundance of water in general but rather when water is available, how it is accessed and how availability matches the needs of the oasis settlement. Information about the nature of non-cultivable land in ancient Kashgar is also needed as much as the study of arable lands. Stein (1907: Vol. I, 76) noted that excessive siltation in the case of the Kurghan Tim

International and regional trade In many parts of India, Buddhist orders chose to locate their institutions in close proximity to trading routes 159

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads (Nagaraju 1981: 12).34 The Kashgar sites appear to repeat this pattern. The passing traffic of merchants, moneylenders, caravan owners, bankers and other commercially minded interest groups would have provided the monks and nuns with essential sources of income both in the form of donations and the sale of goods and services by the monasteries. In Nagaraju’s (1981: 26-27) analysis of 156 temple inscriptions in the rock cut Buddhist temples of western India, 35 per cent of donations came from traders followed by royalty (23 per cent), bureaucrats and clergy (12 per cent), peasants and artisans (10 per cent) and foreigners (8 per cent).35 In 1957 Kosambi (2009: 462) also spoke of the special role that the Buddhist orders of western India played in the regional economy. He even suggested that the monasteries provided loans at interest, using their donated and earned wealth as the basis for further capital accumulation. Dehejia (1972: 144) envisaged the role of monasteries as centres for the redistribution of money and food surpluses to local communities in times of dire need. A special nexus developed between monasteries and traders with the latter selling grains, food and other supplies on behalf of the Buddhist sangha during times of great public scarcity. Ray (1986: 207; 1994: 136-142) has also described similar commercial and social networks.36 Kashgar was a perfect location for this kind of monastic entrepreneurship, being the connecting hub of the northern and southern branches of the Tarim Basin’s Silk Roads and the crossover point between the eastern-western and northern-southern sides of major trading routes. Whether such monastic entrepreneurship developed in Kashgar and how it related to monastic landlordism and relations between the monasteries and local, secular powers, is something that we have yet to establish.

model of ‘low density urban settlements’ that our understanding of all these ancient cultures will remain severely constrained until a holistic analysis of the cultural-political-economic-social environment is completed. Assumptions made by previous generations of scholars are being revised and even overturned. The extent of these civilisations and their impact has been underestimated. In the case of Angkor Wat for example, the narrow focus on Siam Reap had distracted attention from the wider parameters of Khmer civilization. In the case of the Coningham and Gunawardhana project, we now have extraordinarily detailed data and scientific measurements that demonstrate the complexity of the Anuradhapura polity in pre-modern times. Without systematic and multi-disciplinary expertise of the kind that has been brought to bear on Anuradhapura, this would not have been possible. In the case of Sanchi, earlier studies assumed that Buddhist monks did not engage in agricultural labour (although see Chen 1973: 146 ff. for a different view). The methodology developed by Shaw (2009: 233-234) has rewritten the previous history that was based on a narrow interpretation of religious texts that seemed to promote religious meditation at the expense of practical living. Apart from the urgent need for archaeological investigations to resolve some of these issues, we need a detailed hydrology of ancient and modern Kashgar to understand the constraints on agriculture in the oasis and its carrying capacity. Kashgar and Early Human Settlement in the Tarim Basin Some of the world’s most important ancient archaeological sites have been found in the Tarim Basin. At Xiaohe, Keriya, Yuansha, Karadong, Dandan-Uiliq, the Northern Cemetery or Beifang Mudi (see Abuduresule in this volume) and other sites, naturally formed mummies have been re-discovered displaying physical and genetic features that speak to the admixture of the peoples of China with the peoples of the west (see Abuduresule in this volume; Li et al. 2010; Mair 1998; Mallory and Mair 2000). The Xiaohe mummies are amongst the oldest, naturally preserved mummies in the world (Deter-Wolf et al. 2016: Table 1, 3)37 and again attest to the role of Xinjiang as a transition zone between Central Asia and China when it comes to human migration and the transfer of technologies, beliefs and material cultures.

In summary The insights gained from the above studies on India and Sri Lanka, can provide a comparative framework for our study of early Buddhism in Kashgar. We know from the work of these authors and also Fletcher’s One of the most impressive Buddhist rock cut temples is that of Karla. Today, the Karla temple is a long way from modern lines of communication, but in 60-30 BC when it was built (Nagaraju 1981: 227) it was part of a group of three temples that straddled a major trade artery through the mountains of the Western Ghats to the river ports and coast of western India. Siltation converted the river ports into inland towns. The old trade routes have, as a result, long fallen into disuse (Ray 1994; Vicziany et al. 2017). 35  Inscriptions in the Karla rock temple for example, give us the names of some of the donors who financed the construction of the temple – these were lay worshippers or almsmen, a carpenter, a nun, a perfume merchant, the son-in-law of a King, the son of a trader, reciters, mothers/aunts, a physician, and a traders’ association (Kosambi 2009: 463-473). An alternative model of financing the monastic-stupa infrastructure could have been land and other grants to the monks by secular authorities as happened in Anuradhapura (Coningham and Gunawardhana 2013). 36  On the socio-economic interactions see also Heitzman 1984; Chakrabarti 1995; and Morrison 1995 and 1997 all of which focus on Buddhist archaeological and epigraphical evidence. 34 

Unfortunately, no mummies have been found in the Kashgar region so the rich material evidence provided by the graves of the mummies in the other parts of the Tarim Basin, are yet another missing link in our 37  Otzi, the Tyrolean mummy, is the oldest human with tattoos but the oldest mummies date from about 5050 BC and belong to the Chinchorro culture of northern Chile (Deter-Wolf et al. 2016); Interview with Dr Silva, Chile, 17 October 2017.

160

A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times understanding of the historical evolution of this area. However, at Upal, Stone Age tools with estimated dates from 4000 to 3000 BC have been identified (Li 1962; Debaine-Francfort 1988: 9, 12; Xinjiang Institute of Archaeology 1995b: 27-40).38 A number of funerary remains have also been discovered at Xiangbaobao and Xiabandi (Shambabay, Xiabandi / Xiabadi), near Tashkurgan (Taxkorgan) dated from c. 1700 – 1000 BC down to the Qing dynasty (17th to 20th centuries AD). From these we know that groups of Saka horsemen and Qiang, proto-Tibetan tribes, probably intermingled at the fringe of the oasis during the first half of the 1st millennium BC (Chen 1981; Debaine-Francfort 1989: 200-201; Ma and Wang 1994: 210-211; Han 1994: 4; Han 1998: 563; Han 1999: 230-231; Wu 2005; Mallory 2010; Mallory and Mair 2000; photo in Xinjiang Cultural Relics Bureau et al. (eds) 1999: 281; Xinjiang Institute of Cultural Relics and Archaeology, 2012). There is also some archaeological evidence to support the existence of farming people connected to the Karadong culture, stretching from Kucha to Kashgar and the Pamirs from c. 850-750 BC (Mallory and Mair 2000: 145). According to Chen (1981) the Saka tombs dating from c. 500 to 400 BC contain human remains, sheep bones and burial items such as pottery. We can infer that the same Saka horseman who were in the Pamir Karakorum region around the first half of the 1st millennium BC (8th – 4th BC) interacted with other local populations and furthered their presence in the lowlands of the oasis (Narain 1990: 173).39 This might have happened when the higher forest areas were no longer suitable for the intensive pastures needed by their cattle, horses and sheep. The tombs at Xiangbaobao/Xiabandi, on the northern bank of the Tashkurgan (Taxkorgan) River have affinities with the Bronze Age Andronovo culture (Wu 2005: 109-113; Jia et al. 2017), whereas the tombs on the south bank could be from the Qing Dynasty or earlier, in the Islamic period.

Mallory and Mair (2000: 147-148) that ‘the earliest iron artefacts show greatest similarities with those to the west in Ferghana; it [iron] also predates the earliest appearance of iron in China’. The shortest distance from the Ferghana valley to western China is through the Kashgar oasis. Kuz’mina (2008: 114), when speaking of the 13th to the 8th centuries BC speaks of the Bronze Age as a time when the use of metals, horses, chariots and camels spread into Xinjiang from the west. Specifically, she comments on one small discovery to demonstrate the growth of the southern part of the Silk Road: ….The finding of a bispiral pin in Loulan allows one to assume also the establishment of Xinjiang’s southern relations either immediately with Bactria or, which is more likely, with Ferghana, and the inception of the Eastern Route’s southern sector (Kuz’mina 2008: 110). If this regular and extensive migration of peoples, technologies, ideas and livelihoods from Bactria, Ferghana and Sogdiana on China’s western border to sites on the far eastern rim of the Tarim Basin was taking place, it is hard to imagine how that was possible without travelling through the Kashgar oasis at least some of the time. In other words, the lack of graves containing mummies and samples of their material culture in the Kashgar region should not prevent us from imagining the role that Kashgar may have played in this extraordinary interchange of material cultures (see Francfort in this volume on the importance of the southern routes of exchange). If we look at the modern dynamics between the Kyrgyz nomads and the Wakhi agriculturists who live in the Wakhan region that spans western China and Afghanistan, we may infer some trends that may have characterised the early settlements of the Kashgar oasis and the regions at the foot of the mountains. Cribb (1991: 62, 63) argued that as a response to volatile social and environmental conditions, nomadic pastoralism usually alternates with phases of settlement and agricultural specialisation. Remnants of this temporary settlement pattern can be seen even today, with the people living in the Wakhan corridor cultivating barley that is possibly less productive than other grains but more resistant to the harsher conditions of high altitude and water scarcity (Shahrani 2002: 72). Barley for all these reasons is also a less labour intensive and, therefore, a desirable crop.

In the small museum in the compound of the tomb of Mahmud al-Kashgari in Upal (see note 39) there are ancient artefacts, including Bronze Age implements, from excavations carried out in the neighbouring sites. During the late Bronze Age iron appeared in the Tarim Basin c.1000 BC. Iron may have entered the Tarim region, and ultimately Central China, via Kashgar. We surmise this on the basis of a conclusion reached by 38  During our visits in 2005, 2007, 2013 and 2014 we observed material from the Neolithic period onwards displayed in the small museum in the compound of the tomb of Mahmud al-Kashgari. These had been excavated from archaeological sites in the vicinities of Upal. Among these there were microliths from Yakhotan, Yakhurgan and Hoja Khunar, plus copper and iron weaponry from Yakhotan and pottery and other material from the mazars of Hoja Sefuding Buzurkiwa, Layjilgha, Rpulat, Bashsuad and others. 39  The Indo-European Sakas are famous for their distinctive green eyes. In this regard we note that Xuan Zang described the inhabitants of the Kashgar oasis as having ‘tattooed their bodies and they have green eyes’ (Pelliot 959: 210). The people who settled in the Wakhan area had a similar appearance (Beal 1983: 293).

It is also possible to draw some hypotheses about early human settlement patterns in the Kashgar region by extrapolating from data that we do have about the cultivation of wheat on the periphery of western China. Betts speaks of the ‘arc of evidence’ for the arrival of wheat into the southern most western part of 161

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Xinjiang from the south, namely from Pakistan, north Afghanistan, south Tajikistan and eastern Kazakhstan between the 6th and 3rd millennium BC. These scattered sites suggest that in about the 6th millennium BC the culture of wheat growing and cooking was ‘carried over the mountains and into Xinjiang, probably filtering in from the south’ (Personal communication from Professor Betts, 2014). Wheat was also grown on the periphery of the far north of Xinjiang, in the steppes, but this wheat was typically Triticum dicoccum or hulled/emmer wheat while the wheat found in early Chinese sites is Triticum aestivum a free-threshed bread wheat (Betts et al. 2014; Spengler et al. 2014). The latter wheat species requires less preparation before cooking. This same wheat is the type associated with the sites in Pakistan, north Afghanistan, south Tajikistan and eastern Kazakhstan. By c. 2000 BC there is some evidence to show that a wheat culture had established itself amongst the people of the Xiaohe-Northern Cemetery area (see Abuduresule in this volume), so we can imagine that perhaps wheat entered into the Kashgar region some time prior to this. Wheat based cooked foods of various kinds including noodles and dumplings found along the Silk Roads have been dated to the 5th to 3rd centuries BC (Anderson 2010: 5). In other words, the paleobotanical and archaeological evidence thus far suggests that the southern parts of western China played an important role in the transmission of bread wheat cultures into China. This does not, however, exclude the possibility that wheat may also have arrived in China from avenues further north. We believe that there is no need to adhere to old fashioned notions about important innovations having only a single point of origin or entry.

Figure 17. Ruins of a Buddhist stupa and monastery (Toukguzkaznakh), near the tomb of Mahmud al-Kashgari in Upal (photo Di Castro).

the early historical period when Buddhism was already present in Kashgar, we know of the ruins of a Buddhist temple and monastery to the west of Upal on Haizilaitimaomu Mountain (Xinjiang Institute of Archaeology 1995a: 583-587).40 These ruins are still visible in the compound of the tomb of Mahmud alKashgari (Figure 17). Monumental remains such as these, once again suggest that Kashgar’s agrarian system had the capacity to support large populations attracted to centres of Buddhist learning.

Evidence of the cultivation of millet and wheat has been found in the western part of the Kashgar oasis, in the area around Upal (Wupaer). Based on new findings we can assume that some kind of cultivation was practiced at least at the time of the Early-Middle Bronze Age period, c. 1509-1318 BC (Personal communication from Dr Zhao 2013; Dodson et al. 2013: Table 1, 110). Of course, the role of trade in bringing wheat kernels to this region cannot be discounted (see Francfort in this volume). Trade, together with interactions between early agriculturists of the oasis and nomadic pastoral groups, likely happened in the early part of the 1st millennium BC. Although based on the material evidence from Upal (e.g. a Xiongnu type of bronze cauldron in the Kashgar Museum) we can postulate that these interactions occurred during the middle phase of the Iron Age, c. 7th to 5th centuries BC (see also Davis-Kimball 1998; Di Castro 2013: 7). Of course, it is also possible that the early agriculturalists and the nomadic pastoralists may have been the same people – we can no longer accept the old fashioned Marxist divisions of the mode of production as something that progressed along fixed trajectories. Later, during

Climate Change and its Impact on the Evolution of the Kashgar Oasis The human settlement pattern of Kashgar, its history and culture could also be better comprehended by studying the interplay between aridity and water supplies in the very long run. To this end we review various climate change models in this final section of the chapter. The Kashgar oasis is located on the western margin of the Tarim Basin at an average altitude of 1300 m asl. It is an ancient oasis of loess and alluvial soils. This monastery was dated to the 3rd-9th century AD, and described as ‘Tokkuz Kaznak Buddhist Temple’ (Toukguzkaznakh) on a worn out sign that was lying on the ground during our last visit to this site in 2007.

40 

162

A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times According to Hong et al., the evolution of the Kashgar oasis during historic times was driven more by climate change than human activities (Hong et al. 2004). The climate of the Kashgar oasis, as part of the Tarim Basin, has been arid since the late Pleistocene (126,000 to 11,700 years ago) (Zhu et al. 1980). The oasis is fed by surface water and groundwater originating from the snows and glaciers in the surrounding Tianshan, Kunlun and Pamir mountains.

and fall of dynasties. On the latter point, his work suffers from the same weaknesses that we pointed out in our previous chapter in this book where we suggest that many authors unconvincingly correlate prosperity with imperial expansion and centralised states. Might Hsu et al. be on firmer ground when it comes to the issue of mass migrations? Citing the mass migrations of the Germanic speaking groups in the 4th and 5th centuries AD, Hsu et al. suggest that global cooling forced them to move southwards because despite fertile lands, the cold and icy conditions reduced the length of the growing season making it virtually impossible to grow food. Until the 9th century AD when the Germans returned to their original homelands, these abandoned regions remained ‘no man’s land’ (Hsu et al., 1998: 692-693). Similarly, he speculates that the Indo-Europeans abandoned their original homelands around 2000 BC and again about 800 BC in response to climate pressures and migrated further south and east to Greece, Iran, Central Asia and then Xinjiang (Hsu et al., 1998: 694). The cycles of warm/ wet weather and cold/dry conditions in China agrees with the analysis of Kezhen Zhu, the author of the next model that we consider.

Three climate change models are worth considering: – that by Hsu and his colleagues in the 1990s, Zhu’s work in the 1970s and more recently the work of Zhao et al. (2012). The first two models are large-scale descriptions of general patterns of climate change over China as a whole, while the last one reflects historic climate changes in the Kashgar region. Can these models tell us something about the long-term character of Kashgar’s natural resource endowments? Model 1 Global climate change in the last 4000 years The first model is one relevant to not only China but more generally global climate change, with Kenneth Hsu and his colleagues from the Limnogeological Laboratory of the Swiss Federal Institute of Technology relating climate changes to the mass migrations of Europeans, including the Indo-Europeans who started to move into western China about 4000 years ago. He suggests that these migrants may be the same Caucasian types that populated the eastern part of the Tarim Basin and formed the various settlements that have now become famous as the sites of the mummies of the Taklamakan desert. Recent DNA analysis has located the origins of these peoples, prior to their migration into Xinjiang, in the southern parts of Siberia (Li et al. 2010). They may, of course, have been part of earlier mass migrations from further west. The problem with Hsu’s argument is that it does not recognize the mixed nature of the gene pool carried by the mummies. Analysis of the DNA of the oldest mummies of China shows that they contained genetic material from both eastern (i.e. Chinese) and western ancestors and that the admixture probably occurred in southern Siberia during the early Bronze Age and before they reached China (Li 2010: 10).

Model 2 Climate change in China in the last 4000 years Like Model 1, this second model of climate change by Kezhen Zhu generalizes about the alternating wet/ warm periods with cold/dry periods for the whole of China. Kezhen Zhu (the late, highly regarded Chinese climatologist) studied climate changes in China over the last c. 5000 cal. yr BP, basing his conclusions on historic plant phenological records. Unlike Hsu et al., however, he did not relate the impact of global climate change in China with what happened in Europe. Zhu (1972: 186-188), like Hsu et al., identified four warm periods and four cold periods. The four warm periods were as follows: 1.

2. 3. 4.

On the basis of long-term trends in temperatures and whether these were falling or rising, Hsu et al. (1998: 690-691), argued for a distinctive periodicity of about 1200 years in the last 4000 years, with 600 cold years alternating with 600 warm years. The minimum temperatures appeared in the years around 2000 BC, 800 BC, 400 AD and 1600 AD. Colder weather in Europe increased precipitation which made agriculture harder, while in China the effect was to increase aridity which created conditions also unfavourable to growing food for man and beast. Much of Hsu’s paper seeks to relate shifting temperatures to mass migrations and the rise

The first warm period, ca. 5000-3000 BP. In this era, the average annual temperature was higher by 2oC than that of modern times, and the average temperature in January was 3o-5oC higher. The second warm period, 770 BC to the 1st century AD. The third warm period, 600-1000 AD. The fourth warm period began c. 1900 AD and continued to c. 1960 AD. In this time, the snow line on the Tianshan Mountains moved up the mountain by between 40 and 50 metres, the glaciers retreated by between 500 and 1000 metres and the forest line also moved further up onto the mountains.

The coldest periods were around 1000 BC, 400 AD, 1200 AD and 1700 AD, with an average annual temperature being 2oC lower than that of today (Zhu 1972: 186 163

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads According to Zhu (1972: 186), warm and cool periods alternated every 400 to 800 years, and each cycle lasted for about 50 to 100 years. The work of Hong et al. (2004: 549) can be related to some of these cycles: they argue that during the first warm period water resources in the Tarim Basin were abundant and the oasis areas were much more extensive than in modern times. The third warm period again saw an abundance of water resources in the oases (Hong et al. 2004: 550).

• These high mountains harbour glaciers which, under favourable conditions, provide increased meltwaters to the Kashgar oasis; • The Kashgar oasis is fed by a river system characterised by numerous water courses, including the Kyzylsu/Chakmak/Kashgar, the Tuman and the Gaiz, which flow in an easterly direction down from their points of origins high in the Pamirs and West Kunlun Mountains; • The southern Kunlun mountains are so high that they prevent the Indian monsoon from ever reaching Kashgar – hence this monsoon is an unlikely source for any humidity/wetness in Kashgar (Zhao citing Li 1991); • The mid-latitude westerlies are probably the main source of precipitation in the Kashgar region – these winds carry moisture from the Atlantic Ocean, the Mediterranean, the Black Sea and the Caspian Sea (Zhao citing Aizen 2004).

Hsu, Zhu and Hong have all generated climate models that make assumptions about the aggregated nature of climate change throughout China. In models 1 and 2 there is no capacity for identifying regional climate models within China. Nor do these models have any explanatory mechanism for explaining exactly what drove precipitation. By contrast, the third model by Zhao and his colleagues is more sophisticated for the reasons we explain next. Model 3 Climate change in the Kashgar oasis

By matching their analysis of the pollen record with existing work on the mid-latitude westerlies, Zhao et al. confirmed the work of Seong et al. (2009) showing that the movement of the westerlies determined whether the climate of Kashgar was either humid/wet or dry. Seong showed the mechanisms by which this climate cycle worked: when the westerlies moved southwards and increased in strength the amount of snow (precipitation) falling on the three mountain chains surrounding the Tarim Basin was likely to increase. Increased snow falls would have led to an increase in the volume of meltwater flowing via the multiple rivers into the Kashgar oasis. The increased meltwater, in turn, ‘may have caused’ the increases in humidity measured by Zhao et al. during the four humid periods that they identified from their pollen analysis. With increased humidity, the vegetation of the Kashgar oasis also increased. Dry climates were, by contrast, generated by the northerly movement of the westerlies which led to less precipitation or snow falling, reduced meltwater, and increased aridity that then caused the contraction of the vegetation in the Kashgar oasis. The unique locational characteristics of the Kashgar oasis, namely its proximity to the westerlies and the surrounding mountain ranges, suggests that the Kashgar oasis has probably had a climate pattern very different from the other parts of the Tarim Basin.

Zhao and his colleagues (2012) have produced a climate model that is more relevant because it focuses on the micro-climate of a particular locality rather than the whole of China. Moreover, the micro-climate that they describe relates specifically to the Kashgar oasis and it takes into account a much wider range of factors than Hsu and Zhu considered. Most important of all, they have an explanatory hypothesis about the mechanisms that probably generated the climatic oscillations for the last 4000 years, the focus of the work of the authors cited above. Zhao et al. based their analysis on high resolution pollen data sampled at Upal on the outskirts of Kashgar. They showed that there have been four humid/wet periods and three dry periods over the last 4000 yr BP in the Kashgar area as follows: 1. 2. 3. 4. 5. 6. 7.

The 2050-670 BC humid/wet period. The 670 BC-200 AD dry period. The 200-690 AD humid/wet period. The 690-1110 AD dry period. The 1110-1270 AD humid/wet period. The 1270-1400 AD dry period. The 1400-1560 AD humid/wet period. This period corresponded to the Little Ice Age (ca. 1500-1800 AD).

However, it is uncertain whether climate change can be correlated with socio-political structures. Suggested correlations need to avoid making deterministic assumptions. Bearing these constraints in mind, can we surmise, for example, that the humid/ wet period of 200-690 AD coincided with the rise of a flourishing Buddhist culture in Kashgar as we have documented in this chapter and the previous one? Of course, accurate dating of the Buddhist stupas and

What caused these oscillations between humid/wet and dry periods in the history of Kashgar’s climate patterns during the last 4000 years? The following factors were considered: • The Kashgar oasis is surrounded by three groups of mountains- the Tianshan, Pamirs and Kunlun to the north, west and south respectively; 164

A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times monasteries is needed as a first step in speculating about this connection. What if Buddhist communities arose in the Kashgar oasis earlier? Would that not then compel us to argue that human ingenuity in developing irrigation systems to bring water to where it was needed was a more important factor in influencing agriculture than the long dry climates that appear to have typified the period between 670 BC and 200 AD? Exactly how technology might have been used to respond to long-term swings in climate changes over a period of 4000 years is something that could be measured through a careful analysis of the soils of Kashgar. Understanding the nature of this technology and the adaptation of local people to climate change is more important, we submit, that a narrow focus on the rise and fall of empires.

published outstanding research as a result of systematic and multidisciplinary research projects that take into account many overlapping variables. The five sections of this chapter on the archaeological remains of Kashgar, comparisons with research on South and Southeasts Asia, the socio-economic base of Buddhist institutions, the early human settlement patterns and the possible impact of climate change on human endeavour, all call for similar, interrelated methodologies to be applied to reconstruct the pre-history and early history of the Kashgar oasis. Without this, our knowledge of the importance of Xinjiang as a transition zone between the ‘West’ and the ‘East’ remains incomplete. In particular, if Bactria, Ferghana, Sogdiana, Afghanistan and Pakistan played critical roles as centres for the dissemination of peoples, technologies and ideas into western China, then we have a very strong justification for spending time on uncovering how these processes played themselves out in the Kashgar region through which they moved. It is virtually inconceivable that when this process of dissemination spread into the southern and eastern Tarim Basin it did not pass through and touch Kashgar itself. We have a sense of urgency about the task of fully understanding the prehistory and early history of Kashgar because much of the archaeological heritage appears to be vanishing rapidly. That loss is accelerating as Kashgar is now defined as a cross border commercial trading hub, and is coming under intense development pressures with China’s Belt Road Initiative (see Vicziany and Di Castro in this volume; Vicziany and Di Castro 2016). Kasghar’s heritage, moreover, is of global importance, given its central location in Eurasia.

Given the proximity of the Kashgar oasis to the mountain ranges surrounding the Tarim Basin and the meltwater coming from them, variations in the extent of cultivable land might reveal something of the interaction between man and nature. The incremental spread of the Taklamakan desert might, for example, explain why the local keepers of cultural relics today believe that there are other cities buried in the sands adjacent to Khan-oi as mentioned in the previous chapter. But again, without careful archaeological surveys of this area, we cannot separate local legends from the historical facts. How this interaction between human endeavour and climate worked would give us a much better understanding of the natural constraints that shaped Kashgar’s history. Despite the encroachments of the desert, we are unlikely to find the dramatic shifts in river systems or total desertification of the kind that happened on the eastern extremes of the Tarim Basin when the settlements of Xiaohe, Niya and other sites were abandoned. Those dramatic events were possible partly because of the great distance of the rivers of the eastern Taklamakan desert from their sources in the mountains surrounding the Tarim basin. The Kashgar oasis, by contrast, is fortunately located much closer to the water supply than these localities.

Scientific methods for studying in the field, remote sensing and laboratory analysis can yield information that historical documents cannot. Written records that might illuminate the kind of questions that we have discussed in this chapter and the previous one are unlikely to be found in the near future while the existing records, with few exceptions, tend to provide us with an elite view of historical events. Even when we access the great archives of Central Asia, the Middle East and Russia we are likely to find that the written records were typically created by either central or local authorities preoccupied by questions about their own history and power rather than the lives of ordinary people. It is the latter, we have suggested in this and the previous chapter, that need to be understood in order to move beyond elite perceptions of historical reality.

Conclusion In mapping the known archaeological and cultural history of ancient Kashgar, we have identified a range of research questions that are worth addressing if we wish to preserve the special history of the human and natural environment of this oasis town. We note that relatively simple, multidisciplinary methodologies could be used to create a more accurate and holistic picture of the human settlement patterns in this part of western China. Fortunately, there is much to learn from comparative work in Central, South and Southeast Asia where scholars such as Kuz’mina, Shaw, Coningham, Gunawardhana and Fletcher have

A systematic inventory of the scholars and other professionals who have written about the Kashgar region is still lacking with the result that access to field notes, excavation reports and the theoretical frameworks that have been used remain unintentionally 165

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Conyers, L.B. 2012. Interpreting Ground-penetrating Radar for Archaeology. Walnut Creek CA: Left Coast Press Inc. Cribb, R. 1991. Nomads in Archaeology. Cambridge: Cambridge University Press. Dallapiccola, A.L. and Z-A.S. Lallemant (eds) 1980. The Stūpa. Its Religious, Historical and Architectural Significance. Wiesbaden: Franz Steiner Verlag. Davis-Kimball, J. 1998. Tribal Interaction between the Early Iron Age Nomads of the Southern Ural Steppes, Semirechiye, Xinjiang, in V.H. Mair (ed.) The Bronze Age and Early Iron Age Peoples of Eastern Central Asia 1: 238-263. Journal of the Indo-European Studies Monograph. 26. Washington/Philadelphia: Institute for the Study of Man, University of Pennsylvania Museum Publications. Debaine-Francfort, C. 1988. Archéologie du Xinjiang des origins aux Han, Part 1. Paléorient 14: 5-29. Debaine-Francfort, C. 1989. Archéologie du Xinjiang des origins aux Han, Part 2. Paléorient 15/1: 183-213. Dehejia, V. 1972. Early Buddhist Rock Temples: A Chronological Study. London: Thames and Hudson. De Filippi, F. 1932. The Italian Expedition to the Himalaya, Karakoruam and Eastern Turkestan (1913-1914). London: Edward Arnold & Co. Deter-Wolf, A. et al. 2016. The World’s Oldest Tattoos. Journal of Archaeological Science: Reports, h t t p : / / w w w. S c i e n c e d i r e c t . c o m / S c i e n c e / article/pii/S2352409X15301772, doi:10.1016/j. jasrep.2015.11.007. Di Castro, A.A. 2008. The Mori Tim Complex in the Kashgar Oasis. East and West 58: 257-283. Di Castro, A.A. 2013. Stupas and Wine; the Artistic Traditions of the Kashgar Oasis. TAASA Review, The Journal of the Asian Arts Society of Australia 22: 7-10. Dodson, J.R. et al. 2013. Origin and spread of wheat in China. Quaternary Science Reviews 72:108-111. Fletcher, R. 2009. Low-Density, Agrarian-Based Urbanism: A Comparative View. Insights 2: 2-19. Fletcher, R. 2012. Low density, agrarian based urbanism: scale, power and ecology, in M.E. Smith (ed.) The Comparative Archaeology of Complex Societies: 285-320. Cambridge: Cambridge University Press. Fletcher, R. and D. Evans. 2012. The Dynamics of Angkor and its Landscape, in A. Haendel (ed.) Old Myths and New Approaches: Interpreting Ancient Religious Sites in Southeast Asia: 42-62. Melbourne: Monash University Publishing, Monash Asia Series. Fletcher, R. et al. 2015. Angkor War: an introduction. Antiquity 89: 1388-1401. Franz, H.G. 1980. Stupa and Stupa-Temple in the Gandharan Regions and in Central Asia, in A.L. Dallapiccola and Z-A.S. Lallemant (eds) The Stūpa. Its Religious, Historical and Architectural Significance: 3958. Wiesbaden: Franz Steiner Verlag. Gernet, J. 1995. Buddhism in Chinese Society. An Economic History from the Fifth to the Tenth centuries translated

restricted in the case of Kashgar. Our hope is that this chapter and the previous one will refocus attention on this significant oasis and end decades of scholarly neglect and isolation. Acknowledgements We are grateful to Professor Marylin Martin Rhie (Smith College, Mass. USA), Professor Emeritus Colin Mackerras (Griffith University) and Dr Ayxem Eli (University of New South Wales) for their comments on an earlier draft of this chapter. References AFP 2011. Quake hits China’s ancient Kashgar. 12 August, viewed on 19 Sept. 2012, . al-Kashgari, Mahmud 1032. [Edition of 1982], Compendium of the Turkic Dialects, in Robert Dankoff and James Kelly (eds) Sources of Oriental Languages and Literatures, Turkic Sources: 7, Parts 1-3. Cambridge, Mass.: Harvard University Printing Office. Anderson, E.N. 2010. Ancient and Modern Foods from the Tarim Basin. Expedition Magazine: Silk Road Issue 52: 5-6. Bailey, G. and I. Mabbett. 1993. The Sociology of Early Buddhism. Cambridge: Cambridge University Press. Bailiff, I.K. et al. 2013. Luminescence and dating of brick stupas: an application to the hinterland of Anuradhpura, Sri Lanka. Antiquity 87: 189-201. Beal, S. 1983. Si-Yu-Ki: Buddhist Records of the Western World. Translated from the Chinese of Hiuen Tsiang A.D. 629. New Delhi: Munshiram Manoharlal [Reprint of 1st edition of 1884]. Bertrand, A. 2012. Water Management in Jingjue Kingdom: The Transfer of a Water Tank System from Gandhara to Southern Xinjiang in the Third and Fourth Centuries CE. Sino-Platonic Papers April: No. 223, 1-81. Chakrabarti, D.K. 1995. Buddhist Sites across South Asia as Influenced by Political and Economic Forces. World Archaeology 27: 185-202. Chavannes, É. 1903. Documents sur les Tou-Kiue (Turcs) Occidentaux. Paris: Librairie d’Amérique et d’Orient Adrien Maisonneuve. Chen, K. 1973. The Chinese Transformation of Buddhism. Princeton: Princeton University Press. Chen, G. 1981. Ancient Graves of the Pamir Plateau. Acta Archaeologica Sinica (Kaogu xuebao) 2: 199-215 [陈戈 1981. 帕米尔高原古墓. 考古学报 2 期 199215 页]. Coningham, R. and P. Gunawardhana. 2013. Anuradhapura: Volume III: The Hinterland. Oxford: BAR International Series 2568, Archaeopress. 166

A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times by Franciscus Verellen. New York: Columbia University Press [English translation of the 1st French edition of 1956]. Gilliland, K. et al. 2013. The dry tank: development and disuse of water management infrastructure in the Anuradhapura hinterland, Sri Lanka. Journal of Archaeological Science 40: 1012-1028. Hambis, L. (ed.) 1961. Mission Paul Pelliot. I. Toumchouq. Planches. Paris: Documents archéologiques publiés sous les auspices de l’Académie des Inscriptions et Belles-Lettres. Hambis, L. 1964. Sites et monuments de la région de Kachgar et de Toumchouq, in L. Hambis et al. (eds) Mission Paul Pelliot. II. Toumchouq: 1-37. Paris: Documents archéologiques publiés sous les auspices de l’Académie des Inscriptions et BellesLettres. Han, K. 1994. The Study of Ancient Human Skeletons from Xinjiang, China. Sino-Platonic Papers No. 51:113. Han, K. 1998. The Physical Anthropology of the Ancient Populations of the Tarim Basin and Surrounding areas, in V.H. Mair (ed.) The Bronze Age and Early Iron Age Peoples of Eastern Central Asia: 2: 558-570. Philadelphia: Journal of Indo-European Studies Monograph Number Twenty-Six, Institute for the Study of Man, University of Pennsylvania Museum Publications. Han, K. 1999. Physical Anthropological Studies on the Racial Affinities of the Inhabitants of Ancient Xinjiang, in B. Wang (ed.) The Ancient Corpses of Xinjiang: The Peoples of Ancient Xinjiang and Their Culture: 224-241. Urumqi: Xinjiang Renmin Chubanshe (CIP) [韩康信 1999. 新疆古代居民种族 人类学研究, 载于王炳华编 新疆古尸−古代新疆 居民及其文化: 224-241页. 乌鲁木齐: 新疆人民出 版社]. Harmatta, J. 1994. Languages and Literature in the Kushan Empire, in J. Harmatta et al. (eds) History of Civilizations of Central Asia, II: 417-440. The Development of Sedentary and Nomadic Civilizations: 700 B.C. to A.D. 250 Paris: UNESCO. Hedges, R.E.M. and L.M. Reynard. 2007. Nitrogen isotopes and the trophic level of humans in archaeology. Journal of Archaeological Science 34: 1240-1251. Heitzman, J. 1984. Early Buddhism, Trade and Empire, in K.A.R. Kennedy and G. Possehl (eds) Studies in the Archaeology and Palaeoanthropology of South Asia: 121-137. New Delhi: Oxford University Press, IBH/ American Institute of Indian Studies. Hitch, D. 2009. The Special Status of Turfan. Sino-Platonic Papers No. 186: 1-61. Hong, Z. et al. 2004. A preliminary study of oasis evolution in the Tarim Basin, Xinjiang, China. Journal of Arid Environments 55: 545-553. Hsu, K. 1998. Did the Xinjiang Indo-Europeans Leave Their Home Because of Global Cooling?, in V.H.

Mair (ed.) The Bronze Age and Early Iron Age Peoples of Eastern Central Asia, 2 Vols. Journal of Indo-European Studies: 26, 683-696. Philadelphia: Institute for the Study of Man/The University of Pennsylvania Museum Publications. Huang, Y. et al. 2005. Hydrological Process and Characteristics in the Arid Plain Oases, L. Bengtsson et al. (eds) Proceedings of an International Symposium on Sustainable Water Resources Management and Oasis-Hydrosphere-Desert Interaction in Arid Regions. 206-216. Beijing: Chinese Academy of Sciences, [干旱内陆河区水资 源可持续利用和植被-水文-沙漠的相互作用国际 学术研讨会论文集: 27-29, October 2005, ...], viewed on 13 July 2019, . Interview with Dr Veronica Silva, Department of Anthropology, Chilean Natural History Museum, San Diego, Chile, 17 October 2017. Kashi Prefecture Statistics Bureau 2011. Statistics online. [喀什地区统计局 2011, 26 Dec. 2018, . Kosambi, D.D. 2009. Dhenukākata, in B. Chattopadhyaya (ed.) The Oxford India Kosambi. Combined Methods in Indology and Other Writings: 450-475. New Delhi: Oxford University Press (First published in 1957 in the Journal of the Bombay Branch of the Royal Asiatic Society 1955: Vol. 30, 50-71). Kulikova, G. 2016. Source parameters of the major historical earthquakes in the Tien-Shan region from the late 19th to the early 20th century. Potsdam: University of Potsdam, Unpublished PhD dissertation, viewed on 8 July 2019, . Kuz’mina, E.E. 2008. The Pre-history of the Silk Road. Philadelhia: University of Pennsylvania Press. Lattimore, O. 1940. Inner Asian Frontiers of China. New York: American Geographical Society. Li, C. et al. 2010. Evidence that a West-East admixed population lived in the Tarim Basin as early as the early Bronze Age, BMC Biology. 8:15, viewed on 21 May 2019, http://www.biomedcentral.com/17417007/8/15>. Li, J.F. 1991. The Climate of Xinjiang. Beijing: Meteorology Press. Li, Y.1962. A Review of Archaeological Works of the Xinjiang Uygur Autonomous Region. Relics: 7-8, 1115, 80. [李遇春 1962. 新疆维吾尔自治区考古工作 概况. 文物, 7-8期, 11-15页]. Liu, X. 1988. Ancient India and Ancient China: Trade and Religious Exchanges AD 1-600. Delhi: Oxford University Press. Lou, L. 2017. Uncovering the ancient canal-based tuntian agricultural landscape at China’s northwestern frontiers. Journal of Cultural Heritage 23:79-88, http:// dx.doi.org/10.1016/j.culher.2016.04.013. 167

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Ma, Y. and S. Yutang. 1994. The Western Regions under the Hsiung-nu and the Han, in J. Harmatta et al. (eds) History of Civilizations of Central Asia, Vol. II. The Development of Sedentary and Nomadic Civilisations: 700 B.C. to A.D. 250: 227-246. Paris: UNESCO. Ma, Y. and B. Wang 1994. The Culture of the Xinjiang Region, in J. Harmatta et al. (eds) History of Civilizations of Central Asia. Vol. II. The Development of Sedentary and Nomadic Civilisations: 700 B.C. to A.D. 250: 209-225. Paris: UNESCO. Mair, V.H. (ed.) 1998. The Bronze Age and Early Iron Age Peoples of Eastern Central Asia, 2 Vols, Journal of IndoEuropean Studies. Philadelphia: Institute for the Study of Man/University of Pennsylvania Museum Publications. Mallory J.P. 2010. Bronze Age Languages of the Tarim Basin. Expedition Magazine: Silk Road Issue 52: No.3, 44-53. Mallory, J.P. and V.H. Mair 2000. The Tarim Mummies. London: Thames and Hudson. Michell, G. et al. with photographs by J. Gollings 2008. Kashgar, Oasis City on China’s Old Silk Road. London: Frances Lincoln Publishers. Morrison, K.D. 1995. Trade, Urbanism, and Agricultural Expansion: Buddhist Monastic Institutions and the State in the Early Historic Western Deccan. World Archaeology 27: 203-21. Mukhamedjanov, A.R. 1994, Economy and social system in Central Asia in the Kushan Age, in J. Harmatta and B.N. Puri (eds) History of Civilisations of Central Asia 2: 257-282. Paris: UNESCO. Nagaraju, S. 1981. Buddhist Architecture of Western India (c. 250 B.C – c. A.D. 300). Delhi: Agam Kala Prakashan. Narain, A.K. 1990. Indo-Europeans in Inner Asia, in D. Sinor (ed.) The Cambridge History of Early Inner Asia: 151-176. Cambridge: Cambridge University Press. Neelis, J.E. 2011. Early Buddhist Transmission and Trade Networks Mobility and Exchange Within and Beyond the Northwestern Borderlands of South Asia (Dynamics in the History of Religions 2). Leiden: Brill. Oldenburg, S. (1983) [1892]. Kashgar Manuscripts of N.F. Petrovsky (1892), in Buddhist Texts from Kashgar and Nepal, Sata Pitaka Series, Indo-Asian Literatures, Vol. 322: 154-252. New Delhi: International Academy of Indian Culture. [Reprint of first edition of 1892]. Pelliot, P. 1906. Notes sur l’Asie Centrale. Bulletin de l’Ecole française d’Extrême-Orient 6: 255-269. Pelliot, P. 1959. Notes on Marco Polo. I. Ouvrage Posthume publié sous les auspices de l’Académie des Inscriptions et Belles-Lettres et avec le concours du centre national de la recherche scientifique. Paris: Imprimerie Nationale, Librairie A. Maisonneuve [in English]. Personal communications from Professor Yidilisi Abuduresule, Research Professor/Adjunct, Xinjiang Normal University, Urumqi, October 2013. Personal communications from Abduyraehim Qadir, retired Director, Kashgar Museum October 2013.

Personal communications from Dr Keliang Zhao, Institute of Vertebrate Paleontology/ Paleoanthropology and the Key Laboratory of Vertebrate Evolution and Human Origins, Chinese Academy of Sciences, Beijing, December 2013. Personal communications from Professor Alison Betts, Professor of Silk Road Studies, University of Sydney, Sydney, Australia, January 2014. Personal communication from Professor Francesco Sferra, Institute of Oriental Studies, University of Naples, 5 October 2015. Personal communication from Professor Marylin Martin Rhie, Jessie Wells Post Professor Emerita of Art and Professor Emerita of East Asian Studies, Smith College, Northampton, Massachusetts, USA, 10 December 2018. Rapp G.R. Jr, and C.L. Hill 1998. Geoarchaeology. New Haven: Yale University Press. Ray, H.P. 1986. Monastery and Guilds. Commerce under the Sātavāhanas. Delhi: Oxford University Press. Ray, H.P. 1994. The Winds of Change. Buddhism and the Maritime Links of Early South Asia. Delhi: Oxford University Press. Qadir, A. et al. 2002 (Relics Team of Cultural Relics Office in Kashi Prefecture). A Brief Report on the Excavation of Yawaluk, Kashgar in 2000. Cultural Relics of Xinjiang. 67-68: nos. 3-4: 54-57. [喀什地区文 物管理所文物队 2002. 喀什市亚吾鲁克遗址2000 年清理简报. 新疆文物总67-68，2002年3-4期，5457 页]. Rhie, M.M. 1999. Early Buddhist Art of China and Central Asia, Vol. 1. Later Han, Three Kingdoms and Western Chin in China and Bactria to Shan-shan in Central Asia. Leiden: Brill. Sala, R. and J-M. Deom 2008. The 261 Karez of the Sauan Region (Middle Syrdarya). Transoxiana: Journal Libre de Estudios Orientales 13: August, viewed on 2 April 2019, . Sen, T. 2012. The Spread of Buddhism to China: A Re-examination of the Buddhist Interactions between Ancient India and China, China Report 49 (102): 11-17. Seong, Y.B. et al. 2009. Quaternary glaciation of Muztag Ata and Kongur Shan: evidence for glacier response to rapid climate changes throughout the Late Glacial and Holocene in westernmost Tibet. Geological Society of America Bulletin 121: 348365. Shahrani, M.N. 2002. The Kirghiz and Wakhi of Afghanistan: Adaptation to Closed Frontiers and War. Seattle/ London: University of Washington Press. Shaw, J. 2000. Sanchi and its archaeological landscape: Buddhist monasteries, settlements & irrigation works in Central India. Antiquity 74 (286): 775-776. Shaw, J. and J. Sutcliffe 2001. Ancient irrigation works in the Sanchi area: an archaeological and hydrological investigation. South Asian Studies 17: 55-75. 168

A.A. Di Castro, M. Vicziany and X. Zhu: The Kashgar Oasis in Buddhist and Pre Buddhist Times Shaw, J. et al. 2007. Ancient irrigation and Buddhist History in Central India: Optically Stimulated Lumines and Dates and Pollen Sequences from the Sanchi Dams. Asian Perspectives 46 (1): 166-201. Shaw, J. 2009. Stūpas, Monasteries, and Relics in the Landscape: Typological, Spatial, and Temporal Patterns in the Sanchi Area in A. Shimada and J. Hawkes (eds) Buddhist Stūpas in South Asia: Recent Archaeological, Art-Historical, and Historical Perspectives: 114-145. New Delhi: Oxford University Press. Shaw, J. 2016. Buddhist Landscapes in Central India. Sanchi Hill and Archaeologies of Religious and Social Change, c. Third Century BC to Fifth Century AD. London: The British Association for South Asian Studies/The British Academy 2007 Edition republished by Routledge: ProQuest Ebook Central, viewed on 19 May 2019, . Snodgrass, A. 1985. The Symbolism of the Stupa. New York: Cornell University, Ithaca. Stein, A.M. 1907. Ancient Khotan: Detailed Report of Archaeological Explorations in Chinese Turkestan, 2 vols. Oxford: Clarendon Press. Stein, M.A. 1980. Serindia: Detailed Report of Explorations in Central Asia and Westernmost China. Delhi: Motilal Banarsidass. 4 vols. [reprint of First Edition Oxford 1921]. Tamm, E.E. (2012) [2008], ‘The Horse that Leaps Through Clouds’, viewed on 2 April 2019, . Vicziany, M. 2012. Indianisation and the Religious Sites of Southeast Asia, in Civilisational Dialogue between India and ASEAN: 209-218. Delhi: ICCR Patna Conference Proceedings. Vicziany, M. and Di Castro, A.A. 2016. The Hidden Treasures of Buddhist Kashgar. TAASA Review The Journal of the Asian Arts Society of Australia 24: 14-17. Vicziany, M. et al. 2017. Ekveera Devi and the Son Kolis of Mumbai: Have the Kolis Appropriate the Karle Buddhist Chaitya, in J.B. Bapat and I. Mabbett (eds) Conceiving the Goddess: Transformation and Appropriation in Indic Religions: 225-257. Melbourne: Monash Asia Series, Monash University Publishing. Wang, B. (ed.) 1999. The Ancient Corpses of Xinjiang. The Peoples of Ancient Xinjiang and Their Culture. Urumqi: Xinjiang Renmin Chubanshe (CIP) [王炳华编辑 1999. 新疆古尸−古代新疆居民及其文化. 乌鲁木 齐: 新疆人民出版社]. Watt, J.C.Y. et al. (eds) 2004. China. Dawn of a Golden Age 200-750 AD. New York, New Haven and London: The Metropolitan Museum of Art and Yale University Press. Walker, R. 1992. Phosphate Survey: Method and Meaning, in P. Sperry (ed.) Geoprospection in the Archaeological Landscape: 61-73. Oxford: Oxbow Books.

Wintle, A.G. 1996. Archaeologically-relevant Dating Techniques for the Next Century. Journal of the Archaeological Science 23: 123-138. Wu, Y. 2005. New results of the archaeological excavations at Xiabandi Graveyard in Kashgar, Xinjiang. Xiyu Yanjiu 1: 109-113 [吴勇 2005. 新疆 喀什下坂地墓地考古发掘新收获. 西域研究, 1 期, 109-113 页]. Xinhua 2012. Chinese scientists identify largest Neolithic city. China.org: 30 October, viewed on 2 April 2019, . Xinjiang Institute of Archaeology, Social Science Academy 1983. Thirty Years of Archaeology in Xinjiang. Urumqi: Xinjiang Renmin Chubanshe [新疆社会科 学院考古研究所 1983. 新疆考古三十年. 乌鲁木 齐: 新疆人民出版社]. Xinjiang Institute of Archaeology 1995a. Report on the Relics of Artux, Kashgar, Hotan areas during the Qarakhanid dynasty in B. Wang (ed.) (1995) New Achievements in Archaeological Research in Xinjiang during the Span (1979-1989)]: 582: 588. Urumqi: Xinjiang People Press. [新疆考古研究所 1995a. 阿 图什, 喀什, 和田地区喀喇汗朝遗迹调查, in王炳华 (ed.) 新疆文物考古新收获- (1979-1989) 新疆田野 考古报告集: 582-588 页。乌鲁木齐: 新疆人民出 版社]. Xinjiang Institute of Archaeology 1995b. Report on the Mesolithic Relics in New Achievements in Archaeological Research in Xinjiang during the Span (1979-1989)]: 27-40 Urumqi: Xinjiang Renmin Chubanshe. [新疆考古研 究所 1995b. 乌帕尔细石器遗址调查报告, 载王炳 华编辑 新疆文物考古新收获 -(1979-1989) 新疆田 野考古报告集: 27-40页。乌鲁木齐: 新疆人民出版 社]. Xinjiang Cultural Relics Bureau et al. (eds) 1999. A Grand View of Xinjiang’s Cultural Relics and Historic Sites. Urumqi: Xinjiang Art and Photographic Press [新疆 维吾尔自治区文物事业管理局等主编 1999. 新疆 文物古迹大观. 乌鲁木齐: 新疆美术摄影出版社]. Xinjiang Annals Editorial Board (eds) 2007. Xinjiang General Annals Urumqi: Xinjiang People’s Press. [新 疆维吾尔自治区地方志编纂委员会编 2007. 新疆 通志第八十一卷. 乌鲁木齐: 新疆人民出版社]. Xinjiang Institute of Cultural Relics and Archaeology. 2012. Xiabandi Cemetery in Xinjiang. Urumqi: Cultural Relics Press [新疆文物考古研究所 2012. 新疆下坂 地墓地. 乌鲁木齐: 文物出版社]. Zhao, K. et al. 2012. Climatic variations over the last 4000 cal. yr BP in the western margin of the Tarim Basin, Xinjiang, reconstructed from pollen data. Palaeogeography, Palaeoclimatology, Palaeoecology 321322: 16-23, DOI: 10.1016/j.palaeo.2012.01.012. Zhu, K.Z. 1972. A preliminary study on climatic changes in China over the last 5000 years. Acta Archaeologica Sinica 1: 168-189 [竺可桢 1972. 中国近五千年来气 候变迁的初步研究. 考古学报, 1 期: 168-189 页]. 169

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Zhu, Z.D. et al. 1980. Deserts in China. Beijing: Science Press. [朱震达, 吴正, 刘恕, 邸醒民 1980. 中国沙漠 概论. 北京:科学出版社]. Zürcher, E. 1968. The Yüeh-chih and Kaniṣka in Chinese Sources, in A.L. Basham (ed.) Papers on the Date of Kaniṣka: 346-390. Leiden: Brill. Zürcher, E. 2012. Buddhism across Boundaries: The Foreign Input, in J.R. McRae and J. Nattier (eds)

Buddhism Across Boundaries: The Interplay of Indian, Chinese, and Central Asian Source Materials, SinoPlatonic Papers, Philadelphia: University of Pennsylvania.

170

Chapter 10

Unravelling Farming and Metallurgy in Ancient China with Nuclear Science John Dodson, Pia Atahan and Xiaoqiang Li C. Radiocarbon quickly oxidizes to CO2 which then permeates the biosphere. All living matter with a link to the atmosphere develops a 14C content that is in equilibrium with the atmosphere. Once the organism dies and becomes isolated from the atmosphere, its 14C decays away to 14N – completing the cycle. The halflife of 14C is about 5730 years. This forms the basis of radiocarbon dating (Figure 2).

Xinjiang and Gansu are key areas for evidence in building our understanding on the links between eastern and western Asia. Exchanges of agricultural systems, metallurgy and ideas existed thousands of years before the flourishing of the Silk Road in the Han period. Radiocarbon dating is of course a key technique in placing an age on the evidence of these developments, while stable isotope ratios on carbon and nitrogen in animal and human bones shed light on diets, health and even the pace of animal domestication. Trace elements such as lead and strontium in bronze slag, ancient tools and ore can be used to fingerprint the sources of ore and reveal potential exchange routes. In this chapter we showcase examples of nuclear techniques and what they reveal about life in the Bronze Age of northwest China.

14

The Earth, and humans, are constantly bombarded with cosmogenic radiation. In addition to this, the Earth itself contains certain amounts of radioactivity that originates from the planet’s formation, and the decay of which keeps the planet’s core hot. This natural radioactivity is ticking away at half-lives from billions of years to fractions of a second, and these add to the radiation environment including our radiation dose. Accurate measurement of isotopic mass is achieved using accelerators and mass spectrometers. Some decay processes are still captured by measuring the alpha, beta and gamma radiation through a variety of detectors.

About nuclear techniques The precise measurement of the properties of isotopes has a broad application in archaeology and environmental science and we demonstrate the application of these through three studies in northwest China.

In chemistry, isotopes of different mass react at slightly different rates, with heavier isotopes (such as 15 N or 13C) reacting slower than lighter ones (such as 14 N or 12C for these elements). This process is termed isotopic fractionation and results in inhomogeneous distributions of isotopes through physical and biological systems (Figure 3). Through food chains the discrimination against heavier isotopes is compounded with each trophic level, resulting in large differences in

Carbon has 3 major isotopes, each with 6 protons and either 6, 7 or 8 neutrons (Figure 1). The first two of these isotopes are stable (12C and 13C), but the third (14C or radiocarbon) is radioactive. Radiocarbon is formed in the upper atmosphere as cosmic rays (or stripped atomic nuclei arriving mainly from the Cosmos) bombard 14N atoms and turn these to

Figure 1. Diagrammatic representation of the three main isotopes of Carbon. All have 6 protons, making them Carbon, but they have differing numbers of neutrons in the nucleus. Carbon-14 is unstable and decays radioactively (Dodson 2014).

171

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Case Study 1: Bronze in Gansu Early metallurgy was based on metal recovery by heating ores, and the first of these were soft metals. While useful, these required constant upkeep and honing to prolong usage. The advance needed to overcome this was to forge harder and more resilient metals. The first advance came with the production of bronze, a harder alloy made by adding small amounts of arsenic, and later tin and lead, to copper. Modern metallurgy is still focused on the same problem of producing metals with improved characteristics and today, for example, there is a variety of steels based on iron which are designed for specific purposes. Bronze was a significant technological advance over soft copper, and because of the specialised nature of making bronze, an often asked question is where did it originate? The earliest known sites where copper alloys were being exploited are in Mesopotamia, and date to the 4th millennium BC (De Ryck et al. 2005). The earliest copper alloys reported for eastern Asia date to the late 4th and early 3rd millennia BC and come from sites in northwest China (Linduff and Mei 2009; Roberts et al. 2009). By 2500-1900 BC Figure 2. General rules of isotopic fractionation: lighter isotopes bronze was more widespread, occurring move/vibrate faster and form weaker chemical bonds; isotope from Gansu eastwards, through the effects are stronger where the relative mass differences are larger central plains to the eastern coast (Linduff (e.g. between 16O, 17O, 18O); and isotope effects reduce at higher et al. 2000). Regional variations in alloy temperatures (Dodson 2014). composition of bronze objects during this period were affected by natural availability of the ores (Linduff and Mei 2009; Mei 2009 and references therein). The potential of bronze tools to improve farming and hence allow for increased productivity and expanded populations must have been great. The ongoing question for Chinese archaeology is how much of bronze technology was imported and how much was developed independently and, following on from this, how important was it as a factor in creating arable land and transforming the landscape. Sites from the Hexi Corridor of western Gansu help to explain the challenges facing the Bronze Age settlements. We selected this area for study because of its strategic location between the deserts of Mongolia and the Qilian Mountains (see Figure 4). While the Hexi region is arid to semi-arid, there are significant river systems which are fed by snow melt from the Qilian ranges and flow into the desert. The resulting rivers and oases opened up a major transport corridor from ancient times onwards.

Figure 3. Representation of the basis of the radiocarbon dating method (Dodson 2014).

the isotopic make-up of organisms. This is very useful in archaeological studies. 172

J. Dodson, P. Atahan and X. Li: Unravelling Farming and Metallurgy in Ancient China with Nuclear Science Ore and slag samples from Gangangwa and Huoshiliang (Figure 4), and ore from the nearby Dingxin copper mine at Baishantang, were analysed for lead and strontium isotope ratios. While the sample size was small, the isotope ratios in all but one of the samples from the archaeological sites were essentially identical to the ore sample from the Dingxin mine (see Dodson et al. 2009). This is reasonable evidence to suggest the modern mine site was a source of copper for the archaeological sites. The one outlying sample had a very different strontium and lead isotope ratios, suggesting it derives either from another source of ore, or from exchange of artefacts brought into the area.

The Hexi region contains several archaeological sites with complex scatters of pottery, bone, metal slag, ore and other artefacts, but these sites have barely been described. Our radiocarbon ages on seed and charcoal collected from two sites approximately 11 km apart, Gangangwa and Huoshiliang (Figures 4, 5), suggest that human occupation begins around 2135 BC. Occupation of the sites was relatively brief; Huoshiliang was apparently abandoned around 1860 BC as wood supplies were exhausted and the unstable soils led to desert dunes taking over large areas (Dodson et al. 2009; Li et al. 2011). At Gangangwa mounds of sand with charcoal and metal slag are evident and these were apparently ore smelting and bronze manufacturing sites. Our study focused on trace elements, including strontium and lead isotopes, in bronze, slag and ore samples collected at various locations. We also investigated the abundance of selected elements and botanical remains in sediment profiles at the Huoshiliang site and at nearby Tiao Hu.

In summary, bronze manufacture may be more than 4000 years old at Huoshiliang and Gangangwa sites in Gansu – not as old as bronze manufacture in Mesopotamia. At least two sources of ore were used in bronze manufacture at Gangangwa. Case Study 2: Early wheat in China

Bronze artefacts at Huoshiliang are composed predominantly of copper (up to 86.3 per cent) with significant amounts of arsenic (up to 2.4 per cent) and in one case, tin (up to 1.2 per cent) (Dodson et al. 2009). Trace element analysis through the sediment sequence at Huoshiliang site, which accumulated between about 2100 BC to 1700 BC, shows a coincident increase in copper and arsenic at about 1900 BC (see Figure 6). The presence of both copper and arsenic suggest that they were being processed together at the site. The trends in other major metal elements such as zinc, lead, nickel and iron are generally unrelated to the amount of copper or arsenic. A similar analysis was conducted on sediments from a nearby ancient lake system, Tiao Hu. The lake system was present at the time when the region was first occupied and up to about 1700 BC when it dried out and was overwhelmed by a dune field. A contemporaneous peak in concentrations of copper, arsenic, lead, zinc and nickel occurred between c. 2250 and 1750 BC, providing further evidence for a new source of metal elements in the region (Li et al. 2011).

The main cereal crops in China today are rice, millets, maize and wheat. Millet and rice were first domesticated in China, in the Yellow and Yangtze River valleys respectively, during the early Holocene (Bellwood 2005; Crawford et al. 2009; Yasuda 2002). Maize was first domesticated in the Americas at about 6000 cal. yr BP (Piperno and Flannery 2001) and became part of the east Asian cereal suite within the last few hundred years. Wheat was first domesticated in the Fertile Crescent of the Middle East around the beginning of the Holocene (Bellwood 2005) (Figure 4). There are at least two competing theories for the introduction of wheat into eastern Asia. One champions its first appearance in north-eastern China (Crawford et al. 2005), while the other argues that it came into western China through one of the Xinjiang, Gansu or Inner Mongolia regions (e.g. Flad et al. 2010; Dodson et al. 2013). Direct evidence for early wheat in China is based on AMS radiocarbon dating of individual wheat seeds recovered from archaeological remains or sedimentary settings. Many presumed dates which permeate discussion of early wheat in China have been based on associations with pottery, but the links between pottery and wheat remains are not always established. Several radiocarbon analyses on individual wheat grains are now available, and these are probably the best direct evidence about the age of wheat in a location. The oldest of those direct dates are clustered around Gansu and younger ages are evident both west and south of there (Figure 7). The earliest evidence of wheat in China comes from sediments at Xishanping near Dadiwan in Gansu Province. Xishanping includes a well-dated sediment section from a high terrace overlooking the Xi River.

Ores contain many trace elements and several isotopes of some elements can be particularly useful. Lead has many stable isotopes, and 204Pb abundance was set at the formation of the Earth. The other stable isotopes are the end chains of decay series. The 207Pb / 204Pb and 206Pb / 204Pb ratios in an ore are a function of the age of the deposit and the initial 235U and 238U concentrations. Likewise, 87Sr / 86Sr ratios are a function of the origin and weathering history of the materials containing the ore. The stable isotopes of lead and strontium thus form a basis for fingerprinting ores and bronze. 173

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 4. Diagram showing Eurasia and Gansu Province with the sites mentioned in the text.

The section shows millet and rice cultivation occurring over the last 5100 years. Small numbers of wheat and barley seeds are evident in the section and date to as early as 4650 cal. years BP (Li et al. 2007). Although the number of wheat seeds is low in comparison to millet, wheat increases gradually to the top of the cereal profile

while the abundance of millet seeds declines. Rice seeds are relatively constant through the section occurring in small numbers throughout the profile. Radiocarbon dates collected thus far suggest wheat arrived from the west via Mongolia into the Gansu 174

J. Dodson, P. Atahan and X. Li: Unravelling Farming and Metallurgy in Ancient China with Nuclear Science

Figure 5. Photograph showing surface scatter at Huoshiliang site. Inset shows pieces of pottery and occupation deposits visible on the surface (photo by J. Dodson).

Figure 6. Cation profiles in Huoshiliang sediments (Dodson et al. 2009).

region at about 4100 years ago (Dodson et al. 2013). The existing dates suggest that wheat spread from there into Xinjiang and into Shaanxi, Shanxi and Henan over several centuries.

Case Study 3: Food consumption in early China Since ‘You are what you eat (plus a few per mil)’ was coined by Michael DeNiro and Samuel Epstein in 1976, stable isotope analysis has become a widely used tool for reconstructing diets at archaeological sites in China and elsewhere. Stemming from the principle that stable isotope ratios of primary producers are passed up food chains, with a predictable offset or fractionation against the heavier isotope at each trophic level, the technique has proven useful for understanding the characteristics of diet in ancient populations. Bone is particularly useful for palaeodietary investigations, as it is a

Wheat grains of slightly younger age are now known from the Yili Valley of Xinjiang, but clearly more work needs to be undertaken in western Xinjiang to investigate whether the route of wheat to eastern China involved Xinjiang. In any case the data seem to confirm that exchange along one of the ‘Silk Roads’ predated by at least 2,000 years what is commonly recognized as the Silk Road. 175

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 7. Diagram showing the distribution of early wheat sites with robust AMS radiocarbon dating. Sites with wheat discussed in this chapter are represented by the round dots (1-14): 1 Wupaer (Upal), 2 Xiakalangguer, 3 Xintala, 4 Xicaozi, 5 Sidaogu, 5 Huoshaogu, 7 Gangangwa, 8 Huoshiliang, 9 Shaguoliang, 10 Donghuishan, 11 Huangniangniangtai, 12 Qinweijia, 13 Shangguancun, 14 Nansha. Other sites where wheat has been reported are represented by triangles (15-36), 15 Nam river site, 16 Chawuhugou, 17 Gumugou, 18 Yanghai, 19 Luanzanggangzi, 20 Lanzhouwanzi, 21 Tudun, 22 Yanchigucheng, 23 Dalitaiha, 24 Changguo, 25 Fengtai, 26 Xishanping, 27 Wangjiazui, 28 Zhaojialai, 29 Baligang, 30 Zaojiaoshu, 31 Erlitou, 32 Fengzhai, 33 Wangchenggang, 34 Dugangsi, 35 Yingxu, 36 Liangchengzheng.

such as wheat, rice and barley, and most of the wild plants of north-western China, have bulk δ13C values around -24 to -34‰ (Smith and Epstein 1971). Heavy millet consumers can therefore be distinguished by the elevated δ13C values of their body tissue.

material that can persist in depositional environments for long periods of time and has isotope values that directly reflect food consumed by the individual. Information gained through isotopic measurements can be considered a slightly more direct measure of food consumed compared with other common approaches that draw on evidence such as preserved food waste, residues in vessels and tooth wear (Lee-Thorp 2008). Both inorganic and organic components of bone are suitable for stable isotope investigation. Collagen, the main organic component of preserved bone, has stable isotope ratios that predominantly reflect the isotopic composition of dietary protein (Ambrose and Norr 1993; Jim et al. 2004).

As mentioned previously, fractionation occurs up food chains. There is about a +5‰ offset between carbon isotope values of herbivore bone collagen and their plant-based diet, and an additional +1-2‰ for higher trophic levels (Bocherens and Drucker 2003). The offset for stable nitrogen isotopes is about +35‰ per trophic level, plants have δ15N values that are usually around 1-4‰ above atmospheric nitrogen (Bocherens and Drucker 2003). Landscape variation in soil δ15N occurs in response to factors such as salinity (Britton et al. 2008; Heaton 1987), water logging (Billy et al. 2010) and aridity (Murphy and Bowman 2006; Schwarcz et al. 1999). This landscape variation shows up in plants growing in an area. δ15N values of plants in arid areas can exceed 10‰ (e.g., Schwarcz et al. 1999).

Asian millets (Panicum miliaceum and Setaria italica) have carbon isotope values that are distinctly elevated compared with the other key cereals of prehistoric China. This results from differences in photosynthetic pathways used by plants (C4 for millets and C3 for wheat and rice). Bulk δ13C values of the Asian millets are in the range of about -11 to -13‰, while the C3 plants, 176

J. Dodson, P. Atahan and X. Li: Unravelling Farming and Metallurgy in Ancient China with Nuclear Science As noted above, Huoshiliang and Gangangwa sites in Gansu contain abundant pottery fragments, bone, bronze metal working remains, charcoal, microliths and cereal seeds. Also, two human skeletons were found at Huoshiliang and these were examined by Liang Chen at North West University, in Xi’an (Dodson et al. 2012). Faunal bone (identified by Songmei Hu from the Shaanxi Archaeological Institute) was found to include deer, sheep/goat, cattle, dog, rat and pig. Of the 33 bone samples identified from the two sites, 30 samples yielded collagen that was sufficiently well preserved for stable isotope analysis and radiocarbon dating. Descriptions of the collagen preparation and preservation quality tests can be found in Dodson et al. (2012) and Atahan et al. (2011).

Stable carbon isotope evidence reported here show humans at Huoshiliang having diets dominated by the C4 cereal, millet. Wheat, although present at the site, was not a significant dietary staple. Moreover, the lack of offset in δ15N values in human and domestic omnivore fauna compared with herbivore fauna, indicates animal protein was not a major component of the human diet. This picture of Bronze-age human diets in the Hexi Corridor being composed predominantly of millet cereal and with little or no wheat or animal protein foods is intriguing. Additional work at other contemporary sites is needed to test whether this picture is typical for the Hexi Corridor region.

Radiocarbon ages on seeds, charcoal and bone collagen from the surface scatters fall within the range of 2292 to 1758 BC (see Figure 8), and this age-range corresponds with the local Qijia and Siba cultural periods. The two humans were identified as females, approximately 35-40 and 40-45 years old, and showed signs of severe mouth disease and iron deficiency. The date-ranges for the human skeletons are between ca. 1860 to 2020 BC.

The relationship between humans and the environment is a priority area for study. New research has made it clear that environmental determinism for human well-being is too simple a concept to explain much of what we know from the palaeo-environmental and archaeological records. There almost always seem to be feedback mechanisms between societies and environmental settings, and this has probably always been the case since humans learned to manage fire and resources for exploitation. Stable isotopes in organic and inorganic materials contain an abundance of useful information for archaeological research and palaeoenvironmental reconstruction beyond conventional laboratory chemistry methods, yet are still underutilised.

Conclusion

Figure 9 shows the distribution of δ13C and δ15N values for faunal and human samples from Huoshiliang and Gangangwa. High δ13C values among the human, rat and some of the pig and dog samples indicate their diets were dominated by millet or millet-fed animal protein (i.e. meat, dairy or blood foods). Conversely a group of samples with markedly more negative δ13C values is apparent and this group is composed of wild and domesticated herbivores and the remaining pig and dog samples. For this group, the base of the food chain is composed mostly of C3 plants. There is a distinct similarity in isotope values of wild and domestic herbivores, suggesting they were grazing over similar ranges. Furthermore, these ranges appear to have been mostly beyond the agricultural zone, and beyond the range of millet plants that were presumably being cultivated in the wet oasis areas. The high δ15N values of some of the herbivores may be due to their being taken into arid desert regions to graze, away from the wetter agricultural areas.

Radiocarbon is well known in these fields and is an essential tool for establishing an objective chronology. When carefully applied it is also much more reliable than conclusions based on assumptions of association alone. In understanding the origins, transport and evolution of technology and agriculture there is much to gain from the application of nuclear analytical techniques. In addition to those described here, there are other chronological tools that can be applied, for instance bomb-pulse carbon, tritium analysis, other stable isotope measurements, biomarker analysis and the applications of ion beam analysis, neutron activation analysis and synchrotron radiation to materials with cultural significance. An advantage of nuclear tools is that very small amounts of material can be used for analysis, and some methods are virtually non-destructive where valuable or heritage artefacts are involved. A richer story of cultural history and the relationships between human settlement and the environment will be told where nuclear methods are more widely applied.

The difference in δ13C values amongst the pigs and dogs at Huoshiliang and Gangangwa results from differences in their access to millet as a food source, whereby animals with high values (> -12‰) are likely to have been domestic, consuming millet-rich human waste and scraps, while those with low values (< -12‰) are likely to have been living beyond human settlements, consuming wild and mostly C3 foods. Similar isotopic differences have been used to distinguish between wild and domesticated animals at Dadiwan archaeological site, in the Wei River valley (Barton et al. 2009).

North west China and Central Asia are well suited to the application of nuclear techniques as the arid environment can often preserve organic materials better than humid regions. The exchange of 177

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Lab code

Site

Depth (cm)

Sample type

AMS 14C age (BP)

Calibrated age range (2σ)

OZK596

Huoshiliang

0-5

charcoal

3516 ± 52

2009-1693 BC

OZK597

Huoshiliang

40-45

charcoal

3538 ± 47

2015-1745 BC

Huoshiliang

80-85

charcoal

3566 ± 49

2032-1757 BC

OZK599

Huoshiliang

140-145

charcoal

3509 ± 50

1957-1692 BC

OZK600

Huoshiliang

163

charcoal

3601 ± 57

2135-1775 BC

OZK598

OZK603

Huoshiliang

80-100

seed

3636 ± 44

2135-1895 BC

OZM248

Huoshiliang

surface

charcoal

3555 ± 40

2020-1769 BC

OZM249

Huoshiliang

surface

charcoal

3579 ± 38

2032-1777 BC

OZL292

Huoshiliang

surface

bone collagen

3515 ± 45

1892-1859 BC

OZL293

Huoshiliang

surface

bone collagen

3590 ± 45

2016-1996 BC

OZL294

Huoshiliang

surface

bone collagen

3515 ± 40

1896-1862 BC

OZK657

Gangangwa

surface

charcoal

3737 ± 50

2292-1979 BC

OZK658

Gangangwa

surface

seed

3558 ± 47

2026-1758 BC

OZM245

Gangangwa

surface

charcoal

3652 ± 40

2140-1915 BC

OZM246

Gangangwa

surface

charcoal

3744 ± 383

2283-2033 BC

Figure 8. Table showing AMS 14C dates for samples taken from the surface scatters at Huoshiliang and Gangangwa sites (Dodson et al. 2009; Dodson et al. 2011; Atahan et al. 2011).

Figure 9. Diagram showing plot of δ13C and δ15N data measured from herbivore (A & B) and omnivore and carnivore (B) bone collagen samples from Huoshiliang and Gangangwa sites (Atahan et al. 2011).

178

J. Dodson, P. Atahan and X. Li: Unravelling Farming and Metallurgy in Ancient China with Nuclear Science technologies along an east-west axis and the evolution of agricultural systems and cultures is amenable to these subtle yet powerful techniques, and may help to flesh out the story of humankind beyond the records of war and conquest that dominate the earliest written records.

Dodson, J. et al. 2012. Cranial metric, age and isotope analysis of human remains from Huoshiliang, western Gansu, China. Terra Australis 34: 177191. Dodson, J. et al. 2013. Origin and spread of wheat in China. Quaternary Science Reviews 72: 108-111. Dodson, J. 2014. Radioactive carbon in environmental science, in Grice K (ed.) Principles and Practice of Analytical Techniques in Geosciences: 271-284. London: Royal Society of Chemistry. Flad, R. et al. 2010. Early wheat in China: results from new studies at Donghuishan in the Hexi Corridor. The Holocene 20: 955-965. Heaton, T. 1987. The 15N/14N ratios of plants in South Africa and Namibia: Relationship to climate and coastal/saline environments. Oecologia 74: 236246. Jim, S. et al. 2004. Stable carbon isotopic evidence for differences in the dietary origin of bone cholesterol, collagen and apatite: implications for their use in palaeodietary reconstruction. Geochimica et Cosmochimica Acta 68: 61-72. Lee-Thorp, J. 2008. On isotopes and old bones. Archaeometry 50: 925-950. Li, X. et al. 2007. Early cultivated wheat and broadening of agriculture in Neolithic China. The Holocene 17: 555-560. Li, X. et al. 2011. The impact of early smelting on the environment of Huoshiliang in Hexi Corridor, NW China, as recorded by fossil charcoal and chemical elements. Palaeogeography, Palaeoclimatology, Palaeoecology 305: 329-336. Linduff, K. et al. 2000. The Beginnings of Metallurgy in China. New York: Edwin Mellen Press. Linduff, K. and J. Mei 2009. Metallurgy in ancient eastern Asia: retrospect and prospects. Journal of World Prehistory 22: 265-281. Mei, J. 2009. Early metallurgy and socio-cultural complexity: archaeological discoveries in Northwest China, in B. Hanks and K. Linduff (eds) Social Complexity in Prehistoric Eurasia: Monuments, Metals and Mobility: 215-232. Cambridge: Cambridge University Press. Murphy, B. and D. Bowman 2006. Kangaroo metabolism does not cause the relationship between bone collagen δ15N and water availability. Functional Ecology 20: 1062-1069. Piperno, D. and K. Flannery 2001. The earliest archaeological maize (Zea mays L.) from highland Mexico: new accelerator mass spectrometry dates and their implications. Proceedings of the National Academy of Science 98: 2101-2103. Roberts, B. et al. 2009. Development of metallurgy in Eurasia. Antiquity 83: 1012-1022. Schwarcz, H. et al. 1999. 15N enrichment in the Sahara: In search of a global relationship. Journal of Archaeological Science 26: 629-636.

References Ambrose, S. and L. Norr 1993. Experimental evidence for the relationship of the carbon isotope ratios of whole diet and dietary protein to those of bone collagen and carbonate, in P. Lambert and G. Grupe (eds), Prehistoric Human Bone: Archaeology at the Molecular Level:1-37. Berlin: Springer. Atahan, P. et al. 2011. Subsistence and the isotopic signature of herding in the Bronze Age Hexi Corridor, NW Gansu, China. Journal of Archaeological Science 38: 1747-1753. Barton, L. et al. 2009. Agricultural origins and the isotopic identity of domestication in northern China. Proceedings of the National Academy of Sciences 106: 5523-5528. Bellwood, P. 2005. The First Farmers: Origins of Agricultural Societies. Oxford: Blackwell Publishing. Billy, C. et al. 2010. Nitrogen isotopic composition of leached nitrate and soil organic matter as an indicator of denitrification in a sloping drained agricultural plot and adjacent uncultivated riparian buffer strips. Soil Biology and Biochemistry 42: 108-117. Bocherens, H. and D. Drucker 2003. Trophic level isotopic enrichment of carbon and nitrogen in bone collagen: case studies from recent and ancient terrestrial ecosystems. International Journal of Osteoarchaeology 13: 46-53. Britton, K. et al. 2008. Stable isotope evidence for saltmarsh grazing in the Bronze Age Severn Estuary, UK: Implications for palaeodietary analysis at coastal sites. Journal of Archaeological Science 35: 2111-2118. Crawford, G. et al. 2005. Late Neolithic plant remains from northern China: preliminary results from Liangchengzhen, Shandong. Current Anthropology 46: 309-17. Crawford, G. 2009. Agricultural origins in North China pushed back to the Pleistocene-Holocene boundary. Proceedings of the National Academy of Sciences 106: 7367-7372. DeNiro, M. and S. Epstein 1976. You are what you eat (plus a few per mil): the carbon isotope cycle in food chains. Geological Society of America Abstracts with Programs 8: 834-835. De Ryck, I. et al. 2005. An overview of Mesopotamian bronze metallurgy during the 3rd millennium BC. Journal of Cultural Heritage 6: 261-268. Dodson, J. et al. 2009. Early bronze in two Holocene archaeological sites in Gansu, NW China. Quaternary Research 72: 309-314. 179

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Smith, B. and S. Epstein 1971. Two categories of 13C/12C ratios for higher plants. Plant Physiology 47: 380384.

Yasuda, Y. 2002. Origins of pottery and agriculture in East Asia, in Y. Yasuda (ed.) The Origins of Pottery and Agriculture: 15-38. New Delhi: Roli Books.

180

Chapter 11

Early Wheat Cultivation and Plant Use in Xinjiang Prehistory: New Evidence Based on Starch Analysis Peter Jia and Florence Chau The history of plant use in prehistoric Xinjiang is poorly understood as little material has been recovered archaeologically from domestic contexts. Few settlement sites have been excavated, although there has been a strong focus on cemeteries which have generally not yielded botanical material (Jia et al. 2009, 2017; Archaeological Institute 2013; Wang 2014; Xiaohe Archaeological Team 2005; Xinjiang Institute of Archaeology 2014). Despite the limited direct evidence, scholars have assumed that by the Bronze Age, agriculture was already being practiced in the oases and along the mountain foothills and river courses (Chen and Hiebert 1995). In later periods, agriculture is again presumed to have coexisted as a local economic strategy alongside groups of nomadic pastoralists that entered the region (Mei and Shell 1999). More recently there has been considerable discussion about the introduction of wheat into China and the movement of millets out of China towards the west (Betts et al. 2014; Stevens et al. 2016), but most evidence comes from the regions on either side of Xinjiang, in eastern Kazakhstan and the Hexi Corridor (Dodson et al. 2013; Flad et al. 2010; Frachetti and Mar’yashev 2007; Spengler et al. 2014; Spengler and Willcox 2013). Xinjiang has rarely been the focus of this research. In addition, more generally, it remains unclear when, how and which crops or plants other than millets and wheat (Triticum aestivum) were exploited in prehistoric Xinjiang.

New study of plant use in the prehistoric periods in Xinjiang may answer some of these questions and, as plants were shared across cultures, it may also provide more evidence for early cultural interaction between ancient China and the West. Starch residue analysis was used as a tool of archaeological research in Xinjiang in the 1990s, associated with the study of the Chawuhugou cemetery (Xinjiang Institute of Archaeology 1999). Although this was a preliminary study, it was possible to identify starch traces of millets, wheat and barley (Hordeum vulgare) on the inside of a ceramic pot. Starch residue analysis can be also used as a technique to examine the function of stone tools. The presence of ground stone tools on archaeological sites is often taken to be an indicator of an agricultural economy, assuming, without proof, that they were used to process cultivated cereals. This assumption has rightly been challenged recently through starch residue analysis, showing that plants other than cultivated cereals may have also been processed (Liu et al. 2010; Liu et al. 2013; Liu et al. 2011; Yang et al. 2012). A large number of grinding stones have been found at sites across Xinjiang, again originally believed to be direct indicators of a farming economy. Given the now demonstrated variability in the use of stone tools, this assumption has been tested here through starch residue analysis. In this chapter we present the results of starch residue analysis from several archaeological sites in Xinjiang dated from c. 2300 – 1000 BC (Figure 1). This is the first study of its kind from this region.

Despite the lack of real data described above, it is essential for the wider scholarship of the ancient world to understand how wheat farming spread into Central China, where the agricultural economy began to intensify in the mid to late 3rd millennium BC (Betts et al. 2014). Diversifying into wheat cultivation allowed farmers to maintain adequate cropping yields when faced with environmental uncertainty. The spread of wheat-farming into Central China also facilitated the development of social complexity during the late Neolithic and subsequently helped to trigger early state formation in Chinese history (Liu and Chen 2003). Xinjiang is a key region across which these cereal exchanges must have taken place, but many unanswered questions remain in relation to the early interaction between the East and West. One of them is how wheat farming spread from Central Asia through Xinjiang into the Yellow River area (Betts et al. 2014).

Methodology Starch residue analysis involves four major processes: creating a plant reference collection, extracting samples for analysis from both the reference collection and archaeological specimens, preparing samples for the laboratory, and observations of the materials under a microscope. Identification of starch residue requires a relatively complete modern plant reference collection. A database of modern starch granules from local plants is crucial (Field 2006; Fullagar 2006). Stage One of our research began with the collection of 34 species of referent plants from the northern Tianshan regions (Figure 2) including 11 plants of contemporary 181

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 1. Diagram showing sites discussed in this chapter.

economic importance and 23 plants that have no economic value. Starch was extracted from all referent plants. Next, stone tools were selected from four archaeological sites, Kalasu (Xinjiang Archaeological Institute 2008), Sa’ensayi (Xinjiang Archaeological Institute 2013), Luanzagangzi (Jia et al. 2011), and Adunqiaolu (IACASS 2015; Jia et al. 2017) for analysis of starch residue. All these sites are distributed along the Tianshan Mountains. Sa’ensayi, near Urumqi, in the central region of the northern Tianshan, is the earliest, with samples taken from a grave of the Qiemu’erqieke type, one of the early cultural groups of the Xinjiang Bronze Age (Betts et al. 2018; Jia and Betts 2010). A single C14 date places it at 2470 cal. BC (Archaeological Institute, 2013), a date that is compatible with other evidence for the Qiemu’erqieke culture (Jia and Betts 2010). In the next stage of Xinjiang’s Bronze Age, Kalasu has ceramics of Andronovo affinity (Doumani Dupuy et al. 2019) and, on the basis of relative dating, can probably be placed around the mid-2nd millennium, c. 1500 BC. Adunqiaolu also shares cultural affinity with the wider Eurasian Andronovo. A robust sequence of C14 dates for the Adunqiaolu cemeteries and settlement site gave a range of c. 1800 – 1400 cal. BC (Jia et al. 2017). Luanzagangzi, an occupation site, is the latest with

eight C14 samples providing a range of dates from c. 1300 to 900 cal. BC (Jia et al. 2011). Stages Three and Four cover the laboratory processing of both comparative and archaeological samples and microscopy. For two of the sites, the Kalasu settlement site and Saensayi Cemetery, starch residue samples were extracted and analysed from stone tools held in the storage facilities at the Xinjiang Institute of Archaeology and Cultural Relics at Urumqi. For Luanzagangzi and Adunqiaolu, the samples were collected directly from a selection of stone tools at the in-field excavations. In all cases, morphological identification indicated that the selected stone tools were grindstones and pestles, possibly used for processing plant seeds. All the selected tools were cleaned with a nylon brush. Starch free gloves were used throughout the extracting process to avoid any possible contamination (Fullagar 2006). Stage One: Establishing a modern starch reference base for Xinjiang So far, there is no robust modern starch reference database for Xinjiang. For Stage One of our project, it 182

P. Jia and F. Chau: Early Wheat Cultivation and Plant Use in Xinjiang Prehistory

Species

Sample No.

N

Mean Length

SD

Median

Range (μm)

Leguminosae

Phaseolus calcaratus

BXY-11

115

22.97

11.32

23.03

4.25-55.51

Family Economic Leguminosae

Phaseolus vulgaris

BXY-12

164

17.23

8.65

16.15

3.42-40.47

Leguminosae

Pisum sativum

BXY-10

117

21.56

7.15

20.77

6.84-41.28

Leguminosae

Vigna radiate

BXY-05

133

19.17

6.63

19.45

4.80-33.06

Poaceae

Avena sativa

BXY-01

111

6.19

1.65

6.09

2.81-13.39

Poaceae

Hordeum vulgare

BXY-04

114

7.68

6.08

4.72

1.36-28.50

Poaceae

Panicum miliaceum

BXY-14

133

7.48

1.52

7.65

3.20-10.92

Poaceae

Sorghum bicolor

BXY-13

131

11.34

5.59

10.63

1.80-23.07

Poaceae

Triticum aestivum

BXY-08

116

7.14

4.76

5.57

2.07-25.44

Poaceae

Triticum aestivum

BXY-09

179

6.94

5.21

5.30

2.41-26.56

Apiaceae

Ligusticum chuanxiong

JS-29

107

7.17

2.80

6.80

1.80-16.32

Cynomoriaceae

Cynomorium coccineum subsp. Songaricum

BXY-07

123

14.58

5.85

14.47

3.31-25.78

Leguminosae

Astragalus membranaceus var mongholicus

BX-12

102

3.43

1.15

3.38

1.29-6.95

Leguminosae

Glycyrrhiza uralensis

JS-24

120

3.73

1.77

3.19

1.52-10.42

Pinaceae

Pinus koraiensis

RFE-01

105

2.82

0.92

2.63

1.45-5.61

Polygonaceae

Fagopyrum esculentum

BXY-02

113

6.44

2.41

6.07

2.52-12.94

Rosaceae

Crataegus altaica

JS-25

114

4.51

1.46

4.59

1.31-7.73

Scrophulariaceae

Cistanche deserticola

JS-22

159

19.32

10.85

17.02

3.22-58.07

Non-Economic

Figure 2. Table showing maximum length (mm) of starch granules in the comparative reference taxa used for this project (Jia and Chau).

was necessary to collect a starch reference database, gathered gradually over the past few years. The database comprised 34 species, some from Gansu Agricultural University, some from the Botanical Gardens, and some collected directly from the field on the northern slope of the Eastern Tianshan Mountains, near Quanzijie Village in Jimusar County. It comprises a variety of plant species, including cereals, legumes, nuts, and some tuber and stem plants such as Cistanche (Cistanche deserticola) provided by Dr Feng Ying.1

from the surface of the depression. This water can then be sucked into the pipette and put into a plastic tube to store the specimen for later processing in the laboratory. Stage Three: Laboratory protocol for isolating starch granules from archaeological samples It was possible with a small number of samples with very clean water to place them directly on a glass slide ready to be studied under the microscope. The water was mixed with 50% glycerol, covered with a glass sheet and sealed off with nail varnish. For most water samples it was necessary first to float them with heavy agents such as lithium heteropoly tungstates (LST), supplied by Central Chemical Consulting Pty Ltd. The specific gravity (SG) was adjusted to 2.00 with distilled water, and 3 ml of the LST heavy agent was mixed with the sample, and then centrifuged at 1000 rpm for 15 minutes. After centrifugation, the floating material on the surface of the heavy agent was aspirated and placed in a new test tube. This process was repeated four times for each sample.

Stage Two: Residue sample extraction from archaeological samples Each ground stone implement was examined carefully before sampling using a portable microscope (100X) to identify suitable sampling points such as small cracks or holes where starch granules may have been trapped. Samples were collected by employing the ‘release inhalation method’, using a 10 ml pipette to place 2-3 ml of ultra-pure water on the surface of the selected spot on the surface of the stone tool. The water droplets quickly disperse into the cracks or holes. This process is repeated several times until water can be seen seeping

Distilled water was mixed again with the processed sample, which was placed in a centrifuge and spun at 1500 rpm for five minutes. The surface water was then

Dr. Ying Feng at the Natural Museum of Institute of Geography, Xinjiang Branch of Chinese Academy of Sciences identified the botanic names of the reference collection.

1 

183

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads poured off, retaining the residue at the bottom. This process was repeated three times. In the next stage, acetone was used to repeat the above process twice in order to clean out the heavy agent remaining in the sample. Finally, the sample was placed in an open test tube to dry for 12 hours and then mixed with 50% glycerine in preparation for placing on a slide.

differ from those of the others in both shape and formal characteristics (Figures 2-5). The Structure of Starch Granules in the Leguminosae family The starch granules of the Leguminosae family (Figure 4) tend to be oval or circular in shape, with some more kidney shaped (Phaseolus sp., Pisum sativum and Vigna radiata). The hilum is usually located at the centre with radiating long thin grooves. Concentric circles on the surface of Leguminosae starch granules are sometimes visible.

Stage Four: Microscopy A Zeiss Axioskop MAT2 with a HRc digital camera and Zeiss AxiovisionTM digital archiving software were used in this analysis, which was conducted at the Australian Centre of Microscopy & Microanalysis, University of Sydney. Slides were scanned using cross-polarised filters under 10x, 20x and 63x (oil) objectives. Images were taken using Differential Interference Contrast (DIC). DIC, also known as the Nomarski technique, is a method that gives contrast to unstained specimens and requires the use of both polarising filters and Wollaston-Nomarski filters. The resulting image reveals the internal structure of the specimen, which may provide a pseudo-relief shading that appears like surface topography (Barton and Fullagar 2006).

The Structure of Starch Granules in the Grasses family (Poaceae) Starch granules of grasses (Figure 5) are generally circular or nearly circular, with the hilum located at the centre. Many species of Poaceae starch granules, like oat (Avena sativa), broomcorn millet (Panicum miliaceum) and sorghum (Sorghum bicolor), have a polyhedric form. Some Poaceae starch granules also have a radiating hilum point. The polygon may be due to the closely arranged starch granules inside the plant cells. On many starch granules, such as those of sorghum, millet and barley, the hilum points of the surface folds and grooves are visible. On barley, sorghum, and wheat, the starch granules display concentric circles, but this is not the case for oats and broomcorn millet.

Starch grains were initially identified under crosspolarising filters using the birefringence; a property of the highly ordered structure and layering of the amylose and amylopectin molecules within the starch grains. The starch grains exhibit a characteristic extinction cross, a property of the way that the compositional molecules of the starch are arranged and visible as a dark cross over the birefringent background of the starch grain. Maximum length measurements through the hilum of the starch grains were made (Field 2006, Liu et al. 2010), as well as observations of other morphological attributes.

Oat starch granules are the only type in the grass family featuring complex particle structure (Figure 5A). Isolated oat starch granules may be formed during grinding (Figure 5B). Except for barley and sorghum, the sizes of the starch granules of most grass plants are very similar. With wheat, although its sizes are very concentrated, the distance between the largest and smallest ones is rather great. This is probably due to the bimodal distribution of the particles, i.e. the formation of large and small groups (Figure 5G). Wheat starch granules are slightly discoid in shape with a distinct rim which is visible when viewed side on (Figure 5G: indicated by white triangles). These unique typological characteristics of wheat starch granules make them generally easier to identify than those of other species.

Identification of Starches in Reference Collection Analysis of specimens from the reference collection illustrated how starch granules contained in different plants present various distinctive characteristics in size, shape, and structure, features that are critical in enabling identification of species. Generally speaking, the starch granules from economic cereals, medicinal herbs, and pine nuts are small, but legumes (Leguminosae family) and two medicinal herbs, Herba Cynomorii (Cynomorium songaricum), named Suoyang or 锁阳 in Chinese, and Cistanche contain relatively large starch granules. Cistanche starch granules exhibit a broad spectrum of sizes with the large granule dimensions falling between 40µm – 58 µm. Barley starch granules fall into two major types or size groups; one group is relatively small and another quite large. Wheat granules have an unusual discoid formation. The starch granules of legumes such as ricebean (Vigna umbellata), and pea (Pisum sativum)

Collection of Starch Samples from Stone Tools Sample collection from archaeological specimens was undertaken in the period from 2007 to 2010. Only a small number of starch granules was acquired. For example, at Kalasu twenty-three samples were collected from eleven stone tools but only seven samples (29%) were found to contain starch residue. Of these, two samples were too poorly preserved for analysis, leaving five 184

P. Jia and F. Chau: Early Wheat Cultivation and Plant Use in Xinjiang Prehistory

Figure 3. Table showing boxplots of starch granule maximum length of the comparative reference collection data. Numbers of starch granules measured for each sample are in brackets (Jia and Chau).

comparison with modern starches from the project reference collection, the starch granules were given a preliminary identification as wheat, barley and other cereals. The identifiable plants also included the Leguminosae family, possibly such as adzuki bean (Vigna angularis), common bean (Phaseolus vulgaris), or mung bean (Vigna radiata). Some stem and tuber starchy plants like Cistanche and Herba Cynomoriiare were also discovered. Identification of the residues is described below.

samples for processing (09XJ003-2, 09XJ009-1, 09XJ0092, 09XJ010-1 and 09XJ010-2). Cleaning the surface of the stone tools with water after excavation as a normal part of processing the finds may have reduced the recovery of starch.2 Starch residues were recovered from samples collected from four sites (Figures 7, 8). At Luanzagangzi two samples, JQLY07-06 and JQLY07-07, collected from the stone tools of JQLYT1⑧: 2 and JQLYT1⑨: 2 contained starch granules. Starch granules were also found on two grinding stones excavated from Burial No. 38 at the Saensayi cemetery (Jia 2013). At the Kalasu site starch residue was recovered from three grinding stones (07YNKF1: 56, 07YNK④: 1② and 07YNKF1: 59①) excavated from a dwelling (F1). One grinding stone from Adunqiaolu was selected for starch residue sampling. After typological

Wheat Wheat starch residue was recovered from seven samples, among which two were collected from a stone mortar and one from a grinding stone found at Burial No 38 at the Saensayi cemetery (Figure 8 lower row; Figures 9, 10, 11). The other four samples were collected from three grinding stones found in a house pit at the Kalasu site (Figure 12).

2  Six stone artefacts from Saensayi cemetery were not cleaned. Of a sample size of 23, 16 (70%) samples had starch residues.

185

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 4. Starch granules of Leguminosae family: A: 黄芪 (Astragalus membraneus var mongholicus); B, C: adzuki bean (Phaseolus calcaratus); D, E: haricot bean (Phaseolus vulgaris); G: green pea (Pisum sativum); H: mung bean (Vigna radiate); I: 甘草 (Glychirrhiza uralensis) (A, I from tuber; others from seed (photomicrography by Chau using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC)).

Figure 5. Gramineae starch: oat, Avena sativa (A and single grain particles found B); barley, Hordeum vulgare (C, D); millet – Panicum miliaceum (E); Sorghum bicolour (F); wheat, Triticum aestivum (G, H, I) (photomicrography by Chau using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC)).

186

P. Jia and F. Chau: Early Wheat Cultivation and Plant Use in Xinjiang Prehistory

Figure 6. Other species of starch plants: A: Chinese lovage (Ligusticum chuanxing); B: Cynomorium (Cynomorium coccineum subso Songarium); C: Korean Pine (Pinus koraienisis); D: buckwheat (Fagopyrum esculentum); E. Altai Hawthorn (Crataegus altaica); F, G, H, I: desert cistanche (Cistanche deserticola) (photomicrography by Chau using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC)).

Figure 7. Stone and specimen collection site (white square): left from Kalasu site C: 09XJ002; D: 09XJ003; EF: 09XJ003; GH: 09XJ004; right from Adunqiaolu site XBWAF1-14 (photos Jia and Chinese Academy of Social Sciences).

187

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 8. Specimen collection and stone tools: upper 1-2, JQLYT1⑧: 2; 3-4, JQLYT1⑨: 2 from Luanzagangzi site; bottom left 09XJ013-2, 3 and right 09XJ014-1, 2 from burial No 38 at Saensayi cemetery (photos Jia).

Figure 9. Table showing archaeological samples. Boxplots of starch grain maximum length (Jia and Chau).

morphology. A key difference between the two is that wheat starch usually appears as many granules gathered in a cluster, a feature which is rarely seen in barley starch. Under the microscope, a common polarising effect on wheat starch granules is a very clear extinction cross, while barley starch normally exhibits

Barley Some starch granules similar to wheat starch were found on the stone tools from the Kalasu site (Figure 13). However, these have been provisionally identified as barley granules based on some aspects of their 188

P. Jia and F. Chau: Early Wheat Cultivation and Plant Use in Xinjiang Prehistory

Figure 11. Wheat starch recovered from grinding stone (Figure 8 bottom right 09XJ014-1, 2), burial 38 in Saensayi cemetery (photomicrography by Jia using transmitted darkfield microscopes applied with crosspolarised light).

Figure 10. Wheat starch recovered from stone mortar (Figure 8 bottom left 09XJ013-2, 3), burial 38 in Saensayi cemetery (photomicrography by Jia using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC)).

Figure 12. Wheat starch recovered from grinding stone at Kalasu site: A. 09XJ011-2; B. 09XJ010-1; C. starch granules of modern wheat (Trticum aestivum) (photomicrography by Jia: on the left using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC), and on the right using darkfield microscopes applied with cross-polarised light).

189

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 13. Modern reference granules of wheat (Triticum aestivum) (A) and barley (Hordeum vulgare) (B), compared with ancient starch granules found on stone tools (09XJ010-1) (C) believed to be barley starch granules (photomicrography by Jia: on the left using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC) and right on the right using darkfield microscopes applied with cross-polarised light).

190

P. Jia and F. Chau: Early Wheat Cultivation and Plant Use in Xinjiang Prehistory a less distinct dark zone. In addition, wheat starch granules tend to be more rounded, nearly spherical, while barley granules are relatively flat. The Kalasu granules appeared singly, not in clusters, had a dark zone rather than a clear cross and were rounded rather than flattened. On the basis of these characteristics, they were provisionally identified as barley.

concluded that they are likely to be the starch granules of millets, either broomcorn millet or foxtail millet (Setaria italica). Others are similar to the starch granules of oats. Some larger starch granules may be sorghum or buckwheat (Fagopyrum esculentum). While wheat and barley starch granules are common, the number of other cereal starch granules is low, and the features are not obvious, so that identification remains vague. Further evidence of large seeds is needed to support these findings. In addition, the cultivation of sorghum at early archaeological sites is rather controversial (Zhang 1996) and more substantial proof is needed before this can be fully confirmed.

Other cereals We also found small granules with a multilateral/ polyhedral shape. When comparing these with the reference plants in our database (Figures 14-17), we

Beans (Leguminosae)

Figure 14. Modern reference starch granules: left – oats (Avena sativa), right – broomcorn millet (Panicum miliaceum) (photomicrography by Jia: on the left using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC), and on the right using darkfield microscopes applied with cross-polarised light).

Figure 15. Starch granules with multilateral/polyhedral shape found at Kalasu site: 1-5, 7-12 from sample 09XJ010-1; 6 from sample 09XJ003-2 (photomicrography by Jia: on the left using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC) and on the right using darkfield microscopes applied with cross-polarised light).

191

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 16. Modern reference starch granules – sorghum (Sorghum bicolour). (photomicrography by Jia: on the left using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC) and on the right using darkfield microscopes applied with cross-polarised light).

Figure 17. Modern reference starch granules – buckwheat (Fagopyrum esculentum) (photographs by Chau: on the left using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC) and on the right using darkfield microscopes applied with cross-polarised light).

192

P. Jia and F. Chau: Early Wheat Cultivation and Plant Use in Xinjiang Prehistory

Figure 18. (A-E) Ancient starch granules from sample 09XJ 018; (F-H) starch granules with prominent lamellae from sample 09XJ 010; starch granules from the reference collection species: (I) haricot bean (Phaseolus vulgaris); (J) mung bean (Vigna radiata); (K) garden pea (Pisum sativum); (L) adzuki bean (Phaseolus calcaratus) (photographs by Chau using brightfield microscopes applied with Differential Interference Contrast (DIC).

Some starch granules, such as those from sample 09XJ010, are similar to the haricot bean (Phaseolus vulgaris), mung bean, garden pea (Pisum sativum) or adzuki bean (Figure 18). They can be attributed to the bean group, but it is difficult to identify which species they belong to based only on a small sample size.

medicine for thousands of years (CMBA 2014; CMDDHHK 2014a, b). Herba Cynomorii starch granules tend to be round, of medium size and with a centric hilum. The starch granules recovered from Kalasu are similar to Herba Cynomorii starch granules. We noted that the Cistanche starch granules from the reference plants in the database are highly unusual. They have a wide size range from less than 5 microns to over 57 microns. The granules are ovoid in shape with an eccentric hilum. The hila tend to be large, with fissures. The lamellae on these granules are

Medicinal plants Some starch granules point to medicinal plants, for instance Herba Cynomorii (Figure 19) and Cistanche (Figures 20-22), which have been used in Chinese 193

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 19. Possible suo-yang (Songaria Cynomorium) starch granules found on the stone tools at Kalasu site: ancient starch 1, 3, 4 (09XJ010-1), 2 (09XJ009-2), 5, 6 (modern reference) (photomicrography by Jia: on the left using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC) and on the right using darkfield microscopes applied with cross-polarised light).

Figure 20. Starch granules of da-yun (Cistanche deserticola) extracted from modern reference plant (photomicrography by Jia: using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC) except in the case of the one in the middle of the top row which used darkfield microscopes applied with cross-polarised light).

clearly present on the surface and crossed by radiating wrinkles. The starch granules have a compound form comprising a large ovoid granule joining a small

granule, but sometimes a few small ovoid forms are bound to each other. Once the granules have separated, they may exhibit a mark on the surface. 194

P. Jia and F. Chau: Early Wheat Cultivation and Plant Use in Xinjiang Prehistory

Figure 21. Da-yun (Cistanche deserticola) starch granules recovered from samples 11XJ01-03 and 11XJ01-05 collected from a grinding stone XBWAF1-14 from Adunqiaolu site. The two on the left seem flattened, perhaps due to food processing (photomicrography by Jia: using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC) and with cross-polarised light).

Mar’yashev 2007). However, wheat has been found in some abundance at Gumuguo in the Tarim Basin (Flad et al. 2010; Abuduresule et al., in this volume) in contexts dated from the later 3rd millennium BC. Wheat also occurs quite commonly in burials at Xiaohe. The volume of grains recovered and the wide area over which the early Xinjiang pastoralists ranged, far distant from settled agricultural villages, strongly suggests that they were practicing some cultivation of their own. By the mid-2nd millennium BC or earlier, sites with a full agricultural economy are present in Xinjiang, as illustrated at Luanzagangzi (Jia et al. 2011; Zhang et al. 2017). The cereals and legumes identified through starch residue analysis add substantially to this story. The crops identified include wheat, barley, millets, and various beans, suggesting a diversified prehistoric agrarian economy. As all the crops are still cultivated in Xinjiang and Central Asia today it is most likely, particularly for the later periods, that they were locally grown rather than having been imported from elsewhere, but this can only be fully verified when more substantial data become available. Two specific dry land wild plants, Herba Cynomorii and Cistanche, were also probably exploited (Figure 23).

Cistanche has been identified from two sites, Adunqiaolu and Luanzagangzi. Comparing the ancient starch granules with those of reference plants, we find that the ancient granules are varied in form, probably deformed during the grinding process. For instance, granules from Adunqiaolu appears to have been flattened somehow (Figure 21 left), and a few from Luanzagangzi appear to be deformed and also flattened (Figure 22). All these deformations are very likely to be the result of food processing. Discussion Our study has shown that a variety of plants were exploited in Xinjiang from the late 3rd to the end of the 2nd millennium BC. This adds important data to our scanty knowledge of plant use in Xinjiang at this time. Based on paleobotanical data other than starches, in the early part of this period, with the exception of the oasis of Hami, the region was occupied by agro-pastoralists (Betts, this volume). Cereals, wheat and millet, dated to the 3rd millennium BC have been reported from eastern Kazakhstan where they have been interpreted as traded and used in ritual contexts (Frachetti and 195

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads

Figure 22. Ancient da-yun (Cistanche deserticola) starch granules found from stone tools unearthed at Luanzagangzi site (photomicrography by Jia: using transmitted brightfield microscopes applied with Differential Interference Contrast (DIC), and with cross-polarised light).

Date

Adunqiaolu

Saensay Cemetery

Kalasu

Luanzagangzi

2000 BC

1300 BC

2200 BC

1500 BC

Wheat (Triticum aestivum)

√

√

Barley (Hordeum vulgare)

√

√

Oats (Avena sativa)

√

Foxtail millet (Setaria italic)

√

√

Broomcorn millet (Panicum miliaceum)

√

√

Beans (Leguminosae): Haricot bean (Phaseolus vulgaris) Mung bean (Vigna radiata) Garden pea (Pisum sativum) Adzuki bean (Phaseolus calcaratus)

√

Suo-Yang (Songaria Cynomorium)

√

Da-Yun (Cistanche deserticola) Archaeological Culture

√

√ √

Qiemu’erqieke/ Andronovo

Qiemu’erqieke

Andronovo

Luanzagangzi

Figure 23. Table showing plant starches identified from the archaeological sites mentioned in this chapter (Jia and Chau).

Our starch residue analysis has shed light on early wheat and barley consumption. The starch evidence found in Burial No. 38 at the Saensayi cemetery indicates the presence of wheat and barley in the Tianshan region by at least the late 3rd millennium BC. Having a close look at the early phase of the Saensayi cemetery, particularly Burial No. 38, we find some similarities

with the ritual practice and material culture of the Qiemu’erqieke culture found in the Altai region dated to c. 2400 BC (Jia and Betts 2010). This early Bronze Age culture is believed to have links with Xiaohe in the south and Tianshanbeilu in the east (Betts et al. 2018). The wheat starch granule residues suggest a possible route through which wheat came from the west, along 196

P. Jia and F. Chau: Early Wheat Cultivation and Plant Use in Xinjiang Prehistory the Tianshan, and eastward to the Hami region, and then further eastward during the early Bronze Age.

earliest evidence for exploitation of these plants in this region, possibly for both food and medicine.

The starch evidence also shows that during the late Bronze Age wheat and barley, by now certainly cultivated, was supplemented by oats and various beans from Central Asia. In addition, as shown by the starch residue recovered from the Kalasu and Luanzagangzi sites (Jia et al. 2011), the millets from the east, that is the Yellow River area, were present in Xinjiang almost in the same period. The starch evidence shows that prehistoric people in Xinjiang had adapted well to local environments by modifying their economic activities and exploiting plants from different and dispersed regions. For instance, the house types and ceramics from the Kalasu site are very similar to those of the Eurasian Andronovo culture. Although the Andronovo economy is based mainly on agro-pastoralism at the very early stage before expansion, it has traditionally been considered to be a livestock-based economy after expansion to southern and eastern Central Asia (Kuz’mina 2007, 2008). The Andronovo remains found at the Kalasu site present an economy where multi-crop cultivation prevailed. Similar discoveries, although with more limited crop cultivation, have been reported at the Tasbas site of eastern Kazakhstan around 400 km west of Kalasu, suggesting adoption by similar groups of people in diverse environments (Spengler et al. 2014). In the Semirech’ye region in southeastern Kazakhstan, a carbonized wheat grain was discovered in an ash-pit next to the burials at the Begash site dated to the late 3rd to the early 2nd millennium BC (2460-1950 cal. BC) (Frachetti et al. 2010). Together with Kalasu and Saensayi, the Tasbas and Begash sites may be points along a route of eastern expansion of wheat down the Tianshan corridor (Frachetti and Mar’yashev 2007). Herba Cynomorii and Cistanche taxa, which were identified from Adunqiaolu and Luanzagangzi through our starch residue analysis, have been used in Chinese traditional herbal medicine for thousands of years. Herba Cynomorii is one of the most commonly used ‘potency-invigorating’Chinese traditional medicine,3 in pharmaceutical practices prescribed for treating lumbar weakness and enhancing sexual performance for both men and women (Chen 2014). It is mainly distributed across the dry and desert areas of Central Asia and Western China, including Xinjiang, Qinghai, Gansu, Ningxia, Inner Mongolia, and Shanxi province. Cistanche, a common tonic herb which grows in desert areas, is used for improving both memory and sexual potency. It has also been used in Chinese traditional medicine for treating kidney-deficiency syndrome (Qui et al. 2008). The ancient starch granules of these two medicinal plants analysed in our study are the

Conclusion This study of the first stone tool residues found in the Xinjiang region has identified eight utilised plant taxa through analysis of starch granules found on eight grinding stones from four key sites. These new data show that ground stone artefacts were used to process both wild and cultivated plants. The variety of plants identified suggests the practice of multicropping among groups previously assumed to have had a live-stock based economy. The presence of wheat at Saensayi suggests a route for the introduction of wheat into China, while evidence for processing of Herba Cynomorii and Cistanche may indicate a wider use of wild plants as food sources but equally may suggest that they were gathered for medicinal purposes. Traces of wild plant starches on the grindstones show that the presence of such stones is not a definitive sign of agricultural practice and cannot be used as a reliable marker as such. References Barton, H. and R. Fullagar 2006. Microscopy, in R. Torrence and H. Barton (eds) Ancient Starch Research: 47-52. California: Left Coast Press Inc. Betts, A. et al. 2014. The origins of wheat in China and potential pathways for its introduction: A review. Quaternary International 348: 158-168. Betts, A. et al. 2018. A New Hypothesis for Early Bronze Age Cultural Diversity in Xinjiang, China. Archaeological Research in Asia. https://doi. org/10.1016/j.ara.2018.04.001. Chen, J. et al. 2014. Ursolic Acid-Enriched Herba Cynomorii Extract Induces Mitochondrial Uncoupling and Glutathione Redox Cycling Through Mitochondrial Reactive Oxygen Species Generation: Protection Against Menadione Cytotoxicity in H9c2 Cells. Molecules 19: 1576-1591. Chen, W. and F. Hiebert 1995. The late prehistory of Xinjiang in relation to its neighbors. Journal of World Prehistory 9: 243-300. Chinese Medicine Board of Australia (CMBA) 2014. Cross Referenced Nomenclature List of Commonly Used Chinese Herbal Medicines, viewed on 21 March 2019, . Chinese Medicine Division of Department of Health of Hong Kong (CMDDHHK) 2014. Cynomorii Herba, viewed on 24 July 2014. . Chinese Medicine Division of Department of Health of Hong Kong (CMDDHHK) 2014. Cistanches Herba, viewed on 24 July 2014, . Dodson, J. et al. 2013. Origin and spread of wheat in China. Quaternary Science Reviews 2013 (72): 108-111. Doumani Dupuy, P. et al. 2019. Pots and potters of the Bronze Age of north-west Xinjiang. Antiquity 93/371 (2019): 1231–1248. https://doi.org/10.15184/ aqy.2019.136. Field, J. 2006. Reference Collection, in H. Barton and R. Torrence (eds) Ancient Starch Research: 95-113. California: Walnut Creek. Flad, R. et al. 2010. Early wheat in China: results from new studies at Donhuishan in the Hexi Corridor. The Holocene 2010/17: 555-60. Frachetti, M. et al. 2010. Earliest direct evidence for broomcorn millet and wheat in the central Eurasian steppe region. Antiquity 84(326): 993-1010. Frachetti, M. and A. Mar’yashev 2007. Long-Term Occupation and Seasonal Settlement of Eastern Eurasian Pastoralists at Begash. Journal of Field Archaeology 32:221-242. Fullagar, R. 2006. Starch on artefacts, in H. Barton and R. Torrence (eds) Ancient Starch Research: 177-204. California: Walnut Creek. Institute of Archaeology, Chinese Academy of Social Sciences (IACASS), Bortala Mongolian Autonomous Prefectural Museum and Bureau of Cultural Relics, Wenquan County 2015. The Adunqiaolu Site and Cemetery in Wenquan County, Xinjiang. Archaeology 2015(1): 27-32. [中国社会科学院考古研究所、博 尔塔拉蒙古自治州博物馆、温泉县文物局: 《新 疆温泉县阿敦乔鲁遗址与墓地, 《考古》2015 年 1 期, 27-32 页] Jia, P. et al. 2009. Prehistoric archaeology in the Zhunge’er Basin, Xinjiang, China. Journal of Eurasian Prehistory 6 (1-2): 167-198. Jia, P. et al. 2011. New evidence for Bronze Age agricultural settlements in the Zhunge’er (Junggar) Basin. Journal of Field Archaeology 34(4): 269-280. Jia, P. and A. Betts 2010. A re-analysis of the Qiemu’erqieke (Shamirshak) cemeteries, Xinjiang, China. The Journal of Indo-European Studies 38(3 & 4): 275-317. Jia, P. et al. 2017. Adunqiaolu: new evidence for the Andronovo in Xinjiang, China. Antiquity 91 (357): 621-639. Jia, W. 2013. Report for the starch residue analysis of the Saensayi Cemetery. In Xinjiang Institute of Archaeology and Relics: The Saensayi Cemetery of Xinjiang. Beijing: 224-235. Beijing: Relics Press. [贾伟 明: 《萨恩萨伊墓地淀粉残留物分析报告》, 《新 疆萨恩萨伊墓地》224-235 页。2013 年, 北京: 文 物出版社]. Kuz’mina, E. 2007. The origins of the Indo-Iranians. Leiden: Brill. Kuz’mina, E. 2008. The prehistory of the Silk Road. Philadelphia: University of Pennsylvania Press.

Liu, L. and X. Chen 2003. State Formation in Early China. London, Duckworth. Liu, L. and X. Chen 2012. Archaeology of China: From the Palaeolithic to the Early Bronze Age. Cambridge: Cambridge University Press. Liu, L. et al. 2010. What did grinding stones grind? New light on Early Neolithic subsistence economy in the Middle Yellow River Valley, China. Antiquity 84: 816833. Liu, L. et al. 2013. Palaeolithic human exploitation of plant foods during the last glacial maximum in North China. PANS (Proceedings of the National Academy of Sciences of the United States of America) 110 (14): 5380-5385. Liu, L. et al. 2011. Plant exploitation of the last foragers at Shizitan in the Middle Yellow River Valley. Journal of Archaeological Science 38: 3524-3532. Mei, J. and C. Shell. 1999. The existence of Andronovo cultural influence in Xinjiang during the 2nd millennium BC. Antiquity 73: 570-578. Qiu, Y. et al. 2008. Metabolic profiling reveals therapeutic effects of Herba Cistanches in an animal model of hydrocortisone-induced ‘kidney-deficiency syndrome’. Chinese medicine, 3, 3. doi:10.1186/17498546-3-3. Spengler, R. et al. 2014. Late Bronze Age agriculture at Tasbas in the Dzhungar Mountains of eastern Kazakhstan. Quaternary International 348 (2014): 147157. Spengler, R. and G. Willcox 2013. Archaeobotanical results from Sarazm, Tajikistan, an Early Bronze Age Settlement on the edge: Agriculture and exchange. Journal of Environmental Archaeology 18(3): 211221. Stevens, C. et al. 2016. Between China and South Asia: A Middle Asian corridor of crop dispersal and agricultural innovation in the Bronze Age. Holocene 26(10): 1541-1555. Wang, B. 2014. Gumugou. Urumqi: Xinjiang Renmin Press. [王炳华:《古墓沟》2014 年, 乌鲁木齐: 新疆 人民出版社]. Xiaohe Archaeological Team, 2005. New Achievement at Xiaohe Cemetery Archaeology in Lop Nur. Turfan Studies 1:114-125. [新疆文物考古研究所小河考 古队:《罗布泊小河墓地考古发掘的重要收获》, 《吐鲁番学研究》2005 年 1 期, 114-125 页]. Xinjiang Archaeological Institute 2008. Brief report on the Kalasu site, Nileke County, Xinjiang. Xinjiang Relics 3-4: 33-43. [新疆文物考古研究所 2005《尼 勒 克 县喀拉苏遗址考古发掘简报》,《新疆文 物》2008 年 3-4 期，33-43 页]. Xinjiang Archaeological Institute 2013. Xinjiang Saensayi Cemetery. Beijing: Relics Press. [新疆文物考古研究 所: 《新疆萨恩萨伊墓地》, 2013 年, 北京: 文物出 版社]. Xinjiang Institute of Archaeology 1999: Xinjiang Chawuhu – A large-scale clan cemetery excavation report. Beijing: Oriental Press, 431-415. [新疆文物 198

P. Jia and F. Chau: Early Wheat Cultivation and Plant Use in Xinjiang Prehistory 考古研究所 1999: 《新疆察乌呼-大型氏族墓地发 掘报告》。北京: 东方出版社, 431-415 页。 Yang, X. et al. 2012. From the modern to the archaeological: starch grains from millets and their wild relatives in China. Journal of Archaeological Science 39/2: 247-254. Zhang, Z. 1996. Domestic history of Sorghum in Gansu. Journal of agricultural science and technology. 1996(4): 15-16.

Zhang, J. et al. 2017. Cultivation strategies at the ancient Luanzagangzi settlement on the easternmost Eurasian steppe during the late Bronze Age. Vegetation History and Archaeobotany DOI 10.1007/ s00334-017-0608-0.

199

Index

09LE4 Cemetery 21 09LE50 Cemetery 21 09LE8 Cemetery 23 Abbasid 131 Accelerators 171 Aceramic 11, 49 Achaemenid 93, 100, 101-103, 106, 127 Adunquiaolu 2 Afanasevo 97-99, 106 Afghanistan 4, 7, 84, 87, 97-99, 103, 114, 158, 161, 162, 165 Aga’ersen Hoard 12 Agriculture 11, 100, 130, 131, 133, 137, 138, 144, 148, 152, 153, 157, 160, 163, 165, 177, 181 Ahura Mazda 148 Ai Khanum 120 Aketala 15, 92 Ali Arslan Khan 132, 134 Alloy casting 49 Altai Mountains 4, 10, 77 Ammianus Marcellinus 121 Amu-darya 84, 92 An Lushan 125, 126 Anau 84 Andijanians 129 Andronovo 2, 3, 12-15, 62, 67-69, 74, 78, 93, 100, 161, 182, 197 Angkor Wat 5, 152, 157, 159, 160 An-guo 124 Animism 135 Animistic 135 Anthropomorphic stelae 10, 11 Antlers 13 Anuradhapura 5, 152, 155, 156, 160 Apa Khoja 135, 136, 158 Archaeological Institute of the Xinjiang Academy of Social Science 26 Archaeological Institute of Xinjiang Uygur Autonomous Region 29 Achaemenid Empire 93, 102 Arjan 100, 103 Arsenic 172, 173 Artush 131, 132, 133, 136, 137 Australian Aborigines 2 Awl 14 Bactria 4, 99, 102-106, 120, 124, 132, 145, 161, 165 Bactro-Margiana Archaeological Complex (MBAC) 87 Bag 21, 26, 42 Baishantang 173 Balikun (Barkol) 14 Bandykhan 4, 84-87 Banjiegou 14

Barley 6, 11, 100, 114, 157, 161, 174, 176, 181, 184, 185, 186, 188, 190, 191, 195-197 Barbarian 5, 84, 86, 126 Basket 12, 21, 23, 24, 26, 28, 30, 34, 37, 39, 42, 44, 48, 49 Belt Road Initiative 165 Berel 101, 102 Bergman Folke 19, 29 Beshkerem 117, 133, 157 Bibi Maryam 114, 131-133, 136, 157 Black Sea 102, 164 Blood-sweating horses 122 BMAC 87, 93, 99 Boat-shaped coffin 21 Bokhara 129, 131, 136 Bomb-pulse 177 Bone 6, 24, 26, 28, 35, 37, 41, 57, 61, 65-67, 71, 74-76, 86, 93, 178 Bone drill 24 Bone pipe 37 Bortala River Valley 3, 62 Botai Culture 15 Botanical species/macro-remains 86, 173 Bow 21, 77 Britain 19, 136 Bricks / unfired 4, 84, 92, 93 Bronze Age 1-7, 9-15, 19, 48-50, 52-53, 55, 56, 61, 62, 6469, 78, 84, 86, 92, 97, 99, 100, 106, 144, 160-163, 171, 172, 181, 182, 196, 197 Bronze (all others) 5, 12, 22, 35, 37, 42, 48, 49, 57, 61, 62, 64, 66-68, 74-76, 103, 162, 171-173, 177 Buddha 5, 114, 125, 128, 148, 153, 154 Buddhism 120, 124, 125, 127, 128, 132, 143, 148, 149, 152155, 158, 160, 162 Buddhist 5, 104, 114-117, 124-128, 131-133, 138, 143-145, 148, 149, 152, 153, 155-160, 162, 164, 165 Burial 2, 10, 11, 13, 14, 19, 21, 23, 24, 26-29, 31, 34-38, 42, 46-49, 54, 55, 58, 59, 61, 62, 64, 66-68, 70, 71, 74-76, 92, 98-101, 103, 104, 161, 185, 195-197 Burgulyuk 92-94 C14 dating [also see radiocarbon] 11, 37, 38, 42, 182 Camels 77, 161 Canal 84, 87, 133, 148, 149, 156, 157, 159 Cao Quan 125 Carbon 6, 14, 38, 47, 48, 56, 57, 84, 92, 93, 171-173, 174, 176, 177, 197 Caspian Sea 164 Cattle 1, 6, 11, 12, 21, 34, 35, 37, 39, 74, 77, 97, 100, 161, 177 Cattle hides 11, 39 Cattle skulls 34, 35 Cauldron 87, 132, 136, 137, 162 Caves of the Three Immortal Buddhas 153, 154 200

Index Cemetery 2, 10, 11-14, 19, 21-31, 33-50, 55, 56, 58-62, 6569, 71, 74-76, 84, 86, 92, 133, 149, 160, 162, 181, 182, 185, 186, 188, 189, 196 Central Asia/n 4, 6, 7, 12, 50, 53, 62, 64, 69, 84-86, 92, 93, 97-99, 101-103, 106, 113, 114, 120, 122, 125, 127-131, 133, 134, 136, 138, 143, 152, 153, 158, 160, 163, 165, 177, 181, 195, 197 Ceramic/s 11-15, 49, 65-69, 71, 72, 74-80, 84, 87, 92, 93, 98, 181, 182, 197 Cereal 6, 12, 13, 157, 173, 174, 176, 177, 181, 183-185, 191, 195 Ceremonial 49, 99 Chagatai 135 Chakmak River 117, 153, 154, 157 Chang’an 121, 122, 128 Channels 117, 153, 157 Charcoal 173, 177, 178 Chariots 4, 15, 100, 161 Charkilik 158 Chawuhugou 14, 15, 176, 181 Chen Zongqi 24 Chenpan 124, 153 Chilikty 100 China 1, 3, 5-7, 9-11, 13, 15, 19, 52, 53, 64, 92, 93, 97, 99, 100, 102, 103, 105, 113-115, 118-130, 134-138, 143145, 147, 151, 152, 154, 155, 157, 158, 160-165, 171173, 175-177, 181, 197 Chinese chronicles 5, 122, 106 Chinese emperor 113, 121-123, 125, 126, 128-130, 134, 137, 158 Chinese researchers 92 Chinggis Khan 121, 131, 135 Chini Bagh 145 Chisel 12, 57 Christianity 127 Chung [King of Kashgar] 150 Chust Culture 15, 84, 92, 93, 100 Clay shell burial 34 Climate change 144, 162-165 Climatic oscillations 93, 164 Climate patterns 164 Cloak 21, 34 Cloth 2, 21, 23, 24, 28, 37, 48, 76, 125, 127, 133, 155, 157 Cloth doll Clothing 2, 21, 46, 76, 127 Coal 12, 78 Coffin 2, 11, 21, 23, 24, 26, 28, 31, 33, 34, 36-38, 44, 46, 48, 59, 61, 66 Copper 5, 14, 15, 64, 78, 91, 93, 98, 161, 172, 173 Cosmogenic radiation 171 Cotton 121, 130, 156 Da Yuan 123 Dadiwan 6, 173, 177 Dakiyanus 145, 155 Dandan-Uiliq 158, 160 Deer 13, 15, 77, 177 Deer Stone Khirigsuur Complex 13, 15 Diet 6, 11, 106, 171, 175-177

DNA 1, 12, 37, 49, 50, 163 Dombel 158 Domestication 15, 171 Dongheigou 14 Dry faming 10, 14 Dunhuang 133, 143, 155 Earthquake 119, 159 Ecological 93 Elite 4, 5, 101-106, 121, 138, 143, 144, 148, 152, 153, 155, 165 Embroidered 21, 45 Enclosure 60, 62 Endere 144 Environment 2, 5-7, 11, 12, 37, 50, 53, 62, 93, 94, 113, 116118, 131, 137, 143-145, 152, 153, 156, 158-161, 165, 171, 176, 177, 181, 197 Ephedra 3, 24, 26 Eskishahar 5, 150, 151 Eurasia 1, 3-5, 7, 9-15, 19, 49, 50, 52, 53, 62, 64, 67, 78, 93, 97, 100, 101, 106, 165, 174, 182, 197 European 1, 5, 50, 52, 99, 120-122, 130-132, 140, 145, 158, 161, 163 Eylatan culture 93 Faxian 128 Feather 40, 43 Federovo 12, 13, 15, 68 Female symbol 38, 40, 46 Ferghana/Valley 4, 15, 100, 113, 114, 120, 122, 123, 126, 132, 135, 136, 161, 165 Figurines 87, 91 Filippovka 101 First Millennium BC 15, 64, 71, 78, 84, 93, 94, 97, 100, 161, 162 Flax 121 Food chains 121, 175, 176 Fractionation 121, 172, 175, 176 Gandhara 102, 104, 105, 127, 144, 153, 154, 158 Gangangwa 173, 176-178 Gansu 5, 6, 10, 11, 14, 49, 78, 122, 124, 125, 171-174, 177, 183, 197 Gaochang 146 Gilgit 155 Gobi Desert 10 Gold 12, 43, 61, 62, 67, 71, 74-76, 98-100, 105, 122 Gongliu County 12, 74 Google Earth 116-118, 145, 146 Gouge 12 Graeco-Bactrian 103-105 Grape vines 157 Grasslands 9, 13-15, 55, 78 Grave 13, 22, 23, 26-29, 31, 34-36, 42, 46, 48, 49, 54, 59, 61, 62, 64, 66, 67, 69-71, 74-76, 134 Grave goods 2, 3, 6, 12, 49, 66, 67, 70, 71, 74-76 Greek 4, 5, 102, 103, 106 Grinding stone 6, 15, 37, 68, 76, 93, 181, 185, 189, 195, 197 Ground Penetrating Radar – GPR 149, 150 Gumugou Cemetery 19, 26, 28, 46, 47 Gurbantunggut desert 10, 55 201

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Half-life 171 Hami oasis 11-15, 78, 195 Hammer 12, 68, 91 Han 74, 100, 103, 121, 123, 124, 150, 171 Hanshu 122, 123, 150 Han dynasty 74, 122, 123 Hasa Tam 148 Hat 21, 24, 26, 34, 37, 43 Heavenly horses 122 Hede 125 Hedin Sven 19, 24, 48 Hellenic 104, 105 Hemp 121 Hephthalite 125, 128 Hexi Corridor 5, 6, 10, 11, 13, 78, 172, 177, 181 Himalaya Mountains 4, 99, 100, 102, 106, 143 Hoja Khunar 161 Hoja Sefuding Buzurkiwa 161 Hongshankou 14 Horn cup 42 Horse 4, 13, 15, 57, 64, 76, 77, 97, 101, 102, 104, 106, 122, 123, 126, 130, 132, 161 House 2, 3, 12, 14, 24, 46, 49, 52, 54, 55, 57, 58, 73, 74, 114, 115-119, 130, 185, 197 Hou Hanshu 123, 124, 150 Huang Wenbi 19, 24, 48, 64 Hun 106, 122 Hunter-forager-fishers 9 Huntington Ellsworth 28 Huoshiliang 173, 175-178 Hybridization 50 Hydrology 156, 159, 160 Ibn Battuta 135 Idka Khan Mosque 119 Ili (see also Yili) 128, 136 India/Indian 5, 7, 12, 50, 98, 99, 102, 105, 113, 121, 122, 124, 128, 129, 143, 144, 152, 155, 157-160, 164 Indo-European 99, 158, 161, 163 Indus-Baluchistan 4, 97, 98, 102 Indus Civilization 99 Inner Asia 9 Iranian 124, 127, 148, 149 Iron/ Iron Age 3-5, 13-15, 55, 64-69, 74, 77, 78, 84, 97, 101, 102, 162 Irrigation 5, 86, 106, 117, 131, 133, 148, 149, 152, 153, 156-159, 165 Islam 114, 120, 127, 128, 131-137, 149 Islamic 5, 7, 114, 116, 126, 128-132, 134-138, 143, 145, 156, 157, 161 Isotopes 6, 121, 171-173, 176, 177 Isotopic studies 11 Jade 26, 37, 130 Jade axe 37 Jade bead 37 Jamal Qarshi 131 Jibin 127 Jilintai 10 Jimusar 10, 183

Jirentai Goukou 12 Jukti Rashid 132, 134 Kaptar Khana (Pigeon House) 115, 116 Karadong 160, 161 Karakorum 102, 106, 161 Karashar 92 Karasuk 13, 14 Karez 117, 119, 153, 156, 157 Karla 160 Kashgar 4, 5, 7, 15, 113-138, 143-166 Kashgar Oasis 5, 15, 113-116, 122, 125-127, 131-133, 143, 144, 148, 151, 153, 156-159, 161-165 Kashmir 114, 127, 128, 155 Kazakhstan 3, 6, 7, 10, 12, 53, 62, 64, 99-102, 156, 162, 181, 195, 197 Keriya River 2, 11, 19, 36, 46, 48 Khalchayan 104 Khan-oi 5, 114, 116, 117, 145-151, 165 Kharoshti 124 Khoja dynasty 135 Khoja Emin 129, 130 Khotan 104, 106, 113, 123-126, 128, 129, 132-134, 144, 150, 158, 161 Kirghizstan 15 Kizil Debe 116, 145, 159 Knife 14, 15, 37, 93 Koktepa 93 Kongque River 19, 24-26, 28, 29, 37, 48 Kongque River Delta Cemetery 24-26 Kopet Dagh 84 Kucha 113, 123, 125-128, 161 Kuluketage Mountain 28 Kumārajīva 127, 128 Kunlun Mountains 163, 164 Kunlunshan 13 Kurgan 12, 15, 100-104, 106 Kurghan Tim 145, 159 Kushan 103-105, 121, 123-125, 127, 133, 153, 158 Kuxi 13 Kyrgyz 120, 161 Ladakh 99, 102, 128, 130 Landlordism, monastic 153, 160 Lapis 98, 100 LE Fortress 19, 21, 23 Lead 171-173 Leather boots 21, 43 Leather shoes 21, 24, 26 LF Cemetery 19 Li Guangli 122 Liao 131 Linya Culture 49 Liushui cemetery 13 Looting 2, 13, 23, 28, 46 Lop Nur 11, 12, 19, 26, 29, 42, 48, 49 Loulan 26, 28, 124, 158, 161 Loulan Fortress 26, 28 LQ Cemetery 22 LS Cemetery 26 202

Index LT Cemetery 28 Lucerne 123 Machang Culture 11, 49 Maes Titianus 120, 121 Mahāyāna 124, 127, 147, 155 Mahmud al-Kashgari 114, 134, 148, 161, 162 Maize 173 Male symbol 40, 46 Manichaeism 125, 126 Manuscript 128, 145, 155 Marco Polo 121, 135 Marinos of Tyre 120 Mass spectrometer 171 Maternal 50 Maydatepa 4, 84-94 Mazar 131-134, 136, 149, 161 Medicinal Plants 193, 197 Mediterranean Sea 4, 5, 103, 105, 164 Meltwaters 138, 144, 153, 159, 164 Mesopotamia 5, 100, 172, 173 Metallurgy 13, 15, 37, 49, 92, 97, 171, 172 Micro-climate 164 Microlith 37, 64-67, 161, 177 Microscopy 182, 184 Migration 1, 4, 5, 7, 49, 53, 93, 99, 100, 158, 160, 161, 163 Millet 6, 7, 11, 15, 93, 99, 100, 114, 157, 162, 173, 174, 176, 177, 181, 184, 186, 191, 195-197 Mines / mining 5, 64, 91, 130, 131, 137 Ming dynasty 129 Miran 104 Mirshade 87, 92 Mirza Haidar 150 Mitochondrial 50 Monastery 118, 124, 147, 153, 156, 162 Monastic 116, 124, 125, 128, 143, 147, 148, 152, 153, 155, 160 Mongol 104-106, 121, 127, 132, 135 Mongolia 5-7, 13, 53, 55, 57, 103-106, 114, 126, 172-174, 197 Monsoon 153, 164 Mori Tim 5, 117, 119, 145, 147-149, 151, 153, 156, 157 Mortar 15, 91-93, 185, 189 Mosque 119, 143 Mourning 34 Mud brick 84, 86 Murgab 84 Mughal 5, 121, 130 Mummies/mummy 1, 2, 6, 21, 26, 160, 161, 163 Mundigak 84, 98 Mushiston 91 Nashqibandi 129 Nausharo 100 Neolithic 15, 49, 144, 151, 161, 181 Nestorian 131 Nitrogen 150, 171, 176 Niya 122, 124, 144, 158, 165 Noin-Ula 103-106

Nomad 2, 4, 5, 13, 15, 53, 97, 100, 103, 106, 113, 120, 122, 128, 131, 161, 162, 181 Nomadism 4, 97 North west China 171, 172, 177 Northern Cemetery 2, 11, 19, 36-39, 46, 48, 160, 162 Northern Wei dynasty 125 Nuclear techniques 5, 171, 177 Nukti Rashid 132 Oasis 1, 5, 6, 11-15, 50, 78, 87, 93, 94, 98, 100, 106, 113115, 122, 123, 125-133, 137, 143, 144, 148, 151-153, 155-166, 177, 195 Okunevo 11, 97-99 Optical Stimulated Luminescence – OSL 149, 156 Ordam Mazar 131-133, 136, 149 Ordam Padshah 132-134 Ornament 21, 43, 68, 71, 75, 87, 91, 101, 104, 105 Oxus Civilization 84, 99 Oxen 2 Painted pottery 4, 14, 24, 50, 71, 74, 78, 84, 85, 87, 92-94, 98 Pakistan 7, 99, 115, 158, 162, 165 Palaeolithic 9 Pamir Mountains 5, 143, 159, 163 Parkhai-Depe 92 Parthian 102, 103, 105, 123 Parthian Empire 123 Pastoralism 6, 7, 12, 14, 16, 122, 161, 197 Pastoralist 11-13, 15, 97, 122, 162, 181, 195 Pazyryk 102 Pelliot Paul 144-148, 150, 154 Persia 5, 100, 101, 120, 125-127 Persian sources 5 Pestle 15, 91-93, 182 Petroglyph 4, 55, 100, 102 Petrovsky Nikolai 154, 155 Phallic symbol 39, 44 Physical anthropology 37, 49 Pit burials 26, 59, 75 Plate 22, 35, 104, 105 Pocket 14, 15, 24 Pollen analysis 157, 164 Population 4, 5, 10-12, 15, 49, 50, 78, 97, 103, 106, 123128, 133, 144, 149, 152, 153, 155, 157, 161, 162, 172, 175 Populus euphratica 11, 21, 28 Pottery 4, 11, 13-15, 49, 50, 57, 61, 62, 66-68, 71, 74-76, 78, 84, 85, 87-90, 92-94, 98, 99, 102, 145, 150, 155, 161, 173, 175, 177 Pottery black-brown 93 Pottery grey painted 13, 71, 92 Pottery hand-made 66, 68, 74, 84, 85, 87, 93, 94 Pottery wheel-made 87, 90 Prajñāpāramitā sutra 155 Prehistory/prehistoric 3, 6, 9, 10, 15, 49, 52, 53, 64, 65, 69, 75, 77, 80, 91, 176, 181, 195, 197 Ptolemy 120, 121 Qadir Abduyraehim 116, 145 Qanat 153, 157 203

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads Qara-Khanid dynasty 131 Qara-Kitai 131 Qiang 161 Qiemu’erqieke 2, 10-12, 15, 55, 182, 196 Qilian ranges 172 Qing dynasty 129, 137, 161 Qinghai 14, 197 Qiongkeke 3, 65-69 Qocho 146 Quhui 92 Qutayba 113, 114, 128 Radiocarbon dates/dating 6, 38, 47, 48, 56, 84, 93, 171174, 176, 177 Rawak 144, 158 Reeds 2, 26 Religious wars 114, 133, 148, 149, 156 Remote imaging / sensing 147, 149, 150, 157, 165 Republic 53, 92, 98 Reservoir74, 153, 157 Residues 176, 185, 196, 197 Rice 6, 123, 152, 157, 173, 174, 176 Rock art 4, 55, 64, 77, 80, 81, 98, 99, 102 Ruoqiang County 21 Russia 4, 6, 7, 52, 74, 99, 101, 104, 114, 120, 131, 136, 138, 154, 155, 165 Sa’ensayi 10, 182 Saka 74, 103, 105, 106, 120, 161 Salination 117, 138, 156, 157, 176 sanghārāma 148 Sampula 104, 106 Sanchi 5, 152, 153, 155, 158, 160 Sangha 153, 155, 160 Sarazm 4, 98-100 Sasanian 127, 158 Satuq Boghra Khan 114, 128, 131-136 Semirech’ye 10, 12, 53, 62, 74, 102, 197 Serike 120 Settlement pattern 3, 113, 123, 143, 144, 152, 153, 156, 161, 162, 165 Sexual symbol 31, 37 Shaft-hole axe 12 Shahpur I 127 Shamanism 3, 99, 137 Shang Dynasty 15, 100 Shan-shan 124 Sheepskin 26, 28, 38 Shortughaï 99, 100 Shotorak 153 Shule 5, 113, 114, 123, 150, 151, 156 Shurabashat culture 93 Siba Culture 11, 49, 177 Siberia 1, 4, 9, 11, 12, 49, 50, 99, 101, 105, 163 Sickle 12, 15, 87, 93 Silk Road 1, 4, 6, 7, 100, 102, 103, 106, 120, 125, 127, 131, 135, 143, 160-162, 171, 175 Sima Qian 122 Smelting 173 Socketed celt 14

Sogdian 104, 120, 122, 124-130, 145, 161, 165 Sogdiana 104, 120, 161 Spindle whorl 15 Starch/analysis 181 Stein Aurel 19, 113, 115, 144, 146 Steppes 4, 52, 53, 93, 99, 100, 103-105, 122, 162 Stone arrowhead 10, 11, 37 Stone axe 40 Stone bowl 11 Stone cist 10, 59 Stone club 35 Stone figure 26, 62 Stone mace head 41 Stone mould 14, 93 Stone tools 10, 57, 92, 99, 181-185, 188, 190, 194, 196, 197 Stone Tower 120, 121 String 43, 44, 46 Stucco 147, 148 Stupa 5, 114, 116-118, 124, 128, 143-148, 152, 153, 155160, 164 Subeixi 14, 15 Sunburst burial 47, 48 Sung dynasty 127 Surkhandar’ya 93 Sūryasoma 127 Sweden 19 Syr-darya/ river 92 Tacheng 12, 53, 67 Takht-i Sangin 105 Taklamakan Desert 2, 6, 9, 11, 19, 36, 48, 49, 102, 114, 143, 163, 165 Talimu-Kongque River Delta 19 Tang dynasty 125, 126, 148, 150 Tank 157, 158 Tarbagatay range 10 Tarim Basin 1, 11, 13-15, 19, 37, 49, 50, 92, 93, 113, 120125, 128, 131, 135, 148, 158, 160-165, 195 Tarim River 2, 11, 12, 48, 50 Tashkent 92-94, 136 Tashkurgan 13, 120, 161 Tattoo 24, 160, 161 Tax collection 129 Taxila 104, 158 Technological diffusion 13 Tegurman 145, 146, 155 Tendon 44 Textile 4, 38, 45, 102, 104, 106, 130 Theravāda 124, 127, 147 Third National Archaeological Survey 28 Third National Relics Survey 21, 23, 53 Tianshan Mountains 3, 4, 6, 62, 78, 92, 113, 159, 163, 182, 183 Tianshanbeilu 2, 11-15, 19. 49, 50, 196 Tiao Hu 173 Tibetan 9, 10, 12, 13, 122, 126, 128, 161 Tibetan plateau 9, 10, 12, 13 Tillya Tepe/Tillyatepe 84, 103 Tocharoi 120 204

Index Tongtiandong 10 Topa Tim 5, 116, 118, 144, 145, 147-149, 157, 159 Torugart Pass 114 Tuntian 158 Trace elements 171, 173 Trade/traders 4-7, 87, 98, 100, 113-115, 120-122, 124131, 135-138, 143, 144, 152, 153, 157, 159, 160, 162, 195 Transhumant 12, 13, 15 Tree branches 26 Tree trunk 24, 26 Tritium 177 Tumen River 119, 157 Tumuli 13, 75 Turfan 102, 113, 129, 153, 156, 158 Turkic 19, 114, 125, 128, 131, 132, 134, 148, 155 Ukok 101 Upal 5, 114, 115, 144, 161, 162, 164, 176 Uranium 173 Urumqi 10, 113, 117, 128, 137, 144, 182 Utch Meravan 153 Uyghur 1, 92, 126, 128, 130, 135, 145 Uzbekistan 4, 15, 84, 99, 104 Vihāra 147 Wagon 15 Wakhan 161 Wakhi 161 Water / wells 2, 5, 11, 14, 78, 87, 99, 127, 131, 133, 137, 138, 143, 152, 153, 156-159, 161-165, 176, 183, 185 Wencheng Di 125 Westerlies 164 Wheat 5-7, 11, 15, 21, 26, 100, 114, 123, 128, 152, 157, 161, 162, 173-177, 181, 184, 185, 188, 191, 195-197 Wild and domesticated animal diets 177 Wind erosion 31, 54 Wooden anthropomorphic figure 34 Wooden bowl 26 Wooden bucket 42, 49 Wooden comb 24, 26 Wooden figure 22, 45, 46 Wooden pin 21 Wool netting 23 Woollen blanket 26, 37 Woollen cloak 21, 34 Woollen clothes 24, 28, 48

Woollen hat 21, 24, 26, 34 World War II 19 Wudi 122 Xiabandi 13, 67, 161 Xiahoe Culture 2-4, 6, 7, 11, 12, 19, 20, 37, 48-50, 196 Xiangbaobao 13, 161 Xianshuiquan Cemetery 28-30 Xiaohe 1-4, 6, 7, 11, 12, 15, 19, 20, 26, 28-31, 35, 36-50, 160, 162, 165, 181, 195, 196 Xindian Culture 14 Xinjiang 1-7, 9-15, 19-24, 26-50, 52, 53, 62, 64-66, 75, 77, 78, 92-94, 99, 101-104, 106, 112-114, 120, 121, 128131, 133, 136, 137, 143-145, 148, 150, 151, 153-156, 160-163, 165, 171, 173, 175, 181-183, 195, 197 Xinjiang Institute of Relics and Archaeology 19 Xintala 4, 14, 92, 93, 176 Xiongnu 103-106, 122, 126, 162 Xishanping 6, 173 Xuan Zang 128, 143, 147, 157, 158, 161 Yakhotan 161 Yakhurgan 161 Yakub Beg 132, 133, 135-137, 143 Yanbulake 14 Yanqi Basin 14, 50 Yardang 19, 21, 23 Yarkand 125, 127, 129, 137, 155 Yawaluk 5, 145, 146, 155 Yaz I 84-93 Yaz II 87, 92 Yazdepe 84 Yellow River Valley 10, 74, 181, 197 Yengisar 131, 133 Yenissei River 99 Yili River Valley 3, 4, 64, 65, 67, 69 Yili 3, 4, 10, 12-15, 53, 64, 65, 67-69, 71, 74, 76-78, 80, 81, 175 Yuezhi 103-106, 120, 122, 123, 125 Yuqi Moriwan 153 Yusuf Khass Hajib 134 Yutian 125 Zanskar 99, 102 Zaravshan 4, 93, 98, 99 Zhang Qian 122 Zoroastrianism 125, 127

205

The Cultures of Ancient Xinjiang, Western China: Crossroads of the Silk Roads unveils the ancient secrets of Xinjiang, western China, one of the least known but culturally rich and complex regions located at the heart of Asia. Historically, Xinjiang has been the geographic hub of the Silk Roads, serving international links between cultures to the west, east, north and south. Trade, artefacts, foods, technologies, ideas, beliefs, animals and people have traversed the glacier covered mountain and desert boundaries. Perhaps best known for the Taklamakan desert, whose name translates in the Uyghur language as ‘You can go in, you will never come out’, here the region is portrayed as the centre of an ancient Bronze Age culture, revealed in the form of the famous Tarim Mummies and their grave goods. Three authoritative chapters by Chinese archaeologists appear here for the first time in English, giving international audiences direct access to the latest research ranging from the central-eastern Xiaohe region to the western valleys of the Bortala and Yili Rivers. Other contributions by European, Australian and Chinese archaeologists address the many complexities of the cultural exchanges that ranged from Mongolia, through to Kashgar, South Asia, Central Asia and finally Europe in pre-modern times.

Alison Betts, Professor of Silk Road Studies, University of Sydney, has worked on the archaeology of Central Asia for more than two decades and more recently on Xinjiang. Marika Vicziany, Professor Emerita in Arts, Monash University, has specialised during the last four decades in Indian and Chinese culture and socioeconomic change. Peter Weiming Jia, Research Fellow, University of Sydney, has for more than a decade studied the Bronze Age sites of Xinjiang. Angelo Andrea Di Castro, Research Adjunct in Arts, Monash University, has been working on archaeological sites in Italy, Nepal, Australia and China for some three decades

Archaeopress www.archaeopress.com