South Asian Archaeology 2007: Volume I – Prehistoric Periods: Proceedings of the 19th International Conference of the European Association of South Asian Archaeology Ravenna, Italy, 2-6 July 2007 9781407310626, 9781407340340

Table of contents :
Front Cover
Title Page
Copyright
TABLE OF CONTENTS
Dedication
FRANK RAYMOND ALLCHIN (1923 – 2010)
PUBLICATIONS OF FRANK RAYMOND ALLCHIN
Dedication II
GREGORY LOUIS POSSEHL (1941 – 2011)
PUBLICATIONS OF GREGORY L. POSSEHL
Dedication III
SWARAJA PRAKASH GUPTA (1931 – 2007)
PUBLICATIONS OF SWARAJA PRAKASH GUPTA
LET NOT THE 19TH CENTURY PARADIGMS CONTINUE TO HAUNT US !
REFLECTIONS ON “DOTS ON A MAP”
T-SHAPED PILLARS AND MESOLITHIC “CHIEFDOMS” IN THE PREHISTORY OF SOUTHERN EURASIA: A PRELIMINARY NOTE
PROTOHISTORIC TERRACOTTA FIGURINES: CONTINUITY AND CHANGE IN FORM AND STYLE (CIRCA 2000-300 BC)
ASSESSING ANTHROPOGENIC SOIL EROSION WITH MULTI-SPECTRAL SATELLITE IMAGERY: AN ARCHAEOLOGICAL CASE STUDY OF LONG-TERM LAND USE
REJECTING THE STATE: REFLECTIONS ON ANCIENT INDIAN THOUGHT AND THE ORGANIZATION OF HARAPPAN SOCIETY
NOMADIC SETTLEMENT AND ECONOMY OF WESTERN HIMALAYA (UPPER INDUS)
EXPLORING INNER ASIA’S HIGH ALPINE FRONTIER: HIGH ALPINE TRANSHUMANT PASTORALISM, VERTICAL CULTIVATION AND ENVIRONMENTAL ARCHAEOLOGY IN THE LOWER VAKHSH-PANJ CONFLUENCE AND GORNO-BADAKHSHAN REGIONS, SOUTHERN TAJIKISTAN
A VIEW TO THE NORTH: BIOLOGICAL INTERACTIONS ACROSS THE INTER-MONTANEBORDERLANDS DURING THE LAST TWO MILLENNIA BC
ARE THE KHO AN INDIGENOUS POPULATION OF THE HINDU KUSH ? A DENTAL MORPHOMETRIC APPROACH
THE UNESCO INTERNATIONAL SAFEGUARDING CAMPAIGNS FOR MOHENJO-DARO, PAKISTAN. A RETROSPECTIVE
A NEW LATE BRONZE AGE SITE IN SOUTHERN UZBEKISTAN
EXCAVATION AT KANMER, A HARAPPAN SITE IN GUJARAT, INDIA: SOME OBSERVATIONS
TRANSPORT AT ALTYN-DEPE IN THE 3RD MILLENNIUM BC (THE EVIDENCE OF VEHICLE MODELS)
INAA OF AGATE SOURCES AND ARTIFACTS FROM THE INDUS, HELMAND, AND THAILAND REGIONS
IS IT POSSIBLE TO SPEAK ABOUT MIDDLE BRONZE AGE IN SOUTHERN TAJIKISTAN ?
THE LITHIC INDUSTRY FROM HAZAR SUM (SAMANGAN, AFGHANISTAN): NEW PERSPECTIVES AND OLD DATA
HARAPPAN ARCHAEOLOGY IN THE GHAGGAR BASIN WITH SPECIAL REFERENCE TO THE SITES OF GIRAWAD AND FARMANA
AUTOCHTHONOUS LATE BRONZE AGE CULTURES IN CENTRAL ASIA: WHERE ARE THEY?
EXPLORING CERAMIC VARIABILITY IN IRON AGE SOUTH INDIA: SOCIAL AND POLITICAL IMPLICATIONS
ILGYNLY-DEPE: NEW DISCOVERIES
EARLY FARMING CULTURES AT LAHURADEWA: RECENT EVIDENCE IN THE MIDDLE GANGA PLAIN
THREE SEASONS AT KASR AL-KHAFAJI (TOWER 1146) AT BAT, OMAN
THE SO-CALLED STUPA AT MOHENJO DARO AND ITS RELATIONSHIP WITH THE ANCIENT CITADEL
WILL THE MONUMENTS OF THE MIDDLE BRONZE AGE BE DISCOVERED ON THE TERRITORY OF SOUTHERN TADJIKISTAN ?
INTEGRATION BY PRICE IN THE BRONZE AGE
DEFINING THE INDUS SIGN LIST
SATELLITE REMOTE SENSING IMAGERY: NEW EVIDENCE FOR SITE DISTRIBUTIONS AND ECOLOGIES IN THE UPPER INDUS

Citation preview

BAR S2454 2013

VolumePeriods I Prehistoric

SOUTH ASIAN ARCHAEOLOGY 2007 VOLUME I

FRENEZ & TOSI (Eds)

South Asian Archaeology Proceedings of the 19th International Conference of2007 the

B A R Frenez and Tosi 2454 cover.indd 1

South Asian Archaeology 2007 European Association of South Asian Archaeology Ravenna, Italy, 2–6 July 2007 of the Proceedings of the 19th International Conference European Association of South Asian Archaeology Ravenna, Italy, 2-6 July 2007

Volume I

Prehistoric Periods Edited by

Edited by Dennys Frenez Dennys Frenez Maurizio Tosi Maurizio Tosi

BAR International Series2454 BAR International Series 2012 2013

17/12/2012 10:46:29

South Asian Archaeology 2007 Proceedings of the 19th International Conference of the European Association of South Asian Archaeology Ravenna, Italy, 2–6 July 2007

Volume I

Prehistoric Periods Edited by

Dennys Frenez Maurizio Tosi

BAR International Series 2454 2013

ISBN 9781407310626 paperback ISBN 9781407340340 e-format DOI https://doi.org/10.30861/9781407310626 A catalogue record for this book is available from the British Library

BAR

PUBLISHING

TABLE OF CONTENTS

OBITUARIES ALLCHIN, FRANK RAYMOND (1923-2010), by † Gregory L. Possehl

5

POSSEHL, GREGORY LOUIS (1941-2011), by Vasant Shinde & Asko Parpola

13

GUPTA, SWARAJA PRAKASH (1931-2007), by K. N. Dikshit

25

B. B. LAL Let not the 19th Century Paradigms Continue to haunt us!

31

CARL C. LAMBERG-KARLOVSKY Reflections on “Dots on a Map”

47

MASSIMO VIDALE T-Shaped Pillars and Mesolithic “Chiefdoms” in the Prehistory of Southern Eurasia: A Preliminary Note

51

ARUNDHATI BANERJI Protohistoric Terracotta Figurines: Continuity and Change in Form and Style (circa 2000-3000 BC)

59

ANDREW BAUER M. & KATHLEEN D. MORRISON Assessing Anthropogenic Soil Erosion with Multi-Spectral Satellite Imagery: an Archaeological Case Study of Long-Term Land Use

67

PIOTR A. ELTSOV Rejecting the State: Reflections on Ancient Indian Thought and the Organization of Harappan Society

77

R. S. FONIA Nomadic Settlement and Economy of Western Himalaya (Upper Indus) IRENE GOOD Exploring Inner Asia’s High Alpine Frontier: High Alpine Transhumant Pastoralism, Vertical Cultivation and Environmental Archaeology in the Lower Vakhsh-Panj Confluence and GornoBadakhshan Regions, Southern Tajikistan BRIAN E. HEMPHILL A view to the North: Biological Interactions across the Inter-Montane Borderlands during the Last Two Millennia BC BRIAN E. HEMPHILL, IHSAN ALI, SARAH BLAYLOCK AND NIKKI WILLITS Are the kho an Indigenous Population of the Hindu Kush? A Dental Morphometric Approach MICHAEL JANSEN The UNESCO International Safeguarding Campaigns for Mohenjo-Daro, Pakistan. A Retrospective KAI KANIUTH A New Late Bronze Age Site in Southern Uzbekistan

89 107

117

127 139

151

J. S. KHARAKWAL, Y. S. RAWAT AND TOSHIKI OSADA Excavation at Kammer, a Harappan Site in Gujarat: Some Observations

157

LYUBOV’ B. KIRCHO Transport at Altyn-Depe in the 3rd Millennium BC (the Evidence of Vehicle Models)

169

RANDALL LAW, ALISON CARTER , KULDEEP BHAN, ARUN MALIK AND MICHAEL GLASCOCK INAA of Agate Sources and Artifacts from the Indus, Helmand, and Thailand Regions

177

GIOVANNA LOMBARDO Is it possible to speak about Middle Bronze Age in Southern Tajikistan ?

185

ROBERTO MICHELI The Lithic Industry from Hazar Sum (Samangan, Afghanistan): New Perspectives and Old Data

193

VASAN SHINDE, TOSHIKI OSADA, AKINORI UESUGI AND MANMOHAN KUMAR Harappan Archaeology in the Ghaggar Basin with Special Reference to the Sites of Girawad and Farmana

201

T. SH. SHIRINOV & S. R. BARATOV Autochthonous Late Bronze Aged cultures in Central Asia: Where are they ?

217

CARLA M. SINOPOLI Exploring Ceramic Variability in Iron Age South India: Social and Political Implications

233

NATALIA F. SOLOVYOVA Ilgynly-Depe: New Discoveries

243

RAKESH TEWARI Early Farming Cultures at Lahuradewa: Recent Evidence in the Middle Ganga Plain

245

C. P. THORNTON, C. M. CABLE, G. L. POSSEHL Three seasons at Kasr al-Khafaji (Tower 1146) at Bat, Oman

255

GIOVANNI VERARDI & FEDERICA BARBA The So-Called Stupa at Mohenjo Daro and its Relationship with the Ancient Citadel

269

N. M. VINOGRADOVA Will the Monuments of the Middle Bronze Age be Discovered on the Territory of Southern Tadjikistan ?

281

DAVID A. WARBURTON Integration by price in the Bronze Age

287

BRYAN WELLS Defining the Indus Sign List

297

RITA WRIGHT & CARRIE HRITZ Satellite Remote Sensing Imagery: New Evidence for Site Distributions and Ecologies in the Upper Indus

315

This Volume is dedicted to the memory of Professor Frank Raymond Allchin (1923-2010), founder member of the Association of South Asian Archaeologists in Western Europe.

FRANK RAYMOND ALLCHIN (Harrow July 9, 1923 – Cambridge June 4, 2010)

5

FRANK RAYMOND ALLCHIN (1923 – 2010) † Gregory L. Possehl

Allchin’s earliest excavations were at Utnur and Piklihal, two South Indian Neolithic ‘ashmounds’. These he published as monographs. His study of the South Indian Neolithic culminated in one of the most insightful books in South Asian archaeology, titled Neolithic Cattle Keepers of South India: A Study of the Deccan Ashmounds (Cambridge 1963) which brought together the archaeological record, historical data and the ethnographic record in an interpretation of the past. Following his work on the Deccan ashmounds Allchin excavated a number of sites in the Subcontinent including Malwan in Gujarat (India), and Tarakai Qila, Lewan and an Early Harappan mound at Taxila, all in Pakistan. In addition, he conducted archaeological research in Afghanistan, Sri Lanka, Bangladesh and Nepal.

Frank Raymond Allchin (born July 9, 1923) died after a sudden illness on June 4, 2010 in Cambridge, England. Dr. Allchin made an outstanding contribution to South Asian archaeology, in terms of both the substantive part of the field, as well as to its important supporting institutions. Raymond married Bridget Gordon in 1951. She is a prehistorian and was a constant partner in his intellectual and institutional life. We who knew them well thought of them as an inseparable team. He is survived by Bridget and their two children, Sushila and William, as well as several grandchildren. Allchin was born in Harrow and was educated at Westminster School, along with our colleague Ralph Pinder Wilson. In 1944 he was a Lieutenant.in the Royal Corps of Signals serving in India. His principal posting was at Mhow (now Dr Ambedkar Nagar) in Madhya Pradesh, just south of Indore. Mhow was the home of the British Indian Army’s School of Signals. He also served in Agra and Singapore.

Allchin was an outstanding teacher, especially for research students, both archaeologists and art historians. Most of them have found careers in the South Asian field as professors and curators, as well as Directors General of the Archaeological Survey of India and its counterpart in Sri Lanka. They have made their own significant contributions to the scholarship on ancient South Asia. When he arrived in Cambridge he was the only archaeologist teaching full time in the field of South Asia: his effectiveness in this regard is attested by the fact that South Asian archaeology is now taught at Cambridge and Oxford as well as University College London, Durham and Leicester.

Raymond was an avid student of Hindi, and worked hard on his language skills when in India and later in London. He had opportunities to travel in the Subcontinent during his spells of leave in the Army and was much impressed with India’s ancient monuments, places like Sanchi and the Taj Mahal. One of his earliest and most enduring insights into India was his sense that there are strong continuities in Indian life, stretching from prehistoric times to the present, and that the historical and ethnographic record could useful tools for archaeological interpretation.

Raymond co-authored many publications with Bridget, among the most important of which was the trend setting The Birth of Indian Civilization (Penguin 1968). ‘The Birth...’ was a synthetic overview of prehistoric South Asia, the first since S. Piggott’s Prehistoric India published in 1950. The Allchins went on to publish updated syntheses as with The Rise of Civilization in India and Pakistan (Cambridge 1982) and Origins of a Civilization: The Prehistory and Early Archaeology of South Asia (Viking 1997). These are all extraordinary contributions to the field. Notice should also be taken of two seminal works that Raymond edited: The Archaeology of Afghanistan: From Earliest Times to the Timurid Period (Academic Press, 1978, with Norman Hammond) and The Archaeology of Early Historic South Asia: The Emergence of Cities and States (Cambridge, 1995).

By 1950 Allchin had decided to become a student of ancient India and enrolled in London’s School of Oriental and African Studies (SOAS). He graduated with a first in Hindi and Sanskrit. This laid the foundation for his later translations of devotional texts from early Hindi into English - the Kaviavali and The Petition of Ram, both by Tulsi Das (1532-1623) the great Indian philosopher. He received a PhD from SOAS in 1954 working under the guidance of K. de B. Codrington. The title of Raymond’s PhD dissertation was Development of Early Cultures in the Raichur District of Hyderabad. SOAS kept him on as a Lecturer in Indian Archaeology until 1959 when he moved to Cambridge University’s Department of Oriental Studies where he remained for the rest of his career. Raymond was appointed a Fellow of Churchill College in 1963 and Reader in Indian Studies in 1972. In 1989 he retired with the title of Emeritus Reader in South Asian Archaeology. Allchin was made a Fellow of the British Academy in 1981.

Raymond and Bridget were also institutions builders. Their energy and wisdom were key to the establishment of the Association of South Asian Archaeologists in Western Europe, at a meeting held in Cambridge in 1970.

7

† Gregory L. Possehl This organization is now known as the European Association of South Asian Archaeologists and has emerged as one of the cornerstones of our discipline. They also played a key role in the founding of the British Association for South Asian Studies, and its splendid and most useful journal South Asian Studies. And finally, there is their Ancient India and Iran Trust in Cambridge, founded in 1978. The trust was a joint venture between the Allchins, Sir Harold Bailey, Professor Johanna van Lohuizen de Leeuw and her husband Dr. Jan van Lohuizen. The Trust has a very important and growing library, it sponsors intellectual events such as conferences and lectures and has a small, but important grants program. Frank Raymond Allchin was an outstanding scholar, with a deep knowledge of South Asia and ancient India. He made a seminal contribution to our discipline in terms of both his insights into the past, and the institutions he helped to found. He was also a fine human being, and will be much missed for this quality as well.

8

FRANK RAYMOND ALLCHIN (1923 – 2010)

PUBLICATIONS OF FRANK RAYMOND ALLCHIN Collected by Gregory L. Possehl †

Books and Monographs 1954 1960 1961 1963 1968 1978 1979 1982

1986 1989 1991 1995

1997

2003

Development of Early Cultures in the Raichur District of Hyderabad. PhD Dissertation, University of London, School of Oriental and African Studies. Piklihal Excavations, Andhra Pradesh Archaeological Series, 1, Hyderabad, Government of Andhra Pradesh. Utnur Excavations, Andhra Pradesh Archaeological Series 5, Hyderabad, Government of Andhra Pradesh. Neolithic Cattle Keepers of South India: A study of Deccan ashmounds, Cambridge, Cambridge University Press. The Birth of Indian Civilisation (with B. Allchin), London, Penguin. The Archaeology of Afghanistan (with N. Hammond, ed. and contributor), London, Academic Press. A Source-book of Indian Archaeology (with D. K. Chakrabarti), New Delhi, Munshiram Manoharlal, Vol. 1. Indian Monuments through British Eyes, 1780-1980: Handlist of the exhibition in the Fitzwilliam Museum, Cambridge, Fitzwilliam Museum. The Rise of Civilisation in India and Pakistan (with B. Allchin), Cambridge, Cambridge University Press. Lewan and the Bannu Basin: Excavation and survey of sites and environments in North West Pakistan (with B. Allchin, F. A. Durrani, and M. F. Khan, ed. and contributor) Oxford, British Archaeological Reports, International Series, No. 310. The Conservation of the Indian Heritage (with B. Allchin and B. K. Thapar, ed. and contributor), New Delhi, Cosmo Publications. Shahr-i Zohak and the History of the Bamiyan Valley, Afghanistan (with P. H. B. Baker, ed.), Oxford: British Archaeological Reports Number 570, Ancient India and Iran Trust Series No. 1. The archaeology of Early Historic South Asia: The emergence of cities and states, (ed.), Cambridge, Cambridge University Press. Excavations at Malvan; report of the collaboration of the Archaeological Survey of India and Cambridge University in 1970, on the Gujarat Plain (with J. P. Joshi) New Delhi, Memoirs of the Archaeological Survey of India 92. South Asian Archaeology 1995 (with B. Allchin, ed. and contributor) 2 Vols. Delhi: Oxford & IBH. South Asian archaeology, 1971-1991: index of the proceedings of the biennal conferences of the European Association of South Asian Archaeologists (formerly, Association of South Asian Archaeologists in Western Europe), (compiler). Naples: Indices, Istituto Universitario Orientale. Dipartimento di Studi Asiatici, 3. A Source-book of Indian Archaeology (with D. K. Chakrabarti), New Delhi, Munshiram Manoharlal, Vol. 2. A Source-book of Indian Archaeology (with D. K. Chakrabarti), New Delhi, Munshiram Manoharlal, Vol. 3.

Translations 1964 1966

Tulsi Das, Kavitavali, translated with introduction (UNESCO), London, Allen & Unwin. Tulsi Das, The petition to Ram, translation of Vinayapatrika with introduction, notes and glossary, London, Allen & Unwin.

Reports for UNESCO and UNDP 1969

Cultural Tourism in India: its scope and development with special reference to the monumental heritage, UNESCO. Report on the Lumbini Development Project (with K. Matsushita), UNDP.

9

† Gregory L. Possehl Papers (excluding reviews) 1953 1955 1956 1957

1959

1961

1962

1963

1964 1966

1968

1969

1970

1971 1972

‘A flake tool from the Oxus’. Proceedings of the Prehistoric Society, 19: 227. ‘Rock paintings and engravings in Raichur, Hyderabad’ (with D. H. Gordon). Man, 55: 97-99. ‘Stone alignments of Southern Hyderabad’. Man, 56: 133-136. ‘Sanskrit eduka – Pali eduka’. Bulletin of the School of Oriental and African Studies, 20: 1-4. ‘The culture sequence of Bactria’. Antiquity, 31: 131-141. ‘The Neolithic stone industry of North Karnataka region’. Bulletin of the School of Oriental and African Studies, 19: 321-335. ‘Poor men’s "Thālīs": a Deccan potter’s technique’. Bulletin of the School of Oriental and African Studies, 22: 250-257. ‘Upon the contextual significance of certain groups of ancient Indian signs’. Bulletin of the School of Oriental and African Studies, 22: 548-555. ‘Ideas of history in Indian archaeological writing’. In Philips, C. (ed.) Historians of India, Pakistan and Ceylon, Oxford, Oxford University Press: 241-259. ‘The antiquity and methods of gold mining in ancient India’. Journal of the Economic and Social History of the Orient, 5: 195-211. ‘A Neolithic pot from the Billa Surgam caves, Andhra Pradesh’. Antiquity, 36: 302-303. ‘Painted pottery from Patapad in Andhra Pradesh’. Antiquity, 36: 221-224. ‘The archaeology of a river crossing: an account of the Krishna Bridge site’ (with B. Allchin). In Madan, T. N. and G. Sarana (eds) Indian Anthropology, essays in memory of D.N. Majumdar, Bombay: Asia Publishing House: 52-65. ‘The Neolithic stone industry of the Santal Parganas’. Bulletin of the School of Oriental and African Studies, 25: 306-330. ‘Cattle and economy in the Neolithic of South India’. In Mourant, A. E. and F. E. Zeuner (eds) Man and Cattle, London: 149-155. ‘The cultural tradition of India’. The Listener, March: 25-39. ‘An inscribed weight from Mathura’. Journal of the Economic and Social History of the Orient, 7: 201-205. ‘Pottery and head-rests from Narasipur Sangam’. In Sen, D. and A. K. Ghosh (eds) Studies in Prehistory, Robert Bruce Foote Memorial Volume, Calcutta, Firma K. L. Mukhopadhyay: 58-63. ‘The place of Tulsi Das in North Indian devotional tradition’. The Journal of the Royal Asiatic Society of Great Britain and Ireland, No. 3/4 (Oct., 1966): 123-140. ‘Archaeology and the date of Kaniska’. In Basham, A. L. (ed.) Papers on the Date of Kaniska, Leiden, E. J. Brill: 4-34. ‘Explorations in districts Baroda, Broach and Surat’ (with J.P. Joshi). Indian Archaeology 1967-68 – a review, New Delhi, Archaeological Survey of India: 9-13. ‘The social thought of Swami Vivekananda’. Swami Vivekannanda in East and West, London, Ramakrishna Vedanta Centre: 84-105. ‘Dilmun and the gulf of Cambay’. Antiquity, 43: 315-316. ‘Early cultivated plants in India and Pakistan’. In Ucko, P. J. and G. W. Dimbleby (eds) The Domestication and Exploitation of Plants and Animals, London, Duckworth: 323-329. ‘Early domestic animals in India and Pakistan’. In Ucko, P. J. and G. W. Dimbleby (eds) The Domestication and Exploitation of Plants and Animals, London, Duckworth: 317-322. ‘A pottery group from Ayun, Chitral’ (in honour of Sir Harold Bailey). Bulletin of the School of Oriental and African Studies, 33: 1-4. ‘Malvan – further light on the southern extension of the Indus Civilization’ (with J.P. Joshi). Journal of the Royal Asiatic Society, 102 (1): 20-28. ‘The attaining of the void – a review of some recent contributions in English to Virasaiva studies’. Religious Studies, 7: 339-359. ‘A cruciform reliquary from Shaikhan Dheri, Charsada’. In Pal, P. (ed.) Aspects of Indian Art, Leiden, E. J. Brill: 15-26. ‘A piece of scale armour from Shaikhan Dheri, Charsada’. Journal of the Royal Asiatic Society, 102 (2): 113120. 10

FRANK RAYMOND ALLCHIN (1923 – 2010) 1972 1973

1974

1975 1976

1977

1978

1979

1980

1981

‘Malvan’ (with J.P. Joshi). In Deo, S. B. (ed.) Archaeological Congress and Seminar Papers, Nagpur, Nagpur University: 36-42. ‘An inscribed reliquary from Shaikhan Dheri’. In Hartel, H. H. and V. Moeller (eds) Indologen-Tagung, 1971, Wiesbaden, Steiner: 265-272. ‘Problems and perspectives in South Asian archaeology’. In Hammond, N. (ed.) South Asian Archaeology, 1971, Park Ridge, Noyes Press: 1-11. ‘Pottery from graves in the Perumal hills near Kodaikanal’. In Ghosh, A. K. (ed.) Perspectives in Palaeoanthropology (Dharani Sen volume), Calcutta, Firma K. L. Mukhopadhyay: 299-308. ‘Some new thoughts on Indian cattle’ (with B. Allchin). In van Lohuizen-de Leeuw, J. E. and J. J. M. Ubaghs (eds) South Asian Archaeology, 1973, Leiden, E. J. Brill: 71-77. ‘The emergence of civilization in North India’. Encyclopedia Britannica. London, Encyclopedia Britannica. ‘The relationship of early settled communities in peninsular India with those of Late Stone Age huntergatherers’ (with B. Allchin). In Sharma, R. S. and V. Jha (eds) Indian Society, Historical Probings (D.D. Kosambi volume), Delhi, People’s Publishing House: 45-66. ‘The reconciliation of jnana and bhakti in Ramcaritamanasa’. Religious Studies, 12: 81-91. ‘The mysterious Path of Love in Tulsi Das’. Vedanta for East and West, London, Ramakrishna Vedanta Centre, 152: 3-11. Preface for reprint of Annual Reports of the Archaeological Survey of Mysore: 1-3. ‘A modern Indian potter’s technique’. In Chattopadhyaya, D. (ed.) History and Society: Essays in honour of Professor Niharanjan Ray, Calcutta, Firma K. L. Mukhopadhyay: 1-13. ‘Recommendations concerning archaeological sites’. In Sekler, E. F. (ed.) Masterplan for the Conservation of the Cultural Heritage in the Kathmandu Valley, UNESCO, Paris: 147-154. ‘Recommendations concerning moveable cultural property’. In Sekler, E. F. (ed.) Masterplan for the Conservation of the Cultural Heritage in the Kathmandu Valley, UNESCO, Paris: 134-147. ‘Religious symbols and Indian thought’. In Davidson, H. R. E. (ed.) Symbols of Power, Cambridge, The Folklore Society Mistletoe Series: 1-35. ‘Conclusion’ (with N. Hammond). In Allchin, F. R. and N. Hammond (eds) The Archaeology of Afghanistan, New York, Academic Press: 405-414. ‘Introduction’ (with N. Hammond). In Allchin, F. R. and N. Hammond (eds) The Archaeology of Afghanistan, New York, Academic Press: 1-9. ‘Monument conservation and policy in India’. Journal of the Royal Society of Arts, 126: 746-765. ‘Surface collections from Pushkar’. In Allchin, B., A. S. Goudie and K. Hegde (eds) The Prehistory and Palaeogeography of the Great Indian Desert, New York, Academic Press: 331-1+9. ‘India – the ancient home of distillation?’. Man (New Series), 14 (1): 55-63. ‘Stamped Tangas and condensers: evidence of distillation at Shaikhan Dheri’. In Taddei, M. (ed.) South Asian Archaeology, 1977, Naples, Istituto Universitario Orientale: 755-797. ‘A note on the “Asokan” stupas of Patan’. In Dallapiccola, A. L. (ed.) The Stupa, its Religious, Historical and Architectural Significance, Wiesbaden, Frantz Steiner Verlag: 147-156. ‘Archaeological and language evidence for the movement of Indo-Aryan speaking peoples into South Asia’. J. K.R. Cama Oriental Institute, Bombay, 48: 68-102; also published in 1981, in Asimov, M. S., B. A. Litvinsky, L. I. Miroshnikov and D. S. Rayevsky (eds) Ethnic Problems of the history of Central Asia in the Early Period (2nd millennium B.C.), Moscow: 336-349. ‘Textile impressions from the South Indian Iron Age’. P.E.P. Deraniyagala Commemoration Volume, Sri Lanka: 64-67. ‘Antecedents of the Indus Civilization (10th Annual Mortimer Wheeler Archaeological Lecture)’. Bulletin of the British Academy, 66: 135-160. ‘Antiquity and continuity’. In P. Jayakar et al. (eds) The Indian Experience: 125-129. ‘Antiquity of gold mining in the Gadag region’. In Nagaraja Rao, M. S. (ed.) Madhu, Recent Researches in Indian Archaeology and History, Delhi, Agam Kala Prakashan: 81-83. ‘How old is the city of Taxila?’. Antiquity, 56: 8-14. ‘Preliminary report on the Bannu Basin Project (1977-79) – Introduction’. In Hartel, H. (ed.) South Asian Archaeology, 1979, Berlin, Dietrich Reimer Verlag: 217-218. ‘Preliminary report on the excavations at Lewan (1977-78)’ (with J. R. Knox). In Hartel, H. (ed.) South Asian Archaeology, 1979, Berlin, Dietrich Reimer Verlag: 241-244. 11

† Gregory L. Possehl 1981

1984 1985

1986

1989

1990

1992 1993 1998

‘Preliminary report on the excavations at Tarakai Qila (1978-79)’ (with J. R. Knox). In Hartel, H. (ed.) South Asian Archaeology, 1979, Berlin, Dietrich Reimer Verlag: 245-250. ‘The legacy of the Indus Civilisation’. In Possehl, G. L. (ed.) Harappan Civilization, A Contemporary Perspective¸ New Delhi, Oxford & IBH: 325-33. ‘The northern limits of the Harappan culture zone in the light of recent observations’. In Lal , B. B. and S. P. Gupta (eds) Frontiers of the Indus Civilization, New Delhi, Books and Books: 51-54. ‘Guide to the Asokan inscriptions’ (with K. R. Norman). In Allchin, B. (ed.) South Asian Studies, 1: 43-50. ‘The interpretation of a seal from Chanhu-daro and its significance for the religion of the Indus Civilization’. In Schotsmans, J. and M. Taddei (eds) South Asian Archaeology 1983, Naples, Instituto Universitario Orientale, Dipartimento di Studi Asiatici, Series Minor 23: 369-384. ‘Drumbeats from the past’. In Chandra, L. and J. Jain (eds) Dimensions of Indian Art – Pupul Jayakar Seventy, Delhi, Agam Kala Prakashan. ‘Radiocarbon dating of some early sites in North West Pakistan’ (with K. D. Thomas). South Asian Studies, 2: 37-44. ‘The aims of the Seminar’, introductory remarks. In Allchin, B., F. R. Allchin and B. K. Thapar (eds) The Conservation of the Indian heritage, New Delhi, Cosmo Publications. ‘Threats to the conservation of monuments in urban and rural settings’. In Allchin, B., F. R. Allchin and B. K. Thapar (eds) The Conservation of the Indian heritage, New Delhi, Cosmo Publications. ‘Indo-Aryan and Aryan: language, culture and ethnicity’. Ancient Ceylon, 10: 13-23. ‘Patterns of city formation in Early Historic South Asia’. South Asian Studies, 6: 163-174. ‘The end of the Harapp.an Urban Phase and its aftermath’. Ancient Ceylon, 10: 25-40. ‘An Indus ram: a hitherto unrecorded stone sculpture from the Indus Civilization’. South Asian Studies, 8: 53-54. ‘The urban position of Taxila and its place in Northwest India-Pakistan’. In Spodek, H. and D. M. Srinivasan, (eds) Urban Form and Meaning in South Asia. Washington DC: National Gallery of Art. Studies in the History of Art, 31. Center for Advanced Study in the Visual Arts Symposium Papers, 15: 69-81. ‘The interface of archaeology and history’. Man and Environment, 23 (1): 19-36.

12

This Volume is dedicated also to the memory of Professor Gregory L. Possehl. (1941-2011)

GREGORY LOUIS POSSEHL (Tacoma July 23, 1941 – Philadelphia October 8, 2011)

13

GREGORY LOUIS POSSEHL (1941 – 2011) Vasant Shinde & Asko Parpola

Saurashtra and discovered and studied over 450 sites of the Bronze Age culture. He not only studied the determinant factors governing the location of Protohistoric sites in Saurashtra, but also linear, nuclear and hexagonal patterns formed due to the varied distribution of resource patterns. This is the beginning of Settlement Pattern Studies in Indian Archaeology, which was followed by many including me subsequently.

Gregory Louis Possehl was born in Tacoma, Washington State, USA, in 1941. He read anthropology and received his B.A. from the University of Washington in 1964. While working on shell middens for the state highways department, he was introduced to South Asian archaeology after volunteering to work with Professor W. A. Fairservis in a survey of the Quetta Valley of Pakistan. Fairservis encouraged him to return to Washington to graduate school, and he completed his MA in 1967. He left Washington and, following army service, began a Ph.D. at the University of Chicago, focusing on A Study of Prehistoric Gujarat with Special Reference to the Posturban Harappan. Just before the award of his Ph.D. in 1974, Possehl briefly held a lectureship in Anthropology at Vassar College before joining the University of Pennsylvania in 1973 as a lecturer in the Department of South Asian Regional studies and Assistant Curator at the University Museum. He was promoted to associate professor in 1981 and was appointed an associate director of the Museum between 1981 and 1992. Possehl was promoted to a full professorship in the Department of Anthropology in 1989 and held the Chair of the Department between 1995 and 2001. He retired in 2004 but continued to be actively engaged as Professor Emeritus of South Asian Archaeology and Curator Emeritus of the Asia Section.

In his early years at the University of Pennsylvania, Possehl received a fellowship from the American Institute of Indian Studies to examine the Kulli collection made in the early 1900s by the famed explorer Sir Aurel Stein. Two years later he organized a conference on the Harappan culture in Kashmir that resulted in the 1982 publication of Harappan Civilization: A Contemporary Perspective. This was updated in 1993 as Harappan Civilization: A Recent Perspective. Possehl began planning for archaeological fieldwork in Gujarat in western India in 1980, the first by an American in India since 1935. He worked initially at Oriyo Timbo, a monsoon encampment of Late Harappan/Mesolithic pastoral nomads. He is one of the few archaeologists who recognized the role of small site archaeology in the settlement system and hypothesized that there was a symbiotic relationship between the Harappans and the Mesolithic hunter-gatherer on the basis of the data he recovered from such sites. His subsequent extensive excavation at Rojdi produced an exhaustive chronology and established characteristic features of the Harappan culture in Saurashtra. Before commencing the work at Rojdi it was believed, by many archaeologists, that the Harappan phase in Saurashtra was later in date compared to the one at the classical Harappan sites as Harappa and Mohendjo daro (what Possehl called as Sindhi Harappan) and that the Harappans moved into this region later in search of semi-precious stones such as carnelian and agate, the articles made of which had greater demand in Mesopotamia and Egypt. In the first phase of the work at Rojdi, Possehl obtained a series of C14 dates and proved that the Harappan phase in Saurashtra was contemporary to the Sindhi Harappan sites. Possehl, after observing the material culture of the Harappan phase in Rojdi realized difference, mainly in the ceramic assemblage and suggested that the communities of Saurashtra formed a distinctive grouping within the Harappan civilisation and distinguished them by terming them “Sorath”. His excavations also demonstrated that Rojdi, like many other sites, survived the eclipse of the great cities of Mohenjodaro and Harappa and continued to thrive. The Sorath Harappan ceramic differed from the Sindhi Harappan ceramics on account of surface treatment, painted motifs and shapes. Unlike the Sindhi Harappan ware, the Sorath Harappan ware is pale red in colour, has fewer painted

I had known Gregory Possehl (He insisted I should call him Greg) since 1977, when I joined the Department of Archaeology for Master’s programme in Ancient Indian History, Culture and Archaeology at the famed Deccan College, Post-Graduate and Research Institute, Pune, India, from his writings on the Harappan Culture. I had an occasion to meet him personally first in 1979 after I joined the same Institute for my Ph.D. degree course on the Early Settlements in the Central Tapi Basin. Greg was interested in my work because he had already carried out, in the adjoining region of Saurashtr, work on the Harappan settlement pattern and was keen to draw parallels between these two regions on the nature of the settlement pattern and systems. I had a glimpse of his academic achievement and personal warmth in the first meeting itself. Subsequently we became great friends and colleagues until his death. Greg’s association with Deccan College ushered in new era in Indian Archaeology in sixties of twentieth century. Although Greg started his South Asian field career in the Quetta Valley of Pakistan in the 1960s after graduating from the University of Washington, his main fieldwork projects were in western India with a clear focus on the Protohistoric phase. Greg was the first to introduce the concept of Settlement Patterns in Harappan Archaeology, a very important notion of New Archaeology. He carried out extensive and intensive survey of the entire 15

Vasant Shinde & Asko Parpola between Gilund and the Harappan Civilization, as well as with groups farther to the northwest in Afghanistan and Central Asia (Bactia-Margiana Archaeological Complex (BMAC). This work has surely contributed new chapter to the History of Rajasthan. The final Gilund report is being published by the University of Pennsylvania. This report will be dedicated to Gregory Possehl, who is one of the few foreigners, genuinely interested in South Asian history and cultural heritage. He never used his archaeological research work in South Asia for his personal gain. Having worked with so many internationally famed archaeologists in the field, I found Possehl, the most complete field archaeologist. His sound theoretical as well practical knowledge couple with great observations and analytical mind made him such a towering scholar.

motifs, that to mostly geometric confined to the upper half of the vessels and the most dominating shape is convex-sided bowl with our without stud handle. In early 1990s, Possehl excavated Babar Kot, a small fortified Late Harappan site near Rodji for one season. He invited me to participate in this excavation, which provided me proper opportunity to know him personally and discuss many issues related to the Harappan Civilization. The discussion was so stimulating that I developed strong interest in the Harappan archaeology and decided to divert from the Chalcolithic to the Harappan studies. Greg, after one season’s work at Babar Kot, decided to go back to Rojdi for four seasons, when he unearthed a complete plan of a Sorath Harappan village. One of the monographs he published during this period of fieldwork was Harappan Civilization and Rojdi (1989) in collaboration with the Gujarat State Department of Archaeology. In fact all his excavations in Gujarat were carried out in collaboration with the Gujarat State Department of Archaeology.

Having worked along coastal western India for many years, Possehl also took a keen interest in the identification of the lands of Dilmun, Megan and Meluhha, mentioned in Mesopotamian texts as key trading partners within the Arabian Sea. He became directly involved in this study when he joined an international team planning to rebuild one of the famous Black Ships of Magan in 2005. With his fellow archaeologists Professor Maurizio Tosi, of the University of Bologna, and Late Dr. Serge Cleuziou, of Centre National de la Recherche Scientifique, and the Australian naval architect Tom Vosmer, the team used archaeological finds, ancient texts and fragmentary graffiti to allow the reconstruction of a 40ft-long boat made of reed bundles waterproofed with bitumen. Aiming to retrace one of the pioneering voyages of the Bronze Age across the 600 miles separating Oman from Gujarat, the reed boat sank only seven miles off the coast on the first day. Thereafter, Possehl was invited by the Sultanate of Oman to begin fresh excavations at the Unesco World Heritage site of Bat in 2007. Dating to the Bronze Age, the project at Bat allowed Possehl and his team an opportunity to excavate a major complex of conical stone towers and associated cemeteries and habitation sites on the coast, trading communities who developed contemporaneously with those of western India.

By the mid-1990s, Possehl decided to end fieldwork in India and began dismantling his Rojdi storehouse of camping, cooking, and archaeological equipment. He donated all of his supplies to several physicians who were in great need of these materials to help mitigate the ravages of a cyclone that hit Gujarat’s Gulf of Kutch. Having divested himself of his entire archaeological infrastructure, Possehl, of course, soon found himself enticed back into the field by me. I knew his weakness which I exploited to get him agreed to continue work in India. I suggested that several sites on the western periphery of the Harappan region have been excavated to understand their role in the development of the Harappan culture. We need to undertake similar work on the contemporary cultures that flourished on the eastern periphery of the Harappan region to understand their contribution to the development of the Harappan elements, to which he agreed readily. We decided to undertake extensive work at the site of Gilund, which is the largest Chalcolithic site in the Mewar region of Rajasthan and best preserved. He collected funds from different sources such as NSF and private donors including his own brother for the Gilund work. The research work at Gilund that was carried out between 1999-2005, was the most rewarding as it provided an opportunity to understand the development of the Chalcolithic culture in the Mewad region of Rajasthan and their interaction with the Harappan culture. Our work at Gilund demonstrated that the Chalcolithic culture originated locally, but flourished due to their contact with the Harappans. They borrowed necessary technologies and some traditions from the Harappans, which enabled them to flourish between 2500-2000 BCE. The Chcalolithic people developed Reserved-Slip ware, which was later adopted by the Harappans. There were two exciting discoveries at Gilund-the parallel wall structure identified as a public granary and a bin in that building that contained 100 clay seal impressions dating between 2300 and 1800 BC. The designs used suggest links

In addition to publishing monographs on his fieldwork projects in western India, Possehl also edited and wrote several synthetic studies linking the archaeology of western India with the general chronologies for the Harappan Civilization. He also attempted to focus on decipherment of the Harappan script resulting in his 1996 publication of the Indus Age: The Writing System. Mindful of the needs of field archaeologists and researchers Possehl also provided a set of invaluable reference sources, which included a gazetteer of all the excavated sites reported in the volumes of the Archaeological Survey of India’s annual publication Indian Archaeology: A Review (IAR) as well as the publication of a comprehensive list of radiocarbon dates for the region.

16

GREGORY LOUIS POSSEHL (1941 – 2011) Possehl’s contributions to our knowledge of the Protohistoric archaeology of western India were internationally recognised. He was a corresponding member of the Istituto Italiano per l’Africa e l’Oriente; an overseas fellow of Churchill College, Cambridge; a trustee of the Ancient India and Iran Trust, Cambridge; and a visiting professor at Collège de France in 2003. He was also acclaimed in India, receiving a fellowship of the Indian Archaeological Society in 2004 and the President of India’s First Prize for his 1982 book Harappan Civilisation. In Greg’s death, the South Asia has lost its great scholarship and personal friend of many of us. The damage done is unrepeatable. Greg was a great human being. His stay at the Gilund House and in the camp during the course of excavations between 1999 to 2005 was memorable. Greg was suffering from acute chronic asthma but still lived three to four months in the dusty Indian village of Gilund for five seasons during the course of excavations. He never complained of inadequate basic facilities, but always adapted to the situation. He was a true archaeologist who cared only for his work. He made great friendship with the local labourers and farmers and always enjoyed their company in the evening, though they were not able to communicate with language. I must frankly accept that my students and I had great learning experience and we picked up many great qualities from Greg and they will surely help us in our future endeavour. Greg was closely associated with Professor H.D. Sankalia, the founder of the Department of Archaeology at Deccan College. During the course of fieldwork in India for his Ph.D. degree course, he was affiliated to Deccan College under the guidance of Prof. Sankalia and Prof. M.K. Dhavalikar. It was a fine gesture when Possehl responded positively to our request to participate in the Centenary Seminar we arranged in Deccan College in December 2007 in the memory of H.D. Sankalia. He came all the way only for this special seminar on his own expenses. It was his nostalgic visit, which turned out to be his last visit to Deccan College. All my colleagues in Deccan College and outside will always cherish great friendship with Possehl. We salute him for his dedication to the subject and immense contribution to South Asian Archaeology. He has trained many young archaeologists in the United States and some of them are continuing his work in Indian Subcontinent. I have very little idea about his family but I understood from Possehl’s students that he is survived by his companion, Susan West, and by his former wife, and a son and daughter.

17

Vasant Shinde & Asko Parpola

PUBLICATIONS OF GREGORY L. POSSEHL Collected by A. Parpola

Books and Monographs 1966 1974 1976 1979 1980 1982 1984 1985 1986 1989

1990 1991 1992 1993

1994 1996 1999 2002

Pastoral nomadism in prehistoric Iran. Unpublished M.A. thesis, University of Washington. Variation and change in the Indus Civilization: A study of prehistoric Gujarat with special reference to the post-urban Harappan. Unpublished Ph.D. dissertation, Department of Anthropology, The University of Chicago. Ecological backgrounds of South Asian Prehistory (with K. A. R. Kennedy, eds). Symposium convened at the seventy-second annual meeting of the American Anthropological Association, December 2, 1973, New Orleans. South Asia Occasional Papers and Theses, No. 4. Ithaca NY, Cornell University. Ancient cities of the Indus (ed.). Durham NC, Carolina Academic Press; New Delhi, Vikas Publishing House Pvt. Ltd. Indus Civilization In Saurashtra. Delhi, Published on behalf of Indian Archaeological Society by B.R. Publishing Corporation. Harappan Civilization: A contemporary perspective (ed.). New Delhi, Oxford & IBH Publishing Co. in collaboration with with American Institute of Indian Studies; Warminster, Aris & Phillips. Studies in the Archaeology and Palaeoanthropology of South Asia (with Kenneth A. R. Kennedy, eds). New Delhi, Oxford & IBH Publishing Co. and American Institute of Indian Studies. Buddhism: History and diversity of a great tradition, edited by Elizabeth Lyons and Heather Peters with contributions by Chang Ch'eng-mei and Gregory L. Possehl. Philadelphia, University Museum, University of Pennsylvania. Kulli: An Exploration of Ancient Civilization in Asia. Centers of Civilization, No. 1. Durham, NC, Carolina Academic Press. Radiocarbon Dates for South Asian Archaeology. Occasional publication of the Asian Section. Philadelphia, University Museum, University of Pennsylvania. Harappan Civilization and Rojdi (with M. H. Raval, eds., with contributions from Y. M. Chitalwala, Charles Frank Herman, Victoria Stack Kane, Vishnu-Mittre, Steven A. Weber). Leiden and New York, E. J. Brill; New Delhi, Oxford & IBH Publishing Co. Pvt. Ltd. and American Institute of Indian Studies. Harappan Civilization and Oriyo Timbo, edited by Paul C. Rissman and Y. M. Chitalwala with contributions from Gregory L. Possehl et al. New Delhi, Oxford & IBH Publishing Co. Pvt. Ltd. and American Institute of Indian Studies. A report on the excavations at Babar Kot: 1990-91 (with M. H. Raval). Submitted to the Archaeological Survey of India. South Asian Archaeology Studies (ed.). New Delhi, Oxford & IBH Publishing Co. Pvt. Ltd.; New York, International Science Publisher. Harappan Civilization: A Recent Perspective (ed.). 2nd revised edition. New Delhi, American Institute of Indian Studies and Oxford & IBH Publishing Co. Pvt. Ltd. Harappan Studies, vol. I (with Maurizio Tosi, eds). New Delhi, Oxford & IBH Publishing Co. Pvt. Ltd. Radiometric Dates for South Asian Archaeology. An occasional publication of the Asia Section. Philadelphia, The University of Pennsylvania Museum. Indus Age: The writing System. Philadelphia, University of Pennsylvania Press; New Delhi, Oxford IBH Publishing Co. Pvt. Ltd. Indus Age: The Beginnings. Philadelphia, University of Pennsylvania Press; New Delhi, Oxford & IBH Publishing Co. Pvt. Ltd. The Indus Civilization: A Contemporary Perspective. Walnut Creek, CA, AltaMira Press.

18

GREGORY LOUIS POSSEHL (1941 – 2011) Papers and Reviews 1967

1974 1975

1976

1977 1978 1979

1982

1983

1984

‘The Mohenjo-daro floods: A reply’. American Anthropologist, 69 (1): 32-40. Review of Excavations at Langhnaj, 1944-65. Part I, Archaeology, by Hasmukh Dhirajlal Sankalia; Part II, The fauna, by Juliet Clutton-Brock; Part III, The human remains, by Sophie Erhardt and Kenneth A. R. Kennedy (with W. A. Fairservis). American Anthropologist, 69 (5): 536-537. ‘Subsistence pattern of an early farming community in western India’ (with M. K. Dhavalikar). Puratattva, 7: 39-46. ‘The chronology of gabarbands and palas in western South Asia’. Expedition, 17 (2): 33-37. Review of Lothal and the Indus Civilization, by S. R. Rao. Journal of the American Oriental Society, 95 (1): 164-165. Review of South Asian Archaeology, edited by Norman Hammond. American Anthropologist, 77 (3): 683684. ‘Lothal: A gateway settlement of the Harappan Civilization’. In Kenneth A. R. Kennedy and G. Possehl (eds.) Ecological backgrounds of South Asian Prehistory. South Asia Occasional Papers and Theses, No. 4. Ithaca NY, Cornell University: 118-131. Review of A voyage on the Indus, by Alexander Burnes. African and Asian Studies, 11 (1-2): 132. ‘The end of a state and the continuity of a tradition: A discussion of the Late Harappan’. In Richard G. Fox (ed.) Realm and region in traditional India. Duke University Monograph and Occasional Papers Series, Monograph No. 14. Durham, NC, Duke University; New Delhi, Vikas Publishing House Pvt. Ltd.: 234-254. Review of South Asian Archaeology 1973, edited by J. E. van Lohuizen-de Leeuw and J. M. M. Ubaghs. The Journal of Asian Studies, 37 (2): 378-379. ‘Introduction’. In G. Possehl (ed.) Ancient cities of the Indus. Durham NC, Carolina Academic Press; New Delhi, Vikas Publishing House Pvt. Ltd.: vii-xv. ‘Lothal: A gateway settlement of the Harappan Civilization’. In G. Possehl (ed.) Ancient cities of the Indus. Durham NC, Carolina Academic Press; New Delhi, Vikas Publishing House Pvt. Ltd.: 212-218. ‘Radiocarbon dates for the Indus Civilization and related sites’. In G. Possehl (ed.) Ancient cities of the Indus. Durham NC, Carolina Academic Press; New Delhi, Vikas Publishing House Pvt. Ltd.: 358-360. ‘An extensive bibliography of the Indus Civilization including references cited in the text’. In G. Possehl (ed.) Ancient cities of the Indus. Durham NC, Carolina Academic Press; New Delhi, Vikas Publishing House Pvt. Ltd.: 361-422. ‘Pastoral nomadism in the Indus Civilization: An hypothesis’. In Maurizio Taddei (ed.) South Asian Archaeology 1977, vol. I. Istituto Universitario Orientale, Seminario di Studi Asiatici, Series minor, No. 6 Vol. I. Naples, Istituto Universitario Orientale, Seminario di Studi Asiatici: 537-551. ‘Hunter-gatherer/agriculturalist exchange in prehistory: An Indian example’ (with Kenneth A. R. Kennedy). Current Anthropology, 20 (3): 592-593. ‘The Harappan Civilization: A contemporary perspective’. In G. Possehl (ed.) Harappan Civilization: A contemporary perspective. New Delhi, Oxford & IBH Publishing Co. in collaboration with with American Institute of Indian Studies; Warminster, Aris & Phillips: 15-28. ‘Discovering ancient India's earliest cities: The first phase of research’. In G. Possehl (ed.) Harappan Civilization: A contemporary perspective. New Delhi, Oxford & IBH Publishing Co. in collaboration with with American Institute of Indian Studies; Warminster, Aris & Phillips: 405-413. ‘Cambay bead-making: An ancient craft in modern India’. Expedition, 23 (4): 39-47. Review of Revolutions in archaeology, by Bruce G. Trigger and Gordon Childe. Annals of the American Academy of Political and Social Science, 459 (1): 179-180. Review of Archaeological Gazetteer of Afghanistan = Catalogue des sites archéologiques d'Afghanistan, T. 1-2. (Synthèse No. 8.), by Warwick Ball avec la collaboration de Jean-Claude Gardin. Paléorient, 9 (1): 108109. Review of Journal of Intermountain Archaeology, Vol. 1, Nos. 1-2 (with Jeffrey L. Eighmy). American Anthropologist, 85 (4): 951-952. ‘A note on Harappan settlement patterns in the Punjab’. In Kenneth A. R. Kennedy and G. Possehl (eds.) Studies in the Archaeology and Palaeoanthropology of South Asia. New Delhi, Oxford & IBH Publishing Co. and American Institute of Indian Studies: 83-87. ‘Archaeological terminology and the Harappan Civilization’. In B. B. Lal and S.P. Gupta (eds.) Frontiers of the Indus Civilization: Sir Mortimer Wheeler Commemoration Volume. New Delhi, Books & Books on behalf of the Indian Archaeological Society jointly with the Indian History & Culture Society: 27-36. ‘Excavations at Rojdi: 1982-83’ (with Y. M. Chitalwala, Paul C. Rissman and Gail E. Wagner). Puratattva, 13-14: 155-163. 19

Vasant Shinde & Asko Parpola 1985 1986 1987

1988

1990

1991

1992

1993

1994

‘Preliminary report on the second season of excavations at Rojdi: 1983-84’ (with Y. M. Chitalwala, Paul C. Rissman, Gail E. Wagner, Pamela Crabtree and Julia Longenecker). Man and Environment, 9: 80-100. ‘African millets in South Asian prehistory’. In Jerome Jacobson (ed.) Studies in the archaeology of India and Pakistan. New Delhi, Oxford & IBH Publishing Co. and American Institute of Indian Studies: 237-256. ‘Rojdi: The investigation of a prehistoric town in India’. American Journal of Archaeology, 90 (4): 467-468. Review of Pre-Harappan cultures of India and the Borderlands, by Shashi Asthana. Journal of the American Oriental Society, 107 (4): 839. Review of Die Indus-Zivilisation, by Michael Jansen. Journal of the American Oriental Society, 107 (4): 845-846. ‘Indian Archaeology, A Review: Guide to excavated sites 1953-54 through 1983-84’. Puratattva, 18: 113172. ‘Prehistoric cultural regions in South Asia’. In Peter Gaeffke and David A. Utz (eds.) The countries of South Asia: Boundaries, extensions and interrelations. Proceedings of the South Asia seminar, No. 3. Philadelphia, Department of South Asia Regional Studies, University of Pennsylvania. ‘Radiocarbon dates from South Asia’. Man and Environment, 12: 169-196. ‘Some further thoughts on unicorns’. In C. P. Sinha (ed.) Archaeology and art: Krishna Deva felicitation, Vol. I. Delhi, Ramanand Vidya Bhavan: 44-48. ‘The Sorath Harappan: A new regional manifestation of the Indus urban phase’ (with Charles Frank Herman). In Maurizio Taddei (ed.) South Asian Archaeology 1987, Part I. Serie Orientale Roma No. 66, Vol. I. Rome, Istituto Italiano per il Medio ed Estremo Oriente: 295-319. ‘Revolution in the urban revolution: The emergence of Indus urbanization’. Annual Review of Anthropology, 19: 261-282. ‘An archaeological adventurer in Afghanistan: Charles Masson’. South Asian Studies, 6: 111-124. ‘Hasmukh Dhirajlal Sankalia (1908-1989)’ (with Kenneth A. R. Kennedy). American Anthropologist, 92: 1006-1010. ‘A tale of two seals’ (with M. H. Raval). Ancient Ceylon, 7: 157-163. Review of A history of Indian archaeology: From the beginnings to 1947, by Dilip K. Chakrabarti. Journal of the American Oriental Society, 110 (2): 377. ‘A short history of archaeological discovery at Harappa’. In Richard H. Meadow (ed.) Harappa excavations 1986-1990: A multidisciplinary approach to third millennium urbanism. Monographs in World Archaeology, No. 3. Madison, WI, Prehistory Press: 5-11. ‘A house of the Harappan Post-Urban phase: Structures I and IV at Rojdi’. In C. Margabandhu, K. S. Ramachandran, A. P. Sagar and D. K. Sinha (eds.) Indian archaeological heritage: Shri K. V. Soundara Rajan Festschrift, vol. I. Agam Indological Series, No. 11. Delhi, Agam Kala Prakashan: 113-121. Review of An encyclopaedia of Indian archaeology, vol. I-II, edited by A. Ghosh. Journal of the American Oriental Society, 111 (1): 207-208. ‘Walter Ashlin Fairservis, Jr.’. In G. Possehl (ed.) South Asian Archaeology Studies. New Delhi, Oxford & IBH Publishing Co. Pvt. Ltd.; New York, International Science Publisher: 1-12. ‘Toymakers and trade: A notice of early twentieth century commerce between Philadelphia and India’. In G. Possehl (ed.) South Asian Archaeology Studies. New Delhi, Oxford & IBH Publishing Co. Pvt. Ltd.; New York, International Science Publisher: 261-266. ‘The Harappan cultural mosaic: Ecology revisited’. In Catherine Jarrige (ed.) South Asian Archaeology 1989. Monographs in World Archaeology, No. 14. Madison WI, Prehistory Press: 237-244. ‘The chronology of prehistoric India: From earliest times to the Iron Age’ (with Paul C. Rissman). In Robert W. Ehrich (ed.) Chronologies in Old World archaeology. 3rd edition. Chicago, The University of Chicago Press. Vol. I: 465-490; vol. II: 447-474. ‘The Harappan Civilization in Gujarat: The Sorath and Sindhi Harappans’. The Eastern Anthropologist, 45 (1-2): 117-154. ‘The Pre-Harappan period at Prabhas Patan and the Pre-Harappan phase in Gujarat’ (with M. K. Dhavalikar). Man and Environment, 17 (1): 71-78. ‘The date of Indus urbanization: A proposed chronology for the Pre-Urban and Urban Harappan phases’. In Adalbert J. Gail and Gerd J. R. Mevissen (eds.) South Asian Archaeology 1991. Stuttgart, Franz Steiner Verlag: 231-249. ‘Sir Leonard Woolley evaluates Indian archaeology’. Harappan Studies, 1: 1-56. ‘Of men’. In Jonathan Mark Kenoyer (ed.) From Sumer to Meluhha: Contributions to the archaeology of South and West Asia in memory of George F. Dales, Jr. Wisconsin Archaeological Reports, No. 3. Madison, WI, Department of Anthropology, University of Wisconsin: 179-186

20

GREGORY LOUIS POSSEHL (1941 – 2011) 1994

1996

1997

1998

1999

2000

2001

‘Govindbhai-no vadi: A modern farmer's garden in Gujarat and its ethnoarchaeological significance’. In Bridget Allchin (ed.) Living traditions: Studies in the ethnoarchaeology of South Asia. New Delhi, Oxford & IBH Publishing Co. Pvt. Ltd.: 193-204. ‘Excavations at Rojdi, 1992-93’ (with Dinker P. Mehta). In Asko Parpola and Petteri Koskikallio (eds.) South Asian Archaeology 1993, vol. II. Annales Academiae Scientiarum Fennicae No. 271, Vol. II. Helsinki, Suomalainen Tiedeakatemia: 603-614. ‘The Indus Civilization’. Man and Environment, 19 (1-2): 103-113. ‘Meluhha’. In Julian Reade (ed.) The Indian Ocean in Antiquity. London, Kegan Paul International in association with the British Museum: 133-208. ‘Climate and the eclipse of the ancient cities of the Indus’. In H. Nüzhet Dalfes, George Kukla and Harvey Weiss (eds.) Third millennium BC climate change and Old World collapse. NATO ASI, Series No. 1: Global Environment Change, Vol. 49. Berlin - New York, Springer: 193-244. ‘The date of the Nal cemetery’. In C. Margabandhu and K. S. Ramachandran (eds.) Spectrum of Indian Culture (Professor S. B. Deo felicitation volume), vol. I. Delhi, Agam Kala Prakashan: 67-76. ‘Mehrgarh’. In Brian Fagan (ed.) The Oxford Companion to Archaeology. Oxford, Oxford University Press: 436-438. ‘The date of the Surkotada cemetery: A reassessment in light of recent archaeological work in Gujarat’. In Jagat Pati Joshi (ed.) Facets of Indian Civilization: Recent perspectives (Essays in honour of Professor B. B. Lal). New Delhi, Aryan Books International: 81-87. ‘Seafaring merchants of Meluhha’. In Raymond Allchin and Bridget Allchin (eds.) South Asian Archaeology 1995, vol. I. New York, Published for The Ancient India and Iran Trust by Science Publishers, Inc.; New Delhi, Oxford & IBH Publishing Co. Pvt. Ltd.: 87-100. ‘The transformation of the Indus Civilization’. Journal of World Prehistory, 11 (4): 425-472. ‘An Harappan outpost on the Amu Darya: Shortughai, Why was it there?’. Indologica Taurinensia, 23-24: 57-70. ‘Did the Sarasvati ever flow to the sea?’. In C. S. Phillips, D. T. Potts and S. Searight (eds.) Arabia and its neighbours: Essays on prehistorical and historical developments presented in honour of Beatrice de Cardi. Turnhout, Brepols: 339-354. ‘Sociocultural complexity without the state: The Indus Civilization’. In Gary M. Feinman and Joyce Marcus (eds.) Archaic states. School of American Research advanced seminar series. Santa Fe, NM, School of American Research Press; Oxford, James Currey: 261-291. ‘Introduction of African millets to the Indian subcontinent’. In H. D. V. Pendergast, N. L. Etkin, D. R. Harris and P. J. Houghton (eds.) Plants for food and medicine. Kew, The Royal Botanic Gardens: 107-121. Review of The Indo-Aryans of ancient South Asia: Language, material culture and ethnicity, edited by George Erdosy. Journal of the American Oriental Society, 118 (1): 120-121. ‘Prehistoric population and settlement in Sindh’. In Azra Meadows and Peter S. Meadows (eds.) The Indus River: Biodiversity, resources, humankind. Oxford and Karachi, Oxford University Press for the Linnean Society of London: 393-409. ‘The Early Iron Age in South Asia’ (with Praveena Gullapalli). In Vincent C. Pigott (ed.) The Archaeometallurgy of the Asian Old World. MASCA Research Papers in Science and Archaeology, No. 16; University Museum Monograph, No. 89. Philadelphia, The University Museum, University of Pennsylvania: 153-175. ‘The transformation of the Indus Civilization’. Man and Environment, 24 (2): 1-33. ‘Harappan beginnings’. In Martha Lamberg-Karlovsky (ed.) The breakout: The origins of civilization. Peabody Museum Monographs, No. 9. Cambridge, MA, Peabody Museum, Harvard University: 99-112. ‘The drying up of the Sarasvati: Environmental disruption in South Asian prehistory’. In Garth Bawden and Richard Martin Reycraft (eds.) Environmental disaster and the archaeology of human response. Anthropological papers, No. 7. Albuquerque, NM, Maxwell Museum of Anthropology, University of New Mexico: 63-74. ‘The Early Harappan phase’. Bulletin of the Deccan College Post-Graduate and Research Institute, 60-61: 227-241. ‘The Mature Harapopan phase’. Bulletin of the Deccan College Post-Graduate and Research Institute, 60-61: 243-251. ‘Climatic conditions and the rise and fall of Harappan Civilization of South Asia’ (with Vasant Shinde and Yoshinori Yasuda). In Monsoon and Civilization conference abstracts. Kyoto, International Research Center for Japanese Studies: 92-94. ‘The ceramic assamblage in Protohistoric Mewar (Rajasthan), with special reference to Gilund and Balathal’ (with Vasant Shinde and Shweta Sinha Deshpande). Puratattva, 32: 5-24. 21

Vasant Shinde & Asko Parpola 2002

2003

2004

2005

2007

2008

2009 2010

‘The Early Harappan’. In K. Paddayya (ed.) Recent studies in Indian archaeology. Indian Council of Historical Research monograph series, No. 6. New Delhi, Munshiram Manoharlal: 112-128. ‘Fifty years of Harappan archaeology: The study of the Indus Civilization since Indian independence’. In S. Settar and Ravi Korisettar (eds.) Protohistory: Archaeology of the Harappan Civilization. Indian archaeology in retrospect, Vol. II. New Delhi, Indian Council of Historical Research & Manohar: 1-46. ‘Archaeology of the Harappan Civilization: An annotated list of excavations and surveys’. In S. Settar and Ravi Korisettar (eds.) Protohistory: Archaeology of the Harappan Civilization. Indian archaeology in retrospect, Vol. II. New Delhi, Indian Council of Historical Research & Manohar: 421-482. ‘Harappans and hunters: Economic interaction and specialization in prehistoric India’. In Kathleen D. Morrison and Laura L. Junker (eds.) Forager-traders in South and Southeast India: Long term histories. Cambridge, Cambridge University Press: 62-76. ‘Vedic’ (with Michael Witzel). In Peter N. Peregrine and Melvin Ember (eds.) Encyclopedia of Prehistory. Vol. 8, South and Southwest Asia. New York – Dordrecht, Kluwer Academic, Plenum Publishers: 391-396. ‘Indus-Mesopotamian trade: The record in the Indus’. Iranica Antiqua, 37: 325-342. Review of Cracking codes: The Rosetta stone and the art of decipherment, by Richard B. Parkinson. Review of Unlocking the secrets of ancient writing: The parallel lives of Michael Ventris and Linda Schele and the decipherment of Mycenaean and Mayan writing, by Thomas G. Palaima (with Elizabeth Pope and F. Kent Reilly). American Journal of Archaeology, 106 (4): 602-603. ‘The Indus Civilization: An introduction to environment, subsistence, and cultural history’. In Steven A. Weber and William R. Belcher (eds.) Indus Ethnobiology: New perspectives from the field. Lanham MD, Lexington Books: 1-20. ‘Protohistory, 3000-1200 BCE’. In Frederick M. Asher (ed.) Art of India: Prehistory to the present. London & New Delhi, Encyclopedia Britannica. ‘Chiefdoms and kingdoms, 1200-400 BCE’ In Frederick M. Asher (ed.) Art of India: Prehistory to the present. London & New Delhi, Encyclopedia Britannica. ‘The Ahar-Banas complex and the BMAC’ (with Vasant Shinde and Marta Ameri). Man and Environment, 29 (2): 18-29. Review of The flow of power: Ancient water systems and landscapes, edited by Vernon L. Scarborough. Journal of Anthropological Research, 60 (4): 605-607. ‘A report on the excavations at Gilund, 1999-2001’ (with Vasant Shinde). In Catherine Jarrige and Vincent Lefèvre (eds.) South Asian Archaeology 2001, vol. I. Paris, Éditions Recherche sur les Civilisations: 293302. ‘Excavations at Gilund 2001-2003: The seal impressions and other finds’ (with V. Shinde and M. Ameri). In Ute Franke-Vogt & Hans-Joachim Weisshaar (eds.) South Asian Archaeology 2003. Forschungen zur Archäologie aussereuropäischer Kulturen, No. 1. Aachen, Linden Soft Verlag e. K.: 159-169. ‘Mohenjo-Daro: The symbolic landscape of an ancient city’. In Tony Atkin and Joseph Rykwert (ed.) Structure and meaning in human settlements. Philadelphia, University of Pennsylvania Museum of Archaeology and Anthropology: 67-81. ‘The Indus Civilization’. In John R. Hinnells (ed.) Handbook of ancient religions. Cambridge, Cambridge University Press: 418-489. ‘Thoughts on the evolution and history of human populations in South Asia’. In Michael D. Petraglia and Bridget Allchin (eds.) The evolution and history of human populations in South Asia: Interdisciplinary studies in archaeology, biological anthropology, linguistics and genetics. Vertebrate paleobiology and paleoanthropology, No. 4. Dordrecht - New York, Springer: 447-459. ‘The Harappan settlement of Gujarat’. In Elizabeth Stone (ed.) Settlement and society: Ecology, urbanism, trade and technology in Mesopotamia and beyond. Los Angeles, Costen Institute of Archaeology, University of California; Chicago, The Oriental Institute: 297-328. ‘The Middle Asian interaction sphere’. Expedition, 49 (1): 40-42. ‘Indus folklore: An unknown story on some Harappan objects’. In Eric Olijdam and Richard H. Spoor (eds.) Intercultural relations between South and Southwest Asia: Studies in commemoration of E. C. L. During Caspers (1934-1996). Society for Arabian Studies monographs, No. 7; BAR International series No. 1826. Oxford, BAR Publishing: 140-144. ‘Foreword’. In Jagat Pati Joshi Harappan Architecture And Civil Engineering. Infinity Foundation series. New Delhi, Rupa & Co. in association with Infinity Foundation. ‘W. Norman Brown’. Expedition, 50 (2): 33-36. ‘The Indus Civilization’. In John R. Hinnells (ed.) The Penguin handbook of ancient religions. London, Penguin Books Ltd: 418-489. ‘Ernest J. H. Mackay and the Penn Museum’. Expedition, 52 (1): 40-43. 22

GREGORY LOUIS POSSEHL (1941 – 2011) 2010

Review of Corpus of Indus Seals and Inscriptions. Volume 3, New material, untraced objects, and collections outside India and Pakistan. Part 1, Mohenjo-daro and Harappa, by Asko Parpola, B. M. Pande and Petteri Koskikallio in collaboration with Richard H. Meadow and J. Mark Kenoyer. Journal of the American Oriental Society, 130 (2): 271-272. ‘What’s in a name? The “Indus Civilization” onomastics and the rules of science’. In P. M. Kozhin, M. F. Kosarev and N. A. Dubova (Eds.) On the track of uncovering a civilization. A volume in honor of the 80th anniversary of Victor Sarianidi. Sankt-Petersburg: Aletheia: 120-128.

23

This Volume is dedicated also to the memory of Dr. Swaraja Prakash Gupta (1931-2007)

SWARAJA PRAKASH GUPTA (Allahabad December 22, 1931 – New Dehli October 3, 2007)

25

SWARAJA PRAKASH GUPTA (1931 – 2007) K. N. Dikshit

Dr. Swaraja Prakash Gupta was born on 22 nd December, 1931 at Allahabad. His father late Shri Onkar Nath, who was a freedom fighter during Nehru’s time at Allahabad died at an early age and there was none to look after him except his mother, elder brother and maternal grandfather from Itawah (U.P.), who supported his education in the universities and elsewhere. This led him to take deep interest in higher studies and eventually history and archaeology became special area of his interest. He was educated and associated with different universities in India and abroad. He earned his degree of M.A. and L.L.B. from Allahabad, Ph. D. from Delhi and D. Litt. from Magadh Universities. He also obtained Certificate in Museology from ‘École du Louvre and Musée de l'Homme’, Government of France. He secured first position with credit in Diploma in Archaeology from the School of Archaeology, New Delhi and was awarded Wheeler Prize for Excellence in the Field of Archaeology. He was also awarded Maulana Azad and Centenary Commemoration Gold Medal for Excellence in Archaeology. He also obtained Certificate of Merit in Environment Archaeology from the Institute of Archaeology, University of London.

the Dead and Physical Types in Ancient Indian (New Delhi, 1971), Tourism, Museum and Monuments (New Delhi, 1975), Archaeology of Soviet Central Asia and the Indian Borderlands (New Delhi, 1978) 2 volumes, The Roots of Indian Art (New Delhi, 1980) and French Edition (1990), The Indus-Saraswati Civilization (New Delhi, 1996), The Elements of Indian Art (New Delhi, 2002), Cultural Tourism in India (New Delhi, 2002). He also edited several books entitled India’s Contribution to World Thought and Culture (Madras, 1972), Mahabharata: Myth and Reality -Differing Views (New Delhi, 1976), Frontiers of the Indus Civilization (New Delhi, 1983), Kushana Art of Sanghol (New Delhi, 1986) and Masterpiece of the National Museum Collection (New Delhi, 1986) and also reviewed a number of books in different research journals. Dr. S. P. Gupta was a widely travelled scholar covering as many as 32 countries from USA to Europe and Asian destinations. He has contributed more than 100 articles on Indian Art and Archaeology, which are still the source of information on archaeology for the scholars and students alike. Although he was not well, yet he was trying to complete several projects, which were in his hands but couldn’t see the light of the day due to his sudden demise. These comprised of books entitled Atlas of the IndusSaraswati Civilization, Scientific Investigation of the Lost Saraswati River, Elements of Indian Temple Architecture, A Comprehensive Study of Indian Archaeology, and Practical Dictionary of Indian Icons and Modern Museum Management.

After completing his M.A., Dr. S. P. Gupta took training in archaeology at Nagarjunakonda (Andhra Pradesh) and thereafter he joined Archaeological Survey of India in 1955 in the Atlas Branch at Nagpur. He was appointed as Deputy Keeper in the National Museum in 1960 and later rose to the rank of Assistant Director. He was also invited to become the Director of the Allahabad Museum where he served for a period of two years. During his tenure, this museum also attained the status of a National Museum.

Dr. Gupta lived a very simple, almost austere life. He was a strict vegetarian, a teetotaler and never smoked. He was not religious in the conventional sense of the term and was never seen performing any rituals or visiting a temple for offering worship. His religion was doing his duty conscientiously, living a simple and honest life, and helping others.

Dr. S. P. Gupta was a leading art historian, a distinguished scholar of Museology and a well-known expert of Harappan Archaeology. He was never seen thinking and doing anything else, other than archaeology, which led him to dream and build a Centre of Higher Learning, which finally culminated in the development of the Centre for Research and Training in History, Archaeology and Paleo-environment. The Centre presently houses a large number of books on Indian Archaeology and students and scholars from all over the world frequently come to consult the library.

Dr. Gupta’s commitment to archaeology was total. He pursued research for well over half a century with complete dedication, great vigour, and ceaseless hard work. His name will remain immortal in Indian Archaeology on the basis of his abiding contribution in the form of his own research and publications, nurturing Indian Archaeological Society and its journal Puratattva, founding History Today and Journal of Indian Ocean Archaeology.

Dr. Gupta after his superannuation took keen interest in the development of Indian Archaeological Society and Indraprastha Museum of Art and Archaeology in the present premises which got completed in the year 2000.

Dr. Gupta’s noble thoughts, high ideals, love inspiring words and caring advice will continue to guide us in every moment of our lives. He was a great visionary,

Dr. S. P. Gupta wrote several books on Indian Art and Archaeology. Some of his pioneer books are Disposal of 27

K. N. Dikshit philanthropist and a novel soul. With high respect in our heart, one can only say that he was a man Born for Archaeology and the father figure of Modern Archaeology in the country. He remained unmarried and passed away on 3 rd October, 2007. Indian Archaeological Society New Delhi

28

SWARAJA PRAKASH GUPTA (1931 – 2007)

PUBLICATIONS OF SWARAJA PRAKASH GUPTA Collected by K. N. Dikshit Books and Monographs 1971 1972 1975 1976 1978

1979 1980 1983 1986 1989 1996 2002

India’s Contribution to World Thought and Culture, Swami Vivekananda Commemoration Volume (with Lokesh Chandra, eds.). Kanyakumari, Swami Vivekanand Kendra. Disposal of the Dead and Physical Types in Ancient India. Delhi, Oriental Publishers. Tourism, Museums and Monuments: A study of Tourism in India in the light of Indian Museums and Archaeological Monuments. New Delhi, Oriental Publishers. Mahabharata: Myth and Reality Differing Views (ed.). New Delhi, B.R. Publishing Corporation. Archaeology of Soviet Central Asia and the Indian Borderlands Vol. I, (From Paleolithic to Mesolithic). New Delhi, B.R. Publishing Corporation. Archaeology of Soviet Central Asia and the Indian Borderlands Vol. II, (From Mesolithic to Early Iron Age). New Delhi, B.R. Publishing Corporation. The Origin of Brahmi Script. Delhi, D.K. Publications. The Roots of Indian Art. New Delhi, B.R. Publishing Corporation. Frontiers of the Indus Civilization (with B. B. Lal and Shashi Asthana, eds.). New Delhi, Books and Books. Kushana Art of Sanghol. New Delhi, National Museum. Masterpieces of the National Museum Collection. New Delhi, National Museum. An Archaeological Tour Along the Ghaggar Hakra River (ed.). Meerut, Kusumanjali Prakashan. The Indus Saraswati Civilization: Origins, Problems and Issues. Delhi, Pratibha Prakashan. Cultural Tourism in India. New Delhi, D.K. Printworld. The Elements of Indian Art. New Delhi, D.K. Printworld.

Selected Papers 1960 1961 1963 1965 1967

1968

1969

1971 1972

‘Burial customs in Ancient India’. JBRS XLVI (1 4): 84 102. ‘Arrowheads their technology and history’. JBRS XLVII (1 4): 129 142. ‘Indian copper hoards’. JBRS XLIX (1 4): 147 166. ‘Further copper hoards: A reassessment in the light of new evidence’. JBRS LI (1 4): 1 7. ‘Middle Stone Age and Late Stone Age tools, District Mandla, Madhya Pradesh’ (with C. B. Trivedi). IAR 1963 64: 88. ‘Terracotta vessels and figurines from Khotan (Central Asia) in the Hermitage Museum, Leningrad’. JOI XVII (2): 168 180. ‘The mountainous Neolithic Cultures of Central Asia and Northern India’. The Anthropologist XIV: 125 136. ‘Neolithic finds from Lou Lan: Origin and extra¬territorial affinities’. Bharati X & XI: 175 189. ‘New evidence of the Harappa Culture in the Soviet Central Asia’ (with A. Y. Shchatenka). Bharati X & XI: 190 197. ‘Relationship between the Chalcolithic Cultures of India and the Chust Culture of the Farghana Valley, U.S.S.R’. Puratattva 1: 47 53. ‘Review of the problem of relationship between the Chopper Chopping complex of Central Asia and India’. Bharati X & XI: 15 24. ‘Determining the technique of handmade and wheel turned pottery by microscopic analysis’. Puratattva 2: 23 25. ‘Sociology of the Burial practices and its bearing on the problem of Megaliths’. In Seminar papers on the Problem of Megaliths in India, edited by A. K. Narain (ed.), pp. 99 106. Varanasi, Dept. of AIHC and Archaeology, Banaras Hindu University. ‘Gulf of Oman: The original home of Indian Megaliths’. Puratattva 4: 4 18. ‘NBP’. Puratattva 5: 35 36. 29

K. N. Dikshit 1972 1973

1974 1976

1977 1978 1979 1980 1982

1983 1984 1987 1989

1993

1996 1998 2000

2001 2002 2003

2004

2006

‘OCP’. Puratattva 5: 7 8. ‘A model for understanding the first urbanization in India’. Puratattva 6: 42 50. ‘An introduction to models and model making’. In RIA, pp. 359-65. ‘Panel discussion on Protohistoric chrono cultural lacunae’ (with K. N. Dikshit, D. P. Agrawal, F. R. Allchin, G. F. Dales, S. B. Deo, V. N. Misra, M. S. Nagarajarao, S. N. Rao, K. M. Shrimali and B. K. Thapar). In RIA, pp. 494-503. ‘Two urbanizations in India: A side study in their social structure’. Puratattva 7: 53 60. ‘Origin of the form of Harappan Culture: a new proposition’. Puratattva 8: 141 146. ‘The problem of missing link in the process of first urbanization in India’. In Archaeological Congress and Seminar: 1972, edited by U. V. Singh (Ed.) pp. 157 165. Kurukshetra, B. N. Chakravarty University. ‘Painted Grey Ware sites in relation to old river beds in Rajasthan’ (with Shashi Asthana and Amarendra Nath). In EAWI, pp. 79-92. ‘Harappan vis-à-vis Mesopotamian urbanisation and town planning’ (with Shashi Asthana). ME II: 47 50. ‘Baluchistan and Afghanistan: refuge areas or nuclear zones? ’. In EIP, pp. 9 15. ‘Writing Indian art history: a search for valid approach’. Puratattva 11: 68 73. ‘A new tool type of the "Copper Hoards"’. Puratattva 11: 127 128. ‘The alleged Upper Palaeolithic bone Mother Goddess from Belan’. Puratattva 11: 116 117. ‘The Late Harappan: a study in cultural dynamics’. In Harappan Civilization, edited by G. L. Possehl (ed.), pp. 51 59. New Delhi, Oxford and IBH. ‘Herding as the backdrop to the growth of agriculture in West Asia and South Asia’ (with A. Kesarwani). Puratattva 12: 101 111. ‘Internal trade of the Harappans’. In FIC: 417 424. ‘ “Copper Hoard” implements in the National Museum Collection: 1986’. Puratattva 16: 45 46. ‘Cultures: Copper Hoards’. In EIA: pp. 91 93. ‘Cultures: Megalithic’. In EIA: pp. 110 116. ‘Geophysical co-ordinates: River valleys and changes in river courses Holocene’. In EIA: 213 214. ‘Pottery: Pre Harappa’. In EIA: 228 230. ‘Ritual and Cult Objects: Disposal of the Dead’. In EIA: 265 267. ‘Longer chronology of the Indus Saraswati Civilization’. Puratattva 23: 21 29. ‘The Late Harappan: a study in cultural dynamics’. In Harappan Civilization, edited by G.L. Possehl (Ed.), 2nd edition, pp. 51 59. New Delhi, AIIS and Oxford IBH. ‘Fresh look at the Indus Sarasvati chronology: From the Formative to the Mature Period’. In SIC, pp. 63 66. Delhi, B.R. Publishing Corporation. ‘Revised Chronology of the Indus Saraswati Civilization’. In Dating in Indian Archaeology: Problems and Perspectives, edited by T. P. Verma (Ed.), pp. 42 56. Mysore, Bharatiya Itihas Sankalan Samiti. ‘Studies in the Beginning of the Harappan Civilization at the End of the Millennium’. Man and Environment 25 (2): 7-11. ‘The Mahabharata for Harappa’. History Today 1: 46-48. ‘A discussions on River Saraswati in History, Archaeology and Geology’. Puratattva 31: 30-38. ‘Myth of Saffronization of Indian History’. History Today 2: 256-59. ‘New Light on Medieval History: Excavations at Sanjan’ (with Kurush Dalal). HT 3: 99-100. ‘Pro-Minority and Anti-Majority Communal Historians’. HT 3: 46-48. ‘A Report on Atlas of the Indus-Saraswati Civilization’. Puratattva 33: 165-66. ‘Imagining Second Mosque at Ayodhya- A Critique of Irfan Habib's 'The ASI's Report on Ayodhya'’. History Today 4: 59-69. ‘Antiquities from Kamrej Excavations-2003’ (with T. Garge, S. Gupta and A. Geetali). JIOA 35: 209-213. ‘On the Fast Track of the Periplus: Excavations at Kamrej-2003’ (with Sunil Gupta, Tejas Garge, Rohini Pandey, Anuja Geetali and Sonali Gupta). JIOA 1: 9-33. ‘Our Concerns: Our Resolve’. JIOA 3: v-vi.

30

OPENING ADDRESS OF THE 19TH INTERNATIONAL CONFERENCE OF THE EUROPEAN ASSOCIATION OF SOUTH ASIAN ARCHAEOLOGY RAVENNA, ITALY, 2-6 JULY 2007

LET NOT THE 19TH CENTURY PARADIGMS CONTINUE TO HAUNT US ! B. B. Lal

discovered a fortification wall at Harappa and on learning that the Aryan god Indra had been referred to as puramdara (destroyer of forts) he readily pronounced his judgment (Wheeler 1947: 82): “On circumstantial evidence Indra [representing the Aryans] stands accused [of destroying the Harappan Civilization].” In further support of his thesis, he cited certain human skeletons at Mohenjo-daro, saying that these were the people massacred by the Aryan invaders. Thus was reached the peak of the ‘Aryan Invasion’ theory.

Distinguished fellow delegates and other members of the audience, I am most grateful to the organizers of this conference, in particular to the President, Professor Maurizio Tosi, not only for inviting me to participate in this Conference but also for giving me the additional honour of delivering the Inaugural Address. Indeed, I have no words to thank them adequately for their kindness. Perhaps this is the first occasion when a South Asian is being given this privileged treatment by the European Association of South Asian Archaeologists. The conference hall is full of scholars from all parts of the world – from the United States of America on the west to the Land of the Rising Sun, Japan, on the east. All these scholars have contributed in a number of ways to our understanding of the past of South Asia, and I salute them with all the humility that I can muster. However, I hope I will not be misunderstood when I say that some amongst us have not yet been able to shake off the 19 th century biases that have blurred our vision of South Asia’s past. As is well known, it was the renowned German scholar Max Muller who, in the 19th century, attempted for the first time to date the Vedas. Accepting that the Sutra literature was datable to the 6 th century BCE, he gave a block-period of 200 years to the preceding three parts of the Vedic literature, namely the Âraņyakas, Brâhmaņas and Vedas. Thus, he arrived at 1200 BCE as the date of the Vedas. However, when his contemporaries, like Goldstucker, Whitney and Wilson, objected to his ad-hocism, he toned down, and finally surrendered by saying (Max Muller 1890, reprint 1979): “Whether the Vedic hymns were composed [in] 1000 or 1500 or 2000 or 3000 BC, no power on earth will ever determine.” But the great pity is that, in spite of such a candid confession by the savant himself, many of his followers continue to swear by his initial dating, viz. 1200 BCE.

And lo and behold! The very first one to fall in the trap of the ‘Aryan Invasion’ theory was none else but the guru’s disciple himself. With all the enthusiasm inherited from the guru, I started looking for the remains of some culture that may be post-Harappan but anterior to the early historical times. In my exploration of the sites associated with the Mahābhārata story I came across the Painted Grey Ware Culture which fitted the bill. It antedated the Northern Black Polished Ware whose beginning went back to the 6th-7th century BCE, and overlay, with a break in between, the Ochre Colour Ware of the early 2 nd millennium BCE. In my report on the excavations at Hastināpura and in a few subsequent papers I expressed the view that the Painted Grey Ware Culture represented the early Aryans in India. But the honeymoon was soon to be over. Excavations in the middle Gangāvalley threw up in the pre-NBP strata a ceramic industry with the same shapes (viz. bowls and dishes) and painted designs as in the case of the PGW, the only difference being that in the former case the ware had a black or black-and-red surface-colour, which, however, was just the result of a particular method of firing. And even the associated cultural equipment was alike in the two cases. All this similarity opened my eyes and I could no longer sustain the theory of the PGW having been a representative of the early Aryans in India (the association of this Ware with the Mahābhārata story was nevertheless sustainable since that event comes at a later stage in the sequence). I had no qualms in abandoning my then-favourite theory. But linguists are far ahead of archaeologists in pushing the poor Aryans through the Khyber/Bolan passes into India. In doing so, they would not mind even distorting the original Sanskrit texts. A case in point is that of the well known Professor of Sanskrit at the Harvard University, Professor Witzel. He did not hesitate to mistranslate a part of the Baudhāyana Śauta-sūtra (Witzel 1995: 320-21). In 2003 I published a paper in the East

The ultimate effect of this blind tenacity was that when in the 1920s the great civilization, now known variously as the Harappan, Indus or Indus-Sarasvati Civilization, was discovered in South Asia, and was dated to the 3rd millennium BCE, it was argued that since the Vedas were no earlier than 1200 BCE, the Harappan Civilization could not have been Vedic. Further, since the only other major linguistic group in the region was the Dravidian, it was held that the Harappans were a Dravidian-speaking people. Then came the master stroke. In 1946, my revered guru Mortimer Wheeler (later knighted) 31

B. B. Lal familiar with it. I would also like to take this opportunity to heartily congratulate Professor Sarianidi and other archaeologists whose sustained field-work has placed the BMAC on the same high pedestal as occupied by other civilizations of the ancient world.

and West (Vol. 53, Nos. 1-4), exposing his manipulation. Witzel’s translation of the relevant Sanskrit text was as follows: Ayu went eastwards. His (people) are the KuruPancālas and Kūsi- -Videha. This is the Âyava (migration). (His other people) stayed at home in the west. His people are the Gāndhāri, Paraśu and Araţţa. This is the Amāvasava (group).

Characteristic Features of the BMAC First, a map of the region involved is given below (Fig. 2): It is now proposed to discuss, though very briefly, the following features of the BMAC:

Whereas the correct translation is: Ayu migrated eastwards. His (people) are the Kuru-Pancālas and Kāśi-Vihedas. This is the Âyava (migration). Amāvasu migrated westwards. His (people) are the Gāndhāri, Paraśu and Araţţa. This is the Amāvasu (migration).

1. 2. 3. 4. 5. 6. 7.

According to the correct translation, there was no movement of the Aryan people from anywhere in the north-west. On the other hand, the evidence indicates that it was from an intermediary point that some of the Aryan tribes went eastwards and other westwards. This would be clear from the map that follows (Fig.1): Professor Witzel and I happened to participate in a seminar organized by UMASS, Dartmouth in June 2006. When I referred, during the course of my presentation, to this wrong translation by the learned Professor, he, instead of providing evidence in support of his own stand, shot at me by saying that I did not know the difference between Vedic and Classical Sanskrit. Should that be the level of an academic debate? (Anyway, he had to be told that I had the privilege of obtaining in 1943 my Master’s Degree in Sanskrit (with the Vedas included), with a First Class First, from a first class university of India, namely Allahabad.)

town-planning and monumental architecture; the ceramics; stone objects; metal objects; sculptural art; seals and amulets; and finally the chronological horizon.

This seemingly uncalled for exercise is being done in order to demonstrate that (a) the BMAC people were not nomads, as held by my Indian colleagues; and (b) these characteristic features of the BMAC never ever reached east of the Indus up to the upper Gangā-Yamunā doab – an area which was the homeland of the Ŗigvedic people, as is clear from the Nadīstuti verses (10.75.5 and 6) of the Ŗigveda itself. Town-planning and Monumental Architecture The BMAC settlements, by and large, were well planned and contained monumental structures. This is abundantly clear, for example, from the excavated remains at Dashly3, in Bactria, where a ‘cultic centre’ has been found located within a series of three successive fortifications. The cultic centre, circular on plan, with a diameter of 40 metres, was provided with nine square bastions on the exterior (Fig. 3).

To revert to the theory of ‘Aryan Invasion’ The year 1964 saw this theory being pushed back in the reverse gear. George F. Dales published a paper in which, after presenting a thorough analysis of the evidence, he completely lambasted the massacre theory and, in a way, rescued Indra from the charge of having been a marauder. Since then many other scholars (e.g. Danino 2006, Kenoyer 1998, Lal 2002 and 2005, Renfrew 1988, Shaffer and Lichtenstein 1999) have adduced further evidence to add successive nails on to the coffin of the ‘Aryan Invasion’ theory. Unfortunately, however, the ghost of ‘Aryan Invasion’ is not buried deep enough. It is being resurrected in the form of ‘Aryan Immigration’, and in this context the Bactria-Margiana region is said to be the source. Out of the scholars who stand by this rejuvenated thesis, I shall deal here with four representative ones, namely Professors Romila Thapar and R. S. Sharma from India and Professors Asko Parpola and V. I. Sarianidi from the West. But before the views of these four scholars are examined it seems appropriate to spell out in some detail the nature of the Bactria-Margiana Archaeological Complex (BMAC), even though some of the scholars present here may be

In Margiana, excavations at Togolok-1 and Togolok-21 have yielded the remains of multi-roomed temples, of which that at the latter site is more elaborate. Covering an overall area of about 1.5 hectares, the Toglok-21 complex had in the centre a 60 x 50 m. unit enclosed by a 5 m. thick wall which was provided on its exterior with four circular towers, one at each corner and two semi-circular towers, one each abutting the exterior of the eastern and western walls. This central unit had two more enclosure walls at successive distances, which too were provided with circular towers at the corners and semi-circular towers along the walls (Fig. 4). Besides the foregoing temples, at Gonur (also in Margiana) has been brought to light a massive architectural complex of secular nature. Called variously as ‘Citadel’ or ‘Kremlin’, it measures 120 x 115 m. on plan and is enclosed by a fortification wall, having on the exterior a rectangular tower at each corner and four rectangular towers on each side. Here is the plan of the ‘Kremlin’ which incorporates within its premises the king’s palace, audience hall, administrative blocks, garrison complex, etc. (Fig. 5). 32

LET NOT THE 19TH CENTURY PARADIGMS CONTINUE TO HAUNT US ! The Ceramics

Sculptural Art

Many of the BMAC settlements have yielded remains of pottery kilns, having a lower ‘fire-chamber’ and an upper ‘baking chamber’. The pot-forms include, besides others, large carafes with stretched narrow necks, vases with long slim stems and ‘tea-pots’ with spouts. The spouts themselves show a great variety, which includes ‘long tubular’ spouts, ‘bridge’ spouts and ‘trough or channel’ spouts (Fig. 6).

Though we have already dealt with objects of stone and metal and with ceramics, we discuss here separately the art of sculpting in stone, metal and earthenware, since these objects throw valuable light on the artistic acumen of the BMAC people and demonstrate how very individualistic was their art-style. Here is a stone sculpture of a seated lady from Bactria (Fig. 12). In order to bring out the contrast in the portrayal, the sculptor has chosen a blackish stone for the dress with which the major part of the body is covered but a pinkish white one for the head and hands. Also to be noted are not only the herring-bone weave (decoration on) of the garment, but also the details in depicting the hair-style.

Besides these distinctive spouted vessels, there is yet another category which calls for special attention. It is a basin or bowl having a frieze of small animals (sometimes humans as well) running along the rim. Also to be noted are serpents crawling up along the walls of the pot, both on the exterior as well as interior (Fig. 7). Although there is no conclusive contextual evidence to establish their ‘ritual association’, it has been surmised that these bowls may still have had such a use, primarily because the animal-cum-human frieze on the rims would render the bowls rather uncomfortable for the lips to directly drink a liquid from the bowls.

It is not just the human beings that attracted the attention of the BMAC artist, but the animal-world as well. And here are two objects, both from Bactria, one depicting a feline (Fig. 13) and the other a goat (Fig. 14). The goldleaf and inlay-work in the case of the former is indeed superb and so also is the choice of coloured stones in the case of the goat, namely lapis lazuli for the horns, eyes and beard and limestone for the rest of the body. Though termed as ‘the poor man’s medium’ of art-expression, the earthen sculptures are no less remarkable. Take for example, the following figurine from Gonur Depe necropolis in Margiana. It has a slender neck, aquiline nose, prominent eyes, arched eyebrows, receding forehead and a very distinctive head-dress (Fig. 15).

Stone Objects There is a great variety of stone objects in the BMAC: for example, beads and pendants of carnelian, lapis lazuli and turquoise, carved vessels of steatite, ‘diminutive columns’ of multi-coloured stones and seals and amulets of steatite. The illustration that follows shows a cup-onstand and a deep inward-tapering bowl of black steatite.

Seals and Amulets

Both these bear etched geometrical designs on the exterior (Fig. 8). The ‘diminutive columns’, also known as ‘dainty columns’ or ‘dainty pillars’, are generally less than half-metre in height, highly polished and sometimes provided with a vertical groove. While their exact use is debated, it has been suggested by certain scholars that these were used in some kind of ritual. Here is what these look like (Fig. 9).

Seals and amulets are a very significant constituent of the Bactria-Margiana Archaeological Complex, since the individual motifs as well as the narrative scenes depicted thereon throw valuable light on the religious beliefs and practices of the people. The seals were made of metals, such as copper/bronze and silver, while the amulets were usually of stone, mostly black steatite. These latter usually show, amongst other motifs, snakes, scorpions, eagles, two-humped (typically Bactrian) camels, felines, etc. The snake seems to be such a favourite that it is depicted on the ceramic ‘rituals bowls’ as well (already referred to). The metal seals are to be noted, besides other motifs, for geometric ones. But most spectacular are the narratives on the cylindrical seals. We propose to illustrate these later when their narratives will also be discussed.

Metal Objects The BMAC settlements, including graves, have yielded a large number of metal objects, mostly of bronze/copper but sometimes of silver and gold too. Here we illustrate two very distinctive axes. On one of these there sits a human figure at the butt-end, whereas in the other case that position is occupied by an animal (Fig. 10). However, much more remarkable is an axe, made of silver but covered with gold lamina. The decoration showing a winged feline and heads of two eagles would appear to have had some mythological import (Fig. 11). It also seems most likely that this axe was not used for cutting wood or for a similar purpose but may have been ceremonial in nature – perhaps mounted on a specially made staff and held by a person in authority, as a mark of his position.

The Chronological Horizon A good deal of controversy surrounds the origin of the Bactria-Margiana Archaeological Complex, namely whether it was a local development from Namazga V or it was born out of an external impetus. Be that as it may, Carbon-14 dates indicate that Period 1 at Gonur depe may have commenced around 2100 BCE and continued up to 1900 BCE, while Period 2, which is what is known as the BMAC proper, may be ascribed to circa 2000-1700 BCE. 33

B. B. Lal Further, the overall cultural ethos, including the distinctive pottery, of the Gangetic Copper Hoards is totally different from that of the Bactraia-Margiana Archaeological Complex and that the former cannot be derived from the latter. But more strange is the argument that the occurrence of a single antennae-hilted sword in Bactria would entitle that region to be the ‘motherland’ of the Gangetic Copper Hoard people who produced these copper weapons and other associated objects in hundreds, if not thousands. If this logic is stretched further, I will not be surprised if one day Parpola comes out with the thesis that the Harappan Civilization too originated in Margiana, because in that region (at Gonur) has been found one steatite seal bearing typical Harappan inscription and motif (Fig. 18), unmindful of the fact that such seals constitute an unalienable part of the Harappan Civilization.

An Analysis of the Various Views In the earlier part of this address we had stated that we shall examine the viewpoints of Professors Romila Thapar and R. S. Sharma on the one hand and of Professors Asko Parpola and Viktor Sarianidi on the other. Thus, we begin with the views of the first two scholars. Views of Romila Thapar and R. S. Sharma Having failed to establish an ‘Aryan Invasion’ of India, Professor Thapar comes out (1989-91: 259-260) with a new theory, viz.: “If invasion is discarded then the mechanism of migration and occasional contacts come into sharper focus. The migrations appear to have been of pastoral cattle-breeders who are prominent in the Avesta and R.igveda”.

If, following the footsteps of Parpola, I were to say that the find of the well known seal of the ‘Persian Gulf’ style at Lothal in Gujarat establishes that the Persian Gulf Culture (which abounds in such seals) originated in Gujarat or, again, if I said that the occurrence of a cylinder seal at Kalibangan in Rajasthan entitles Rajasthan to be the ‘motherland’ of the Mesopotamian Culture (wherein cylinder seals are found in large numbers), I am sure my learned colleagues present here would at once get me admitted to the nearest lunatic asylum. One finds yet another amusing example of a similar kind of unbridled imagination when Parpola calls the ground-plan of the palace at Dashly-3, datable to circa 2000 BCE, “the prototype of the later Tantric man.d.alas/yantras”. He then goes on to add: “That the religion of the Dasas [who are mentioned in the R. igveda and whom he identifies with the Bactria-Margiana people] was an early form of Sыaktism is also suggested by the ground plan of the palace of Dashly-3 in Bactria closely agreeing with the later Tantric man.d.ala..” (ibid.: 52). For the sake of unambiguity, I reproduce now the drawings of the Dashly-3 Palace and the Mahakali yantra (Fig.19), as published by Parpola himself (ibid.: 62), and leave it to the learned scholars to decide whether they too would like to accompany Parpola in crossing this 4000year-old and 4000-kilometre-long bridge along with Parpola. Must we really indulge in such a kite-flying just to support our preconceived notions?

Following faithfully the footsteps of Thapar and amplifying her stand, Professor Sharma avers (1999: 77): “the pastoralists who moved to the Indian borderland came from Bactria-Margiana Archaeological Complex or BMAC which saw the genesis of the culture of the Ŗigveda”. Both Thapar and Sharma are even now labouring under the 19th century belief that the Vedic Aryans were nomads. But have they even once cast a glance at the make-up of the Bactria-Margiana Archaeological Complex. As would have been absolutely clear by now, the BMAC is a fully developed civilization with all the trappings of urbanism. How can then Thapar and Sharma devalue the Bactria-Margiana people and call them ‘pastoral cattle-breeders’? Just to fit into their preconceived notion that the Ŗigvedic Aryans were ‘nomads’? Views of Asko Parpola In his paper, ‘Margiana and the Aryan Problem’, Asko Parpola states (1993: 47): “These excavations at Mehrgarh, Sibri, Nausharo and Quetta have conclusively shown that immigrants bringing with them an entire new cultural complex have settled in Baluchistan, with close parallels in Gurgan, south Turkmenistan, Margiana and Bactria of the Namazga V-VI period”. Whereas certain parallels between the Quetta-Sibri finds and those from the Bactria-Margiana regions are acceptable, one is really baffled by the succeeding statement of Parpola, namely: “A newly found antennae-hilted sword from Bactria paralleling those from Fatehgarh suggests that this same wave of immigrants may also have introduced the Gangetic Copper Hoards into India”. (Fig.16).

Views of V.I. Sarianidi It has been claimed by certain scholars, including Sarianidi, that the BMAC people were the forebears of the Iranians and Indo-Aryans. This conclusion seems to have been drawn primarily on the following four counts: viz.

I am sure Parpola is aware of the fact that the Copper Hoards of the Gangetic Valley, as would be seen from the illustration that follows (Fig. 17), include many other very distinctive types, such as anthropomorphic figures, harpoons, shouldered axes, etc. which have never been found in Bactria.

1. 2. 3. 4.

34

fire-worship temples of BMAC; supposed use of soma/homa in BMAC rituals; mistaken identity of a horse’s skeleton as evidence of the aśvamedha; and cult motifs on the BMAC glyptics.

LET NOT THE 19TH CENTURY PARADIGMS CONTINUE TO HAUNT US ! samples from the site [forwarded by Dr. Fred Hiebert of Harvard University], which were subjected to pollen analysis at the Department of Botany, University of Helsinki […] The largest amount of pollen was found in the bone tube (used for imbibing liquid?) from Gonur 1, but even in this sample, which had been preserved in a comparatively sheltered position when compared with the other investigated samples, only pollen of the family Caryophyllaceae was present. No pollen from ephedras or poppies was found and even the pollen left in the samples showed clear traces of deterioration (typical in ancient pollen having been preserved in a dry environment in contact with oxygen). Our pollen analysis was carefully checked for any methodological errors, but no inaccuracies were found”.

We shall now examine, though briefly, the evidence in respect of these four claims. Fire Worship It has been argued that the BMAC temples were devoted to fire-worship and that since this kind of worship constitutes the main religious base of the Zoroastrians, they and the BMAC people shared a common ancestry. This is what Sarianidi says (1993a: 679) in this context: “A building unearthed at Gonur temenos and conveniently called a fort has a cross-shaped general plan with twelve corner towers. It most closely resembles the outer contours of the indisputable fire temple at Tepe Nush-i-Jan. A difference of minor importance only consists of the fact that while the towers are round at Gonur, they are square at Tepe Nush-i-Jan” (Fig. 20). But he immediately admits the weakness of the comparison by stating: “Unfortunately, the fort (at Gonur temenos) appears to be unfinished so we do not know its inner construction, but apparently it too was a fire temple”. What is the great point in forcing such a comparison when the evidence itself so weak?

The Aśvamedha The discovery, in the cemetery area at Gonur, of the skeleton of a horse, with its head missing, has led Sarianidi to postulate that it is a case of the Aśvamedha (horse-sacrifice); and since the Aśvamedha is a ritual mentioned in the Ŗigveda, he argues that the authors of the Bactria-Margiana-Archaeological-Complex must have been the ancestors of the Ŗigvedic Aryans. Before we examine the validity of such a conclusion, let us have a look at the photograph of the skeleton, published by Sarianidi himself (Fig. 22). It would be seen from the photograph that there is no clear outline of a pit in which the horse is supposed to have been regularly buried. Further, the skeleton lies hardly a few centimeters below the ground-level. Thus, there could have been many other reasons for the head to be missing, such as erosion through natural agencies or subsequent human interference. And no less important is the fact that the skeletal remains do not conform to the manner in which the horse had to be sacrificed in the Aśvamedha. There are two Sūktas in the Ŗigveda, viz. 1.162 and 1.163, which are devoted to the Aśvamedha and lay down how the horse, tied with ropes and accompanied by a goat, had to be taken to the sacrificial altar and then, after some rituals, had to be sacrificed.

In order to show that the Indian subcontinent was also involved in such a fire-worship, Sarianidi adds (ibid.: 676): “An example is the fire temple in the lower town of Mohenjo-daro; its basic plan comprises a ‘courtyard encompassed by corridors’ (Dhavalikar and Atre 1989)” (Fig. 21). In their article, Dhavalikar and Atre have gone into a series of conjectures, there being no down-to-earth evidence for actual fire worship in that complex. On the contrary, Marshall (1931, Vol. I: 202) takes this structure to have been a normal residential house. But in his desperation to bring India as well into the orbit of a Zoarastrian kind of fire worship, Sarianidi has completely lost sight of the fact that the house-complex at Mohenjodaro belongs to the Indus Civilization of the 3 rd millennium BCE whereas the BMAC is much later, dated to the 2nd millennium BCE. Thus, if at all the comparison made by Sarianidi is accepted, the direction of movement will have to be from India to Central Asia and not viceversa. The Soma/Homa

I quote below two of the verses, viz. 1.162.18 and 1.162.19, which are very relevant here. “The axe penetrates the thirty--four ribs of the swift horse, the beloved of the gods, (the immolators), cut up (the horse) with skill, so that the limbs may be unperforated and recapitulating joint by joint.. (18) There is one immolator of the radiant horse, which is Time: that are two that hold him fast: such of thy limbs as I cut up in due season. I offer them, made into balls (of meat), upon the fire. (19) The foregoing description makes it abundantly clear that in the case of the Asыvamedha the horse had to be cut up into parts, ‘recapitulating joint by joint’ .. There is hardly any evidence of such a cutting up in the case of the horse’s skeleton discovered at Gonur. Then why be so imaginative as to call this skeleton as evidence of the Aśvamedha and thereby draw an uncalled for conclusion

It has been claimed that there occurred the remains of ephedra and poppy in the temple of Togolok-21 in Margiana and since ephedra has been thought to be identical with the soma/homa of the Ŗigveda/Avesta, the BMAC people must have been the ancestors of the IndoIranians. There are two snags in this thesis. In the first place, not all experts agree that soma/homa is nothing but ephedra. But what is more important is that Harri Nyberg, a well known authority on the subject, after a thorough examination of the Togolok evidence, writes (1995: 402): “remains of ephedras have also been reported from the temple-fortress complex of Togolok 21 in the Merv oasis (ancient Margiana – Parpola 1988; Meier-Melikyan 1990) along with the remains of poppies. […] In 1990 I received some 35

B. B. Lal that the BMAC people were the ancestors of the Ŗigvedic Aryans?

Finally, the most crucial aspect of the issue The following three maps (Figs. 26, 27 & 28), not drawn by me but published by Sarianidi himself (1993b: Figs. 2, 3 and 5) relate to the spatial distribution respectively of ‘the motif of man-bird with hit animals’, ‘the motif of acrobats jumping over bulls’, and ‘miniature columns and cult vessels with depiction of snakes and animals on the rim’.

Motifs on BMAC glyptics The fourth argument that has been pressed into the service of the supposed BMAC=Aryan equation is that the motifs on the BMAC seals compare with certain motifs on the Syro-Hittite glyptics and since there occur on some Boghaz Qui tablets the names of Vedic deities, viz. Indra, Mitra, Varuņa and Nāsatya, the Boghaz Qui Aryans must be at the root of BMAC ethnic make-up.

A careful look at these maps would make it abundantly clear that the glyptic motifs shown on the first two maps occur from the Bactria-Margiana region on the east to the Syro-Hittite region on the west but do not travel southwards in the direction of Afghanistan or Baluchistan. It is only the miniature columns and bowls bearing on their rim animal-and-snake motifs that find their way into Baluchistan. But in no case did any of the above-noted motifs, columns or snake-decorated bowls find their way east of the Indus up to the upper reaches of the Gangā-Yamunā doab which, as spelt out in the Nadīstuti Sūkta of the Ŗigveda itself (10.75.5-6), was the region occupied by the Ŗigvedic Aryans.

To quote Sarianidi (op.cit.: 677): “Since it is Mitanni texts that contain the oldest mention of Aryan deities, there cannot be any doubt about the connection of the Mitanni empire with the so-called Aryan problem. As the replication of Mitanni art in Bactria and Margiana is clearly not coincidental, we are justified in connecting the tribes migrating into Central Asia and the Indus Valley with the settlement process of the Aryan or Indo-Iranian tribes”. Elsewhere Sarianidi goes into the details of these Syro-Hittite vis-à-vis Bactria-Margiana glyptic parallels. For example, he states (1993b:12-13):

The only exceptions to the foregoing distribution-pattern are some seals/seal-impressions from Chanhu-daro and Gilund (Possehl 2004: figs. 7 and 15). Sometimes, a double-spiral-headed copper pin from Chanhu-daro (Mackay 1976, reprint, p. 195, pl. LXVIII, 9) and a twoanimal-headed antimony stopper-rod, also of copper, from Harappa (Vats 1974, reprint, p. 390, pl. CXXV, 36) are also brought into the discussion. But let it be remembered that all these are only peripheral to the most characteristic and core-items of the Bactria-Margiana Archaeological Complex. Anyway, it would be simply ridiculous to ascribe these few objects to a migration of the BMAC people (cf. Gupta 2006, under print). Have we not in the past explained the occurrence of some items of a given culture-complex in another complex by means of trade/exchange/casual gift or a similar mechanism: for example, the occurrence of Harappan seals, etched carnelian beads, etc. in Mesopotamia, Iran and even Central Asia by trade and not by migration of the Harappan population? Then why invoke the migration of the BMAC people to explain the presence of some seals/seal-impressions, etc. at stray Indian sites?

In this connection worthy of utmost attention is the impression of a cylinder seal on one of the Margianian vessels, found... at Gonur. The central figure of a frequently repeated frieze composition is a standing nude anthropomorphic winged deity with an avian head holding two mountain goats by the legs... Such anthropomorphic winged and avian-headed deities are represented fairly fully in the glyptics and on the seals of Bactria. These Bactrian images find the most impressive correspondence in Syro-Hittite glyptics. If the fact that it’s for the Mittani kingdom that the names of Aryan deities are evidenced is taken into account the importance of the Bactrian-Margianian images will become obvious in the light of solving the Aryan problem on the basis of new archaeological data. While one has little hesitation in accepting the above-noted SyroHittite vis-a-vis Bactia-Margiana parallels, what indeed is the basis of connecting these motifs with the Aryan gods, viz. Indra, Mitra, Varuņa and Nāsatya? (cf. Fig. 23).

In the context of the debate whether the R.igvedic people were indigenous or invaders/immigrants from outside, the evidence of two sister disciplines, namely human biology and human genetics, must also be brought into the picture. After a thorough examination of the relevant human skeletons, Hemphill and his colleagues (1991) categorically pronounced: “As for the question of biological continuity within the Indus Valley, two discontinuities appear to exist. The first occurs between 6000 and 4500 BC... and the second occurs at some point after 800 BC.” In other words, there was no entry of a new set of people between 4500 and 800 BCE, much less of Aryan invaders/immigrants. In recent years a great deal of genetic research has been carried out which too throws valuable light on this issue; and I quote here

Does Sarianidi think that the ‘standing nude anthropomorphic winged deity with avian head’ and holding animals by their tails in each hand represent one of the above-mentioned Vedic gods Indra, Mitra, Varun.a, Na -satya? Likewise, what precisely Aryan is there in the following narrative portrayed on another cylinder seal? (Fig. 24). Perhaps one fine morning someone might be tempted to designate the scene depicted on the next seal as “The offering of Soma to Indra”, where Indra is the central figure seated on a chair and his devotees are offering the soma in cups, the beverage itself being stored in the jar. (Fig. 25). 36

LET NOT THE 19TH CENTURY PARADIGMS CONTINUE TO HAUNT US ! Sanghamitra Sahoo et al. (2006: 843-48): “The sharing of some Y-chromosomal haplogroups between Indian and Central Asian populations is most parsimoniously explained by a deep, common ancestry between the two regions, with the diffusion of some Indian-specific lineages northward. The Y-chromosomal data consistently suggest a largely South Asian origin for Indian caste communities and therefore argue against any major influx, from regions north and west of India, of people associated either with the development of agriculture or the spread of the Indo-Aryan language family”.

Mackay, E. J. H. (1976), reprint. Chanhu-daro Excavations 1935-36. Delhi: Bhartiya Publishing House. Marshall, J. (1931). Mohenjo-daro and the Indus Civilization. 3 vols. London: Arthur Probsthain. Muller, F. Max. (1890), reprint 1979. Physical Religion. New Delhi: Asian Educational Services. Nyberg, H. (1995) ‘The Problem of the Aryans and the Soma: The Botanical Evidence’. In Erdosy, G. (ed.) The Indo-Aryans of Ancient South Asia. Berlin: Walter de Gruyter: 382-406. Parpola, A. (1993) ‘Margiana and the Aryan Problem’. IASCCA Information Bulletin, 19. Nauka: 41-62. Possehl, G. L. et al. (2004) ‘The Ahar-Banas Complex and the BMAC’, Man and Environment, 29 (2): 18-29. Renfrew, C. (1988) Archaeology and Language. New York: Cambridge University Press. Sahoo, Sanghamitra, et al. (2006) ‘A Prehistory of Indian Y Chromosomes: Evaluating Demic Diffusion Scenarios’. PNAS,103 (4): 843-848. Sarianidi, V. I. (1993 a) ‘Margiana and the Indo-Iranian World’. South Asian Archaeology, Vol. II: 667680. Sarianidi, V. I. (1993 b) ‘Margiana in the Ancient Orient’. IASCCA Information Bulletin, 19. Nauka: 5-28. Sarianidi, V. I. (2002) Margush: Ancient Oriental Kingdom in the Old Delta of the Murghab River. Ashgabat. Shaffer, J. G. and Lichtenstein, D. (1999) ‘Migration, Philology and South Asian Archaeology’. In Bronkhorst, J. and Deshpande, M. M. (eds.) Aryan and Non-Aryan in South Asia: Evidence, Interpretation and Ideology. Cambridge: Harvard University. Sharma, R. S. (1999) Advent of the Aryans in India. New Delhi: Manohar Publishers. Thapar, R. (1988-91). In Journal Asiatic Society of Bombay, Vol. 64-66: 259-260. Vats, M. S. (1974), reprint. Excavations at Harappa. Delhi: Bhartiya Publishing House. Wheeler, R. E. M. (1947) ‘Harappa 1946: The Defences and Cemetery R 37’. Ancient India, 3: 58-130. Witzel, M. (1995) ‘R.igvedic history: poets, chieftains and polities’. In Erdosy, G. (ed.) The Indo-Aryans of Ancient South Asia. Berlin: Walter de Gruyter: 307-352.

Scholars have already abandoned (though after much dithering) the ‘Aryan Invasion’ theory. Is it not high time to re-think and shelve the newly hugged-to-the-chest ‘Bactria-Margiana Immigration’ thesis as well? At the end, it needs to be emphasized that the purpose of this address was not to criticize Professor X or Professor Y. Far from it. The whole emphasis has been on demonstrating how the 19th century paradigms are still dominating our thinking, thereby producing a very blurred vision of South Asia’s past. Can’t we begin thinking afresh in this 21st century?

Bibliographical References Dales, G. F. (1964) ‘The Mythical Massacre at Mohenjodaro’. Expedition, 6 (3): 36-43. Danino, M. (2006) L’Inde et l’invasion de nulle part. Paris: Le Belles Lettres. Dhavalikar, M. K. and Shubhangana Atre (1989) ‘The Fire Cult and Virgin Sacrifice: Some Harappan Rituals’. In Kenoyer, J. M. (ed.) Old Problems and New Perspectives in the Archaeology of South Asia, Wisconsin Archaeological Reports, Vol. 2: 193-205. Gupta, S. P. (under print) ‘Did the BMAC ever cross the Indus?’ Paper presented at an International conference held at Vadodara in 2005. Hemphill, B. E. et al. (1991) ‘Biological Adaptations and Affinities of Bronze Age Harappans’. In Meadow, R. H. (ed.) Harappa Excavations 1986-1990. Madison, Wisconsin: Prehistory Press: 137-182. Kenoyer, J. M. (1998) Ancient Cities of the Indus Civilization. Karachi: Oxford University Press and American Institute of Pakistan Studies. Lal, B. B. (2002) The Sarasvati Flows On: The Continuity of Indian Culture. New Delhi: Aryan Books International. Lal, B. B. (2005). The Homeland of the Aryans: Evidence of Rigvedic Flora and Fauna & Archaeology. New Delhi: Aryan Books International. Ligabue, G. and Salvatori, S. (1989) Bactria: an Ancient Civilization from the Sands of Afghanistan. Venezia. 37

B. B. Lal

Fig. 1 Migration of people as mentioned in the Baudhāyana Śrautasūtra.

Fig. 2 – Region of the Bactria-Margiana Archaeological Complex.

38

LET NOT THE 19TH CENTURY PARADIGMS CONTINUE TO HAUNT US !

Fig. 3 – Temple at Dashly-3, Bactria.

Fig. 4 – Temple at Togolok-21, Margiana.

Fig. 5 – Citadel at Gonur, Margiana.

39

B. B. Lal

Fig. 6 – Pottery vessels from Margiana.

Fig. 8 – Carved stone vessels from Gonur-depe, Margiana.

Fig. 7 – Cult vessel from Togolok-1 Temple.

Fig. 9 – Stone ‘dainty columns’ from the temple, Togolok-21.

40

LET NOT THE 19TH CENTURY PARADIGMS CONTINUE TO HAUNT US !

Fig. 10 – Bronze axes from Bactria.

Fig. 12 – Composite stone figurine from Bactria.

Fig. 11 – Silver ceremonial axe with parts in gold lamina, Bactria.

Fig. 13 – Chlorite and gold leaf representation of a feline, with semi-precious stone inlay.

Fig. 14 – Limestone goat with horns, eyes and beard in lapis lazuli.

Fig. 15 – Terracotta figurine from Gonur-depe necropolis.

41

B. B. Lal

Fig. 16 – Antennae-hilted swords of copper: from Bactria (a) and from Fatehgarh (b).

Fig. 17 – Copper-hoards from the Gangetic valley, India.

42

LET NOT THE 19TH CENTURY PARADIGMS CONTINUE TO HAUNT US !

Fig. 18 – Harappan seal from Gonur and its impression.

Fig. 19 – Plan of palace at D ash ly-3 (left); Plan of Mahakali yantra (right).

Fig. 20 – Mohenjo-Daro: “Fire Temple” ?

Fig. 21 – Fort at Gonur (a) and Temple at Nush-i-Jan (b).

43

B. B. Lal

Fig. 22 – Burial of a horse in Gonur-depe.

Fig. 23 – Impression of cylinder seal from Gonur-1.

Fig. 24 – Cylinder seal from Togolok-21 and its impression.

Fig. 25 – Scene on a cosmetic flacon, Bactria.

44

LET NOT THE 19TH CENTURY PARADIGMS CONTINUE TO HAUNT US !

Fig. 26 – Spread of the motif of man-bird with hit animals.

Fig. 27 – Spread of the motif of acrobats jumping over bulls.

Fig. 28 – Spread of ‘miniature columns’ and their probable prototypes in the Syro-Hittite world.

45

OPENING ADDRESS OF THE 19TH INTERNATIONAL CONFERENCE OF THE EUROPEAN ASSOCIATION OF SOUTH ASIAN ARCHAEOLOGY RAVENNA, ITALY, 2-6 JULY 2007

REFLECTIONS ON “DOTS ON A MAP” Carl C. Lamberg-Karlovsky

the southwestern and northwestern regions. Central Asia and the Arabian Peninsula were separate constellations little known, veritable regions of terra nullius. The first stage in the archaeological (cultural?) integration of larger regions involved specific archaeological projects, initially dedicated entirely to the documentation of their regional significance. These archaeological excavations provided the necessary database to move to the second stage: the integration of a regional interaction of distinctive sites. The first stage involved the excavations at Susa by Jean Perrot, at Shahr-i Sokhta by Maurizio Tosi, at Tal-I Iblis by Joe Caldwell, at Bampur by Beatrice de Cardi, at Hili by Karen Frifelt and Serge Cleuziou, at Ra’s-al Jinz by Maurizio Tosi, at Tepe Yahya by Lamberg-Karlovsky and at Mohenjo-Daro by Michael Jansen. By the late 1970s Victor Sarianidi, I. N. Masimov and Ahmed Ali Askarov were writing a new chapter for the Bronze Age of Central Asia. All were involved in the discovery and excavation of the Bactrian Margiana Archaeological Complex (BMAC, aka the Oxus Civilization). Some of the above archaeologists were in contact with each other in a collaborative sense. By no means were all involved in a singular effort to relate the sites and regions to each other. Most progressed as before, namely: “My site-my world: a self-sufficient reality”.

In today’s world much is made about the respect for ‘cultural diversity’ and the importance of ethnic distinction and pride while, on the other hand, it is said we live in the benefits of a homogenizing ‘global village’. This apparent paradox, if not dialectical contradiction, is replayed in the manner we view the third millennium land mass ranging from Mesopotamia to the Indus Valley and from Central Asia to the Arabian Peninsula. On the one hand, we hold that each archaeological site, region, and interaction sphere, is to be appreciated for its uniquely distinctive nature while, on the other hand, the so-called ‘Middle Asian Interaction Sphere’, was a Bronze Age global village, an integrated whole. Distributed throughout this vast landscape are shared resources, of limited distribution, i.e. lapis, carnelian, turquoise, tin; shared aspects of technology from stone drill heads, etched carnelian beads, to techniques of ceramic and metal production; as well as shared iconographic elements depicted on steatite vessel and cylinder seals. We have been inordinately successful in documenting the wide ranging commerce and cultural interaction that united this region. Yet we despair in understanding the structure of the commerce that integrated the region. In fact, there is little effort even made to address this issue. Most of the items we document as evidence for the interaction are elite items. Does this suggest that what we view as trade and commerce is, in fact, gift exchange between the elites? And what then of markets? To what extent was there entrepreneurial trade directed by merchants? And were they private or state? Sponsored? Within a market economy? Such questions are never addressed in discussing the MAIS. The emphasis is upon placing ‘dots on the map’, supposedly documenting the presence of something foreign, coming from a specific region, thereby indicating an ‘interaction’ between site A and site B, or region C and D.

It is important to place into context two unfolding phenomena of the early to mid 1970s. First, the Uruk Expansion was a dominant concern in the archaeology of Mesopotamia. Excavations at Habuba Khabira, Godin Tepe and Susa were but three of the early sites in the 1970s that documented the unexpected and wide scale spread of the Uruk culture. Later debate implicated the Uruk Culture as involved in colonialist and imperialist expansion, dedicated to the control of commerce and trade routes. Even more surprising was the discovery of a slightly later Proto-Elamite Expansion that imitated the Uruk Expansion. The first evidence for this Proto-Elamite expansion came in 1971 with the completely unexpected discovery of Proto-Elamite tablets at Tepe Yahya. Tepe Yahya was located more than 1000 kilometers to the East of where a consensus believed the origin of the ProtoElamite Culture originated, i.e. Susa (Khuzistan). The first of 17 tablets was discovered in a trench being excavated by Martha Prickett. The discovery of the Proto-Elamite settlement at Tepe Yahya was followed by Proto-Elamite tablet(s) recovered (once again) from Susa,

There is little ‘social archaeology’ in our discussions of MAIS. Perhaps, the materialist emphasis is to be expected, for only one generation has been involved in its documentation and the very concept of the MAIS did not exist a decade ago. We may cite the birth of MAIS, although still unnamed, in the 1960s and 70s and align it within the context of specific research projects. In the 1960s little was known of the archaeology of the Iranian Plateau and almost all that was known was centered in 47

Carl C. Lamberg-Karlovsky surveyed entire regions. Even more dots can be placed on the map relative to ceramics, metals, cylinder seals, specific resources, styles, and artifacts discovered as intrusive on distant sites. It suffices to write a stand-alone article on the style of a specific flower recovered from the UAE and compare it to its point of origin: the BMAC. The most distant parallels are favored while its more likely derivation from nearby SE Iran, where such items have also been recovered, is overlooked. Connectivity and interaction over the most distant points are favored in demonstrating the vastness of the interactive network. This distorts the fact that like politics most economic relations are mostly local!

and from Choga Mami, Shahri- Sokhta, Tal-i Malyan and most recently at Tepe Sofali, expanding the Proto Elamite Expansion northward to the Tehran Plain. The connectivity of sites, the extent and degree of their interaction, was a new and unanticipated consequence of the 1970s. It needed to be addressed. In 1969 Maurizio Tosi stood in the doorway of my office and announced himself: “I am Maurizio Tosi and I am excavating Aratta”. It was the start of a fourty year relationship, at times volatile but always rewarding and productive. Maurizio was to visit our excavations at Tepe Yahya and put us into contact with a number of his Italian colleagues in a lasting and productive collaboration. We decided to collaborate on a synthetic article and to that end Maurizio came to Harvard. We spent several months crafting an article that related the two sites we were excavating and integrating them into their larger relations to the regions from Central Asia to the Arabian Peninsula and from Mesopotamia to the Indus Valley. The results hoped for was a descriptive and illustrated framework of “Interaction Spheres”. The concept was adopted from Joe Caldwell who also coined the term the “New Archaeology” the theoretical fashion of the 60s and 70s. As more data accumulated and the complexity of the extensive interaction that characterized this vast landscape unfolded Phil Kohl was later to include the Caucasus within the framework of interaction while escalating the nature of connectivity by being the first to adopt Wallerstein’s concept of “World Systems” and apply it to the third millennium of the Near East. Thus, first we had “interaction spheres”, then ‘World Systems’ and now ‘Middle Asian Interaction Sphere”. We are good at offering names for the dots on the map but not so good at offering an understanding of the social processes that placed the dots there.

Nevertheless, the material evidence behind the dots on the map makes for a most impressive continental scale of cultural interaction. A linkage of connectivity, that we may refer to as a ‘chaining effect’ linking distinctive interaction spheres, began in the mid third millennium and reached across the continent of Asia: each interaction sphere in contact with its neighbor! Thus, the cultures of the eastern shores of the Arabian Peninsula were in contact with both Mesopotamia and the Indus Valley (which, in turn, were in contact with each other); Mesopotamia and the Indus Valley were in contact with the diversity of cultures on the Iranian Plateau; while the Indus Valley and the cultures of the Iranian Plateau were in contact with the BMAC of Central Asia. The above constitutes the world of MAIS. Extending beyond the extensive connectivity of MAIS is the evidence for BMAC contacts with the varied cultures of the Eurasiatic steppes, i.e. represented by the diversity of the steppe cultures of the generic Andronovo Culture of the second millennium. The geographical reach of the steppe cultures extended to distant Xinjiang where there is further evidence for contacts with the heartland of China and the Erlitou Culture. Needless to say the material evidence for the connectivity of these interactions spheres is entirely variable. The evidence for cultural interaction among the distinctive archaeological cultures within MAIS is fairly robust. As we move eastward from the BMAC the evidence for connectivity and interaction among the sparsely settled pastoral nomads inhabiting the vast terrain of the Eurasiatic steppes becomes more ephemeral. The links within the ‘chaining effect’ are not bound as strongly and the material evidence is more ephemeral as one crosses the more than 10.000 km divide of the Eurasiatic steppes. Nevertheless, the recent evidence put forth by Karl Jettmar and Michael Frachetti even indicates that the enormity of the mountain massifs of the Pamirs, Hindu Kush and Himalaya were insufficient barriers in preventing cultural interaction in the late Bronze Age.

I continue to resist the merits of applying ‘World Systems’ to the connectivity that brought this vast area into a shared relationship and I continue to prefer the more neutral term ‘interaction spheres’. World Systems Theory (WST) as crafted by Immanuel Wallerstein insists on three assumptions. None of which, I have previously written, are characteristic of the Bronze Age of the Greater Near East: 1) The core dominates the periphery, whether by military means, organizational superiority, or ideological agency; 2) the core exploits the periphery by asymmetric trade; extracting valuable raw resources from the periphery while dumping their cheap goods on them; 3) the core dominates the political structures of the periphery by its dominance in the organization of trade and exchange. This essentialist perspective, a distinctly Marxian one, is demonstrably wrong when applied to the ancient Near East and perhaps to all pre-industrial societies. It does, however, have considerable merit when applied, as crafted by Immanuel Wallerstein, to the emergence of capitalism between 1450 and 1550. Today, when compared to the 1960s and 70s, we have traveled a great distance. Our archaeological data base and our theoretical perspective have been greatly expanded. We have placed a large number of dots on the map with regard to numerous new sites excavated and

It is time to ask the question what do all these dots mean. What inspires the interaction? What effects does interaction bring about? How does interaction differ in different regions and in time and space? What causes interaction to terminate? What is the nature of its agency? One might say that we have completed the agenda of “Processual” or the ‘New Archaeology’, namely, the scientific documentation and gathering of a data base 48

REFLECTIONS ON “DOTS ON A MAP” through systematic excavation and survey and its articulation with severable hypotheses, i.e. interaction spheres, world systems, colonialism, et al. We can now turn to ‘Post-processual Archaeology’. What do the dots on our maps mean? We need to address the political, economic, even ideological ramifications of this wide scale geographic interaction. Does the wide scale distribution of the ‘Intercultural Style’ mean that from Mari to Mohenjo-Daro there was a shared knowledge, and understanding of the meaning, of its iconography? The presence of BMAC materials on numerous sites of the Iranian Plateau and the Indus offer stark contrast to the fact that not a single BMAC object is to be found in Mesopotamia (while neighboring Susa has many BMAC artifacts). What are we to make of the fact that when compared to other ‘interaction spheres’ the Indus Valley has a poverty of foreign goods? While sites on the eastern shores of the Arabian Peninsula have artifacts, granted few in number, from every ‘interaction sphere’. The asymmetry of cultural interaction that is the variable distribution of dots on the map in space and time that characterized the different ‘interaction spheres’ are trying to tell us something. It is time we address the social, as well as the material evidence, in our body of knowledge. Lew Binford once told me that I had no right to be spending NSF funds on the excavation of an unknown region of Iran. He held to the belief that no meaningful hypothesis could be generated in a condition characterized by an inadequate data base. His approach to science was that hypotheses had to be formulated first and the gathering of a data base was to be “tested” against the hypotheses formulated. However, a new data base can generate new hypotheses and hypotheses can give new meaning to a data base. Both are viable approaches, neither is dominant. This was one of the seminal discoveries of the emergent scientific revolution of the 18th century Enlightenment. We have done well in gathering a descriptive data base and constructing the architecture of ‘interaction spheres’, ‘World Systems’ and ‘Middle Asian Interaction Spheres’ now it is time to address explanation by attempting to develop an understanding of the far more difficult social processes underlying the meaning of these dots on the map.

49

OPENING ADDRESS OF THE 19TH INTERNATIONAL CONFERENCE OF THE EUROPEAN ASSOCIATION OF SOUTH ASIAN ARCHAEOLOGY RAVENNA, ITALY, 2-6 JULY 2007

T-SHAPED PILLARS AND MESOLITHIC “CHIEFDOMS” IN THE PREHISTORY OF SOUTHERN EURASIA: A PRELIMINARY NOTE Massimo Vidale

sculptures oddly looking like wooden ‘totem-poles’ (totempfahl) of much later tribal contexts of the northern american continent. Nothing similar had been previously reported from any near-eastern or eurasiatic site.

The Urfa revolution First of all, I would like to point out how far the recent discoveries ⅲ‘’by Harald Hauptmann, Mehmet Özdogan, Klaus Schmidt and their collegues in the valley of Urfa and in south-eastern Turkey in general (Fig. 1) have changed our understanding of the beginning of food production and the Near east and southern Eurasia (Badisches Landesmuseum 2007; Özdogan and Basgelen 1999; Hauptmann, Schmidt 2007). The digs at Göbekli Tepe and Nevali Çori, exactly at the inflection point of the traditional geography of the so-called ‘Fertile Crescent’, have revealed a series of huge monumental sculptures designed and carved to be erected in series within special round or square buildings, rebuilt in the same places for centuries. These buildings were used between the late 10th and the 9th millennium BC. The sculptures were often broken and re-employed in the inner filling layers. Because of the presence of basements interpreted as altars or podia for statues, the presence of several fragments of larger-than size sculptures in the round, interiors painted white and red, and of stone benches running around the inner perimeters, these constructions at Nevali Çori and Göbekli Tepe are interpreted as cultic buildings. Remnants of similar constructions were identified at the nearby site of Çayonu as well.

The most common type of sculpture are T-shaped pillars. They end in rectangular or semicircular expansions; both at Nevali Çori and in some pillars at Göbekli, such horizontal arms of the ‘T’ were conceived by the carvers as representing human heads, as the pillars bear on the sides the representation of bent arms. Thus some Tshaped pillars are definetly anthropomorphs. In the round or elliptical buildings, perimetral pillars thuss look to the centre, i.e. to the two central pillars supporting the full weight of the roof. Rows of cup-marks appear exclusively on the upper surface of the ‘head’, out of sight from below, providing one of the few dating terms for this most enigmatic component of Eurasiatic rock art. Other pillars bear on the end animal figures, carved in bas-relief on the sides or fully and massively sculpted in the round. Sometimes the animal sculptures or their fragments are highly stylized and the protrayed species cannot be confidently identified. A clearly antropomorphic small-scale specimen from Kisilik and a simpler miniature pillar at Göbekli demonstrate that ‘T’shaped pillars were also reproduced in small-scale stone replicas, and this points to a widely acknowledged symbolic value of such megaliths also outside the original context of the buildings. Also the famous natufian bone sculpture from Mugharet el Uad (about 9000 BC), previously considered a ‘sickle handle’ might be the miniature replica of one of such pillars, crowned by an animal.

The largest sculpted monoliths (an unfinished pillar at Göbekli was almost 7 m high) were apparently used as pillars supporting elements for stone or wooden roofs. Wooden beams might have been placed on the perimetral ‘capitals’ and possibly converged towards couple of central pillars, the most visible architectural component of these buildings. Other T-shaped pillars are crowned by images in the round and evidently were not used to support the roof.

This almost unbelievable profusion of prehistoric iconography was evidently deeply rooted in contemporary material culture: from the sites of Tell Qaramel, Jerf el Ahmar, Körtik Tepe (10.000-9000 AC) and from contemporary levels of Göbekli come highly elaborated carved chlorite vessels and stone tablets with complex incised geometric and animal patterns closely resembling seals or painting stamps. Some carved vessels are comparable with simpler marble pieces previously found at south-west in contemporary natufian contexts. Others are simply astonishing, like a chlorite vessel from Körtik Tepe with a row of scorpions alternating with

Pillars represent, both in highly schematic features or in less artificial conventional traits, human figures, isolated or variously combined with animal forms and icones (vulture and other birds, snake, jackal or fox, lion or panther, bull, wildboar, scorpion, but also sheep, onagers, reptiles, insects), often combined with abstract signs. Human figures and animals, in some instances, are combined together in interlocked three-dimensional stone 51

Massimo Vidale Megaliths might have been carved, transported and dedicated in the frame of a complex system of ritual obligations among various lineages competing for preeminence and political authority. Economically, such efforts should have been supported by substantial amounts of food and beverages stored and re-destributed in the occasion of collective meetings or celebrations. The key function of the ‘cultic’ buildings would have been “to reduce the factors of social conflict” (Frangipane 2000: 216) by the means of massive investments in virtual ritualized clashes, competing, at the same time, for prestige and the preferential access to vital economic resources.

snakes rising towards the edge (Badisches Landesmuseum 2007, cat. 150: 303), coming 6000-7000 years before the same linked images or symbols on chlorite vessels of the Iranian plateau. By the early ceramic neolithic, around 7500 BC, baskets and unbaked clay containers, carved wooden vessels (like those found at Chatal Höyük), elegantly decorated and plain vessels carved in limestone and chlorite, the earliest pottery and lime plaster vases (in ‘white ware’) could be used by the Levant communities to express their internal status and even subtle rank differences. Going back to the Urfa sites, besides vessels and carved tablets, several types of semiprecious stone beads, buttons and highly refined polished marble bangles are on record, hinting to the presence in the largest sites of a social élite sponsoring the trade of exotic raw materials and specialized craft production.

Sites such as Göbekli (where these buildings could be about 20) and Nevali Çori might have represented main central locations periodically visited by tribal groups of hunter-gatherers, and later by early agriculturalists and breeders, assuming in the course of time a leading position for exchange and barter of valuable goods. Their strong influence might be reflected in the similarities between the Urfa material culture and some natufian finds (incised stone vessels and miniature ‘T’-shaped pillars).

Mesolithic chiefdoms? Pre-state social organizations are notoriously a slippery ground of investigation, and the neo-evolutionary social typologies so far proposed as steps of ascending scales of complexity have been often criticized on several theoretical and factual grounds. Which type of social organization lies behind the early holocenic archaeological record of Urfa?

Amazingly, the economy of Göbekli (about 9500-8800 AC), according to Klaus Schmidt (2007), was entirely based upon hunting gazelles, aurochs, boars, wild goats, onagers, deer and foxes, and on the gathering of local wild cereals (wild einkorn being the main collected resource). There would be no evidence of domesticated plants or animals. That a non food-producing economy was capable of supporting such labour investements is highly significant. So far, no obvious large-scale storage facility comparable to those excavated at the later sites of Çayonu and Nevali Çori itself (grill buildings) was found at Göbekli. The abundance of fresh food in this region, at least at given seasonal intervals, should have been overwhelming. Such a ‘late palaeolitic welfare’, furthermore, might well explain the flourishing of monumental or defensive architecture at Jericho, the rise of lime technology in the Siro-palestinian coast (limeburning kiln in the Hayonim cave, ca. 10.500 BC) and its early use in floor plasters of elitarian buildings at Jericho and other sites of the Levant.

A careful attitude would imply to suspend our judgement, advocating a general lack of information and the possible presence of an hitherto unknown organizational model. But if we have to judge on the normative basis of the extant anthropological knowledge, this is the record of a system of powerful chiefdoms, rather then the result of more egualitarian, pan-tribal associations like those described by E. R. Service (1982: 22-27). Social inequality is revealed by the precious vessels and ornaments so far described. The effects of the struggle among chiefs and lineages for leadership, their capability of managing both non specialized and specialized craft manpower, and concentrating and redestributing resources to their own benefit (and to the benefit of their lineages) are in full sight. Even at a very superficial level, the only possible explanation for these culting buildings is that ‘T’-shaped pillars and the other megaliths represented ancestral heroes or animal totems, or, better, ritual associations among these fictitious entities. In cultic building B of Göbekli the 2 central pillars have the same fox bas-relief: the most logical conclusion (or the least unlikely hypothesis?) is that it was the cultic seat of the ‘fox’ lineage or clan, to whom the other lineages, let us imagine, paid homage by dedicating other pillars. The pillars with sequences of superimposed animal bas-reliefs found at Göbekli might symbolize marriages, ritual links or alliances among clans. Ultimately, the record of Urfa might match the nature of chiefdom-level organizations in terms of social inequality, central leadership (perhaps based upon religious authority and monopoly of the most important rituals), asymmetrical control of precious goods and exotic status symbols, management of collective manpower for monumental ritual constructions aimed at celebrating the ancestors of the chiefs.

At Nevali Çori the cultic buildings (8400-8100 BC) would attest similar concerns, organization, social practices and perhaps symbolism in the context of a well developed agricultural economy based upon einkorn, 2grained wheat, and legumes such as lentils, peas, vetch, and the gathering of pistachios, almonds and grapes and other wild species. Hunting provided most of the wild species traditionally exploited in the region, with a gradual decrease, in time, of gazelle. Goats and sheep were domesticated. Both sites were finally deserted before the end of PPNB, around 8000 BC. Thus, in such light, the ‘mesolithic’ chiefdoms of Göbekli seem to have evolved, apparently in similar forms, into neolithic organizations equally centered on the worship of mythical ancestors and public competition for supremacy within the tribes. 52

T-SHAPED PILLARS AND MESOLITHIC “CHIEFDOMS” IN THE PREHISTORY OF SOUTHERN EURASIA: A PRELIMINARY NOTE are made of a whitish crystalline limestone, quite uncommon in Swat.

The archaeological record of south-eastern Turkey thus reveals that highly ranked, hyerarchical societies interpreted the local introduction of food-producing. Agriculture, in this case, could have been adopted for purely political reasons: the purpose of increasing the economic potential of some clans while rivalling with their opponents in food redistribution and monumental carvings. Later, the sudden impact of agriculture condemned to its extinction such archaic forms of redistribution and its related values. The chiefs of Urfa might have discovered 10.000 before the scholars of the ‘Technology and Culture’ milieu (Staudenmaier 1988) that you cannot change technical systems without transforming society. As Alessandro Vanzetti pointed to me, the whole process implies a noticeable demographic growth at the beginning of the Holocene and its climatic optimum (you need crowds of hunters-gatherers to collect the required food and beverages and to erect the cult buildings), while agriculture itself, in turn, would have soon triggered a further population increase. In this context, competition for supremacy might have cohincided with fights for the control of highly productive land, an uncommon condition for huntergatherers societies.

Detailed observation shows that both were the upper part or ‘capitals’ of ‘T’-shaped pillars, crowning a shaft to be placed into the ground. The original shape of Panr 1 (Fig. 2) difficult to perceive because while the upper part is finished and undamaged, the opposite face shows a large, oblique face that probably cancelled a damaged surface. Its section is a kind of flattened hexagon. It has 2 main decorated sides. One has a flat surface with a double carved frieze made of 2 irregular chains of 7-8 carved lozenges, separated by an horizontal ridge bending down to the left. On the rear one sees the same pattern of 2 superimposed registers separated by a horizontal protruding ridge. The upper oneshows a grid including a double series of squares, ending to the right in a zig-zag line, while the lower register hosts winding line resembling a winding snake. Panr 2 (Fig. 3), when observed from above appears elongated, with 2 rounded extremities; the view from the short sides emphasizes an egg-like section. One extremity is thicker and shows 2 large round, shallow incised eyes. It obviously represents a animal head. The eyes are joined by a V-shaped design, departing at the rear of the head down to a ridge in relief (a ‘crest’?) running all along the back of the ‘animal’, perhaps ending in a tail. The opposite side is intensively damaged by a series of strong blows repeated from the back to the head. The muzzle of the animal is elongated downward but short, almost cylindrical, with a slightly rounded outer contour. It is underlined a ridge in low relief that runs horizontally and divides the flank of the animal in 2 parts. On the opposite flank, a horizontal line divides the body in 2 registers. There is a panel bearing a snake motif very similar to that observed Panr 1. Another important feature of this sculpture is a well-designed curve that underlines the body of the animal, kindly sloping from the rear towards the front part of the animal. It suggests an enlarged rib gage, or, possibly, the presence of extended front paws, in the hint of a dynamic posture. Perhaps the animal was represented as running.

At any rate, the seed of social inequality, in spite of the old marxist legends on a ‘primitive communism’ might have been planted much earlier. In spite of the levelling effect played by redistribution in chiefdom-level societies described by modern etnographers, social inequality would be straightly rooted in archaic forms of political competition, before the much later opportunity of asymmetrically managing stored foodstuffs by the means of administration (in the sense of Frangipane 2000). What was the role of previous epipalaeolithic bands in this evolution? And was it limited to south-eastern Turkey, or did it affect other regions of Eurasia as well? T-shaped pillars in Swat, NWFP, Pakistan Once stated that whatever we have discussed so far, in terms of archaeological record, is a unicum, a recent discovery (or re-discovery) in the valley of Swat (Pakistan) might be relevant to the discussion (Vidale 2006b). In 2002 Luca Olivieri pointed to my attention two large stone sculptures found in the early years of the Italian activities in Swat and presently kept at the Museum of Saidu Sharif (V 504 and V 503, respectively called in this paper Panr 1 and Panr 2 after the presumed findspot near Mingora). Olivieri was the first to notice the similarity between the Urfa and Swat megaliths, when at the beginning I had a conservative attitude. The two stones come from a secondary context: they had been reemployed within a recent masonry wall near a modern islamic cemetery near the Buddhist sacred area of Panr, in the hill north-east of the Mingora. The stones were totally different from anything seen in Swat or in the nearby areas. Nobody had any idea about them, and eventually they were forgotten in a dark corner of the Museum’s stores. In 2003, Olivieri and I decided to draft and document in detail the stones. Both measure about 80 cm in length, 40 cm in height and 20 cm in thickness, and

In term of tool marks, both stones were treated with a simple point whose extremity had a distal diameter of 5 mm or less. The impacts of this tool are clearly visible within the grooves carved in different parts of the two stones as well as on the general exterior. Flat surfaces, chipped and polished rather than cut with a saw, show faint polishing traces. Surfaces, otherwise, were shaped and finished by the means of the same tool used for carving the incised lines, with a pecking technique that gave to the sculptures a peculiar coarse appearance. Panr 1 and 2 were crafted with a very limited technical kit, possibly including only a point, a hammer and an abrading tool. There is no evidence whatsoever of flat chisels or any other tool that might safely be interpreted as metallic, nor any trace of drilling. I checked carefully the surfaces for any evidence of painting residues, but found none. 53

Massimo Vidale In the later centuries of PPNB, the archeological record points to a general conversion of the ancestors’ veneration in new forms: the detachment from the skeleton, manipulation and public show of the ‘resurrected’ skulls; and the making and exhibit of series of possible ancestors’ statues 1-1.5 m high (Rollefson 1984). In both cases, such rituals involved making and applying lime plasters on specific supports. Both statues and skulls, judging from their size and findspots in the premises of private dwellings and semi-public spaces, had a more restricted public. Remodelled skulls, at least in some cases, were mounted on small spirally woven mat bats, and displayed in temporary shrines among dwellings and domestic spaces. Lime plaster statues, on account of their extremely fine face details and clumpsy rendering of the rest of the body, were evidently meant for wearing scarfs or turban, precious cloths and valuable ornaments. Possibly they were displayed for short times near the houses of community chiefs or big men, or in semi-publi spaces such as courtyards and small squares, like very similar figures of ancestors (sometimes containing actual skeleton parts and skulls) were mounted near the door of chief’s houses in the New Hebrides and other Oceanian islands. Both lime plaster statues and remodelled skulls were fragile, and had a limited intentionality of duration. Both statues and skulls, after use, often seem to have been respectfully disposed by carefully burying them in pihoards, probably not far from their contexts of display. The recent find at Chatal Höyük of an exceptional burial where the deceased holds in his arms a red-painted remodelled skull adds another ritual model to the behavioural range previously known.

Could the stones from Swat belong to the same early holocenic chronological horizon of the Urfa megaliths, and to a similar context of representation? Admittedly, their secondary find contexts does not help. But the Swat specimens are similar in being, most probably, capitals of ‘T’-shaped pillars, and having the same size; the pecked surface is identical; Panr 2 would share with some of the scultures in the round from Göbekli a complete animal figure (perhaps in running position) on the top, and the emphasis on snake images. Even the lozenge-net pattern in Panr 2 might be a simplified rendering of the impressive ‘woven snakes carpet’ that appears on the sides of a ‘T’-shaped pillar at Göbekli. The 2 stones, and particularly Panr 2, might share some stylistic feature with one of the few lat palaeolithic figurated artefacts, found in the Subcontinent, the pebble from Ak Kupruk (Dupree 1972), while the zig-zag lines and lozenge net might recall the geometric designs on the mesolithic flint core found at Chandravati (Sonawane 1987). Thus, the possibility of an affinity is real. In the immediate future, our mission will check the area of the old find at Panr, in the hope of finding more specimens or any possible trace of a relevant archaeological site. The consequences of a positive identification would be manifold. If the Panr stones actually are the oldest large stone sculptures of the Indopakistani subcontinent, we might have similar social organizations, and the same ritualistic behaviour, at the beginning of the Holocene, at the two opposed extremities of the Iranian plateau (Fig. 4). The traditional geometry of the ‘Fertile Crescent’ would appear meaningless; the Iranian plateau, once more, would play the role of a totally unknown world, and we should finally agree with J. Diamond that the large-scale transition to food production in southern Eurasia in the 9 th millennium BC depended upon the ecological constraints of the huge belt of arid reliefs and semi-arid valleys running longitudinally from Taurus to the Hindukush.

Always at Chatal Höyük, the renewed excavations lead by I. Hodder (1999) indicate that the relief plaster sculptures of ‘ritual’ rooms (ca. 6500-6000 BC) had meaningful connections with the cycles of use and duration of the whole house. “…The life of the houses in which the art occurred may relate to the life cycles of extended families…(relief sculpture)…seems to have been used and destroyed as the house was used and abandoned, as the family heads grew from young to old. The destruction or recovery of relief sculptures from central points in abandoned buildings perhaps suggests a concern with a passing on of authority, rights of access or ancestral ties” (1999: 162). Hodder thus proposes to link ritual plaster sculptures with the time-span of one or two generations. It would be very hard to state wether plaster statues had similar or shorter intentionality of duration, but the association with life-cycles of extended families is intriguing.

Evolution of the ancestors’ images What is really astonishing, at present, is what we could label the ‘intentionality of duration’ of the Urfa megaliths. The ‘T’-shaped pillars are not contained within the architecture: they are the cultic buildings. Their erection was the clear-cut expression of political projects meant to last for generations: in other words, and ideally, they are adapted to the mythological projections and to the ideology of powerful clans and their chiefs. With the dissolution of their social context, both cultic buildings and megaliths could not be transformed or recycled; they could only be abandoned. According to Klaus Schmidt (2007), the cultic buildings of Göbekli were intentionally infilled with their loads of figurated pillars and a new building was constructed on such ‘platform’. This is why Schmidt seems to think to these constructions in terms of funerary ‘mounds’ rather than as rooms, in any case devoted to the cult of the dead (totenkult).

In the past, Ruth Amiran (1973) stressed the basic similarity between the famous ivory figurine from Beersheba (Israel), dated to a chalcolithic horizon, with the equally famous stetatite figurine found at Tepe Yahya (Kerman, Iran) (Lamberg-Karlovsky, Meadow 1970). The Beersheba ivory has a general stylistic similarity with the PPNB plaster statues from Ain Ghazal and Jericho, and the holes surrounding its face, evidently meant to host hair, feathers or vegetal fibers are a direct link with a similar stone aceramic neolithic skeletal mask 54

T-SHAPED PILLARS AND MESOLITHIC “CHIEFDOMS” IN THE PREHISTORY OF SOUTHERN EURASIA: A PRELIMINARY NOTE a large mammal) well fits the ideology of a well organized society of hunter-gatherers. A true break comes in Stage 2, when still highly visible, but much more manageable human ancestors figures required the collaboration of various crafts and specialized knowhows (reeds and cords for the vegetal core of the statues, kiln construction and lime burning for making plaster, modelling, treating of bitumen for eyelids, use of shells for the eyes, pigments for painting, plus the presumed use of cloths and ornaments: see Kingery et al. 1988, 1992). Both the much smaller size and a quite lower cost in terms of energy input would set the plaster statues in the context of a more limited group interaction, perhaps to be identified in extended family groups. Their material nature warranted a greater flexibility in terms of production, movement, use and obliteration. Finally, in Stage 3 the same figures further shrunk in size and were carved in valuable exotic materials, possibly to be used, handled and inherited for longer times within more restricted familiar groups, possibly, in the late chalcolithic, within nuclear families.

found in the cave of Nahal Hemar (Israel). More recently (Vidale 2006a) I proposed that the lime-plaster statues of Jericho and Ain Ghazal of PPNB could be viewed as the direct forerunners of later neolithic and chalcolithic anthropomorphic statuettes most likely representing ancestors: besides Beersheba and Tepe Yahya, the chalcolithic inventory might include the anthropomorphic statuettes in limestone or marble found at Tell-es Sawwan (Iraq), one of the earliest sites in Mesopotamia where both architecture and settlement lay-out, for the 6th millennium BC, might suggest institutional redistribution. The stone figurine of Tepe Yahya was found in a small room (presumably used for storage) of a residential building, whereas the stone statuettes of Tell-es Sawwan came from graves. In both cases we are dealing with private contexts of display and fruition, possibly within to the semantic sphere of nuclear families. Are the ‘T’-shaped pillars from Urfa (10.000-8500 BC), the late PPNB plaster statues from the siro-palestinian strip (8000-7000 BC), and the chalcolithic statuettes (6000-4500 BC) images of idealized ancestors? Do they represent 3 stages of a single evolutionary process? This conjecture and its implications are summarized in the enclosed Tab. 1.

This conjectural, very sketchy reconstruction is based upon a set of data so scanty and uncertain that, like in other fields of human and social evolution, one might feel free to fill gaps and outline reconstructions following any different preconceived view. This scheme only describes the possible evolution of some representations of ancestors at the western extremity of the Iranian plateau. If the Swat stones are actually comparable to those from Urfa, there is no evidence, in the archaeological record of neolithic Mehrgarh, of similar images. But everything indicates that our knowledge of the Subcontinent, in the early Holocene, is still paradoxically limited.

The gap for the 7th millennium is real – for example, the superabundant art and iconography of Chatal Höyük seems to have quite different concerns – but the question would definetly deserve further definition. As a preliminary comment, I would like to stress that the huge pillars of Stage 1, in their absolutistic meaning (‘Our founder was the first living creature, we will last forever’) as well as in their manufacturing technology (depending on simple, intensive and collective efforts, like the hunt to

Table 1 – Conjectural reconstruction of the evolution of ancestrors’ images in the Levant and southern Eurasia, ca. 10.000-4500 BC.

Stage 1 (10.000-8000 BC) Type of ancestor image Raw material Context of display Suggested social form Intentionality of duration

Stage 2 (8000-7000 BC)

Stage 3 (6000-4500 BC)

Limestone megaliths

Lime plaster statues

Semiprecious materials figurines

Locally available

Local but transformed

Exotic, valuable

Public

Public, semi-public?

Restricted, private?

Tribes and clans

Extended families

Restricted familiar groups?

“Unlimited”

Limited

Long

Strategy of alienation

Massive burial, breakage

Curated burial

Hereditary transmission

Manufacturing technology

Hypertrophic, very simple

Relatively complex

Simpler

55

Massimo Vidale Schopen, G. (2005) Fragment and Figments of Mahāyāna Buddhism in India. Honolulu Spink, W. (2005-2008) Ajaṇṭā: History and Development. Vols.1-5. Leiden. Walser, J. (2004) Nāgārjuna in Context. New York. Zwalf, W. (1996) A Catalogue of the Gandhāra Sculpture in the British Museum. Vols. 1-2. London.

Bibliographic References Bautze Picron, C. (2000) ‘Nasik: the late Mahāyāna caves 2, 15, 20 & 23-24’, in G. De Marco & M. Taddei (eds), SAA 1997, 1201-227. Roma. Behl, B. (1998) The Ajanta Caves. New York. Berkson, C. (1986) The Caves at Aurangabad. Ahmedabad. Brancaccio, P. (2000) ‘The Buddhist Caves at Aurangabad: The Impact of the Laity’, Ars Orientalis Supplement I: 41-49. Brown, R. (1990) ‘Lajjā Gaurī in a Buddhist Context at Auranagabad’, The Journal of the International Association of Buddhist Studies 13 (2): 1-16. Burgess, J. & Fergusson, J. (1988) The Cave Temples of India. Delhi. Bhandaranayake, S. (1974) Sinhalese Monastic Architecture. Leiden. Foucher, A. (1917) ‘The Great Miracle at Śravasti’, in The Beginnings of Buddhist Art, 148-84. Paris. Franz, H.G.(1981) ‘Ambulatory Temples in Buddhism and Hinduism’, in H. Hartel (ed.) SAA 1979, 44958. Berlin. Gokhale, S. (1991) ‘Ajanta: The Center of Monastic Education’, in R. Parimoo (ed.) The Art of Ajanta, New Perspectives. Delhi. Gomez, L. & Silk, J. (eds) (1989) The Great Vehicle: Three Mahāyāna Buddhist Texts. Ann Arbor. Gomez, L. (1996) The Land of Bliss: The Paradise of the Buddha of Measureless Light. Honolulu. Hartel, H. & Yaldiz, M. (1982) Along the Ancient Silk Routes: Central Asian Art from the West Berlin State Museum. New York. Huntington, J. (1980) ‘A Gandhāran Image of Amitāyus’ Sukhāvatī’, Annali dell’Istituto Universitario Orientale 40: 652-72. Kim, I. (1997) The Future Buddha Maitreya: An Iconological Study. Delhi. Lamotte, E. (1970-1981) Le Traité de la Grande Vertu de Sagesse. Vols 1-5. Louvain. Litvinsky, B.A. (ed.) (1996) History of Civilizations of Central Asia. Vols. 1-6. Paris. Prematilleke, R. & Silva, R. (1968) ‘A Buddhist Monastery Type of Ancient Ceylon Showing Mahayanist Influence’, Artibus Asiae XXX (1): 61-84. Rhi, J. (2003) ‘Early Mahāyāna and Gandhāran Buddhism: An Assessment of the Visual Evidence’, The Eastern Buddhist XXXV (1-2): 152-90. Schlingloff, D. (1999) Guide to the Ajanta Paintings. Delhi. Schopen, G. (1997) Bones, Stones, and Buddhist Monks. Honolulu.

56

T-SHAPED PILLARS AND MESOLITHIC “CHIEFDOMS” IN THE PREHISTORY OF SOUTHERN EURASIA: A PRELIMINARY NOTE

Fig. 1 – Map of the “Fertile Crescent” with the location of Göbekli Tepe, Nevali Çori and other sites mentioned in the text. (Badisches Landesmuseum 2007).

Fig. 2 – Drawing of Panr 1, most probably part of a T-shaped pillar (from Panr, Swat, Pakistan, presently at the Archaeological Museum of Saidu Sharif).

57

Massimo Vidale

Fig. 3 – Drawing of Panr 2, the capital of another T-shaped pillar (from Panr, Swat, Pakistan, presently at the Archaeological Museum of Saidu Sharif).

Fig. 4 – The valleys of Urfa (Turkey) and Swat (NWFP, Pakistan) are separated by about 3000 km as the crow flies.

58

PROTOHISTORIC TERRACOTTA FIGURINES: CONTINUITY AND CHANGE IN FORM AND STYLE (CIRCA 2000-300 BC) Arundhati Banerji

In the beginning of second millennium BC, a number of regional cultures1, designated as chalcolithic, emerged in different parts of the Subcontinent (Table 1).2 These chalcolithic cultures were located on the tributaries of main rivers, not far from the hills in a semi-arid region. Their economy was based on farming, animal husbandry, hunting and fishing. These cultures were characterized by a distinctive ceramic industry3 and artifacts of stone and copper4. These cultures flourished mostly in north-west (Swat valley), Eastern Baluchistan, Punjab, Haryana, Uttar Pradesh, Central India, western India, Rajasthan, Gujarat and Maharashtra. In Central India and Maharashtra these settlements were succeeded either by the black-and-red ware phase followed by the blackslipped or directly the Northern Black Polished Ware (NBPW) phases and in north, there are several phases represented by OCP of the upper Ganga plain, black-andred ware, Painted Grey Ware (PGW), pre-NBPW cultures which were widely distributed from the lower Ganges to the Punjab. 5

protohistoric period elephant becomes relatively common, especially in the regions of the upper or mid Ganges (Banerji 1994)6. On the other hand, in case of female forms, an evolution could be noticed from the simple, cruder, featureless variety to the specimens with broad hips decorated with a girdle in addition to some other anatomical details (Banerji, in press). Archaeological evidence of terracotta finds from the above mentioned socio-economic strata indicates a wide distribution (Table 1). In almost all the regions, clay was fashioned in a number of forms. The early specimens are all hand-made, pressed into the desired shapes in round without possessing any definite shape or form. The anatomical details where indicated are expressed by appliqué pellets, bands, incisions or grooves and pinholes. At a slightly later stage punches were introduced to emphasize both, ornaments as well as anatomical details. Most of these specimens are characterised by appliqué, incised, pin-hole and punched decorations, reflecting the social milieu of the artists. At times, some examples were also finished by some kind of colour while a few others bear traces of some paintings. Majority of these examples lack ornamentation as the decoration is yet to take a definite form. The primary point that strikes us is the nudity and only the occasional use of ornaments. The clay used for modelling was free from pebble and was also not brought from the alluvial plains as found in case of the later examples. Sometimes the use of tempering material like ash, sand, rice-husk or cattle-dung is evident. The paste was made after kneading the clay and in majority of cases, these were left simply sun-dried. Both human and animal forms were made, depending on the usage and the requirements of a particular community or a society.

This paper deals with the study of protohistoric terracottas from the sub-continent to ascertain how forms continue and change with the time, space and shift in various stages of culture. The subject chosen is with special reference to the human and animal forms from the northern and western India (Table 1). In these regions, bull is found to be the most commonly depicted animal showing a variation in style particularly noticeable in the depiction of horns, hump and various other physical features (plates and figs.). Towards the later stages of this 1 The decline of the Harappan Civilisation was followed by the emergence of regional copper using cultures some of which heavily depended on the use of stone tools, though they also used copper as permitted by their economy. These cultures were devoid of elaborate town planning, monumental buildings, drainage, unified system of weights and measures in addition to the writing of the Harappans. This phase is termed as post-Harappan Chalcolithic (Allchins 1968: 179). It emerged on the scene as a consequence of the end of urban centres and change in socio-economic formation indicating the movement of people all over the region. This post-Harappan period certainly marks a new socio-economic and cultural stage distinct from the Harappans.

The study of terracotta art forms of northern and western India during the protohistoric period reveals the following three main stages of development between the Harappan and the Mauryan times, based on the typological continuity and change in form and style (Table 2).

2

Stage 1: Post-Harappan and Chalcolithic cultures Stage 2: Early Iron Age cultures Stage 3: Early NBPW cultures

Peasant agricultural communities settled in almost all the regions of Indian sub-continent. The settlements are located in different ecological zones or regions, represented generally by low mounds of small size, grouped around central sites which were relatively prominent. 3 Red ware of relatively inferior quality formed the most characteristic ceramic with variations in shapes and painted motifs marked by profuse decoration either by painting or by incision.

The Stage 1 (Banerji 1994) is marked by the postHarappan cultures, known from Gandhara Grave culture (Period I) of northwestern India (Stacul, all entries), Pirak

4 The chalcolithic people were pastoral agriculturists and favored growth of specialisation and invented metallurgy, based on copper, tin and an alloy bronze. 5 All these assemblages could be categorised as the Gangetic Civilisation for they present almost a continuous process which finally ushered as an urban state.

6 The post-Harappan artistic traditions are products of agrarian society showing the disappearance of urban cities, which explains the popularity of bull in this stage.

59

Arundhati Banerji culture of eastern Baluchistan7, Jhukar culture of Sind, Cemetery H culture of central Punjab8, Bara culture of the Indo-Gangetic divide, OCP culture of the GangaYamuna doab, Lothal B, Surkotada I C, Prabhas and the Lustrous Red Ware cultures of Gujarat, Ahar culture of southeastern Rajasthan, Malwa culture of Central India and Jorwe culture of the Deccan9.

anthropomorphic (Pl. 5) figurines (Joshi 1993), a snakehood - like (Pl. 8) object (Sharma 1987) from the Bara cultures of Haryana ,Punjab and Uttar Pradesh; a female form with bald head and perforated shoulders (Pl. 6) from the OCP level at Lal Qila (IAR 1969-70); other animal figurines from Lothal B and Rangpur (Rao 1985, 1963) besides the manifold forms of bull and corpulent human male-female forms from the Ahar (Sankalia 1969), Malwa (Ansari 1975 and Wakankar 1989) and Jorwe (Dhavalikar 1988) are noteworthy (Pl. 1.9; Figs. 24, 7, 10) for studying the continuity and change in terracotta representations14. Late Jorwe level also revealed human (Fig. 3) and animal forms with slight improvement in quality. As is evident from our study, the tradition of art in this stage shows a discontinuity in its original form but it marks an emergence of a new pattern much cruder and simpler in type (Banerji 1994). This points to a continuity in the technique of modelling by hand besides the contents emphasizing thereby an existence of a pattern of life which was already very old. The Stage II is marked by the introduction of iron or the diffusion of iron technology. The Gandhara Grave 15 and Pirak16 cultures reveal continuity from the earlier stage, elsewhere it is represented by the advent of the PGW culture in north India17 and the manifestation of blackand-red ware18 and black-slipped ware culture19 on its periphery in Uttar Pradesh and Rajasthan. As also the black-and-red ware cultures of Malwa, Gujarat and the Deccan20.

In Stage 1 two main features were marked viz., limited occurrence and preference for animal forms of which bull10 is most common (Pl. 1-3; 9 and Figs. 4, 7, 8, 16). These specimens are shown with prominent hump and horns (Banerji 2006). Bull figurines are indicative of a variety in the depiction of hump and horns (Pls. 1-3, 9; Figs. 4, 7, 16). Although both humped and humpless varieties with long and short horns are reported (Fig. 11). Other forms of animals comprise horse and doublehumped camel (Figs. 5-6; Pl. 4)11. Human forms from this stage are simple, conventional type, with corpulent body and stumpy limbs12. Little emphasis is given in depicting the physical features (Figs. 2; Pls. 5-7, 10-11). Pinching, incisions and appliqué methods are used to indicate the parts of the body besides some other details like ornaments13. The violin-shaped female form evolved during the last phase of the post-Harappan and chalcolithic cultures of western India (Fig. 1). Other important forms like wheeled animals (Pl. 3), 7 In eastern Baluchistan (Pakistan), a distinct material culture was unearthed at Pirak in Kachi plain. In southern Sind, a grey ceramic was found labelled under the site name of Jhangar, also reported at Amri and Lohumjodaro. Some important discoveries have also been made in the upper Indus Valley. In the frontier region of Pakistan, a large number of sites were located in Swat, Dir, Chitral and Peshawar Valleys represented by extensive cemeteries showing inhumation, cremation and urn burial along with funerary pottery and objects of copper, bronze, shell, terracotta and stone. In the final phase, iron appears with no significant change in the material culture.

14 The expansion of peasant communities into the virgin soil of central India and the Deccan and the relative affluence attained by the Ahar and Malwa cultures speak of the profusion of terracotta figurines in these regions. 15

In Swat valley in the north-west at Katelai I, Loebanr I, Barama, Timargaha, Thana, Butkara II and Balambat (a settlement site), iron occurs in the final phase of the graves along with grey pottery and skeletal remains of horse. A single object of iron was reported from Period V from Katelai (grave 126) along with a grey vase. No iron object was found in the earlier deposits elsewhere. In the 3 rd period (Stacul’s Period VII) a number of iron implements, advanced pottery and burials were found which continue right up to the historical period.

8 Since the field work has not been done uniformly in all the regions, very little or nothing is known about the terracotta art of post-Zhob culture of northern Baluchistan, post-Kulli culture of southern Baluchistan, Cemetery H culture of Central Indus Valley, Jhukar and Jhangar cultures of Sind, Prabhas culture of Saurashtra and Kayatha culture of central Malwa.

16 Pirak is an important site showing a transition from copper to iron technology. The introduction of iron indicated a gradual process of evolution before the end of Period II around c. 1000 BC – the frequency of which increases only in Period III around c. 800 BC.

9

These cultures can be broadly distinguished by the predominance of exotic elements in the regions of Gandhara and Kachi plain: the local or composite culture with distinct foreign elements, such as the Jhukar cultures of Sind, the degenerate rather rustic form of the post-Harappan cultures primarily derived from the Harappan socio-cultural milieu, as the Bara, OCP, Lothal B cultures, the regional cultures of south-east Rajasthan, Malwa and the Deccan.

17

This culture is distinguished from the preceding post-Harappan cultures (Bara and OCP) by the use of iron, horse and a grey coloured pottery bearing broad similarities with the Gandhara grave culture of north-west Pakistan and south-east Afghanistan. PGW Culture zone is bounded by the Sutlej-Beas in the west, Thar desert in the south, the river Chambal in the south-east and the Siwaliks in the north.

10

Bull in Rajasthan, Malwa, Deccan, Gujarat represented by Ahar, Malwa, Jorwe and the Lustrous Red Ware cultures reveal indigenous origin with varied strands, colonising new areas.

18 This was given a distinction of an independent culture which succeeds the OCP settlements in northeastern Rajasthan and southern Uttar Pradesh. It has two different chronological contexts firstly in a prePGW horizon secondly in association with PGW assemblage.

11

In Kachi plain at Pirak (Banerji 1994), camels with two humps, painted in bichrome colour scheme were represented with wheels. Also interesting are the forms of she-camel and cow with udders besides horse of different sizes found with projected long neck, occasionally painted and provided with a hole for the tail. Another addition is that of horse riders with non human face. Occurrence of certain human forms with headdress bearing resemblance to the Zhob female forms (Banerji 1987) is also interesting.

19 In central Ganga basin there is presence of a deposit bearing blackslipped ware pottery below the NBPW deposit and its associated finds such as black-and-red ware, occasionally painted (comparable to those reported from the Vindhyas), coarse red and plain grey wares in the middle and lower Ganga valleys besides iron objects, short blade industry, bone tools and terracotta objects.

12

In Gandhara Grave culture of Swat Valley human figurines are flat and schematised (Banerji, in press).

20 Pre-NBPW phase occurs in these regions with sporadic evidence of iron at Nagda and Eran in central India. In Maharashtra evidence of a late phase of Jorwe culture exists but iron was reportedly missing. In

13

These forms are reported from Sibri, Pirak, Navdatoli and Nagda (Banerji 1994).

60

PROTOHISTORIC TERRACOTTA FIGURINES: CONTINUITY AND CHANGE IN FORM AND STYLE (CIRCA 2000-300 BC) The last Stage (III) points to an upsurge in production 22. This early NBPW phase is characterised by improved quality with an increase in the number of animal figurines which were produced in greater number in this phase. As far as the technique is concerned, the incision, punched circlets etc., are more frequently used for decoration. Details are marked by pinching, incision, appliqué pellets etc. Stylisation noticed in the earlier examples also gradually decreases.

The art tradition of early Iron Age marks a change from the preceding stage by integrating several elements found in the local art forms particularly in the form of decoration from Gandhara Grave (Banerji 2005) and the Pirak cultures (Jarrige 1979). The forms of earlier type continued along with unbaked clay objects, which are marked by exaggeration of forms. The animal forms like horse and camel (Jarrige et al., 1976), though stylised, still continue (Fig. 11). Most important ones are the flat forms of human.

Animal figurines (Pl. 7-8) represent bull, elephant, horse, ram, dog, deer or stag (Gaur 1983 and Narain et al. 1963) and birds (Joshi et al. 1976-77). Bull figurines are popular besides elephant. Of the human forms, female with heavy hips are most common (IAR 1967-68, 197980 and Gaur 1983), but these do not appear to be naturalistic representations (Pl. 9, 11). Noteworthy finds are the composite forms of bird (IAR 1964-65) or snake with body that of a human being (Pl. 9-12; Fig. 14). The snake-hooded specimen with broad hips, slender waist, elaborate decoration and appliqué pellet eyes (Banerji in press) forms one of the unique finds of this phase.

Iron using Gandhara Grave culture (Dani 1967; Banerji 1994) revealed a flat schematised form with pinched face and stumpy, outstretched arms; in addition to a human figurine with chhannavira besides pinholes in the lower part which were used to indicate the ornaments (Figs. 12, 15). Another type viz., the fan-shaped female form from Zarif Karuna (Khan 1979) is also of special importance. These figurines are available in both red and grey colours. Animal figurines include humped bull, cat with pinched mouth, dog, horse, she-camel with udders ,two wheeled camel (Jarrige 1976), ram’s head and horn (Dani 1967). Bull figurines21 have conjoined legs, appliqué hump and tail (Dani 1967); at times hump becomes prominent (Khan 1979). Lines in black colour appear on the ram’s head (Dani 1967), horns and snout.

What distinguishes the terracotta art tradition of the three stages is primarily the variation of forms and qualitative change in the delineation of figurines. Our study suggests that the occurrence of bull in a variety of forms all over the region throughout the period serves as a link between the terracotta art traditions of different stages (Table 2) 20. The figurines from this period certainly are important for the study of stylistic evolution, particularly modelled in terracotta and of stone sculptures in general. It is true that with the discovery of figurines in the aceramic Neolithic levels, the beginning of plastic art in Indian sub-continent has been pushed back to the seventh millennium BC and it would thus be misleading to consider the history of Indian sculpture to begin with the Mauryan period.

Some new forms in red and grey colours were introduced in the PGW level. These show the continuation of pinching style and ornamentation by pierced holes, incisions or appliqué methods. There are deep grooves and punched circlets to decorate the animal figurines. Common types (Pls. 1-4; Fig. 17) were wheeled ram and bull apart from horse, dog, deer, pig, birds (Banerji 1994). Ram has twisted horns. Both human male and female forms occur (Banerji 1994). The female forms are characterised by broad hips (Pl. 6). Other significant inclusion is that of a violin-shaped female with an incision in the lower half form (Banerji 1994: pl. 88). A bust shows composite features (Pl. 5) - body of human female and head that of an animal (Sahi 1978).

The study of protohistoric figurines further indicates that there was an art-period intervening between the Indus Valley and the Mauryan periods with a continuity in technique, style and motifs particularly in the modelling of female forms, but with certain modifications. The find spots of these terracottas also serve as an indication that the centre of terracotta art shifted gradually towards east to the Ganga valley, central and eastern India where the cultures developed under a village ethos. This entire period speaks of a stage of readjustment by a social class or a group who settled in isolation in several pockets to the rural setting without loosing its absolute identity as reflected by the artifacts and techniques that survived with many changes in a restricted manner.

Female with broad hips (Fig. 13) adorned with an appliqué girdle became a unique innovation of the preNBPW deposit (Narain et al. 1977) along with a violin shaped form shown with navel quite distinctly (Banerji 1986). The depictions of animals include elephant, bull, dog, cat ,deer, fox, etc. (Banerji 1986 and Narain et al. 1977). The depiction of elephant is naturalistic. Importantly the specimens are highly polished, at times also decorated with paintings. Other forms are more stylised and decorated with the help of appliqué bands, incisions and punched circlets.

This is indicative of a changing pattern and gradual transformation of the society which also influenced the art forms. The figurines were not utilised as a commodity product, but in all probability, formed an essential part in

Gujarat a long hiatus appears between the end of chalcolithic period and the beginning of early Iron Age occupation.

22

This phase is significant that follows the PGW, Black-and-red ware and black-slipped ware levels in northern and western India with economy based on agriculture, cattle keeping, hunting, fishing and crafts.

21

The graves of Timargarha are devoid of any example of bull but the settlement site at Balambat revealed figurines of bull. The bull is also not found in the Swat cemeteries.

61

Arundhati Banerji day-to-day life of the people. These figurines also do not appear to be the product of any specialised craftsman or an artist as in case of the later art forms.

Rao, S. R. (1963) ‘Excavations at Rangpur and other Explorations in Gujarat’. Ancient India 18-19. New Delhi: 5-207. Rao, S. R. (1985) Lothal and the Indus Civilization. Bombay. Rao, S. R. (1979) ‘Lothal A Harappan Port Town 195562’. MASI 78 (I-II). New Delhi. Sahi, M. D. N. (1978) ‘New Light on the Life of the PGW People as Revealed from Excavations at Jakhera, District Etah’. Man & Environment II: 101-103. Sankalia, H. D. (1969) Excavation at Ahar-Tambavati 1961-62. Poona. Santoni, M. (1984) ‘Sibri and the South Cemetery of Mehrgarh: Third millennium connections between the northern Kachi plain (Pakistan) and Central Asia’. In Allchin, B. (ed.) South Asian Archaeology (SAA) 1981. Cambridge: 52-60. Sharma, Y. D. (1987) ‘Fresh Light on the Bara Culture from Mahorana’. In Pande, B. M. and Chattopadhyay, B. D. (eds) Archaeology and History, Essays in memory of Shri A. Ghosh. Delhi: 157-176. Silvi Antonini C. & Stacul G. (1972) The Protohistoric Graveyards of Swat (Pakistan), Rep Mem, VII, Rome. Stacul, G. (1969) ‘Excavations near Ghaligai (1968) and Chronological Sequence of Protohistorical Cultures in the Swat Valley (West Pakistan)’. East and West (EW), 19 (1-2): 44-91. Stacul, G. (1974) ‘New Archaeological Evidence on North-west Indo-Pakistan c. 3rd-1st millennia B.C.’. EW, 24 (3-4): 239-243. Stacul, G. (1976) ‘Excavation at Loebanr III (Swat, Pakistan)’. EW, 26 (1-2): 13-30. Stacul, G. (1977) ‘Report on the Excavations at Aligrama (Swat, Pakistan) 1974’. EW, 27 (1-4): Fig. 42. Stacul, G. (1978) ‘Excavations at Bir-kot-Ghundai (Swat, Pakistan)’. EW, 28 (1-4): 137-150. Stacul, G. (1979) ‘Early Iron Age in the North-east of Sub-continent’. In Agrawal, D. P. and Chakrabarty, D. K. (eds) Essays in Indian Protohistory. Delhi: 341-345. Stacul, G. (1979a) ‘The Black-burnished Ware Period in the Swat Valley (c. 1700-1500 B.C.)’. SAA 1977/2, M. Taddei (ed.). Naples: 661-673. Stacul, G. (1979b) ‘The Sequence of the Protohistorical periods at Aligrama (Swat, Pakistan)’. SAA 1975, J.E. Van Lohuizen De-Leeuw (ed.). Leiden: 88-90. Stacul, G. (1980) ‘Bir-kot-Ghundai (Swat, Pakistan): 1978 Excavation Report’. EW, 30 (1-4): 55-65. Stacul, G. (1980a) ‘Loebanr III (Swat, Pakistan): 1979 Excavations Report’. EW, 30 (1-4): 67-76. Stacul, G. (1984) ‘Cultural change in the Swat Valley and beyond c. 3000-1400 B.C.’. SAA 1981, B. Allchin (ed.). Cambridge: 205-212. Stacul, G. (1989) ‘Continuity of Forms and Traditions at Bir-Kot-Ghundai, Swat’. SAA 1985, Kren Frifelt and Per Sorensen (eds.). London: 321-326. Stacul, G. (1993) ‘Kalako-deray, Swat: 1989-1991 Excavation Report’. EW, 43 (1-4): 69-94.

Bibliographical References Allchin, Raymond and Bridget (1968) The Birth of Indian Civilization. Harmondsworth, Middlesex. Ansari, Z. A. (1975) Excavations at Kayatha. Poona. Banerjee, N. R. (1987) ‘Nagda 1955-57’. Memoirs of the Archaeological Survey of India (MASI), 85. New Delhi. Banerji, Arundhati (1994) Early Indian Terracotta Art (including its notes and bibliography). New Delhi. Banerji, Arundhati (2006) ‘Terracotta Female Figurines from Protohistoric Swat (with all the entries in Notes and Bibliography)’. Ancient Pakistan, Vol.XVII,Peshawar, Pakistan ,83-89. Banerji, Arundhati (2006) ‘Bull in Banas Culture’. In Arundhati Banerji (ed.) Harismriti Studies on Art, Archaeology and Indology. New Delhi: 217-223. Banerji, Arundhati (2006-7) ‘Representation of Snake Goddess in Eastern India’. Journal of Bengal Art,Vol.11-12, Dacca, Bangladesh ,121-126. Banerji, Arundhati (2008) ‘Archaic Human Forms in Terracotta from the Early NBPW Deposit’. In Devangana Desai and Arundhati Banerji (eds) Kaladarpana. The Mirror of Indian Art, papers presented in memory of Shri Krishna Deva: New Delhi. Banerji, Arundhati (2008) Terracotta Female Forms Tradition Innovation and Evolution. New Delhi (in press). Chakravarty, K. K., Wakankar V. S., and Khare M. D. (1989) Dangwada Excavations. Bhopal. Dani, A. H. (1967) ‘Timargarha and Gandhara Grave Culture’. Ancient Pakistan III. Pakistan: 1-55. Dani, A. H. (1967) ‘Report on the Excavation of Balambat Settlement Site’. Ancient Pakistan III. Pakistan: 237-284. Dhavalikar, M. K., Sankalia H. D. and Ansari Z. D. (1988) Excavations at Inamgaon I (I and II). Pune. Gaur, R. C. (1983) Excavations at Atranjikhera Early Civilization of the Upper Ganga Basin. Delhi. Jarrige, J. F. (1979) Fouilles de Pirak. Paris. Joshi, J. P. (1993) ‘Excavation at Bhagwanpura 1975-76 and Other Explorations and Excavations 1975-81’. In Haryana, Jammu & Kashmir and Punjab, MASI 89. New Delhi. Joshi, M. C. and Margabandhu C. (1976-77) ‘Some Terracottas from Excavations at Mathura: A Study’. Journal of Indian Society of Oriental Art(ns) VII. Calcutta: 16-31. Khan, G. M. (1979) ‘Excavation at Zarif Karuna’. Pakistan Archaeology 9. Pakistan: 1-94. Narain, A. K. and Roy T. N. (1968) Excavations at Prahladpur 1963. Varanasi. Narain, A. K. and Singh P. (1977) Excavations at Rajghat 1957-58, 1960-65, I-II. Varanasi. 62

PROTOHISTORIC TERRACOTTA FIGURINES: CONTINUITY AND CHANGE IN FORM AND STYLE (CIRCA 2000-300 BC) Stacul, G. (2005) ‘Human Figurines From Early Swat( 2nd-1st millennia B.C.)’. SAA 2001, Vol. I. Catherine Jarrige and Vincent Lefevre (eds.) Paris: 303-309. Tusa, S. (1981) ‘Notes on Some Protohistoric finds in the Swat Valley (Pakistan)’. EW, 31 (1-4): 99-120. Wakankar, V. S. (1967) ‘Kayatha Excavation Number’. Journal of Vikram University. Ujjain.

63

Arundhati Banerji

Table 1 – Stages of Culture in Northern and Western India

Kabul Valley Mundigak VII – Iron becomes common VI – Iron in early levels

North-West (Swat Valley) Gandhara grave (Iron, Grey Ware, skeletal remains of horse)

Baluchistan Rana Ghundai IV-V and Dabarkot / Shahi Tump Cemetery (layers of burning) Pirak III (grey pottery, iron) Pirak II (red ware with iron in late levels)

V – Chalcolithic IV – Harappan

Gandhara grave (pre-iron) Neolithic

West Punjab Gandhara grave Cemetery H Harappa Pre-Harappan

Pirak culture (pre-iron) Zhob-Kulli Chalcolithic Neolithic (ceramic) Neolithic (aceramic)

East Punjab & Haryana

Doab

Central Doab

Sind

Gujarat

Early NBPW,

Early NBPW

Early historical

NBPW, B&RW

BSW, PGW

PGW, B&RW, BSW

Early NBPW, BSW

Jhangar

(iron)

Bara

Bara, OCP

Jhukar

LRW (iron from late levels)

Harappan

Harappan

Pre-Harappan

Pre-Harappan

Lothal B, Surkotada IC, Prabhas Rangpur IIB

Rajasthan Early historical

Northern Rajasthan NBPW

Malwa Early NBPW

Deccan B&RW (Iron)

(NBPW)

BSW/PGW

B&RW, BSW

Jorwe

Ahar IC (LRW) Ahar IA-B

OCP/Bara

(Iron in late levels)

Malwa

Malwa, Ahar, Kayatha

Daimabad culture Post-Harappan Savalda

* B&RW (Black-and-Red Ware), BSW (Black-Slipped-Ware), LRW (Lustrous-Red-Ware), NBPW (Northern-Black-Polished-Ware), OCP (Ochre-Coloured-Pottery), PGW (Painted-Grey-Ware).

Table 2 – Typological Continuity and Change in Protohistoric Terracotta Figurines Continuity

Change

1. Modelling by hand

1. Forms are much cruder and simpler in style

2. Pinching technique used for body details

2. Difference both in quality and quantity from the earlier periods, i.e., in the Harappan terracottas

3. Use of appliqué pellets, strips, nail marks, occasional pinholes making a rosette design for body details or decoration

3. Improper baking

4. Faint traces of paintings on the figurines from Dadheri (Fig. 8) and Rangpur like Kulli and Mehi (Banerji 1994).

4. Lack of anatomical details except for the bull 5. No artistic merit in the depiction excepting for those found towards the close of the third stage.

64

PROTOHISTORIC TERRACOTTA FIGURINES: CONTINUITY AND CHANGE IN FORM AND STYLE (CIRCA 2000-300 BC)

Fig. 1 - Violin-shaped form Bilwali; Fig. 2 - Female form, Inamgaon; Fig. 3 - Female form Inamgaon; Fig. 4 - Bull Kayatha; Fig. 5 - Camel Pirak; Fig. 6 - Horse Pirak; Fig. 7 - Bull Kayatha; Fig. 8 - Bull Dadheri; Fig. 9 - Horse Kausambi; Fig. 10 - Bull Inamgaon; Fig. 11 - Camel Pirak; Fig. 12 - Human form, Timargarha; Fig. 13 - Female form, Rajghat; Fig. 14 - Female bust, Kausambi; Fig. 15 - Female form Loebanr; Fig. 16 - Humped bull, Hulas; Fig. 17 – Bull, Rajghat; Fig. 18 - Elephant, Rajghat; Figs. (1-8, 10 and 16: Stage-1; 11-13, 15, 17: Stage-2; 9, 14 and 18: Stage-3).

65

ASSESSING ANTHROPOGENIC SOIL EROSION WITH MULTI-SPECTRAL SATELLITE IMAGERY: AN ARCHAEOLOGICAL CASE STUDY OF LONG-TERM LAND USE Andrew M. Bauer & Kathleen D. Morrison

Entering the tropics from the cooler latitudes, one enters a geomorphically entirely new setting. Not only the climate, fauna, and flora, but also the human race, speech, social structure, and cultural achievements in this region are completely foreign that one adjusts but slowly (Büdel 1982: 120).

Introduction to the Study Region: Previous Research and Geomorphology The region analyzed below is situated in the vicinity of the ancient city of Vijayanagara and includes a study area of approximately 40 km2, the majority of which is on the north side of the Tungabhadra River. Most of this area has been intensively surveyed by Morrison and Sinopoli’s Vijayanagara Metropolitan Survey between 1988 and 1998 (Sinopoli & Morrison 2007). The VMS was primarily concerned with settlement, land use, and the organization of production in the hinterland of Vijayanagara during the city’s reign as an imperial capital from the 14th to mid 16th century AD. Towards this end, the VMS intensively surveyed over 300 km2 outside of the Vijayanagara urban core, recording more than 1000 sites. While the vast majority of the sites discovered by the VMS dated to the Vijayanagara Period, many were also attributed to earlier periods, and included artifact scatters, iron producing locales, rock shelters, megalithic complexes, and settlements. Of particular note are a series of five Iron Age2 settlements situated on inselberg hills. In this short paper, we largely limit our investigations to such Iron Age sites.

Despite his insistence on the fundamental ‘otherness’ of tropical peoples, Büdel, a German geomorphologist famous for his paradigmatic theory of ‘double planation’, makes a critical point about our understanding of geomorphic processes. Büdel’s double planation, still used (in modified form) to explain the development of regions characterized by nearly level plains punctuated with isolated hills, typical of tropical regions, and especially many parts of South India, emerged from his research on the Tamilnad Plain of India. As he implies, geologists and other scholars have adjusted ‘but slowly’ to the differences evident between tropical and temperate environments. Here, however, we wish to take Budel’s argument a step further, re-integrating those very human inhabitants into our understanding of long-term environmental process, in the end insisting that if tropical geomorphology requires additional models for understanding change, then so does archaeology here – and elsewhere. Although the importance of long-term human action on vegetation, faunal distributions, and even soils is well-accepted, if not fully explored empirically, we make the point here that social institutions and cultural practices can play an active role in shaping the earth’s surface itself – even on the scale of the massive hills and expansive plains that characterize much of southern India.

The South Indian Iron Age has been the subject of scholarly research for nearly two centuries. However, until recently most research on the Iron Age has focused on megalithic burials, which appear to have been constructed across much of the peninsula during this period. The results of the VMS suggest that settlement during the Iron Age was spatially diverse. Within the study area, village sites show occupational ranges from over 60 ha to under 5 ha in extent. In addition to size variation, there also appear to be some differences in the activities that occurred at certain settlements. For example, some sites show evidence for specialized iron production, while others have distinctive grinding slicks indicative of grain processing3. Within the Southern Deccan region of the subcontinent more generally, only a few Iron Age settlements have been systematically excavated. From those, it appears that agricultural

In this paper we assess the distribution of soil in an area of northern Karnataka that was densely occupied during both prehistoric and historic periods. More specifically, we present a land classification of ASTER 1 satellite data in combination with surface observations to document the proportions of exposed rock and remaining soil cover on residual hills within a 40 square kilometer study area, most of which is in Karnataka’s Koppal district (Fig. 1). Through an examination of soil cover in relation to archaeologically documented sites and features, we suggest how ancient land use activities impacted the distribution of soils and shaped the environmental history of the region more generally. Before doing so, however, it is necessary to briefly contextualize the study region, ancient land use, and other physical and geomorphic processes operating in the region.

2 Archaeological discrimination of the Neolithic Period (3000-1200 BC), Iron Age (1200-500 BC) and Early Historic Period (500 BC – AD 500) sites based on surface evidence alone is somewhat problematic (see A. Bauer et al. 2007; Boivin et al. 2005; Devaraj et al. 1995; Morrison 2007). However, radiometric evaluations from excavated contexts at several of the sites identified by the VMS, and discussed here, have yielded primarily Iron Age results (Sinopoli et al. 2003). 3

1

Note that the material remains for these activities are not necessarily mutual exclusive by site. Moreover, such grinding features are reported on Neolithic sites throughout the Southern Deccan region (Boivin 2004; Boivin et al. 2005; Fuller et al. 2001), and the dating of such features devoid of stratigraphic contexts remains difficult.

Advanced Spaceborne Thermal Emission and Reflectance Radiometer.

67

Andrew M. Bauer & Kathleen D. Morrison production was perhaps diversified and intensified during the period. Cultigens such as barley, wheat, rice, millets and a variety of pulses have been found in Iron Age contexts at several sites in the region at large (Kajale 1984, 1989). At the moment, it is difficult to characterize the entire range of Iron Age agricultural production. However, Morrison (2009) has suggested that the concentration of settlements close to perennial water sources, such as the Tungabhadra River, may reflect an increased reliance on floodwater rice production. Moreover, preliminary associations between Iron Age sites and small scale reservoirs, as well as modifications to naturally occurring weathering pits and rock pools, suggest that more sophisticated water management began in this period (Bauer 2007; Bauer et al. 2007; Bauer & Morrison 2008; Morrison 2009).

variations of two-stage models, of which Budel’s theory of ‘double planation surfaces’6 remains the most prominent. Through the dual actions of basal weathering and surface erosion, residual hills are created on time scales in the magnitude of millions of years 7. In short, the land is constantly lowered as surface materials are slowly removed and the basal weathering surface moves downward. As the ground surface lowers, unweathered basal knobs and core stones become exposed and stacked as residual hills. Once rocks are elevated above the surrounding mantle they are more stable and less prone to weathering due to the absence of moisture retaining soil. Nevertheless, fresh gray granite does oxidize to a reddish-brown color within the study area. In general, this process appears to be on the order of magnitude of thousands to millions of years; however, the rate varies according to a number of factors. For example, it is clearly facilitated by the growth of biofilm covers of bacteria, which retain moisture close to the rock surfaces in certain settings. Indeed, many boulders in the study area are stained red and black from both oxidation and biofilm cover. Often the hills in the study region show a horizontal band dividing oxidized and unoxidized rock of granitic composition on their slopes, which generally can be taken as evidence of more and less recent exposure. In other words, boulders toward to the top of the residual hills – the highest and earliest to be exposed by planation – tend to show higher degrees of oxidation than those toward the hill bases.

Within the study region, Iron Age pastoral activities have been clearly demonstrated by the analysis of faunal remains excavated by Morrison and Sinopoli’s current EHLTC project at Kadebakele. R. Bauer’s (2007) analysis of faunal remains suggests variability in the consumption and use of wild and domesticated animals across the site. Among domesticates, zebu cattle and water buffalo comprised the most significant proportion of the sample. Fusion data and analyses of tooth wear indicate that most of these animals lived until advanced age, possibly reflecting pastoral subsistence marked by a dietary reliance on secondary products rather than meat. However, it is worth noting that patterns of burning across the site suggest that cattle were occasionally prepared via roasting, perhaps in ritual feasting events. Nevertheless, osteological signatures also indicate that some individuals were used for traction and possibly agricultural labor into advanced age; hence, characterizations of the Iron Age subsistence as a mixed agro-pastoral regime seem most appropriate.

On several hills exposures of weathered residual material have been observed relatively high on the outcrop, and even show in situ core stones. However, the majority of the soil on the inselbergs not covered by archaeological sediments is a thin layer of material on local occurrences of gentle slopes or flat basin areas. These soils are generally composed of angular coarse-sandy material of weathered rock that has been transported a short distance from the surrounding granite. A thin layer of accumulated organic-rich humus can often be found in these contexts. While these soils do support a variety of vegetation, they are generally very thin, with total depths of less than 15 to 20 cm. At their base there is often an abrupt transition to minimally weathered oxidized rock and in locations where these soils have been stripped, relatively flat areas of exposed rock are left behind (Fig. 2).

The study region receives an average of 550 mm of rain per year (India Meteorological Department 1981), with most precipitation falling in the summer months during the southwest, or advancing, monsoon. Geomorphologically, the region can be described as a (partly) stripped etchsurface4, or more commonly peneplain, containing several types of residual inselberg hills composed primarily of Precambrian granitic to granodioritic gneiss – commonly referred to simply as granite. The residual hills of the Tungabhadra Corridor likely owe part of their relief to tectonic activity5; however, in general, they are products of long periods of weathering and erosion. The most accepted explanations for such plain and residual hill development are

Methods and Analysis As indicated above, the primary landforms that characterize the study region are erosional features congruent with geomorphic models of planation due to weathering and soil removal. Changing conditions and human activities, however, will alter the rates of surface

4 See Thomas (1994: 299-309) for a detailed discussion of the variable terms ‘etchplain’, ‘etchsurface’, ‘peneplains’ and ‘pediplains’. Here, we use the term ‘etch’ surface to emphasize that the geomorphology of study region appears to have resulted primarily from long processes of deep chemical weathering and subsequent erosion (see also Adams 1975).

6 Büdel’s theory of ‘double planation surfaces’ was first presented as ‘doppelten Einebnungsflachen’ (Büdel 1957) and later synthesized more generally in Klima Geomorphologie (Büdel 1977, 1982). Also see Bremer & Sander 2000; Thomas 1994; Twidale 1981.

5

Radhakrishna and Vaidyanadhan (1997) have suggested that the presence of developed etchplains at different elevations in Karnataka is the result of very ancient tectonic uplift.

7 According to Bremer and Sander (2000), it can take an estimated 2-6 million years to lower a plain by 100 m.

68

ASSESSING ANTHROPOGENIC SOIL EROSION WITH MULTI-SPECTRAL SATELLITE IMAGERY AN ARCHAEOLOGICAL CASE STUDY OF LONG-TERM LAND USE erosion and basal weathering responsible for these landforms. In the Tungabhadra study area, increased rates of surface stripping on pediment slopes and hills will outpace weathering. Thus, on hills of increased stripping we would expect less soil cover on the inselbergs and slopes and a greater exposure of weathered and unweathered rocks. In other words, more bare rock and less soil cover on the hills will reflect relative different degrees of stripping, which might also be associated with specific forms of land use that facilitate soil stripping. Alternatively, human activities that contribute to sedimentation – for example, the building of retention walls and the accumulation of habitation debris – might also be reflected by different degrees of sediment cover. Thus, we might also expect hills that show dense prehistoric occupation to contain more sediment cover and evidence less exposed residual rock – given the extent to which habitation debris and archaeological sediments have accumulated on them. These expectations can be assessed at the regional level with multi-spectral satellite imagery. Here we rely on ASTER satellite data to preliminary assess geomorphological differences between hills that show evidence for substantial prehistoric occupation and those that do not. However, systematic sampling of sedimentation and weathering profiles is necessary to build on this preliminary assessment.

94% of 183 ground-truthed points were classified correctly – a figure clearly acceptable for preliminary analysis9 (Figs 3-4). Given the expectations for assessing differential soil stripping highlighted above, the proportional composition of the surface materials of the hills and their upper pediment and talus slopes was computed10. The results were grouped according to whether or not hills showed evidence for occupation during the Iron Age. Figure 5 summarizes the proportional results accordingly (Fig. 5). Settled hills contain fewer pixels dominated by soil cover and more pixels dominated by exposed granite than unsettled hills. An examination of the 95% confidence interval for proportional differences between settled and unsettled hills suggest these disparities are not random, and that heavily occupied areas contain approximately ~10% less soil cover. In short, this observed pattern indicates that soil stripping has been greater on hills that were heavily occupied during the Iron Age than those that were not. Obviously, correlation does not equal causation. Hence, the suggestion that this pattern is a product of human activities should be taken with some skepticism, and is further discussed below. Discussion: Soil Distributions, Iron Age Pastoralism, and Landscape Modifications

Agro-

An approximate 10% difference in soil cover and exposed rock indicated by the satellite classification does not suggest that Iron Age land use was the dominant geomorphic agent in shaping these hills; indeed, millions of years of preceding weathering and sheet erosion are chiefly responsible for these geological products. However, the differences in soil cover and exposed rock between hills that were settled during the Iron Age – and it is important to note, possibly to a lesser extent during

Given the spectral properties of the primary surface constituents of the study region, an ASTER land classification based on visible, near-infrared, and midinfrared spectral reflection will distinguish between areas dominated by vegetated soil or sediment cover and bare oxidized or un-oxidized rocks. Using GRASS software, a surface classification was produced through a cluster analysis with 25 classification signatures8. These signatures were then used to assign each pixel into a class based on a maximum-likelihood discriminate analysis. Once completed, the classification map was compared to field notes to match the 25 computed classes with the dominant surface materials in that pixel. Subsequently, the number of classifications were reduced by grouping like cluster classes into the following categories; (1) water, (2) oxidized granite, (3) unoxidized granite, (4) biofilm covered rock, (5) organic and clay-rich irrigated soils, (6) vegetated red-brown sandy soils, and (7) freshly quarried, highly reflective granite. Once cluster classes were combined, the resultant final classification was compared to GPS observation points and field notes to assess its accuracy. It is possible that mixed-pixels have biased the analysis; however, a comparison between the final classification and field observations suggest that the classification is reasonably accurate. Indeed, more than

9 To be clear, the 94% correspondence between classified surface predications and ground-truthed observation points should not be read as a 94% correspondence at the pixel level across the entirety of the classification map. More detailed ground based reference mapping and systematic sampling is necessary to fully assess such ‘wall-to-wall’ comparisons (cf. Lillesand & Kiefer 1994). Rather, the correspondence indicates that 94% of the total GPS observation points that have been collected to date were correctly classified, which suggests the classification is relatively accurate. However, it should be noted that the 183 GPS points reported here are not evenly distributed across the region, but rather cluster on or around Iron Age sites visited during archaeological work. 10 To compare the differential degree of soil stripping on and around the residual hills in the study region it was necessary to isolate hills and the upper portions of their associated pediment slopes to create grouped, statistically viable samples. We categorized a Shuttle Radar Topography Mission (SRTM) digital elevation model of the study area using Wood’s (1996) GRASS module (r.param.scale) – designed to extract geomorphic parameters from a DEM by fitting a biavariate quadratic polynomial to a neighborhood matrix. The resultant ‘peaks’ and ‘ridges’ of Wood’s (1996) algorithm corresponded to residual hills and inselbergs, while ‘plain’ corresponded to the expansive peneplain and wash basins that characterize the region. We then isolated the boundary lines between ridges and planes. This effectively produced an outline around the base of each hill. Each hill feature was then extended by a uniform distance of 45 m to ensure that the geographical features being compared also included the upper portion of the pediment and talus slopes adjacent to each hill.

8 Ideally, a supervised classification seeded with ground-truthed pure pixels (i.e., training areas) would be preformed. However, at the time of analysis only a limited number of observations in all surface classes were available, and few could be verified as ‘pure pixels’. Nevertheless, a comparison of the resultant classification with ground-truthed observations indicates that the results of the cluster based classification reported here are reasonably accurate, and certainly allow for an assessment of preliminary pattering among surface materials.

69

Andrew M. Bauer & Kathleen D. Morrison the Neolithic Period also – and those that were not, suggests that prehistoric land use had some impact on erosional and depositional processes. Soil studies targeting animal grazing, traction, and certain agricultural practices such as clear-cutting vegetation and tillage, as chief causes of soil erosion provide useful starting points for thinking about the relationship between prehistoric occupation of these hills and apparent differences in soil stripping (cf. Ayala & French 2005; Mendiola 2004; Montgomery 2007). Vegetation clearance from animal grazing and agriculture tillage on and around the slopes of occupied hills could certainly have produced conditions that facilitated soil removal, particularly in the summer months when torrential monsoon rains might have encountered cleared and tilled ground. Forthcoming pollen analyses to document the vegetational history of the region will help to clarify these explanations, as the pollen data will allow for an assessment of how ancient activities impacted vegetation.

soils and exposed rock. This correspondence is further complicated by the fact that one might expect pronounced human effects on soil distributions around the base of the hills to potentially alter slope values by either amplifying the angle between lowering pediment slopes and the rock walls of residual hills, or decreasing the angle in locations where increased sedimentation has occurred. Moreover, increases or decreases in water runoff due to differences in the amount of sediment and soil available for infiltration on each hill will further affect stripping and slope development. In other words, it is possible that slight differences in slope degrees evidenced between occupied and unoccupied hills are also related to human activities. We are currently unwilling to eliminate the proposition that Iron Age inhabitants occupied more developed inselbergs with already less soil cover to explain differences between settled and unsettled hills. However, several additional points suggest that a priori differences in inselberg morphology cannot account entirely for the observed pattern of the distribution of soils and exposed rock in the study region. First, hills that were heavily occupied during the Iron Age show higher proportions of both oxidized granite and ‘fresh’ un-oxidized rock than hills that were not occupied during the Iron Age, which suggests more ‘recent’ exposure. Second, and most importantly, soil retention walls and check-dams are often found on Iron Age sites within the study area, implying that ancient inhabitants recognized soil erosion in areas of intensive land use (Fig. 6). Although it is clearly difficult to date such features, it is noteworthy that they are found on many Iron Age sites in the region and in association with Iron Age materials. For example, at Kadebakele there are a series of retention walls on the southern slope of the site leading down from the welldated Iron Age occupation area near the summit of the inselberg. Hence, it appears likely these features were constructed at some point during the site’s Iron Age occupation. While ancient inhabitants would have almost certainly conceptualized soil movement differently than a modern soil scientist, they nevertheless appear to have made efforts to retard it by bolstering slopes and building retention walls to inhibit the movement of soil materials. Clearly much more systematic work must be done on these retention features (including absolute dating where possible); however, current associations imply that the movement of stripped materials downslope was occurring during the period when the hills were heavily occupied. Hence, it should not be surprising that the results of the remote sensing analyses suggest that prehistoric activities contributed to the distribution of the soils in the study region.

In the meantime, however, equally parsimonious explanations for the discrepancies between soil cover and exposed rock between settled and unsettled hills cannot be entirely ruled out. One obvious explanation is that prehistoric inhabitants selectively occupied morphologically distinctive residual hills, perhaps with already less soil cover and more exposed rock. Several simple metrics allow for a partial evaluation of the morphological differences between occupied an unoccupied hills. For example, a comparison of the distribution of elevation cells from an SRTM digital elevation model shows an overall similar pattern between hills that were occupied and unoccupied during the Iron Age – though the curve for occupied hills is less smooth due to a much smaller sample size. In short, unoccupied hills have a slightly higher range of values than occupied hills; however, the mean elevations for occupied and unoccupied hills are nearly identical, separated by only 2 meters – 441 for unoccupied hills and 443 for occupied hills – suggesting that there is not a great difference in elevation between the hills considered above. However, the distribution of slope degrees on each hill derived from the same elevation model suggests there is a slight difference in the morphology of occupied and unoccupied hills, which might be related to differences in relief and soil cover. More specifically, cells of unoccupied hills yield a slightly lower mean slope (~2.0 degrees difference) than occupied hills, which is statistically significant according to the 95% confidence intervals relying on cell counts. However, the distributional curve for slope degrees of both unoccupied and occupied hills are very similar, both skewed to the right with similar ranges and means. While occupied hills show slightly higher peaks than unoccupied hills in the number of cells between 20 and 30 degrees, these slight differences in the slope distributions might be a product of the relatively much smaller sample size of settled hills – despite statistical significance. Thus, while there does appear to be a minor difference in the slope-morphology of settled and unsettled hills, it is difficult to attribute these small variations to the differential distribution of

Given this latter observation, it is important to note that the archaeological debris of occupation on many settled hills actually contributed greatly to the areal extent and depth of soil and sediment cover. In other words, the trash and building materials of ancient inhabitants have increased sedimentation in many localized settings on the inselbergs, and in some cases at their bases. The fact that heavily occupied hills still evidence proportionally less 70

ASSESSING ANTHROPOGENIC SOIL EROSION WITH MULTI-SPECTRAL SATELLITE IMAGERY AN ARCHAEOLOGICAL CASE STUDY OF LONG-TERM LAND USE shaping their material environments, which included expanding rock pools where less xeric vegetation grows (Bauer & Morrison 2008), as well as creating retention walls where down-slope moving sediment accumulated, likely facilitating agro-pastoral production. In other words, we are not arguing that Iron Age inhabitants degraded their material environment and resource base within the study region; rather, they recreated it, and arguably expanded it according to social strategies of production.

area of soil cover than unoccupied hills may imply that Iron Age activities had a slightly greater effect on the distribution of soils in the region than this preliminary analysis suggests. Along similar lines, the entirety of this study area was heavily occupied during the Late Middle Period, when the region was home to the capital of the Vijayanagara Empire. Comprehensive survey by Morrison and Sinopoli’s VMS project, in combination with pollen and sediment analysis by Morrison (1995), has shown that throughout this later period much of the study area was intensively used for settlement, agriculture, and animal grazing. Hence, one might expect the impact of later land use to obscure the effects of prehistoric activities on the local geomorphology. That there is still a pattern suggesting greater soil stripping on hills heavily occupied during the Iron Age suggests that the 10% difference in area soil cover between occupied and unoccupied hills might actually underestimate the role of prehistoric occupation in shaping this material environment.

Conclusion It is clear from surface observations as well as the results of the remote sensing analysis presented above that the current morphology of many of the hills and geomorphic features in the study region can be attributed to dynamic interactions between prehistoric inhabitants and complex environmental processes; they are not simply timeless products of a deep geologic past. Prehistoric inhabitants appear to have expanded rock pools, constructed small reservoirs where fine sediment and water accumulate, created weathering pits and grinding holes, and facilitated the movement of soil materials down slope while also trapping sediment behind retention walls for agropastoral production. Similarly, they re-worked regional vegetation patterns, both by design and by accident. Importantly, these activities were concurrent parts of strategies of (re)producing social relationships and symbolic resources. In other words, social, symbolic, and material resources were not depleted, but rather extended through the historical production of landscapes. In this sense, Iron Age inhabitants could accurately be described as producers of land; they were no less geomorphic agents than the processes of differential weathering and erosion described by Büdel.

Since prehistoric habitation debris has actually increased the amount of sediment and soil cover in specific locales on many of the occupied hills, it is reasonable to suggest that the differences in soil cover between settled and unsettled inselbergs can largely be attributed to soil removal on sloped areas of occupied hills that were seemingly used by inhabitants, but not directly settled on. This is further suggested by the fact that the proportional differences in freshly exposed rock appear to be particularly pronounced on the lower slopes of the occupied hills in the study area. These areas could likely have been used for grazing, or were perhaps even cultivated. Indeed, scholars have suggested that prehistoric inhabitants likely harvested rainfall runoff by focusing agricultural activities around the bases of the residual hills in the region (e.g., Morrison 2009). Likewise, soil retention features are found in such settings. Hence, the initial suggestion that animal grazing and agricultural tillage practices on areas of the hills and their lower slopes have contributed to the observed pattern of soil distributions in the study area appears to have some credence. However, it is import to reiterate that this in only a preliminary conclusion, and that forthcoming pollen analyses, botanical studies, and additional geoarchaeological observations will clarify this picture.

Bibliographical References Adams, G. (ed.) (1975) Planation Surfaces, Benchmark Papers in Geology, vol. 22. Dowden, Pennsylvania. Ayala, G. & French, C. (2005) ‘Erosion modeling of past land-use practices in the Fiume di Sotto di Troina River Valley, north-central Sicily’. Geoarchaeology: An International Journal, 20: 149-167. Bauer, A. M. (2007) ‘Producing “Nature” in early South India: Relating rock pools, weathering pits, and soil distributions to place-making in Iron Age northern Karnataka’. 106th Annual Meeting of the American Anthropological Association. Bauer, A. M.; Johansen, P. G. & Bauer, R. L. (2007) ‘Toward a political ecology in early South India: preliminary considerations of the socio-politics of land and animal use in the Southern Deccan, Neolithic through Early Historic Periods’. Asian Perspectives, 46: 3-35.

In the meantime, we cannot rule out the possibility that ancient inhabitants occupied more developed inselbergs with already less soil cover. However, it appears that prehistoric activities associated with stock herding and agricultural production contributed to some soil stripping evidenced on parts of the settled hills and their lower slopes. Given this latter interpretation, it is important to avoid the trope that social strategies of stock-herding are generally destructive to the environment by promoting over-grazing and resource depletion (Gadgil & Guha 1992; Hardin 1968). To be clear, we are not suggesting that Iron Age agro-pastoral activities were ‘profligate’ uses of material resources (sensu Gadgil & Guha 1992). Iron Age inhabitants appear to have been active in 71

Andrew M. Bauer & Kathleen D. Morrison Bauer, A. M. & Morrison, K. D. (2008) ‘Water management and Reservoirs in Southern India and Sri Lanka’. In Selin, H. (ed.), Encyclopedia of the History of Science, Technology, and Medicine in Non-Western Cultures: 2207-2214. New York. Bauer, R. L. (2007) Animals in Social Life: Animal Use in Iron Age Southern India, VDM Verlag Dr. Muller, Saarbrucken. Boivin, N. (2004) ‘Rock art and rock music: petroglyphs of the South Indian Neolithic’. Antiquity, 78 (229): 38-53. Boivin, N., Korisettar, R. & Fuller, D. (2005) ‘Further research on the Southern Neolithic and the Ashmound Tradition: The Sanganakallu-Kupgal Archaeological Research Project interim report’. Journal of Interdisciplinary Studies in History and Archaeology (Allahabad) 2 (1): 63-92. Bremer, H. & Sander, H. (2000) ‘Inselbergs: geomorphology and geoecology’. In Porembski, S. & Barthlott, W. (eds), Inselbergs: Biotic Diversity of Isolated Rock Outcrops in Tropical and Temperate Regions. Berlin: 7-35. Büdel, J. (1957) ‘Die “doppelten Einebnungsflächen” in den feuchten Tropen’. Zeitschrift für Geomorphologie, 1: 201-288. Büdel, J. (1977) Klima-Geomorphologie, Stuttgart. Büdel, J. (1982) Climatic Geomorphology, Princeton. Devaraj, D. V.; Shaffer, J. G.; Patil, C. S. & Balasubramanya (1995) ‘The Watgal excavations: an interim report’. Man and Enivronment XX (2): 57-74. Fuller, D. Q.; Korisettar, R. & Venkatasubbaiah, P. C. (2001) ‘Southern Neolithic Cultivation Systems: A Reconstruction based on Archaeobotanical Evidence’. South Asian Studies, 17:171-187. Gadgil, M. & Guha, R. (1992) This Fissured Land: An Ecological History of India. New Delhi. Hardin, G. (1968) ‘The tragedy of the commons’. Science, 162 (3859): 1243-1248. India Meteorological Department (1981) Climatological Atlas of India: Part A (Rain Fall). Government of India, New Delhi. Kajale, M. D. (1984) ‘Appendix: Botanical remains from Veerapuram’. In Sastri, T. V. G., Kasturi Bai, M., Rao, J. V. P. (eds) Veerapuram: A Type Site for Cultural Study in the Krishna Valley, i-xi. Hyderabad. Kajale, M. D. (1989) ‘Ancient grains from excavations at Hallur (1976 Season), District Dharwar, Karnataka’. Bulletin of the Deccan College Research Institute 47-48: 123-128. Lillesand, T. M. & Kiefer R. W. (1994) Remote Sensing and Image Interpretation (3rd Edition). New York. Mendiola, M. Q. (2004) ‘Highland grassland vegetation in the northwestern Andes of Argentina: Vegetation structure and species composition in relation to grazing’. Mountain Research and Development, 24: 243-250.

Montgomery, D. R. (2007) ‘Soil erosion and agricultural sustainability’. Proceedings of the National Academy of Sciences of the United States of America, 104: 13268-72. Morrison, K. D. (1995) Fields of Victory: Vijayanagara and the Course of Intensification, Contributions of the University of California Archaeological Research Facility no. 53. Berkley. Morrison, K. D. (2009) Daroji Valley: Landscape, Power, and Place in the Making of a Dryland Reservoir System. Delhi. Radhakrishna, B. P. & Vaidyanadhan, R. (1997) Geology of Karnataka. Bangalore. Sinopoli, C. M. & Morrison, K. M. (2003) ‘Early Historic landscapes of the Tungabhadra Corridor’, Paper presented at the 32nd Annual Conference on South Asia, 24-26 October, Madison. Sinopoli, C. M, & Morrison, K. D. (2007) The Vijayanagara Metropolitan Survey: Volume 1, Museum of Anthropology (University of Michigan Memoirs, 41). Ann Arbor. Thomas, M. F. (1994) Geomorphology in the Tropics: A Study of Weathering and Denudation in Low Latitudes. New York. Wood, J. (1996) The Geomorphological characterisation of Digital Elevation Models. Dissertation, Department of Geography, University of Leicester, U.K. Twidale, C. R. (1981) ‘Granitic inselbergs: Domed, block-strewn and castellated’. The Geographical Journal, 147 (1): 54-71.

72

ASSESSING ANTHROPOGENIC SOIL EROSION WITH MULTI-SPECTRAL SATELLITE IMAGERY AN ARCHAEOLOGICAL CASE STUDY OF LONG-TERM LAND USE

Fig. 1 - Location map and SRTM digital elevation model.

Fig. 2 - Thin organic rich soil undergoing erosion.

73

Andrew M. Bauer & Kathleen D. Morrison

Fig. 3 - Aster based land surface classification.

Fig. 4 - Classification error matrix.

74

ASSESSING ANTHROPOGENIC SOIL EROSION WITH MULTI-SPECTRAL SATELLITE IMAGERY AN ARCHAEOLOGICAL CASE STUDY OF LONG-TERM LAND USE

Fig. 5 - Proportional composition of surface materials.

Fig. 6 - Retention feature associated with an Iron Age settlement.

75

REJECTING THE STATE: REFLECTIONS ON ANCIENT INDIAN THOUGHT AND THE ORGANIZATION OF HARAPPAN SOCIETY Piotr A. Eltsov

branches of research in several fields of knowledge. Ancient history and archaeology would suffer the most, for the states of ancient China, Mesoamerica, Mesopotamia, South Asia, the Middle East, and even of Medieval Europe would become the fruits of scholarly imagination. Amusingly, neither archaeologists nor political scientists refer to the works of each other. Both sides seem either to ignore or to be ignorant of the research undertaken by their opponents. To set the ground for my own argument, it is therefore worth revisiting a few defining characteristics of the term ‘state’ in Western intellectual tradition.

The identity of the Harappan civilization remains one of the most controversial issues in South Asian protohistory. It is still debated whether the Harappans lived in a state, chiefdom, or a unique sociopolitical system; whether caste, hierarchical ranking, or heterarchical relations characterized their social structure; and whether it is appropriate to allude to Vedic literature as a legitimate source for the reconstruction of the Harappan ethnolinguistic identity. This article addresses the issue of the state in the context of Harappan archaeology and ancient Indian sociopolitical thought. It argues that the concept of the state and the related neo-evolutionary models of culture change are unable to create an adequate picture of the Harappan sociopolitical organization. It proposes that some of the ancient Indian sociopolitical ideas provide a better option for explaining the historical uniqueness of Harappan society.

The Latin word ‘status’ and its linguistic equivalents which are usually seen as the predecessors of the modern concept of the state are found in a large variety of texts dating to the Middle Ages, Renaissance, and the Age of Reformation. Historians, philosophers and political scientists debate on how, when, and in which exact contexts this term obtained its modern conceptual meanings. One must say that there is a bit of irony in this debate. Medievalists see the emergence of the term ‘state’ in the works of Thomas Aquinas, Erasmus, and other great theoreticians of the Middle Ages (Kantorowicz 1957; Post 1964; Harding 1994, 2002). Specialists in the Renaissance tend to discover all the great ideas in the writings of humanists, and trace the emergence of the concept of the state from the books of Niccolò Machiavelli (Dowdall 1923; Meinecke 1924; Cassirer 1946). British political scientists look for precise definitions in the works of their compatriot Thomas Hobbes (Dyson 1980; Skinner 1978, 1989). Thus, not unlike the stereotypical archaeologist who tends to discover all of the important inventions and transitions in the cultural sequence of the site that he or she excavates, scholars writing on the emergence of the term ‘state’ seem to find all the answers in the texts that they know the best.

The Idea of the State One of the most striking features of the theory of the state is its incredible intellectual malleability. The notions of the state held in different fields of knowledge are not only different, they are mutually exclusive. The majority of archaeologists specializing in early complex societies believe that the first states emerged in the Near East in the late 4th millennium BC. Consequently, they find archaic states throughout the entirety of the ancient world; i.e. from China to Mesoamerica and from the Near East to Sub-Saharan Africa. Moreover, the roots of statehood are looked for in the cultures that precede the first civilizations by thousands of years. One archaeologist has suggested that ‘most questions about the origins of the state could be answered very simply’, by discovering the origins of Homo sapiens sapiens; and even that ‘it has been always accepted that the domestication of plants and animals was of central importance in the transformation of society to eventual statehood’ (Gamble 1986: 22, 26). Typical of archaeology, this line of thinking goes back to the ideas of V. Gordon Childe, who aside from having defined the key transitions in prehistory, drew a variety of conceptual links between sedentism, agriculture, cities, states, and the idea of civilization.

Turning to the concept of the state, one finds even a more intensive discussion between scholars of different orientations. The philosopher Peter Steinberger conceives of the state as an idea, which is ‘embedded in a complex, multi-layered theory of human thought and our relationship to the world’. The state, in Steinberger’s definition, is ‘a structure of self-understanding. It tells us who we are’ (Steinberger 2004: 324). Archaeologist Bruce Trigger, on the contrary, defines the state simply as ‘a politically organized society that is regarded by those who live in it as sovereign or politically independent and has leaders who control its social, political, legal, economic, and cultural activities’ (Trigger 2003: 92). Such a simplification of the concept of the state stems

Meanwhile, there is an academic tradition, which treats the state as a phenomenon that first appeared in the sixteenth century AD. The political scientist Andrew Vincent for example asserts that ‘there is no Greek or medieval State’ and there is a great deal of forms of sociopolitical organizations ‘to be included under the rubric of a “stateless” society’ (Vincent 1987: 10). Following Vincent, we would need to discard whole 77

Piotr A. Eltsov from the never-ending efforts of cultural evolutionists to discern the course of historical processes (for an illustrative and recent restatement of the evolutionary theory, see Claessen 2000); and reflect the understandable desire by archaeologists to link the differentiated, internally specialized, and system-like concepts of the state with settlement hierarchies defined through archaeological surveys. Research by the archaeologist Henry Wright exemplifies this approach at its best. Defining the state as ‘a sociocultural system in which there is a differentiated, internally specialized, decision making sub-system’, Wright finds an easy way to connect archaeological data directly to the theory of the state (Wright 1978: 56). All that one needs to discern the state then is a multi-tier settlement hierarchy, i.e., the existence of several levels of settlements increasing in size and located within a regular distance of each other.

of world history’ (der Geist, der sich im Prozesse der Weltgeschichte seine Wirklichkeit gibt) (Hegel 1821: 249-250). Remarkably, Hegel was well aware of the abstract nature of his ideas; in the notes to Philosophie des Rechts he clarified that as ‘an overwhelmingly perpetual and necessary being of the Geist’ (das an und für sich ewige und notwendige Sein des Geistes), the idea of the state does not have much to do with the historical origin of the state in general or with any particular state with its rights and customs (Hegel 1821: 242-243). Hegel’s students later added that when defining the idea of the state, ‘one must not have in his mind any particular state, or particular institution, but must rather reflect on the idea, this actual God, by itself’ (Hegel 1833: 320). Ironically, this idea of actual God was not that different from the idea of the state by the evolutionists. In both cases, the state was the essence of history, the realization of cultural change, and the unifying core of sociopolitical diversity. When Karl Marx famously noted that Hegel’s dialectics were ‘standing on its head’ and had to be ‘turned right side up’, he clearly recognized the value of Hegelian ideas for the materialist interpretation of historical process (Marx 1872: 822).

Meanwhile, there are many studies of archaic states that give no reference to theory. Some authors seem to take the idea of the state for granted and believe that it is possible to study concrete ‘states’ without getting into the intricacies of the theoretical legacy of the idea of the state. Taking into consideration the intellectual history of the idea of the state, I believe that most of such studies are implicitly evolutionary, hierarchical, linear, and teleological; for just to accept the idea of the state as a tool for the analysis of diverse forms of sociopolitical organization implies at least four assumptions. First, having been seen by the generations of scholars as a complex form of sociopolitical organization, the state has become an element of linear schemes of political transformation. This linearity implies that the forms of sociopolitical organization that preceded the state or coexisted with the state were not as complex as the state. Second, in spite of its tremendous diversity and malleability, the concept of the state is closely associated with the evolutionary theory (e.g., with the progressions of savagery-barbarism-civilization or band-tribechiefdom-state), and it is doubtful that either anthropologists or historians will ever be able to separate the two. Third, the use of the concept of the state in the analysis of diverse forms of sociopolitical organization throughout the world from the late 4th millennium BC until the present day implies a very high degree of universality in human history; for example that the sociopolitical systems of Sumer, ancient Egypt, and Classical Maya shared something essential in common, not only with one another but with the political cultures of 18th century France, 20th century Russia, and 12th century Japan. Fourth, the concept of the state implies the existence of hierarchical relations between social groups and the functioning of a governing body. In other words, state-level societies are not only socially divided but are also structured along hierarchical principles with a fulltime government solidifying these principles.

Fortunately or unfortunately, the history of humankind is more prosaic and less teleological. As far as the phenomenon of the state is concerned, it is clearly not as universal as some of the great thinkers believed, and there must have existed, in the past, other forms of complex sociopolitical organization, of which we know very little and which deserve to be conceptualized in terms different from those of the traditional Staatswissenschaft or any of its modified versions (for example, Southall 1988). To renounce the concept of the state as a tool for the study of early complex societies would, of course, require the efforts of many specialists. Such efforts are underway (Yoffee 2005, Pauketat 2007) and this article is hoped to provide a new perspective to this important undertaking. A particularly inspiring example of how the alternatives to the state are being gradually embraced by academics can be found in the writings of anthropologist Peter Skalnik. Initially, Skalnik co-edited two well-acclaimed volumes that dealt with states in ancient history and made an impact on the archaeology of complex societies (Claessen and Skalnik 1978a, b; 1981). The concept of the state as proposed in these volumes was evolutionary, structural and above all under the heavy influence of Marxism. Yet, soon after the publication of the second volume and after conducting much ethnographic research in Africa, Skalnik refuted the universal applicability of the concept of the early state (Skalnik 1983: 11). Even Claessen who continues to believe that the term ‘early state’ remains useful for the conceptualization of ancient sociopolitical structures (Claessen 2000: 187) has recently admitted that it is ‘no more than a scholarly construction’, which ‘does not necessarily conform to any actual state, but rather brings together the most important characteristics of early states in a coherent scheme’ (Claessen and Oosten 1996: 2-3, 7). The question then arises whether the gap between this coherent scheme and the actual archaic polities provides us with anything other but a structure for the sake of the structure, which, in

Considering these four traits together, one cannot help concluding that to those who use the concept of the state universally and throughout human history, it remains, implicitly or explicitly, the Hegelian Spirit, or in Hegel’s own words, the Geist ‘which realizes itself in the process 78

REJECTING THE STATE: REFLECTIONS ON ANCIENT INDIAN THOUGHT AND THE ORGANIZATION OF HARAPPAN SOCIETY civilization had monumental structures as well. Perimeter walls, public buildings, water reservoirs, and most importantly mud-brick platforms required massive consumption of labor and certainly qualify to be called ‘monumental architecture’. The sites of Mohenjo-Daro and Dholavira provide, by far, the best examples of such structures. Nonetheless, after reviewing the available data, one is left with the impression that the message that the Harappan architecture was meant to convey was fundamentally different from that of other ancient civilizations. First, no Harappan structure was as visually imposing and intimidating as some of the Mesopotamian, Mesoamerican, or Egyptian edifices. Second, the monumentality was characteristic mostly of walls and platforms. Occasional public buildings, such as the ‘granaries’ of Harappa and Mohenjo-Daro, are poorly understood and cannot be properly visualized. Third, despite the reasonable claims that some houses of Mohenjo-Daro and other sites may have belonged to the local nobility, the overall message that the Harappan domestic architecture conveys is that of relative homogeneity and the lack of monumental expressions of wealth.

essence, is not that different from Hegel’s Geist or Max Weber’s Idealtyp. The Puzzles of Harappan Archaeology As long as the Indus writing is not deciphered, our main source of understanding the Harappan society remains the archaeological record. Yet archaeological record is ambiguous and can be very misleading. The neoevolutionists have debated whether the Harappan society should be characterized as a chiefdom (Malik 1968; Fairservis 1984, 1992) or a state (Jacobson 1986; Kenoyer 1994, 1997, 2000). Those who reject the neoevolutionary paradigm have pointed out that the Harappans may have lived in a society that was very complex and unique yet was neither state nor chiefdom (Possehl 1998, 2002). Yet, in one way of another, all the authors refer to the same set of puzzling traits that can be summarized as follows: Settlement Layout. One of the most distinctive features that define the layout of Harappan settlements is segregation. Most of the sites are divided into two or more sectors encircled by walls. Mohenjo-Daro is a controversial case, yet recent research by Italian archaeologists shows that some or possibly all the mounds of Mohenjo-Daro had perimeter walls (Leonardi 1988). In addition, most of the Harappan sites reveal occupational deposits beyond the city walls. In other words, these sites have sectors not protected by walls but symbolically separated from the rest of the site. Another distinctive feature of Harappan settlements is the elevation of certain sectors and structures by means of mud and mud-brick platforms. These platforms can be seen as parts of fortification complexes as they are connected to the perimeter walls both technically and functionally. As I have argued elsewhere, the function of walls and platforms on Harappan sites was predominantly symbolic (see Eltsov 2008): they created threedimensional and segregated worlds characterized by restricted access and hidden monumentality. No military, ecological or any other practical reason would justify the splitting of the western sector at Kalibangan, the creation of a long passage in front of the northern gate at Dholavira, or the restriction of access to certain areas in Harappa. Likewise, no ecological concerns would require the construction of platforms that had no structures on top of them (Kalibangan), were different in height (MohenjoDaro), or covered only partial areas of the site (Banawali).

Territorial Expansionism. It has been argued that soon after their initial formation the earliest civilizations underwent similar patterns of territorial expansion. The discovery of Harappan sites in Northern Afghanistan and Makran Coast show that the earliest civilization of South Asia was not the exception. In the third millennium BC, the Harappan civilization was in fact the largest in the world. This unprecedented geographical span may have reflected a variety of sociopolitical and ideological processes. Yet one thing is clear. The Harappan culture could not develop simultaneously from Gujarat to Northern Afghanistan and territorial expansionism must have played an important role in the formation of the Harappan style and identity. Moreover, the difference of material cultures that characterize the Harappan outposts indicates that the nature of Harappan expansion may have followed different patterns in different areas (see Dales & Lipo 1992: 156; Ali & Rahim 2001). As I have recently suggested, the lack of certain artifacts and technologies on some outposts may be indicative of the absence of sociopolitical structures and mechanisms for which these artifacts and technologies stand (Eltsov 2008: 142-144). The nature of the Harappan territorial expansion thus is an important yet poorly explored topic. The least that it shows is that territorial expansionism was a key element of its sociocultural and political identity.

Monumental Architecture. Since V. Gordon Childe concluded that monumental architecture was one of the defining traits of the ancient city (Childe 1950), archaeologists have written extensively on its role in the emergence of ancient civilizations, and for the most part concluded that monumental architecture reflected a variety of socio-cultural process (e.g., the crystallization of power, the consumption of labor, the formation of local identities, and the growth of religious domination) and was a key constituent in the process of visualization of ancient civilizations (see Trigger 1990). The Harappan

Intercultural Relations. The concept of the interaction sphere as defined in Anglophone literature by Joseph Caldwell (1964) was among the themes that occupied the minds of archaeologists in the second half of the 20 th century. By the time the earliest civilizations emerged in Egypt and Mesopotamia, the entirety of the Eurasian World was interconnected. Yet, one is perplexed by the degree, to which the Indus civilization was involved in the interaction spheres of the Larger Near East, Central Asia, and the Eurasian Steppes. A significant amount of Indus material has been found in the Arabian Gulf, 79

Piotr A. Eltsov think in this direction. No proper cemetery has been discovered in Mohenjo-Daro, nor at several other important sites. The cemeteries excavated in Harappa, Kalibangan, Lothal, and most recently Dholavira posit a number of intriguing questions. First, the number of buried individuals is strikingly lower than the number of residents of the relevant sites. Second, most of the buried individuals display signs of a stable group of population which did not suffer nutritional stress. Third, in spite of some variations in the layout and content of individual graves, the Harappan burials strike the observer as lacking lavish expressions of social distinction. Even seals, inscribed objects, and other possible valuables are absent in the graves of the Harappans. The questions then arise as to whether being ‘properly buried in the ground’ was a social privilege in the Harappan society, whether there was another way of disposing the dead in the Harappan cities, and whether the Harappan ideology imposed restrictions on the display of wealth. The answers to these questions lie within some of the cultural sequences that preceded the emergence of the Harappan civilization. The cultural sequence of Mehrgarh, for example, reveals intriguing changes in burial customs. The graves of the Neolithic period display clear signs of differentiation. Simplicity and modesty are characteristic of funeral rites during the Chalcolithic period. The cemetery of the occupational period, which immediately precedes the Harappan civilization, is the most modest (Sellier 1987). Thus, intriguingly, with the growth of complexity, the display of lavish wealth in burial ceremonies becomes for some reason unacceptable.

Mesopotamia, and Central Asia. The Indus or Indus-like seals, beads, figurines, black-slipped jars, inscribed shards, metal implements, weights, and other artifacts of Indus origin have been found west of the Indus. In addition, the Mesopotamian texts of the Early Dynastic period report a variety of products as coming to Mesopotamia from Meluhha, a toponym often associated with the Harappan civilization. On the other hand, it is commonly known that there is very little material of western origin in the Indus civilization. The Bactrian hoard discovered in Quetta (Jarrige 1987) and the occasional objects of Central Asian origin found in the upper levels of several Harappan sites date to the last centuries of the Harappan civilization, the time when the Harappan world had begun to undergo drastic sociocultural changes. Religious Practices. It has been noted that the cities of the Harappan civilization oddly lack archaeologically identifiable temples and lavish religious paraphernalia. The only structure that could be called monumental and is most likely associated with ritual practices is the Great Bath of Mohenjo-Daro. Since the time of John Marshall, scholars have therefore argued for the domestic nature of Harappan religion and pointed to similarities between the Harappan and later historical beliefs and practices. The evidence referred to in support of these arguments usually includes a standard set of archaeological features; viz., wells and water devices, hearths, ornaments and jewelry, figurines, iconographic motifs, burials, and the alleged ritual objects. The ritual practices that are hypothetically inferred on the basis of these traits include: 1) The worship of proto-Shiva and the beginnings of Shaivaism; 2) The dichotomy of the male god and the female goddess and the beginning of Shaktism; 3) The notion of ritual purity; 4) The sanctity of water, and the performance of ritual ablutions; 5) The practice of Yoga; 6) The cult of mother goddesses; 7) The worship of fire; 8) The sacredness of certain animals, possibly of a cow, and the nagas (snakes); 9) The worship of human genitalia (linga and yoni); 10) The worship of trees, particularly of the papal tree; 11) The performance of human and animal sacrifices, and last but not least; 12) The making of an intoxicating concoction and the burning of incense. As I have suggested elsewhere, the three-dimensional settlement layout reinforced by segregation, restrictions of access, and the elevation of certain areas can also be seen as a reflection of ritual practices (Eltsov 2008). The best examples of such threedimensionality are the Stupa Mound at Mohenjo-Daro and the southern section of the western mound at Kalibangan.

Artistic Traditions. The Harappan artistic traditions are expressed most vividly in figurines, sculptures, masks, personal ornaments, glyptics, and pottery. Depending on one’s definition of art, one could also consider puppets, toy carts, whistles, dice and other decorated and miniature objects so typical of the Harappan material culture as forms of artistic expression. Technologies involved in the making of jewelry are particularly stunning; some cannot be replicated. Raw materials brought from distant sources indicate the value of some objects, while the amounts of jewelry displayed on terracotta figurines point to their wide use in daily life. Based on the study of involved techniques and trade routes, Kenoyer has argued that seals, ornaments, and pottery may have served as symbols of wealth and status (Kenoyer 2000). Nevertheless, the nature of Harappan artistic expressions remains puzzling. First, the Harappan sites have produced a very small amount of high quality sculpture. All pieces of sculpture come from Harappa, Mohenjo-Daro, and Dholavira; while the overall stylistic and artistic diversity of these pieces stand in striking contrast to their number. Second, there is a clear lack of what one could call the monumental and visually imposing artistic expressions of religious, social, and political nature. For example, there is no truly monumental sculpture, typical of other ancient civilizations. Neither is there any lavish sociopolitical or religious paraphernalia. Given that the skills involved in the production of some objects of Harappan art were highly sophisticated, it is doubtful that the Harappans

Funeral Rites. To the archaeologist, burials are of great importance because they convey invaluable information on the perception of death, attitude to afterlife, and social organization. At the same time, burials can be very deceptive. As we know from ethnographic and historical examples, modest burials may be the products of socially complex societies that for some ethical or religious reasons place restrictions on the display of wealth. The excavated burials of the Harappan society lead one to 80

REJECTING THE STATE: REFLECTIONS ON ANCIENT INDIAN THOUGHT AND THE ORGANIZATION OF HARAPPAN SOCIETY unique society. None of the four traits highlighted above as characteristic of the idea of the state seem to be helpful for understanding the puzzles of Harappan archaeology. The linear nature of the concept of the state and its close association with the evolutionary theory prevents one from thinking about the Harappan society in its own terms. There was no straightforward progression from the pre-Harappan to Harappan and even more so from the Harappan to post-Harappan cultural entities. Neither was there a simplistic correlation between the sociopolitical organization of the Harappans and other complex cultural entities that existed in the Eurasian world in the third millennium BC. A high degree of universality of the concept of the state is not particularly enlightening either, as the least that the aforementioned puzzling traits demonstrate is the uniqueness of Harappan society on the grand scale of historical process. The presence of social hierarchy is a more complicated issue. Given its scale and complexity, the Harappan society must have been divided in a large number of professional, sociopolitical, and religious groups. These groups were augmented by a multitude of identities and local worlds that are characteristic of any complex human organization. It has been argued that some of the Harappan artifacts ‘reflect the distribution of identical, shared symbols along the vertical socioeconomic axis’ (Kenoyer 2000: 104). Even if this is the case, the question remains whether the existence of such a socioeconomic axis shows sufficiently the existence of the state. Archaeologists have convincingly argued that heterarchical relations may maintain complex sociopolitical organization without a clearly defined government and with some elements of hierarchical structures (see Ehrenreich et al 1995). The groups, identities, and other local worlds of Harappan society were likely structured in this manner as well.

were unable to create monumental sculpture, elaborate relieves, and lavish paraphernalia. Moreover, monumentality was characteristic of other expressions of Harappan culture. For example, platforms and walls were monumental. In delineating space in three dimensions, they created mesocosms with restricted access to certain areas (for an analytical analogy, see Eliade 1949, 1957). None of these monumental endeavors however found a clear expression in art. Whatever its social or political function was, the Harappan art was lacking in terms of monumental appearance. The art of Neolithic Catal Huyuk was visually more dominant and intimidating. The Transmission and Preservation of Information. The role of literacy in the genesis of complex societies has been the focus of a long-standing debate among anthropologists and historians. Anthropologist Goody and his followers believe that writing had, ‘historically, been responsible for the evolution of new forms of discourse… that reflected a new, more subjective and reflective frame of mind’; and most importantly that it led to the ‘new forms of social organization, of states rather than tribes’ (Olson 1991: 251; also see Goody & Watt 1963; and Goody, 1968, 1987, 2000). The opponents of Goody argue that orality enabled by highly sophisticated mnemonic devices could be a very effective tool for attaining complex forms of sociopolitical organization. Moreover, the written texts in certain cultural environments can be more malleable than oral traditions, while oral traditions can be more fixed than the written texts (see Falk 1990; Fuller 1984; Olson & Torrance 1991; Rocher 1994; Witzel 1995). For understandable reasons, the example of ancient South Asia is brought up by both sides of the debate frequently and vigorously, in particular with regards to the cultural traditions of early historic South Asia. Meanwhile, the role of orality in the functioning of the Harappan society remains barely explored (Malik 1968). As the story stands now, the Indus Script is undeciphered and it is still debated whether it was a genuine writing system (Farmer et al 2004). The genesis and maintenance of the Harappan sociopolitical system must nevertheless have required the existence of elaborate techniques to preserve and transmit information. If Jansen is right in asserting that in order for Mohenjo-Daro to have been built, ‘any earlier settlement must have been abandoned and the design of the city must have been perfectly planned by master builders’ (Jansen 1994: 271), the level of societal organization and technical knowledge of the people of the Indus Valley in the 26th century BC must have been very high. Given that there is no evidence of script prior to the construction of Mohenjo-Daro, this assertion poses an intriguing puzzle. Could the technology, social organization, and political culture necessary for the rapid construction of the city such as Mohenjo-Daro be elaborated, preserved, conveyed, and implemented without a script?

Ancient Indian Sociopolitical Thought Ancient India is known for its sophisticated sociopolitical ideas that are found in a variety of Sanskrit and Pāli texts, e.g., in the Buddhist Canon, Kauṭilīya Arthaśāstra, the Law Books of Manu and Yājñavalkya, Mahābhārata, Valmīki Rāmāyaṇa, Nītisāra, Nītivākyāmṛta and Śukranītisāra. These ideas have been the focus of numerous publications including several substantial monographs (Altekar 1949; Drekmeier 1962; Ghoshal 1959; Hillebrandt 1923; Law 1921; Rau 1956; Scharfe 1989; Sharma 1968; Spellman 1964). Most authors have emphasized that aside from reflecting critical realities of their time, ancient Indian sociopolitical thought attained a high level of abstraction, theorizing, and intellectualism. The investigation of this thinking has followed two main directions; i.e., the study of specific ideas and theories as the things in themselves; and the study of specific ideas and theories in comparison with other political ideas and theories of the pre-modern world. As a conceptual source for the writing of South Asian history, ancient Indian sociopolitical thinking has nonetheless received minimal attention; most studies produced by anthropologists and historians in the past have relied heavily on the theories of western secular science, which emerged and developed in the context of modernity. There is, therefore, an

Taking into careful consideration each of these puzzling traits, one cannot help but question whether it is appropriate to use the concept that reflected primarily the intellectual realities of west-European political tradition in the analysis of a very ancient, very complex, and very 81

Piotr A. Eltsov focus of much research. The theories of the Arthaśāstra that deserve particular attention are the daṇḍanīti, saptāṅga and maṇḍala. The daṇḍanīti is a teaching on the mechanisms of coercion and power. The saptāṅga is the definition of polity. The maṇḍala theory conceptualizes peer-polity interaction. Other themes that are important in the Arthaśāstra are the varṇāśrama dharma, the idea of kingship and government, the separation of politics from theology, the concept of vārtta, and suchlike. Many similar ideas are found in the Manu- and Yājñavalkyasmṛti, the texts that in many respects continue the line of the Dharmasutras and deal in detail with the varṇāśrama dharma. The teachings of the daṇḍanīti, saptāṅga, and maṇḍala are discussed in the Manu and Yājñavalkya, yet with less detail than in the Arthaśāstra. The theory of kingship, distinction between sacred and temporal power, the differentiation of the Dharmaśāstra and Arthaśāstra as sources of sociopolitical theory, and possibly the idea of organic unity of society are also of great significance in these texts. The great Sanskrit Epics present the next important source of sociopolitical ideas in the Sanskrit literary tradition. The dharma and in particular rājadharma become the key themes of the Mahābhārata and Rāmāyaṇa. The norms of the daṇḍanīti as a broadly defined science of government are discussed in great detail as well. The idea of kingship is presented most vividly in the Rāmāyaṇa as a highly idealized story of the reign of Rāma (rāmarājya). A strong ethical component or the relations of politics to ethics is of importance in both the epics, yet is particularly noticeable in the Rāmāyaṇa. One could even argue that the Rāmāyaṇa advocates a kind of contractual relation between the ruler and the ruled; as despite the semi-divine nature of the ruler, his obligations and severe penalties for wrongdoing are clearly stated. The later legal texts of the so-called Nīti tradition develop the ideas presented in the Śāstras and Epics. The Nītisara of Kāmandaka is repetitive of the Arthaśāstra. A particular emphasis is placed on the idea of kingship and daṇḍa as the foundation for the proper functioning of society. The theories of saptāṅga and maṇḍala are reiterated quite literary from earlier sources. The separation of morals from politics is also quite typical of the Nīti texts and, in this sense, they follow the Arthaśāstra quite literary. One would need to emphasize however that the Machiavellian separation of morals from politics does not imply the acceptance of absolute and unlimited authority. No text of Brahminical tradition had ever advocated the idea of authority similar to that of European absolutism or archaic despotism. Even in the Rāmāyaṇa, known for its glorification of the reign of Rāma, authority had limited nature and under exceptional circumstances people were allowed to overthrow and even slay the king.

evident lack of research that would approach the intellectual legacy of ancient Indian sociopolitical ideas as a source for the creation of our own analytical models. When approaching the study of the Harappan civilization, this lack is particularly striking and unfortunate. As I have argued elsewhere, the sociopolitical ideas of ancient India may present a viable alternative to the epiphenomenal ideas of western academia and fill the intellectual vacuum that is sensed in the study of the earliest phases of South Asian history (Eltsov 2008). Fundamentally different from the a-priori identifications of literary and political traditions with archaeological data (a method that is so vigorously employed in recent writings by some Indian archaeologists and historians), the testing of ancient ideas and theories against archaeological data considers the indigenous models of thinking yet does not imply any direct ethnolinguistic or religious correlations. Even a brief assessment of ancient Indian sociopolitical ideas from this point of view demonstrates their incredible intellectual depth and value as a potential source for our own theorizing. The earliest reflections of sociopolitical thinking in South Asia are found in Vedic literature. The Vedic concepts that have been brought up as socially important are the classification of society into four varṇas; the rule of cosmic law (rta and vrata); the idea of temporal kingship (rāja); the interaction of Kṣatriyas and Brahmins (brahmakṣatra); and the separation of sacred and temporal powers. In addition, it has been argued that a crucial role in the genesis of Vedic civilization was played by the śrauta ritual, which was responsible for the key reorganization of the Vedic society (Witzel 1997: 48). One could add that some of the Vedic rituals, the oral means of their preservation, and the concept of yajña can be seen as peculiar forms of sociopolitical thinking as well. The rituals of aśvamedha and rājasūya, for example, reflect the shaping of the concepts of world conquest and the world ruler. The sound and its proper transmission play a key role in the performance of yajña (Falk 1990; Rocher 1994: 8, 11). Above all, orthopraxy appears to be more important than orthodoxy, or the techniques of performance overpower the dogmas of belief. Buddhist literature contains yet another rich set of sociopolitical ideas, many of which display a remarkable degree of abstract theorizing and conceptual depth. One of the most important achievements of the early Buddhist texts is an elaborate story of the origin of the world. The concept of dhamma, the theory of the two domains of power (religious and secular), the rejection of the Brahmanical concept of daṇḍanīti, and the idea of the world ruler are among the other noteworthy sociopolitical ideas. Last but not least, Buddhism stood for the ethical or even moral foundations of political practice. Ethics in Buddhism were above the politics, and no Machiavellian purpose could justify their violation. The most elaborate political ideas of ancient India are nonetheless found in the Śāstras. In spite of being the product of Brahminical ideology, many of these ideas attain a high level of abstraction and secularism. The Kauṭilīya Arthaśāstra contains, by far, the largest amount of such ideas and understandably has been the

Clearly, it is impossible in such a brief article to make a proper assessment of ancient Indian sociopolitical thought. What I would like to do instead is to highlight several themes that unite different religious and literary traditions and to discuss whether any of these themes may be conceptually valuable for the analysis of archaeological data. In the case of ancient Indian 82

REJECTING THE STATE: REFLECTIONS ON ANCIENT INDIAN THOUGHT AND THE ORGANIZATION OF HARAPPAN SOCIETY structure. What is nonetheless important is that the relationship between the two powers is among the themes that had occupied the minds of ancient Indian theologians and social thinkers from the earliest times.

sociopolitical thought, there seem to be at least seven themes that deserve attention. 1. The varṇāśrama system and the varṇāśrama dharma play a key role in all the traditions of sociopolitical thinking of ancient India. Even the Buddhist canon, which rejects divine creation and daṇḍanīti, pays a great deal of attention to the varṇāśrama. Kingship, according to the Buddhism, is above the varṇāśrama, yet the varṇāśrama is clearly recognized. From the theoretical point of view, this leads to at least two observations. First, the varṇāśrama arguably implies the pre-dominance of heterarchical rather than hierarchical relations in the organization of society. Second, the primary role in the formation of social groups and other identities is played primarily by ideational (non-economical) factors.

5. Beginning with the Vedic rituals of aśvamedha and rājasūya, the ideas of conquest and expansion are expressed throughout early Brahminical and Buddhist literature. The Buddhist concepts of the world ruler and the world conqueror, the śāstric theories of the vijigīṣu and maṇḍalas, the glorification of the rāmarājya, and even the very cause of the Mahābhārata war are the expressions of the idealistic concept of territorial expansion. In the Brahmanical tradition, the best example of this concept is the praxis of aśvamedha. In the Buddhist tradition, it is expressed in the term ‘chakkavattī’ (the world ruler). In both traditions it is visualized as a wheel. In the Śvetāśvatara Upaniṣad, the wheel is technically described and philosophically construed. As the allegory proceeds, “in this great wheel of brahman in which all living beings [are encircled] and which exists in everything, a goose roams around. Viewing atman distinctly from the impeller and pleased by him, he [the goose] proceeds to immortality” (Śvetāśvatara Upaniṣad 1.6.) In a similar way, the concept of chakkavattī, which literally means ‘the turner of the wheel’, conveys the metaphoric equation of the wheel with the world.

2. In spite of its significant diversity from one tradition to another, the concept of authority in ancient Indian sociopolitical thought has a limited and semi-contractual nature. Neither absolutism of European monarchies, nor the unconditional power of the Egyptian pharaohs has obvious parallels in ancient India. The king, similar to his subjects, must follow his dharma; and the Machiavellian logic of the Arthaśāstra that may justify the means of achieving the proper purpose is not the same as the absolute power that justifies and deifies any activity of the ruler or the ruling elite. In fact, the strict compliance with the dharma affects the king in ways similar to those that the dharma of his subjects prescribes.

6. The concept of the polity in ancient Indian texts is inseparable from the teaching of polity interaction. A good example of how close to one another these two teachings are is the interrelation of the maṇḍala and saptāṅga theories. The last of the seven constituent elements (‘limbs’) that define the polity according to the saptāṇga theory is ‘the ally’. The purpose of the maṇḍala theory is to convey the importance of proper polity interaction. Both theories thus reflect a geopolitical layout, which is defined by a set of relatively small and interconnected polities, not by a unitary state. Remarkably, the concept of the world-ruler that is particularly typical of Buddhist thought is not in conflict with this layout; for despite having developed an idealistic concept of world conquest, the ancient Indian texts did not have a clear theory of the unitary and imperial state. What the archaeologist Colin Renfrew has termed ‘peer polity interaction’ seems to have dominated ancient Indian sociopolitical thinking (Renfrew & Cherry 1986).

3. Most expressions of sociopolitical thinking in ancient India are circumscribed by ethical norms. The Dharmasūtras unequivocally state that it is the king’s obligation to protect his people. On some occasions, the king is required to reimburse the stolen property of his subjects from his own treasury, or to pay a penalty if he does not properly perform the daṇḍa. The ethics of politics is one of the main themes in the Buddhist canon. The Jātakas, for example, pay a great deal of attention to the description of disasters that happen when the ruler becomes unrighteous and no longer follows his dhamma. The Dīgha-Nikāya unequivocally states that the chakkavattī (the world ruler) must be particularly righteous. Many texts of the Brahminical tradition emphasize the role of ethics in political practice as well. 4. One of the key themes in early Brahmanical thought is the dialectics of secular and religious authority, or the socalled concept of the two powers. The term ‘brahmakṣatra’ which appears in the Brāhmanas presents the initial formation of this concept. Its further elaboration is found in the Dharmasūtras, Śāstras, Epics, and Purāṇas. The Buddhist tradition also pays some attention to this issue, yet evidently presents it from a very different perspective. There has been much debate in the Indological literature on how the relation between Brahmins and Kṣatriyas are presented in different texts. It is usually agreed that while the Brahminical tradition emphasizes the pre-eminence of Brahmins, the Buddhist tradition envisions kingship as a keystone of societal

7. There has been a long-standing debate among anthropologists and Indologists on the relation between textuality and orality in ancient South Asia. I tend to follow those scholars who believe that orality has played an extremely important role not only in the transmission and preservation of sacred information, but also in the transmission and preservation of legal information (Rocher 1994; Falk 1990; Olson & Torrance 1991; Witzel 1995). One could add that being such a key element of religious, sociopolitical and cultural life, orality in the South Asian context can be seen not only as 83

Piotr A. Eltsov expression of new identities and world views, the three dimensional layout of sites must have been created under the supervision and control of secular authorities. Responsible for the performance of domestic rituals and possibly for the reinforcement of ideological norms (particularly of their ethical component), the religious authority, on the other hand, may have left no material expression.

a key device for the preservation and transmission of information, but as a conceptual framework. Long after writing was invented and began to be utilized for various purposes, orality continued to play a crucial role in almost every aspect of life in South Asia. Anti-Hegel or the Constructive Suggestions The rejection of the neo-evolutionary theory as a framework to explain the genesis of ancient civilizations is not new. There has been a good deal of research showing the uniqueness of cultural trajectories that the earliest civilizations pursued. Yet unfortunately, much of this research stops short at proposing alternative models. This article hopefully sets the grounds for moving beyond the mere criticism of the theory of the state and for finding a way of dealing with the emergence of one civilization on its own terms. Thinking of the theoretical subtext that some of the sociopolitical ideas found in ancient Indian texts convey, one cannot help but be reminded of the puzzles that the archaeology of Harappan civilization poses. Let us consider some of these puzzles in conjunction with the aforementioned ideas (see Table 1).

Mechanisms of power backed by ideological norms were also likely to be responsible for territorial expansion. The mere fact that the Harappan civilization was the largest in the 3rd millennium BC posits a number of intriguing questions. Was the expansion part of Harappan ideology? Did the trajectories of expansion follow different patterns in different areas? What were the intellectual and historical roots of the theories of world conquest and the world ruler and was there any connection between these theories and the political praxis of the Harappan civilization? Although archaeology does not provide easy answers to any of these questions, it is tempting to hypothesize that expansionism had played a crucial role in the formation of Harappan ideology; for one thing is clear: expansionism was a key element of both the Harappan archaeological identity and ancient Indian sociopolitical thought. Moreover, the lack of evidence of warfare, the enormous geographical span of the Harappan material culture, and the diversity of areas to which the trajectories of expansion were directed point to the ideational and peaceful nature of expansion. Each of these traits also indicates that the Harappan polity was most likely de-centralized, lacked unitary government, and consisted of competing entities. Remarkably, the geopolitical premises of maṇḍala and saptāṅga theories come in handy as a tool for understanding this type of polity and polity interaction.

The varṇāśrama implies the importance of ideational and heterarchical principles in the shaping of social system. It is not unreasonable to suggest that similar principles were characteristic of the Harappan society. Kenoyer has argued that some of the Harappan artifacts (in particular, seals, ornaments and pottery) served as symbols of social status and hierarchy (Kenoyer 2000). Even if this was the case, these artifacts do not seem to provide sufficient evidence for the hierarchy being the defining principle of social relations. The social system of the Harappan society was clearly very complex. The construction of massive platforms, the establishment of founder’s settlements, the circumvallation of sites by walls, and the making of highly precious objects must have required control of labor and a certain degree of hierarchical relations. Peculiar urban segregation, the lack of lavish paraphernalia, and relative modesty of burials nonetheless indicate that the proper metaphor for the Harappan social system is a low-standing tree with many long meandering branches, and not the upright ladder of social hierarchy.

Last but not least, orality explains several puzzles of Harappan archaeology. What was conveyed in other civilizations through intimidating architecture, luxurious burials, lavish ritual paraphernalia, monumental art, and long written texts could be conveyed by oral means in the Harappan culture. Even the pace of urbanization and, in particular, the events that led to the sociopolitical and cultural efflorescence in the middle of the 3 rd millennium BC may have been triggered by highly developed oral culture. The technology, social organization, and political culture necessary for the construction of Mohenjo-Daro and other founder’s settlements, as well as for the implementation of many other drastic innovations that took place in the 26th century BC could hardly be implemented without literacy or a highly sophisticated culture of preserving information by oral means. As we know, there is no evidence of the Indus script prior to the Mature Harappan period. Moreover, the doubts are still raised on the relation of this script to any spoken language. Given the lack of visually imposing expressions of power, it seems reasonable to suggest that orality played a paramount role in the formation and functioning of Harappan culture (Malik 1968). The power of the spoken word could have excluded the necessity to build temples, palaces, and tombs.

The lack of lavish material objects and foreign goods as well as the modesty of burial patterns may also be indicative of the semi-contractual institutions of power circumscribed by ethical norms. Changes in funeral ceremonies from the Neolithic to Chalcolithic period in Mehrgarh seem to contain keys to the origins of the ideology that was responsible for the formation of this type of power. Remarkably, the Harappan authorities were powerful enough to maintain urban segregation and to reinforce complex sociocultural relations; yet aside from perimeter walls and massive platforms they did not leave any visually imposing objects. Given this, it seems reasonable to suggest that the crystallization and separation of religious and secular powers had taken place by the beginning of the mature Harappan period. Having no clear religious subtext and being the 84

REJECTING THE STATE: REFLECTIONS ON ANCIENT INDIAN THOUGHT AND THE ORGANIZATION OF HARAPPAN SOCIETY Claessen, H. J. M. and Oosten J. G. (1996) ‘Discussion and considerations’. In Claessen, H. J. M. and Oosten J. G. (eds) Ideology and the Formation of Early States. Leiden: 359-405. Claessen, H. J. M. and Skalnik P. (1978a) ‘The early state: models and reality’. In Claessen, H. J. M. and Skalnik P. (eds) The Early State. The Hague: 637-650. Claessen, H. J. M. and Skalnik P. (eds) (1978b) The Early State. The Hague. Claessen, H. J. M. and Skalnik P. (eds) (1981) The Study of the State. The Hague. Dales, G. F. and Lipo C. P. (1992) Explorations on the Makran Coast, Pakistan: A Search for Paradise. Berkeley. Dowdall, H. C. (1923) ‘The word “state”’. The Law Quarterly Review, 39: 98-125. Drekmeier, C. (1962) Kingship and Community in Early India. Stanford. Dyson, K. H. F. (1980) The State Tradition in Western Europe: A Study of an Idea and Institution. Oxford. Ehrenreich, R. M., Crumley C. L., and Levy J. E. (eds) (1995) Heterarchy and the Analysis of Complex Societies. Arlington. Eliade, M. (1949) Le Mythe de l'Eternel Retour: Archétypes et Répetition. Paris. Eliade, M. (1957) . Paris. Eltsov, P. A. (2008) From Harappa to Hastinapura: A Study of the Earliest South Asian City and Civilization. Boston. Fairservis, W. A. (1984) ‘Harappan civilization according to its writing’. In Allchin, B. (ed) South Asian Archaeology 1981. Cambridge: 154-161. Fairservis, W. A. (1992) The Harappan Civilization and its Writing. New Delhi. Falk, H. (1990) ‘Goodies for India – literacy, orality, and Vedic culture’. In Raible, W. (ed) Erscheinungsformen kultureller Prozesse: Jahrbuch 1988 des Sonderforschungsbereichs “Übergänge und Spannungsfelder zwischen Mündlichkeit und Schriftlichkeit”. Tübingen: 103120. Farmer, S., Sproat R., and Witzel M. (2004) ‘The Collapse of the Indus script thesis: the myth of a literate Harappan civilization’. Electronic Journal of Vedic Studies, 11(2): 19-57. Fuller, C. J. (1984) Servants of the Goddess: The Priests of a South Indian Temple. Cambridge. Gamble, C. (1986) ‘States in history’. In Hall, J. A. (ed), States in History. New York: 22-47. Ghoshal, U. N. (1959) A History of Indian Political Ideas: The Ancient Period and the Period of Transition to the Middle Ages. Bombay. Goody, J. (ed) (1968) Literacy in Traditional Societies. London.

In conclusion, one must note that it is unfortunate that while most historians had long ago abandoned the language of evolutionary theory, anthropological archaeologists continue to beat a dead horse. The indiscriminate labeling of ancient civilizations as chiefdoms and states is like the judging of nonrepresentational art by the standards of socialist realism. Stretching the term ‘state’ to the farthest end and presuming it to be as semantically inclusive as any complex sociopolitical organization does not provide a better option either; for misusing the concept of the state, one disregards the tremendous intellectual legacy that it has left in several languages and intellectual traditions. From the point of view of both the humanistic Begriffsgeschichte and case-oriented anthropology, it seems to be much more productive to define the analytical language that reflects the ancient past in more appropriate terms. Ancient Indian sociopolitical thought may help us create such a language for the study of the Harappan civilization. Acknowledgements The thoughts expressed in this article were shaped during the course that I taught at the Freie Universität in spring of 2007. Therefore, first of all I must thank Alexander von Humboldt Foundation for generously funding my current research and Professor Harry Falk for kindly inviting me to work and teach at the Institut für die Sprachen and Kulturen Südasiens in Berlin, Germany. I also would like to thank all the students who attended my course in Berlin and whose comments and questions stimulated my own thinking about the Harappan civilization. For the invitation to participate at the Biannual Meeting of European Association of South Asian Archaeologists in Ravenna, Italy, in July 2007, I am personally indebted to Professor Maurizio Tosi, Dennys Frenez, Barbara Cerasetti and other Italian colleagues. My wife Vitessa Del Prete proofread the final version of this article.

Bibliographical References Altekar, A. S. (1949) State and Government in Ancient India. Delhi. Ali, I. and Rahim G. (2001) ‘Excavations at Jhandi Babur’. Ancient Pakistan, 14. Caldwell, J. R. (1964) ‘Interaction spheres in prehistory’. In Cadwell, J. R. and Hall, R. L. (eds) Hopewellian Studies. Springfield: 133-143. Cassirer, E. (1946) The Myth of the State. New Haven. Childe, V. G. (1950) ‘The urban revolution’. Town Planning Review, 21(1): 3-17. Claessen, H. J. M. (2000) Structural Change: Evolution and Evolutionism in Cultural Anthropology. Leiden.

85

Piotr A. Eltsov Goody, J. (1987) The Interface between the Written and the Oral. Cambridge. Goody, J. (2000) The Power of the Written Tradition. Washington. Goody, J. and Watt I. P. (1963) ‘The consequences of literacy’. Comparative Studies in Society and History, 5: 304-345. Harding, A. (1994) ‘The origins of the concept of the state’. History of Political Thought, 15(1): 57-72. Harding, A. (2002) Medieval Law and the Foundations of the State. Oxford. Hegel, G. W. F. (1821) Grundlinien der Philosophie des Rechts. Berlin. Hegel, G. W. F. (1833) Grundlinien der Philosophie des Rechts oder Natturrecht und Staatswissenschaft im Grundrisse. [Werke. Vollstandige Ausgabe. Band 8]. Berlin. Hillebrandt, A. (1923) Altindische Politik: Eine Übersicht auf Grund der Quellen. Jena. Jansen, M. (1994) ‘Mohenjo-Daro, type site of the earliest urbanization process in South Asia: ten years of research at Mohenjo-Daro, Pakistan, and an attempt at a synopsis’. In Parpola, A. and Koskikallio, P. (eds) South Asian Archaeology 1993, vol. 1. Helsinki: 263-280. Jarrige, J. F. (1987) ‘Der Kulutrkomplex von Mehrgarh (Periode VIII) und Sibri: Der Schatz von Quetta’. In Jansen, M. (ed), Vergessene Städte am Indus: Früher Kulturen in Pakistan vom 8-2 Jahrtausend v. Chr. Mainz am Rhein: 102-111. Kantorowicz, E. H. (1957) The King’s Two Bodies: A Study in Medieval Political Theology. Princeton. Kenoyer, J. M. (1994) ‘The Harappan state: was it or was not it?’. In Kenoyer, J. M. (ed), From Summer to Meluhha: Contributions to the Archaeology of South and West Asia in Memory of George F. Dales, Jr. Madison: 71-80. Kenoyer, J. M. (1997) ‘Early city-states in South Asia: comparing the Harappan phase and the early historic period’. In Nichols, D. L. and Charlton, T. H. (eds), The Archaeology of City-States: CrossCultural Approaches. Washington: 51-70. Kenoyer, J. M. (2000) ‘Wealth and socio-economic hierarchies of the Indus Valley civilization’. In Richard, J. and van Buren, M. (eds) Order, Legitimacy and Wealth in Ancient States. Cambridge: 88-109. Law, N. N. (1921) Aspects of Ancient Indian Polity. Oxford. Leonardi, G. (1988) ‘New problems of surface Archaeology: sampling in HR east area of Moenjodaro (Pakistan)’. In Jansen M., Tosi M., and Leonardi G. (eds) Interim Report: Reports on Field Work Carried out at Mohenjo-Daro, Pakistan, 1983-1986 by the IsMEO-Aachen University Mission, vol 3. Aachen: 7-92.

Malik, S. C. (1968) Indian Civilization, The Formative Period: A Study of Archaeology as Anthropology. Simla. Marx, K. (1872) ‘Nachwort zur zweiten Auflage’. In Marx, K. Das Kapital: Kritik der politischen Oekonomie. Hamburg: 813-822. Meinecke, F. (1924) Die Idee der Staatsräson in der neueren Geschichte. München. Olson, D. R. (1991) ‘Literacy as a metalinguistic activity’. In Olson, D. R. and Torrance, N. (eds) Literacy and Orality. Cambridge: 251-270. Olson, D. R. and Torrance, N. (1991) Literacy and Orality. Cambridge. Pauketat, T. R. (2007) Chiefdoms and Other Archaeological Delusions. Lanham. Possehl, G. L. (1998) ‘Sociocultural complexity without the state: the Indus civilization’. In Feinman, G. M. (ed) Archaic States. Santa Fe: 261-291. Posshel, G. L. (2002) Indus Civilization: A Contemporary Perspective. Walnut Creek. Post, G. (1964) Studies in Medieval Legal Thought, Public Law and the State, 1100-1322. Princeton. Rau, W. (1956) Stadt und Geselschaft im atlen Indien. Wiesbaden. Renfrew, C. and Cherry J. F. (eds) (1986) Peer-Polity Interaction and Socio-Political Change. Cambridge. Rocher, L. (1994) Orality and Textuality in Indian Context. Phildalephia. Scharfe, H. (1989) The State in Indian Tradition. Leiden. Sellier, P. (1987) ‘Mehgarh: Grabstätten und Bestattungsritus’. In Jansen M., Mulloy M., and Urban G. (eds) Vergessene Städte am Indus: Früher Kulturen in Pakistan vom 8. – 2. Jahrtausend v. Chr. Mainz am Rhein: 83-94. Sharma, R. S. (1968) Aspects of Political Ideas and Institutions in Ancient India. Delhi. Skalnik, P. (1983) ‘Questioning the concept of the state in indigenous Africa’. Social Dymanics, 9 (2): 1128. Skalnik, P. (1987) ‘On the inadequacy of the concept ‘traditional state’ (illustrated with ethnographic material on Nanun, Ghana)’. Journal of Legal Pluralism, 25, 26: 301-325. Skalnik, P. (ed.) (1989) Outwitting the State. New Brunswick. Skinner, Q. (1978) The Foundations of Modern Political Thought. Cambridge. Skinner, Q. (1989) ‘The State’. In Ball T., Farr J., and Hanson R. L. (eds) Political Innovations and Conceptual Change. Cambridge: 90-131. Southall, A. (1988) ‘The segmentary state in Africa and Asia’. Comparative Studies in Society and History, 30: 52-82.

86

REJECTING THE STATE: REFLECTIONS ON ANCIENT INDIAN THOUGHT AND THE ORGANIZATION OF HARAPPAN SOCIETY Witzel, M. (1997) ‘Early Sanskritization: origins and development of the Kuru state’. In Kolver, B. (ed) Recht, Staat, und Verwaltung in klassischen Indien. The State, the Law, and Administration in Classical India. Munchen: 27-52. Wright, H. T. (1978) ‘Toward an explanation of the origins of the state’. In Cohen, R. and Service, E. R. (eds), Origins of the State: The Anthropology of Political Evolution. Philadelphia: 49-68. Yoffee, N. (2005) Myths of the Archaic State. Cambridge.

Spellman, J. W. (1964) Political Theory of Ancient India: A Study of Kingship from the Earliest Times to circa AD. 300. Oxford. Steinberger, P. J. (2004) The Idea of the State. Cambridge. Trigger, B. G. (1990) ‘Monumental architecture: a thermodynamic explanation of symbolic behavior’. World Archaeology, 22 (2): 119-132. Trigger, B. G. (2003) Understanding Early Civilizations: A Comparative Study. Cambridge. Vincent, A. (1987) Theories of the State. Oxford. Witzel, M. (1995) ‘Early Indian history: linguistic and textual parameters’. In Erdosy, G. (ed) The IndoAryans of Ancient South Asia: Language, Material Culture and Ethnicity. Berlin: 85-125.

87

Piotr A. Eltsov

Table 1. - Selected Sociopolitical Ideas in the Ancient Indian Literature

Varṇāśrama system

VL

DS

KA

BC

MS

YS

MB

RM

KN

SN

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Vārtta Dharma

+ +

+

+

+

+

Daṇḍanīti Conquest

+

+

+

+

+

+

+

+

+

+

Temporal vs. sacred

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

+

Concept of kingship Saptāṅga theory

+

+

+

+

+

+

+

Naṇḍala theory

+

+

+

+

+

+

+

Ethical foundations

+

+

+

+

+

+

+

+

+

+

The world ruler

+

The five powers

+

VL Vedic Literature

KA Kauṭilīya Arthaśāstra

MS Manusmṛti

MB Mahābhārata

KN Kāmandaki Nītisāra

DS Dharmasūtras

BC Buddhist Canon

YS Yājñavalkyasmṛti

RM Valmīki Rāmāyaṇa

SN Śukranītisāra

88

NOMADIC SETTLEMENT AND ECONOMY OF WESTERN HIMALAYA (UPPER INDUS) R. S. Fonia

discoveries in the region. Contemporary life pattern of the nomads of the Himalayas also help us to understand the centuries old customs and traditions preserved from the ancient past. When we take vast span of time into consideration, we also find here a considerable cultural continuity from prehistoric times to the historic period which also covers a vast geographical area from Nubra to Dhauli valley. Secondly, my intention is also to identify correctly the nomadic people with their age old customs and also to bring to light the historical records which have so far not been studied thoroughly and methodically. Situated on the ideal geographic conditioned, the nomads of the Western Himalaya played considerable role in making the Cultural and Historical processes of various epoch. Unfortunately we do not find any references on them even in the school texts and university courses. And lastly, I have intended to study the development of culture in the context of trade and commerce in so far as they are reflected in archaeological remains and oral traditions. The present study is aimed to understand the nomadic settlement and economy during prehistoric times in the Himalayas.

The Upper Indus region situated on the extreme west of the Himalayas is an interesting area for archaeological and anthropological studies. The Cultural Vestiges is preserved in a major way in its archaeological remains and monuments of varied nature. The range of relics of the past, from this region is indeed vast. Although monuments belonging to the medieval times are well known, but very little is known about the prehistory of the earlier phase. The recent archaeological discoveries in this region indicate the archaeological potentialities of the region and produced evidence to understand the various stages in the evolution of transhumant practices to the barter of the sustenance commodities in the background of the local environment. The commercial and cultural contacts with the neighboring regions in the lower Indus, Tibet and Central Asia also helped in the development of local communities and growth of urban centers in the region. The Himalayas consist of mainly three mountain ranges extending from west to east, which is the outer Himalaya, the middle Himalaya and the Great Himalayas. These mountainous terrains enclose between them a number of beautiful elevated valleys, lush meadows, lakes and snow clad peaks which give rise to three principle river systems, viz, the Indus-Sutlej in the north-west, the Ganga-Kali in the centre and the Brahmaputra in the east. The present study is primarily concerned with the western Himalaya which forms natural boundary between India and Tibet. The natural division of the western Himalayan region comprises the Nubra Valley on the Sheyok and Nubra rivers, Ladak on the Indus, Zanskar on the Zanskar river, Purig, Suru, Sod and Drass along different branches of the Drass river. Similarly Kashmir valley along the Jhelum river, Lahul-spiti, Kistwar, and partly Jammu lie along the Chandrabhaga (Chenab) river, Kinnaur along the Sutlej, Nilang along the Bhagirithi, Juhar along the Gori, Darma on the eastern Dhauli and Byans on the Kali, and Niti-Mana valley along the Dhauli, Alaknanda and Saraswati rivers. The valleys of these rivers are the main lines of drainage. These rivers also facilitate, rather than inhibit the movement of the nomadic people during their seasonal migrations. They also determine not only the areas where nomads may live, but also roots of communication between Tibet and Central Asia. The climate and physiographic conditions determine the economy of the people, the types of animals, their migration and seasonal settlement patterns and many other aspects of life.

In carrying out the present study I have largely depended on my own extensive field work in the remote areas of Himalayas. The investigations and intensive surveys carried out in the region after region is based on study and analysis of many prehistoric sites, stone artifacts, rock-shelters, cave-dwellings, rock-engravings, sculptures, historical monuments, personal diaries, trade agreement letters, important items of trade like Pashmina wool, Gold dust, Salt, Borax, Musk, herbs, silajit (local medicine of obscure nature used in India since time immemorial) and Yak tails, oral traditions and folklores. What I have thus attempted to do here is to corroborate the evidence collected during field work with evidence derived from variety of records. Dating of the archaeological remains and structures has been made on the basis of my personal observation of material and geological formations, method of construction and style with reference to the published and unpublished document bearing on the history and culture of the region. Further scientific research is required to confirm the relationship of archaeological findings and historical records. Geologically, the Himalayan basin falls in the Ladakh Kashmir, Lahaul - Spiti, Kumaon - Garhwal and Nepal Bhutan regions which were once probably all connected as a single unit. On the basis of the geological character four longitudinal geological belts are identified as Tibetan Tethyian, Higher Himalaya, lower Himalaya and Sub Himalaya. In fact, Himalaya being the youngest mountain system in the world is very interesting area for geological and archaeological research (Fig.1a & b).

The main objective of the present study on the subject are three fold: Firstly, I intend to discuss the detailed study and documentation of prehistoric sites and remains, already discovered in number of sites in the Upper Indus which provide important material for conducting future 89

R. S. Fonia Terrace deposits are quite common all along the Indus river and its tributaries in the western Himalayas but more so where the valleys have broadened out. These levels of terraces have been identified by Burbank and Fort (1985). The lowest terrace at +20 meter is a simply alluvial or rock terrace related to the present river. The second terrace consists of coarse alluvial fan conglomerates presumably related to peri-glacial fan advance as these are occasionally overlain by moraines and lie between +200 m and +400 from the present river bed. Brookfield considered these terraces to have been formed the last 100 kyr (Brookfield 1993). Thus the basins created by internal drainage, the lake deposits on the Indus belt, terraces and lacustrine deposits belonging to Quaternary deposits have all proved quite significant to understand the evidence of the early Human culture and to locate the centre of civilizations of the region.

Birbal Sahni Institute of Palaeobotany, Lucknow reported the evidence of temporary settlements in the form of hearths near Gaik, and the carbon date comes to about 6,710 BP +130 (Sharma, Rajagopalam and Choubey 1989). Again a group of Geologists, from the Geological Survey of India, while investigating in 1985 the problem of Neogene/Quaternary successions of Upper Indus system in Ladakh, found evidence of early man in the region. They collected stone artifacts from stratified context found in association with Indus terraces near village Nurla around Khalsi (Personal correspondence with G.S.I. Team, Ladakh Project; Sahni 1936: 58). Archaeology has provided earliest evidence of Human activities in the valley through explorations. In the year 1990-1996, some stone tools along with rock boulders engraved with hunting scene of ibex were noticed by me on the left bank of the Indus. In the course of archaeological studies on Upper Indus (Fig. 2), some Palaeolithic artifacts have been discovered from the terraces of Tirisa Lake and Hunder Dok in Nubra valley (Fonia 1993) (Fig. 3). Many prehistoric sites have also been noticed in the remote areas of Ladakh, i.e., Zanskar, Changthang, Batalik, Da-Hanu and Kargil area. These sites can provide archaeological data only after small scale excavation are conducted here to understand the typology and stratigraphy. The stone tools comprise cleaver, scrapper, handaxes and flakes. These are fabricated on chert, granite or diorite.

The geological studies concerning the ice age associated with human culture conducted by H. De Terra and Peterson (De Terra & Patterson 1939) and explorations conducted by Francke (Franke 1930) and Birbal Sahni provide some clues about the existence of early man in the area of our study, but no further investigations was carried out by the succeeding scholars in this direction. Birbal Sahni (1936) has drawn an important conclusion that the cultural contact between India and China probably existed since the dawn of human existence. Long before the man conquered the ocean intercourse between these two ancient countries was possible by the direct route across the Himalayas which, during Paleolithic and Neolithic times, were possibly not so high as to form an effective barrier. By the middle of the twentieth century, reports on the Stone Age culture of the region began to appear. Palaeolithic tools have been known to occur from the Potwar region in Pakistan, Tajakistan in Central Asia, Kashmir and all along the Himalayan foot hills.

The archaeological evidence at present is not sufficient to give complete account of the prehistoric culture of man in the region, but most of the material is confined to its primitive rock engravings. The primitive rock engravings are widely distributed in the western Himalaya all along the ancient routes connecting Ladakh with Baltistan, Zanskar, Changthang in the east, Nubra in the north and further Lahul Spiti in the south-east and Kashmir in the west. These rock engravings are the visible form of art left by the nomadic people in the region and are mainly seen on the high mountain grazing grounds, ancient routes all along river valleys, lakes and lush meadows. These engravings depict a common theme of hunting scenes, grazing livestock, later fighting scenes, folk dances and religious figures etc (Fig. 4-6). The animals represented here are ibex, markhor, yak and camel. These carvings testify the ancient occupation of the remote valleys of the western Himalaya proving that from very early times the higher region of the Himalayas were inhabited. The most important factor to be ascertained and identified is the origin and chronology of these engravings and their comparative analysis with rock engravings from the northern areas of Pakistan, TransPamir, and Xin-Jiang regions. The prehistoric finds provide information for reconstructing the history of the earliest people who lived in these higher reaches of the Himalayas. They were the first to exploit the natural resources and leave behind a tradition that had great bearing on the subsequent period of history. Evidences of early community life are noticed in the rock engravings and paintings. The common themes of these primitive engravings are the hunting

Archaeological findings were reported by Archaeological Survey of India from Burzahom (Indian Archaeology. A review 1959-60; 1960-61; 1961-62; 1962-63) and Gufkral. Both the sites in Kashmir have left valuable clues to work further into Upper Indus (Ladakh) in order to ascertain its origin and cultural link with Central Asia. Archaeological discoveries from Mehrgarh, situated on the foot of Bolan Pass have furnished sufficient evidence to understand the various stages in the evolution of the food producing communities on the basis of local environment (Jarrige & Lechevallier 1979). Further investigation on the Upper Indus (Ladhak region) is also essential to understand the transition from food gathering to food producing stage. Besides in view of the common artifactual remains such as harvester and Jade beads recovered from the Neolithic sites North China (Chang 1968) and Burzhom and Kashmir the mechanism of cultural contact during this period is to be understood. The possible route of movement of people and ideas in the Kashmir and the Upper Indus valley was through a series of passes which connects Gilgit Valley with Xinjiang (China). In the year 1980, researches from Wadia Institute of Himalayan Geology, Dehradun and 90

NOMADIC SETTLEMENT AND ECONOMY OF WESTERN HIMALAYA (UPPER INDUS) region. Here, limited excavations were taken by the author on behalf of the Archaeological Survey of India (1995-1996) (Fonia 1997) that revealed artifacts like chisel, celt, ring stone, quern, pounders, and a few number of unfinished artifacts like flakes, cores, debitage etc. Although the tools are not polished, but typologically they resemble with the artifacts of Burzahom and Gufkral limited number of ill fired red ware showing medium to thick fabric was also encountered, the shapes including bowl, vase and long necked jar deserve mention. Coarse grey ware also came across representing flat base of the pots. The shapes resemble with the handmade ware found from Gufkral and Burzahom. On the basis of a mud floor and a line of boulders around the oval shaped floor it has been suggested that inhabitants lived on mud floors prepared on the ground supported by thatched roofs. The characteristic feature of the Malpur Neolithic is the complete absence of bone tool industry.

scene, mode of transport and groups of ibex and other animals. The hunting scenes found in the entire Ladakh region represent the socio-economic life of the prehistoric people who depended on hunting for their livelihood. Prof. Dani in an attempt to date these engravings, associates them with the pre or early Holocene age when the present dry bed had running water rivulets (Dani 1987). On the basis of terraces of River Indus it is noticed that the Indus River was flowing about 20 m above the present level. The depiction of animals near the grazing grounds shows that this human group lived in pastoral stage (Fig. 7). On the whole, this was the stage in between food gathering and food producing stage and the food producing stage of the early pastoral nomads. The economy of these nomads was basically pastoral and barter system of essential commodities. The engravings of ibex provide the earliest information on the totemistic cult, having ibex as the sacred animal because, according to local legend, it was considered as a fertility Goddess and later as one of the incarnations of Lord Buddha. Its horns are still kept in the graves as a symbol of protection till recent times by the locals.

As we come closer to the pre-historic period, the evidence multiplies, the depiction of human and animal life acquires a wider range, the symbols increase in number, the religious meaning become clearer and at the same time new monumental remains in the form of grave stones, circles and monolithic rocks indicate presence of new people, new contacts and new system of socioeconomic behavioral pattern. Here, by now, we find a meeting ground of the primitive and the settled, the nomads and cultivated groups, the hunters and the pastoral people, and above all we can visualize the growth of a political order in which a technically advanced nomadic tribe emerges as masters of the rest of the population and establishes its historic identity under the name of Daradas and Khasas mentioned in the ancient classical literature.

The engraving of ibex in folk pattern on the smooth black surface of rock (Fig. 8) also suggests the concept of mountain worship by these early people. The practice of the worship by the women of Ladakh of these rock engravings of ibex, for begetting a child is in vogue even today. The depiction of palm and foot impression (Fig. 9) symbolizes the religious and ritual practices. Simple bow and arrow are the only representation of weapons in the engravings (Fig. 10). Later, these rocks were selected for the engravings of giant human figures. A. H. Dani notes “The rock art can be studied in its local geographical environment but more than it can be placed in a wider prospective of rock art as known from the surrounding areas in the Trans-Pamir side, Xin-Jiang, Lhasa, Ladhak, Kashmir and Swat. The first is that the examples of rock art, spread over the great length of this extensive zone, are inter related and second is that there is a continuity in their rock art from early times right in to the historic period in such a way that we can well speak with language of cultural continuity from pre-history to history” (Dani 1987). From Burzhom in Kashmir two engraved stone slabs were found in the late Neolithic level. The first one was upside down while the other was in a damaged condition. The engraving depicts a hunting scene showing a stag being pieced from behind with a long spear and struck from the front by an arrow by two hunters the upper portion of the slab shows two suns and a dog. The other stone slab depicts an incomplete patter identified as a tectiform. The design on the slab appears to be a highly stylized and abstract representation of a hut with a domed roof; the body portions the animal and its tail (Pande 1972) (Fig.-11 a & b).

The Proto-historic phase in the Western Himalayan regions associated with the relatively advanced culture of megalithic pit burials, bronze and iron technology, earthen vessels, beads and utensils, weapons, introduction of food production and domestication of animals. The temporary hearths, cave dwellings near sal lakes, and sulphur and soda mine in Changthan area belonging to the Neolithic cultural phase show cultural and commercial links with Tibet and with the areas on the northwest and south of the Himalayas. Various sites of this phase in the region suggest a new living system based on productive economy, by developing terrace fields near river banks. Natural springs also played an important role in the settlement. The remains of stone walls, built close to rock-shelters, and remains of castles on the top of hillocks bear upon the social and political dimensions of the society, i.e., the residences of the common people and of the heads of the community or chiefs along with the provisions of defense. Number of archaeological remains in Ladakh region shows gradual improvements in the pastoral and agricultural economy, gradually adopting to trance human practices of stockbreeding and semi-nomadic way of life. They migrate with their herds during every summer to the higher reaches of pastures and after sowing the small cultivated fields around their dwelling, return to collect

Malpur, situated at a distance 28 km north of Jammu on the left bank of Ranbir canal, drawn from the river Chenab is a lesser known neolithic settlement of this 91

R. S. Fonia region have roughly the same single growing season) except to remain confined within rock shelters near hot spring. Their annual movement is restricted to 60-70 kms. In fact, they try to minimize the movement, contending that it may weaken their livestock and increase mortality. The livestock must forage for 8-9 months on dead plants left standing at the end of the growing season. The limiting factor is the amount of vegetation left at the end of summer, which must be sufficient to sustain livestock until the next year’s vegetation growth begins.

their harvest from autumn to winter. They built their winter dwellings in winter grazing areas and summer settlements in summer pastures. The trading nomads perform their long journey to Tibet and Central Asia during summer season and during winter roam around the lower valleys to collect food grains in exchange with the Tibetan products mainly wool and salt. A. H. Francke (Franke 1914), on the basis of the above practice and belief, identified these early people, who carved the figure of ibex, as Daradas but it is clear from its various representations and the geographic environment that these people were much older than the Daradas. It can be said that the Daradas could have developed their art, and carried this tradition to Tibet with the spread of religion. This fact is also substantiated by the findings of later superimposition of religious figures of chhodten (stupa), sun, moon and swastika over these figures. Their temporary and nomadic character is evident from the findings of rock shelters, seasonal camping grounds, temporary hearths etc. until they adopted the mode of settled life. The early people to migrate in the region in search of pastures may have been pastoral nomads known as Brogkpas from lower Indus and nomadic herdsman known as the Changpas from Tibet. From the archaeological evidence it is clear that some nomads from the neighboring areas used to trade with Changpas during the summer season. Although Pashmina wool and salt are the important products of Chang thang, nothing is known about the nomadic life of the Changpas and about their system of rearing animals and system of distribution and redistribution of trade. The system which they inherited through generated are discussed below on the basis of my observations made in the course of my stay with Changpas for five seasons.

Under these conditions these nomads move between two encampments: a main home base-three season encampment, i.e., used in winter, spring and summer and the sites of their tents which they use as their bases during their major migrations and reside at these encampments in handmade tents until late November when the forage is just about to be exhausted. Then they return with their sheep and goats to the main home base encampment and use the remaining vegetation until the next growing season. Complete dependence on livestock is the unique feature of their way of life, distinguishing it from that of villagers in the Himalayan region. Livestock products directly provide them with food, clothing and shelters and, indirectly, through barter system, yielding grain, tea, sugar, utensils, implements and clothing. The archaeological remains of rock shelters and cave shelters and cave dwellings near salt lakes, borax points and wool mandis also suggest the existence of primitive barter system of goods among nomads of Central Asia, Nubra, Zanskar, Baltistan, Lahul-Spiti, Tibet, Kashmir, Gilgit, Garwhal and Kumaun. The Chang thang region of Ladakh and Tibet produces the best quality of wool. The wool from Chang thang has great demand in Kashmir valley. The mineral produce of the region are borax, sulphur and salt. Borax is thrown out on the beds of stream by many hot springs at various temperatures from 80 °C upwards. Sulphur is found along both banks of rivulets at Pugga. Salt is found in the region at Tsokar Lake and near Tirisa Lake. The nomads pitch tents near salt lake. The salt is broken into small pieces by horn of Yak, and then it is packed into wooden saddle bags. The mouth of saddle bags is sewn and is piled on one side.

Changpas are the nomadic herdsmen residing in the Chang thang area which lies south-east of Ladakh. They are basically producer of wool and pass on their products to other nomads. Their economy is very simple and is based on the household raising of sheep, goats and yaks under a natural system of pastoral production. Their livestock’s are not fed with fodder or grains but they survive exclusively by grazing on range forage. This complete reliance on natural vegetations, however, creates difficulties because the Chang thang’s location on the high altitude region permits only a single limited growing season from June to October. Despite this, they not only manage to maintain forage fodder for their livestock throughout the year but also preserve their customs, culture and language.

The yak is the most useful and essential animal of the region. It grazes up to the height of 18,000 feet in winter in the areas where temperature falls to minus 40 °C to minus 60 °C. It is the only animal that can carry the load of nomads, heavy and bulky tents, and their soft wool is used for ropes and blankets. Their hide is used for soles of boots and the animal also provides large quantities of meat. The female yak provides large quantities of milk throughout the year. Sheep, goat and camel are also domesticated for mode of transport in the region. All along known history, there existed long standing trade contacts among the nomads of western Himalayas and Tibet. To facilitate trade transaction, a sot of forwarding region had developed here. The inhabitants of there

By the end of October the grasses and the sedges in the region stop growing. New vegetation does not appear until the next April-May and there are no other areas where grass is available in winter. As a result, these nomads have no option but to remain within the region unlike the Gujjars or the Gaddis who move hundreds of kilometers in winter to lower regions where fresh grass is growing. Further, they cannot escape the harsh upland long winter climate (because all the adjacent areas in the 92

NOMADIC SETTLEMENT AND ECONOMY OF WESTERN HIMALAYA (UPPER INDUS) Some writers state that the Khasas occupied the entire Himalayan belt from Khashgar, Kashmir to Khasi hills beyond Assam. One thing which is significant is that the Khasas as a whole played a very interesting role in the later history of the western Himalayan area. Rahul Sanskritayana tried to strengthen the theory of Khasa expansion in the Himalayas by referring to some old graves including one at Lippa in Spiti valley of Himachal Pradesh, which on the basis of his own examination, appeared to be of those of the Khasas. In the course of archaeological explorations, some graves came to my notice at Padam in Zanskar, Matho near Leh and Nyoma at Chang thang.

forwarding regions acted as middle man between the food providing region and wool producing regions. The articles which formed their way to Chang thang and Tibet consisted of food grains, cotton, gunny bags, utensils from Sub Himalayan regions, dry fruits, silk, Saffron from Kashmir, wool, precious stones, herbs, gold dust, musk, salt, borax and pack animals from Tibet. Ancient literary texts mention the name of some tribal groups like Kiratas, Kinnaras, Gandharvas and Nagas inhabiting in the highest Himalayan region much before the arrival of pastoral folk of the Aryans the descends of whom are now known as Khasas and Dardas. Thus, one can say that the region has served as the generator of population movements within the Himalayan land mass. The history of the ancient Himalayan world was shaped to a great extent by the succession of people who spilled out of Central Asian steppes, first to the north-western mountainous tracts and then gradually to the elevated valleys, meadows and plateaus of the vast area extending from Kashmir to Kangra and then to Kumaon and Garwhal in search of richer pastures and trade contacts. In order to survive in these hostile environs with the limited supply of natural resources the people of the Himalayas adopted the practice of semi-nomadism as according to the geographic conditions and the cycle of seasons. In this process there developed a cultural symbiosis between the hunters and the pastoralists, the primitives and the settled, the nomads and the Buddhist monks which is attested by the number of rock engravings and monuments in the region as discussed above.

The most important region in the western mountain range is Afghanistan which was once united with the North Western frontier province of India during Kushan Empire. It was approachable via Bolan and Khyber Pass. From archaeological point of view, this mountain range is the most interesting and most important part of the mountain zone. As number of trade route passes through this mountain range to Central Asia and China on one hand and to Persia and the west on the other. The passes and routes from these mountain ranges come down into the province of Sind and Punjab. Sind is situated on lower Indus between the mountain of Baluchistan on the north-west and the Thar Desert on the south-east. Like Sind the southern Punjab is situated between the western mountains and desert and the northern Punjab abuts upon the foothills of the main Himalayan ranges. Punjab literally means the land of five rivers. The main tributaries of the river Indus flows across the vast alluvial plains from their sources in the Himalayas, carrying water from the melting snow and from the monsoon rains. Thus the location of Punjab serves as a highway between the Kabul valley (one of the most important entries into India from the north-west) and the fertile plains of GangaYamuna Doab. This is the region which produced the Harappan Culture during 3rd millennium BC and further cemented the cultural and commercial contacts with western Himalayan regions.

Archaeological remains and ancient literary texts also reveal the interaction of various groups of people who migrated in the region in search of pasture and shelter. The archaeological data collected from the protohistoric graves and artifacts from Swat were analyzed and examined by scholars. These show that the newcomers introduced compound technology of megalithic buildings and ritual practices of pit-burials. They have been identified by G. Tucci (Tucci 1977) and Karl Jettmar (Jettmar 1967) with Daradas, in Ladakh and Himachal with the Khasas by Rahul Sanskritayan. However, in the local tradition or languages, the Daradas are not seen to be the people of the region. In Ladakh, they are known as Brogkpas, and in northern Pakistan as Shinas and Yaksuns.

Ladakh situated on the upper Indus in the extreme west is the meeting place of several routes coming from Yarkand via Karakoram pass, Rudok in Tibet via Changla pass and Lhasa via Gartok in Tibet all along the Indus River. The six passes from Uttrakhand Himalaya and passes from Lahul, Spiti, Bushahar, meets the Leh-Lahasa route at Gartok. Thus the physical condition of the western Himalaya clearly shows that the Himalayan economy was dependent upon its trade with Tibet, and Central Asia and China. The trade played vital role not only in preserving its centuries old socio-economic customs but it also gave its continuity and consistency throughout its history. The primitive form of trade was conducted through barter of sustenance commodities among the nomadic people living off the land in adjacent ecological zones but its passes and routes have opened the land for outside interaction which enabled them to conduct long distance trade with Tibet, Central Asia and China. The long distance trade based on a variety of pastoral and agricultural life style also helped in the development of

So far as the Uttrakhand Himalayas is concerned, Tucci’s view is quite relevant and it gives a rough sketch of time limit of the constantly migrating hordes of tribes in Himalayas during the 2nd millennium BC. Among such people, the Khasas constitute one of the predominating tribes about whom it is believed that they were the offshoots of the Indo-Aryans migrating form Central Asia. B. K. Thapar notices a general break up and movements to torch-bearers of Western Asiatic Civilization during the first quarter of the 2nd millennium BC (Thapar 1970). This movement was more or less a large scale migration of the people in various directions. 93

R. S. Fonia dependent on other essential commodities like food grains from the Himalayan borderland people.

the local communities and villages but also growth of and urban centers in the region. As such there was growth of commercial center at Yarkand, Khotan, Rudok, Gartok, Leh, Srinagar and Shikapur.

They survive not only by barley, salt or rice but there were further supplies of other day to day necessities like cooking oil, spices, sugar, tea, cotton, cloth etc. Besides, during ancient times, in the absence of treasury and bank, the major safe investment was made in procuring gold, silver and precious stones, silk, brocade, namdas, carpets and shawls were considered as the items of pride and luxury for the rich man. Basic needs were met from their fields, livestock and wool from local trade but luxurious items were procured from long distance trade from Yarkand and western Tibet.

The stone artifacts, rock shelters, primitive rockengravings, seals, letters of agreement and oral traditions are sources which testify the existence of trade relationships among the nomads of Ladakh and their contacts with Tibet and Yarkand. The outlines of the religious figures over the primitive rock engravings symbolize the interaction between nomadic traders and the Buddhist monks who followed each other into Tibet, Central Asia and China. Initially barter system of trade relations developed between nomadic herds men of Chang thang who were rich in livestock such as wool, meat, dairy products and salt and the peasants of lower area all along Indus i.e. Nyemo, Domkhar, Da-Hau, Sod, Kargil, Bodh-Kharbu, Chigtan and Lamayuru, Zanskar being away from the trade routes provided mode of transport in the form of animal and human labor.

The significance of Himalayan trade with Tibet and Central Asia not only lies in items of exchange but also encouraged cultivation of crops, animal husbandries and also ensured promotion of art, architecture and folk industries. It also makes an attempt to provide a survey and an understanding of the socio-economic aspects of western Himalaya within historical framework. Archaeological research conducted in the western Himalayan region proves that the nomadic settlement took place here during the early phase of the Neolithic period. The transition from food gathering to food production is marked by two productive system i.e. cultivation of crops and animal rearing. Thus it can be presumed that the barter of sustenance commodities came into existence with the practice of cultivation of crops in lower Indus valley and animal rearing in Chang thang in Upper Indus during Neolithic period. This has played significant role in making the cultural and economic history of the region. Contemporary life pattern of the nomadic communities has also helped in understanding the cultural continuity from the earlier to recent times in the region.

At the sustenance level, the food grains grown in the valleys of the Indus river were exchanged with wool and salt from the Chang thang region of Ladakh and western Tibet. The entire trade was conducted on the basis of mutual trust. Initially it was held on the basis of oral agreement. Although no seals have been noticed in the region to witness the trade agreement between the trading partner of Tibet and Ladakh, but as for local tradition the primitive form of trade agreement was known as Singchyad (piece of wood). It was usually a piece of wood or stone, broken into two pieces marked with their identity retained by each party in order to be tallied to ensure the trade contact. This was followed by some rituals to strengthen the relationship and also to gain confidence as mark of reputation for honesty. Later when the composition of trade items increased and also within the event of development of long distance trade, it was recorded on written promissory note by giving details and transaction of trade.

Therefore the present archaeological finding in the western Himalaya and number of sites discovered in the northern Pakistan, north of China and Central Tibet establishes cultural and commercial links between western Himalaya and Central Asia since prehistoric times and opens a new prospect to conduct multidisciplinary studies and to ascertain the antecedent stage of these cultures. We hope that the further continuing investigation in the area will testify the reliability of the archaeological findings and historical records in the region.

Pashmina wool was the main item of trade transacted between western Tibet and Kashmir via Ladakh. It is the woolen fabric woven form the raw Pahsmina and is grown in the rarified climate of trans-Himalayan region by a particular breed of domestic goat. There is another lucrative material locally known as toos or this is a fibre softer and finer than pashmina. It is believed in Ladakh that such finest and warmest fibre toos was taken from the unicorn animal or ibex or Tibetan antelope. The earliest depiction of unicorn is found on the Harappan seals. It is prized for its pashm but also for its meat. Seven Hedin noticed the practice of hunting the antelope by the Tibetan nomads.

As to how old and rich is the antiquity of the Upper Indus region of the western Himalaya is yet to be ascertained. For understanding the age old customs, traditions and trade pattern of the nomads of the Himalayas, the area needs to be visited again and again, because this cannot be discussed and read unless it is seen and felt.

The trade with Tibet was a fundamental part of life for the people of the western Himalayan borderland. For example the nomadic herdsmen of Chang thang were rich in wool, salt, borax and dairy products but were 94

NOMADIC SETTLEMENT AND ECONOMY OF WESTERN HIMALAYA (UPPER INDUS) Bibliographical References Brookfield, M. E. (1993) ‘Quartanary Deposits along the Indus Suturp Zone and Evolution of the Himalayan Rivers’. Current Science, 64 (11 &12): 904-907. Burbank, D. W. and Fort M. (1985) ‘Bed Rock Control Glacial limits; Examples from the Ladhak and Zanskar Ranges’. North West Himalaya, India, Journal of Geology, 31 (108): 143-149. Chang, K. C. (1968) The Archaeology of Ancient China, 2nd ed., New Haven, Yale University Press. Dani, A. H. (1987). History of Northern Areas of Pakistan (Islamabad): 96. De- Terra, H. & Patterson T. T. (1939) Studies on the Ice age of India and Associated Human culture, Washington: No 439. Fonia, R. S. (1993) ‘Ladakh Corridor to Central Asia’. Journal of Central Asian Studies, , University of Kashmir: 35-40. Fonia, R. S. (1997) A Report on ‘Excavation at Malpur, Akhnoor, distrit Jammu’, submitted to the Director General, ASI. Franke, A. (1930) ‘Notes on collection of stone Implements from Ladhak’. Indian Antiquity, 32: 384-391. Franke, H. (1914) ‘The antiquities of Indian Tibet’. Archaeological Survey of India, Series XXXVIII, New Delhi-Calcutta: 69. Jarrige J. F., Lechevallier M. (1979) ‘Excavation at Mehrgarh, Baluchistan: Their significance in the Prehistorical context of the Indo-Pakistan Borderlands’. In South Asian Archaeology, 1977: 463535. Jettmar, K. (1967) ‘Ethnological Research in Dardistan’, Preliminary report. Pande, B. M. (1971) ‘Neolithic Hunting Scene in a stonelab from Burzahom, Kashmir’. Asian Perspectives, 14: 134-138. Pande, B. M. (1972) ‘A Neolithic ‘Tectiform’ from Burzahom, District Srinagar, Kashmir’. Indian Anthropological Society Journal, 7 (2): 175-177. Sahni, Birbal (1936) ‘The Himalayan Uplift since the advent of Man; Its culthistorical significance’. Current Science, 5 (10). Sharma K. K., Rajagopalam G. and Choubey V. M. (1989) ‘Radio Carbon Dating of charcoal found at Pre-Indus Civillization, Fire place, Upper Indus Valley Ladhak’. Current Science, March 20: 306308. Thapar, B. K. (1970) The Aryans: A Reappraisal of the problem. India’s contribution toward Thought and Culture, Vivekananda Commemorative Vol. Lokesh Chandra ed. Madras: 171. Tucci, G. (1977) ‘On Swat: The Dards and Connected Problems’. East and West, 27: 99-103. 95

R. S. Fonia

Table 1 - List of Prehistoric Sites 1. Tirisha lake, Nubra, Leh

2. Hunder Dol, Nubra, Leh

3. Cave Shelter, Gaik, Leh

4. Rock Shelter, Kesar, Leh

5. Narla, Kargil

6. Malpur, Akhnoor, Jammu

Table 2 – List of sites with rock engravings BALTIK REGION, DISTRICT LEH

NUBRA REGION, DISTRICT LEH

No.

No.

Site name

Geo-coordinates

Site name

Geo-coordinates

1.

Achni Thang

34° 30΄ 30΄΄ N; 76° 37΄ 30΄΄ E

1.

Khardong

34° 24΄ 10΄΄ N; 77° 39΄ E

2.

Biamah

34° 36΄ 30΄΄ N; 76° 31΄ 15΄΄ E

2.

Khallar

34° 29΄ 30΄΄ N; 77° 42΄ E

3.

Chuli Chang

34° 30΄ N; 76° 19΄ E

3.

Diskit

34° 33΄ 10΄΄ N; 77° 33΄ 30΄΄ E

4.

Darchik

34° 38΄ N; 76° 26΄ E

4.

Hundar

34° 35΄ 20΄΄ N; 77° 27΄ 50΄΄ E

5.

Baldes

34° 36΄ 15΄΄ N; 76° 30΄ 30΄΄ E

5.

Skampuk

34° 37΄ 30΄΄ N; 77° 26΄ 20΄΄ E

6.

Garkun

34° 38΄ N; 76° 26΄ E

6.

Skuru

34° 40΄ 30΄΄ N; 77° 16΄ 30΄΄ E

7.

Gurgundo

34° 39΄ 15΄΄ N; 76° 20΄ 30΄΄ E

7.

Tirat (Tirit)

34° 32΄ 30΄΄ N; 77° 38΄ 30΄΄ E

8.

Hanu Thank

34° 34΄ N; 76° 20΄ 30΄΄ E

8.

Sumur

34° 37΄ 20΄΄ N; 77° 37΄ 10΄΄ E

9.

Hanu Yogma

34° 36΄ N; 76° 37΄ E

9.

Tiger (TiKhar)

34° 38΄ 20΄΄ N; 77° 37΄ E

10.

Lhasthang

34° 36΄ 30΄΄ N; 76° 31΄ E

10.

Panamik

34° 47΄ 30΄΄ N; 77° 32΄ 30΄΄ E

11.

Sanjak

34° 35΄ N; 76° 31΄ 30΄΄ E

11.

Murginala

34° 45΄ 50΄΄ N; 77° 31΄ 50΄΄ E

12.

Skyurbanchan

34° 26΄ N; 76° 43΄ E

12.

Yensa, Densa

34° 48΄ 30΄΄ N; 77° 30΄ 10΄΄ E

13.

Dumkhardo

34° 23΄ 30΄΄ N; 76° 46΄ E

14.

Takmaching

34° 23΄ N; 76° 46΄ E

CHANGTHANG REGION, DISTRICT LEH

ZANSKAR REGION, DISTRICT KARGIL

No.

No.

Site name

Geo-coordinates

Site name

Geo-coordinates

Likche Niornis

33° 43΄ 40΄΄ N; 76° 57΄ 30΄΄ E

Phey

33° 31΄ N; 76° 44΄ E

33° 26΄ N; 78° 12΄ 30΄΄ E

Tonde

33° 30΄ N; 77° 00΄ E

Zagla

33° 41΄ 30΄΄ N; 76° 59΄ E

Drangtse Nyoma

34° 12΄ N; 77° 20΄ 20΄΄ E

Kidmag

33° 23΄ 40΄΄ N; 78° 16΄ 30΄΄ E

Sakti

33° 59΄ 40΄΄ N; 77° 48΄ E

96

NOMADIC SETTLEMENT AND ECONOMY OF WESTERN HIMALAYA (UPPER INDUS)

Fig. 1a - Map Showing Longitudinal Belts of the Himalayas

Fig. 1b – North-South Cross-Section of the Himalayas.

97

R. S. Fonia

Fig. 2a – Archaeological Site in Upper Indus (Ladakh)

Fig. 2b – Hunder Cave, Nubra, Ladakh

98

NOMADIC SETTLEMENT AND ECONOMY OF WESTERN HIMALAYA (UPPER INDUS)

Fig. 2c – Flake core, Tirisa Lake, Nubra Valley, Ladakh (left); Flake core, Hundar Cave, Nubra, Ladakh (right).

Fig. 3 – Prehistoric site, Tirisha lake, Nubra Valley, Ladakh.

99

R. S. Fonia

Fig. 4 - Rock engravings. Hunting scenes.

Fig. 5 – Rock engravings. Ibex and other animals.

Fig. 6 – Rock engravings. Human Activities

100

NOMADIC SETTLEMENT AND ECONOMY OF WESTERN HIMALAYA (UPPER INDUS)

Fig. 7a – Nomadic settlements.

Fig. 7b – Nomadic seasonal settlements.

Fig. 8 – Engraving of ibex in folk pattern on the smooth black surface of rock.

101

R. S. Fonia

Fig. 9 – Depiction of palm and foot impression.

Fig. 10 – Simple bow and arrow are the only weapons represented in the engravings.

102

NOMADIC SETTLEMENT AND ECONOMY OF WESTERN HIMALAYA (UPPER INDUS)

Fig. 11a – Hinghly standardized and abstract representation of a hut with a domed roof.

Fig. 11b – Cyclopean wall enclosing a shrine.

103

R. S. Fonia

Fig. 12 - Sulphure mines, Pugga, Changthang, Ladakh.

Fig. 13 – Structure showing the residence of a nomadic chief. Likche, Changthang, Ladakh.

104

NOMADIC SETTLEMENT AND ECONOMY OF WESTERN HIMALAYA (UPPER INDUS)

Fig. 14a - Engraving depicting Trident and Swastik symbol. Nyoma, Changthang, Ladakh.

Fig. 14b - Engraving depicting symbol of Sun and Swastik. Gurukun, Ladakh.

105

EXPLORING INNER ASIA’S HIGH ALPINE FRONTIER: HIGH ALPINE TRANSHUMANT PASTORALISM, VERTICAL CULTIVATION AND ENVIRONMENTAL ARCHAEOLOGY IN THE LOWER VAKHSH-PANJ CONFLUENCE AND GORNO-BADAKHSHAN REGIONS, SOUTHERN TAJIKISTAN Irene Good

areas, the Zerevshan and Hissar Valleys to the north and the narrow river valleys of Kafirnigan, Vakhsh, Kyzylsu and tributaries Obikik and Yavansu to the south. From a broad scale view, Tajikistan is like Afghanistan: a crossroads between the Khorasan and Seistan regions to the south and west, the variegated steppe and foreststeppe lands to the north, the desert oasis regions of the Tarim Basin to the east, and the Indus Valley to the southeast34. Although the high mountain plateaus of the Pamirs and Hindu Kush have been widely assumed to be barriers to human movement and interaction, in fact there has been activity in prehistoric times in even the most remote parts of this whole area (see, for example, Sinor 1990; Stacul 1987; Jettmar 1959; Mason et. al 1939). There is evidence of single grave burials, ancient mines and petroglyphs deep in the Pamirs (see Masov et al. 2005; Ranov and Bubnova 1961) dating back to the third and second millennia BC.

Models of interaction in Eurasian archaeology, particularly for the bronze and iron ages, have centered on large-scale movements of peoples explained by principally economic or climatic drivers, for example in the exchange of exotic goods, or of the expanding steppe zone due to increased aridity. These models have relied on a simplistic ‘steppe nomad’ versus ‘settled farmer’ dichotomy, with mixed agro-pastoral systems consistently overlooked. Tribal exogamy and gender and age-based activities also contribute to the complexity of interaction in prehistoric Eurasia. This paper considers social dimensions of interaction and in particular discusses the importance of highland and high alpine landscapes, not as strictly wild space but also as an emerging cultural space, especially as mining metal ores and precious stones increased. This highland and alpine space became an integral region of interaction in Central Asia, with remnants of ancient lifeways still found among rural Pamiri peoples today.

The region of present-day Tajikistan has, in fact, played a critical role in the development of complex societies within Central Asia. From the widespread distribution of lapis lazuli across western Asia, found principally from the region of Badakhshan (Casanova 1999; 1989; Amiet 1986; Tosi 1974) but also from the Syludyanka River near Lake Baikal and from the Khorog region in the Pamirs (Good 2007); and from very early evidence for mining tin in the Zherevshan (Borrofka et. al 2002), we can assert that the particular characteristics of intermontane landscape within Tajikistan have played a pivotal role in the interaction between peoples and in the development of culture histories in the Transoxiana.

The archaeology of southern Tajikistan is poorly understood at present, (Fig. 1) and that of the Pamirs is virtually unknown. New investigation33 of prehistoric sites within Tajikistan with an environmentally based research design is beginning to greatly enhance our understanding of the cultural interaction between prehistoric peoples within Central Eurasia. As part of a project jointly conducted under the auspices of the Tajik Academy of Sciences, the National Museum of Antiquities of Tajikistan, Harvard University’s Department of Anthropology and the Peabody Museum, we have begun a series of archaeological and ethnographic field reconnaissance studies in order to produce valuable data on land use within both highland (below timberline) and high alpine (above timberline) landscapes in southern Tajikistan. This paper is a brief discussion of some initial findings.

The Pamirs and deeply cut river valleys of Vakhsh and Kafirnigan are difficult and complex landscapes, yet have invited human occupation for well over 10 millennia. The eastern half of Tajikistan is the High Pamir, with both plateau (to the north and east) and deeply cut montane valley systems (to the south and west). The western half of Tajikistan is also geographically divided into two

There are specific ecological distinctions between two types of highland zone: a midland-highland regime as exemplified in the Qurghonteppa region of the Vakhsh Valley, and a highland-high alpine regime, as exemplified by the western Pamirs and the Wakhan Corridor in Gorno-Badakhshan. Within these highland zones there are distinctive, geographically defined subregions in southern Tajikistan; from low, flat, narrow alluvial floodplains in braided rivers to steep escarpments, scrub and parkland slopes. In some areas semi-arid and poorly covered lithosols dominate; in other areas hills are lush with vegetation by constant moisture. Water is one of the two prime factors of this vegetational mosaic; the other is elevation. Within the Wakhan Corridor, an even more complex landscape emerges,

33

34

The Vertical Mosaic: Interplay of Micro-Environments

Research supported by generous funding from the American School of Prehistoric Research.

Unlike Afghanistan, however, Tajik archaeology has not been severely disrupted by looting, the ravages of civil war or the Taliban.

107

Irene Good their flocks and herds to high ground (up to 4000 m) 35, while women, small children and older people stay in summer huts making yoghurt and tending fields (Fig. 5, 6). In winter, they reside in hamlets down below in the deeply cut river valleys. People move an average of approximately 200 km. from winter to summer venues 36.

including forested mountains, waterfalls, fast-moving streams flowing into high velocity rivers, slopes of glacial till with terraced fields and footpaths surrounding small (still unmapped) hamlets (Fig. 2, 3). The Vertical Archipelago Model Traditional highland cultivation and animal husbandry practices within southern Tajikistan and GornoBadakhshan echo the vertical archipelago of the preColumbian Andes (Murra 1968; Quattrin 2001). A vertical economy is one in which a highland-lowland interface is fully integrated and interdependent. Vertical transhumance is one agent by which this economic interdependence forms. Within a vertical economy model, human adaptation to vegetational differentiation is based on altitude, and a relatively small area can carry a highly dense mosaic of ecozones. This model allows us to view these complex landscapes not as wilderness but as cultural space; and as space that is comprised of local nodes in a regional network which is in turn connected with other regions, either by contiguous populations or by individuals or groups moving through them. Testing this model requires high resolution in research design, aimed toward recovering seasonality data as well as anthropogenic effects on a highly varied landscape.

The second type of vertical movement is a diurnal, shortrange shift. This type is typical in the highlands of Khatlon. In the hills east of the Qurghonteppa region, throughout the year cattle and flocks of sheep and goat are let out during the day to pasture in the hills and are looked after by men and boys, and they return at dusk each day. In the uplands east of Qurghonteppa (up to ca. 2000 m) mixed herds graze all year long unless there is snow cover. Fodder is collected for weak and sick animals as well as for winter when snow is on the ground. Here the vegetation varies from open grassland to patches of remnant oak-pistachio forest with xerophytic plants such as Cedrus spp. growing around intermittent water sources. It is hypothesized here that these agro-pastoral lifeways have a deep antiquity, and that their distinction leaves a discernable signature in anthropogenic soils. These preliminary studies are helping to frame specific questions about land use, plant use and the impact of grazing practices. These studies will help in the interpretation of archaeological data of ephemeral sites, land use and seasonality, allowing us to consider these hitherto unexplored highland and high alpine landscapes as an integral developmental aspect of prehistoric Transoxiana.

Preliminary Observations Sheep, goat and cattle are today the principal livestock in Tajikistan. Tajik and Pamiri pastoral lifeways have been witnessed in fall and spring, with close attention to both plant and animal use, in the hilly region between Qurghonteppa (in the Vakhsh Valley) and Kulyab in Khatlon, and also in the Wakhan. There are two kinds of vertical movement. The first type is seasonal transhumance (i.e. for the duration of the summer months). This type is typical in the Pamirs. Above timberline (ca. 2500 m) are dry steppe-like thin soils holding xerophytic ground-covering plants used for summer grazing (Fig. 4). The Wakhan Corridor is host to the highest elevation cattle grazing in the world. Mixed sheep/goat herds are tended by shepherds and cattle are left alone to graze in summer on sometimes very steep slopes. At first glance the hillsides appear covered with wild waste vegetation, but in nearly every place where plants can take hold, there are cultigens or domesticates tended by local people (pomegranate, pistachio, apricot, birch). Shepherds supplement their summer livelihood by collecting rhubarb, wild mushrooms and other easily cultivated plants to sell or barter with people on the road. Brush and wood is collected by children on a daily basis for cooking, and fodder is collected for cattle. Langar, a small town at the mouth of the Wakhan Corridor, is poised at the crux between the western Pamir and the eastern High Pamir and Murghab Plateau. In Langar, grazing and herding practices consist of cattle, sheep and goat grazing in the high meadows, often on steep inclines of between 40-55°. Shepherds (men and older boys) bring

Past Environments, Past Lifeways The farming cultures of Vakhsh and Beshkent (ca. 14001100 BC) have been described as ‘interacting’ with the so-called steppe bronze pastoralists as evidenced by the occasional presence of bronze objects, particularly in Beshkent graves, that are similar to ones from to the north at Zaman Baba (Litvinsky and P’iankova 1994; P’iankova 1994, 1996b; Vinogradova 2004). These farming communities are thought to have developed out of the agricultural cultures of southeastern Uzbekistan, in particular the late phase of Sapalli culture. I would argue that the development of bronze cultures in southern Tajikistan is indigenous. Vakhsh and Beshkent people had interaction with cultures to the west (Sapalli, Molali) and also to the north (Sarazm, Zaman Baba) but are 35 Parkes (Parkes 1984) has suggested that highland pastoralists all across Eurasia have a highly dichotomized worldview principally based on gender role differences; that women tend small farms in the valleys and men herd animals in the mountains. These gender-role differences are highly pronounced, to the point of antagonism. Within the male (high elevation) domain is purity, represented by the ibex or Markhor goat, solitude, the fragrant juniper plant, and the colour blue. These aspects are juxtaposed against the female (low elevation) domain, where impurity is symbolized by sheep, village life, the madder plant, and the colour red. 36 Personal communication, R. Nuroloev, my informant in 2005, a Wakhi from Ishkashim.

108

EXPLORING INNER ASIA’S HIGH ALPINE FRONTIER: HIGH ALPINE TRANSHUMANT PASTORALISM, VERTICAL CULTIVATION AND ENVIRONMENTAL ARCHAEOLOGY IN THE LOWER VAKHSH-PANJ CONFLUENCE AND GORNO-BADAKHSHAN REGIONS, SOUTHERN TAJIKISTAN Mediterranean, in a web of complex exchange networks that became the template for the later historic Silk Road. This project aims to recover evidence for these humanenvironmental interactions in prehistory in this poorly understood region, particularly regarding seasonal vertical transhumance. It is hypothesized here that the highland peoples of the Wakhan and Vakhsh regions, contrary to conventional thought, have been by tradition connected to a wider web of human-ecological zones, in a vertically structured, ecozonally ordered interaction that not only predates the Silk Road but has played a significant role in shaping it. This participation in exchange with the wider world, particularly from the third millennium through the second, has been hitherto unrecognized in the archaeological record due to poorly documented chronologies and emphasis on the more westerly sites, and from illicitly excavated materials from Afghanistan. Now we are beginning to develop a more cohesive picture of the development of the Oxus civilization, and the hope here is to continue an integrated program of archaeological recovery with emphasis on environmental aspects, by examining landscapes to better understand the nature of integration between pastoralism and agriculture; the nature of highland agriculture and its antiquity, and the nature of the so-called ‘steppe bronze’ culture, its lifeways, and the extent of catchment areas for procuring food and craft resources. By using an environmental approach to the research design of both landscape survey and excavation, we hope to answer important questions concerning the early interaction of peoples and the prehistoric development of cultures in Central Asia.

distinctive primarily because of an adaptation to highland landscape and also through connection with high alpine peoples further to the east. We are now beginning to see evidence for earlier antecedents of Vakhsh culture (see Lombardo 2011; Vinogradova 2011), which may or may not have connection with the very archaic Hissar Neolithic. Highland, as well as high alpine zones, were utilized in southern Tajikistan in antiquity as they are today. I hypothesize that the early inhabitants of these highland zones participated in a mixed agro-pastoral economy based on vertical transhumance (see Khazanov 1984). This in turn developed increasingly specialized secondary product use from differentiated herd compositions, particularly regarding animal fiber production (see Good 2006). I hypothesize that this economy of agro-pastoral production preceded the specialized nomadic pastoralism of the steppes; and indeed that pastoral nomadism developed out of it. Clear evidence of high alpine resource use comes from second millennium finds in the Tarim Basin in Xinjiang. Cashmere was found in cloth from lowland oasis sites along the southern rim of the Tarim Basin ca. the midsecond millennium BC (Debaine-Francfort 2002; Good 1998). Unlike other animal fibers, cashmere is not from a specific breed but rather from domestic goats specifically adapted to a high altitude regime. The peoples from Chärchän and Djamboulaq in the southwestern Tarim Basin had access to products from a high alpine environment, either as seasonal transhumants themselves or through a network of economic interaction with highlanders. This fact has prompted a closer examination of highland and high alpine environments as areas of not only resource extraction but also of resource production.

Acknowledgements

In order to test these hypotheses, a comprehensive research program of ecology, ethnography and archaeology (including new series of radiocarbon dates) has begun in order to recover data on several fronts, natural and cultural, contemporary and diachronic, with a particular focus on verticality. Evidence for seasonality, for anthropogenic alteration of the flora, and for finetuned distinctions in grazing patterns between sheep, goat and cattle are currently being investigated by employing quantifiable environmental data from pollen, recovered from a variety of contexts with tailored sampling. Changes in land use can be evaluated through careful microstratigraphic study of pollen rain from both pristine and anthropogenically altered contexts, particularly regarding arboreal to non-arboreal plant ratios through time. This study will greatly increase our understanding of human interaction in Eurasian prehistory, as it addresses regions assumed to be barriers to movement.

The author wishes to thank: Karl Lamberg-Karlovsky, Saidmurod Bobomuloev, Abdurauf Razzakov, Sharof Kurbanov, Natalya Vinogradova, Giovanna Lombardo, Yuri Kutimov, Alexei Netch, Mirali Kaganov, Ruslan Nuriloev, Markus Hauser, Myra Bubnova, Gary Bowersox, Zsolt Kasztovszky, Judit Zoefoeldi, and Hafiz Faizulaev.

Bibliographical References Amiet, P. (1986) L’Age des Echanges des Inter-Iraniens 3500-1700 av. Paris: JC. Editions de la Réunion des Musées Nationaux, 1986. Askarov, A. (1978) Drevezemledl’cheskaya kul’tura epokhi bronzy juga Uzbekistana. Tashkent. (1973) Sapallitepe. Fan, Tashkent. (unpublished translation by H. Michael). Banning, E. B. (1996) ‘Highlands and Lowlands: Problems and Survey Frameworks for Rural Archaeology in the Near East’. Bulletin of the American Schools of Oriental Research, 301: 2545.

Summary This research program is based in part on the premise that the isolated, fertile deep-cut mountain valleys of southern Tajikistan participated in a highly integrated vertical economy in prehistory; and that later trade links brought these communities into a larger sphere of interaction with the ancient civilized world, from China to the 109

Irene Good Good, I. (2006) New Views of the Tarim Mummies of Xinjiang. Lecture delivered at the East Asian Archaeology Seminar Series, Harvard University (May 2006). Good, I. (1998) ‘Bronze Age Textiles from the Tarim Basin: The Cherchen Evidence’. In Mair, V. (ed) Bronze and Early Iron Age Peoples of Eastern Central Asia, Journal of Indo-European Studies and the University of Pennsylvania Museum Publications. Hiebert, F. and Shishlina, N. (1998) ‘Between the Steppe and the Sown’. In Mair, V. (ed) Bronze and Early Iron Age Peoples of Eastern Central Asia, Journal of Indo-European Studies and the University of Pennsylvania Museum Publications. Isakov, A. I. (1996) ‘Sarazm: An Agricultural Center of Ancient Sogdiana. The Archaeology and Art of Central Asia - Studies from the Former Soviet Union’. Bulletin of the Asia Institute New Series, 8: 1-12. Jarrige, J. F. (1988. ‘Les Styles de Geoksyur et de Quetta et la Question des Rapports entre les Régions au Nord et au Sud de l’Hindu Kush a fin du 4 e et au debut du 3e millénaires’. In L’Asie Centrale et ses Rapports avec les Civilizations Orientales, des Origines a l’Age du Fer. Actes du Colloques Franco-Sovietique. Memoires de la Mission Archaeologique Francaise en Asie Centrale, Tome I: 95-115. Jettmar, K. (1959) ‘Ethnological Research in Dardistan’. Proceedings of the American Philosophical Society, 105 (1): 79-97. Khazanov, A. (1984) Nomads and the Outside World, 2nd edition, Madison: U. of Wisconsin Press. Khlopina, L. I. (1972) ‘Southern Turkmenia in the later Bronze Age’. East and West, 22 (3-4): 199-214. Kohl, P. (2002) ‘Archaeological Transformations: Crossing the Pastoral/Agricultural Bridge’. Iranica Antiqua, 37: 151-190. Kohl, P. (1992) ‘Central Asia (Western Turkestan): Neolithic to the Early Iron Age’. In Enrich, R. (ed) Chronologies in Old World Archaeology, 3rd Edition, University of Chicago Press: 179-194. Kohl, P. (1984) Central Asia: Palaeolithic Beginnings to the Iron Age. Reehereche sur les Civilizations. Paris: CNRS. Koryakova, L. (1994) ‘Central Asia in the Bronze Age: Sedentary and Nomadic Cultures’. Antiquity, 68: 355-372. Litvinsky, B. A. and P’iankova, L. T. (1992) ‘Pastoral Tribes of the Bronze Age in the Oxus Valley (Bactria)’. In Ahmad Hasan Dani, Masson V. M., Harmatta J., Litvinovsii B. A., Bosworth C. E. (eds) History of Civilizations of Central Asia. vol. I. The Dawn of Civilization from Earliest times to 700 BC, Paris: Unesco: 379-395.

Barger, E. (1939) ‘Exploration of Ancient Sites in Northern Afghanistan’. The Geographical Journal, 92 (5): 377-391. Borrofka N., Cierny J., Lutz J., Parzinger H., Pernicka E. and Weisberger G. (2002) ‘Bronze Age Tin from Central Asia: Preliminary Notes’. In Boyle K., Renfrew C. and Levine M. (eds) Ancient Interactions: East and West in Eurasia, Cambridge: McDonald Institute Monograph: 135159. Casanova, M. (1999) ‘Le Lapis-lazuli dans l'Orient ancien’. In Cornaline et pierres précieuses. La Méditerranée, de l'Antiquité à l'Islam, Actes du colloque établie par A. Caubet, Paris 1995. Louvre conférences et colloques. Documentation Française, Paris: 189-210. Casanova, M. (1989) ‘The Lapis Lazuli Sources in the Ancient East’. South Asian Archaeology 1987. Debaine-Francfort, C. and Abduressul, I. (2000) Keriya, Mémoires d’un Fleuve- archaeologie et civilization des oasis du Taklamakan, France: Editions Findakly. Doebrich, J. and Wahl, R. (2006) The Geologic and Mineral Resource Map of Afghanistan. Presented at the annual meeting of the American Geological Society, Philadelphia. Duprée, L. (1972) ‘Prehistoric Research in Afghanistan’. Transactions of the American Philosophical Society, New Series, 62 (4). Finkelstein, I. and Perevolotsky, A. (1990) ‘Processes of Sedentarization and Nomadization in the History of Sinai and the Negev’. Bulletin of the American Schools of Oriental Research, 279: 67-88. Frachetti, M. and Chippindale, C. (2002) ‘Alpine space, Alpine time and Prehistoric Human Experience’. In Chippindale, C. and Nash, G. (eds) European Landscapes of Rock-Art, London: Routledge: 116143. Frankfort, H. P. (2005) ‘La Civilization de l’Oxus et des Indo-Iraniens et Indoaryens en Asie Centrale’. In Fussman G., Kellens J., Francfort H.P., Tremblay X. (eds) Aryas, Aryens et Iraniens en Asie Centrale. Paris: College de France Institut de Civilization Indienne fs. 72: 253-332. Gardin, J. C. (1998) Prospections Archéologiques en Bactrianne Orientale (1974-1978), Vol. 3. Description des sites et notes de synthese. Memoires de la Mission Archaeologique Francaise en Asie Centrale tome IX. Good, I. (2007) Sourcing Beauty: Exploring Inner Asia’s High Altitude Frontier (unpublished paper for the 2007 meetings of the Society for American Archaeology, Austin, TX. Symposium entitled ‘Inter-regional Interaction in Neolithic and Bronze Age China: A Multidisciplinary Approach’ chaired by Y.T. Li).

110

EXPLORING INNER ASIA’S HIGH ALPINE FRONTIER: HIGH ALPINE TRANSHUMANT PASTORALISM, VERTICAL CULTIVATION AND ENVIRONMENTAL ARCHAEOLOGY IN THE LOWER VAKHSH-PANJ CONFLUENCE AND GORNO-BADAKHSHAN REGIONS, SOUTHERN TAJIKISTAN Lombardo G. (2012) ‘Is it possible to speak about Middle Bronze Age in Southern Tajikistan?’. In Frenez D. and Tosi M. (eds) South Asian Archaeology 2007. Vol. 1, Prehistoric Periods. Proceedings of the 19th International Conference of the European Association of South Asian Archaeology July 2007. (Ravenna, Italy, 2-6 July 2007) BAR International Series. Oxford, BAR Publishing: 185-192.

P’iankova, L. T. (1996b) ‘Southwestern Tajikistan in the Bronze Age. Central Asia: the Beginnings’. International Association for the Study of Ancient Civilizations of Central Asia Bulletin, 1: 34-46. P’iankova, L. T. (1994) ‘Central Asia in the Bronze Age: Sedentary and Nomadic Cultures’. Antiquity, 68: 355-372. P’iankova, L. T. (1989) Drevnie skotovody juzhnogo Tadzhikistana (po materialam mogil’nika epokhi bronzy ‘Tigrovaja Balka’). Dushanbe: Donish. Quattrin, Dale W. (2001) Prehispanic Chielfdoms in the Valle de la Plata, vol. 4: Vertical Economy, Interchange, and Social Change During the Formative Period. Memoires in Latin American Archaeology 11. University of Pittsburgh Latin American Archaeology Publications and Universidad de los Andes, Bogotá. Ranov V., and Bubnova M. (1961) ‘Uncovering the History of the Roof of the World’. American School of Archaeology, 65 (4): 396-397. Stacul, G. (1987) Prehistoric and Protohistoric Swat, Pakistan (c300-1400 BC.). IsMEO. Reports and Memoires, Vol. 20. Rome. Toderich K., Tsukatani T., Abbdusamatov M., Rakhmatullaev R., Latipov R., Khujanazarov T. (2006) A Farm in Kumsangir of Tajikistan: A Perspective of Water/Land Use Along the Pyandzh River. Kier Discussion Paper Series, Kyoto: Institute of Economic Research, Discussion Paper 619. Tosi, M. (1974) ‘The Lapis Lazuli Trade Across the Iranian Plateau in the Third Millennium BC’. In Gururajamanjarika. Studi in onore di Giuseppe Tucci, Vol I. Naples: 3-22. Vinogradova, N. M. (2004). Southwest Tajikistan in the late Bronze Age. Moscow: Ran. (in Russian). Vinogradova, N. M. (2001) ‘Südtadzhikistan in der Spätbronze- und Früheisenzeit’. In Migration und Culturtransfer. Der Wandel vorder-und zentralasiatischer Kulturen im Umbrach vom 2. Zum 1 vorchristlichen Jahrtausand. Akten des Internationalen Kolloquiums Berlin, 23, Vol. 6: 199-219. Vinogradova, N. M. (1996) ‘Kangurt Tut: The Ancient Burial Ground in Southern Tajikistan’. International Association for the Study of Ancient Civilizations of Central Asia Bulletin, 20:171-194. Vinogradova, N. M. (1994) ‘The Farming Settlement at Kangurt Tut (South Tajikistan) in the Late Bronze Age’. Archaeologische Mitteilungen Aus Iran, 27: 29-47. Vinogradova, N. M. and Kuz’mina E. E. (1996) ‘Contacts between the Steppe and Agricultural Tribes of Central Asia in they Bronze Age’. In Hiebert F. T. and DiCosmo N. (eds) Between Lapis and Jade: Ancient Cultures of Central Asia. New York: Sharpe Books: 29-54.

Mashkour, M. (2003) Tracing ancient 'nomads': isotopic research on the origins of vertical 'transhumance' in the Zagros region. NATO Advanced Research Workshop on Impact of the Environment on Human Migration in Eurasia. St. Petersburg, Russia. Mason K., Wright P., Emanuel W.V., de La Valette J., de B. Codrington K., Barger E., (1939) ‘Exploration of Sites in Northern Afghanistan: Discussion’. The Geographical Journal, 91 (5): 392-398. Masov, Rahim, M., Bobomuloev S., and Bubnova M., (eds) (2005) The National Museum of Antiquities of Tajikistan: The Album. Dushanbe: Donish Institute of History, Tajik Academy of Sciences and ACTED. Masson, V. M. (1992) ‘The Bronze Age in Khorasan and Transoxania’. In Ahmad Hasan Dani, Masson V. M., Harmatta J., Litvinovsii B. A., Bosworth C. E. (eds) History of Civilizations of Central Asia. vol. I. The Dawn of Civilization from Earliest times to 700 BC, Paris: Unesco: 225-246. Murra, J. (1984) ‘Andean Societies’. Annual review of Anthropology, 13: 119-141. Murra, J. (1980) The Economic Organization of the Inka State. JAI, Greenwich, CT. Murra, J. (1972) ‘El control vertical de un máximo de pisos ecológicos en la economia de las sociedades andinas’. In Murra J. V. (ed) Visita de la Provincia de León de Huánuco II. Huánuco, Peru: Universidad Hermilio Valdizán: 427-476. Murra, J. (1968) ‘An Aymara Kingdom in 1567’. Ethnohistory,15 (2): 115-151. Molodin, Vjaceslav I. (2001) ‘Bronzezeit im Berg-Altai’. Eurasia Antiqua, 7: 1-51. Núñez M., Okkonen J., Frachetti M., & Daly P. (1997) ‘GIS where altitude is a function of time’. In Johnson, I. & North, MacLaren (eds) Archaeological Applications of GIS: Proceedings of Colloquium II, UISPP XIIIth Congress, Forli, Italy, September 1996, Series 5. Sydney: University Archaeological Methods. Parkes, P. (1987) ‘Livestock Symbolism and Pastoral Ideology Among the Kafirs of the Hindu Kush’. Man, New Series, 22 (4): 637-60. P’iankova, L. T. (1996) ‘Bronze Age Cultures of the Steppe and Highlands in Central Asia’. In Between Lapis and Jade. Ancient Cultures of Central Asia. New York: Sharpe Publishing. 111

Irene Good Vinogradova, N. M. (2011) ‘Will the Monuments of the Middle Bronze Age be Discovered on the Territory of Southern Tadjikistan? ’. In Frenez D. and Tosi M. (eds) South Asian Archaeology 2007. Vol. 1, Prehistoric Periods. Proceedings of the 19th International Conference of the European Association of South Asian Archaeology July 2007. (Ravenna, Italy, 2-6 July 2007) BAR International Series. Oxford, BAR Publishing: 281-286.

112

EXPLORING INNER ASIA’S HIGH ALPINE FRONTIER: HIGH ALPINE TRANSHUMANT PASTORALISM, VERTICAL CULTIVATION AND ENVIRONMENTAL ARCHAEOLOGY IN THE LOWER VAKHSH-PANJ CONFLUENCE AND GORNO-BADAKHSHAN REGIONS, SOUTHERN TAJIKISTAN

Fig. 1 – Satellite image of southern Tajikistan with distribution of known bronze period sites. There are also numerous sites to the east in the GornoBadakhshan dating to the second millennium BC; including petroglyphs, evidence for mining, cave shelters, and burials with parallels to those in the Tarim Basin. (I. Good)

Fig. 2a – Kulyab region in October: Hilly terrain with evidence for cattle grazing (Photo I. Good).

Fig. 2b – Kulyab region in October: View of Kara Birdeh settlement site (Photo I. Good).

113

Irene Good

a)

b)

Fig. 3 – Terraced landscapes in Wakhan (view of the Afghan sidea), an apricot orchard in Hissar (b), and a household in the highland village of Vrang (c). c)

Fig. 4 – June cattle herding in the uplands near Langar, Wakhan (ca. 4000 m elevation) (photo I. Good).

114

EXPLORING INNER ASIA’S HIGH ALPINE FRONTIER: HIGH ALPINE TRANSHUMANT PASTORALISM, VERTICAL CULTIVATION AND ENVIRONMENTAL ARCHAEOLOGY IN THE LOWER VAKHSH-PANJ CONFLUENCE AND GORNO-BADAKHSHAN REGIONS, SOUTHERN TAJIKISTAN

Fig. 5 – Summer huts for making yoghurt, Hissar Mountains south of Aini (photo I. Good).

Fig. 6 – Mazar (shrine) wall embedded with Argali sheep horns in Wakhan. The wild Argali is venerated in the Pamirs. Parkes (1987) noted in his observation of the Kalasha Kafirs in Chitral, worship of the Markhor goat and ibex. This worship is a part of an ancient pre-Islamic pastoral symbolism that is part of what Jettmar (1959) had originally termed the ‘Dardic Complex’ in the Hindu Kush, in which the wild ibex and Markhor goat are considered sacred, due to the high elevation of their habitat. On the Pamiri side, the veneration of Argali sheep (by Ishmaili Muslims) can be seen as a parallel to this pastoral symbolism, and as remnant of pre-Islamic practice (photo I. Good).

115

A VIEW TO THE NORTH: BIOLOGICAL INTERACTIONS ACROSS THE INTER-MONTANE BORDERLANDS DURING THE LAST TWO MILLENNIA BC Brian E. Hemphill

he (1998: 565-568) claims that colonization from the Russo-Kazakh steppe generally co-occurred with a movement of East Asian populations into the eastern part of the Tarim Basin. Han suggests that peopling of the Tarim Basin ended around 500 BC with a later migration of a ‘Pamir-Ferghana type’ population from the west. Hence, Tarim Basin populations feature early contributions from ‘Europoid’ populations from the Russo-Kazakh steppe and ‘Mongoloid’ populations from East Asia, coupled with a later contribution from populations of the Pamir-Ferghana region.

Schoolchildren in the United States are taught that initial contacts between East and West occurred in the 13 th century AD with the famous travels of Marco Polo to the court of Kublai Khan (Komroff, 1930). By contrast, popular perception in the East holds that contacts stem from the travels of Zhan Qian to the West in 132 BC (Di Cosmo, 1996). Recent archaeological discoveries in Central Asia suggest that even this date is far too recent (Barber, 1999: 207-208). In the 1970s, silk was discovered accompanying a burial at Sapalli tepe in Uzbekistan, and Askarov (1973: 133-134) interpreted this finding as proof of contact between East and West as early as the 2nd millennium BC. Later that decade, discovery of woolen textiles as well as evidence of the cultivation of wheat and barley at Qäwrighul in Xinjiang provided additional evidence of early contact between East and West (Mallory and Mair, 2000: 136-140).

Of crucial importance to Han’s tripartite model are human remains from Yanbulaq. According to Han (1990, 1998: 561), it is here that evidence of initial contact between the peoples of East and West may be found, for his analysis of 29 relatively complete crania from this site led him to identify eight as ‘Europoid’ and the remaining 21 as ‘Mongoloid’.

Qäwrighul represents just one example of Late Bronze/Early Iron Age sites that have been discovered in a series of oases along the periphery of the Taklamakän Desert (An Zhimin, 1992). This study focuses on the sites of Krorän, Qäwrighul and Alwighul, but of special importance is the site of Yanbulaq, located at the eastern periphery of the Tarim Basin. Excavation of these sites yielded not only woolen textiles, but also wheeled vehicles and metal objects, none of which appears to have any local antecedents (Kuzmina, 1998). Perhaps of greatest interest was the discovery of a series of unintentionally mummified individuals. These discoveries led many, such as Mair (1995: 289) to query; who were these people, and where did they come from? The best-known explanation has been offered by Mallory and Mair (2000), who assert the Late Bronze/Early Iron Age populations of the Tarim Basin were colonists from the Russo-Kazakh steppe who came to this unoccupied region. Their model calls for two waves of immigration; an early immigration of Afanasievo peoples, followed by a later immigration of Andronovo populations. Mallory and Mair offer myriad evidence in support of their model including linguistic evidence for the presence of an undifferentiated Indo-European language known as Tocharian, the presence of bronze objects bearing close similarities to those found on the Russo-Kazakh steppe, and iconography that depicts individuals of distinctly non-East Asian appearance.

The purpose of this study is to test Han’s tripartite model for Tarim Basin population origins by testing two assertions fundamental to this model. First, there is evidence of two biologically very different populations (‘races’) at Yanbulaq. Second, these remains from Yanbulaq provide evidence of early contact between populations of East and West. That is, given the attributions made by Han (1990), one set is affiliated with ‘Europoids’ from the Russo-Kazakh steppe while the other shares closest affinities with ‘Mongoloids’ from East Asia. Materials and Methods The basis for assessing biological affinities of the crania recovered from Yanbulaq is eight cranial measurements, two of the vault and six of the face, taken in accordance with the standards established by Martin (1928). Though far from ideal, this battery represents the best compromise between two opposing factors: the number of measurements forming the basis of comparison, and the number of individuals represented by a complete set of data. With this battery of eight measurements, 23 of the 29 Yanbulaq crania (16 ‘Mongoloid’, 7 ‘Europoid’) were retained for multivariate comparison. The Yanbulaq cranial series is compared to 29 other cranial series comprising a total dataset of 1505 individuals (845 males, 660 females) ranging in antiquity from the Neolithic to modern (see Hemphill and Mallory 2004: Table 1). The dataset encompasses nine samples from the Russo-Kazakh steppe, eight eastern Central Asian (Xinjiang, China) and East Asian samples, seven samples from western Central Asia (Turkmenistan,

The most extensive attempts to identify the biological origins of the Bronze Age Tarim Basin skeletal remains (crania) have been by Han Kangxin. In 1998, Han offered what may be referred to as the ‘tripartite model’. Han asserts that initial peopling of the Tarim Basin may be attributed to immigration of pastoralist populations from the Russo-Kazakh steppe, but unlike Mallory and Mair, 117

Brian E. Hemphill results not only fail to identify the presence of two biologically distinct populations at Yanbulaq, they also confirm that the effects of sex dimorphism are effectively controlled for by this method of size standardization.

Uzbekistan), three samples from the Iranian Plateau (Iran), and three samples from the greater Indus Valley (Pakistan) (see Hemphill and Mallory, 2004: Figure 1). Different statistical approaches are used to address the two assertions fundamental to Han’s tripartite model. The assertion that two very different biological populations are present at Yanbulaq (Han, 1998: 561) is tested by size adjusting each individual. This is accomplished by dividing each measurement by the geometric mean of all eight measurements by individual (Jungers et al., 1995). This removes the effect of differential size between individual crania, which is especially important when both males and females are compared, so that biological differences due to geographic origin are not conflated by sex dimorphism. Once adjusted for overall size, crania are submitted to Ward’s (1963) hierarchical cluster analysis to assess the patterning of intersample similarities. Han’s second assertion that the remains recovered from Yanbulaq provide evidence of contact between East and West is tested by quantifying pairwise differences between samples with Mahalanobis generalized distance. Patterning of inter-sample affinities is simplified with two different multivariate techniques: neighbor joining cluster analysis (Saitou and Nei, 1987) and principal coordinates analysis (Hair et al., 1971). In the latter analysis results are ordinated in threedimensional space, and a minimum spanning tree (Sneath and Sokal, 1973) is imposed to ease interpretation of intersample associations.

East Meets West at Yanbulaq? Contrary to the expectations of Han’s model, a neighborjoining analysis among males (Fig. 2a) reveals that ‘Europoid’ males from Yanbulaq share closest affinities to ‘Mongoloid’ males from this same site, followed by secondary affinities to males from other Tarim Basin Bronze Age sites. Han’s (1998: 561) assertion that the ‘Mongoloids’ from Yanbulaq may be identified as related to Khams Tibetans also appears unsupported, for of all ‘Mongoloid’ samples included in this comparison, the inhabitants of Yanbulaq are most dissimilar to Tibetans. There also appears no basis to identify the inhabitants of Yabulaq as ‘Europoids’, for none of the ‘Europoid’ males from the Russo-Kazakh steppe or males associated with ‘steppe-influenced’ archaeological cultures (KOK, TMM; see Alekshin, 1986; Masson, 1992) exhibit any affinity to the alleged ‘Europoid’ males from Yanbulaq. Rather, apart from other Bronze Age samples from the Tarim Basin, males from Yanbulaq share closest affinities to males from the greater Indus Valley. Principal coordinates analysis yields similar results (Fig. 2b). ‘Europoid’ and ‘Mongoloid’ males from Yanbulaq share closest affinities to one another and to males from Qäwrighul, the earliest of the Tarim Basin sites. None of the Tarim Basin sites include males with affinities to East Asian males. Rather, males from later Tarim Basin sites, such as Alwighul and Krorän, show closest affinities to western Central Asian males from Sapalli tepe and Djarkutan. Likewise, none of the males from Yanbulaq, including those identified by Han as ‘Europoids’, exhibit affinities to males from the Russo-Kazakh steppe. Again, secondary affinities of Yanbulaq males, and males from Qäwrighul, occur with males from the greater Indus Valley.

Results Expectations of Han’s model are straightforward. First, if two biologically very different populations were present at Yanbulaq, hierarchical cluster analysis should provide a profound separation between ‘Europoid’ and ‘Mongoloid’ individuals. Second, if ‘Europoids’ and ‘Mongoloids’ were present at Yanbulaq, ‘Europoids’ from Yanbulaq (YANE) should be identified as most similar to ‘Europoids’ from the Russo-Kazakh steppe and ‘Mongoloids’ from Yanbulaq (YANM) should be identified as most similar to ‘Mongoloids’ from East Asia. Individuals from Yanbulaq should exhibit no affinities to individuals from western Central Asia, the Iranian Plateau or the Indus Valley since Han’s model does not call for people inhabiting these regions to have played any role in the formation of the population at Yanbulaq.

Although Han (1990) classified all of the nearly complete female crania recovered from Yanbulaq as ‘Mongoloid’, neighbor-joining cluster analysis (Fig. 3a) reveals these females share no affinities to other females from the Tarim Basin or to modern East Asian females. Yanbulaq females also exhibit no affinities to ‘Europoid’ females of the Russo-Kazakh steppe or to females associated with ‘steppe-influenced’ artifacts. Instead, females from Yanbulaq share nearly equal affinities to females of western Central Asia, the greater Indus Valley, and the Iranian Plateau. This pattern is confirmed by principal coordinates analysis (Fig. 3b).

Two Biologically Distinct Populations at Yanbulaq? Each individual cranium was coded by sex, specimen number, and ‘race’ identification made by Han (1990). The dendrogram obtained from hierarchical cluster analysis of size-standardized individual crania recovered from Yanbulaq is provided in Figure 1. The results show no systemic separation between those crania identified by Han (1990) as ‘Europoid’ and those identified as ‘Mongoloid’. Instead, the seven ‘Europoid’ crania are distributed across five of the six clusters. Likewise there is no separation between male and female crania. Such

Discussion Five significant findings may be identified from these results. First, hierarchical cluster analysis fails to identify the presence of two biologically very different populations at Yanbulaq. Second, there is no evidence that significant gene flow from Russo-Kazakh steppe 118

A VIEW TO THE NORTH: BIOLOGICAL INTERACTIONS ACROSS THE INTER-MONTANE BORDERLANDS DURING THE LAST TWO MILLENNIA BC weakens the association between tomb type and racial type and makes a distinction based on relative status and/or wealth far more convincing.

populations led to the establishment of the population at Yanbulaq, or any of the Tarim Basin populations tested here. Third, there is also no evidence that East Asian populations, including Tibetans, played any significant role in the establishment of the population at Yanbulaq. Fourth, both neighbor-joining cluster analysis and principal coordinates analysis indicate that Yanbulaq males share closest affinities to males from the Indus Valley. Fifth, in contrast to males, females from Yanbulaq appear to have shared equally close affinities to females from western Central Asia, the Iranian Plateau and the greater Indus Valley.

Turning to a more fundamental issue, modern biology is predicated on the population concept, which emphasizes that variation among individuals of the same species exists within any population (Futuyma, 1997). In the ‘bad old days’ of physical anthropology, some workers (cf. Hooton, 1938) attempted to identify racial types that possessed certain ‘hallmark’ traits. Such efforts failed miserably. Unfortunately, Han’s (1990, 1998) identification of ‘Europoids’ and ‘Mongoloids’ at Yanbulaq is grounded in this same typological thinking that has proven to be of so little utility elsewhere.

Why is there no evidence of two ‘races’ at Yanbulaq? The dual race assertion was advanced by Han with little ancillary support. The strongest support comes from inhumation context. Three inhumation patterns were identified at Yanbulaq and designated as types I, II and III. Mallory and Mair (2000: 141) note that Han found type I tombs were exclusively Mongoloid while type II and III tombs show the presence of Caucasoids. The chronology of the tomb types is unclear, but one type II tomb has been identified as stratigraphically later than a type I tomb. Mallory and Mair (2000:141-2) seize upon this difference to claim, “if the distinctions are chronological … then we may be witnessing the movement of Caucasoid populations into a territory in which Mongoloid populations had already established themselves from the east”. However, Chen and Hiebert (1995: 262) observe that, in addition to the more than 100 bronze artifacts associated with these tombs, a few iron objects were also recovered and all derived from type I tombs. Using the same logic as Mallory and Mair, this finding could be interpreted as evidence for an earlier ‘Caucasoid’ presence at Yanbulaq followed by a subsequent ‘Mongoloid’ presence.

Why is there no evidence for gene flow, let alone outright colonization, from the Russo-Kazakh steppe? The battery of artefactual, linguistic and iconographic evidence marshaled in support of steppe influences is impressive (Mallory and Mair, 2000), but movement of artifacts and language distributions do not necessarily equate to an actual movement of people. As noted by Lamberg-Karlovsky (2002: 63), “ethnicity and language are not easily linked with an archaeological signature”. The same may also be said for genetic affinity; that is, ethnicity and language are also not easily linked to a genetic signature.

As provocative as such findings seem the association between tomb type, ‘race’, and chronology is ambiguous, for two important points are unresolved. First, how many burials in type I tombs were recovered but not subjected to Han’s craniometric analysis because of incompleteness and hence not assigned to a ‘racial type’? Second, although it is clear that individuals identified by Han as ‘Caucasoid’ were found in type II and III tombs, did these tombs also include individuals might have been identified as ‘Mongoloids’ had they been more completely preserved?

An array of recent genetic studies among living Central and South Asians reveals just how tenuous the relationship is between ethnicity, language and genetic affinity. A comparison of genetic distances obtained from Y-chromosome variations and geographic distances by Zerjal and coworkers (2002: 474) found cases where there was a great deal of heterogeneity among populations who share either geographic proximity or linguistic similarity as well as cases of genetic similarity among populations who either speak very different languages or who occupy widely separated regions. Similar results were found south of the Hindu Kush by Qamar and coworkers (2002: 1112). These researchers found that, apart from the Hazara who were identified as quite distinct, all other groups exhibited a general similarity to one another, despite the inclusion of a population that speaks a linguistic isolate (the Burushos), one that speaks a Dravidian language (the Brahui) and one that speaks a Sino-Tibetan language (the Balti).

Chen and Hiebert (1995:262) assert that the while the typology of the tombs appears clear, the artifact assemblages associated with each are similar. From this they suggest variation in tomb type reflects factors other than chronology and they offer social or economic status as possible influences. Discarding a chronological distinction between tomb types, Mallory and Mair (2000: 142) postulate an alternative scenario with “Mongoloids burying their dead with many grave goods in larger chambers while Caucasoids adopted single-grave burial in less elaborate tombs”. Yet, not knowing the degree of exclusivity of Han’s racial types to tomb types seriously

Iconographic and linguistic evidence for the presence of Tocharian-speakers within the Tarim Basin postdates the initial peopling of the Tarim Basin by many centuries (Mallory and Mair, 2000: 122). Hence, absence of evidence for gene flow or even outright colonization of the Tarim Basin from the Russo-Kazakh steppe may simply be a consequence of a settlement pattern in which the most valued localities were already occupied by sedentary agriculturalists, who practiced a farming and herding economy similar to that found in western Central Asia, when steppe pastoralist populations first appeared in this region of eastern Central Asia (Barber, 1999: 150; 119

Brian E. Hemphill (Mei-cun, 1992: 93). Drawing on historical documents, Barber (1999: 19) calls for an even later date (120 BC) for initial substantial movement of Han Chinese, citing efforts to open up regular trade with populations located even further west.

Chen and Hiebert, 1995: 257). Indeed, it is intriguing that prized settlement locations mirror those preferred in Turkmenistan and Uzbekistan (Chen and Hiebert, 1995: 262). Mair (1995: 299) posits that the inclusion of small, finely-woven baskets, which often contain grains of wheat not only serves to link the earliest burials to later burials but also reveals that agriculture was an important aspect of their lives. Further, he asserts the early inhabitants of the Tarim Basin enjoyed a peaceful and egalitarian lifeway, for their graves show little evidence of weaponry and none for social stratification (Mair, 1995: 301-302). These are dramatic differences from graves from the Russo-Kazakh steppe and in greater accordance with graves associated with the BactriaMargiana Archaeological Complex (BMAC) to the west (Askarov, 1973; Askarov and Abdullaev, 1983) or those of the greater Indus Valley to the south (Allchin and Allchin, 1982; Dani, 1967).

A second potential reason this research failed to identify the presence of ‘Mongoloids’ at Yanbulaq is because Khams Tibetans, identified by Han as the purported source for the ‘Mongoloids’ at Yanbulaq, may be profoundly different cranially from the other Mongoloid samples included in this analysis. Many years ago, Morant (1924: 2) provided an analysis of Khams Tibetan crania curated at the British Museum. Morant found these crania to be distinct from all neighboring Asiatic peoples, including western and central Tibetans. Nevertheless, he also found evidence for admixture with southern Chinese and with other Tibetans, such that these two populations exhibited closest affinities to Khams Tibetans (Morant, 1924: 5). However, in a surprising comment Morant (1924: 40) noted, “it is probable that the Khams Tibetans have intermingled to a certain degree with the Chinese and Tibetans of the A type [western and central Tibetans], but they have preserved predominantly the characteristics of an entirely different stock which resembles the Moriori [aboriginal inhabitants of the Chatham Islands east of New Zealand believed to be related to the Maori] as much as any of the Asiatic races”. Though such affinities appear initially improbable, it may be they represent very ancient connections that date to the initial dispersal of modern Homo sapiens throughout Central and Southeast Asia (cf. Passarino et al., 1996). Such a possibility requires further examination.

Why is there no evidence for gene flow or colonization of Mongoloid populations into the Tarim Basin? Han (1998: 561) claimed that two races were present at Yanbulaq and that those identified as ‘Mongoloid’ shared especially close affinities to Khams Tibetans. Yet, regardless of statistical approach used or which sex forms the basis of comparison, none of the inhabitants of Yanbulaq exhibit any affinity to Han Chinese, Nepalese, or Tibetans. There are at least two possible reasons to account for such results. First, it may be that substantial immigration of East Asians into the Tarim Basin simply occurred after the first half of the first millennium AD, the period represented by the latest of the samples included in this analysis. Although Han (1998:568) admits the timing of initial appearance of ‘Mongoloids’ in the Tarim Basin is unclear, he nevertheless claims that they appeared around 1350 BC, or a bit earlier, and mainly in the eastern portion of the basin, such as at Yanbulaq.

How can it be that males from Yanbulaq exhibit affinities to males from the Indus Valley? It should be remembered that the Great Silk Road wasn’t a single track running from East to West, but a comprehensive network with branches to Siberia in the north and to the Indian subcontinent in the south (Kuzmina, 1998: 63-64; Mallory and Mair, 2000: 51-52). Indeed, as pointed out by Chen and Hiebert (1995: 248249) not only is the Tarim Basin linked by wide passes to eastern Kazakhstan to the northwest and to the Altai region of southern Siberia to the north, but wide passes also link the Tarim Basin to the Ferghana Valley to the west and to the Indus Valley of Pakistan via the Karakoram Pass to the south. Since this network of interregional trade and communication existed during the first half of the first millennium AD, it may be that these connections were of even greater antiquity. Mair (1995: 296-299) claims the Bronze Age inhabitants of the Tarim Basin are likely ancestors to the Tocharian speakers, whose presence was first detected through decipherment of manuscripts dating to the sixth to eighth centuries AD. Tocharian monks are depicted in Buddhist cave sites near Kucha and Mair emphasizes that they are portrayed as possessing such non-East Asian physical features as fair hair, lightly pigmented eyes and narrow noses. A characterization of these monks is given by Mallory and Mair (2000: 248-249) in a subsection

Other researchers are less sanguine about an early appearance of East Asian populations. Mair (1995: 292294) envisions a three-part scheme for the peopling of the Tarim Basin during the Bronze Age. He posits that the third and final population movement involved ‘Mongoloid’ peoples from the East who entered the region even later than the second movement, an entry of ‘Caucasoid’ peoples from the west and southwest sometime after 500 BC. Mai-cun (1992) notes that while there is clear evidence for contacts between the Tarim Basin and points east during the Shang dynasty, this contact was exclusively from west-to-east. Mei-cun asserts that this is poignantly attested by small finds recovered from the tomb of Fu Hao at An-yang in 1976. Over 1600 small finds were recovered and of these more than 750 were jade carvings. All of these jade objects were made from Khotan jade from the area of Mount Kunlunshan. By contrast, archaeological investigations in the Tarim Basin consistently demonstrate that objects commonly associated with Han populations of China, such as silk fabrics and lacquer wares, do not appear until a millennium later, during the 7th to 3rd centuries BC 120

A VIEW TO THE NORTH: BIOLOGICAL INTERACTIONS ACROSS THE INTER-MONTANE BORDERLANDS DURING THE LAST TWO MILLENNIA BC Years ago, Stacul (1992) discovered a series of artifacts in northern Pakistan he interpreted as evidence of reciprocal contacts across the Karakoram and Hindu Kush Mountains between the inhabitants of western China, western Central Asia and northern Pakistan. Stacul assigned these finds to Swat Periods III and IV, which based on their stratigraphic position and associated radiocarbon dates, derive from the first half of the second millennium BC. Similarly, Barber (1999: 66) notes that Good determined the presence of cashmere in the dress worn by a mummy recovered from the Tarim Basin. The presence of cashmere indicates that local groups either kept or had access to goats from the south, from Kashmir in northern Pakistan and northwestern India. Together, these finds suggest contacts between the Indus Valley, Central Asia and western China may have occurred earlier than the first millennium BC and may have continued throughout the period encompassed by the Tarim Basin cranial series analyzed in the current study.

entitled ‘Knights with Long Swords’, and although they are careful to note the non-European aspects of these portrayals in both cases, clearly the implication of this iconographic evidence is clear-gene flow, if not outright colonization, from Russo-Kazakh steppe populations. In a similar vein, Barber (1999: 72) asserts that a substantial number of Uyghurs present today throughout Xinjiang also possess these features, which she attributes to “a legacy of old intermarriages with the ancient early arrivals from the west”. In an attempt to reinforce her claim, Barber (1999: 94) draws attention to Stein’s identification of the mummy found 15 miles from Loulan Station as being of the Homo alpinus type. However, it should be noted this alleged racial type was not associated with Russo-Kazakh steppe populations, but with the present population of the Tarim Basin and with the highland populations of the Pamirs to the west and the Hindu Kush to the southwest (Stein, 1933:155). A second aspect of Tocharian iconography also raises questions about possible connections between the Tarim Basin and populations located south of the Hindu Kush. Mair (1995: 299) notes, that in addition to possessing non-East Asian physical features, the Buddhist monks depicted in the cave sites near Kucha are dressed in silks and they bear an Indian religious mark (a tilaka) on their foreheads. It has always been assumed that the silks worn by such monks in post-Han times was of Chinese origin, but discovery of silk in pre-Han period contexts outside China raises the question of whether alternative sources of wild silk may have been developed and used before silkworms were domesticated in China (Good, 1995: 959). Good (nd) asserts there are several economically viable wild species of silkworm, some of which are not indigenous to China, but may be found in South Asia, the eastern Mediterranean, and even Mongolia. In Central Asia, pre-Han silk has been found at Niya, Lop Nor and Loulan in the Tarim Basin, at Ukok in the Altai, and at Pazyryk in southern Siberia (Good, 1995: 960). Hence, to unquestioningly assume that such silk is of Chinese origin is premature.

How can it be that Yanbulaq females have equidistant affinities to females from Central Asia, the Iranian Plateau, and the Indus Valley? Recent studies of mitochondrial DNA (mtDNA: Comas et al., 1998) and Y-chromosome (Pérez-Lezaun et al., 1999; Zerjal et al., 2002) variation reveal that living Central Asians are extremely heterogeneous and yield an interesting pattern. When affinities through the male line are assessed from Y-chromosome variations, Central Asian populations are marked by high degree of structure, in which nearly 24% of all variation occurs between populations. By contrast, when affinities are traced through the female line from mtDNA virtually all differences occur within populations, while there is little to no differences across populations (Yao et al., 2000). Comas and coworkers (1998) investigated differences in mtDNA among Kazakhs, Kirgiz and Uyghurs. They found linguistic affiliation did not contribute to genetic variation. A subsequent investigation of Y-chromosome short tandem repeats in these same individuals by PérezLezaun and coworkers (1999: 215) yielded evidence of differential mobility by sex across linguistic boundaries. Specifically, they found that 16% of Y-chromosome haplotype diversity could be attributed to language group differences, while differences in mtDNA proved negligible (-0.5%). As such, it appears that among modern Central Asians - at least among Kazakhs, Kirgiz and Uyghurs - female migration may overcome linguistic barriers with greater facility than male migration. Yao and coworkers (2000: 177) analyzed mtDNA diversity among populations of Xinjiang, Qinghai and Yunnan provinces. Like Comas and coworkers (1998), they also found few differences across samples. Based on the results of the study by Pérez-Lezaun and coworkers (1999), Yao and coworkers attributed the lack of differentiation to high rates of female migration.

Good (1995: 960-961) points out a crucial difference in the processing of Chinese and non-Chinese silk. Chinese silk from domesticated Bombyx mori cocoons is processed by degumming the silk by boiling the cocoon before the moth emerges. This permits the cocoon to be unraveled in one continuous filament and hence, no spinning is required. However, the silkworm dies. On the other hand, mukta silks were developed in India to avoid religious taboos against the destruction of life in order to produce ungummed silk. As Buddhists, it is highly unlikely the Tocharian monks wore robes of degummed silk obtained through a process that sacrificed the lives of the silkworms. Rather, it is more likely they wore robes made from mukta silks of possible South Asian origin. Discovery of pre-Han silk at Sapalli tepe (Askarov, 1973) and its recent discovery at Harappa (Good, nd) raises the possibility that at least some, if not all, of the pre-Han silk found in Central Asia may have originated in South Asia.

Thus, those ethnic groups of living Central Asians that have been genetically tested share a common pattern. Data collected within villages that encompass a 121

Brian E. Hemphill Futuyma, D. J. (1997). Evolutionary Biology. 3rd Edition, Sinauer Associates, Sunderland, MA. Good, I. (in press) ‘When East Met West: Interpretive Problems in Assessing East-West Contact and Exchange in Antiquity’. In Betts, A. and Kidd, F. (eds) Vth ICAANE Congress Proceedings, Ancient Near Eastern Monograph Series, Peeters, Louvain. Good, I. (1995) ‘On the Question of Silk in pre-Han Eurasia’. Antiquity, 69: 959-968. Hair, J. F., Anderson R. E., Tatham R. L. and Grablowsky B. J. (1971) Multivariate Data Analysis, PPC Books, Tulsa. Han, K. (1990) ‘The Study of Racial Elements of Bones from the Yanbulaq Site of Qumul, Xinjiang’. Acta Anthropologica Sinica, 5: 371-390. Han, K. (1998) ‘The Physical Anthropology of the Ancient Populations of the Tarim Basin and Surrounding Areas’. In Mair V.H. (ed) The Bronze Age and Early Iron Age Peoples of Eastern Central Asia. Volume 2, University of Pennsylvania Museum Publications, Philadelphia: 558-570. Hemphill, B. E. and Mallory J. P. (2004) ‘HorseMounted Invaders from the Russo-Kazakh Steppe or Agricultural Colonists from Western Central Asia? A Craniometric Investigation of the Bronze Age Settlement of Xinjiang’. American Journal of Physical Anthropology, 124 (3): 199-222. Hooton, E. A. (1938) Up from the Ape, Macmillan, New York. Jungers, W. L., Falsetti A. B. and Wall C. E. (1995) ‘Shape, Relative Size, and Size-adjustments in Morphometrics’. Yearbook of Physical Anthropology, 21: 137-161. Komroff, M. 1930. Introduction, pp. v-xxxii in: The Travels of Marco Polo, Komroff M, ed., Boni and Liveright, New York. Kuzmina, E. E. (1998) ‘Cultural Connections of the Tarim Basin People and Pastoralists of the Asian Steppes in the Bronze Age’. In Mair, V.H. (ed) The Bronze Age and Early Iron Age Peoples of Eastern Central Asia, Volume 1, University of Pennsylvania Museum Publications, Philadelphia: 63-93. Lamberg-Karlovsky, C. C. (2002) ‘Archaeology and Language: The Indo-Aryans’. Current Anthropology, 43 (1): 63-88. Mair, V. H. (1995) ‘Prehistoric Caucasoid Corpses of the Tarim Basin’. Journal of Indo-European Studies, 23(3-4): 281-307. Mallory, J. P. and Mair V. H. (2000) The Tarim Mummies, Thames and Hudson, London. Martin, R. (1928) Lehrbuch der Anthropologie. Zweiter Band: Kraniologie, Osteologie, Gustav Fischer Verlag, Jena.

substantial number of individuals from a single ethnic group exhibit considerable variation in mtDNA, but intervillage comparisons show little variation. The opposite pattern is found for the Y chromosome. That is, among males there is little intravillage variation in the Ychromosome, but profound intervillage differences. Since mtDNA is inherited exclusively through one’s mother, while the Y chromosome is inherited from one’s father, such disparate patterning is the logical outcome of a marital system in which women move out of their father’s village upon marriage to the village of their new husband, while males remain in the villages of their fathers, their paternal uncles and grandfathers (Seielstad et al., 1998). It may be that the social system found today in Central Asia (Pérez-Lezaun et al., 1999: 215), in which females serve as a sort of social mortar connecting together isolated groups of related males, was also present some four millennia in the past. If so, this would explain why females from Central Asia in general, and females from Yanbulaq in particular, exhibit equidistant affinities to females from an array of adjacent regions while their male counterparts do not.

Bibliographical References Alekshin V. A. (1986) ‘The Origin of Neolithic-Bronze Archeological Cultures of Central Asia (Based on Burial Customs Evidence)’. Information Bulletin of the International Association for the Study of the Cultures of Central Asia 11: 90-97. Allchin, B. and Allchin R. (1982) The Rise of Civilization in Indian and Pakistan, Cambridge University Press, Cambridge. An Zhimin (1992) ‘Bronze Age Communities in Eastern Parts of Central Asia’. In Dani, A. H. and Masson, V. M (eds) History of Civilizations of Central Asia, volume I, UNESCO, Paris: 319-336. Askarov, A. A. (1973) Sapallitepa, FEN, Tashkent. Askarov, A. A. and Abdullaev, B. N. (1983) Dzharkutan, Fen, Tashkent. Barber, E. W. (1999) The Mummies of Ürümchi, W.W. Norton, New York. Chen, K. and Hiebert, F. T. (1995). ‘The Late Prehistory of Xinjiang in Relation to its Neighbors’. Journal of World Prehistory, 9 (2): 243-300. Comas, D., Clafell F., Mateu E. (and 9 others) (1998) ‘Trading Genes along the Silk Road: mtDNA Sequences and the Origin of Central Asian populations’. American Journal Human Genetics, 63: 1824-1838. Dani, A. H. (1967) ‘Timargarha and the Gandharan Grave Culture’. Ancient Pakistan, 3: 1-407. Di Cosmo, N. (1996) ‘Ancient Xinjiang between Central Asia and China: The Nomadic Factor’. Anthropology and Archaeology of Eurasia, 34 (4): 87-101.

122

A VIEW TO THE NORTH: BIOLOGICAL INTERACTIONS ACROSS THE INTER-MONTANE BORDERLANDS DURING THE LAST TWO MILLENNIA BC Masson. V. M. (1992) ‘The Decline of the Bronze Age Civilization and Movements of the Tribes’. In Dani, A. H. and Masson, V. M. (eds) History of Civilizations of Central Asia, Volume I, UNESCO, Paris: 337-356. Mei-cun, L. (1992) ‘Tocharian People: Silk Road Pioneers’. In Tadao, U. and Toh, S. (eds) Significance of Silk Roads in the History of Human Civilizations, Senri Ethnological Studies No. 32, National Museum of Ethnology, Osaka: 91-96. Passarino, G., Semino O., Bernini L. F. and SantachiaraBenerecetti A. S. (1996) ‘Pre-Caucasoid and Caucasoid Genetic Features of the Indian Population, Revealed by mtDNA Polymorphisms’. American Journal of Human Genetics, 59: 927-934. Pérez-Lezaun A., Calafell F., Comas D. (and 9 others). (1999) ‘Sex-specific Migration Patterns in Central Asian Populations, Revealed by Analysis of Ychromosome Short Tandem Repeats and mtDNA’. American Journal of Human Genetics, 65: 208-219. Qamar, R., Ayub Q., Mohyuddin A. (and 6 others) (2002) ‘Y-chromosomal DNA Variation in Pakistan’. American Journal of Human Genetics, 70: 11071124. Saitou, N. and Nei M. (1987) ‘The Neighbor-Joining Method: A New Method for Reconstructing Phylogenetic Trees’. Molecular Biology and Evolution, 4: 406-425. Seielstad, M., Minch E. and Cavalli-Sforza L. L. (1998) ‘Genetic Evidence for a Higher Female Migration Rate in Humans’. Nature Genetics, 20: 278-280. Sneath, P. H. A. and Sokal R. R. (1973) Numerical Taxonomy. W.H. Freeman, San Francisco. Stacul, G. (1992) ‘Further Evidence for the ‘Inner Asia Complex’ from Swat’. In Possehl, G. (ed) South Asian Archaeology Studies, Science Publishers, New York: 111-122. Stein, M. A. (1933) On Central Asian Tracks. Macmillan, London. Ward, J. H. (1963) ‘Hierarchical Grouping to Optimize an Objective Function’. Journal of the American Statistical Association, 58: 236-244. Yao, Y.-G., Lü X.-M., Luo H.-R., Li W.-H. and Zhang Y.-P. (2000) ‘Gene Admixture in the Silk Road Region of China: Evidence from mtDNA and Melanocortin 1 Receptor Polymorphism’. Genes and Genetic Systems, 75: 173-178. Zerjal, T., Wells R. S., Yuldasheva N. and Ruzibakiev R. (2002) ‘A Genetic Landscape Reshaped by Recent Events: Y-chromosomal Insights into Central Asia’. American Journal of Human Genetics, 71: 466-482

123

Brian E. Hemphill

Fig. 1 – Hierarchical cluster analysis of individual crania recovered from Yanbulaq with Ward’s method (1963). Branch points are Euclidean distances. Individual crania are coded by sex, specimen number, and “race” identification according to Han (1990). Example: F21M= Female, Specimen Number 21, identified by Han (1990) as “Mongoloid.”

124

A VIEW TO THE NORTH: BIOLOGICAL INTERACTIONS ACROSS THE INTER-MONTANE BORDERLANDS DURING THE LAST TWO MILLENNIA BC

a)

b)

Fig. 2 – Patterning of intersample affinities and differences among males obtained from the diagonal matrix of pairwise Mahalanobis generalized distances. (a) neighbor-joining cluster analysis, (b) principal coordinates analysis. Sample abbreviations: Russo-Kazakh steppe and “steppe affiliated”: ADM= Andronovo/Minusinsk, AFA= Afanasievo/ Altai, AFM= Afansievo/Minusinsk, AND= Andonovo/Kazakhstan, KAM= Karasuk/Minusinsk, KOK= Kokcha III, SAMB= Samtavro B, TMM= Tigrovaja and Makoni Mor; Eastern Central Asia and East Asia: ALW= Alwighul, AYG= An-yang, HAI= Hainan, KRO= Krorän, NEP= Nepalese, QAW= Qäwrighul, TIB= Tibetans, YANE= Yanbulaaq “Europoids”, YANM= Yanbulaq “Mongoloids; Western Central Asia: ALT= Altyn depe, DJR= Djarkutan, Djarkutan period, GKS= Geoksyur, KAR= Kara depe, KUZ= Djarkutan, Kuzali period, MOL= Djarkutan, Molali period, SAP= Sapalli tepe; Iranian Plateau: SHS= Shahr-I Sokhta, TH2= Tepe Hissar, period II, TH3= Tepe Hissar, period III; Greater Indus Valley: CEMH= Harappa, Cemetery H, HAR= Harappa, Cemetery R37; TMG= Timagarha.

125

Brian E. Hemphill

a)

b)

Fig. 3 – Patterning of intersample affinities and differences among females obtained from the diagonal matrix of pairwise Mahalanobis generalized distances. (a) neighbor-joining cluster analysis, (b) principal coordinates analysis. Sample abbreviations same as in Figure 2, except YAN= Yanbulaq (“Mongoloids” only).

126

ARE THE KHO AN INDIGENOUS POPULATION OF THE HINDU KUSH ? A DENTAL MORPHOMETRIC APPROACH Brian E. Hemphill , Ihsan Ali, Sarah Blaylock & Nikki Willits

Archaeological Complex (BMAC) urban centres, crossed the Hindu Kush, established the Gandharan Grave Culture, and brought Indo-Aryan languages to the Indian subcontinent.

When attempting to understand the lifeways of past peoples researchers are immediately confronted with issue of identity; yet “identity” is a complex concept. Archaeologists (i.e., Binford 1971) have sought to approach the concept of identity by operationalising Goodenough’s (1965) concept of social roles. Goodenough, drawing from the seminal work of Linton (1936), notes that individuals can occupy a number of roles. Some are selected strategically by the individual, but others are assigned. Often these identities reflect inequities in individual status that may signal differences along lines of social class, gender, and ethnicity. It is in regard to the latter, especially in situations accompanied by abrupt change in material culture, where the question “Who were these people?” becomes especially important.

Distinguishing Cultural from Biological Transmission Despite such machinations, the fundamental dilemma remains - how to differentiate culture change through the diffusion of ideas from that due to the actual movement of people. The best means for such differentiation are those that can detect biological changes in the alleged recipient population. For if abrupt cultural changes are the consequence of the appearance of outsiders, either through the process of outright colonization as proposed by Hiebert (1994) or through significant gene flow as proposed by Parpola (1996), the biological input of these outsiders should be detectable.

Wheeler (1968: 132-3) attributed what he perceived as an abrupt coup de grâce to an already degenerate and enfeebled Harappan Civilization to the entrance of invaders bearing a foreign culture into the Indus Valley. In similar fashion, Hiebert and Lamberg-Karlovsky (1992; Hiebert 1998) interpreted the sudden appearance of Central Asian artefacts at sites on the Iranian Plateau as signalling the presence of Central Asians occupying distant outposts for purposes of conquest or trade.

The development of DNA amplification through PCR offers an opportunity to examine the very genetic composition of past populations. Such sophisticated techniques allow the biological anthropologist to compare the genetic profile of alleged donor and recipient populations, and determine whether archaeological evidence of significant shifts in material culture is accompanied by concurrent biological evidence of profound genetic change. Through the use of mtDNA and Y-chromosome variations, DNA studies can elucidate the differential roles played by men and women in such events (Stoneking 1998). Indeed, such analyses among living Central Asians reveal a consistent pattern: within a village there is considerable diversity in mtDNA, but few Y-chromosome variations. By contrast, when compared across villages, there is little variation in mtDNA, but profound heterogeneity in Y-chromosomes (Comas et al 1998; Yao et al 2000). Such patterning reflects a marital system where males remain in their natal villages, while females move from their birth village to the village of their new husbands upon marriage (Pérez-Lezaun et al 1999: 214; Stoneking 1998: 220; Yao et al 2000). Hemphill’s (in this volume) analysis of phenetic affinities among Bronze Age skeletal samples from Central Asia indicates this type of social system may be of considerable antiquity.

The great difficulty faced by archaeologists who attempt to attribute changes in material culture to population movements is separating alterations due to the diffusion of ideas from those stimulated by the actual movement of people. A number of solutions have been proposed. Hiebert (1994: 67-68) argues for a direct colonization of the oases of Bactria and Margiana by populations from the Kopet Dagh foothill plain based on the common occurrence of an entire “complex” of material artefacts. He reasons, not illogically, that the greater the number of artefact categories found in common in two widely separated areas, the less likely such co-occurrence is due to mere chance. This is no doubt true, but to conclude such co-occurrence signals the movement of a population, or even of individuals, is premature as Hemphill’s (1999) craniometric analysis of “Bactrian” burials at the Iranian site of Tepe Hissar clearly demonstrated. Even more tenuous are attempts to appeal to some scheme of “elite dominance” to account for dramatic shifts in material culture. A well-known example is Renfrew’s (1996: 76-81) invocation of elite dominance to account for the eastward spread of agriculture and Indo-European languages from an Ürheimat in Anatolia. In another formulation, Parpola (1995: 8-10) attempted to account for the introduction of proto-Indo-Aryans into the Indus Valley. He claimed proto-Indo-Aryans of the Biskent culture subjugated the much larger population of the Bactrian Margianan

Although DNA analysis offers great promise, several problems constrain its utility (especially in South Asia), for addressing questions of prehistoric population interactions and movements. Only two will be mentioned here. First, DNA yields imprecise coalescence estimates (Quintana-Merci et al 2001: 540; Zerjal et al 2002). Consequently, attempts to determine when two living populations last shared a common ancestor - the situation 127

Brian E. Hemphill , Ihsan Ali, Sarah Blaylock & Nikki Willits in a case of direct colonization - from a comparison of DNA profiles yields estimates that often encompass several millennia. While this problem is avoided when DNA is extracted from the ancient skeletons recovered from archaeological sites, the second difficulty may then be encountered. Obviously, for DNA to be extracted, organic material (collagen) must be present. Unfortunately, every attempt to amplify DNA from ancient bones in South Asia has failed because the depositional environment resulted in the complete destruction of all organic collagen.

The Kho: A Case Study in Population Origins Here we present a preliminary case study in which several alternative hypotheses for the origins of a specific population, the Kho of Chitral District, NWFP, Pakistan are tested. This represents an initial effort in a larger, more ambitious project undertaken by Ali and Hemphill to conduct the first comprehensive dental survey of the living peoples of northern Pakistan. It is our hope that our archaeological colleagues will acknowledge the relevance of this research for addressing critical, unresolved issues in South Asian prehistory.

If DNA analysis is either too imprecise when extracted from living populations, or simply possible in the case of ancient skeletons from South Asia - it begs the question – is it impossible to use biological data to distinguish changes in material culture due to the diffusion of ideas from those due to the physical movement of people? There is a solution and this is provided by dental analysis.

The Kho represent the numerically dominant ethnic group of Chitral District (Young et al 2000). Most Kho are agriculturalists who cultivate small fields of wheat and rice, as well as a few scattered orchards of apples and pomegranates. Today the Kho are a devoutly Muslim group who speak Khowari, a language attributed to the Dardic branch of Indo-European (but see Morgensteirne 1961).

The dental elements of numerous animals, including humans, have been the subject of intensive study for centuries. Long before Darwin, natural scientists recognized that various animals - both living and palaeontological - could be classified by their dental remains. Under Darwin’s theory of evolution, close parallels between dental structure and relatedness made good sense given the central concept of descent with modification from a common ancestor and, with the rediscovery of genetics, it was found that both tooth size and morphology are conservative anatomical structures under strong genetic control (Scott & Turner 1997; Townsend & Brown 1978, 1979).

Three Models Several different theories have been offered to account for the peopling of the Indian subcontinent and expectations for the biological affinities of the Kho may be inferred from each. Proponents of the first and historically most influential Aryan Invasion Model, such as Erdosy (1998), Kuzmina (1998), and Parpola (1995), note the known similarities among “eastern” IndoEuropean language groups - Dardic, Indo-Iranian and Indo-Aryan. They offer an array of corroborative archaeological evidence; assert that Indo-Aryan languages entered the Indian subcontinent during the 2 nd millennium BC by horse-mounted invaders from the steppes of Central Asia, and maintain the invader’s descendants subsequently spread Vedic culture, first to the upper Doab region of north India, then throughout the rest of the Indian peninsula. The alleged source of this migration is of some disagreement. Kuzmina prefers a direct connection to Andronovo steppe populations, while Erdosy and Parpola view the urban populations of Bactrian and Margiana - under the direction of Andronovo-associated Bishkent culture overlords - as the source.

Teeth offer two crucial advantages for deciphering patterns of interaction among past populations. First, because teeth are composed of the hardest structures of the body, they preserve under depositional environments that compromise other hard tissues. This has been demonstrated by Lukacs and coworkers (1985), whose studies of the dental remains from Mehrgarh yielded insights into the occupant’s health status and biological origins (Lukacs & Hemphill 1991) that could not be ascertained from the badly crushed skeletal remains. Second, teeth represent the only directly accessible hard tissues in living individuals. Because of this, the teeth of living people can be measured easily and their morphology assessed allowing researchers to collect large, highly accurate dental samples. This permits large datasets of known males and females to be analyzed and compared, both to other living populations and to populations of the past. Given the inability to extract DNA from ancient skeletons, as well as the general paucity of skeletal remains in South Asia, collection and analysis of dental casts from members of living ethnic groups offers the greatest potential for establishing a baseline of human biological variation from which culture change due to diffusion may be distinguished from that caused by population movement.

A second model, offered by Kennedy and coworkers (1984), may be described as the Long-Standing Continuity Model. Proponents of this model contend that the population history of South Asia is one in which there were no major post-Pleistocene population movements, either internally or into the subcontinent from outside. Consequently, patterning of biological affinities should be a simple reflection of isolation-by-distance (Sokal & Wartenburg 1983). That is, due to the constraints of travel, marital partners should be preferentially recruited from nearby. Hence, with the passage of time, populations closest both temporally and geographically should be most similar biologically.

128

ARE THE KHO AN INDIGENOUS POPULATION OF THE HINDU KUSH ? A DENTAL MORPHOMETRIC APPROACH Tooth measurements were size corrected by standardizing measurements against individual geometric means in samples of living populations and against sample means in prehistoric samples (Jungers et al 1995). Differences between samples were quantified with generalized Euclidean distances and patternings of inter-sample differences were simplified with cluster analysis using Ward’s minimum variance technique (Ward 1963) and with principal co-ordinates analysis (Gower 1966).

Proponents of the third model, which may be identified as the Early Intrusion Model suggest that populations of the Indian subcontinent have been affected by at least two significant population incursions from the west (Lukacs & Hemphill 1991; Hemphill et al 1992). The initial incursion may be associated with the introduction of Dravidian languages, appears to have been the more influential genetically, and likely occurred between 6000 and 4500 BC. Another wave, possibly associated with the introduction of Indo-European languages, appears to have had far less genetic impact on South Asian populations, and may have occurred no earlier than the latter half of the 1st millennium BC.

Cluster analysis (Fig. 1a) reveals that the most profound separation occurs between prehistoric Central Asians and all South Asians, both living and prehistoric, except for the Kho. Remaining samples fall into two clusters that may be identified as living inhabitants of the Indian peninsula versus prehistoric inhabitants of the Indus Valley. The only exceptions to this pattern are the postHarappan samples from Timargarha and Sarai Khola, which appear as distant outliers to living peninsular Indians, and the Late Jorwe inhabitants of Inamgaon who possess closer affinities to prehistoric inhabitants of the Indus Valley than to living inhabitants of peninsular India. Initial inspection suggests language family is ineffective for distinguishing among living inhabitants of the Indian peninsula, for Dravidian-speaking caste Hindus from Andhra Pradesh (GPD, PNT) exhibit closer affinities to Indo-Aryan speaking populations than to Dravidian-speaking Chenchu tribals. Yet, the peripheral association of Indo-Aryan speaking Bhil tribals to their adjacent Hindu caste counterparts, Garasias and Rajputs, suggests that a biological segregation of tribal and caste Hindu populations in peninsular India - regardless of linguistic affiliation - may be obscuring biological differences between linguistic groups.

These diverse scenarios have specific implications for the predicted biological origins of the Kho. If the Aryan Invasion Model is correct, the Kho - occupying the territory intermediate between source and recipient populations - ought to represent the direct lineal descendants of these Indo-Aryan invaders from Central Asia. As such, the Kho should be biologically identical to their source populations in Central Asia and very similar to their sister descendant populations found in the upper Doab of north India. If the Long-Standing Continuity Model is correct, the Kho are indigenes of the Hindu Kush Mountains. As such, they should exhibit closest biological affinities to populations most proximate to them both in time and space. If, however, the Early Intrusion Model is correct, the Kho as a living population found in the extreme northwestern extent of South Asia, should be the product of some degree of gene flow between Central Asian immigrants and local populations that took place several millennia ago. Consequently, the Kho should exhibit distant and nearly equivalent affinities to Central Asians, to populations of the upper Doab of north India, and to the most recent of the archaeological samples from the Indus Valley. The current research attempts to ascertain the greatest parsimony between the pattern of phenetic affinities yielded by dental data and the three models identified above to account for South Asian population origins and subsequent interactions.

A clear pattern emerges from principal co-ordinates analysis (Fig. 1b). All prehistoric Central Asian samples occupy the left side of the array, prehistoric inhabitants of the Indus Valley occupy the centre, while living inhabitants of peninsular India are on the right. This plot indicates the Kho have little affinity to Central Asians. In fact, the Kho share only distant affinities to any of the other samples included in this analysis, for even a slight change in the path setting for the minimum spanning tree shifts this tenuous connection with Central Asians to the Chalcolithic inhabitants of Mehrgarh.

First Analysis: Odontometrics The first phase of this study is based on dental casts of 104 schoolchildren from Booni, a village located 80 kilometres north of the district capital, Chitral. Mesiodistal lengths and buccolingual breadths of all permanent tooth crowns, except third molars, were measured according to the standards of Moorrees (1957). Results were compared to 19 samples of prehistoric and living Central and South Asians, the latter of which encompasses both Indo-Aryan and Dravidian-speakers (Table 1). The comparative sample includes 1,734 individuals, the majority of which are living individuals of known sex and linguistic affinity. Together, these samples encompass individuals from Central Asia, the Indus Valley, western India and southeast India.

With more scrutiny, language provides a clearer separation among living inhabitants of peninsular India, for the three Dravidian-speaking samples exhibit closest affinities to one another, as do the four Indo-Aryan speaking samples, with one notable exception - the mixed urban caste sample from Pune, Maharashtra (RAS). Numerous linguistic indicators reveal that Dravidian languages were once spoken in west-central India, especially western Maharashtra, long after the initial appearance of Indo-Aryan languages (Southworth 1995: 271). The phenetic position of this mixed caste sample could reflect a base population with close affinities to Dravidian-speaking populations to the south that experienced little subsequent gene flow from Indo-Aryan speaking populations. The connection between this 129

Brian E. Hemphill , Ihsan Ali, Sarah Blaylock & Nikki Willits occupies a basal position on the branch extending to the lower left that encompasses the samples of living IndoAryan-speaking inhabitants of Maharashtra. Timargarha links to these samples via the Neolithic inhabitants of Mehrgarh proximally and the prehistoric inhabitants of Inamgaon distally. Harappa occupies a basal position on the branch toward the upper left. This branch includes the Dravidian-speaking tribal Chenchus proximally, followed by rather close affinities between their caste Hindu counterparts (PNT, GPD), the mixed urban caste sample from Pune (RAS) and the mixed urban caste sample from Calcutta (RAS).

sample and the Neolithic sample from Mehrgarh runs counter to such a scenario, but as with the Kho, a slight change in the minimum spanning tree setting shifts this tenuous connection to Inamgaon, the exact result expected if this reconstruction is valid. Second Analysis: Dental Morphology Phase two of this study is based on dental casts of 145 schoolchildren from Chitral. Morphology was scored as 17 tooth-trait combinations and trait frequencies among the Kho were compared to 19 samples of prehistoric and living South Asians. Morphological traits were scored in accordance with the Arizona State University system (Scott & Turner 1997). The comparative sample includes 1,594 individuals, the majority composed of living individuals of known sex and linguistic affinity. This sample includes prehistoric individuals from Central Asia and the Indus Valley, as well as living individuals from southeast India. Importantly, this sample also encompasses living west-central Indians and Bengalis, but lacks Gujaratis (see Table 1).

A broadly similar patterning of intersample associations is identified by multidimensional scaling (Fig. 2b). Once again, prehistoric Indus Valley samples dominate the centre of this array; with one exception, the Chalcolithic inhabitants of Mehrgarh. This sample serves to link the living Kho to prehistoric Indus Valley samples, albeit at a distant remove. Mirroring results obtained from principal coordinates analysis of odontometric data, a slight shift in the path setting for the minimum spanning tree shifts the link between the Kho and the Chalcolithic inhabitants of Mehrgarh to prehistoric Central Asians via the Molali period sample (MOL).

Trait frequencies were compared with Smith’s mean measure of divergence statistic corrected with Freeman and Tukey’s angular adjustment (Green & Suchey 1976). This distance measure normalizes trait frequency distributions and is especially useful when dealing with low and high frequency traits, as well as cases with missing observations and/or the small sample sizes often encountered when incorporating archaeologically derived dental samples (Sjøvold 1977). Patterns of inter-sample differences can be simplified in various ways, including the cluster and principal co-ordinate analyses used previously for odontometric data. Two alternative techniques were used for the dental morphology data in an effort to demonstrate the robusticity of inter-sample relationships and to confirm that such relationships are not solely a product of the simplifying technique employed. These techniques include neighbour-joining cluster analysis (Saitou & Nei 1987) and nonmetric multidimensional scaling (Kruskal 1964).

Remaining samples are arranged in the same macroregional aggregates described in other analyses. That is, living Indo-Aryan-speaking samples from westcentral India (MDA, MRT, MHR) exhibit closest affinities to one another and link to other samples via the prehistoric samples from Inamgaon, the Neolithic inhabitants of Mehrgarh, and the post-Harappan sample from Timargarha. The only exception is the mixed urban caste sample from Pune (RAS). As previously identified, this sample exhibits closer affinities to Dravidianspeaking caste Hindus from Andhra Pradesh (GPD, PNT) and to Indo-Aryan-speaking inhabitants of Calcutta (BNG) than to the other samples from Maharashtra. As indicated by neighbour-joining cluster analysis, prehistoric Central Asians link to the other samples included in the current analysis (except the Kho) via the latest of the Indus Valley samples, Sarai Khola.

The dendrogramme obtained from neighbour-joining cluster analysis (Fig. 2a) corroborates the macroregional segregation of samples identified from odontometric data, but the separation is not clear-cut. Once again, prehistoric inhabitants of the Indus Valley generally occupy the centre of the array. This is especially the case for the samples from Sarai Khola, Timargarha, and Harappa, which occupy basal positions on each of the three primary branches that define this dendrogramme, respectively.

Discussion Abrupt changes in the archaeological record have prompted researchers to offer population movement and even outright population replacement to account for such changes. When evidence for actual physical movement of populations appears lacking, some - such as Renfrew (1987) - have turned to “elite dominance” to effect such change. When faced with a lack of evidence for physical population movement or even evidence of a change in material culture, some have sought nevertheless to invoke outside intrusion. Parpola’s model (1995; see also Erdosy 1995) for the introduction of Central Asians into the Indian subcontinent is a particularly tenuous case in point. With no evidence for a replacement of BMAC material culture within the urban centres of Bactria and Margiana with those of Andronovo steppe nomads, Parpola and Erdosy nevertheless identify these urban

Sarai Khola occupies a basal position on the branch extending to the right that encompasses prehistoric Central Asians, the living Kho and the Chalcolithic inhabitants of Mehrgarh. Corroborating the principal coordinate analysis of the odontometric data, this array depicts the Kho as possessing only distant affinities to Central Asians and slightly closer affinities to the Chalcolithic inhabitants of Mehrgarh. Timargarha 130

ARE THE KHO AN INDIGENOUS POPULATION OF THE HINDU KUSH ? A DENTAL MORPHOMETRIC APPROACH resolutions. Either the Kho represent a highly isolated and peripheral population of the rugged Hindu Kush highlands that has experienced very little gene flow with other populations, or they are the product of a more recent population movement (post-1st millennium BC) into the most northwestern periphery of South Asia from populations located outside southern Central Asia. Indeed, the Kho themselves claim recent ancestry from the Kafirs of Afghanistan (Jan 1996) and such a possibility cannot be ruled out by this analysis.

populations as the agents of change into South Asia. Semifacetiously dubbed the “Kulturkugel” by Mallory (1998: 192-194), their model calls for domination of the urban BMAC population by an Andronovo-affiliated elite, who in turn remarkably adopt all the material trappings of their subjects. Retaining their martial traditions and their religious deities, these urban-charged elites allegedly crossed the Hindu Kush to conquer the cities of the Indus Valley and beyond. Clearly, any model invoking “elite dominance” is inherently weak. In the former case, those seeking to invoke elite dominance assert that shifts in material culture are the consequence of foreign intrusion even when no evidence for intrusion exists in the form of population movement. In the latter, those invoking a Kulturkugel go one step further, arguing for an implausible foreign incursion despite any observable indication of population movement or introduction of material culture. By advancing models that rely on negative evidence, the burden of proof for acceptance of such extraordinary claims must be higher (Sagen 1995). Hence, models that invoke population movement must be substantiated by biological evidence and those that appeal to an ethnic-based shift in political power must demonstrate biological evidence commensurate with such a usurpation of authority by an intrusive elite. Since archaeologically-derived skeletal collections currently known in South Asia are far too small and incomplete to investigate potential inter-ethnic differences within an archaeological site, biological data from living South Asians is crucial for establishing a baseline of biological variation against which important issues in South Asian archaeology may be addressed.

As such, the biological origins of the Kho remain enigmatic, for none of the three proffered models definitively identify the Kho’s likely progenitors. Further fieldwork, dental cast collection and data analysis are required to resolve the intriguing, but as yet unresolved, role played by this ethnic group located along the borderlands between Central and South Asia. While dental analysis has been unsuccessful so far in identifying the likely ancestors of a specific living ethnic group located along the northwestern periphery of South Asia, this does not imply that analyses incorporating dental observations among members of both past and present populations do not provide valuable insight for evaluating models offered by archaeologists and others for reconstructing the population history of South Asia. These analyses reveal that neither odontometry nor dental morphology offers any support for the Aryan Invasion Model. Though the latest of the prehistoric Indus Valley samples, Sarai Khola, consistently exhibits closer affinities to prehistoric Central Asians than their predecessors, this affinity occurs too late in time and is unaccompanied by similar connections to subsequent living populations. The absence of affinities between prehistoric Central Asians and living inhabitants of Gujarat and Maharashtra indicates that the establishment of Vedic culture and Brahminical Hinduism did not involve large-scale population intrusion from Central Asia.

Three models have been offered to account for South Asian population origins. Each has different expectations for the pattern of biological affinities to be found among the ethnic groups of South Asia, both prehistoric and living. The Kho, an agricultural ethnic group of Chitral District, Pakistan served as a test case to evaluate the congruence between the expectations of these models and the observed patterns of biological affinities found among South Asians. Permanent tooth measurements and morphological trait frequencies formed the basis for comparison.

Patterns of phenetic similarities likewise fail to support the Long-Standing Continuity Model. Inter-sample affinities are not a mere consequence of isolation by geographic and temporal distance. This disparity is especially marked between the Neolithic and Chalcolithic inhabitants of Mehrgarh, as well as by the phenetic positions of the mixed urban caste samples from Pune and Calcutta.

Results of the current study indicate that, rather than showing close affinities to prehistoric Central Asians and contemporary north Indians as called for by the Aryan Invasion Model, the Kho exhibit only distant affinities to these groups. Likewise, instead of exhibiting moderate and equidistant affinities to prehistoric Central Asians, later period prehistoric Indus Valley dwellers and living north Indians, as expected under both Long-Standing Continuity and Early Intrusion Models, the Kho show little or no affinities to these groups. Instead, the marked by prehistoric inhabitants

The patterning of phenetic affinities among prehistoric and living South Asians offers some support for the Early Intrusion Model, but in a fashion somewhat different from its original formulation. As expected, dental morphology analysis indicates a profound separation between the Neolithic and Chalcolithic inhabitants of Mehrgarh. However, there is no support for the claim of a Dravidian intrusion into the Indus Valley prior to the mid-5th millennium BC. No analyses demonstrate affinities between the Chalcolithic inhabitants of Mehrgarh and Dravidian-speaking samples from southeast India. However, dental morphology analysis yields a pattern that ties living Dravidian-speaking

rather isolated phenetic position of the Kho, nearly equivalent but distant affinities to Central Asians and the Chalcolithic of Mehrgarh, suggests two possible 131

Brian E. Hemphill , Ihsan Ali, Sarah Blaylock & Nikki Willits Gower, J. C. (1966) ‘Some distance properties of latent root and vector methods used in multivariate analysis’. Biometrika, 53: 325-338. Green, R., Suchey, J. (1976) ‘The use of inverse sine transformation in the analysis of non-metrical data’. Aermican Journal of Physical Anthropology, 45: 61-68. Hiebert, F. T. (1994) Origins of the Bronze Age Civilization in Central Asia. Peabody Museum of Archaeology and Ethnology, American School of Prehistoric Research Bulletin No. 42. Cambridge. Hiebert, F. T. (1998) ‘Central Asians on the Iranian Plateau: a model for Indo-Iranian expansionism’. In Mair, V. H. (ed.) The Bronze Age and Early Iron Age Peoples of Eastern Central Asia, vol. I. University of Pennsylvania Museum Publications. Philadelphia: 148-161. Hiebert, F. T., Lamberg-Karlovsky, C. C. (1992) ‘Central Asia and the Indo-Iranian borderlands’. Iran, 30: 1-15. Hemphill, B. E. (1999) ‘Foreign elites from the Oxus Civilization? A craniometric study of anomalous burials from Bronze Age Tepe Hissar’. American Journal of Physical Anthropology, 110: 421-434. Hemphill, B. E., Lukacs, J. R., Rami Reddy, V. (1992) ‘Tooth size apportionment among contemporary Indians: factors of caste, language, and geography’. Journal of Human Ecology, 2 (Special Issue): 231-253. Jan, S (1996) ‘History and Development of the Kalasha’. In Bashir, E., ud-Din, E. and I. (eds) Proceedings of the Second International Hindu Kush Cultural Conference. Oxford University Press. Karachi: 239-246. Jungers, W. L., Falsetti, A. B, Wall, C. E. (1995) ‘Shape, relative size, and size-adjustments in morphometrics’. Yearbook of Physical Anthropology, 38: 137-161. Kennedy, K. A. R., Chimet, J., Distotell, T., Myers, D. (1984) ‘Principal-components analysis of prehistoric South Asian crania’. American Journal of Physical Anthropology, 64: 105-118. Kruskal, J. B. (1964) ‘Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis’. Psychometrika, 29: 115-129. Kuzmina, E. E. (1998) ‘Cultural connections of the Tarim Basin people and pastoralists of the Asian steppes in the Bronze Age’. In Mair, V. H. (ed.) The Bronze Age and Early Iron Age Peoples of Eastern Central Asia, vol. 1. University of Pennsylvania Museum Publications. Philadelphia: 63-93. Linton, R. (1936) The Study of Man. Appleton. New York. Lukacs, J. R., Hemphill, B. E. (1991) ‘The dental anthropology of Prehistoric Baluchistan: a morphometric approach to the peopling of South Asia’. In Larsen, C. S. and Kelley, M. A. (eds)

samples to one another and to mature-phase Harappans. Such a pattern is in accordance with the expectations of the Early Intrusion model if the entrance of Dravidianspeaking populations into the Indus Valley occurred later, at some point between the late 5th and early 3rd millennia BC. However, the failure of odontometry to identify a similar pattern suggests the relationship between prehistoric Indus Valley populations and living Dravidian-speaking populations requires additional examination. The research design used here intentionally attempts to maximize diversity in the results obtained. Two biological systems of variation were assessed, one of which involves continuous variation and assessment through parametric statistical procedures, while the other involves discontinuous variation along an ordinal scale and assessment through non-parametric statistical procedures. Two distinct base samples of Kho, alternative data reduction procedures, and different sets of comparative samples all offer potential sources for diversity in the results obtained. While some differences in results were found, the overall patterning of affinities among samples proved remarkably consistent. Such consistency suggests these results are likely the consequence of the biological differences separating these samples and are not artefacts of either the statistical procedures employed or the specific samples included in the analysis. As such, the analytical techniques employed by dental anthropologists using data from both ancient and living populations demonstrates the potential to unravel the mysteries surrounding the peopling of South Asia.

Bibliographical References Binford, L. R. (1971) ‘Mortuary practices: their study and their potential’. In Brown, J.A. (ed.) Approaches to the Social Dimensions of Mortuary Practices. Society for American Archaeology Memoirs 25. Washington, DC: 6-29. Comas, D., Calafell, F., Mateu, E., Pérez-Lezaun, A., Bosch, E., Martínez-Arias, R., Clarimon, J., Facchini, F., Fiori, G., Luiselli, D., Pettener, D., and Bertranpetit, J. (1998) ‘Trading genes along the Silk Road: mtDNA sequences and the origin of Central Asian populations’. American Journal of Human Genetics, 63:1824-1838. Erdosy G. (1995) ‘Language, material culture and ethnicity: Theoretical perspectives’. In Erdosy, G. (ed.) The Indo-Aryans of Ancient South Asia. Walter de Gruyter: Berlin: 1-31. Goodenough, W. (1965) ‘Rethinking ‘status’ and ‘role’: toward a general model of the cultural organization of social relationships’. In Blanton, M. (ed.) The Relevance of Models for Social Anthropology. Praeger, ASA Monographs No. 1. New York: 1-24. 132

ARE THE KHO AN INDIGENOUS POPULATION OF THE HINDU KUSH ? A DENTAL MORPHOMETRIC APPROACH Townsend, G. C., Brown, T. (1978) ‘Heritability of permanent tooth size’. American Journal of Physical Anthropology, 49: 497-504. Townsend, G. C., Brown, T. (1979) ‘Family studies of tooth size factors in the permanent dentition’. American Journal of Physical Anthropology, 50: 183-190. Ward, J. H. (1963) ‘Hierarchical grouping to optimize an objective function’. Journal of the American Statistical Association, 58: 236-244. Wheeler, R. E. M. (1968) The Indus Civilization. 3rd Edition. Cambridge University Press, Cambridge. Yao, Y-.G., Lü, X-.M., Luo, H-.R., Li W-.H., and Zhang, Y-.P. (2000) ‘Gene admixture in the Silk Road region of China: Evidence from mtDNA and melanocortin 1 receptor polymorphism’. Genes and Genetic Systems, 75: 173-178. Young, R., Coningham, R., Batt, C., Ali, I. (2000) ‘A comparison of Kalasha and Kho subsistence Patterns in Chitral, NWFP, Pakistan’. South Asian Studies, 20: 133-142. Zerjal, T., Wells, R. S., Yuldasheva, N., Ruzibakiev, R. (2002) ‘A genetic landscape reshaped by recent events: Y-chromosomal insights into Central Asia’. American Journal of Human Genetics, 71: 466-482.

Recent Advances in Dental Anthropology. Alan R. Liss. New York: 77-119. Lukacs, J. R., Retief, D. H., Jarrige, J-.F. (1985) ‘Dental disease in prehistoric Baluchistan’. National Geographic Research, 1: 184-197. Mallory, J. P. (1998) ‘The Bronze Age and Early Iron Age peoples of eastern Central Asia’. In Mair, V.H. (ed.) The Bronze Age and Early Iron Age Peoples of Eastern Central Asia, vol. 1. University of Pennsylvania Museum Publications. Philadelphia: 175-201. Moorrees, C. F. A. (1957) The Aleut Dentition. Harvard University Press. Cambridge. Morgenstierne, G. (1961) Dardic and Kafir languages. The Encyclopaedia of Islam, New Edition, vol. 2 (25). E. J. Brill. Leiden: 138-139. Parpola, A (1995) ‘Formation of the aryan branch of Indo-European’. In Blench, R. and Spriggs, M. (eds) Language and Archaeology, vol. 3: Combining Archaeological and Linguistic Aspects of the Past. Routledge. London: 1-27. Quintana-Murci, L., Krausz, C., Zerjal, T., Sayar, S. H., Hammer, M. F., Mehdi, S. Q., Ayub, Q., Qamar, R., Mohyuddin, A., Radhakrishna, U., Jobling, M. A., Tyler-Smith, C., McElreavy, K. (2001) ‘Ychromosome lineages trace diffusion of people and languages in Southwestern Asia’. American Journal of Human Genetics, 68: 537-542. Renfrew, C. (1986) Archaeology and Language. Cambridge University Press. Cambridge. Renfrew, C. (1996) ‘Language families and the spread of farming’. In Harris, D. R. (ed.) The Origins and Spread of Agriculture and Pastoralism in Eurasia. University College Press. London: 70-92. Sagen, C. (1995) The Demon-Haunted World. Random House. New York. Saitou, N., Nei, M. (1987) ‘The neighbor-joining method: a new method for reconstructing phylogenetic trees’. Molecular Biology and Evolution, 4: 406425. Scott, G. R., Turner II, C. G. (1997) The Anthropology of Modern Human Teeth: Dental Morphology and its Variation in Recent Human Populations. Cambridge University Press. Cambridge. Sjøvold, T. (1977) ‘Non-metrical divergence between skeletal populations: the theoretical foundation and biological importance of C.A.B. Smith’s mean measure of divergence’. Ossa [Suppl.], 4: 1-133. Sokal, R. R., Wartenburg, D. E. (1983) ‘A test of spatial autocorrelation analysis using an isolation-bydistance model’. Genetics, 105: 219-237. Stoneking, M. (1998) ‘Women on the Move’. Nature Genetics, 20: 219-220. Southworth, F. C. (1995) ‘Reconstructing social context from language: Indo-Aryan and Dravidian prehistory’. In Erdosy, G. (ed.) The Indo-Aryans of Ancient South Asia. Walter de Gruyter. Berlin: 258-277. 133

Brian E. Hemphill , Ihsan Ali, Sarah Blaylock & Nikki Willits

Table 1 - Living and Archaeologically Derived Dental Samples

Living

Abb.

Metrics

Morphology

N

Nmax

Central Asians

GPD

177

178

Altyn Depe

ALT

25

Pakanati Reddis

PNT

184

182

Geoksyur

GKS

64

Chenchus

CHU

196

194

Sapalli Tepe

SAP

49

43

Djarkutan

DJR

69

39

Dravidian Speakers Gompadhompti Madigas

1

Prehistoric

Abb.

Metrics

Morphology

N

Nmax

Indo-Aryan Speakers

Kuzali

KUZ

42

24

GUJARAT

Molali

MOL

73

41

Vaghela Rajputs

RAJ

190

Garasias

GRS

207

South Asians

Bhils

BHI

208

Neolithic Mehrgarh

NeoMRG

42

49

Chalcolithic Mehrgarh

ChlMRG

28

25

Harappa

HAR

26

33

70

Timargarha

TMG

21

25

MAHARASHTRA Mixed Caste, Pune

RAS

70

Marathas

MRT

199

Sarai Khola

SKH

25

15

Mahars

MHR

195

Inamgaon

INM

38

41

Madia Gonds

MDA

169

Mixed Caste, Calcutta

BNG

73

Kho

KHO

Total Prehistoric

502

335

TOTAL

1838

1730

WEST BENGAL

Total Living

1

104

136

1336

1395

Nmax refers to the maximum number of individuals for which any of the tooth-trait combinations were possible to score.

134

ARE THE KHO AN INDIGENOUS POPULATION OF THE HINDU KUSH ? A DENTAL MORPHOMETRIC APPROACH

Fig.1 – Location of odontometric and dental morphology samples employed in the current analysis. Abbreviations are from Table 1. The site of Mehrgarh (MRG) encompasses two temporally distinct samples; a Neolithic period sample (NeoMRG), and a Chalcolithic period sample (ChlMRG). The site of Djarkutan (DJR) yields three temporally distinct samples; a Djarkutan period sample (DJR), and a Kuzali period sample (KUZ), and a Molali period sample (MOL).

135

Brian E. Hemphill , Ihsan Ali, Sarah Blaylock & Nikki Willits

a)

b) Fig. 2 – Patterning of intersample affinities and differences obtained from the diagonal matrix of Euclidean distances based on size standardized tooth size measurements. (a) hierarchical cluster analysis with Ward’s (1963) minimum variance technique, (b) principal coordinates analysis.

136

ARE THE KHO AN INDIGENOUS POPULATION OF THE HINDU KUSH ? A DENTAL MORPHOMETRIC APPROACH

a)

b)

Fig. 3 – Patterning of intersample affinities and differences obtained from the diagonal matrix of standardized Smith’s mean measure of divergence values obtained from the frequencies of 17 dental morphology variables. (a) neighbour-joining cluster analysis, (b) nonmetric multidimensional scaling..

137

THE UNESCO INTERNATIONAL SAFEGUARDING CAMPAIGNS FOR MOHENJO-DARO, PAKISTAN. A RETROSPECTIVE Michael Jansen

Foreword

If we differentiate the Mohenjo-Daro Campaign complex into hierarchic levels we can observe:

This is a short version of an extended research dealing with the UNESCO International Safeguarding Campaigns for Mohenjo-Daro1, Pakistan, in which the author as researcher and later as member of the UNESCO International Consultative Committee (ICC) has been involved for more than 25 years. To evaluate a campaign of such a calibre means to research into a complex programme with several parameters. In addition, most of the documents are still restricted and only available in the UNESCO archives. This is one of the reasons, why so little information of this campaign has been made available publicly. A final publication is still pending. For the evaluation of the Campaign some internal spade work has been made like the report by Felipe Vargas 1988, W.S Logan’s draft of the Up-Dated Master Plan in 1993 and the evaluation of the Campaign by A. Ardeleanu 1995. With this retrospective a first attempt is made to inform about this complex process. I am thankful to UNESCO and the Government of Pakistan for their support. All documents are available through the UNESCO Documentation Centre, Paris.

1. the UNESCO-World Community level (1979-1997); 2. the UNESCO-Government of Pakistan level (1973ongoing); 3. the UNESCO-United Nations Development Programme (UNDP) Japan Trust Fund (JPT) Government of Pakistan level (1992-1998); 4. the Government of Pakistan-Authority for the Preservation of Mohenjo-Daro (APM) level (19741997); 5. the APM- projects (ground water control, river control, conservation, landscaping, tourism) level (till 1997); 6. the APM- Mohenjo-Daro Conservation Cell (MCC) Department of Archaeology and Museums (DAM) relation level (1992-1998). Within these quite complex structures and relations we find UNESCO institutions like the International Consultative Committee, ICC (from 1979-1997), the Technical Consultative Committee, TCC (since 2003), the Authority for the Preservation of Mohenjo-Daro, APM and its Executive Board, EB (from 1974 till 1997), the National Fund for Mohenjo-Daro (since 2001) and the international UNESCO consultants who functioned diastructural within these hierarchic systems.

Mohenjo-Daro may be a unique example of demonstrating the great importance of the interrelation between archaeological research and archaeological conservation. It also shows, after first high tech attempts on the late 60s, based on the conservation believe of those days, that conservation is a highly sensitive issue primarily based on micro activities as a result of careful observation and alternating experimentation based on highly detailed documentation both, of the structures themselves but also on the documentation of the conservation process.

In January 2006 the Board of Governors for the National Fund for Mohenjo-Daro along with the Technical Consultative Committee (TCC) had been dissolved, but was re-established in January 2007. With the dissolving of the Physical Heritage Department within UNESCO and its integration into the World Heritage Centre along with the decentralization of central responsibilities for Mohenjo-Daro from Paris to the UNESCO regional office in Islamabad, Pakistan, another major change in the general management structure has taken place. Since January 2007 no meeting of the Technical Consultative Committee has taken place.

Today we can proudly announce that the Bronce Age City of Mohenjo-Daro is safed. We have a good hope, as announced in 2007 during the SAA conference in Ravenna by the Director General, Dr Fazil Dad Kakar, that the urgently needed archaeological research will continue soon I am especially thankful to Dr Hideo Noguchi, former coordinator of the program at UNESCO, Paris, with whom I could work for several years in this project. My deep respect and thanks go to Justice Abdul Kadir Shaikh, former judge of the Supreme Court and Chairman of the Authority for the Preservation of Mohenjo-Daro (APM), Pakistan.

The Bronze Age City of Mohenjo-Daro and its Cultural Significance With a complex site like Mohenjo-Daro a complex inside terminology is connected. Regarding the naming of the different excavated areas the site map may assist in their location (see Fig. 2). The abbreviations of the different excavation areas derive in most cases from the excavator´s names which are: SD = Siddiqui, DK = Dikshit, VS = Vats, HR = Hargreaves, MN = Moneer.

1

The traditional spelling according to the first excavation reports was Mohenjo-Daro, which was changed in the 60s to Mohenjo-Daro. The author keeps in principle the original spelling.

139

Michael Jansen The Indus Civilization with these of Egypt and Mesopotamia is one of the three earliest of mankind. While the latter two civilizations have been known throughout the history the knowledge about the Indus Civilization seems to have disappeared with its existance, as no direct records whatsoever survived in later cultures of this area. Only by chance excavation in the year 1922, almost simultaneous at Harappa in today's Punjab and Mohenjo-Daro, Sind, brought to light the exciting remains of this third early civilization of mankind (Fig. 1). Major excavations particularly at Mohenjo-Daro in the following years quickly showed the picture of a large urban compound, densely covered by hundreds of houses built of burnt brick of today's size and proportion. A lower but larger area with a predominantly residential character (Fig. 2) was separated by a non inhabited area of approx. 200 meters from a smaller but high place in the west with a large bath (Fig. 3), individual bathrooms, an ‘assembly hall’ and, in later excavations a so called granary. To the big surprise of the excavators neither a temple complex nor a palace was found.

history of sedentary life in Pakistan can be established back to almost the 8th millennium BC, with a continuous development from village cultures to the large urban cities of the Indus Civilization. It is nowadays also understood that the Indus Civilization developed out of the earlier cultures of the early 3rd millennium BC like the Amri, Kot-Diji- and Sothi Culture in the Indus valley proper. In the later stage around 2100 BC the Indus people were in contact with Mesopotamia and the Arabian Peninsula.

Excavations at Harappa were not as successful, as the site had been robbed of most of its burnt bricks by British railway constructors for the package of the railway Lahore- Multan shortly after 1857.

Mohenjo-Daro the history of discovery

They had a high standard of living and complex socioeconomic forms of organization, also expressed by a form of writing not yet deciphered. The reasons for the final decline and collapse of this civilization remains, one of many questions of the big riddle, as yet un-answered. Mohenjo-Daro is till today the largest of the Indus valley cities and seems to have been its centre, from which the large state was managed. Therefore it remains the most important archaeological source for the study of one of our earliest civilizations of the world.

Mohenjo-Daro was first visited and described in 1912 by the archaeologist D. R. Bhandarkar. Ten years later, in the winter 1922/1923 his successor as director of the ‘Western Circle’, R. D. Banerji, started with excavations on what he had thought to be one of the largest Buddhist stupas in Sind. At first glance his assumption was justified in excavating a monastery with a stupa in its central courtyard. This he could date into the 2nd century AD by coins of the Kushana king Vasudeva I, found in a pot in one of the cells.

In 1925 the Director General, Archaeological Survey of India, Sir John Marshall, invited the American Archaeologist Ernest H. Mackay to continue the excavations at Mohenjo-Daro. The Indian archaeologist N. G. Majumdar started after 1929 to explore the region of Sindh and the western areas of Manchhar Lake and Sind Kohistan. There he discovered several important prehistoric sites like Amri, Damb Bhuti and Ali Murad. He also identified Chanhu Daro as an Indus Civilization site. Parallel to his ongoing research Sir Aurel Stein started 1926 to carry out further exploration in the Zhob valley, Waziristan and Gedrosia. Later research was carried out in the Bahawalpur district. At the end of the twenties not only was a first chronology established, dating the Indus Civilization into the third millennium BC, but also a first interregional network of sites had been formulated from the Baluchistan mountains to the river basin of the Indus and showing the complexity of a widely distributed cultural phenomenon.

Digging deeper he very soon came across hitherto unknown artifacts, especially small square seals with an unknown script. In the same year, similar seals had been found at Harappa, 600 kilometers north in the Punjab. This discovery was the re-birth of the Indus Civilization. Today 100.000 square meters or 10% of Mohenjo-Daro have been excavated. Probably Mohenjo-Daro is still bigger as the ancient ruins are not limited to the today visible mounds. It continues underground probably up to the Indus river. This would make Mohenjo-Daro the largest known bronze age city of the world covering more than 4 million square meters.

A new impetus appeared with Mortimer Wheeler, who in 1944 became Director General of the Archaeological Survey of India. His short but very intensive presence in the Subcontinent not only reformed and modernized the Archaeological Survey of India but also led to further major discoveries at Harappa and Mohenjo-Daro. At Harappa he excavated 1945 the fortification of the citadel and in Mohenjo-Daro 1950 the ‘granary’. After partition in 1947 the inter-regional research, both in the Indian Union and Pakistan was intensified.

The major excavations were carried out between 1924 and 1931. A smaller excavation at Moneer site was carried out 1936. Mortimer Wheeler excavated 1950 the so called granary west of the Great Bath. The American archaeologist George Dales in 1965 was the latest archaeologist to have carried out excavations at MohenjoDaro (UPM or Dales area west of HR area) (Fig. 2). From 1979 till 1987 a German team headed by the author has been reviewing all excavations. In an additional joint project the German team and a team from Italy (IsMEO) carried out a complete surface survey and studies into the horizontal-vertical extent of the site.

Today more than 1500 sites belonging to the Indus Civilization are known, covering in the north-west of the Indian sub-continent approx. 1 million square kilometres, most of it in Pakistan. Recent research has shown that the 140

THE UNESCO INTERNATIONAL SAFEGUARDING CAMPAIGNS FOR MOHENJO-DARO, PAKISTAN. A RETROSPECTIVE DK-G area to trace the vertical extent of the site marked by a deep digging of more than 6 meters with more than 150.000 cubic meters excavated ground (DK-G south) and marked by several test trenches outside the daro (Mackay, 1937).

A short description of the site. The Citadel (Fig. 2) Arriving from the entrance to Mohenjo-Daro the citadel is the first prominent feature to be seen. Rising as a gigantic platform for more than twenty meters above today's alluvial ground it must have even more prominent 4500 years ago, when the surrounding plain was more than seven meters lower. Ascending the hill to the left we find the Buddhist monastery with the stupa in the centre of the courtyard dated to the 2nd century AD. From here we have an overwhelming view not only over the whole citadel but also over the major parts of the lower city to the east. The Great Bath at the foot of the stupa is one of the largest structures of the city measuring approx. 25 x 70 meters. Its centre consists of a courtyard surrounded by a cloister with big square pillars and a large basin in the middle, measuring 7x12 meters and entered by two flights of steps in the north and the south. A corbelled arch covers an almost 1.8 meter high drain to the west, which cuts a corner of the large granary. North of the Great Bath one finds a row of drained bathrooms which are interpreted as ‘priest baths’. In the south of the citadel, in L-area, you find to your right a row of square pillar bases belonging to the ‘assembly hall’.

3. The research under Sir Mortimer Wheeler 1950 in the west (REM-site) and south-east of the citadel (ACC site or ‘fortification’) to trace the fortification of Mohenjo-Daro (Wheeler, 1968). 4. The research of George Dales 1964-1965 (UPM or Dales site) to study the slope situation west of HR area in the search for the port (Dales, 1965). 5. The research of the German Research Project Mohenjo-Daro from 1979 till 1987 which concentrated on the complete re-documentation of the excavated remains (Jansen, 1983a, 1984b) and on a complete surface survey (Aachen University – Ismeo Rome Mission). In turn of this research the Moneer Area was ‘rediscovered’ as being excavated in 1936 and never published (it dropped out of the general mind). Through this research new maps were made available which form the basis of today's research and of the conservation in situ. This research had mayor consequences for the UNESCO programme at Mohenjo-Daro. We know today that the whole surface of Mohenjo-Daro is of archaeological importance which has direct consequences for the landscaping programme (restricted planting of trees, careful cover by grass) and for further the removal of ‘debris’ as was proposed in the first master plan 2. It also became clear, that the traditional method applied by the Department of Archaeology and Museums to introduce damp prove courses by inserting concrete slabs into the bases of the walls results in heavy destruction: several brick layers of the original wall have to be removed, often whole walls collapsed before finalization of the treatment. Finally, by drilling and careful studies of reports it became obvious that the anticipated depth of primary anthropogenic deposits (settlement horizons) were not as deep as anticipated by the drilling profiles obtained: the city had been constructed on gigantic platforms to save it from the annual inundation of the river Indus. Brick pieces and pottery in the drilling profiles showed secondary deposits in clay pits and dig holes for the construction of the city (Jansen, 1985).

The lower city is transversed by its ‘backbone’, the ‘First Street’. The lower city, like the citadel, was obviously planned. All houses are more or less oriented towards the cardinal directions (N-S, E-W). Most houses have their own well, indicating not only a huge water consumption but also a high standard of civil engineering. No other civilization of the world at that time had such a sophisticated system of water supply and drainage. In almost every street and lane one comes across the drains running along one of the sides. Most impressive stand the walls in DK-G area excavated between 1927 and 1931 by Ernest H. Mackay. Here we find more than three major building phases of the city which can be studied by the many walls showing entrances and drain outlets, sometimes at a considerable height. Wells stand like pre-industrial chimneys after having been excavated up to six meters height. They descend to 20 meters below the present surface. In a remote area, DK B, the famous ‘priest-king’ was found in 1925. The excavation reports of Marshall and Mackay are full of exciting stories about the discoveries of more than 50.000 registered objects, amongst them thousands of seals, terracotta figurines, household objects, bronze figurines and jewellery. All in all approx. 300 excavated house-units can be counted. The calculated total length of all excavated walls comes to approx. 50.000 meters (50 km or 35 miles!). The archaeological research carried out at Mohenjo-Daro since its discovery 1922 can be divided into five phases:

The interim results at present point towards a gigantic system of substructures which seem to have been erected in a perfectly planned manner to support the habitation area on top. In the citadel area a mud brick platform approx. 200 x 400 meters large could be traced to at least seven meters below the present surface 3. First research around the lower town proved to be similarly successful. Though several years of research by different nations and

1. The initiative research under Sir John Marshall marked by large horizontal excavations during which most of the nowadays' visible structures in the northern citadel and the southern lower town with VS and HR area were excavated (Marshall, 1931). 2. The research under Ernest H. Mackay after 1926 in 141

2

In several cases primary archaeological ground had been removed.

3

Research of Aachen University Mission 1987, unpublished.

Michael Jansen different scientists have been invested, we still know very little about this, probably largest bronze age city of the world. As will be shown scientific conservation can only be executed in coordination with research. Mohenjo-Daro is the only city in the world supplying such a wealth of information about 3rd millennium life of mankind. Therefore, as a world heritage monument belonging to the whole mankind it has to be scientifically preserved to enable further research and to enlighten future generations.

The history of conservation of Mohenjo-Daro is as old as its history of the discovery. Soon after the first excavations had been taken place after 1923 the need for conservation became evident: the exposed structures began to deteriorate due to salt action in the bricks (Fig. 4). The conservation activities can be studied in the Annual Reports of the Archaeological Survey of India (ARASI) since 1926 onwards (Urban, 1983). The ARASI report of 1930-1934 mentions under Ernest Mackay, special officer for exploration (and excavator of the DKG area) the need of replacing mud mortar and to exchange deteriorated bricks.

The question of chronology

The ARASI 1934-1935 reports the replacement of walls in the Stupa area:

Starting with horizontal by chance excavations in the early twenties Mackay continued 1927 with a systematic approach to build up a complete chronology without being able to reach virgin soil4.

‘A sum of 5710 Rps provided under the heading exploration was also utilized in repairing the monastery at the Stupa site at Mohenjo-Daro... The walls of monastic cells have been restored to the different heights so as to show their outer as well as cross walls, which, having perished down to the floor level, could hardly be traced before conservation’(ARASI 1934-35:14ff).

A second attempt by Wheeler 1950 to reach the lowest levels also failed. Wheeler added to the already known structures of the citadel the ‘granary’5. Wheeler's search for the fortifications within the citadel area was not as successful as expected. He finally was unable to prove that the discovered structure in the south-east of the citadel was, as discovered in Harappa 1946, the fortification. From nowadays's point of view his most important discovery was a big mud brick wall west of the ‘granary’ starting immediately below the surface (48.5 m amsl) and continuing down for more than seven meters (Wheeler 1968: 45). Dales' attempt to build up a new chronology and to solve the problem of the sloping area west of HR was interrupted by the Pakistan-Indian war in 1966.

The ARASI 1936-1937 mentions: ‘The exposed walls are gradually being reduced to powder on account of the action of salt except where over-burnt bricks were used in the original walls which were the least affected. Gaining by this knowledge, the new bricks were manufactured slightly over-burnt in the kiln as these are expected to resist the action of salt more than the average brick... In repairing the buildings ordinary bricks have been replaced by slightly over-burnt bricks’ (ARASI 1936-37:21)

Though researching only for less than three months he collected excellent data which 1986 have been published (Dales & Kenoyer 1986). Generally speaking, the traditional chronology is out-dated and only could be established by further excavations.

A first systematic summery of activities which had been carried out till 1970 and a first proposal for a scheme for the archaeological remains at Mohenjo-Daro was presented in 1970 by the former Director of Archaeology, Dr. F.A. Khan. As major agents of destruction Khan mentions water logging and salinity. He mentions the rise of the water table and the change of the course of the river Indus. A first systematic summery of parameters was presented in the report of the Plenderleith Mission in 1964.

The needs for conservation However, Mohenjo-Daro is permanently endangered. Since 1979 an international campaign for the safeguarding of Mohenjo-Daro has been launched under the patronage of UNESCO. The 4500 years old walls will disappear within the near future, if the human race does not conserve his most precious property. Only the original wall structure is a primary source of history which is worth it to be conserved scientifically. Just by replacing bricks we are faking history and waste money (Jansen, M. 1995).

4

Mackay's chronology consisted of three periods (early, middle and late) differentiated into three phases (I, II, III) each based on a horizontal growth model. 5

All his unpublished plans of the excavation are due to the help of Dr. Hammond, London and Professor Tosi, Rome available with the Aachen University Mission.

142

1.

The depth of the oldest foundations below the ground level estimated to 30 feet seems to be rather high. It is recommended that three or four percussion borings be executed at selected spots in order to locate the deepest formerly inhabited areas;

2.

The Indus River bed has probably risen some 15 feet since the Harappa period; the ground water table has also risen considerably since ancient times. However, no proof has been found that the water table is still rising. At the moment, there seems to be a more or less fixed equilibrium, with yearly variation of about 8 feet;

THE UNESCO INTERNATIONAL SAFEGUARDING CAMPAIGNS FOR MOHENJO-DARO, PAKISTAN. A RETROSPECTIVE 3.

a. b.

c. d.

4.

5.

programme was not directly connected to the conservation programme and has to be dealt with separately.

The rapidly increasing deterioration of baked brick walls is caused by the influence of salt crystallization near the surface of the bricks. There are four main sources of the salts:

In 1987 when the Aachen University Mission to Mohenjo-Daro carried out their condition report (Jansen, M. 1987) Prof. Oswald, member of the team, observed, that the soil humidity in the highest parts of the citadel was as high as in the more than 16 m lower plain area. He observed6 that the soil humidity is not directly dependent from the subsoil water table with the consequence that lowering of the water table by pumping would have no direct effect on the soil humidity7. Coming back to the arguments of those days, the relation: depth of walls underground - depth of ground water table salts transport - destruction of bricks by crystallization of salts in bricks became the primary basis for the execution of the ground water control as can be studied in the Master Plan 1975 and can be studied in all reports till today.

Ground water rising by capillary action (capillary fringe about 8 feet); Horizontal movement of the ground water through the walls at those places where there is a different ground level on the two sides of the walls; Salts brought into the area by rain and dust; Salts which have accumulated in the walls, partly during the thousand years of habitation and partly since the excavations were started. The four sources of deterioration will have to be dealt with separately.

Before a decision is made to execute the proposed measures, certain tests and detailed investigations should be carried out. These tests include, among others, 3 to 4 percussion borings and leaching tests;

Based on the 1987 condition report of the whole site, including all walls and the topography and on consecutive actions undertaken by UNESCO (Dr. Noguchi) a completely new strategy could be developed for the early 90s. Accordingly the inter-relational actions were defined as follows and based on following observations:

If the results of these tests prove to be satisfactory, a decision should be made whether a ‘minimum or maximum’ programme for the protection of the ruins should be adopted. The minimum programme consists of measures to protect only the presently excavated area. The maximum programme provides for the possibilities of deeper excavations and for excavations of the yet untouched area to the south of the city.

‘The anthropogenic material identified in depths up to 60 feet (approx. 18 meters or 30 m amsl, today's surface of the plain is 47 m amsl) does not represent primary occupation levels but secondary ones. Lowest primary occupation levels may be expected down to approx. 7 to 9 m below the present surface’ (Jansen, M. 1995).

In the light of the conclusions, the mission recommended the following measures to be taken to check the four main sources of salts:

Major capillary destruction of bricks takes place below 32,4 °C (dehydrate state thenardite (Na2SO4 density 2.68) hydrated state mirabilite (NA2SO4 . 10H2O, density 1.46) with the help of water. The water source is most probably atmospheric e.g. dew and not the rising humidity in the walls. Therefore we have to differentiate between water as a salt transporter and water as a catalyst for the change of the mineral phases thenardite/ mirabilite of the sodium sulphate (Goudie, 1977; Ludwig, 1983).

a. Lowering the ground water table; b. Digging of trenches behind walls or filling low lying streets with fresh river sand; c. Regular cleaning and maintenance of walls where salts outcroppings occur; d. A well organized and systematic programme of leaching the salts from the walls; e. Prohibiting rice cultivation east of the Dadu canal (Khan, F.A.1970:15-16).

Inter-relational actions

This, rather unknown publication by F.A. Khan is in so far of interest as it antedates the Master Plan by two years. The pathological arguments, listed here, have remained more or less the same for all conservation actions till the middle of the 1980s when new research made different actions necessary. They were based on the interrelation: depth of walls underground - depth of ground water table - salts transport - destruction of bricks by crystallization of salts in bricks.

According to observations the inter-relational action is as follows: soil humidity (capillary action of silt, clay etc.) rather independent (within range of 10-20 metres?) from depth of ground water table - depth of walls underground max 7-9 m, capillary action in walls (bricks plus

6 This observation has been proved by measurings of MCC the soil humidity at a depth of -0,5m below surface is approx. 23% all over the site.

The consequences for the conservation programme were (as listed in the Master Plan) lowering of the water table, installation of damp prove courses (DPC), exchange of salty earth, reduction of aerosols etc. The river protection

7 Goudie (1977:75) had made similar observations in 1977 and had doubted already in those days the functioning of the pumping scheme.

143

Michael Jansen horizontal layers of mud mortar8 every course of bricks) unknown - depth of ground water table - salts transport by capillary action in soil plus walls - evaporation of water close to soil - wall surfaces plus crystallization of salts - destruction of bricks by hydration - de-hydration process of sodium sulphate under specific conditions of temperature and atmospheric humidity.

found a chance to be employed by the Government on permanent base. Technically speaking the following results were obtained and new methods introduced. Once the relation groundwater- salt transport- salt crystallisation thenardite - mirabilite process was fully understood, it became clear that the original brick structures could not be saved only by the applied methods so far. The solution finally was found in a simple but very effective method: The treatment and coverage of the original brick surfaces by a thin sacrificial layer of mud slurry (Fig. 5). This traditional method with very low material and labour costs was the optimum solution to the complex problem. Tests within the UNDP and later workshops clearly showed the success, based on:

According to this interrelation the destruction process only starts with the hydration - de-hydration process of the sodium sulphate in the capillaries of the bricks and not with the transportation of salts by capillary actions. Another dangerous agent of destruction is the erosion through rain water. Research results (Jansen, M. 1987) have shown that water erosion was already a problem in ancient days if not to say the problem of ancient days. Extensive repair works with mud bricks, fillings of stagnant water holes with debris in streets and court yards, gigantic collections of industrial waste and pot sherds (south of HR area) for the consolidation of the alluvial urban ground show that already 4500 years ago the maintenance and conservation of urban structures and houses was a major concern of the citizens of MohenjoDaro.

1. topping the walls with layers of earth bricks (protection against mechanical and thermal stress); 2. protecting the sides of the walls with clay slurry (sacrificial layer for the crystallisation of salts and protection against thermal stress and re-pointing the endangered walls; 3. special control and protect the bases. Regular monitoring, of course, was an absolute must.

As more than 50% of the urban structures in MohenjoDaro consist of mud brick and earth structures, the conservation of the city means not only the conservation of burnt brick fabric but also of the earthen structures.

This basic treatment of the wall surfaces with a slurry also could allow the groundwater to flow in the walls, as the rising water itself did not endanger the original fabric. Also the primary crystallisation of the sulfates in the sacrificial layer surface caused no damage. The actual destruction of the original brick capilaries would have taken place through re-solution with water of the salts below 32 degrees in the change of the thenarditemirablite phase.

Consequences The extended understanding of the pathological conditions of the site resulted in further investigations into potential treatments. With the enormous increase of costs for the running of the ground water control scheme arguments for a mid and long term strategy for the protection of the Bronze Age city of Mohenjo-Daro had to be developed. Based on the extended observations of the groundwater- salt transport- brick destruction chain parameters had to be worked out jointly with the cost factors on a long term strategy for the Government of Pakistan.

As this water as catalyst mostly in the winter period comes in Mohenjo-Daro from the dew in the air and not from the ground, none of the traditional techniques like lowering of groundwater table or insertion of damp proof layers in the walls would have prevented the walls from this destruction. With the full cover of the walls with the sacrificial layers of clay, finally the original fabric was safe. With these new recognitions, in the early 90s the Campaign was newly designed and experts were accordingly trained by UNESCO in different workshops.

The additional funding of a training project through UNDP was dearly welcomed. UNESCO jointly with the Government of Pakistan developed a training programme for young Pakistani professionals to specialise in conservation of archaeological sites. With several earth conservation workshops, taken care of by experts like Richard Hughes, UK, and the trainees were introduced to the sensible topic of how to deal with such a large site. The training programme consisted of scientific documentation, conservation of earth structures, site management and history and ethics of conservation along with ICOMOS/ UNESCO guidelines. Unfortunately, none of the trainees, after the completion of the scheme,

As the river training programme had been successfully completed and the groundwater control scheme had become of secondary importance, only the landscaping scheme remained to be completed. Based on several studies for a master plan and also on the new recognition that there is no need anymore for prohibiting any irrigation close to the site, a new landscaping concept should be developed and implemented. In 1997 UNESCO and the Government of Pakistan officially declared the Campaign as successfully completed. The remaining funds were partly returned by the Pakistani Government to the UNESCO Headquarters

8

First tests have shown that horizontal layers of mud (clay, silt?) mortar may under specific conditions (water saturation) function as water transport blockers.

144

THE UNESCO INTERNATIONAL SAFEGUARDING CAMPAIGNS FOR MOHENJO-DARO, PAKISTAN. A RETROSPECTIVE and partly turned into a National Trust Fund to be further used and spent on the conservation on Mohenjo-Daro as one of the most important Bronze Age cities of the world.

Dames W., Jacobs M., Wanzke H. (1984) “Close Range Aerial Photogrammetry in Archaeology”. In Jansen, M. Urban, G. 1984: Dwars, H. and Verhay Ltd. NEDECO (1968) Desalinization of the Monuments of Mohenjo Daro. UNESCO Serial No. 654/BMS. RD/CLT Paris. Fosberg, R. (1979) Informal Report Smithsonian Institution Mohenjo-Daro Report, UNESCO. Fosberg, R. (1980) Technical Report The Plant Ecosystem for Mohenjo-Daro, UNESCO. Serial No. FMR/CC/CH/80/189 (RP/1977-78/4.121.6). Goudie, R. S. (1977) “Sodium Sulphate Weathering and the Disintegration of Mohenjo-Daro, Pakistan”. Earth Surface Processes, 2: 75-86. Gussone, R. (1983) “Mineralogical Investigations on 2 Stone Specimens from Mohenjo Daro”. In Jansen, M. & Urban, G. (1983) Dokumentation in der Archäologie. Techniken, Methoden, Analysen. Veröffentlichungen der Seminarberichte vom 5. 6. Dezember 1981, Aachen: 165-169. Hughes, R. (1992) Archaeological Remains: Conservation Experiments. Restricted Technical Report FIT/536-PAK-71. Serial No. FMR/CLT/CH/92/219 (FIT). Hughes, R. (1993) Archaeological Surface Burial and Wall Stabilization by the Use of Geotextils/ Geogrids at Mohenjo-Daro. Internal UNESCO Report. Jansen, M. & Urban, G. (eds) (1983) Dokumentation in der Archäologie. Techniken, Methoden, Analysen. Veröffentlichungen der Seminarberichte vom 5. 6. Dezember 1981, Aachen. Veröffentlichungen des Geodätischen Instituts der RWTH Aachen, Nr. 34. Jansen, M. & Urban, G. (eds) (1984) Interim Reports Vol. 1. Reports on Field Work Carried out at Mohenjo Daro Pakistan 1982 83. IsMEO Aachen University Mission. Aachen. Jansen, M. & Urban, G. (eds) (1985) Mohenjodaro. Data Collection, Vol I. Fieldbooks and Concordance of HR Area. Leiden. Jansen, M. & Urban, G. (eds) (1987) Interim Reports vol. 2. Reports on Fieldwork Carried out at Mohenjo Daro, Pakistan, 1983-84. IsMEO Aachen University Mission. Aachen. Jansen, M. (1982) Preservation as a Way of Planning. In Buschmann, K. H. (ed) (1982) International Seminar Preservation & Documentation of Traditional Architecture 24-27 March 1982, Karachi: 71-77. Jansen, M. (1983) “Preliminary Results of Three Years Documentation in Mohenjo Daro”. In Jansen, M. & Urban, G. (eds) (1983) Dokumentation in der Archäologie. Techniken, Methoden, Analysen. Veröffentlichungen der Seminarberichte vom 5. 6. Dezember 1981, Aachen: 21-36.

As such with the case of Mohenjo-Daro we do not only have the story of a discovery, the history of the probable capital of a Bronze Age Civilisation but also the history of over 30 years of conservation science, application, ideology and management practices. As mentioned in the beginning: with the development of the new treatment conceptions today Mohenjo-Daro seems to be safed. We obviously have not only found the right recipe but this recipe is applicable, manageable and financiable. Meanwhile the UNESCO ban for excavations has been lifted and there is good hope for future new research programs. May the case of Mohenjo-Daro help to improve archaeological conservation as a major issue of future cultural tourism.

Bibliographical References Abbott, J. (1924) Sindh. A Re Interpretation of the Unhappy Valley, Bombay. Alcock, L. (1952) “Exploring Pakistan`s Past: The First Year`s Work”. Pakistan Quarterly, 2 (1): 12 16 ARASI Annual Reports of the Archaeological Survey of India. Calcutta. Ardeleanu, A. (1985) “Brief History of Excavations in HRArea”. In Jansen, M. & Urban, G. (eds) (1985) Mohenjodaro. Data Collection, Vol I. Fieldbooks and Concordance of HR Area. Leiden: X-XV. Ardeleanu, A. (1995) Evaluation Report on MohenjoDaro. Internal Report, UNESCO. Bhandarkar, D. R. (1912) “Excavation”. In Progress Report of the Archaeological Survey of India: Western Circle, for the Year Ending 31 March 1912 : 4-5. Calcutta Buschmann, K. H. (ed) (1982) International Seminar Preservation & Documentation of Traditional Architecture 24-27 March 1982, Karachi. Clifton, J. R. (1980) Preservation of Structural Remains at Mohenjodaro. Restricted Technical Report RP/1929 80/ 4/7. 6/05 UNESCO Serial Number FMR/CC/ CH/80/104. Paris Cunningham, A. (1875) “Harappa”. In ASIM 1872, 73: 105-108. Dales, G. F. (1964) “The Mythical Massacre of Mohenjo Daro”. Expedition, 6 (3): 37-46. Dales, G. F. (1965) “New Investigations at Mohenjo Daro”. Archaeology, 18 (2): 145-150. Dales, G. F. (1966) “The Decline of the Harappans”. In S Am, 214 (5): 93-98. Dales, G.F. & Kenoyer, M. (1986) Excavations at Mohenjo-Daro, Pakistan: The Pottery. Philadelphia.

145

Michael Jansen Jansen, M. (1984a) “Theoretical Aspects of Structural Analyses for Mohenjo Daro”. In Jansen, M. & Urban, G. (eds) (1984) Interim Reports Vol. 1. Reports on Field Work Carried out at Mohenjo Daro Pakistan 1982 83. IsMEO Aachen University Mission. Aachen: 39-62. Jansen, M. (1984b) “Architectural Remains in Mohenjo-Daro”. In Lal, B. B., Gupta, S. P. (eds) Frontiers of the Indus Civilization. New Delhi. Jansen, M. (1985) “Mohenjo Daro, cite de l'Indus”. La Recherche, 163 (16): 166-176. Jansen, M. (1987) Evaluation Report Constructural remains Mohenjo-Daro. UNESCO Internal Report. Jansen, M. (1989) “Water Supply and Sewage Disposal at Mohenjo Daro”. World Archaeology, 21(2):177192. Jansen, M. (2003) Moenjo-Daro Pakistan, The Problem of Adobe & Brick Conservation; 9th International Conference on the Study and Conservation of Earthen Architecture. Bam, Iran. Jansen, M. (ed) (2004) Revised Evaluation Report on structural conditions. UNESCO. Khan, F. A. (1970) Preservation of Mohenjo-Daro. Karachi Khan, A. N. (ed) (1975) Proceedings of International Symposium on Mohenjo-Daro 1973 Lambrick, H. T. (1964) Sind: A General Introduction. Hyderabad , 2nd Edition, 1975. Ludwig, U. (1983) Die Zerstrung von Ziegel durch Salze. In Jansen, M. & Urban, G. (eds) (1983) Dokumentation in der Archäologie. Techniken, Methoden, Analysen. Veröffentlichungen der Seminarberichte vom 5. 6. Dezember 1981, Aachen: 153-165. Logan, W. S. (1993) Updated Masterplan 1993-2003. Internal UNESCO report. Mackay, E. J. H. (1935) Early Indus Civilization, London. Mackay, E. J. H. (1938) Further Excavations at Mohenjo-Daro, Delhi. Marshall, Sir J. (1924) “First Light on a Long Forgotten Civilization”. ILN, 20 (9): 528-532, 548. Marshall, Sir J. (1931) Mohenjo-Daro and the Indus Civilization, London. Mughal, R. (1971) The Early Harappan Period in the Greater Indus Valley and Northern Baluchistan (ca. 3000 2400 B. C.). Ph. D. Dissertation University of Pennsylvania. Ann Arbor, Michigan. Naqvi, S. A. (1973) “Five Thousand Year Old City of Mohenjo-Daro”. AC,5: 24-31. Oswald, R. (1987) Informal Report. UNESCO - Project: Save Mohenjo-Daro- Classification of Damage Occurring at the archaeological site of MohenjoDaro, Aachen, 30.09.1987.

Plenderleith, H. (1964) UNESCO Mission of Experts. Preservation of Monuments at Mohenjo-Daro 1964. (Plenderleith team report). UNESCO, Bibliotheque Centrale, microfiche 75FR 0115. Praccia, S. (1987) “Surface Analysis of Pottery Manufacture Areas at Mohenjo-Daro”. In Jansen, M. & Urban, G. (eds) Interim Reports vol. 2. Reports on Fieldwork Carried out at Mohenjo Daro, Pakistan, 1983-84. IsMEO Aachen University Mission. Aachen Raikes, R. L. (1963) “The End of the Ancient Cities of the Indus Civilization in Sind Baluchistan”. Am An, 65 (3): 284-299. Raikes, R. L. (1965) “The Mohenjo Daro Floods”. Antiquity, 39: 126-203. Raverty, H. G. (1979) The Mihran of Sind, Lahore. Rehmatullah, Ch. (1975) “Leaching Experiments to Eliminate Salt Effects at Mohenjo-Daro”. In Khan, A. N. (1975) Proceedings of International Symposium on Mohenjo-Daro 1973:100-107. Tabata, J. (1989) Informal Paper. The Mohenjo-Daro Area Landscape Planning Project. Paris, UNESCO. Urban, Th. (1983) “Some Problems Concerning the Evaluation of Wall Structures in Mohenjo-Daro”. In Jansen, M. & Urban, G. (eds) (1983) Dokumentation in der Archäologie. Techniken, Methoden, Analysen. Veröffentlichungen der Seminarberichte vom 5. 6. Dezember 1981, Aachen: 21-37, Aachen. Vargas, F. (1988) Draft Report. Mohenjo-Daro International Campaign Remarks and Observations an a Safeguarding Programme, by Felipe Vargas, Consultant CC/CH, UNESCO, Paris,1988; MOFESI.HD. Wheeler, Sir M. (1953) The Indus Civilization, Cambridge . Wheeler, Sir M. (1968) The Indus Civilization. 3rd Edition, Cambridge.

146

THE UNESCO INTERNATIONAL SAFEGUARDING CAMPAIGNS FOR MOHENJO-DARO, PAKISTAN. A RETROSPECTIVE

Mohenjo-Daro

Fig. 1 – Satellite image of the North-west Indian Subcontinent with the location of Mohenjo-Daro.

147

Michael Jansen

Fig. 2 – Topographic map of Mohenjo-Daro showing the excavated areas and the collector drain with pump stations of the UNESCO Campaign.

Fig. 3 – The ´Citadel´ area from west with the ´Great Bath´ in front and the Kushana stupa remains in the back. Walls are treated with clay slurry.

148

THE UNESCO INTERNATIONAL SAFEGUARDING CAMPAIGNS FOR MOHENJO-DARO, PAKISTAN. A RETROSPECTIVE

Fig. 4 – A wall conserved in 1976 without clay slurry showing the destruction of brick capillaries both in original and replaced bricks.

Fig. 5 – Walls in DK-G area with different clay treatments.

149

A NEW LATE BRONZE AGE SITE IN SOUTHERN UZBEKISTAN Kai Kaniuth

opinions exist as to their interpretation2. Absolute chronology initially depended on external relations. In the case of pottery, comparisons abound with other Namazga VI groups in Turkmenistan and North Afghanistan, but these are not dated more securely; other small find categories are not sensitive enough for fine-dating. A series of 26 radiocarbon dates from Dzarkutan established a time bracket of the 20th-15th centuries BC, but these samples have not yet been published with reference to certain ceramic assemblages, so we lack a good resolution within this 500-year span (Görsdorf and Huff 2001).

Southern Uzbekistan in the Late Bronze Age – the Sapalli Culture Today, as in ancient times, the people living on the territory of the Republic of Uzbekistan are concentrated in the riverine oases of the country’s east. Here, the Syrdarya, Zerafshan and Surkhandarya rivers emerge from the mountainous zone of the Pamir-Alai and Western Tian-Shan ranges, and eventually flow into the deserts and semi-deserts of the Turanian basin. The southernmost of these oases is irrigated by the Surkhandarya river and its tributaries, whose waters discharge into the Amudarya.

When it comes to settlements, we have a couple of plans of monumental complexes which have been quite reasonably compared to contemporary Qalas, but this term is purely descriptive, and, just like the interpretive terms ‘temple’, ‘fortress’ etc.. based upon purely formal analogies. Complete inventories (including the pottery assemblages) which might shed some light on room function have not been published, and this is true for contemporary buildings in Margiana and North Afghanistan, as well. A functional analysis of an entire building complex – or at least a representative portion of one – is therefore currently the major desideratum in the study of the Sapalli Culture.

During the Late Bronze Age, the period of approximately the 20th-15th centuries BC, the Surkhandarya region was settled by people whose material culture is known under the term Sapalli Culture (literally ‘pottery culture’, named after the eponymous site of Sapallitepe). It forms the easternmost wing of the Namazga cultural sphere, and thus the northeastern boundary of the Ancient Near Eastern world1. The Sapalli Culture has been studied since the late 1960s, mostly through the efforts of teams from the Institute of Archaeology of the Uzbek Academy of Sciences, headed by Ahmadali Askarov, who directed excavations at the major sites of Sapallitepe and Dzarkutan (Askarov 1973; Askarov 1977; Askarov and Abdullaev 1983; Askarov and Shirinov 1993; for summaries see Askarov 1981; Kohl 1984; Kaniuth 2006).

Sapalli Culture sites are found all along the western half of the Surkhandarya plain. Several sites are known from surveys, but only three have been investigated so far: Of the eponymous site of Sapallitepe, only the central building complex covering a surface of 0,7 ha, with 148 rooms and 138 burials inside the built-up area, are known (Askarov 1973; Askarov 1977). Sapallitepe dates to the 20th and 19th centuries BC. The twin sites of Dzarkutan and Bustan, on the other hand, comprise both architectural complexes and large necropoleis, dating 1900-1500 BC (Askarov and Abdullaev 1983; Askarov and Shirinov 1993; Huff 2001). The Molalitepe has only been sounded (Beljaeva and Khakimov 1973), and all remains here would appear to belong to the 17 th-16th centuries BC. Still within the Surkhandarya plain but on Tajik territory is the late Sapalli Culture site of Tandyrjul, whereas the earlier grave of Zardca Khalifa in the Zerafshan region with its typical LB I ceramic forms cautions against considering the question of the regional extent of the Sapalli Culture settled (Bobomullaev 1997). In the light of this distribution, it was not altogether unexpected, when a new Sapalli Culture site was

When reviewing our knowledge of the Sapalli Culture, it must be pointed out first, that our information derives primarily from the comparatively abundant burial remains. More than 1100 interments have been excavated at Dzarkutan, with a further 50 from the neighbouring site of Bustan, and 138 burials were studied in Sapallitepe. These figures do not include a large cemetery currently excavated by Samarkand University, about which up to date figures are not available but which yielded at least a further 350 graves. Of these ca. 1700 burials, an estimated 5 % are published in full, and these are the basis for our relative chronological framework and nearly all of our inferences on the structure of Sapalli Culture society. The data has been ordered through typological approaches and with the help of seriation, and conflicting

1

The question of Southern Tajikistan as a contact zone between the Near Eastern and Eurasian macroregions will not be discussed here (but see Vinogradova 2004 for an excellent summary of the archaeological evidence pertaining to the Late Bronze Age).

2 The familiar division of successive Sapalli-, Dzarkutan-, Kuzali-, Molali- and Bustan stages was found to lack precision during the campaigns of the German Archaeological Institute at Dzarkutan, and a new system of two periods, Late Bronze (LB) I and II, each internally subdivided into phases, was proposed. For details of this new chronological system, which is endorsed here see Teufer 2005.

151

Kai Kaniuth discovered in the summer of 2006 in the western piedmont zone of the Kugitangtau mountains, close to the village of Pashkhurt (Fig. 1).

Tocharistan Expedition’s 2006 sounding. The cultural remains are up to 3 m thick, but work has not yet begun on the highest part of the site.

Tilla Bulak

The architecture consists of mudbrick walls, sometimes on pebble foundations; stone boulders were used to secure the slopes of the hill; brick sizes vary between 24 × 48 × 12 and 28 × 56 × 14 cm. Remains of the second building phase were discovered on a length of some 30 m, and their identical orientation indicates that a certain degree of planning was involved in their construction. All of this is in line with the information we have from other sites such as Sapallitepe and Dzarkutan. While the architectural remains do not extend down the steep slopes of the southern spur, we are not sure about the limits of occupation in the north. So far, settlement remains were discovered in all trenches except for the lowest sounding of the step trench.

The Pashkhurt area was first surveyed by Edvard Rtveladze in the 1960s (Rtveladze and Khakimov 1973) and archaeological work has since been carried out at two pre-Islamic to Medieval sites just north of the village, Dabil-Kurgan and Ghisht Tepe (Mkrtychev, Bolelov and Ilyasov 2005). During his foot survey, Rtveladze had concentrated on the valley floor, along the major watercourses and so missed the settlement of Tilla Bulak which lies just inside a chain of hills separating the Pashkhurt area from the Surkhandarya plain proper. As it turns out, Tilla Bulak is situated in a very favourable niche, commanding good views all the way to the Kugitangtau in the west, but also controlling a small valley eastwards, which is well watered by a perennial spring and constitutes one of the few ways of access (through small trails) from Pashkhurt down into the Surkhandarya plain.

The pottery has been classified according to wares and types. While neither study has been concluded, the following picture can be outlined: The fine wares – predominantly with a greenish hue, but also including red and grey wares – make up more than ninety percent of the assemblage. They appear to be all of local production, while the cooking pots were manufactured of clay from a different source4. The types differ slightly from the ones we know from the Surkhandarya plain graves, but they coherently point to the incipient phase of the Sapalli Culture: Spouted bowls, teapots and stemmed bowls with S-shaped rims are typical forms, and together with the presence of grey ware, which also appears in the early site of Sapallitepe, they suggest an LB Ia date for the settlement at Tilla Bulak5. A radiocarbon date of 3615 +/23 (Hd-26067; 2032-1905 cal. BC at two standard deviations) for a context linked to the earliest constructional phase ties in well with the previously suggested 20th/19th century date for this period.

The site was discovered in 2006 by Sergej Bolelov, a member of the Tocharistan Expedition, who sounded the place and suggested a Late Bronze Age date. Since April 2007, Tilla Bulak is being investigated by a joint expedition of Ludwig-Maximilians-Universität, Munich (LMU), the Tocharistan Expedition of the National Institute for the Study of the History of Art (TE) and Termez University (UT)3. Tilla Bulak is a settlement on a natural hillock, rising up to 12 m above the plain to the west, and some 20 m above a small valley to the east, which is easily irrigable with water from the ‘Gold Spring’ just north of the site. The tepe forms a plateau slowly but steadily rising from north so south, with rather steep flanks on its western, eastern and southern sides. The highest point of the site is thus close to its southern end (Fig. 2).

Whether the variations we pick up with respect to the Surkhandarya Plain ceramic material are chronological (does Tilla Bulak partially predate Sapallitepe?), regional (may differences be due to stronger influences from either Margiana or North Afghanistan?) or functional (do settlements generally contain material different from grave assemblages or are we dealing with specialised functions of the contexts excavated in Tilla Bulak?) will be among the questions to be addressed in the future.

A four-week preliminary campaign was carried out in March and April 2007, during which an area of 80 square meters was sounded down to the first architectural remains, including a step trench and a deep sounding to test the overall depth of the deposit (Fig. 3). The primary goals were to determine as far as possible the approximate size of the settled area and to gain a basic understanding of the occupational history of the site. Particular emphasis was placed on the recovery of a pottery sample to enable a secure dating of the site.

Only a few small finds can be mentioned here (Fig. 4): The co-occurrence of leaf-shaped and triangular arrowheads is interesting in the light of what we know of flint arrowheads in the region, even though in principle both types are present in Sapallitepe (Baratov 1986; Shirinov 1986: fig. 1.28). Likewise, the potmark of a bow and arrow is of a type uncommon in the Sapalli Culture repertoire (but see Askarov 1973: Fig. 36).

In the 2007 soundings, two architectural strata were identified, with a third stratum being present in the 3

The preliminary campaign of 2007 was funded by the Gerda-HenkelFoundation. Generous support was given by the German Archaeological Institute. Participants of the 2007 campaign were: J. Ilyasov (TE), M. Teufer, E. Roßberger, A. Kurmangaliev (LMU), Sh. Shaydullaev, A. Shaydullaev (UT).

4 These preliminary results of the pottery analysis were communicated by Malgorzata Daszkiewicz (Warsaw), and Gerwulf Schneider (Berlin). 5

152

For the types defining the Late Bronze I and II stages see Teufer 2005.

A NEW LATE BRONZE AGE SITE IN SOUTHERN UZBEKISTAN Avanesova, N. A. (1995) ‘Bustan VI, une nécropole de l’Age du Bronze dans l’ancienne Bactriane (Ouzbékistan méridional): témoignages de cultes du feu’. Arts Asiatiques 50: 31-46. Baratov, S. R. (1986) ‘K voprosu tipologicheskoj klassifikatsii kamennikh nakonechnikov strel epokhi bronzy Srednej Azii’. Istorija Materialnoj Kul’tury Uzbekistana 20: 19-28. Beljaeva, T. B. and Khakimov, Z. A. (1973) ‘Drevnebaktrijskie pamjatniki Mirshade’. In Pugachenkova, G. A. (ed.), Iz antichnoj kul’tury Uzbekistana, 35-51. Tashkent. Bobomulloev, S. (1997) ‘Ein bronzezeitliches Grab aus Zardza Khalifa bei Pendzikent (Zerafshan-Tal)’. Archäologische Mitteilungen aus Iran und Turan 29: 121-134. Görsdorf, J. and Huff, D. (2001) ‘14C-Datierungen von Materialien aus der Grabung Džarkutan, Uzbekistan’. Archäologische Mitteilungen aus Iran und Turan 33: 75-87. Huff, D. (2001) ‘Bronzezeitliche Monumentalarchitektur in Zentralasien’. In Eichmann, R. and Parzinger, H. (eds), Migration und Kulturtransfer. Der Wandel vorder- und zentralasiatischer Kulturen im Umbruch vom 2. zum 1. vorchristlichen Jahrtausend, 181-198. Bonn Kaniuth, K. (2006) Metallobjekte der Bronzezeit aus Nord-Baktrien. Mainz. Kaniuth, K. (2007) ‘Tilla Bulak 2007. Bericht über die erste Kampagne’. Archäologische Mitteilungen aus Iran und Turan 39: 31-47. Kaniuth, K. (2009) ‘Tilla Bulak 2008. Vorbericht zur zweiten Kampagne’. Archäologische Mitteilungen aus Iran und Turan 41: 75-96. Kaniuth, K. (2010) ‘Tilla Bulak 2009. Vorbericht zur dritten Kampagne’. Archäologische Mitteilungen aus Iran und Turan 42: 129-163. Kohl, P. L. (1984) Central Asia. Palaeolithic beginnings to the Iron Age. Paris. Mkrtychev, T. K. and Bolelov, S. B. and Ilyasov, D. (2005) ‘Issledovanija na juge Uzbekistana’. Arkheologicheskie Otkrytija 2004 goda: 524-528. Rtveladze, E. V. and Khakimov, Z. A. (1973) ‘Marshrutnye issledovanija pamjatnikov Severnoj Baktrii’. In Pugachenkova, G. A. (ed.), Iz antichnoj kul’tury Uzbekistana, 10-34. Tashkent. Shirinov, T. (1986) Orudija proizvodstva i oruzhie epokhi bronzy sredneaziatskogo mezhdurechja. Tashkent. Teufer, M. (2005) ‘The Late Bronze Age chronology of Southern Uzbekistan. A reanalysis of the funerary evidence’. In Franke-Vogt, U. and Weisshaar, H.J. (eds), South Asian Archaeology 2003, 199-209. Aachen. Vinogradova, N. M. (2004) Jugo-zapadnyj Tadzhikistan v epokhu pozdnej bronzy. Moscow.

Summary and outlook Tilla Bulak is a stratified Late Bronze Age site with a relatively short occupational history, very much like that of Sapallitepe, some 25 miles downstream. Three architectural phases can so far be differentiated. Its architecture may be comparable to that of better-known settlements in the Surkhandarya Plain in that some degree of planning seems to have been involved in the initial construction. The tepe has a surface of c. 1 ha, of which at least 0,5 ha were settled. While we are fairly certain about the eastern, southern and western limits of the site, which appear to closely follow the contour of the hill and do not extend down the slope, the northern boundary of the settled area was not reached. As to absolute chronology, the pottery recovered and the dates obtained through radiocarbon dating of charcoal samples do indicate a date within the Late Bronze I stage, and more closely the LB Ia phase (20th - 19th centuries BC), the very beginning of the Sapalli Culture. Our plans for the future are to widen the horizontal exposure of the excavation with the aim of making a functional analysis of the settlement. Stratigraphic analysis must further be systematically combined with the recovery of radiocarbon dates to complement the evidence we have from Dzarkutan and Sapallitepe and to put our chronology of the Sapalli Culture on a more secure footing. Also, the evidence from Tilla Bulak will enable us to assess issues of regional diversity within the Sapalli Culture during Late Bronze I, its formative and – in terms of interregional connections – most interesting phase. Addendum Preliminary reports have meanwhile appeared on the campaigns 2007-2009 (Kaniuth 2007; 2009; 2010). The reader is kindly referred to these publications for further reference.

Bibliographical References Askarov, A. A. (1973) Sapallitepa. Tashkent. Askarov, A. A. (1977) Drevnezemledel’cheskaya kul’tura epokhi bronzy Juga Uzbekistana. Tashkent. Askarov, A. A. (1981) ‘Southern Uzbekistan in the Second Millennium BC’. In Kohl, P. L. (ed.), The Bronze Age civilization of Central Asia. Recent Soviet discoveries, 256-272. Armonk. Askarov, A. A and Abdullaev, B. (1983) Dzharkutan. Tashkent. Askarov, A. A. and Shirinov, T. (1993) Rannaja gorodskaja kul’tura epokhi bronzy Juga Srednej Azii. Samarcand. Askarov, A. A. and Shirinov, T. (1994) ‘The palace, temple and necropolis of Jarkutan’. Bulletin of the Asia Institute 8: 13-25. 153

Kai Kaniuth

Fig. 1 – Distribution of Sapalli Culture sites.

Fig. 2 – Tilla Bulak seen from the West.

154

A NEW LATE BRONZE AGE SITE IN SOUTHERN UZBEKISTAN

Fig. 3 – Contour plan of Tilla Bulak. The areas excavated in 2006 and 2007 are shaded.

Fig. 4 – Small finds from Tilla Bulak 2007

.

155

EXCAVATION AT KANMER, A HARAPPAN SITE IN GUJARAT, INDIA: SOME OBSERVATIONS J. S. Kharakwal, Y. S. Rawat & Toshiki Osada

site is locally known as Bakar Kot. Aludawaro Vokro, a seasonal nala, originating in the hillocks located hardly two km west and north-west of the site, drains the area around the ancient site. The settlement is barely five kilometers away from the Little Rann of Kachchh and the landscape around the site is composed of Jurassic piedmont having current bedded ferrogenous sanstone, limestone and variety of shale. And it is located about one kilometer off the national highway numbered 15.

Summary Kanmer, a small fortified Harappan settlement, is located about 23 km southwest of Surkotada, another excavated site in Kachchh, Gujarat, India. The ongoing excavation at the site is a joint venture of Indo-Japanese collaboration and this paper is a summary of results of three season's work completed so far. Our small scase excavation has revealed a five-fold cultural sequence i.e., Early Harappan, Mature Harappan, Late Mature Harappan, Historical phase and Medieval. The fortification is squarish on plan inside which are remains of a planned Harappan settlement. The Mature Harappan deposit has been subdivided as Period IIa and II b due to introduction of some new ceramic types found in the later. Regular burning activity is evident in the northern and eastern sector of the site during Pd II b as the evidence of several episodes of ash dumping on either side of the fort wall would indicate. The defence wall was made quite strong for some reason as its width (about 18m) at the middle level suggests. A variety of beads of terracotta, semi precious stone, paste, steatite, faience and shell, weights of agate, a terracotta seal impression and a sealing are some of the minor antiquities. Evidence of domesticated plants and animals has been found right from Early phase at the site.

The vegetation available here is characterised by tropical thorny forest. Azadirachta indica (neem), Piloo or Zar tree, Calotropis sp., Acacia sp., Tamarindas sp., Mangifera sp. and Ricinus communis are common plants of Kachchh. The main land of Kachchh like the eastern margin of the Little Rann, grows a variety of grasses supporting the livestock, such as bokana (Cressa cretica), kharidhar (Aeluropus flariddum), lapdi (Aristida redaets), soma (Enimochloe colonum), jinko soma (Panicum flaridum), mancho (Dhetyloclemium egyptium), mano (Chlarismonm), dhaman (Cenchrus ciliaris) and zinzvo (Andropogan pumlis) grow immediately after the first monsoon showers (Bhan 1994: 73).

The Harappan site of Kanmer is situated close to the Little Rann of Kachchh in Rapar taluka in the Kachchh district of Gujarat (Fig. 1). The peninsula of Kachchh is largely composed of marine and estuarine sediments and has a variety of rocks ranging from middle Jurassic to recent times. The area is marked by low lying bare rocky hills, sandy plains and vast tracts of naked tidal mud flats, in the northern and eastern parts, popularly known as the Great Rann and the Little Rann. The entire region of Kachchh is drained by seasonal rivulets and the annual precipitation does not exceed more than 300 mm. The monsoon showers are received between June and August. In its eastern neighbourhood the alluvial plains of north Gujarat are drained by the rivers Banas, Saraswati and Rupen and their tributaries, which flow into the Little Rann of Kachchh. The inter-dunal depressions in these plains accumulate rainwater, thus supporting livestock.

In fact out of 555 known Harappan and Chalcolithic settlements in Gujarat as many as 63 are located in Kachchh (Possehl 1999; Seth et al. 2007). Among them, only a few sites, such as Dholavira (IAR 1967-68: 14-16; Bisht 1989a, b, 1991, 1997), Surkotada (Joshi 1990) and Junikuran (Pramanik 2003-2004), have been subjected to horizontal excavation, whereas Desalpar, Pabhumath, Sikarpur were subjected to small scale excavation. Of course, Shikarpur is now being re-excavated by M. S. University of Baroda. These sites include both urban and post urban settlements. Majority of the sites are located in the eastern part of the peninsula, which is believed to be an arm of the sea in the past (Gupta and Pandya 1980). It is held that the accessibility of marine resources, semiprecious stones and arable land were major factors for the expansion of enterprising Harappans in Gujarat. Though researches on the Harappans in Gujarat have unfolded various new facets e.g., elaborate water structures, dock yard (?), diverse craft industries, regional variation in architecture, ceramic and so on, there are still various challenging issues yet to be addressed, like the decline of the Harappans.

Kanmer was discovered by R. S. Bisht and his colleagues of Archaeological Survey of India way back in the eighties of the last century (IAR 1985-86: 15-19). The ancient site is located close to the modern village of Kanmer. A truncated pyramid shaped iselberg next to the site marks the location very well from a distance and the

Kanmer was chosen for excavation in order to understand regional variation in architecture and ceramic assemblage, the major reasons for de-urbanisation and to develop a fuller understanding of the Harappans in Kachchh. Besides, our interest is to understand the role of non-Harappan and non urban or contemporary rural folks

Kanmer (23°23’N; 70°52’E)

157

J. S. Kharakwal, Y. S. Rawat & Toshiki Osada Y 30 and Z 30). Only a very small area of the lower level could be excavated. No structure or floor level has been identified in this deposit. The cultural material is mainly represented by ceramic assemblage which is composed of a variety of Red Ware i.e., Coarse Red, Cream slipped, Red Slipped, Black slipped and un-slipped, besides bichrome. The pottery is mostly prepared on slow wheel, though fragments thrown on fast wheel as well as handmade are also available. Except Coarse Red the fabric of this assemblage varies from fine to medium and most of the sherds are found painted. The Cream Slipped pottery, on the basis of surface treatment, can be further subdivided into two categories fine and coarse. The fine variety is decorated with variety of beautiful black paintings. In the case of the Red Ware with black slip the core of majority of sherds is grey. As none of the sherds of this ware with oxidised core was discovered in the first season, we thought it is perhaps Grey Ware. Majority of potsherds of this ware have been decorated with white bands on the neck or shoulder or middle part of the body on which a variety of red paintings are executed. In case of bichrome ware often the white background has peeled off and the paintings in such case are visible only under wet conditions. Similar decoration pattern has been noticed in the Red Slipped ware. The common paintings in these varieties include fish scale, series of strands arranged horizontally and vertically, bands of wavy lines and chess board pattern. The shapes are represented mainly by bulbous and spherical jars with short and concave neck, bowls with stand in Red slipped Ware and bichrome, with flared sides and incurved rim in cream slip. Some deep bowls/ dishes of Red slipped and Black Slipped Ware are carinated externally. The cream slipped pottery, coarse red ware and bichrome appear to be local or regional pottery types, identical to Anarta assemblage (Ajitprasad and Sonawane 1994). Among these, the bichrome disappears in the subsequent cultural levels. The other wares of this assemblage continue but with some change in each case either in shape or texture or surface treatment. Though small quantity of Mature Harappan artefacts were also found from this level, the characteristic S shaped or perforated jars, terracotta cakes, weights or beads of typical Harappan types were conspicuous by their absence. Therefore, we have identified this deposit as Early Harappan. This assemblage is perhaps similar to stage one and two of Dholavira (Bisht 1989-90:76) and somewhat similar to Surkotada Pd IA (Joshi 1990).

like Anart or Ahar in the urbanisation process and the hinterland trade mechanisms. The Harappan archaeology has been dominated by excavations of large sites, and as a result the role of smaller settlements like Kanmer has not been fully understood. The main mound at Kanmer is small in size measuring 115m (E-W) x105m (N-S) and looks roughly a square from top (Fig. 2). It rises like a small plateau from a horizontal plane up to a height of about eight meter (i.e., 20m above the mean sea level), hence visible from a considerable distance (Kharakwal et al. 2005, 2007). Close examination revealed that barring corners and slopes, the mound was largely intact for systematic excavation. A few stone walls were also visible on the eastern margin of the site, whereas the western and southeastern slopes were littered with huge boulders, perhaps rolled down members of a fortification. The periphery of the mound-top being part of the once high fort wall remained slightly higher than the inner area which shows a slight depression sloping down towards south central direction. On account of such surface features controlled digging was started in 2005-06, which allowed us to identify the following, fivefold cultural sequence: Pd I. Early Harappan (KMR I) Pd II. Mature Harappan (KMR II) Pd III. Late Mature Harappan, (KMR III) Pd IV. Historic (KMR IV) Pd V. Medieval. (KMR V) It is interesting that the site has yielded a long cultural sequence with two short gaps, one between the Late Harappan and Historic and the other between the Historic and Medieval phases. Such long cultural sequence indicates that the location of the site remained very significant right from Harappan period for trade, administrative or for some other reasons. Therefore to understand its function (s) three field seasons of work has been completed at the site and the work may continue for a few more seasons. The find of a massive rampart and remains of several structures and floor levels belonging to different cultural periods, excellent evidence of storage pits, a hoard of steatite beads, a furnace, and ceramic indicating intra site and long distance cultural or trade contacts are some of the interesting discoveries at the site (Kharakwal et al. 2007, Kharakwal et al. 2009). At present attempts are made to understand the nature and history of the rampart, craft areas and sources of various kinds of raw material that was used at the site by the Harappans and subsequent cultures. Results of the last three seasons (2005-08) are summarised in the following pages.

Period II The second cultural period was represented by the Mature Harappan artifacts, fortification and structures. In the small cuttings in the south eastern, north western and north eastern areas of the mound both the inner and as well as outer faces of fortification were found. In the southeastern and northeastern areas the inner faces were found just below the surface whereas in the northwestern area it was found buried under the Historical deposit. A square (five into five meter) mud platform plastered with white shale was found at the south eastern corner, which might have served as the inner bastion. At this corner,

Period I Period I was mainly identified on the basis of about 20 cm thick earliest cultural deposit overlying the sterile layer i.e., brown sandy clay (locally known as lilva) on the bed rock in south central part of the mound (trenches 158

EXCAVATION AT KANMER, A HARAPPAN SITE IN GUJARAT, INDIA: SOME OBSERVATIONS in the northern area towards the end of their settlement at Kanmer. Similar episodes of ash were also found on the northern- eastern slope of the mound and along the outer face and addition. While cutting down in the eastern and northern slopes, such dumping of ash yielded mainly Mature Harappan pottery belonging to Period II B. It appears that after reducing the thickness of the wall they used the available space right upon the top of the older rampart for constant burning of fuel for some purpose yet to be ascertained. The stratigraphy of dumping of ash on either side of the wall suggests that it was constantly thrown from top of the wall.

internally, the southern wall (oriented roughly E-W), exposed up to a length of 12.50 m, was found joining the eastern wall at an obtuse angle. In this area a small Mature Harappan structure joining the fort wall at right angle and a street running along the wall were found underlying a Late Mature Harappan structural complex. The inner faces of eastern and southern wall were made of semi dressed stones. From the Mature Harappan level were found a terracotta sealing, a few terracotta cakes, beads of steatite and agate and a few weights of agate. The sealing has a unicorn (!) and a few Harappan letters. The fortification walls show batter on either faces. The inner face has been exposed near the southeastern and northeastern corner up to a depth of 1.50 m. In the north western area three courses of this face (60 cm height) were identified. Similarly, the outer face has been exposed in the north eastern (up to a length of 14 m), north western (up to a length of 9 m) and in the south eastern corner (up to 17.50 m). A few trenches were also laid on the northern, eastern and southern slopes to find out continuation and exact orientation of the wall. These limited cuttings revealed that there are two parallel faces. It was noticed that the thickness of the outer face or veneer of northern wall gradually increased up to 1.10 m from east to west. However, at the northwestern corner the addition was found missing. At the southeastern corner also the thickness was not uniform as the result the inner face appeared to form an angle of 15 degree with the outer one. This kind of evidence indicates some kind of alteration and change in orientation too. Further excavation will give the correct idea of planning. As both faces, in case of northern wall, have been found resting on the same level, it is tentatively presumed that the outer most face of the rampart may have been erected after a short time gap due to some serious problem with the original one. Was it a problem created by an earthquake? Thus the total width of northern wall has been found about 18 m (at the middle level) and it has survived up to height of about 5m.

In the north eastern corner of the mound, close to the inner face of older wall, was found an impressive bulb shape furnace on plan with central cylindrical column (diameter 31 cm, depth 35 cm) (Fig. 3). This column has opening only from top and was filled with whitish material ashy in appearance. The furnace was ovoid in shape having longer axis in N-S orientation measuring 1.40 m whereas the E-W length was only 96 cm. The clay walls of the furnace were barely 4 cm thick. The area between column and outer clay wall was filled with ash. A small squarish platform made of flat stones and roughly square in shape was found adjacent to and associated with the furnace. The identification of the furnace was done mainly on the basis of its shape and associated burnt clay and other material. It appears that the temperature raised in the furnace may have been more than 500° C as the wall of the cylindrical column, the outer wall and the earth around were burnt red. Besides, the heat generated from the furnace also damaged the face of the fort wall. Several tubular faience beads were found from the dumped ash around the furnace and near the squarish platform. This furnace was also found underlying the dumping of ash along with the inner face of rampart. Similar dumping of ash has also been recorded from the middle town of Dholavira. Joshi (1990) has also recorded large scale burning at Surkotada, which he interprets some kind of destruction in the settlement. In an exploration the authors have noticed similar dumping of ash at Pabhumath, another Harappan site located in the same region. It seems that this kind of ash in various episodes may have been generated due to some craft activity.

In the northeastern corner two successive stages of construction of fortification have been identified. The latest one was found just below the surface and the earlier one at about 1.70 m below and underlying the former. It seems that in order to increase the height of the rampart during the later part of their occupation the Mature Harappans reduced its width by shifting the inner face almost 10 m further north. The outer face of this part of the rampart has eroded therefore we do not know the exact width of this phase. However, taking into consideration the inclination of surviving outer faces underneath, the width of the last phase may not have been more than six meter. This new inner face was in fact erected right on the core of the earlier rampart and has been identified as the third constructional phase of the rampart. The inner face, exposed up to a length of 7.50 m, was found buried under more than one meter thick deposit of ash.

The index trench, located in the sloping south central part of the mound, has yielded a fivefold cultural sequence of the site. As many as four Mature Harappan structures were partially exposed in this trench. Among these, a rectangular room was identified, the northern wall (oriented roughly E-W) of this which exposed up to a length of 3.80 m and the western one (oriented roughly NNW-SSE) up to a length of 3.50 m (see Fig. 4). These walls join at right angle in the SW part of trench Z30. This structure has survived only up to a height of 90 cm with six courses. A small rectangular platform or veranda was attached to this room, which is located along a street roughly oriented east-west. The structure MH3 was found resting on an earlier street, a structure and four different successive floor levels. This clearly suggested that the older street was about a meter to the south of the later

The Harappans dumped huge quantity of ash in various episodes all along the inner and outer faces of the rampart 159

J. S. Kharakwal, Y. S. Rawat & Toshiki Osada basis of absence of the Mature Harappan pottery, weights and planning of settlement. This complex was found resting right on a Mature Harappan street and superimposed by a dark greyish layer, which represents break between the Historic and Late Harappan deposit. As many as three rectangular rooms, located next to each other of this complex, were partially exposed. The first room was found close to the south eastern corner of the rampart, and the other two to the west of the first one. Among these, only one of the rooms (330x2.84 m) was exposed completely. The southern wall of all three rooms was common, which was found parallel to the inner face of the rampart. There was a narrow space between this structural complex and inner face of rampart which measured about 1.50 m. The width of this space increases gradually towards the south eastern corner of the rampart. It was found filled with debris of the wall. In the eastern were found four successive floors, out of which the lowest or earliest was mature Harappan and other three belonging to the Late Mature Harappan phase.

one. This kind of evidence suggests that some change in the planning took place in the subsequent structural phases of the settlement. Inside this room was found a beautiful terracotta sealing (stamped by a squarish Harappan seal) and an inscribed sherd of Red Ware. This sealing is quite interesting as it has a perforation in the center for hanging as a pendent (Fig. 5). And on the reverse were found a few engraved Harappan letters. Underneath the older street deposit was found a small amount of Early Harappan material mixed with the Mature Harappan assemblage in layers 12, 13 and 14. Thus during period II was raised a planned settlement inside the aforesaid strong rampart. The core of this rampart was made of mud and mud bricks and its both exterior as well as interior faces were finally veneered with very large and medium sized semi-dressed and dressed stones. The wall shows a constant battering and thus looks very impressive. Though the southeastern and northwestern corners appear to join at obtuse angles and northeastern and southwestern corners at acute angles, the fort may have been originally squarish. The mound looks squarish from top as the contour map also revealed. The Mature Harppan streets and structures, so far exposed in the central and south eastern area, appear parallel to the southern wall of rampart. It appears that the rampart survived even after the decline of the Harappans and dictated the subsequent planning of the Late Mature Harappan, Historical and Medieval settlements at the site.

To prepare the floors the Late Harappans first leveled the ground by placing yellowish clayey sand (locally known as pila lilva) mixed with fine clay from nearby ponds (locally known as khotda). On top of this was placed less than five centimeter thick bed of brown clayey sand (locally known as lilva), which was properly rammed. And finally a thick plaster of white shale mixed with cow dung was applied. Sometimes the plaster is as thick as 15 cm. This technique of floor making was also identified in the Mature Harappan levels at the site. Of course the Late Harappan floors are thinner in comparison to the Mature Harappan ones. The local villagers informed that the same tradition is continuing in the village even now. There are also instances when the Harappans have laid more than 20 cm thick deposit of brown sandy soil at the base of the floor. Perhaps this kind of deposit prevented rats and other insects from entering the house. As this is a known knowledge in the village, therefore this kind of evidence indicates the continuity of traditional knowledge of floor making technology. On these floors were found parallel sided querns, hammer stones, complete or broken pots of the Gritty Red Ware, and bowl fragments of the white painted Black-and-Red Ware, Buff Ware and fine Red Ware. Apart from these, beads of faience, steatite and shell were interesting discoveries. The lowest or earliest floor yielded a few Mature Harappan pot sherds, two weights of chert and bladelets of agate and a few beads of steatite and faience. Besides, a cooking hearth (U shaped) was found associated with this floor, the diameter of which was found 30 cm. It has survived up to a height of 30 cm.

The Mature Harappan (Pd II a) ceramic assemblage is composed of a variety of Red Ware (Fig. 6). This ware has been treated with red, chocolate and cream slip and a variety of paintings with both the geometrical and nongeometrical ones executed on the outer surface. The Red Slipped ware can be termed as polytone as it is available in various shades of slip e.g., dark brown (generally with chocolate colour), light red with pinkish appearance and sometimes looks like brick red. The dark brown slip is relatively thicker and often has peeled off the surface. The other shades were perhaps prepared by smearing the colour on the external surface. The coarse red ware is often decorated with deep incised designs confined to the neck and shoulder portions. A very small percentage of Buff Ware, and Reserve slipped ware has been discovered. In the Mature Harappan (Pd II b) all the wares of Pd IIa continue, but the quality of surface treatment show some change. In this phase the quantity of Buff Ware suddenly increases, besides new pottery types like Ahar type white painted Black and Red Ware (Misra et al. 1997), Sorath Harappan types and Gritty Red Ware is introduced at the site. The Coarse Red Ware of Pd I is now replaced by the Gritty Red Ware.

It seems that the Late Mature Harappans initially used the existing rampart as a defense wall, but subsequently this function ceased for some reason yet unknown to us. Towards the end of their occupation they erected their structures right on top of the rampart or sometime robbed the veneer stones of the wall for making their houses. A small bulbous pot (broken) of Red Ware with narrow opening and flared mouth was found full of beads of steatite and faience on the floor of a house located right

Period III Period III is represented by the Late Mature Harappan deposit at Kanmer. A large structural complex of this period was partially exposed in south eastern area. The identification of this complex was done mainly on the 160

EXCAVATION AT KANMER, A HARAPPAN SITE IN GUJARAT, INDIA: SOME OBSERVATIONS sand and ash at the base. Some such pits also yielded ritualistic pots. Some of them yielded charred grains too. Therefore it appears that they might have been used for storing cereals. A few small stone structures are also visible at several places on top of the rampart, which seems to be built during this period. The thickness of medieval settlement doesn't exceed more than 40 cm anywhere in the mound, which suggests a short period occupation. The ceramic assemblage of this period was mainly represented by the Red and Grey Ware. Sometimes the Red ware, which also dominates the assemblage, is decorated with red paintings on the outer surface.

on top of rampart (in trench GG 30 and GG 31). It was buried barely 15 cm below the present surface. The total number of beads counted was 11707 out of which 27 were faience ones. Many pottery types of the Mature phase e.g., the Red Slipped Ware, the Buff Ware, the Cream Slipped Ware, the Reserve Slipped ware though continued in the late phase but in general the quality of their surface treatment had declined, design pattern and shapes had changed. For example, the fine quality of the Red Ware with red slip has turned ocherous type towards the end of the occupation and it outnumbers all other types. Geometrical designs were preferred on the pottery. The Anart type coarse Red Ware was absent from this deposit. However, the Ahar type Black-and-Red Ware the Gritty Red Ware continued in this deposit with other major types. A new pottery type was introduced during this phase i.e., made of white shale. Only a few jar fragments of this ware have been found in our limited digging. It has been treated with pink slip or designs. Very often the surface treatment of this entire Late Mature Harappan assemblage reminds the Early Harappan pottery tradition of the site.

Minor Antiquities Among the noticeable antiquities are the large number of beads of steatite, faience and agate. Terracotta cakes were found distributed throughout the deposit. They are smaller in size in the Mature Harappan deposit, whereas their size increased in the late levels. Other small finds include beads of gold, terracotta bangles, long tubular beads and shall objects. Beads

Period IV

A variety of beads of gold, semi precious stones like lapis lazuli, agate, carnelian and blood stone, paste, faience, steatite, terracotta, shell and bone have been discovered from the Mature and Late Harappan levels. In the Historical and medieval levels, however, the majority of beads are of terracotta, semi precious stones and shell. Occasionally gold and ivory have also been used.

This cultural period was represented by remains of the Historical period. One large structural complex of this period has been noticed so far in the north western part of the mound (Kharakwal et al. 2007). The thickness of historical period habitation deposit is hardly one meter in the northern and eastern parts of the mound whereas in the central part it is more than two meters thick. Perhaps the settlement was laid on an earlier steep slope. This cultural period is identified on the basis of discovery of a terracotta sealing with Brahmi characters, Roman amphorae, the Red Polished Ware, the Rangmahal type Red Ware, a variety of beads of semiprecious stones and iron objects. The discovery of such ceramic assemblage suggests that during the early centuries of Christian era i.e., Historical times Kanmer again emerged as an active center of trade. It may surely be connected to sites like Vadnagar in Mehsana district in the north Gujarat and several contemporary coastal sites such as Dwarka. In fact an identical ceramic assemblage has been discovered from such sites. The site of Vadnagar, currently under excavation by Y. S. Rawat of State Department of Archaeology of Gujarat, was an important city during the Historical times and the entire ceramic assemblage of period IV at Kanmer is identical to an early phase of that site.

The majority of Harappan beads are disk type, bicone and tubular. Beads of steatite outnumber all others as more than 15000 have been discovered so far. They have been recorded as micro-beads, mini disk type, disk type, cylindrical or tubular, segmented, wafer type and triangular ones. Except for one hoard from a house of the Late level most of them have been found from streets. Some of the tubular beads have been decorated with circular marks. In the Late Harappan phase the quantity of tubular beads decreases and sometimes they have been coated with green or blue colour, perhaps to look like faience. Faience beads are disk type, short tubular or tubular and blackish, greyish and greenish in appearance. The blackish or greyish examples appear to be the result of ill or over firing. Most of the tubular ones are twothree centimeter in length and have been discovered from north eastern area. It is likely that they were manufactured in the northern area of the mound, which strongly appears to be used for craft activities. The paste beads though poor in quantity compared to steatite, their shapes and appearance is quite attractive. Most of them are either black or red or white and globular, disk type or cylindrical in shape.

Period V This last cultural phase at the site is represented by a thin cultural deposit of layer three in the central area. A few large fire places have been found in this level. However, in south eastern and north western area of the mound over a dozen circular storage pits (silos) have been found. The walls of these pits were made of fine clay obtained from some nearby lakes. Most of them had alternate layers of

The beads of semi precious stone have been identified as carnelian, agate, lapis lazuli, chalcedony, serpentine and bloodstone. The site yielded raw material of agate at the site besides, chipped, rough outs, grinded, unpolished 161

J. S. Kharakwal, Y. S. Rawat & Toshiki Osada The Harappan Phase

bead blanks. All these findings indicate that these beads were manufactured at the site. Perhaps the raw material was brought from Mardak Bet located about 20 km east of Kanmer in the Little Rann. Except for lapis lazuli, source of raw material for all these beads types could have been the Little Rann and its adjacent areas. Lapis may have been brought either from Afghanistan. The beads of carnelian are disk type, globular, bi-cone and tubular and etched ones. These beads were drilled after polishing as the find of many broken but polished examples would indicate. In the Late phase the quality of surface treatment of such beads was inferior. A few examples of shell, bone and metal ones have also been found from the Mature and Late levels.

In the Late Mature Harappan phase domestic animals predominate like in the other cultural phases at the site. Among the domestic animals the cattle and buffalo constitute a majority and many of them bear cut marks or are charred suggesting that they were consumed. The wild animals represented in the Late phase include a large bovid, wild pig, antelopes, deer, carnivores and small mammals like hare and rodents (9.60%). In the Mature Harappan phase 96% faunal remains belong to cattle and buffalo (81.09%), followed by sheep and goat (13.93%). Besides these, wild animals like nilgai (blue bull), fourhorned antelope, a suidae species and the porcupine were identified.

The discovery of a large number of drill bits from Kanmer together with raw material and unfinished beads clearly suggests that one of the important functions of the site was bead manufacturing.

Thus as many as twenty-four species were identified, out of which the domestic mammals were represented by seven species (cattle, buffalo, sheep, goat, pig, dog and cat) wild mammals were represented by 14 species (Nilgai, wild pig, antelopes (Blackbuck, Chinkara and Four-horned antelope), deer (Sambar, Chital, and Mouse deer), a felid species, porcupine, hare and rodents (house rat and desert rat). Besides these, a few birds, reptiles, fish and shells were also identified.

Faunal Remains The faunal analysis was initiated by P. P. Joglekar of Deccan College, Pune and later Pankaj Goyal also joined him. Though faunal remains were recovered from almost all the trenches, they were not evenly distributed. In most of the trenches bones were encrusted with soil, but they were able to identify cut marks on bones and charred bones in some cases. In areas like the northern and southern slopes of the mound, in the ash dumps (from layer one to six), few or no bones were found. They have identified several animal taxa, which include mammals, birds, fish, reptiles and molluscan species. Among the domestic animals, cattle, buffalo, sheep, goat and pig were identified. More than a dozen wild animals were identified in the collection, including the nilgai, antelopes, deer, carnivores and rodents.

Evidence of charring, butchering and cut marks has been found on a large number of bones which perhaps indicate that these animals may have been part of their diet. The relative proportion of charred bones was larger in the Mature Harappan phase compared to the Late Harappan and Historic. Some of the bones of cattle/buffalo and sheep/goat were fire hardened and it is likely that they may have been used as tools. Goyal and Joglekar (in press) conclude that “domestic animals were the most important resource to the animalbased subsistence throughout the occupational history. The higher percentage of domestic animals like cattle, buffalo, sheep and goat may suggest that stock raising and pastoralism was an important component in the subsistence activity of the community in all the periods”. They further point out that the proportion of wild mammals used in the late phase of the Harappans was more than any other cultural phase, clearly indicating an increase on the dependence on the wild mammals. Of course one has to keep in mind that in all our trenches we have exposed Late Mature Harappan deposit more than any other cultural phase. Eventually the faunal assemblage of this phase outnumbers all other phases. Therefore, this conclusion, as the authors have pointed out, is tentative.

Historical Phase In the Historic phase seven domestic mammals have been found- cattle, buffalo, goat, sheep, pig, dog, and the cat. Buffalos were consumed during this phase as revealed by the presence of two completely charred bones of this animal. Bones of wild-mammals include antelopes (Gazella bennetti, Antilope cervicapra and Tetracerus quadricornis), deer (Axis axis) and hare from this phase. Besides, bones of birds and reptiles were also identified. Break A dark greyish soil layer (two) in the south eastern and north eastern area of the mound and layer six in the central part represented break between the Historic and the Harappan phase. As many as eleven species have been identified from this phase, which include cattle, buffalo, sheep, goat, pig, dog, horse and wild-mammals like Sambhar, house rat and hare and domestic fowl. It is likely that some of this faunal material may have percolated down from the sealing layer (one). On the other hand the possibility of mixing up of this material with the Late Harappan deposit also exists.

Palaeobotanical Investigations In order to understand subsistence of the Harappans, Anil K. Pokharia (Birbal Sahni Institute of Palaeobotany, Lucknow, India) retrieved botanical remains by water floatation technique. He was assisted by several research students of the Rajasthan Vidyapeeth at the site. For wet floatation water was brought from a distance by bullock 162

EXCAVATION AT KANMER, A HARAPPAN SITE IN GUJARAT, INDIA: SOME OBSERVATIONS this phase has been recovered only from the central part of the mound so far. The Mature Harappan deposit is changed as Period II, which is further sub-divided in two parts Pd IIa and Pd IIb according to appearance of new ceramic at the site. It seems during this period a planned settlement was built secured with a massive and strong fortification. Did they store something precious inside? Towards the end of the Mature Harappan phase the hight of rampart was further raised but its width was reduced. The Late Harappan buildings have been found overlying the Mature Harappan remains. After the decline of the Late Harappans the site got covered with a thick layer of dark grey soil. It is often loose and sandy in nature. The long spell of abandonment after the historical period again got deposition of fine silt mixed with clay indicating that aeolian soil getting deposited in a large depression and later on worked by rain water. Our archaeological diggings and the results of digital recording indicates that the rampart is squarish on plan and not oriented to the cardinal direction. In fact its N-S wall is 30 degree off the magnetic north. The total thickness of the wall was measured about 18 m in the middle level of its height. It is likely that the Harappans did not maintained the fort wall towards the end as they finally they raised their houses right on top of the rampart.

cart as there is no source of water at the site. Initially quantitative control on sampling was attempted, but gradually it was realised that some strategy is required for varied context. For example from floors or general habitation levels ten ghamelas (about 50 kg) soil was sampled for floatation. Ghamela is a deep basin kind of metal utensil generally used for transporting or throwing soil. From hearth/ fire places were sampled two ghamelas and from burnt patches five ghamelas. One ghamela contains about 5 kg of soil. Soil samples for wet flotation were selected from trenches S 21 (in the northwestern area) and GG 29, 30, 31, HH 30 (in the southeastern area), BB 11 (north eastern corner), II 20 (eastern slope). Each sample was poured into the body of water in a tub that was agitated so that light material is buoyed to the surface and skimmed off through a 25 mesh sieve. Floatation allowed recovery of all size and classes of botanical material preserved in the sediment, making qualitative and quantitative analysis possible. A large amount of carbonised material was gleaned from well stratified trenches. A sizable amount of botanical material was found in utterly fragile, highly burnt and mutilated state of preservation. The inferences of the study, therefore, are based on a small fraction of the material. The remains were sorted out under the stereo-binocular microscope and thereafter cleaned in acid-alcohol (glacial acetic acid 10% + ethyl alcohol 50% in equal volume). The identification is based on the morphological details preserved in the carbonised grains and seeds by comparing them with the corresponding parts of extant plants of the same species. These remains represent the accidentally scorched portion of a larger amount.

Outside the rampart area no other settlement like a lower town could be located. Of course there is a very thin deposit of Late phase to the east of the main mound. This eature also makes the site unique. From the Harappan deposit considerable quantity of nodules of agate, perhaps brought from Mardhak Bet, located about 20 km to the northeast of Kanmer have been found. It is likely that this raw material was used for making variety of beads, blades and weights. The faunal and floral remains clearly indicate that agriculture and cattle breeding were among the important occupation at the site. It is likely that the Harappans were exporting finished beads as well as raw material to other sites from Kanmer.

In the first field season several cereals and pulses such as barley (Hordeum vulgare), bread-wheat (Triticum aestivum), dwarf-wheat (Triticum sphaerococcum), rice (Oryza sativa), field-pea (Pisum arvense), grass-pea (Lathyrus sativus), and green-gram (Vigna radiata) were identified besides cotton (Gossypium arboretum / herbaceum), some weeds and wild taxa (see Kharakwal et al. 2007). From limited samples the remains of jowarmillet (Sorghum bicolor), pearl-millet (Pennisetum typhoides), sesame (Sesamum indicum), and horse-gram (Macrotyloma uniflorum) were among the new finds. Remains of weeds and other wild taxa were also encountered reflecting the ecological conditions and ground vegetation. Details concerning the plant remains with their archaeological provenience are summarised in the following Table 1.

Acknowledgements The entire team of Kanmer Archaeological Research Project (KARP) sincerely thanks the Archaeological Survey of India for giving us permission to continue our studies at the site. We thank various authorities of the Government of Gujarat and JRN Rajasthan Vidyapeeth, Udaipur for encouraging and helping us while conducting field research in Kachchh. The authors would also like to express sincere gratitude to Prof. D. P. Agrawal, Dr. R. S. Bisht, Prof. B. S. Garg, Prof. Lokesh Bhatt, Sri V. N. Maira, Dr. R. S. Fonia, for their guidance and Dr Piyush Bhatt, Sri V. K. Uppal and M. S. Rawat for helping at various stages. We also thank Mr. Lal Chand Patel, Mr. Hansmukh Seth, Mr. Rajesh Meena, Mr. Suresh Meena, Mr. Asif Hussain, Mr. Krishna Pal Singh, Mr. Sameer Vyas, Mr. Bhim Raj Varhat, Dr. Manikant Shah, Drs. Shahida Ansari, Alok Kanungo, Miss Amrita Sarcar, Miss Anjana Reddy and Mr. Pankaj Goyal, Mr. Sanjeev Kataria, Dr. Teramura, Drs. Akinori Uesugi, Kondo, Miahara for participating in the excavation. Our

Conclusion Majority of work in the first two field seasons was concentrated on the rampart. On the available evidences three distinct phases of the construction of this magnificent structure at Kanmer have been identified. The site has yielded a fivefold cultural sequence beginning with the Early Harappan and until the decline of the Harappans. The Early Harappan settlement was raised on a thin soil cover on the bed rock. The deposit of 163

J. S. Kharakwal, Y. S. Rawat & Toshiki Osada Joshi, J. P. (1972) ‘Exploration in Kutch and excavation at Surkotada and new light on Harappan migration’. Journal of the Oriental Institute 22 (12): 137-143. Joshi, J. P. (1990) ‘Excavation at Surkotada 1971-72 and Exploration in Kutch’. New Delhi: Archaeological Survey of India. Kharakwal, J. (2005) ‘The Indus Civilization: an overview’. In Osada, T. (ed.), Occasional Paper, Kyoto: Research Institute for Humanity and Nature: 41-85. Kharakwal J. S., Rawat Y. S., and Toshiki Osada (2005) ‘Harappan sites in Kachchh and new opportunities of tourism, in Heritage Tourism’. Government of Gujarat: 35-43. Kharakwal J. S., Rawat Y. S., and Toshiki Osada (2007) ‘Kanmer: A Harappan sites in Kachchh, Gujarat, India’. In Osada, T. (ed.), Occasional Paper 2, Kyoto: Research Institute for Humanity and Nature: 21-46. Kharakwal, J. S., Y. S. Rawat and T. Osada (2009) ʻExcavations at Kanmer: A Harappan Site in Kachchh, Gujaratʼ. Puratattva, 39: 147-164. Misra, V. N., Shinde V. S., Mohanty R. K., Lalit Pandey and Jeewan S. Kharakwal. (1997) ‘Excavations at Balathal, District Udaipur, Rajasthan (1995-97), with special reference to Chalcolithic Architecture’. Man and Environment, 22(2): 3559. Possehl, G. L. (1992) ‘The Harappan Civilization in Gujarat: the Sorath and the Sindhi Harappans’. The Eastern Anthropologist, 45 (1-2): 117-154. Possehl, G. L. (1999) The Indus Age. Delhi. Pramanik, Subhra (2003-2004) ‘Excavation at Juni Kuran: 2003-2004 a preliminary report’. Puratattva, 34: 45-67. Seth, H., Patel L.C. , and Varhat B. (2007) ‘Harappan sites in Gujarat’. In Osada, T. (ed.), Occasional Paper 2, Kyoto: Research Institute for Humanity and Nature: 111-126.

understanding about the site would have remained very superficial, had scientists like Dr. P. P. Joglekar, Mr. Pankaj Goyal, Dr. Pokharia and Prof. T. Uno and his team not joined us in the field and contributed. We therefore record special thanks to them.

Bibliographical References Agrawal, D. P. (2007). Indus Civilization. Delhi: Aryan Books International Agrawal, D. P. and J. S. Kharakwal (2003) Bronze and Iron Ages in South Asia. Delhi: Aryan Books International. Ajitprasad, P. and V. H. Sonawane (1994).ʻThe Harappa Culture in North Gujarat: a Regional Perspectiveʼ. In Possehl, Gregory L. and Vasant Shinde (eds) Harappans and Others in Gujarat. Delhi: OxfordIBH: 129-139. Bhan, K. K. (1994) ‘Cultural development of the prehistoric period in North Gujarat with reference to Western India’. South Asian Studies, 10: 71-90. Bisht, R. S. (1991) ‘Dholavira: A new horizon of the Indus civilization’. Puratattva, 20: 71-82. Bisht, R. S. (1989a) ‘A new Model of the Harappan town planning as revealed at Dholavira in Kutch: A surface study of its plan and architecture’. In Bhaskar Chatterjee (ed.), History and Archaeology (Prof. H.D. Sankalia felicitation volume), Delhi: Ramanand Vidya Bhawan: 397-408. Bisht, R. S. (1989b) ‘The Harappan colonisation of Kutch: an ergonomic study with reference to Dholavira and Surkotda’. In Krishan Deva and Lallanji Gopal (eds.), History and Art. Delhi: Ramanand Vidya Bhawan: 265-272. Bisht, R. S. (1997) ‘Dholavira excavations: 1990-1994’. In Joshi, J.P. (ed.), Facets of Indian Civilization: Recent Perspectives, New Delhi: Aryan Books International:107-120. Goyal, Pankaj and Joglekar, P. P. (in press) ‘Report on the faunal remains recovered from Kanmer, Gujarat, during the second field season (200607)’. In Osada, T. (ed.), Occasional Paper 4. Kyoto: Research Institute for Humanity and Nature. Gupta, S. K. and Pandya, S. (1980) ‘Kutch Harappanstheir communication routes- a discussion’. Vidya 23 (1): 39-46. IAR. Indian Archaeology - A Review. Annual Publication of the Archaeological Survey of India, New Delhi. Joglekar, P. P. (2007) ‘Report of the faunal remains recovered from Kanmer, Kachchh, Gujarat, during the first season (2005-06)’. In Osada, T. (ed.), Occasional Paper 2, Kyoto: Research Institute for Humanity and Nature: 47-76.

164

EXCAVATION AT KANMER, A HARAPPAN SITE IN GUJARAT, INDIA: SOME OBSERVATIONS

Fig. 1 – Location map of Kanmer.

Fig. 2 – Contour map of Kanmer.

165

J. S. Kharakwal, Y. S. Rawat & Toshiki Osada

Fig. 3 – Furnace with central column.

Fig. 4 – Mature Harappan structure and street in the central part of the mound.

166

EXCAVATION AT KANMER, A HARAPPAN SITE IN GUJARAT, INDIA: SOME OBSERVATIONS

Fig. 5 – Seal impression from Kanmer.

Fig. 6 – Mature Harappan jars and bowls.

167

TRANSPORT AT ALTYN-DEPE IN THE 3RD MILLENNIUM BC (THE EVIDENCE OF VEHICLE MODELS) Lyubov’ B. Kircho

materials – lapis lazuli, alabaster and metals. Via the mediation of populations of the Central-Asian and Iranian regions these materials were exported to various centres of the Near and Middle East (Amiet 1986). Secondly, during the epoch of the formation of the early urban centres in Central Asia, the rising subculture must have had certain uniformity (with participation of the Geoksyur and Mundigak cultural traditions). Along with connections between the peoples of southern Turkmenistan, south-eastern Iran and southern Afghanistan, that unity is reflected, as it seems, in a similarity of the house-building traditions, collective sequential burials in mudbrick chambers and close parallels between the compositions of the grave offerings and types of seals.

In the history of the development of the early farming culture in southern Turkmenistan of the 5 th - 3rd millennium BC, three phases of active interactions with sites and cultures of Iran, Afghanistan, northern Baluchistan, the valley of the Zeravshan River and even some remoter regions of the Ancient East have been distinguished (Kircho 2008). The first phase (middle and second half of the 4th millennium BC) is marked by innovations in the complexes of the Anau culture (Namazga II period) suggesting that three interrelated processes were taking place there. These processes, conditioned probably by the Mesopotamian colonization of the Late Uruk period (Lamberg-Karlovsky 1985: 56–58) were: 1) an indirect cultural impact of the most advanced regions of Asia Minor; 2) beginning of the long-distance trade, mainly in lapis lazuli; and 3) intensification of the contacts and folk movements of the populations of central and, particularly, south-western Iran where the proto-Elam civilization was rising. With the latter process, establishing of the Geoksyur complex was probably related. In turn, some groups of the population of southern Turkmenistan were moving to the east and north-east: a number of sites with pottery and a figurine of the Late Namazga II type have been recorded in the northern area of the ancient Murghab delta (Masimov 1979: 75, Fig. 2) whereas the earlier complexes of Sarazm in the upper reaches of the Zeravshan River suggest that this settlement of ancient metallurgists was founded with the direct participation of south-Turkmenian communities (Kircho 2007).

The third period of interactions is dated to the second half of the 3rd millennium BC when large settlements of the early urban type arose in southern Turkmenistan. The evidence from Altyn-Depe of the end of the Early Bronze Age and Middle Bronze Age (Late Namazga IV – Namazga V periods) presents numerous indications of contacts both with the main centres of the ancient oriental civilizations (the Indus and Mesopotamo-Elamite circle) and with all the surrounding cultures and raw-material regions in general (Kircho 2005a: 518–519; Masson 1988: 97-110). These contacts probably reflected the introduction of the population of South Turkmenistan into the system of international trade during the peak of the Harappan civilization. Exactly the active interactions with the ancient oriental centres had in the late 3rd millennium BC induced along with a high technological level also the rise of a proto-state of the ancient eastern type with its capital at Gonur-Depe in Margiana.

During the second stage (late 4th – first centuries of the 3rd millennium BC) in the objects of cult or prestige, it is possible to trace the impact of the culture of Mesopotamia and Elam (the iconography of a number of anthropomorphic statuettes, alabaster vessels and figurines, and the imprint of a cylindrical seal of the Jemdet-Nasr period). At the settlements of the central section of the piedmont zone of the Kopet Dagh, the complex of painted ware with zoomorphic motifs of the Kara-Depe style was forming under the influence of the pottery of central and northern Iran. Simultaneously, folks from the south-eastern Caspian area with their grey ware moved as far as the Ashgabat oasis. At the same time, the population of south-eastern Turkmenistan with the culture of the Geoksyur type infiltrated south-eastern Iran (at Shahr-i-Sokhta) and had extensive contacts with the people of Mundigak in southern Afghanistan. Ample evidence on the interactions of southern Turkmenistan of the period suggests evidently two major processes. Firstly, those were the contacts and folk movements that resulted of the proto-Elam expansion and were related with mining and trading in valuable mineral raw

Among the most important problems in studies of the cultural interactions is that of the transport since thesystematic regular contacts, particularly trade, would have been impossible without any effective transporting means. The first evidence of the appearance of wheeled transport in southern Turkmenistan is dated to the second half of the 4th millennium BC. In the cultural layers of early agricultural sites of the Middle Eneolithic period (Namazga II period), terracotta objects have been revealed which by a number of scholars are considered as wheel models. These have been found at Ak-Depe (Kircho 1999: Fig. 22, 1), the Northern Hill of Anau, Kara-Depe, in the layers of the Yalangach and Early Geoksyur period at Ilgynly-Depe and sites of the Geoksyur oasis (Aina-Depe, Akcha-Depe, Geoksyur 1, Geoksyur 9) (Fig. 1, 19-26). They were uncovered, in addition, in the Late Eneolithic layers at Kara-Depe, 169

Lyubov’ B. Kircho Altyn-Depe and Chong-Depe (Fig. 1, 15–18). These objects are round articles 7–10 cm in diameter. In their centre there is a large hole (the ratio of the diameters of the objects and holes amounts to about 1 : 5 – 1 : 3) and a single-sided protrusion by which possibly the one-sided boss of the wheel was rendered. A number of wheel models show traces of the protrusions on the two sides (Fig. 1, 22, 26). It is probable however that these earliest wheel models reflect the appearance of the most ancient disc wheels fixed rigidly on the axle and revolving together with it (Kozhin 1986: 186). The models dated to the Namazga II period have fairly strong difference from the younger clay and terracotta models of wheels common in the layers of the Late Eneolithic period (period of Namazga III: c. 3100–2800/2700 BC) at AltynDepe (horizons 9–13) and Kara-Depe (horizon 1A) (Fig. 1, 9–14). This difference is both in construction and in dimensions. Wheel models of the Namazga III period have a two-sided boss in the centre, their diameters usually do not exceed 4–7 cm, whereas the ratio between their diameters and the diameters of the holes is 1: 8 – 1: 10. Probably these ratios corresponded already to those of the real wheels. The large dimensions of the articles of the Namazga II period and their wide central holes suggest that the vehicle models themselves during that period were fairly large and possibly made of wood.

13) and stone tools. Noteworthy is that in the same partition there was a miniature hearth on a square adobe base suggesting that this part of the house as well as the complex of the clay objects found in it were used for ritual purposes. In addition, models of one-axle carts with a single pole or two ducts for the shafts are represented at Altyn-Depe among the finds of the Middle Bronze Age (Fig. 2, 5, 6). As carting and ploughing animals, it seems, exclusively bulls were used in southern Turkmenistan in the 4th – first half of the 3rd millennium BC. At least only bull figurines of that period from Kara-Depe show traces of fixation of the harness. These traces are transversal holes in the withers or hypothetical representation of harness in the form of strips of paint on the muzzle and back of the animal (Fig. 2, 18, 19). Bull figurines with holes or deep dimples in the withers are found at Altyn-Depe and with representation of harness at Ulug-Depe also during the second half of the 3rd millennium BC (Fig. 2, 2–4, 7, 8, 20). Some of them have a clearly expressed hump (Fig. 2, 7) rendering possibly the appearance of some zebu-like cattle. Nevertheless, no osteological evidence that there were humped bulls-zebu at the sites of the Eneolithic and Bronze Age periods in southern Turkmenistan has been found (Kasparov 2006: 57). One of the figurines of the early Namazga V period with a hole in the upper part of the hump represented probably a camel (Fig. 2, 1).

Among the series of finds from Altyn-Depe of the Late Eneolithic period, along with numerous wheel models there are clay and terracotta models of complete carts (Fig. 2, 14–17). These are fairly small (occasionally even miniature) oval or nearly rectangular objects with a weighty basket, low sides, one transversal hole for the axle and a deep longitudinally slanted duct for inserting the draught pole. In other words, we are dealing with models of one-axle double-wheeled carts with a single pole slanted upwards. Fastened to the pole was probably the yoke put on the draught animals. Thus the team of the animals differed in no way from the paired ploughing team represented on the famous staff-head from Tepe Hissar (Schmidt, E.F. 1937: Pl XLVIII, H 4885). Some of the carts were probably roofed. On a number of the cart models there are upright ducts at the edges. The rods inserted into these holes probably served as the supports of the roof or as the base for the high (possibly wattled) sides of the basket (Fig. 2, 14, 14A). The models of oneaxle carts from Altyn-Depe have parallels at some sites of the Indus Valley civilization — in Chanhudaro and Lothal (Childe 1951: Pl. IX, b, c; Rao 1985: Pl. CCXXI, B). However, those parallels are dated from the second half of the 3rd millennium BC.

Breeding of two-humped camels in the region under consideration in the 5th – 4th millennium BC is attested by separate bones among the osteological collections from the Northern Hill of Anau and Chong-Depe (Ermolova 1985: 86; Kasparov 2006: 62, 63). As early as the second half of the 3rd millennium BC, domesticated camels were already used as draught animals. The most ancient among the reliably documented finds are the small heads of a draught camel found in the 4th horizon of Altyn-Depe (ca. 2400 BC, Figs. 3, 8) and the painted one at TaichanakDepe (Figs. 3, 10). Similar heads of camels, or occasionally bulls, with a dowel for inserting into the front side of a cart model are widely represented in assemblages of the late Namazga IV and Namazga V periods both at Altyn-Depe (Figs. 3, 1–9) and at most of the excavated sites in southern Turkmenistan (Kuz'mina, Lyapin 1980: Fig. 1, 4; Lisitsyna 1978: Fig. 6, 2; Masimov 1976: Fig. 16, 11; Masson 1988: Pl. XII, 10– 12; XXX, 2–4; Shchetenko 1968: Fig. 11, 2). Almost all of the heads have through holes in the forepart of the muzzle for threading the harness. In other words the system of controlling the animals had been changed. Moreover, at Altyn-Depe figurines of two-humped camels have been found in the levels of the Middle Bronze Age along with the heads mentioned above (Fig. 3, 11–13). Notable evidence on the importance of camel in the culture of southern Turkmenistan of the late 3 rd – early 2nd millennium BC has been obtained at GonurDepe. Here, in the elite “royal burials” were found gold and silver vessels and a pin with representations of camels (Sarianidi 2005: 193, 210, Fig. 56, 95–97). The complete skeleton of a camel has been uncovered beside

Single-axle cart models continued in South Turkmenistan throughout the Early and Middle Bronze Age (Namazga IV period: ca. 2750–2350 BC, and Namazga V: about 2350–1900 BC). In the 4th building horizon of AltynDepe (late Namazga IV: ca. 2400 BC), in a small partition at the end of the corridor of a three-roomed house was found a clay model of a one-axle cart with two longitudinal ducts (probably for insertion of shafts) (Fig. 2, 9, 9A). Uncovered nearby were a figurine of a draught bull, a wheel and some other clay objects (Fig. 2, 8, 10170

TRANSPORT AT ALTYN-DEPE IN THE 3RD MILLENNIUM BC (THE EVIDENCE OF VEHICLE MODELS) the remains of a cart in the “royal” tomb no. 3200 (Dubova 2004: 267, Fig. 21, 22).

centuries of the 2nd millennium BC (Dubova 2004: 276– 279, Fig. 35–37, 40–42).

In the end of the Early Bronze Age (about 2400 BC) when Altyn-Depe was at its economical peak, new constructions of cart models appeared. Of the last centuries of the 3rd millennium BC particularly characteristic are two-axle four-wheeled wagons (Fig. 3, 15–17, 21). These include e.g. a wagon model in the form of a flat platform with traces of fixation of two animal heads (Fig. 3, 15). We are dealing here evidently with a representation of a real wagon of a considerable carrying capacity. The most widespread were four-wheeled highsided wagons drawn by a single animal (Fig. 3, 16) (Ganyalin 1959: 34, table V; Lisitsyna 1978: fig 6, 4. 5; Masson 1988: Pl. XIX, 1; XLIII, 1, 2; Schmidt, H. 1908: 172, Fig. 419, Pl. 47, 11; Udeumuradov 1993: Fig. 22, 6). Sometimes the models of two-axle wagons have two ducts for the shafts (Fig. 3, 21). In addition, a miniature model of a roofed one-axle cart has been found (Fig. 3, 14).

The functions of vehicle models and wheeled zoomorphic objects are still arguable. A find of the remains of a twowheeled cart model in the grave of a child 9–10 years old at Altyn-Depe (Kircho 2005: 417; Masson 1988: Pl. XLIII, 3, 4) seemingly suggests their use as toys. At the same time, plentiful evidence from various regions of the ancient world attests to the prestigious and symbolical value of vehicle models in funerary rites (Masson 1976: 165, 166) so that the ritual meaning of the models is quite possible. In any case, the morphology of these objects reflects more or less exactly the features of some transporting means which really existed. Analysis of the quantitative distribution of the models of wheels or vehicles throughout the area excavated in different stratigraphic horizons of Altyn-Depe shows that the number of such finds is the greatest in the layers of the end of the Late Eneolithic period (horizons 10 and 9 of Altyn-Depe). For the Early Bronze Age this coefficient decreases almost 4 times, while for the late Namazga IV and Namazga V period again it rises three times.

In the same period, a new technique of putting through the axles of the vehicles appeared. On the cart models of the Late Eneolithic period and numerous models of oneor two-axle carts of the Bronze Age, the axles were put through a hole in the bottom of the basket. During the Late Namazga IV period, terracotta models appeared with separately modelled loops (oval or nearly rectangular) attached to the lower body of the basket by means of special dowels-protrusions (Fig. 3, 18–20). The axles on which the wheels were revolving were now put through such loops. In the opinion of Elena E. Kuz'mina, that technique of attaching axles “does not repeat the construction features of real vehicles, but rather imitates two ways of attaching axles to metal models” from the Near East (Kuz'mina, Lyapin 1980: 40, 41).

Thus the appearance of the two basic types (one-axle and two-axle) of vehicle models at Altyn-Depe and thedistribution of their finds generally correspond to the greatest intensity of contacts of the inhabitants of southern Turkmenistan with the contiguous regions in the late 4th and first centuries of the 3rd millennium BC as well as in the second half of the latter millennium. The relative rareness of finds of vehicle models dated to the beginning and middle of the Early Bronze Age (about 2750–2450 BC) possibly is a reflection of some transport crisis in the second quarter of the 3rd millennium BC. During that period, under the conditions of the growing aridity of the climate the use of vehicles drawn by bulls may have become impossible as long-distance transport. It seems that exactly the aridization came to be the reason for domestication of camel – the main transporting animal in Central Asia since the second half of the 3rd millennium BC.

Moreover, by means of separately modelled loops, the axles, on which the wheels were revolving, were attached not only to the vehicle models but also to animal figurines, zoomorphic vessels or zoomorphic vehicles. The wheeled figurines and vessels with zoomorphic details had appeared about 2500 BC (Fig. 3, 24, 25). However, the most widespread at Altyn-Depe these wheeled vessels and vehicles in the form of animals, as well as vessels with zoomorphic spouts, became during the Middle Bronze Age (Fig. 3, 22, 23, 26–28) (Masson 1970: Fig. 12, 1). It is noteworthy that nearly all such objects show exclusively representations of either camels (Masson 1988: Pl. XX, 4) or bulls (Fig. 3, 23, 26, 27).

Bibliographical References Amiet, P. (1986) ‘L’âge des échanges inter-iraniens 3500–1700 avant J.-C’. Notes et Documents des Musée de France 11. Paris. Childe, V. G. (1951) ‘The First Waggons and Carts – from the Tigris to the Severn’. Proceedings of the Prehistoric Society XVII: 177–194. Dubova, N. A. (2004) ‘Mogil'nik i tsarskiy nekropol' na beregakh Bol'shogo Basseyna Severnogo Gonura [Graveyard and royal necropolis on the Large pool’s banks of North Gonur]’. In Kosarev M. F., Kozhin P. M., Dubova N. A. (eds) U istokov tsivilizatsii [Near the Sources of Civilizations], Moscow: 254–281.

In younger assemblages from Altyn-Depe, wheel models with relief details on the planes are found. These probably represent composite or light-weight wooden wheels with slots (Fig. 1, 1, 2, 5, 6). Composite cart wheels are known by finds from burials of the Early Dynastic period in Ur and Susa (Littauer, Crouwel 1979: 15–36, fig 3, 5, 8). At Gonur-Depe, wagons with four wooden composite wheels, very similar to the Susa ones, have been found in the «royal» tombs of the 3 rd – first 171

Lyubov’ B. Kircho Ermolova, N. M. (1985) ‘Novye issledovaniya ostatkov mlekopitayushchikh iz eneoliticheskogo poseleniya Anau [New investigations of the mammal remains from the Eneolithic settlement of Anau]’. Izvestiya AN Turkmenskoy SSR [Proceedings of the Academy of Sciences of the Turkmenian SSR], 1. Ashkhabad: 85–87. Ganyalin, A. F. (1959) ‘Altyn-Depe (po materialam rabot 1953 g.) [Altyn-Depe (on the materials of excavations in 1953)’. Trudy Instituta istorii, arkheologii i etnographii AN Turkmenskoy SSR [Transactions of the Institute for the History, Archaeology and Ethnology of the Academy of Sciences of the Turkmenian SSR], V: 30–44. Kasparov, A. K. (2006) Skotovodstvo i okhota epokhi neolita – paleometalla v Yuzhnom Turkmenistane (razvitie strategii ispol'zovaniya zhivotnykh resursov) [Cattle-breeding and hunting in Neolithic and Paleometal epoch in Southern Turkmenistan (strategy of use of animal resources – its change in time)]. St.-Petersburg. Khlopin, I. N. (1969) Pamyatniki razvitogo eneolita Yugo-Vostochnoy Turkmenii [Remains of the Developed Eneolithic in South-Eastern Turkmenia]. Leningrad. Kircho, L. B. (1999) K izucheniyu pozdnego ehneolita Yuzhnogo Turkmenistana (osnovy klassifikatsi raspisnoy keramiki i neopublikovannye materialy poseleniya Ak-depe) [Studies of the Late Eneolithic of Southern Turkmenistan (Basic Classification of Painted Pottery and Unpublished Materials from the Settlement of Ak-Depe)]. St.Petersburg. Kircho, L. B. (2005a) ‘Zaklyuchenie [Conclusion]’. In Masson V. M., Berezkin Yu. E. (eds) Khronologiya ehpokhi pozdnego ehneolita — sredney bronzy (pogrebeniya Altyn-depe) [The Chronology of the Late Eneolithic — Middle Bronze Ages in Central Asia (Burials of AltynDepe)]. St.-Petersburg : 512–519. Kircho, L. B. (2005b) ‘Pogrebal'nyy inventar' Altyn-depe [Grave offerings from Altyn-Depe]’. In Masson V. M., Berezkin Yu. E. (eds) Khronologiya ehpokhi pozdnego ehneolita — sredney bronzy (pogrebeniya Altyn-depe) [The Chronology of the Late Eneolithic — Middle Bronze Ages in Central Asia (Burials of Altyn-Depe)]. St.-Petersburg: 347–421. Kircho, L. B. (2007) ‘Drevnie svyazi naseleniya Yuzhnogo Turkmenistana i doliny Zeravshana (nachalo formirovaniya torgovykh putey v Sredney Azii) [Ancient connections between South Turkmenistan and the Zeravshan valley (the formation of the first trade routes in Middle Asia)]’. Zapiski IIMK RAN [Transactions of the Institute for the History of Material Culture], 2. St.-Petersburg: 193–208. Kircho, L. B. (2008) ‘Kul'turnye vzaimodeystviya kak faktor stanovleniya rannegorodskoy tsivilizatsii (po materialam Yuzhnogo Turkmenistana V–III

tys. do n. e.) [Cultural Interactions as a Factor in the Development of the Early Urban Civilization (on the basis of Southern Turkmenistan materials 5th – 3rd millennium BC)]’. Vestnik SanktPeterburgskogo Universiteta [Bulletin of the St.Petersburg University], series 2, History, 3. St.Petersburg: 143–154. Kozhin, P. M. (1986) ‘Pervye povozki (First wagons)’. Voprosy istorii [Problems of History], 7. Moscow: 185–189. Kurbansakhatov, K. (1987) Eneolit Anau [Eneolithic of Anau]. Ashkhabad. Kuz'mina, E. E. & Lyapin A. A. (1980) ‘Novye dannye o rasprostranenii kolesnogo transporta v Turkmenii [New evidence of the spread of wheel transport in Turkmenia]’. Novye issledovaniya po arkheologii Turkmenistana [New studies on the archaeology of Turkmenistan]. Ashkhabad: 34–44. Lamberg-Karlovsky, C. C. (1985) ‘The longue durée of the Ancient Near East’. De L’Indus aux Balkans, Recueil Jean Deshayes. Paris: 55–72. Lisitsyna, G. N. (1978) Stanovlenie i razvitie oroshaemogo zemledeliya v Yuzhnoy Turkmenii [The rise and development of the irrigated agriculture in South Turkmenistan]. Moscow. Littauer, M. A. & Crouwel, J. H. (1979) Wheeled Vehicles and Ridden Animals in the Ancient Near East. Leiden/Köln. Masimov I. S. (1976). Keramicheskoye proizvodstvo epokhi bromzi v iuzhnom Turkmenistane [The Bronze Age ceramic production in South Turkmenistan], Ashkhabad. Masimov I. S. (1979) ‘K voprosu ob osvoenii nizoviy Murgaba drevnezemledel'cheskimi plemenami [On the problem of occupation of the lower reaches of the Murghab by the early agriculturalist tribes]’. Izvestiya AN Turkmenskoy SSR [Proceedings of the Academy of Sciences of the Turkmenian SSR], 2. Ashkhabad: 85–87. Masson, V. M. (1960) ‘Kara-depe u Artyka [Kara-Depe near Artyk]’. Trudy Yuzhno-Turkmenistanskoy arkheologicheskoy kompleksnoy ekspeditsii [Transactions of the South-Turkmenistan Archaeological Interdisciplinary Expedition], X. Ashkhabad: 319–463. Masson, V. M. (1970) Raskopki na Altyn-depe v 1969 g. [Excavations at Altyn-Depe in 1969]. Ashkhabad. Masson, V. M. (1976) Ekonomika i sotsial’nyi stroy drevnikh obshchestv [Economy and social organization of ancient societies]. Leningrad. Masson, V. M. (1988) ‘Altyn-depe’. University Museum Monograph 55. Philadelphia. Rao, S. R. (1985) Lothal. A Harappan Port Town. 1955– 62, II. New Delhi. Sarianidi, V. I. (1960) ‘Eneoliticheskoe poselenie Geoksyur [Eneolithic settlement of Geoksyur]’. Trudy Yuzhno – Turkmenistanskoy arkheologicheskoy kompleksnoy ekspeditsii 172

TRANSPORT AT ALTYN-DEPE IN THE 3RD MILLENNIUM BC (THE EVIDENCE OF VEHICLE MODELS) [Transactions of the South-Turkmenistan Archaeological Interdisciplinary Expedition], X. Ashkhabad: 225–318. Sarianidi, V. I. (1965) Pamyatniki pozdnego ehneolita yugo-vostochnoy Turkmenii [Late Eneolithic Sites in South-Eastern Turkmenia]. Moscow. Sarianidi, V. I. (2005) Gonur-depe. Gorod tsarey i bogov [Gonur-Depe. City of Kings and Gods]. Ashkhabad. Schmidt, E. F. (1937) Excavations at Tepe-Hissar, Damghan. Philadelphia. Schmidt, H. (1908) ‘The Archaeological Excavations in Anau and Old Merv’. Explorations in Turkestan, 1, , ed. R. Pumpelly, Washington: 83–210. Shchetenko, A. Ya. (1968) ‘Raskopki poseleniya epokhi bronzy Taychanak-Depe [Excavations of the Bronze Age site of Taychanak-Depe]’. Karakumskie drevnosti [Kara-Kum Antiquities], 2. Ashkhabad: 18–24. Udeumuradov, B. N. (1993) Altyn-depe i Marghiana: svyazi, khronologiya, proiskhozhdenue [AltynDepe and Margiana: connections, chronology, origin]. Ashkhabad.

173

Lyubov’ B. Kircho

Fig. 1 – South Turkmenistan. Wheel models of the second half of the 4 th and 3rd millennium BC (8, 10, 12 – unbaked clay, the rest – terracotta and ceramic): 1–6 — Middle Bronze Age (Namazga V period); 7, 8 – Early Bronze Age (Namazga IV period); 9–18 — Late Eneolithic period (Namazga III); 19–26 – Middle Eneolithic period (Namazga II). 1–13, 17 – Altyn-Depe (1, 3–6 – horizon 1; 2 – surface finds; 7–13 – horizons 7–13; 17 – horizon 12); 14–16, 19 – Kara-Depe (14, 15 – horizon 1A; 16 – horizon 1B; 19 – horizon 3); 18 – Chong-Depe; 20 and 23 – Ilgynly-Depe, horizons II and III; 21 – Northern Hill of Anau; 22 – Geoksyur 9; 24 – Geoksyur 1, horizon 5; 25 – Akcha-Depe; 26 – Aina-Depe (3 – according to Udeumuradov 1993; 18 and 24 – according to Sarianidi 1965 and 1960; 19 – according to Masson 1960; 21 – according to Kurbansakhatov 1987; 22, 25, 26 – according to Khlopin 1969).

174

TRANSPORT AT ALTYN-DEPE IN THE 3RD MILLENNIUM BC (THE EVIDENCE OF VEHICLE MODELS)

Fig. 2 – South Turkmenistan, figurines, statuettes and carts of the late 4 th – 3rd millennium BC (4, 6, 15–20 – terracotta, the rest – unbaked clay): 1–6 – Middle Bronze Age (Namazga V period); 7–13, 20 – Early Bronze Age (Namazga IV period); 14–19 – Late Eneolithic period (Namazga III). 1 – figurine of a camel (?); 2–4, 7, 8, 18–20 – bull figurines: 5, 6, 9, 14–17 – models of one-axle carts; 10 – wheel from a cart model; 11–13 – anthropomorphic statuettes. 9A, 14A, 16A – cart models (partial reconstruction). 1–17 – Altyn-Depe (1–3, 5 – horizon 3; 4, 6 – horizon 1; 7 – horizon 5; 8–13 – horizon 4; 14 – horizon 10; 15, 16 – horizon 11; 17 – horizon 12); 18, 19 – Kara-Depe, horizon 1A (according to Masson 1960); 20 – Ulug-Depe (according to Lisitsyna 1978).

175

Lyubov’ B. Kircho

Fig. 3 – Altyn-Depe and Taichanak-Depe (10), terracotta figurines, carts and vessel details of the second half of the 3 rd millennium BC: 8, 10, 17, 18, 21, 24, 25 – the end of the Early Bronze Age (Late Namazga IV period)1; the rest – Middle Bronze Age (Namazga V period). 1–8, 10 – heads of camels and of a bull (9) from cart models; 11–13 – camel figurines; 14 – model of one-axle cart; 15–17, 21 – models of two-axle carts; 18–20 – cart details; 22, 25, 28 – fragments of zoomorphic vessels and a bull figurine on wheels; 23, 24, 26, 27 – zoomorphic spouts and detail of vessels. 1, 3, 4, 7, 11, 13–16, 19, 20, 23 – horizon 1; 16 – horizon 2; 2, 17, 21, 22, 28 – horizon 3; 8 – horizon 4; 5, 6, 12, 18, 26, 27 – from the surface of the site; 24, 25 – horizon 5 (2, 15, 17 – according to Masson 1988; 10 – according to Shchetenko 1968; 13 – according to Masimov 1976).

1 Painted fragments of terracotta figurine and cart models (10, 17, 18, 21) were found in the Namazga V horizons, but they were produced in the Late Namazga IV period.

176

INAA OF AGATE SOURCES AND ARTIFACTS FROM THE INDUS, HELMAND, AND THAILAND REGIONS Randall Law, Alison Carter, Kuldeep Bhan, Arun Malik & Michael D. Glascock

shapes of the cavities in which they formed) may be carried away by fluvial action and end up reconsolidated in conglomerate or loose in the beds of rivers, streams, nalas and wadis. Such secondary contexts are sometimes far removed from original host formations and can potentially contain agates from multiple geologic occurrences located across an enormous geographic area. It is for reasons such as these that some scholars feel identifying the sources of agate artifacts might be problematic (see discussion by Vidale 2000: 42).

Agate was one of the ancient world’s premier prestige goods, especially the red-orange variety known as carnelian. The stone was utilized by and traded between societies from Africa to eastern Asia (Inizan 1993; Insoll et al. 2004; Theunissen et al. 2000). In this paper, we present the results of a series of instrumental neutron activation analyses (INAA) of agate samples and artifacts from sources and/or sites in the Indus, Helmand, and Thailand regions. This study represents the beginning of a broad-scale, long-term project aimed at identifying Old World agate sources and the regional and inter-regional trade networks through which this important stone was exchanged in both raw and finished form.

Geologically speaking, agate is not a particularly uncommon rock. Figure 1 is a map on which a few dozen or so of the more notable occurrences in Asia are identified (details including citations for these can be found in Law 2011, Chapter 8). However, good agate – i.e, that which ancient lapidaries would have found suitable for beadmaking – is not widely available. Nodules of the size and quality required to make Harappan-style long-barrel carnelian beads are, in fact, extremely rare.

Our main goal at this initial stage was to determine if agate sources could be effectively differentiated from one another using INAA-derived data. To this end, sets of samples representing multiple geologic occurrences were analyzed at two different research reactors – the University of Wisconsin’s Nuclear Reactor (UWNR) and the University of Missouri Research Reactor (MURR). It was found that good-to-excellent (≈ 85 % to 95 %) statistical separation between different sources could be achieved when the INAA results from these facilities were evaluated using canonical discriminant analysis (CDA). A set of agate artifacts from the Indus Civilization sites of Harappa and Nagwada along with a set of carnelian beads of unknown archaeological provenience were also analyzed. Although the geologic dataset they were compared to is, at present, limited, the results both support and challenge past assumptions regarding the acquisition of agate in proto-historic South Asia.

There have been two recent geologic provenience studies involving agate. Theunissen and others (2000) employed PIXE/PIGME in an effort to shed light on the historic era carnelian trade in Southeast Asia. Their dataset consisted of 51 agate artifacts (beads and manufacturing debris) from sites in Sri Lanka, Thailand and India and two geologic samples from a Thai quarry. Insoll and others (2004) used UV-LA-ICP-MS to examine the carnelian trade between India and Africa. Their dataset contained 13 samples (archaeological debris and modern raw material) from the Ratanpur region of Gujarat, India and 13 artifacts (ornaments and debris) from sites in western Africa. Each study was laudable in that the methods employed were non-destructive (or minimally destructive) and produced useable data on range of major and minor elements. The results of both, however, were equivocal.

We begin with brief overviews of agate, agate occurrences and the analytical methods employed previously and in this study. The UWNR and MURR analyses are then discussed separately in turn. The different datasets and research questions examined are outlined in each of those sections. Lastly, the implications of this study for the future of agate provenience research are discussed.

Although some meaningful groupings were observed among samples when each dataset was examined using principal component analysis and/or cluster analysis, regional differences were difficult to define and evaluate. The problem, in our opinion, lay not so much with the methodologies that were employed but with the datasets that were examined. Theunissen and others’ dataset was almost entirely made up of agate artifacts while that analyzed by Insoll’s group contained geologic samples from what was essentially a single extensive deposit (Ratanpur). In order to begin to confidently source agate artifacts it is necessary to compare them to a large number samples collected from multiple geologic occurrences. This is the approach we have taken in the

Agate, agate occurrences and previous provenience studies Agates are translucent microcrystalline sedimentary rocks that form when silica precipitates into cavities within some type of host rock. The mechanisms behind their growth, frequent banding and other distinctive characteristics are subjects of much debate (see Moxon 1996 for more on these matters). As they erode from their host rocks, which can also be sedimentary but are more commonly volcanic, agates (many retaining the nodular 177

Randall Law, Alison Carter, Kuldeep Bhan, Arun Malik & Michael D. Glascock The region could aptly be called the ‘Saudi Arabia’ of agate. Lastly, Gujarat was a historically important source area. Greek (McCrindle 1885: 77, 334), Mughal (Khan 1756: 250) and early European colonial (Barbosa 1517: 66-7) records all make reference to the agate resources there. The city of Khambhat (Cambay) has been a major center for the manufacture of agate ornaments since at least the 16th century (Arkell 1936) and the traditional methods still employed there have been the subject of several ethnoarchaeological studies (Kenoyer et al. 1991; Possehl 1981; Roux 2000).

current study. Moreover, we have employed an analytic technique (INAA) that, while destructive, has a long history of success in provenience studies of archaeological stone and a method of data evaluation (CDA) that is especially well-suited to differentiating between known geologic sources and assigning artifacts to them. Instrumental neutron activation analysis and canonical discriminant analysis INAA is a highly accurate and precise method for quantifying the elemental compositions of materials. Archaeologists around the world have long employed it in efforts to determine the proveniences of a wide range of artifacts (see Glascock and Neff 2003 for a detailed account of this technique and its application). In brief, INAA involves the irradiation (or activation) of elements within artifacts and/or source samples by exposing them to a neutron flux. Following varying periods of decay, the gamma ray emissions they produce are detected and counted. After the results are screened of elements that failed to be detected in all samples or had high count-rate standard deviations, the data are evaluated.

It has long been assumed (from Pascoe 1931: 681 to Vidale 2000: 42) that, within Gujarat, Harappan agate primarily came from the extensive deposits of the Ratanpur area in the southern part of the state. However, noting that there are also deposits in the Kutch area, Shereen Ratnagar recently asked (2004: 146) ‘did the Harappan inhabitants of Dholavira know of these sources?’ This was a good question. Agate occurrences in northern Gujarat should have been far more accessible to the peoples of that city than the Ratanpur deposits, which are located hundreds of kilometers to the southeast. Although no artifacts from Dholavira were available for this analysis, it was reasoned that if Gujarat was the principal region from which Indus Civilization peoples obtained agate resources, then it should be possible address the question through provenience analyses of artifacts from other Indus sites in Gujarat and the Indus Valley proper. For this study, 24 agate artifacts (debris fragments and broken beads) from Harappa in the Punjab and three from Nagwada in Saurashtra region of Gujarat were subjected to INAA at the UWNR. These were compared to agate samples collected from three deposits in Gujarat as well as a set of agate artifacts from the site of Shahr-i-Sokhta, which served as proxy samples for a source in eastern Iran.

For this study, CDA was deemed to be the most effective statistical method with which to use multivariate INAA data to differentiate agate sources from one another and to assign a possible provenience to agate artifacts. During CDA (see Baxter 1994 for a full discussion of this method), linear combinations of variables called discriminant functions are generated that produce a maximum degree of separation (discrimination) between various defined groups of cases, which in this instance are the individual sets of samples collected from different agate sources. Discrimination success is evaluated by a cross-validation technique in which each case is left out its group in turn and compared to the dataset as an ungrouped case. A percentage is generated based on the number of cases that were correctly assigned to the groups to which they actually belong. Agate artifacts are plotted as ungrouped cases and assigned to the group whose center (or centroid) in multidimensional space they are nearest. The data are displayed on a bivariate plot using the first and second discriminant functions.

Agate sources sampled in Gujarat Among the low hills around the village of Ratanpur, Bharuch District, Gujarat, there are hundreds (if not thousands) of agate mining pits and shafts sunk into the Miocene conglomerate called the Babaguru Formation. Although these workings are often referred to as the ‘Rajpipla’ deposits/mines (as they were within the confines of that princely state prior to 1947), ‘Ratanpur’ is a more appropriate designation (Ball 1886: 238). All occur within 15 kilometers of Ratanpur village and the hilltop tomb/shrine of Gori Pir (or Baba Ghor) – a Muslim saint who is said to have come from Africa in the 15th century and established bead-making operations at nearby settlements such as Limodara (Kenoyer and Bhan 2005). Samples for this study were collected from pits, shafts and tailings along a zone that extended from the base of Gori Pir Hill three kilometers toward the southeast. Because the Babaguru Formation is a secondary context agate deposit, an attempt to assess geochemical variation across that zone was deemed to have little utility. All samples from this occurrence are, therefore, treated as coming from a single source.

UWNR study: Agate artifacts from Harappa and Nagwada compared to agate samples from Gujarat and eastern Iran The Indian state of Gujarat is widely believed to have been an important source area, perhaps even the primary source area, for the agates used by peoples of the Indus Civilization (c. 2600 to 1900 BC). There are many good reasons for this. Firstly, Indus peoples were present there, often in very close proximity to some significant occurrences (Fig. 2). In fact, ornamental stones were probably among the resources (some other being marine shell, salt and pasturage) that attracted them to the region in the first place. Secondly, although occurrences of agate can be found in many parts of Asia, the extent, diversity and sheer richness of sources in Gujarat is unparalleled. 178

INAA OF AGATE SOURCES AND ARTIFACTS FROM THE INDUS, HELMAND, AND THAILAND REGIONS dataset. Iran is an underappreciated potential source area. The ‘most celebrated Iranian agate localities are in the central and eastern’ part of the country (Nazari 2004: 21). Around the Khur area, primary context banded agate nodules occur within tuffaceous andesite (ibid.). Extensive secondary context deposits can be found in Iran’s broad salt deserts (dasht) and inland deltas. The explorer Henry Savage Landor marveled (1902: 79) at the ‘handsome agates’ spread across the wastes of the Dasht-e Lut. Ali Hakemi wrote (1997: 15) that ‘carnelian is found in considerable quantities in the Lut flood plain’. Finally, Maurizio Tosi noted (1969: 374) that ‘with regard to cornelian … numbers of little pebbles of this stone, with a diameter often exceeding 3 cm, may be collected along the dried out beds and ancient branches’ of the Helmand River delta near the site of Shahr-iSokhta. For this study, a set of 14 agate flakes and nodule fragments from that site were used as proxy samples for a source in eastern Iran.

There are numerous agate of occurrences in central and eastern Kutch (Geological Survey of India 2001: 47; Merh 1995: Fig. 17; Wynne 1872: 72-3) that would have been directly accessible to Indus Civilization peoples settled in northern Gujarat. A source near Khandek village was brought to our attention by R. S. Bisht, the excavator of the Harappan city of Dholavira, which is located some 70 km to its west-northwest on the island of Khadir. Ravaji Solanki – the local stone expert (pattarwala) at Dholavira – provided directions to Khandek and his brother Narsingh, who resided there, guided co-author Randall Law to the source itself. A pavement-like layer of loose agates (natural carnelian, yellow-brown agate, clear chalcedony, moss agate) and other microcrystalline silicates (red, green, brown and variegated jaspers) covers an area of two or three hectares just east of the village. This source is located around five kilometers from the small fortified Indus Civilization settlement of Surkotada (Joshi 1990). Although no prehistoric workings or cultural materials were identified, numerous ‘window’ flakes (pieces of cortex that were struck from nodules in order to observe the quality of the agate inside) were found that indicate it had been exploited for materials at some time in the past.

UWNR study results Twenty samples each from the Ratanpur, Mardak Bet and Khandek sources along with the 14 proxy source samples from Shahr-i-Sokhta were analyzed at the UWNR. From the INAA data that were returned, ten elements 1 – Al, Co, Cr, Eu, Fe, La, Na, Sb, Sc, and V – were selected for use in CDA (Fig.3). Good separation between the three Gujarati sources and the Iranian proxy source was achieved. Exactly 85.1% of leave-one-out cross-validated grouped geologic cases were classified correctly. Most of the misclassification (overlap) that occurred was among the Gujarati sources. Only one sample from Shahr-iSokhta (S-i-S_14, noted on Fig. 3) was classified as belonging to a Gujarati source (Mardak Bet) when it was cross-validated. Overall, however, it is clear that the Shahr-i-Sokhta agates are geochemically distinct from those found in Gujarat. When the artifacts from Harappa and Nagwada were plotted as ungrouped cases all but four from Harappa (AH-1, 2, 9 and 20) were assigned by CDA to either Ratanpur (n = 3), Mardak Bet (n = 13) or Khandek (n = 7). Although agate samples from many other regions will eventually need to be analyzed, these results seem to confirm what scholars have long assumed – that Indus peoples mainly used Gujarati agate. The four artifacts from Harappa assigned to the Shahr-i-Sokhta group do suggest that residents of the site sometimes utilized agate from sources in other areas. However, until samples from actual deposits near Shahr-i-Sokhta can be analyzed (instead of artifacts) it would be premature to firmly state that those sources were in eastern Iran.

The agate deposits on the island (bet) of Mardak, in the salt marsh southeast of Kutch known as the ‘Little Rann’, can be difficult to reach due to seasonal flooding of the area surrounding them (Trivedi 1964: 11). A sampling trip by authors Law and Arun Malik in early 2003 ended with both researchers stuck in the mud within sight of the island. A second attempt by Law, Malik and Dr. Kuldeep Bhan later that same year just prior to the summer monsoons was successful. Mardak Bet is a thinly-shaped, east-west oriented island around 12 km in length with a maximum width of about 1.25 km. The agate beds are found in two main areas. The most extensive is located near the island’s constricted mid-section, which Malik designated ‘nana.’ Another occurs 3 km to the east, around the base of its highest hill (≈ 40 m above the salt flats), which was designated ‘mota.’ A wide range of microcrystalline silicates are found at both locations. Brownish-gray agate is by far the most abundant type but nodules of natural carnelian, clear chalcedony and moss agate are not uncommon. Red, green, yellow-brown and variegated jaspers (including bloodstone) are also found. No prehistoric settlements are known to exist on Mardak Bet and no clearly ancient workings in the island’s agate beds were identified during our short visits to them. Mining pits and sorting areas related to modern extraction activities have likely obscured any evidence of earlier ones. There are, nonetheless, indications that ancient peoples did exploit these deposits. Numerous agate and jasper flakes, some with a heavy patina suggesting great antiquity, were found on the hillside at ‘mota’ Mardak Bet.

Next, the three agate sources in Gujarat were compared to one another alone, without the Shahr-i-Sokhta proxy source-group (Fig. 4). Although good separation between the three was achieved, it was only slightly better than that for the original CDA. Exactly 86.7% of leave-oneout cross-validated grouped cases were classified correctly this time whereas 85.1% were classified

Iranian sources In order for us to best assess whether or not Harappans were acquiring agate from Gujarat it was necessary to have samples from different region in the geologic

1

179

These data can be found in Law 2008: appendices 8.1 through 8.5.

Randall Law, Alison Carter, Kuldeep Bhan, Arun Malik & Michael D. Glascock were producing polished stones and cabochons from the local agate to use as ring settings (Nigel Chang personal communication 2007). However, during Carter’s visit only one villager was still practicing this craft. Twenty samples from Ban Khao Mogul were sent to MURR for INAA.

correctly when the S-i-S source-group was included in the analysis. This again shows that nearly all of the overlap (misclassification of grouped cases) in the dataset is among Gujarati sources rather than between them and the Iranian artifact/proxy source samples. In this instance, the misclassifications were among the Mardak Bet samples (one was predicted to belong to Ratanpur) and the Ratanpur samples (one was predicted to belong to Mardak Bet and three to Khandek).

A set of beads from Afghanistan Unfortunately, no agate artifacts from Southeast Asian sites were available for this analysis. However, Dr Mark Kenoyer kindly provided a large number of carnelian beads he obtained from Haji Ashoor – an Afghani jeweler working in Peshawar, Pakistan who deals in ornaments from Afghanistan. The age and provenance of the 184 heavily worn examples given to us was not known. Therefore, prior to INAA, a morphological study of these beads was conducted in the hope that it would provide clues as to when and where they were made.

On Figure 4, the Harappan artifacts (excluding the four had previously been predicted to belong to the Iranian proxy source-group) are plotted as ungrouped cases in relation to the three Gujarati source-groups. In this instance, 14 were assigned to Mardak Bet, seven to Khandek, and only two (one each from Harappa and Nagwada) to Ratanpur. These results, although preliminary, suggest that, contrary to what many scholars have assumed, Indus Civilization peoples were not heavily exploiting the Ratanpur area agate deposits. Instead, most of their agate appears to have been derived from sources in northern Gujarat.

One of the best ways to narrow down when and where a bead was made is by looking its drill hole. As Kenoyer (1992: 86) notes, ‘the drill hole reflects an important cultural choice that represents very different technologies’. During the Harappan period, most beads were perforated using either chert or a type of flint clay known informally as ‘Ernestite’. However, during the Early Historic period (beginning c. 600 BC) drills tipped with diamonds were introduced. These allowed for faster and easier perforation of hard stone like agate-carnelian and are still used by beadmakers today. Based on research by Kenoyer, Vidale, and Bhan (1994) there are two primary diamond-drilling techniques. In the first, two diamonds are affixed to the drill tip (double-diamond drilling). This technique is unique to South Asia and generally results in a larger hole. The second technique employs a drill tipped with one diamond (single-diamond drilling). Historically, it is practiced in Central Asia and results in a comparatively smaller hole.

MURR study: Agate beads from Afghanistan compared to samples from Gujarat, Iran and Thailand For the next phase the project, we shifted our analyses to the University of Missouri Research Reactor (MURR) in order to take advantage of the longer INAA count times employed at that facility. The geologic dataset was expanded to include samples from an agate source in Thailand. In addition, 15 beads from Afghanistan of unknown provenance were analyzed. The agate source at Ban Khao Mogul, Thailand Agate beads recovered at archaeological sites in Southeast Asia are often assumed to have originated from sources in India and are widely believed to be among the first signs of contact between those two regions during the early Iron Age period (c. 500 BC to AD 500). Some scholars treat such beads as proxies for Indian political and cultural influence (Francis 1996; Glover 1989). However, others suspect that many of the agate ornaments found in mainland Southeast Asia may have been produced by indigenous groups rather than imported from South Asia (Theunissen et al. 2000). One possible raw material source for local bead manufacturing that has been discussed in the literature (Glover 1989; Theunissen et al. 2000) is found at Ban Khao Mogul in Lopburi Province, Thailand. The deposit there consists of small nodules of agate and other microcrystalline materials eroding from a small limestone outcrop. Two agate samples from Ban Khao Mogul were included in the study by Theunissen and others (2000) discussed earlier in this paper. Although those results were inconclusive, they did suggest that some agate beads from Southeast Asia may not have originated in India. In February 2007, Alison Carter traveled to Ban Khao Mogul in order to collect additional samples for a more thorough analysis. The age of the quarry at this location is not known. As recently as 10 years ago, several villagers

The 184 beads were analyzed and classified (by Carter) using a bead coding system developed by Kenoyer for use at Harappa. Characteristics such as shape, length and width measurements were recorded. The interior perforations were measured and impressions of the drill holes were made. The preliminary results indicated that the beads were made using both single and doublediamond tipped drills. Fifteen examples were sent to MURR for INAA. Ten were selected from the doublediamond drilling group and five were chosen from the single-diamond drilling group. We hypothesized that the double-diamond drilled beads would likely be more compositionally related to the Gujarati sources, as they were made using the South Asian technique, whereas the Central Asian-style single-diamond drilled beads might be more closely related to geologic samples from the Iranian proxy source. MURR study results Along with the 20 Ban Khao Mogul samples and the 15 Afghani beads, sets of agate from Mardak Bet (n = 14), 180

INAA OF AGATE SOURCES AND ARTIFACTS FROM THE INDUS, HELMAND, AND THAILAND REGIONS Ratanpur (n = 15) and Shahr-i-Sokhta (n = 15) were analyzed at the MURR. From the data that were returned, ten elements2 – Al, Ce, Co, Fe, La, Mn, Na, Sb, Sc, and Sm – were selected for use in CDA (Fig. 5). An outstanding degree of separation between the different sources was achieved with 95.1% of cross-validated grouped geologic cases classifying correctly. Most of what little misclassification occurred was between the two Gujarati sources. The Thai samples are highly distinct, which has exciting implications for future studies of long-distance trade between India in Southeast Asia. Although many more sources will need to be analyzed (especially those in eastern India), our results suggests that, using INAA, archaeologists working in Southeast Asia should be able to clearly differentiate between agate beads imported from South Asia and those being made from local stone. Understandably, researchers might be reluctant to subject rare and/or finished objects to such a destructive analytical technique. However, it may be possible to source agate artifacts using other nondestructive (or less destructive) methods (such as PIXE/PIGME or UV-LA-ICP-MS) provided that a suitable compositional database of different geologic sources against which to compare them is established.

is possible to differentiate samples from three agate deposits in Gujarat reasonably well and that, contrary to the expectations of many scholars, Harappan beadmakers were probably obtaining much of their raw material from sources in the northern part of that state rather than from Ratanpur. An excellent degree of discrimination was achieved when the Gujarati deposits were compared to a set of artifacts presumably from sources in eastern Iran. This suggests that it is possible, at the very least, to confidently assign a regional geologic provenience to Harappan agate. The results of the MURR study lent further support to this conclusion. Agate from deposits in Gujarat, the group of Iranian artifacts and the Thai samples were all shown to be highly distinct from one another. There is very good reason to expect that when the geologic dataset is eventually expanded to include samples from occurrences in Sindh, Balochistan, Afghanistan, Arabia, Tibet and Central Asia it will also be possible to differentiate those source areas and assign a regional geologic provenience to agate artifacts. Acknowledgements The authors wish to thank Drs Richard Meadow and J. Mark Kenoyer – co-directors of the Harappa Archaeological Research Project, for providing us access to the agate artifacts from Harappa and to Dr. Fazal Dad Kakar – Director-General of Archaeology and Museums, Government of Pakistan, for allowing us to analyze those artifacts in the United States. In addition, we are deeply grateful to Professor Maurizio Tosi and Dr. Massimo Vidale for supplementing this study with agate fragments from the site of Shahr-i-Sokhta. Thanks also to Dr. Nigel Chang, Dr. Bill Boyd, and Sompong Paekosae. Special thanks to Robert Agasie and Kevin Austin at the University of Wisconsin’s Nuclear Reactor Laboratory. Support for this project was provided by the United States Department of Energy Reactor Sharing Program, the Wenner-Gren Foundation (Gr. 7066), the Bead Society of Greater Chicago, the Ruth Dickie Graduate Women in Science Grants-in-Aid program, and the National Science Foundation (BCS-0327246 & BCS0504015).

When the 15 Afghani beads were compared to the geologic sources/proxy sources, eleven were predicted to belong to one of the Gujarati deposits – nine to Ratanpur and two to Mardak Bet. This is an interesting reversal of the pattern seen during the Harappan Period when all but two agate artifacts assigned to occurrences in Gujarat seem to have come from deposits in that northern part of that state. It makes sense though that a large number of the 15 diamond-drilled beads would have come from Ratanpur in the south, which by all accounts was a far more important agate source during the historic period. Also interesting is the fact that of the four beads (labeled on Fig. 5) assigned to the Shahr-i-Sokhta group, three were single-diamond drilled. Agate sources in the Helmand region or related ones elsewhere in Iran or Afghanistan would have been among the nearest to the places where beadmakers were employing this Central Asian drilling technique. No beads were assigned to Ban Khao Mogul, which was both expected and reassuring. It is highly unlikely that South or Central Asian beadmakers would have imported agate all the way from Southeast Asia when there were closer and richer sources. The fact that none of the Afghani beads even remotely resembled the Thai deposit lends further support to our finding that agates from these different regions are chemically distinct from one another.

Bibliographical References Arkell, A. J. (1936) ‘Cambay and the Bead Trade’. Antiquity, 10: 292-305. Ball, V. (1886) ‘A Geologist's Contribution to the History of Ancient India’. Journal of the Royal Geological Society of Ireland (1882-1884), Volume VI (New Series): 215-62. Barbosa, D. (1517) A description of the coasts of East Africa and Malabar in the beginning of the sixteenth century. London. Baxter, M. J. (1994) Exploratory multivariate analysis in archaeology. Edinburgh. Francis, P. (1996) ‘Beads, the Bead Trade, and State Development in Southeast Asia’. In Chutiwongs,

Conclusions The results of our study bode very well for future research of this kind. The UWNR analysis showed that it 2

These data may be downloaded at:

http://spreadsheets.google.com/pub?key=pNc4HHJLeWSAnL1nns0uPFg

181

Randall Law, Alison Carter, Kuldeep Bhan, Arun Malik & Michael D. Glascock N. (ed.) Ancient Trades and Cultural Contacts in Southeast Asia. Bangkok. Geological Survey of India (2001) Geology and Mineral Resources of Gujarat, Daman & Diu. Geological Survey of India Miscellaneous Publication, No. 30 (Part XIV): 1-102. Glascock, M. D. and Neff H. (2003) ‘Neutron activation analysis and provenance research in archaeology’. Measurement Science and Technology, 14 (9): 1516-1526. Glover, I. C. (1989) Early Trade Between India and South-East Asia; A Link in the Development of a World Trading System, 16. Center for South-East Asian Studies. Hakemi, A. (1997) ‘Shahdad: Archaeological excavations of a Bronze Age center in Iran’, Reports and memoirs 27. Rome. Inizan, M.-L. (1993) ‘At the Dawn of Trade, Cornelian from India to Mesopotamia in the Third Millennium: The Example of Tello’. In Gail A. J. and Mevissen G. J. R. (eds) South Asian Archaeology 1991. Stuttgart: 121-34. Insoll T., Bhan K. K., Polya D. A. and Jarvis K. (2004) ‘Towards an Understanding of the Carnelian Bead Trade from western India to Sub-Saharan Africa: The Application of UV- LA-ICP-MS to Carnelian from Gujarat, India and West Africa’. Journal of Archaeological Science, 31(8): 1161-1173. Joshi, J. P. (1990) ‘Excavations at Surkotada 1971-72 and Exploration in Kutch’. Archaeological Survey of India, New Delhi. Kenoyer, J. M. (1992) ‘Ornament Styles of the Indus Tradition: Evidence from recent excavations at Harappa, Pakistan’. Paléorient, 17 (2): 79-98. Kenoyer, J. M. and Bhan K. K. (2005) ‘Sidis and the Agate Bead Industry of Western India’. In CatlinJairazbhoy A. and Alpers E. A. (eds) Sidis and Scholars. London: 61-80. Kenoyer, J. M., Vidale M. and Bhan K. K. (1991) ‘Contemporary Stone Bead Making in Khambhat India: patterns of craft specialization and organization of production as reflected in the archaeological record’. World Archaeology, 23 (1): 44-63. Kenoyer, J. M., Vidale M. and Bhan K. K. (1994) ‘Carnelian Bead Production in Khambhat India: An Ethnoarchaeological Study’. In Allchin B. (ed.) Living Traditions: Studies in the Ethnoarchaeology of South Asia. New Delhi: 281306. Khan, A. M. (1756) Supplement to the Mirat-i-Ahmadi: with explanatory notes and appendices. Translated from the Persian of Ali Muhammad Khan by Syed Nawab Ali and Charles Norman Seddon. Baroda. Landor, A. H. S. (1902) Across Coveted Lands: A Journey from Flushing (Holland) to Calcutta, Overland. London.

Law, R. W. (2011) Inter-Regional Interaction and Urbanism in the Ancient Indus Valley: A Geologic Provenience Study of Harappa's Rock and Mineral Assemblage. Kyoto. McCrindle, J. W. (1885) Ancient India as described by Ptolemy. London. Merh, S. S. (1995) Geology of Gujarat. Geological Society of India, Bangalore. Moxon, T. (1996) Agate: Microstructure and Possible Origin. Auckley. Nazari, M. (2004) ‘Agates & Geodes from the Khur Area, Central Iran’. Australian gemologist, 22 (1): 21-28. Pascoe, E. H. (1931) ‘Minerals and Metals’. In Marshall, J. H. (ed) Mohenjo-Daro and the Indus Civilization. London: 674-685. Possehl, G. L. (1981) ‘Cambay Beadmaking: An Ancient Craft in Modern India’. Expedition, 23 (4): 39-47. Ratnagar, S. (2004) Trading Encounters: From the Euphrates to the Indus in the Bronze Age. New Delhi. Roux, V. (2000) Cornaline de l'Inde: des pratiques techniques de Cambay aux techno-systáemes de l'Indus. Paris. Theunissen, R., Grave P. and Bailey G. (2000) ‘Doubts on diffusion: challenging the assumed Indian origin of Iron Age agate and carnelian beads in Southeast Asia’. World Archaeology, 32 (1): 84105. Tosi, M. (1969) ‘Excavations at Shahr-i-Sokhta: Preliminary report on the second campaign, September-December 1968’. East and West, 19 (3-4): 283-386. Trivedi, R. K. (1964) ‘Agate Industry of Cambay’, Census of India 1961. Volume 5, Part 7-A (1). Government of India, Delhi. Vidale, M. (2000) The Archaeology of Indus Crafts: Indus craftspeople and why we study them. Rome. Wynne, A. B. (1872) ‘Memoir on the Geology of Kutch’, Memoirs of the Geological Survey of India, Vol. IX (Part 1): 1-293

182

INAA OF AGATE SOURCES AND ARTIFACTS FROM THE INDUS, HELMAND, AND THAILAND REGIONS

Fig.1 - Select agate sources and archaeological sites in Asia.

Fig. 2 - Select agate sources and archaeological sites in Gujarat, India.

183

Randall Law, Alison Carter, Kuldeep Bhan, Arun Malik & Michael D. Glascock

Fig. 3 - CDA of Harappan agate artifacts compared to Gujarati and Iranian sources.

Fig. 4 - CDA of select Harappan agate artifacts compared to Gujarati sources.

Fig. 5 - CDA of Afghani agate beads compared to Gujarati, Iranian and Thai sources.

184

IS IT POSSIBLE TO SPEAK ABOUT MIDDLE BRONZE AGE IN SOUTHERN TAJIKISTAN ? Giovanna Lombardo

where moulds for similar axes and a cast specimen have been found (Amiet 1986: 41-42 and 238, Fig. 7, Sialk III 4-5 Period), in Susa (Deshayes 1960: Pl. XXXVI, n, 5, Pl. LX, n. 10, n. 2098, probable Susa II-IIIA, end of the 4th millennium BC), at Shahdad (Hakemi 1997: 636, Gp 2, A Cemetery, gr. 159, second half of the 3 rd millennium).

It has always been stated that the arrival of the population of the Sapalli farming culture in Southern Tajikistan had taken place in the middle of the 2nd millennium BC, in the Molali-Bustan phases of the culture, (Vinogradova 1993: 289-290; Vinogradova 2002: 71-72). The materials from the excavated sites appeared to be consistent with this statement: mostly Late Bronze –Early Iron Namazga VI pottery was found and the settlement pattern, with only farming villages and necropolises, was characteristic of the Namazga VI Period lack of urban centres.

Axes-hatchets of the same type as the Sharshar one have also been found in Central Asia since the Namazga III period at Sarazm (Isakov 1991: Fig. 78, nn. 1, 2), in the area around Pianjikent (Chernykh 1992: 175, Fig. 64, n. 10, stray find; Kuzmina 1966: Pl. I, n. 8, from Iori-sai, probably of the end of the 3rd - beginning of the 2nd millennium) and finally, at Sapalli Tepe (Askarov 1977: Pl. XXXI, n. 2, Sapalli phase).

In her relation here Dr Natalia Vinogradova describes a group of objects, some of which are stray finds, while some other come from excavations, from various areas of Southern Tajikistan: three copper axes (Fig. 1, ns 3-5), a dagger or knife (Fig. 1, n. 6), a stamp seal (Fig. 1, n. 7), two silver bracelets (Fig. 1, ns 1-2); a pottery vessel of square form, perhaps a house model (Fig. 2, n. 7), a group of three pottery vessels of the Sapalli-Djarkutan phases, which were the furniture of a burial in the cemetery of Krug-sai (Fig. 2, ns 1-3), a poorly baked clay female figurine which was discovered in a hearth near the burial, among burnt bones (Fig. 2, n. 4). These objects, are now in the Kuljab Museum, except the stamp seal which is in the Khulbuk museum (Vinogradova, 2012).

The most ancient comparable specimen for the axe from Arakchin (Fig. 1, n. 5) are in the metallurgy of the KuroAraxes culture of Transcaucasia (end of the 4 th-first half of the 3rd millennium BC: Chernykh 1992: 62, Fig.19, n. 20, from Brdazdori, n. 21, from Yalbuz, n. 22 from Kulkakebi, stray finds); the type is documented also in the Caucasus, in the Majkop culture (Chernykh 1992: 70, Pl. 2, n. 1; 71, Fig. 23, n.1, from the Kurgan in the town of Nalchik; 82, Fig. 27, ns 1- 4, moulds for axes, cemetery of Lebedi, Kuban) and in the lower Dnepr basin, in the Pit Grave (Yamnaja) focus (Chernykh 1992: 89, Fig. 29a, n. 21, from Berkhedneprovsk, stray find, n. 23, from Staika, stray find, second half of the 3rd beginning of the 2nd millennium BC). Finally, the type is also documented at Troy III (Schaeffer 1948, vol. II: Fig. 167, nn. 1-3, from Hoard 2, around 2300-2200 BC).

If we consider, among the metal objects, the group formed by the three copper axes and the dagger (Fig. 1, ns 3-6) we can observe that the types to which they belong are spread in a vast area including to the West south-eastern Europe, particularly the Carpatho-Balkan region and the lower Dnepr and Don basins, the Volga Ural region and the area to the North of the Black and Azov Seas, Caucasus and Transcaucasia, Iran and Central Asia to the East.

The third axe, from Sangvora (Fig. 1, n. 4), belongs also to a widely spread typology: like the two others it is documented in south-eastern Europe, Caucasus, Transcaucasia, south-western Iran and central Asia. The most ancient known specimen of this type dates back to the end of the 4th – beginning of the 3rd millennium BC and comes from Susa (Deshayes 1960: Pl. XXIII, n. 2, 1536). In the same period this axe can be found in Transcaucasia, in the Kura-Araxes culture and under the influence of Kura Araxes it was imitated in Anatolia (Chernykh 1992: 62, Fig. 19, n. 19, from Zemi-Avchala, Transcaucasia, 63, Fig. 20, n. 4, from Dinar, Turkey). The typology still existed in the middle-second half of the 3rd millennium in the Kuban basin (Chernykh 1992: 75, Fig. 24, n. 26, 28, stray finds from the Kuban river and from Khashi, Majkop culture). In Transcaucasia this axe is present also in a later period, in the first half - middle of the 2nd millennium in the Trialeti culture (Chernykh 1992: 114: Fig. 37, nn. 2,4, from the Ureki Hoard in western Georgia).

The axe-hatchet from Sharshar (Fig. 1, n. 3) is documented in the Tripolye culture in Moldavia (end of the 4th – first three quarters of the 3rd millennium BC, Chernykh 1992: 38, Fig. 10, n. 10 from the Karbuna hoard, 42, Fig. 12, ns 1, 4 from Brinzeny, stray find, probable second half of the 3rd millennium BC); in the river Kuban basin of Northern Caucasus, in the Majkop culture of the second half of the 3 rd millennium (Chernykh 1992: 47, Fig. 14, n. 2 from Ust Labinskaya). This axe-hatchet continued to be produced in southeastern Europe, in the 2nd millennium, in the area of the Dnepr-Desna culture, in the region to the North of the Black Sea (Sulimirsky 1970: 192: Fig. 46, last row, 197, Fig. 48, last row). The type was also present in northern Iran, in the second half - end of the 4th millennium at Tepe Ghabristan, 185

Giovanna Lombardo same pattern is also in the glyptic of Tepe Giyan V B/C, as some stamp seals show the same design with asymmetrical cross (Rashad 1990: 39, 41-42 and abb. 3, ns 75, 85 from Tepe Giyan VB/C, 4000/3900 – 3500 BC). According to Susanne Baghestani this type of seal can be dated, with reserve, to the beginning of the 2nd millennium (Baghestani 1997:134-135), or perhaps we could hypothesize, also a wider range of time, between the end of the 3rd and the beginning of the 2nd millennium.

The dagger (Fig. 1, n. 6) is documented in the same regions but there is a great number of comparable specimens in Central Asia, especially in the area of the Bactria-Margiana archaeological complex (BMAC). It has been found also in south-eastern Iran, at Shahdad (Hakemi 1997: 640, Fig. Gq 8, from Cemetery A, Grave 081, dagger similar to ours except for the tang, the extremity of which widens up at the sides). One of the most ancient of these specimen has been found in Sarazm (Isakov 1991: Fig. 77, n. 1, Namazga III Period); in Bactria the type is well documented at Sapalli Tepe and in Djarkutan, in the Sapalli and Djarkutan phases (Kaniuth 2006: 96, n. 97, 97, Fig. 99 from Djarkutan, Djarkutan phase; 97, n. 101 and Fig. 101 from Sapalli Tepe, Sapalli phase) and in the plundered graves of northern Afghanistan (Pottier 1984: 137, Fig. 3, n. 18, 138, Fig. 4, n. 19, 185, Pl. III, n. 19); a dagger in relation can also be found in Margiana at Togolok 1 (Sarianidi 1981: 179, Fig. 8, n. 1, Namazga V Period). Outside the BMAC this dagger is present in the Early Bronze Age cemetery of Parkhai II, in the Sumbar valley (Khlopin 1986: 121, abb. 107, n. 2). Comparable specimen have been found in the Pit Grave (Yamnaya) culture (second quarter - end of the 3rd millennium) in the Volga-Urals region and in the lower Dnepr area (Chernykh 1992: 87, Fig. 28, ns. 11-12 respectively from Katovska, kurgan 5, and from the Pokrovsky cemetery, kurgan 17, burial 1; 89, Fig. 29, n. 9, from Maiorovska, kurgan 4, burial 5, n. 13, from Nova Nikolaevka, secondary kurgan burial). The type is documented also in Southern Caucasus, in the Bedeni culture, at the middle-third quarter of the 3rd millennium (Chernykh 1992: 105, Fig. 33, n. 7 from Khramebi, Kurgan 1, n. 10, from the kurgan of Bakurtsikhe, specimen in relation) it can still be found in the Catacomb culture, in the first half of the 2nd millennium in Kalmykia, (Lower Volga Area and Northern Caucasus), and in the region of the northern Black Sea, in the Dnepr–Don metallurgic focus (Chernykh 1992: 127, Fig. 44, n. 26, in relation, from Lola, in the Kalmykian focus, kurgan 4, burial 7; 129, Fig. 45, n. 12, from Voikovo, in the Dnepr-Don basin).

There are no precise comparisons for the incised silver bracelets from the necropolis of Parkhar, in the Kyzylsu valley: there are no similar specimens, only single elements which appear separately on different kinds of objects. There are no analogies, for example, for the whole decorative pattern but only for parts of it: the four petals rosette, the only well documented pattern on the bracelets, is a known in the 4th millennium glyptic of Mesopotamia and Susiana (Amiet 1961: 40 and Pl. 30, ns. 478, 479, this from Iran, 482, from Tello, 489 from Susa: according to Amiet the typology of these seals, in glazed steatite, was originary of south-eastern Iran; Amiet 1972: I, 153, n. 1121, see also 154, ns 1133-1134, II, Pl. 26, ns 1121, 1133-1134, Protoelamite Period 3100-2800 BC); the four petals rosette is present, in Central Asia, on the small chlorite cosmetic bottles from southern Bactria (Pottier 1984: 54, Fig. 20, penultimate row, the 3 rd from the left, 55, Fig. 21, penultimate row, the 4 th from the left), or sometimes on stone stamp seals-amulets, as on a specimen from Margiana (Sarianidi 1986: 222, at the bottom of the page). A four petals rosette is also incised on a finger ring from Early Dynastic Mesopotamia (Musche 1992: Taf. XXXI, n. 3.2 from Girsu), this is the only pattern documented in iconography. As I told, there is no precise parallel for the decoration of these bracelets. The only object with incised decoration which associates the rosettes with geometrical patterns is a silver diadem from the necropolis of Gonur, on which are represented three multi petal rosettes, a dotted zig-zag pattern and two double dotted parallelograms (Sarianidi 2001: 62, Fig. 28) but from the iconographic and also technical point of view it has little in common with our specimen an cannot be compared to our bracelets. The only other element which has a comparison is the double spiral fastening which can be compared with that of a bracelet from the cemetery of Bylim in the Caucasus (Motzenbäcker 1996: 40, Abb. 15, n. 23, datable between the end of the Middle and the beginning of the Late Bronze Age). It is therefore difficult to date the bracelets from Parkhar: for the presence of the rosette pattern they could be dated between the end of the 3rd and the beginning of the 2nd millennium, the spiral fastening, however, is similar to that on the later bracelet from Bylim. A possible dating could be to the first centuries of the 2nd millennium.

There are some comparisons in Central Asia for the bronze compartimental seal in the Khulbuk museum, decorated with the design of an asymmetrical cross made of parallel segments in a round field (Fig. 1, n. 7), particularly in southern Bactria, in the Dashly oasis, the same pattern, in a square field, is also present on some seal impression from Shahdad (Sarianidi 1977: Fig. 55, n. 11; Amiet 1990: 168, Fig. 13,l; Salvatori 1990, 183 and color Figs. 55, the last seal to the right and 56, the last to the right; see also Baghestani 1997: 348, abb. 108, ns 553 and 555 from a private collection, probably found in Dashly 3, beginning of the 2nd millennium BC, 229, abb. 75, ns 368-370 from Shahdad, end of the 3rd millennium; Masson 1988, Pl. XVII, n. 7, square shape, from Altyn Tepe; Sarianidi 1981: 179, Fig. 8, n. 8 from Margiana). There are comparable stone seals at Susa B (Rashad 1990: Taf. 30, ns 792, in gray stone, 793, in steatite; see also Taf. 43, n. 1127, from the antique market, Ubaid Period). Other parallels have been found in the Middle bronze Age II, at Alishar (Baghestani 1997: 134). The

We shall now examine the pottery. The ‘square vessel’ (Fig. 2, n. 7), perhaps the stylized representation of a house, has many comparisons in the Zamam Baba culture, dated to the Namazga V Period, as Vinogradova has written (see Vinogradova’s paper here; Amiet 1990: 161), there are moreover other comparable specimen in southern Bactria, in the same age (Sarianidi 1986: 107, 186

IS IT POSSIBLE TO SPEAK ABOUT MIDDLE BRONZE AGE IN SOUTHERN TAJIKISTAN ? 110; Amiet 1990: 160, Fig. 2a-b, 161). A vessel of the same type, only with four compartments, is documented at Shahdad (Hakemi 1997: 580, Fig. Dg 1).

and Margiana (for n. 8 see Götzelt 1996: Taf. 38, n. 512 from Altyn Tepe; for n. 9 see Sarianidi 1990: 277, Pl. LII, n. 15 and Pl. C, n. 1, from Togolok 21, end of Namazga V-beginning of Namazga VI; for n. 11 see Götzelt 1996: Taf. 77, n. 979 w from South Kelleli oasis, ‘Complex 2’; Sarianidi 1990: 207 Pl. II, n. 1, from Auchin 1; for n. 10 see Götzelt 1996: Taf. 44, n. 578 c from Togolok 1, n. 580 a from Adži Khui; Masson 1988: Pl, XXXIV, n. 6 from Altyn Tepe; Udemuradov 1993: 49, Fig. 17, in relation; Sarianidi 1990: 276, Pl. LXXI, n.7, in relation for the shape but the spout in this specimen is open.).

All the three vessels from the cemetery of Krugsai (Fig. 2, ns 1-3) are comparable to the pottery assemblage of the sites of the Dashly oasis (Sarianidi 1976. 60, Fig. 37, n. 11, for n. 2, n. 5 for n. 3; Figs 40, n. 13 and 43, n. 19, although the foot is a little taller and slender, for n. 1); the same types can be found in the necropolis of Gonur and in Altyn Tepe (Sarianidi 2001: 150, Pl. 15, group 1, the first row of vessels and the last two on the right, 152, Pl. 17, group 2, the 3rd from the right for n. 1, 150, Pl. 15, group 2, the last row for n. 2, 151, Pl. 16, group 4, for n.3; Götzelt 1996: Taf. 16, n. 208 for n. 8, from Altyn Tepe).

The last of the considered group of objects is a vessel made of onyx (Fig. 2, n. 6), in the tradition of the vessels in semipretious stones spread all over the Middle and Near East particularly in the 3rd millennium BC, also characteristic of the Bactrian civilization and, in general, of the Namazga culture. Comparable vessels, although slightly squatter in the shape than our specimen, have been found at Sapalli Tepe (Askarov 1977: Pl. XXIV, ns 1-2, Sapalli phase). After having examined our objects in order to understand if they can be set in the Central Asian Middle Bronze Age context, and to what extent, what can we answer to the question “Can we speak about Middle bronze Age in Southern Tajikistan ?”.

The clay female figurine (Fig. 2, n. 4), discovered outside the burial where the three vessels were found is compared by Yusupov, as Vinogradova writes, with the clay figurines from the Geoksjur oasis, of the Namazga II-III Period (Vinogradova, 2012) there is some other similar specimen (Kohl 1984: Pl. 5a, the two ‘foot shaped’ figurines, from Geoksjur I, Namazga II-III Period) a seated female figurine from Sarazm could also be comparable, although the stylization of the head is different (Isakov 1991: Fig. 68, n. 1, Sarazm I, 3500-3300 BC).

Among the metal objects the three axes (Fig. 1, ns 3-5) belong to very ancient typologies, in some cases dating back to the Chalcolithic and Early Bronze Ages, which did not originate in Central Asia but have been brought there. Their area of origin is, as we have seen (see before), a vast area comprised between South-eastern Europe, and the Iranian Plateau. The closest analogies are with the axes of the Carpatho-balkan area, the region to the North of the Black Sea, the Caucasus, Transcaucasia, the areas of Susa, in South-western Iran and Tepe Ghabristan, in northern Iran. The date of the comparable specimen ranges between the second half of the 4 th and the 3rd millennium BC, some later specimen reach the first centuries of the 2nd millennium. The axes belonging to the typology of the one from Shar-Shar, for example, which have been found in Central Asia, are too numerous to think that they are only imports: the mould for a variant with shaft hole of this type has been found at Dashly 3 (Sarianidi 1978: 188, Fig. 3, 189, Fig. 4). Moreover some of these axes, from Central Asia, like those from Sarazm, can be dated to the last centuries of the 4th - firsts of the 3rd millennium BC, a very ancient documentation of the presence of this typology in Tajikistan. The type of the Shar-Shar axe becomes rather common in the metallurgy of Bactria and of the nearby regions by the end of the 3rd millennium.

For the vessel in the Kulyab Museum, (Fig. 2, n. 5) a spread type in the Namazga V horizon, there are several comparisons in Altyn Tepe (Kircho 2004: 143, Fig. 1, Pottery, first row, the last one; Masson 1988: Pl. XXXIII, n. 20; Götzelt 1996: Taf. 55, n. 722 n, q, Taf. 59, n. 757). In the 2007 excavations, made by Vinogradova in the site of Gelot, in the Kuljab district, a catacomb burial (burial 2) containing Namazga V pottery has been discovered (see further, Fig. 2, ns 8-11 and Vinogradova, 2012) and this is a documentation of the presence of the Namazga V-Sapalli people in Southern Tajikistan at a previous time than the Late Bronze Age. With regard to these vessels we can say that there are comparisons for each of them in the BMAC area, many of them in Bactria, particularly in the area of the Sapalli culture, at Sapalli Tepe itself, Djarkutan I, at Dashly 1, and 3 in the south (for n. 10 see Askarov 1973: 161, Pl. 22, n. 10, in relation; for n. 11, Askarov 1973: 155, Pl. 15, ns 5, 10 from Sapalli Tepe; Shirinov, Baratov 1997: 69, Abb. 2, n. 8, from the necropolis 4c of Djarkutan, Shirinov 2002: Djarkutan I phase; 116, Abb. 71, n. 5, from the necropolis, Djarkutan I phase; Sarianidi 1976: 29, Fig. 18, n. 4, Fig. 19, n. 8 from Dashly 1, Fig. 33, n. 9 from Dashly 3, Sapalli-Djarkutan I phases; for n. 9 see Shirinov 2002: 118, Abb 73 B, ns 5-7, from the settlement, Djarkutan I phase; some specimen in relation can be found also more to the North in the Zeravshan valley (Bobomullaev 1997: 124, Abb. 2, n. 1, a vessel from a grave in Zardča Chalifa (end of the 3 rd- first quarter of the 2nd millennium BC), although larger than our n. 8). Several comparisons for every type of the group can be found, of course, also in Southern Turkmenistan

The other two axes were perhaps imports from southeastern Europe, Caucasus or Transcaucasia, their types are scarcely documented in Central Asia but belong to a period ranging from the second half of the 4 th and the beginning of the 3rd millennium and cannot be later than the first quarter of the 2nd millennium, so the presence of the axes from Arakchin and Sangvora in Southern Tajikistan can be part of the evidence of an earlier occupation of the area. 187

Giovanna Lombardo but also for the ethnic reality that the Sapalli population found there, where the presence of the nomadic tribes of Andronovo and of the other related ‘steppe’ cultures was stronger than in Uzbekistan and Afghanistan: this new situation could have changed more than an aspect of the original Sapalli culture. In any case, whether the urban settlements are to be found or not, what is really important is that the culture in its different aspects, in the pottery technology, in the metallurgy and in the art and handicraft is still recognizable as belonging to the Namazga-Sapalli horizon, we could say that there is a Tajik aspect of the Sapalli culture (Lombardo, in Vinogradova, Lombardo, Götzelt, forthcoming; Lombardo 2007, forthcoming).

The type to which the dagger (Fig. 1, n. 6) belongs is documented, as we have seen (p. 3), between the second quarter and the end of the 3rd millennium, in the Pit Grave culture of the Volga-Urals and Lower Dnepr area, also in the second quarter of the 3rd millennium in the Bedeni culture of Southern Caucasus and in the first half of the 2nd millennium in Kalmykia and NorthernCaucasus. This type of dagger is also widely spread in Central Asia, since the Namazga III Period at Sarazm and in the whole BMAC. The bronze stamp seal, the onyx vessel and the square ‘house model’, too, seem to date back to earlier periods then the Late Bronze Age, i.e. at least to the end of the 3 rd and the beginning of the 2nd millennium, and the clay female figurine, finally, is connected to Namazga II-III Periods: it is difficult to interpret the presence of this figurine, in the Namazga V context of the cemetery of Krug-sai, unless we venture the hypothesis that there are deeper levels in the site, dating back to Namazga II-III Period.

The answer to our previous question could be that “yes, there is some evidence of a Middle Bronze Age occupation of Southern Tajikistan” but, with the scarce documentation we have at present, we are not able to trace a picture of the cultural development in the region before the Namazga VI Period. We are only at the beginning of our research: to obtain such a picture will be necessary to continue the excavations in Gelot, in the whole Kuljab district in general, and in the other areas of Southern Tajikistan where the stray finds have been brought into light.

To sum up, we have on one side single objects, all of them stray finds, dating from Namazga V or perhaps even earlier, scattered in the territory of Southern Tajikistan: without claiming to date some of the most ancient back to the Copper Age, we would say that their presence in various sites of Southern Tajikistan is meaningful, at least in regard to the hypothesis of the peopling of the area by the Sapalli groups at an earlier age; on the other side we have a number of items, like the two silver bracelets and the pottery vessels, which come from regular excavations and are part of grave furniture. Leaving aside the bracelets, of which is difficult to establish a date, the vessels belong to a precise context of Namazga V Period. They are, of course, the main ground to suppose that there has been an occupation during the Middle Bronze Age, at least, in the Kuljab area, which is not so distant from Northern Afghanistan and from the Dashly oasis where the sites have levels of the Sapalli – Djarkutan phases, i.e. of Namazga V. How far this Namazga V occupation extends in Southern Tajikistan we cannot say at the moment. It will be difficult, in any case, to establish exactly when the farming tribes arrived in Tajikistan, with the few documents we have, but it seems reasonable to hypothesize that the moment of this arrival could have been at least during the Namazga V period, i.e. during the last centuries of the 3 rd and the firsts of the 2nd millennium BC.

Acknowledgements I would like to thank Dr. Vinogradova for letting me use the pictures of the Gelot campaign.

Bibliographical References Amiet,

P. (1961) La glyptique Mésopotamienne archaïque. Paris Amiet, P. (1972) La Glyptique Susienne. Mémoires de la Délegation archéologique en Iran, Vol. 43. Paris. Amiet, P. (1986) L’âge des échanges inter-iraniens 35001700 avant J.C., Paris. Amiet, P. (1990) ‘Antiquities of Bactria and outer Iran in the Louvre Collections’. In Ligabue G., Salvatori S. (eds), Bactria, the Ancient Oasis Civilization from the Sands of Afghanistan, Venice: 159-180. Askarov, A. A. (1973) Sapalli Tepa. Tashkent. Askarov, A.A. (1977) Ancient Farming Culture of the Bronze Age in Southern Uzbekistan. Tashkent (in Russian). Baghestani, S. (1997) Metallene Compartimentsiegel aus Ost-Iran, Zentralasien und Nord China. Rahden/Westfalien. Bobomullaev, S. (1997) ’Ein bronzezeitliches Grab aus Zardča Chalifa bei Pendžikent (Zeravšan-Tal)’. Archäologische Mitteilungen aus Iran und Turan B 29: 121-134. Chernykh, E. N. (1992) Ancient Metallurgy in the USSR. The Early Metal Age. Cambridge.

The Namazga V and the Sapalli cultures were both urban civilizations and, up till now, no urban centres have been discovered in Southern Tajikistan, this could mean either that further excavations are necessary to find them or that the Sapalli culture, characterized in Southern Uzbekistan and Northern Afghanistan by the imposing settlements of Sapalli, Djarkutan, Molali, Dashly 1 and 3, had undergone a deep change in the passage to Tajikistan. This had probably happened after the arrival of the farming groups in the new country, which was definitely different from the original homeland, not only for the nature of the territory, more mountainous in Tajikistan, 188

IS IT POSSIBLE TO SPEAK ABOUT MIDDLE BRONZE AGE IN SOUTHERN TAJIKISTAN ? Deshayes, J. (1960) Les outils de bronze del’Indus au Danube (IV e – II millénaire), Paris. Goetzelt, T. (1996) Ansichten der Archaeologie SudTurkmenistans bei der Erforschung der Mittleren Bronzezeit (Periode Namazga V). Deutsches Archaeologisches Institut. Berlin. Hakemi, A. (1997) Shahdad, Excavations of a Bronze Age Center in Iran. Rome. Isakov, A. I. (1991) Sarazm. Dushanbe. (in Russian) Kaniuth, K. (2006) Metallobiekte der Bronzezeit am Nordbaktrien. Berlin. Kircho, L. B. (2004) ‘Formation of the Most Ancient Protourban Civilizations of the Bronze Age of Central Asia (on Altyn Depe Materials)’. In Margiana Archaeological Expedition (eds), Near the Sources of Civilizations. The Issue in Honor of the 75-Anniversary of Viktor Sarianidi, Moscow: 142-160. (in Russian) Khlopin, I. N. (1986) Jung Bronzezeitliche Graeberfelder in Sumbar – Tal, Sudwest Turkmenistan, AVA Materialen 35, Muenchen. Kohl, Ph. L. (1981) (ed.) The Bronze Age Civilization of Central Asia. Recent Soviet Discoveries. New York. Kohl, Ph. L. (1984) Central Asia. Palaeolithic Beginnings to Early Iron Age. Paris. Kuzmina, E. E. (1966) ‘Metal Objects in the Neolithic and Bronze Age in Central Asia’. Svod Archeologishesky Ishtosnikov, 4-9. Moscow. (in Russian) Isakov, A. I. (1991) Sarazm. Dushanbe. (in Russian) Ligabue G., Salvatori S. (eds) (1990) Bactria, the Ancient Oasis Civilization from the Sands of Afghanistan. Venice Lombardo, G. (forthcoming) ‘Typologische und Vergleichende Fundanalysen’. In Vinogradova, N. M., Lombardo, G., Götzelt, Th, et alii (eds) The Ackerbausiedlung und Nekropole von Kangurttut, Archälogische Untersuchungen in Kangurttut (Südwest-Tadžikistan). Lombardo, G. (2007) ‘La dialettica tra le diverse culture nel Tajikistan meridionale nell’Età del Bronzo Tardo – Ferro I (secondo quarto del II - inizi del I millennio a.C)’. In Fedi, P. (ed.) “…. Alla maniera di “ Convegno in ricordo di Maria Teresa Lucidi” 28-30 maggio 2007, Dipartimento degli studi orientali, Facoltà di Lettere e Filosofia, Università di Roma “La Sapienza” (forthcoming) Masson, V. M. (1988) Altyn Depe, Philadelphia Motzembaecker, I. (1996), Sammlung Kossnierska – Der Digorische Formenkreis der Kaukasischen Bronzezeit. Berlin. Musche, B. (1992) Vorderasiatischer Schmuck von den Anfangen bis zur Zeit der Achaemeniden (ca 10.000-330 v. Chr.). Leiden. Pottier, M. H. (1984) Matèriel Funéraire de la Bactriane mèridionale de l’Âge du Bronze. Paris.

Rashad, M. (1990) Die Entwicklung der vor-und frügeschlichenStempelsiegel in Iran: Im vergleich mit Mesopotamien,Syrien und Kleinasien – von ihren Anfängen bis zum Beginn der 3 Jahrtausend v. Christ. Berlin. Salvatori S. (1990) ‘Early Bactrian Objects in Private Collections’. In Ligabue G., Salvatori S. (eds), Bactria, the Ancient Oasis Civilization from the Sands of Afghanistan. Venice: 181-187. Sarianidi, V. I. (1976) ‘Issledovanie pamjatnikov Dashlinskogo Oazisa’. In Kinglikova, I, T. (ed.) Drevnjaja Baktrija, 1. Moscow. Sarianidi, V. I. (1977) ‘A Bactrian Centre of Ancient Art’. Mesopotamia XII, Firenze: 97-110. Sarianidi, V. I. (1978) ‘Les plus anciennes haches de l’Afghanistan’. Sovetskaya Archeologhia, 2: 186194. (in Russian) Sarianidi, V. I. (1981) ‘Margiana in the Bronze Age’. In Kohl, Ph. L. (ed.) The Bronze Age Civilization of Central Asia. Recent Soviet Discoveries. New York: 165-193. Sarianidi, V. I. (1986) Die Kunst des Alten Afghanistan, Siegel, Keramik, Architektur,Kunstwerke aus Stein und Metall. Leipzig. Sarianidi, V. I. (1990) The Antiquities of the Country of Margush. Ashkhabad. (in Russian) Sarianidi, V. I. (2001) Necropolis of Gonur and Iranian Paganizm. Moscow. Schaeffer, C. (1948) Stratigraphie comparée de l’Asie occidentale. Oxford. Shirinov, T., Baratov S. (1997) ‘Bronzezeitliche Grabstaetten aus der Necropole Dzarkutan 4c (Süd Uzbekistan)’. Archäologische Mitteilungen aus Iran und Turan, B. 29: 65-120. Shirinov, T. (2002) ’Die Frühurbane Kultur der Bronzezeit im südlichen Mittelasien. Die vorgeschichtliche Siedlung Džarkutan’. Archäologische Mitteilungen aus Iran und Turan, B 34: 1-170. Sulimirsky, T. (1970), Prehistoric Russia, an Outline, Dublin. Udemuradov, B. N. (1993), Altyn-depe i Margiana: Svjazi, chronologija, proischoždenie. Ashkhabad. Vinogradova, N. M. (1993) ‘Interrelation between Farming and “Steppe” Tribes in the Bronze Age South Tadjikistan’. In Gail A. J., Mevissen G. J. (eds) South Asian Archaeology 1991. Proceedings of the Eleventh International Conference of South Asian Archaeologists in Western Europe, Stuttgart: 289-301. Vinogradova, N. M., Lombardo G. (2002) ‘Farming Sites of the Late Bronze and Early Iron Ages in Southern Tajikistan’. East and West, 52, nn.1-4: 71-125. Vinogradova, N. M. (2012) ‘Will the Monuments of the Middle Bronze Age be Discovered on the Territory of Southern Tadjikistan? ’. In. Frenez D. and Tosi M (eds) South Asian Archaeology 2007. Vol. 1, Prehistoric Periods. Proceedings of the 189

Giovanna Lombardo 19th International Conference of the European Association of South Asian Archaeology July 2007. (Ravenna, Italy, 2-6 July 2007) BAR International Series. Oxford, BAR Publishing: 281-286.

190

IS IT POSSIBLE TO SPEAK ABOUT MIDDLE BRONZE AGE IN SOUTHERN TAJIKISTAN ?

Fig. 1. – ns 1-2 : Silver bracelets with incised decoration, from the necropolis of Parkhar, Kyzylsu valley, Southern Tajikistan . n. 3: Copper axe-hatchet from Shar-Shar. n. 4: Copper axe from Sangvora. n. 5: Copper axe from Arakchin. n. 6: Bronze knife or dagger from Kuljab. n. 7: Bronze stamp seal from the kishlak of Boljuan, Khulbuk Museum

191

Giovanna Lombardo

Fig. 2 – ns 1-3: Pottery vessels from the cemetery of Krugsai. n. 4: Baked clay figurine from the cemetery of Krugsai. n. 5: Vessel from the Kuljab Museum. n. 6: Onyx vessel from the Kuljab Museum. n. 7: pottery model of a house from the Kuljab Museum. ns 8-11: Pottery vessels from a burial in the necropolis of Gelot.

192

THE LITHIC INDUSTRY FROM HAZAR SUM (SAMANGAN, AFGHANISTAN): NEW PERSPECTIVES AND OLD DATA Roberto Micheli

The Italian Archaeological Mission of IsMEO began its activities in Afghanistan in 1957. Till the end of the 1970s the excavations were carried out in the Ghazni province and focused on the Buddhist sacred area of Tapa Sadar (Verardi & Paparatti 2005; Filigenzi 2010) and on the Ghaznavid palace of Rawza (Giunta in press). Prehistoric research was limited to the first part of the 1960s and geographically centred on the northernmost slopes of the Hindu Kush mountain range in northern Afghanistan (Micheli 2010); it was focused on the Palaeolithic period and consisted in archaeological surveys in the Hazar Sum area (Castaldi 1963; Puglisi 1963) and excavations at the Darra-i Kalon rock-shelter (Mussi 1979).

The archaeological site A preliminary presentation of the results of the Italian activity at Hazar Sum was published in 1963 by Puglisi and Editta Castaldi on the quarterly East and West, but, unfortunately, no comprehensive analysis ever followed. Puglisi described Hazar Sum as a vast archaeological site of an urban type whose ruins cover an area of about 35 hectares, with a length of 700 metres along the northsouth axis and a width of about 500 metres along the eastwest axis. The archaeological research allowed to discover a large number of prehistoric stone tools and many remains attributed mainly to the Kushan and Islamic periods, among which megalithic monuments, dry wall features and more than 200 multi-roomed and multi-storied rock cave dwellings, where many paintings, bas relieves, and petroglyphs were found (Ferrandi 2010). The latest frequentation of the site coincides with the extensive destruction following the Mongol invasion in the first half of the 13th century AD; hereafter, the occupation has been occasional.

This paper presents a brief summary of the results at Hazar Sum enriched with new data: in fact, in 2006, during the moving of the IsIAO (former IsMEO) ‘Centro Scavi’ in Rome from the old building to the new, a box was found in the storehouse holding many prehistoric stone tools recovered during the early 1960s in the Hazar Sum area, and not yet studied. The analysis of these artefacts offers a reassessment of old data in a new perspective.

The characteristic cave dwellings of Hazar Sum find architectural comparisons with those of a more famous site located not far from Haibak, namely the Buddhist complex carved into the bedrock at Takht-i Rustam (Mizuno 1962) during the Kushan-Sasanian period, dated to the 4th and 5th century AD, with a later occasional occupation comprised within the 11th and the 12th century AD during the Ghaznavids and Seljuks (Ball 1982: 266). As it is known, the cave temples in Afghanistan are few and they are concentrated in the Jalalabad, Haibak and Bamyan regions perhaps because in the other provinces there were no suitable cliffs in which caves could be hollowed out. However, such architecture, particularly the Bamiyan and Haibak examples, is geographically the westernmost of all those known in Central and South Asia (Higuchi & Barnes 1997: 283, Fig. 2). Therefore, it is clear the importance of the rock cave dwelling complex of Hazar Sum, which will certainly require further research as soon as the political situation in Afghanistan improves.

The geographical setting Between 1961 and 1962, a team guided by Salvatore M. Puglisi carried out some surveys in the Hazar Sum valley in northern Afghanistan, about 16 kilometres north of Haybak city, now called Samangan, capital of the same province. The area is located at around 1000 metres above sea level, and dominates the valley of the Samangan river (Fig. 1.1). Hazar Sum was situated along the old caravan route stretching from Balkh to Bamiyan and entering the Kabul valley after crossing the Shibar or Panjifilan passes. It was an important junction on the old route which branched in the direction of the Kunduz valley and was presumably one of the ways connecting ancient India to the regions of Central Asia. The area is now a desert valley with a roughly south-north orientation formed by the erosion of the sand and limestone river banks. Nevertheless, the nature of the sediments favours the formation of numerous rock-shelters and caves along the valley sides that may have been employed by prehistoric peoples. Besides, the abundance of architectural and archaeological remains of later periods proves that the Hazar Sum area was in the past a good place for settlements along an important caravan route (Puglisi 1963).

Old data from the field: the chipped stone industry recovered in the 1960s The lithic artefacts found in the 1960s were surface findings recovered in the bottom valley during two survey campaigns, in the space of one or two kilometres around Hazar Sum. Puglisi’s note is brief and it does not offer a detailed study of the materials recovered; he described the industry as mainly of Clactonian aspect with some Levallois and Mousterian elements (Puglisi 1963: 5-6, figs. 3-6). The Clactonian types are obtained 193

Roberto Micheli retouched edges (Fig. 2.4-11), while the artefacts identified as true tools are only few and feature, among them, burins (Fig. 3.1), borers (Fig. 3.2-3), transversal and lateral scrapers (Fig. 3.4, 7-9, 10-11), points (Fig. 3.5) and notched tools; some picks of large size complete the collection (Fig. 1.2-3). The retouch can be generally direct or indirect and the edges often present a denticulate outline. Nevertheless, many specimens present an irregular arrangement of the detaches along the edges that appears to be more the product of chance than the result of intentional working; in several artefacts such pseudoretouch was produced after the formation of the patina and in some cases it caused notches and a denticulate outline of the edge. It is clear that the high incidence of pseudoretouch indicates a secondary deposition context of the artefacts on the surface and the exposure to weathering agents and agricultural practices (Micheli 2010).

from large flakes or thick laminar flakes of medium and small size. The flint used for these implements is darkbrown with some traces of patina. The artefacts present a large bulb of percussion with the typical, slight, radial fissures. The angle formed by the flake surface and the striking platform is normally obtuse. The edges present clear vigorous retouches that sometimes form denticulation; in some cases inverse retouches are documented. The most common tools are side-scrapers, discoidal scrapers, knives and few points on flakes; Puglisi recognized also some oval-shaped bifacial and a few thin blades delicately retouched that can be attributed to the Neolithic technology, because the flint used for the latter is grey or yellow-brown in colour and, therefore, it is different from that of Clactonian types. He also noted that numerous flint nodules split in half or having an initial flaking could be related to stone tools manufacturing and proved that this activity was made on the spot. The stone tools were indeed made of the locally abundant nodular flint quarried out from the outcrops along the sides of the wadi channels.

The series counts several artefacts that prove different stages of the working sequence on site. There are some totally corticated pieces and many flakes and blades partially corticated that show the first stage of cortex removing (Fig. 2.7, 9; 3.3). There are no true cores, although a subpyramidal preform is present (Fig. 2.1) along with some artefacts related to the reshapening of the cores, such as a rejuvenation of a portion of the platform and the detaching surface, three tabular rejuvenation flakes (Fig. 2.2) and one crested blade (Fig. 2.3).

On the basis of the artefacts typology and the raw material used, Puglisi supposed that most of the chipped stone industry of Hazar Sum could be ascribed to a very old occupational phase of the area between the Lower and the Middle Palaeolithic, while only few pieces could be assigned to a later frequentation during the Neolithic. His discovery was therefore particularly important at that time, because it increased the data on the Palaeolithic period in Afghanistan which until then was largely unknown.

Comparisons and closing remarks The absence of any clear reference context for the findings, the small number of peculiar specimens and the unusual morphology of the stone artefacts make the chronological and cultural determination of the Hazar Sum series difficult. Nevertheless, there are some objects that we roughly attribute to the Palaeolithic and Neolithic periods, while most of the series is unfortunately not so well-defined.

New data from the storehouse: the chipped stone industry recently uncovered in Rome The box found in 2006 counts 181 stone artefacts held in plastic bags labelled with the Hazar Sum name, but giving no information about the exact place where the stone artefacts were found and their arrangement on the surface. We suppose that such series had been collected in the same area where the materials published in 1963 were found and that it formed a part of the collection not yet studied in detail.

The peculiarity of the Hazar Sum lithic industry raises several questions that are difficult to answer and verify, considering the current situation of Afghanistan; in fact, it is likely that we will have to wait for a long time before we can resume prehistoric research there. However, we are inclined to believe that this archaeological assemblage is the result of several frequentations of the area over a long period of time, linked to high quality flint quarrying activities. This is suggested by the presence of some picks related to flint extraction, of several objects proving that the working process took place in situ and of a large number of blanks resulting from the huge exploitation of the local flint and showing pseudoretouch due to the disruption caused by natural agents or agricultural activities.

The new series includes stone artefacts made of different varieties of cryptocrystalline flint whose colour varies from dark grey or dark brown to light brown or buff colour. The implements have generally fresh edges, even if there is evidence of more or less intense alteration due to weathering effects or soil acidity in the form of a whitish patina (Bassetti 2010). The technological observations made by Puglisi in the 1963 are generally correct, but some further comments are needed. The débitage is simple, often perfunctory, and not easy to define. If we consider the artefacts preserving an intact butt, we can note that the majority has a natural or flat platform and the bulbs are often prominent. These data suggest that a simple technology was employed, probably based on the use of direct percussion with no preparation of the striking platform. The series is characterized by a high number of flakes, laminar flakes and blades with

Although Richard Davis (1978: 44) observed that “none of the artefacts presented by Puglisi were necessarily manufactured by any of Clactonian, Levallois or Mousterian techniques” and an on-the-spot inspection by 194

THE LITHIC INDUSTRY FROM HAZAR SUM (SAMANGAN, AFGHANISTAN): NEW PERSPECTIVES AND OLD DATA Franco 2008: Fig. 5.8-9), could be a clue in confirming this hypothesis.

him in the Hazar Sum area did not reveal any certain evidence of Middle Palaeolithic industries, we think, instead, that the new series includes some artefacts that could be actually attributed to such phase, as, for instance, an atypical Levallois point (Fig. 3.6), some transversal and lateral scrapers and in particular a flake scraper that seem to have been manufactured by the Levallois technique (Fig. 3.7), not so common in this part of Asia. In fact, although several sites have been classified as Mousterian or Middle Palaeolithic, only one in Afghanistan can be surely assigned to this phase, namely the Darra-i-Kur rock shelter in the Badakhshan province. The layer dated to around 30,000 years BP produced hundreds of stone tools made of a variety of basalt of classic Middle Palaeolithic types obtained with the Levallois technique. A characteristic of this assemblage is the high incidence of blades, but irregularly or partially retouched artefacts were also frequent. There appear to be only rare side scrapers, no bifacial tools and few Upper Palaeolithic types (Dupree & Davis 1972: 3031). Other attestations of the Levallois technique are known from the Ferghana basin in Uzbekistan (Davis & Ranov 1999: 190), from Punjab and Sindh in Pakistan (Salim 1986; Biagi 2006; Biagi & Starnini 2011) and from several sites in India (James & Petraglia 2005: S910).

We therefore think that the Hazar Sum lithic series documents several exploitation phases of the flint outcrops from the Middle Palaeolithic onwards, and it is possible that such exploitation continued until the last century, when flint was no longer used to manufacture stone tools as in prehistoric times, but rather was employed by nomadic pastoralists for the production of flint-lock and gun-lock, as it also happened in other places of Eurasia where rich outcrops of high quality flint were located. This overview shows how rich of potential prehistoric remains Afghanistan is and how much work has yet to be done. The new data about Hazar Sum expand the previous results published by Puglisi, confirming some of his observations and providing a starting point for future investigation in that region aiming at discovering new archaeological sites and at improving our knowledge of the Prehistory of those northern territories. Afghanistan is a fascinating country and it is a promising land for Asian prehistoric record. Let us hope that the political situation will soon improve, so that it will be possible to carry out further archaeological research, and, most important, so that there will be peace and security for Afghan people.

Indeed, it is difficult to compare the series with other Prehistoric assemblages of Afghanistan, since the Hazar Sum stone artefacts cannot be attributed to other Palaeolithic phases and only two regular blades with parallel sides and trapezoidal section might be ascribed to a Neolithic technology (Fig. 2.10-11). However, such series does not feature any characteristic types of the Upper Palaeolithic, as documented, for instance, at the Kara Kamar rock shelter, layer III, where the stone tools count notched pieces, retouched blades, end-scrapers and carinated scrapers re-used as bladelets cores (Coon 1957; Davis 1978: 50-55). Nor it features any types of EpiPalaeolithic sites, as recognise at Aq Kupruk II and III (Dupree & Davis 1972: 14-30), Darra-i Kalon (Mussi 1979; Micheli 2010), and Kara Kamar, layer I (Coon 1957); here, the stone industry counts burins, endscrapers of various types and denticulated tools and points, and was characterized by a high incidence of very fine microblades obtained from bullet-shaped cores that prove a very sophisticated technology based on the pressure technique and possibly involving heat treatment, which improves the débitage of some flint rocks as attested in several prehistoric sites of South Asia (Inizan & Lechevallier 1996).

Acknowledgements I would like to thank Anna Filigenzi, Director of the Italian Archaeological Mission of IsIAO in Afghanistan, and Massimo Vidale, Head of the Prehistoric Section of IsIAO ‘Centro Scavi’, for giving me the opportunity to study the Hazar Sum materials; Giovanni Boschian from the University of Pisa and Giampaolo Dalmeri from the Museo Tridentino di Science Naturali of Trento for their useful advice.

Bibliographical References Ball, W. (1982) Archaeological Gazetteer of Afghanistan. Paris. Bassetti, M. (2010) ‘The Lithic Industry of Hazar Sum: A Brief Geological Note’. In Filigenzi, A. and Giunta, R. (eds) The IsIAO Italian Archaeological Mission in Afghanistan 1957-2007. Fifty Years of Research in the Heart of Eurasia, Proceedings of the Symposium, Rome (January 8th 2008), Conferenze 21. Rome: 19-22. Biagi, P. (2006) ‘The Levalloisian assemblages of Sindh (Pakistan) and their importance in the Middle Palaeolithic of the Indian subcontinent’. In Archaeology of the Early North-eastern Africa, Studies in African Archaeology 9. Poznán: 10051017. Biagi, P. and Franco C. (2008) ‘Ricerche archeologiche in Balochistan e nel Sindh meridionale (Pakistan)’. In Gelichi, S. (ed) Missioni

Davis (1978: 44) documented, as Puglisi did, a tremendous amount of chipped debris of both Palaeolithic and post-Palaeolithic frequentation by people who used the excellent nodular flint that outcrops on the sides of the wadi channels in the Hazar Sum valley. The recovery of some picks which are probably related to the quarrying of raw material, as documented in other prehistoric sites in Pakistani Sindh connected to the exploitation of flint (Negrino & Mukthiar Kazi 1996: figs. 19.2, 20; Biagi & 195

Roberto Micheli James, H. V. A. and Petraglia, M. D. (2005) ‘Modern Human Origin and the Evolution of the Behaviour in the Late Pleistocene Record of South Asia’. Current Anthropology, 46: S3-S27. Micheli, R. (2010) ‘Prehistory of Afghanistan: The Italian Contribution and Research Perspectives’. In Filigenzi, A. and Giunta, R. (eds) The IsIAO Italian Archaeological Mission in Afghanistan 1957-2007. Fifty Years of Research in the Heart of Eurasia, Proceedings of the Symposium, Rome (January 8th 2008), Conferenze 21. Rome: 1-17. Mizuno, S. (1962) Haibak and Kashmir-smast, Buddhist Cave Temples in Afghanistan and Pakistan Surveyed in 1960 (in Japanese, with English summery). Kyoto. Mussi, M. (1979) ‘Darra Kalon Rock-shelter (Northern Afghanistan): Levels I – IV’. East and West, 29 (1-4): 189-220. Negrino F. and Mukhtiar Kazi, M. (1996) ‘The Palaeolithic Industries of the Rohri Hills (Sindh, Pakistan)’. Ancient Sindh, 3: 7-38. Puglisi, S. M. (1963) ‘Italian Archaeological Mission in Afghanistan. Preliminary Report on the Researches at Hazar Sum (Samangan)’. East and West, 14 (1-2): 3-12. Salim, M. (1986) The Middle Stone Age Culture of Northern Pakistan. Lahore. Verardi, G. and Paparatti, E. (2005) ‘From Early to Late Tapa Sardar. A Tentative Chronology’. In Callieri, P. and Filigenzi, A. (eds) Maurizio Taddei Memorial Volume, East and West 55 (1-4): 405444.

archeologiche e progetti di ricerca e scavo dell’Università Ca’ Foscari Venezia, VI Giornata di Studi (Venezia, 12 May 2008). Venezia: 9-18. Biagi, P. and Starnini, E. (2011) ‘Neanderthals at the South-Easternmost Edge: the Spread of Levalloisian Mousterian in the Indian Subcontinent’. In Biró, K. T. and Markó, A. (eds) Emlékkönyv Violának. Papers in Honour of Viola T. Dobosi. Budapest: 5-14. Castaldi, E. (1963) ‘Preliminary report on the researches at Hazar Sum (Samangan): Italian archaeological mission in Afghanistan’. East and West, 14 (3-4): 183-205. Coon, C. S. (1957) The Seven Caves: Archaeological Explorations in the Middle East. New York. Davis, R. D. (1978) ‘The Palaeolithic’. In Allchin, F.R. and Hammond, N. (eds) The Archaeology of Afghanistan from earliest times to the Timurid period. London: 37-70. Davis, R.D. and Ranov, V.A. (1999) ‘Recent Work on the Palaeolithic of Central Asia’. Evolutionary Anthropology, 8 (5): 186-193. Dupree, L. and Davis, R. D. (1972) ‘The Lithic and Bone Specimens from Aq Kupruk and Darra-I-Kur’. In Dupree, L. (ed), Prehistoric Research in Afghanistan, Transactions of the American Philosophical Society, 62 (4): 14-32. Ferrandi, M. (2010) ‘Out of History: Themes and Symbols in the Hindu Kush at the Fringes of the Main Cultural Traditions’. In Filigenzi, A. and Giunta, R. (eds) The IsIAO Italian Archaeological Mission in Afghanistan 1957-2007. Fifty Years of Research in the Heart of Eurasia, Proceedings of the Symposium, Rome (January 8th 2008), Conferenze 21. Rome: 23-40. Filigenzi, A. (2010) ‘The Buddhist Site of Tapa Sardar’. In Filigenzi, A. and Giunta, R. (eds) The IsIAO Italian Archaeological Mission in Afghanistan 1957-2007. Fifty Years of Research in the Heart of Eurasia, Proceedings of the Symposium, Rome (January 8th 2008), Conferenze 21. Rome: 41-58. Giunta, R. (2010) ‘New Epigraphic Data from the Excavations of the Ghaznavid Palace of Mas‘ūd III at Ghazni (Afghanistan)’. In Callieri, P. and Colliva, L. (eds) South Asian archaeology 2007: Proceedings of the 19th Meeting of the European Association of South Asian Archaeology in Ravenna, Italy (2nd-6th July 2007). Vol. II, Historic periods. Oxford. Higuchi T. and Barnes, G. (1995) ‘Bamiyan: Buddhist cave temples in Afghanistan’. World Archaeology, 27(2): 282-302. Inizan, M.-L. and Lechevallier, M. (1996) ‘Deux traditions techniques de la pierre taillée dans le sous-continent indo-pakistanais : le débitage par pression e le traitement thermique’. Paléorient, 22 (1): 145-152.

196

THE LITHIC INDUSTRY FROM HAZAR SUM (SAMANGAN, AFGHANISTAN): NEW PERSPECTIVES AND OLD DATA

Fig. 1 – Sketch map with the localization of the Hazar Sum area (1) and picks (2 and 3) (Drawings by R. Micheli).

197

Roberto Micheli

Fig. 2 – Hazar Sum: subpyramidal preform (1), tabular rejuvenation flake (2), crested blade (3), retouched or pseudoretouched flakes (3-7), retouched or pseudoretouched blades (8-10) and fragmented blade with retouch on fracture (11) (Drawings by R. Micheli).

198

THE LITHIC INDUSTRY FROM HAZAR SUM (SAMANGAN, AFGHANISTAN): NEW PERSPECTIVES AND OLD DATA

Fig. 3 – Hazar Sum: transversal burin (1), borers (2 and 3), transversal scraper (4), point (5), atypical Levallois point (6), scraper on Levallois flake (7), transversal scrapers (8 and 9) and lateral scrapers (10 and 11) (Drawings by R. Micheli).

199

HARAPPAN ARCHAEOLOGY IN THE GHAGGAR BASIN WITH SPECIAL REFERENCE TO THE SITES OF GIRAWAD AND FARMANA Vasant Shinde , Toshiki Osada , Akinori Uesugi & Manmohan Kumar

processes from the Early to the Mature phases (developed stage) and from Mature to Late phases (declined phase) have been identified, but the exact nature and reasons of their transformation is yet to be understood. A cursory glance at the distribution of the Harappan settlements suggests a heavy concentration of the Early and Mature Harappan and sporadic of the Late Harappan in the Ghaggar basin. This pattern suggests that the river was quite active during the first two phases and went dry in the last phase. The sites of Girawad and Farmana are ideal to pursue some of the problems mentioned above.

Introduction The Ghaggar/Hakra basin in the Northwest part of Indian Subcontinent was one of the most favoured zones of the Harappans right from the early phase dated roughly between the last quarter of the 4th to the beginning of the 3rd millennia BCE. Part of the river that flowed through the present political boundary of India is called Ghaggar and the same river is termed as Hakra when it enters into Pakistan. The Hakra river basin in Cholistan has been extensively and intensively surveyed by Rafique Mughal as a result more than 400 sites have been reported from that region (Mughal 1997). Sir Aurel Stein began survey in the Ghaggar basin in the late 1930s and subsequently A. Ghosh, J.P. Joshi, Y.D. Sharma, Suraj Bhan, R.S. Bisht, etc. and State Departments of Haryana and Punjab carried out sporadic work intermittently. The most systematic work till this date is that of Katy Frenchman (1972) who carried out systematic survey in parts of the Ghaggar basin for her Ph.D. research. Approximately 350 sites have been reported from the Ghaggar basin so far. While understanding settlement pattern of this region, one can see many gaps. These gaps are not because there were no sites, but because of the biased research strategy. As a part of this project, efforts were made to undertake systematic survey in the area previously not explored and also visit the known sites and document them systematically by collecting surface samples. This has enabled placing the sites in proper perspectives and reconstruction of various features like the pattern of their settlements, site typology, and relationship between sites, resource exploitation patterns, demographic and functional aspects.

The Ghaggar river basin lies mostly in the states of Haryana and Rajasthan. The parts of the Ghaggar basin region fall in a subtropical, semi-arid to sub humid, continental and monsoonal type of climate. The geological formation of the state of Haryana ranges from Precambrian to the recent times and can be divided into Aravalli system, Siwalik system and alluvial plains. Aravalli hills are the oldest formations present to the south and west of Chautang basin in present Bhiwani, Mahendragarh and Gurgoan districts. They are composed of quartzite, quartzitic sandstone, mica schist and crystalline limestone. The soils of this area have been formed by aeolian and fluvial processes. The main physiographic units in this area are Chautang Flood plains and Aeolian plains. The relict course of the river Chautang are filled-up with sediments causing differences in the physiographic positions. All the rivers but Yamuna in the state of Haryana are semi-perennial. The Ghaggar River system (which also includes Chautang) emerges from lesser Himalayas and piedmonts and passes through the Indian states of Haryana and Rajasthan and then enters into the Bahawalpur region of Pakistan. The river Ghaggar-Hakra has been identified as the ancient Saraswati and Chautang as Drishadvati very often referred to in the Rig Vedic period.

The earlier belief that the Harappan Civilization was homogenous has turned out to be a myth. Within the Harappan regions are found manifestation of the regional variations (domain according to Possehl 2002). It was noticed that the material culture associated with the Harappan culture at Rojdi showed some difference compared to that found in the Sindh-Baluchistan region, termed by Possehl as ‘Sindhi Harappan Domain’. This was found true for the whole Saurashtra region. This difference was treated as a regional variation of the Harappan culture in Saurashtra and termed as Sorath Harappan (Possehl and Raval 1989). Similar regional difference in the material culture, more particularly in the ceramic assemblages of the Harappan sites in the Ghaggar basin is visible. One of the aims of this research is to understand the causes of this variation in the Ghaggar basin. From the beginning of the Harappan culture in the middle of 4th millennium BC until its end in the middle of second millennium BC, continuous cultural processes were taking place from time to time. The

The archaeological survey carried out by the authors in 2007 in parts of Hanumangarh and Ganganagar districts of Rajasthan and Bhiwani and Rohtak districts of Haryana have recorded some of the sites with the help of GPS. Surprisingly, all the sites near Anupgarh area are actually located in the Ghaggar river course (Shinde et al. 2008). This is very interesting and suggests that the river Ghaggar has dried much before the emergence of the Harappan culture in this area. The true picture will emerge only when the all the known sites are visited and recorded with GPS. Excavations at Girawad (GRW) The site of Girawad (28 o 58’ 41” N 76 o 28’ 47” E), one of the important Early-Harappan sites in Haryana was 201

Vasant Shinde , Toshiki Osada , Akinori Uesugi & Manmohan Kumar and well made plastered floor inside. It is 2.20 m long and 3.75 m wide and is 10 cm deep. There are a few postholes on the periphery of this feature, indicating the presence of superstructure on this. Feature no. 4 was added to the pit-dwelling could be an extension of Feature nos. 37, 38A could be identified as garbage pits of this complex as they are small and irregular, circular on plan with rough sides and bottom. There is a shallow and elongated depression (Feature 37a) which could be meant for accommodating saddle quern. Feature no. 38B of this complex is the base of a round-based water storage pot. Feature nos. 5, 6, and 38, cylindrical in shape and small in dimensions, with sides and bottom smooth and plastered, could be storage pits of the complex.

selected for excavation mainly to salvage the remains, which otherwise would have been lost forever. The ancient site known as ‘kheri’ lies roughly 3 kms to the east of the present village Girawad, which falls in the jurisdiction of Meham Block of Rohtak District, Haryana. The village is roughly 22 km north-east of Meham town and about the same distance to the south-west of Rohtak, the district headquarters. The ancient site, circular on plan, is an extensive single culture one spread over an area of roughly 8 hectares (Fig. 1). When the team visited the site in early March 2007 a number of features, including a part of the mud platform, pits, large burnt patches, brick attribution and even a complete plan of couple of fire places (kilns) were visible at the base of the site. Most of the features visible on the base surface of the site are mostly confined to the southern side of the metal road. It was this area which was selected for rescue excavation.

Pit- Dwelling Complex No. 2: Feature nos. 7, 7A, 8, 8A & 2 form another complex at Girawad. Feature no. 7, irregular in shape and shallow with well made floor, could be identified as a pit-dwelling of the complex. Feature No. 2A (Tr. Nos.7 and 12), cylindrical with a diameter of 70 cm and a depth of 15 cm could be a shallow storage pit of the complex. Feature No. 2 (Tr. No.7) located 35 cm towards the north western side of 2A is circular (diameter 1.20m) and a very shallow 10 cm depth. As it is badly eroded, its exact use cannot be identified. Feature nos. 7A, 8 and 8A, cylindrical in shape and having smooth sides and bottom, could be storage pits of this complex.

The cultural sequence of the site cannot be reconstructed at this stage as the entire habitation deposit has been removed. However, it is clear from the ceramic assemblages that the first settlers were the Early Harappan Hakra people. The term Hakra culture was first used by Rafique Mughal for the early culture of the Cholistan region (1997). However, though principally the ceramic found at Girawad resembles the one found in Cholistan, there is a lot of variation in their surface treatment and shapes. This has been treated as regional variation of the Hakra culture and therefore the earliest phase at Girawad has been termed as Regional Hakra Culture Tradition. The first settlers at the site built a very strong mud platform, in which different features were made (Fig. 2). The mud platform was visible over an area of 50 m (EW) x 30 m (NS). All the features that have been excavated belong to the earliest occupation of the site. Based on comparative analysis with the site of Kunal, the site of Girawad can be dated to the middle of 4th millennium BCE. The outline of the features became visible immediately after scraping the surface of the area selected for excavation. The features which have been excavated in the designated area include circular or irregular oval shallow or deep pit-dwellings sometimes with post-holes on their peripheries, garbage and storage pits, small circular shallow depressions possibly for accommodating large jars, post-holes, full pots and storage jars survived in their bases and fire places. The functional analysis of these features is based on their nature, location and content. It was observed that the pitdwellings excavated formed different complexes and each complex consisted of features like storage and garbage pits, sometimes storage jars and even shallow pits identified as bases of large round-based pots. In all 60 different features were excavated and they were found to be associated with thirteen different pit-dwelling complexes identified in the excavated area. The following are the different complexes (Fig. 3).

Pit- Dwelling Complex No. 3: Feature nos. 3 and 9 are two circular on plan forming one big pit-dwelling in the shape of English numeral ‘8’. The other features of this complex could not be detected as they most probably lied to the north and northeast area, which is unexcavated. The diameter of feature no. 3 is 2 m whereas that of no. 9 is 3.05 m. The maximum depth of the pit dwelling along the western edge of pit no. 9 is 20 cm. Pit-Dwelling Complex No. 4: (pit nos. 11, 15, 12, 13, 17, 17A) - Feature 11, a large pear-shaped pit with perfectly vertical sides and flat bottom, can be identified as the main pit-dwelling of this complex. It is 2.90 m long and 2.20 m wide with a depth of 72 cm. The other features associated with this could be identified as storage pits (Nos. 15, 17, 17A), a depression for accommodating a round-base storage jar (No. 12) and a garbage pit (No. 13). Pit- Dwelling Complex No. 5: (Nos. 14B, 14, 14A, 14C, 40) - Feature 14B has been identified as a pit-dwelling of this complex. The other features of this complex include depressions for holding a round-base pot (No. 14C, 14A), garbage pits (No. 14, 40) and a storage jar. The dwelling pit of the complex is small. It is quite likely that the open space available within this complex may also have been used for carrying out lot of domestic activities. Pit-Dwelling Complex No. 6: (Nos. 39, 19, 18 and 14) Feature No. 39 appears to be a dwelling pit of the complex as it is the largest among the pits included in this complex and symmetrical. It is oval on plan and measures 2 m by 1.65 m and is 12 cm in depth. The sides are

Pit-Dwelling Compex No. 1: This complex consists of Feature Nos. 1, 4, 5, 6, 37, 37a, 38, 38a and 38b. Feature no. 1, oval in shape, has been identified as a shallow pitdwelling mainly on the basis of its relatively large size 202

HARAPPAN ARCHAEOLOGY IN THE GHAGGAR BASIN WITH SPECIAL REFERENCE TO THE SITES OF GIRAWAD AND FARMANA may have been an extension of pit-dwelling no. 27 as they share an expansion of the same floor level. These pits with smooth floor and bottom could be identified as pit-dwelling of the complex. Feature No. 27A, an oval in shape, measures 1.20 m by 1.10 m and has a depth of 15 cm. The sides and bottom are well-made and plastered and could have been used for domestic activity. Feature No. 26A a semi-circular on plan has a well-made base which could have been used for some domestic activity.

perfectly vertical and the floor is horizontal and well made. Feature Nos. 18 and 19 perfectly cylindrical (diameter. 80 cm) with a flat base and their sides and the bases plastered with clay appear to be storage pits of the complex. The Feature no. 14 appears to be a garbage pit of the complex as it contains pottery and bones and irregular, circular in shape. Pit- Dwelling Complex No. 7: (Nos. 20, 20A, 21, 16, & 17B, 22) - Feature Nos. 20 and 22 could be pit-dwellings of this complex as they are well-made with post-holes along their periphery. The surface is quite uneven. It is made of rammed earth but has developed depressions at places. Futures 17B, 20A and 21 could be storage pits as they are small cylindrical and No. 16, a small roughly circular pit, which is connected to Feature no. 22 may have been used for collecting dirty water resulting from cleaning of the Feature no.22.

Pit- Dwelling Complex No. 11: (Nos. 35A, 35, 34, 33A, 33B, 33) - Features 35A and 35 could be the main dwelling pits of the complex as they are symmetrical and have well-made sides and bases. Feature 34, irregular in shape could be a garbage pit. Features 33 and 33A cylindrical in shape may be the storage pits of the complex. In the open space of the complex is a depression for a round pot (33B). Pit -Dwelling Complex No. 12: (pit nos. 36, 47, 48) Only three features of this complex have been excavated. Feature 48, partially excavated, could be the main pitdwelling of this complex as its sides are lined with burnt clay lumps and appears large circular in shape. This complex may belong to a potter as Feature 36 is a pottery kiln of the complex. Feature No. 47, oval in plan has been excavated to a length of 1.50 m. The pit has an uneven but hard floor at a depth of 12 cm and may have been used for some domestic activities.

Pit-Dwelling Complex No. 8: (Nos. 42 - 43) - Feature No. 42 (Trench no. 24) that resembles a horse-shoe shape, measures 2.95 m by 1.60 m. It is not a pit dwelling in the real sense but can be termed as sunken-dwelling as it is hardly 3 cm deep. Three postholes are located on its periphery and it has a rammed and smooth floor. Feature No. 43 (Trench Nos. 24 and 30) located by the western side of the feature no. 42, is a perfectly cylindrical pit dug into the platform and has a diameter of 1.10 m and a depth of 50 cm. The sides and the bottom, which is slightly concave, are perfectly symmetrical and smoothened by plastering with clay. This can be identified as the storage pit of this complex.

Pit- Dwelling Complex No. 13: (Nos. 23, 50, 50A, 50B, 46) – This complex may belong to a potter as there is a pottery kiln (Feature 23) located in it. The other features of this complex include a pit-dwelling (No. 46), storage pits (Nos. 50, 50A and 50B) and a storage jar.

Pit-Dwelling Complex No. 9: (Nos. 28, 29, 31, 45, 44) Feature Nos. 28 and 29 are sunken dwelling of this complex as they are separated by a small ridge which may have acted as a partition between them. Feature 28 is slightly oval on plan oriented 30o northeast-southwest. It is 3.95 m long and the breadth varies from 1.95 m towards eastern end to 75 cm towards the western end. The feature no. 29 is made of two circles one large (dia. 1.50 m) and one small (dia. 80 cm). Both are connected to each other and therefore together appear to form a slightly bigger sunken dwelling of the complex. Feature No. 31, may have been used for carrying out some domestic activities as it has a well-made floor. Feature Nos. 44 and 45, cylindrical and oval in shape respectively, may have been storage pits of the complex.

Feature No. 46, though disturbed to a great extent, appears to be a large circular pit, only a small crescentshaped portion of which is visible. It is lined with 10 cm thick burnt clay plaster. This could be identified as a pitdwelling of the complex. The storage jar of the complex is crushed into pieces but the circumference of the pit which is 65 cm could define the size of the jar. Feature No. 50 is an oval pit measuring 1.20 m by 1.05 m. It has a depth of 45 cm and could be identified as a storage pit. Feature No. 50A is a small, circular (dia. 45 cm) and shallow (depth 10 cm) pit. This appears to be the depression used to fit-in the base of the storage jar. Feature No. 50B (Trench no. 27) is a cylindrical pit with a diameter of 50 cm and a depth of 35 cm. The sides are perfectly vertical and have a slightly concave base; both are well made and plastered suggesting it a storage pit of the complex.

Pit -Dwelling Complex No. 10: (Nos. 27, 27A, 27B, 32, 26, 25, 26A) – Feature No. 27 lined with clay lumps is oval on plan and measures 4 m in length (northwest) and 2.50 m (southeast) in width. Feature No. 27B (Trench 30), circular on plan with a diameter of 80 cm and 12 cm deep, is located immediately to the northeast of feature no. 27, which is directly connected to this feature as there is no separating line between them. Feature Nos. 25 and 26 (trenches 29, 36 and 37) are two circular pits joined together and form one long elongated dwelling. It measures 4.60 m by 2.50 m. Along the periphery are found a few post-holes. Feature No. 32 (Trench 37) the eastern edge of which merges with the floor of feature 27,

Pottery kilns Three pottery kilns have been discovered during the course of excavation at Girawad. One is a small open kiln and two are closed kilns. They are as follows: Pottery Kiln 1- Feature No. 36 is a small pottery kiln roughly oval but slightly irregular in plan. It is N-S 203

Vasant Shinde , Toshiki Osada , Akinori Uesugi & Manmohan Kumar Survey on the other side of the road at Girawad revealed the presence of mud-brick wall. The wall made in the typical Early Harappan ration of 1:2:3 appear to be of a large structure as it is quite wide (1.20 m). It is survived to a length of almost 3 m. The remains of bricks were found at the base of the habitation and thus contemporary with those pit-dwelling complexes found on the other side of the dividing road. This evidence suggests that the Regional Hakra Culture Tradition people at the site have also built mud-brick houses. It is quite likely that the area south of the road was occupied by the craftsmen, particularly the potters whereas the people of the higher social ranking lived on the northern side of the road. The complexes occupied by the craftsmen were provided with storage facility. There was greater need of storage in the dwellings of the craftsmen as they were paid in kinds for their services to the community.

oriented measuring 1.70 m by 1.45 m. It is survived to a depth of 35 cm. It has a concave base slightly deep in the centre; the base and the sides of the pit are burnt red. It contains a large number of broken pots of different varieties, burnt clay lumps and circular clay cushions. One of the clay cushions discovered in it has a diameter of 25 cm. It appears that the large storage jars were placed vertically and were supported by these cushions for firing. The ‘house complex’ presumably of a potter is located 30o to its north-western side and is represented by two pits numbered 47 and 48 which have been partially excavated. These pits are located at a distance of 2.10 m (Fig. 4). Pottery Kiln 2 – Feature No. 23 represents one of the closed pottery kilns of the Early Harappan phase at Girawad. This pottery kiln built in situ in a pit, is circular on plan and has a long rectangular fire chamber. It can be associated with the Pit-Dwelling Complex No. 13. The diameter of the circular portion of the kiln is 1.30 m. It was built of clay coils vertically placed along its periphery, the average thickness of which is 9 cm. The wall has existed to a maximum height of 22 cm. The fluted inner surface of the fire place is exactly similar to that of the Pottery Kiln No. 3 (Feature No. 30). Inside the kiln, almost in the centre, are located two rectangular slabs placed horizontally. The eastern slab measures 45 cm in length, 27 cm in width and is 13 cm thick. The other slab located on the western side at a distance of 33 cm is exactly parallel to the one on the eastern side but has only partially survived. The opening of this pottery kiln on the west is 40 cm wide. The inner surface of the opening was plastered. The fire chamber, made of rough rectangular clay lumps, is 1.10 m in length and 55 cm broad (Fig. 5). Inside the kiln were found burnt lumps of the clay wall, which contain impressions of rice, wheat and barley.

Girawad pottery The actual habitation deposit exposed belonged to the earliest phase of the occupation at the site with Pit or subterranean dwellings belonging to the Regional Hakra Culture Tradition of the Early Harappan phase. The ceramic assemblage at the site of Girawad constitutes a Red ware industry with varied surface treatment ranging from appliqué to incised decorations and a chocolate/black slip on the body. The ceramic industry as a whole seems to be closely related to the Hakra wares of the Cholistan desert (Mughal, 1997). The assemblage has been classified on the basis of the differential surface treatment since the fabric, finesse and shapes are similar throughout the industry. The Grey ware forms a very small part of the assemblage probably less than 1% while the bichrome (black and white) painted red ware accounts for an even smaller quantity. The vessels are made of medium fine fabric without any gritty inclusions though the presence of tempering in the form of fine sand is clearly visible. The clay is well levigated as it gives a compact feel to the surface and the core of the pots. The vessels are made on a slow wheel with uneven striation marks in different parts and use of luting techniques is visible in the manufacture of the complete vessel. Handmade vessels are also quite common using the ring method followed by scooping using a bamboo strip as the paring marks are clearly visible. In some cases the rims are wheel made with perfect striations luted to an uneven body with the finger impressions clearly visible. The vessels are light weight, medium to thin bodied except for storage jars which are heavy and have a thick body. The various wares identified include Mud Appliqué, Chocolate/Black slipped ware, Red ware, Incised Ware, Grooved or Cordoned, Periano reserve and Bichrome painted red ware. The basic shapes include medium to large globular pots with everted, rounded, outcurved and outturned rims with wide mouth or narrow mouth and very short neck. A large number of basins and troughs have also been found in this category including U-shaped large bowls/basins, handmade with a fluted rim and thick mud coating till the rim (Fig. 7).

Pottery Kiln 3 (Feature No. 30) - This pear-shaped pottery kiln is circular (dia. 1.50 m) on plan with a long fire chamber (1.05 m long) located toward its northeastern side. The fire chamber near its junction with the circular position is 90 cm wide and towards its end it is 45 cm wide. The total length of the feature from the tip of the fire chamber to the circular end is 2.42 m. The kiln was made in situ in a pit and had a 10 cm thick mud wall made of vertically placed clay coils, which is survived to a depth of 50 cm. Inside the kiln is found a debris of a burnt clay wall, suggesting it had high wall above the ground. These burnt clay lumps found inside have traces of wheat, barley and rice, which is confirmed by Archaeo-botanists Drs. Saraswat and M. D. Kajale (personal communication). This could be identified as pottery kiln as similar features have been reported from a number of pre/early-Harappan sites in India and Pakistan. In the centre of the kiln is a large roughly circular clay stem, 65 cm in diameter, made of two course of bunshaped clay lumps set in mud mortar and plastered with clay (Fig. 6). This stem may have been for supporting vessels inside the kiln.

204

HARAPPAN ARCHAEOLOGY IN THE GHAGGAR BASIN WITH SPECIAL REFERENCE TO THE SITES OF GIRAWAD AND FARMANA Earlier survey carried out by Vivek Dangi (2006) have brought to light objects like a couple of gold beads, a copper arrow head, a copper chisel, fragment of fish-hook and many other undefined copper objects, 20 beads of steatite, etc. The excavations have yielded terracotta globular beads, crude bone tools, perforated and unperforated pottery discs, terracotta bangle fragments, stone sling balls, copper rods, stone querns and pounders and terracotta cakes (Shinde et al. 2008).

small storage pit which was at a later time used for dumping materials. It appears that the lower half of all three pots was probably placed in a depression. The surface of the clay lined depression in which the individual pots were placed is burnt red. It is quite likely that an attempt was made to harden the surface of the recess by burning them.

Excavation at Farmana (FRN)

Along the northern section was found an oblong pit (the major portion of the pit lies to the north of the Trench which is left undug). It is located 1.50 m to the east of the western section. The pit is oriented 40 0 southwest northeast. It is 3.20 m long and excavated to a width of 90 cm. The pit was dug by the earliest occupants of the site at the base of layer (8) into the natural soil. The sides are perfectly vertical and smooth whereas the base is slightly concave in the middle. Two post-holes were noticed along the southern margin of the pit (Fig. 9). Taking into account the existence of post-holes and wellmade sides and bottom, it appears to be a pit-dwelling, exactly similar to those found in the earliest occupation level at Kunal.

Pit - Dwelling

The ancient site (29 o 02’ 22”N 76 o 18’ 021”E) that falls in the jurisdiction of three different villages; Farmana, Seman and Bhaini Chandrapal (Bada Bahen) all falling in Meham block of Rohtak district is locally known as Darksh Kheda. Since major portion of the sites lies in the jurisdiction of Farmana village, it is treated as a part of that village (Fig. 8). Farmana is quite big in size (18 ha) and qualifies to be termed as a town or small city. The site is extremely rich in cultural material and appears to have played an important role, both in social and economic organizations of the Harappans. One of the aims of the excavation was to study the cultural development from early-Harappan to Mature Harappan in this region and study the contribution of the Early Harappan in the making of Harappan regional manifestation in the Ghaggar basin. The excavations revealed two distinct cultural periods Period I - Early Harappan (Regional Hakra Culture Tradition and Period II - Mature Harappan. The extant habitation deposit at the site is 2 m, in which were found 8 habitation layers. On the basis of the nature of the layer and the content it appears that layers (7), (7a) & (8) belong to the Regional Hakra phase (total thickness 50 cm). Layer (6) appears to be a transition from Hakra to Mature and the first five layers belong to the Mature Harappan period.

Pottery The repertoire of pottery found in the Regional Hakra Culture Tradition level at the site is small compared to that of Girawad. The nature of pottery and certain shapes are exactly similar to that of the Girawad Regional Hakra Culture Tradition pottery and therefore the pottery from Farmana has not been elaborately discussed here. The Hakra Culture pottery in general is made on either slow turned table or wheel, either treated with slip or unslipped, made from fine clay and well fired. The various wares found at the site include wares like Mud Appliqué, Incised (either shallow or deep), Chocolate Slipped, Plain and Painted Red. In all these different categories of the shapes that are common include vases, bowls, globular pots, small globular pots with handles and cups. Storage jars are rarely found. The characteristic shapes in the Mud Appliqué ware include medium size storage jars and small globular pots with slightly flared out or everted rim. The Incised ware at the site is represented in two distinct categories - superficially incised decorations in combed pattern and deeply incised haphazardly decorated patterns. The shapes found in this category include wide mouthed basins, lota-shaped pots, handled globular pots and convex-sided bowls. The incised decorative linear or curvilinear patterns are found mostly on the outer surface but in case of wide-mouthed pots on inner surface too. It was also observed that the incised patterns in both the varieties of this ware are confined to the upper half, between the rim and waist portions and sometimes they are bound by horizontal black painted bands. The Chocolate-slipped ware, one of the characteristic of this culture, is found in considerably good quantity. The chocolate slip on the external surface is confined to the upper half of the vessel. The common shapes include globular pots, straight-sided vases, lota and bowls. The Red ware, both plain and painted, is close

Period I – Regional Hakra Culture Tradition Remains of Regional Hakra Culture Tradition were excavated in the Index Trench (No. 3) at the site. The presence of Regional Hakra Culture Tradition was confirmed on account of the presence of ceramic assemblage and pit-dwellings similar to the one found at Girawad and Bhirrana in the Fatehabad District of Haryana (Rao et al. 2004). Excavation in the Index trench no. 03 has yielded some features, one typical pit-dwelling and pottery usually associated with the Regional Hakra Culture Tradition. They are as follows: Features in Layer (8) in Trench 03 Three storage pots and one small cylindrical storage pit were discovered in the western half of this trench. Two pots and the storage pit in the section are straight in a line. All the three pots are globular medium of Coarse Red variety. The cylindrical storage pit has a diameter of 37 cm and is 50 cm deep. Sides are perfectly vertical and the bottom slightly concave. This could be identified as a 205

Vasant Shinde , Toshiki Osada , Akinori Uesugi & Manmohan Kumar survived to a thickness of 50 cm. Two courses of mudbricks of the wall have survived. One post hole on the inner side of the wall is noticed. The western wall of the structure which has also survived in good condition is 3.85 m long and 40 cm broad. It has survived to a total thickness of 22 cm. This wall continues beyond its corner towards the northern side suggesting the presence of multiple rooms.

to the Harappan Red ware. It is sturdy, made from very fine clay and well baked. Some pots are painted in black with simple geometric bands. Small vases, globular pots, storage jars, basins and occasionally bowls are some of the important types in this ware (Fig. 10). It should be stated that most of the Hakra pottery from Period I continues, albeit in small quantity, in Period II (Mature Harappan) at the site. The other material remains found in this period include micro and disc beads of steatite, terracotta beads and shell and terracotta bangles.

The floor of the structure is well made with bricks and clay. Inside the structure almost in the centre a typical Harappan steatite seal was discovered. The seal has a symbol of the water buffalo and a few Harappan letters on top (Fig. 12). Outside the northern wall of the structure are remains of a large dish on stand and a couple of medium sized globular pots. In all probability this is a domestic structure, of which only one chamber has been excavated.

Period II - Mature Harappan The Mature Harappan period at the site marks the culmination of the cultural process at the site, which is reflected in their settlement pattern and cultural material. The entire 18 ha area of the site was occupied during this period suggesting expansion of the population. Elaborate remains of well-built and planned brick structures, superior ceramic assemblages, presence of seal and sealing and rich cultural material excavated in a limited area, all point towards the flourishment in this period. The Mature Harappan phase can tentatively be dated to between 2500-2200 BC on the basis of seal and ceramic assemblages found in the excavations.

Structure No. 2 (Trenches 01, 01A, 01A1, 01B, Sealed by layer 1) (Fig. 13) This structure, rectangular in plan, is a large multiroomed brick structure, which is oriented 30 0 NW-SE. The orientation of this structure is in alignment with Structure 1 which belongs to the early Mature Harappan phase. This structure is divided into two parts by a partition wall. On the north-eastern side of the structure is a broad and large rectangular platform. At a later stage, the original walls were thickened by adding more brick courses on the outer side. Also in a later stage the western wall was extended further to form another complex. Two construction phases of this structure can be clearly identified.

Structures The remains of a few mud-brick structures excavated from the Mature Harappan levels at the site point to the adoption of a modicum of planning. The brick-size used is typical Mature Harappan in the ratio of 1:2:4. Two structural phases have been excavated. Structure 1 belongs to the Early Mature phase (c. 2500-2300 BC) and Structure 2 of the late Mature phase (2300-2200 BC). These structural phases have been identified on the basis of their relative stratigraphical positions and nature and sizes of the structures. The other features like construction method, brick size, nature and the type of raw material used for construction are common in both the phases.

Early construction phase The original structure lies over an area of 6.20 m by 4.20 m. All the four walls of the structure are intact and have been exposed. The structure has been divided into two parts by a partition wall. The chamber located on the northern side has been named 2A and one on the southern side 2B. The partition wall is 2.65 m in length (east-west direction) and 50 cm in breadth. Two brick courses set in yellow clay mortar of the wall are visible. The wall has existed to a thickness of 25 cm.

Structure No. 1 (Fig. 11) At the base of layer (3) a rectangular mud brick structure oriented 300 in the southeast - northwest direction was encountered in Trench 2. It is survived only by its foundation and lowermost brick course of the wall. The structure is almost complete as all its four walls have been excavated. From the inner side it measures 3.60 m (North-South) by 2.60 m (East-West). The northern wall of the structure, which is survived by a single course, is 2.20 m long, 30 cm wide and is survived to a thickness of 34 cm (inclusive of foundation). Two post-holes are visible on top of the northern wall.

Structure 2A: The northern part of the structure termed 2A is roughly squarish in plan measuring 3.05 m (northsouth) by 2.80 m (east-west) from the inner edge of the walls. This relatively large room may have been used as a dwelling. Structure 2B: To the south of the partition wall is located a small chamber-like room measuring 2.80 m (east-west) by 1.20 m (north-south) from the inner edge. This component of the structure appears to be a bath-room of the house as the eastern part of this chamber over an area of 1.10 m (east-west) by 1.20 m (north-south) is lined with wedge-shaped burnt bricks and one vertical course along the inner phase of the partition wall, which is plastered with 7 cm thick coating of fine yellow clay.

The southern wall has been exposed along the southern section. There appears to be an entrance (65 cm wide) in the southwest corner of the structure in the southern wall. The total length of this wall is 2.65 m and is 35 cm broad. The eastern wall, 4.45 m in length and 70 cm broad is 206

HARAPPAN ARCHAEOLOGY IN THE GHAGGAR BASIN WITH SPECIAL REFERENCE TO THE SITES OF GIRAWAD AND FARMANA almost parallel and symmetrical to the platform 1 and structure 2. A very small portion measuring 1.45 m (north-south) by 1.50 m (east-west) has been excavated in the north-western quadrant of the trench O1E. The platform was destroyed partly by a later pit. Since a maximum portion of this platform lies further south in an unexcavated area the other details and dimensions cannot be given at present.

Thus this part of the room appears to have been made water proof. In the remaining portion of this room covering an area of 1.48 m (east-west) by 0.96 m (northsouth) was prepared a floor of mud bricks. There appears to be a small mud platform by the side of the bathroom which was also plastered with a coating of yellow clay. Along the edge of the partition wall at a distance of 45 cm from the inner face of the western wall was made a small bench-like feature of a single vertical course of mud brick. It runs parallel to the wall to a length of 1.05 m and 27 cm breadth.

Pottery and other artefacts The Mature Harappan pottery found at the site looks closer to the classical Harappan pottery in terms of the fabric and shapes. The painted motifs, however, are much fewer compared to those found on the classical Harappan pottery. Besides the typical squarish steatite seal with incised water buffalo symbol and letters, the discovery of a clay sealing embossed with a unicorn and Harappan letters (Fig. 14), are some of the noteworthy finds of this season. A large number of terracotta bangle fragments, a few shell and faience bangle fragments, beads of lapis lazuli, steatite, plain and also etched carnelian, agate, paste, shell and terracotta, a few copper fragments, bones tools, large number of triangular terracotta cakes, mustika, grinding stones, etc. form the repertoire of the artefacts found in the excavation.

Later Phase of 2A (Tr. 01 and 01A): In the later phase of structure 2A a lot of new additions were made. The western wall was thickened further by adding three horizontal courses of brick on the outer edge. This added wall (55 cm broad) does not end at the south-western corner of the original structure (early phase) but continues further towards the south. It is exposed to a total length of 13.20 m. A similar form of extension is noticed in the northern wall. The additional wall incorporated to the original in the later phase has survived to its full width (55 cm). The total width of the wall now is 1.30 m. Platform No. 2: Parallel to the eastern wall of Structure 2 on the outer side a thick mud platform was added. It is 5.32 m long and its maximum width is 2.55 m. Four courses of bricks of the platform are visible in the eastern section. The edge of the platform is nicely made by arranging bricks vertically all along the edge, whereas inside some lines are arranged vertically and some horizontally. Two large postholes are found in the platform. This platform can be associated with Structure No. 2.

Concluding Remarks The site of Girawad is extremely interesting as this appears to be close to the eastern periphery of the Hakra Culture. The ceramic found in the Hakra levels show regional variations and therefore this has been termed as Regional Hakra Tradition Culture. We are treating the Hakra phase as Early Harappan in this region simply because most of the features of the ceramic assemblage continue into the Mature Harappan. This is the second Hakra Culture site in Haryana after Bhirrana which has been excavated systematically. Though carbon 14 dates are not available, it could be dated to the middle of the fourth millennium BC on the basis of its comparison with Bhirrana. The data on architecture from Bhirrana and Girawad suggests that the Hakra culture people lived in mostly pit-dwellings, which were shallow. Each pitdwelling complex consisted of other features like storage pit, garbage pit, water storage pits and storage pots. It appears that the open space available in each cluster was used for carrying out some domestic activities. The discovery of pit-dwellings and also traces of mud-brick structures at Girawad is significant as it indicates that both pit-dwellings and permanent structures in the form of rectangular mud-brick architecture were in use simultaneously. The pit-dwelling complexes excavated at Girawad were most probably occupied by the craftsmen, mainly potters as the discovery of three pottery kilns would indicate. The area on the other side of the road where the remains of mud-brick structures were noticed could have been occupied by the higher ranking people. This is the indicator of the presence of a class-structured society. The discovery of three pottery kilns in such a small excavated area in Girawad suggests that it was one of the major pottery manufacturing sites during the Hakra

Structure 2C: This structure labelled 2C is a room added to the original structure at a later stage. The floor of this room is well rammed and plastered, and is located at a depth of 40 cm from the top of the surviving platform. Platform No. 1: To the southern side of structure 2 was added a roughly squarish platform, which is oriented in alignment with the main structure. The brick lining along the periphery is arranged vertically all along. This platform is separated from structure 2 by a gap of 60 cm. It appears that this platform was built simultaneously along with structure 2 as the original western wall continues further to form the original western edge of the platform. The dimension of the platform measures 5.55 m (east-west) by 4.22 m (north-south). A later pit measuring 1.78 m (north-south) by 2.65 m (east-west) has damaged a part of the platform. The exact function of these two platforms cannot be easily identified as the contents have been lost because of the disturbances caused by ploughing. However considering the modern parallels it is quite likely that lot of activities were carried out on this platform. Platform No. 3 (Trench 01E): To the south of platform 1 at a distance of 43 cm is located yet another platform 207

Vasant Shinde , Toshiki Osada , Akinori Uesugi & Manmohan Kumar Culture in the middle of the 4th millennium BC. It is quite likely that Girawad supplied pottery to a number of other small sites located in its proximity. A few sherds of the Periano Reserve pottery, usually confined to the Baluchistan area, found at Girawad suggest that the people of Girawad had cultural or trading contact with far off regions.

(Occasional Paper 3), Research Institute for Humanity and Nature, Kyoto, Japan: 77-158.

The site of Farmana is an ideal candidate to study cultural processes from the Early-Harappan to the Mature Harappan culture as it has the required cultural sequence. The Regional Hakra culture tradition found at the base of the site is similar in all respect to that found at Girawad. There are indications that the Early-Harappan culture in this region has contributed substantially to the making up of the Harappan Civilization in this region. The continuation of the traits of the Early-Harappan into the Mature Harappan at the site could have resulted in the formation of a regional manifestation of the Harappan Culture in the Ghaggar basin. This aspect will be dealt in great details in future. The Mature Harappan phase at Farmana appears to be extremely rich as the discovery of structural remains, pottery and seals would indicate. The glimpses of modicum of planning and strong well-built brick structures at the site surely indicate that the Mature Harappan phase at the site was most prosperous. Extremely thick walls of the structure are also indicator of the presence of probably double storey houses at Farmana. Farmana appears to have been developed into a town in the Mature Harappan times from its modest beginning. It could have played an important role in the region, which needs to be investigated in future.

Bibliographical References Dangi, V. (2006) Settlement Pattern of Meham Block (Rohtak). Unpublished M.Phil Dissertation, University of Kurukshetra. Frenchman, K. (1972) Prehistoric Pottery Industries along the “Lost” Saraswati River of the Great Indian Desert. Unpublished Ph.D. Thesis, University of Pune. Mughal, M. R. (1997) Ancient Cholistan: Archaeology and Architecture. Lahore, Ferozsons. Possehl, G. L. (2002) The Indus Civilization: A Contemporary Perspective. New York, Altamira Press. Possehl, G. L. and Raval, M. H. (1989) Harappan Civilization and Rojdi. Delhi, Oxford and IBH and the American Institute of Indian Studies. Shinde, V., Osada T., Sharma M. M., Uno T., Maemoku H., Uesugi A., Sinha Deshpande S., Shirvalkar P., Kulkarni A., Sarkar A., Reddy A., Rao V. and Dangi V. (2008) ‘Exploration in the Ghaggar Basin and excavations at Girawad, Farmana (Rohtak District) and Mitathal (Bhiwani District), Haryana, India’. In Osada T. and Uesugi A. (eds) Linguistics, Archaeology and the Human Past 208

HARAPPAN ARCHAEOLOGY IN THE GHAGGAR BASIN WITH SPECIAL REFERENCE TO THE SITES OF GIRAWAD AND FARMANA

Fig. 1 – General Plan of Girawad.

Fig. 2 – General view of the excavated remains.

209

Vasant Shinde , Toshiki Osada , Akinori Uesugi & Manmohan Kumar

Fig. 3 – General plan of the excavated remains.

210

HARAPPAN ARCHAEOLOGY IN THE GHAGGAR BASIN WITH SPECIAL REFERENCE TO THE SITES OF GIRAWAD AND FARMANA

Fig. 4 – Pottery kiln 1 (feature n°36).

Fig. 5 – Pottery kiln 2 (feature n°23).

211

Vasant Shinde , Toshiki Osada , Akinori Uesugi & Manmohan Kumar

Fig. 6 - Pottery kiln 3 (feature n°30).

Fig. 7 – Pottery from Girawad.

212

HARAPPAN ARCHAEOLOGY IN THE GHAGGAR BASIN WITH SPECIAL REFERENCE TO THE SITES OF GIRAWAD AND FARMANA

Fig. 8 – General view of Farmana.

Fig. 9 – Remains from Period I at Farmana.

213

Vasant Shinde , Toshiki Osada , Akinori Uesugi & Manmohan Kumar

Fig. 10 – Pottery from Farmana, surface collection.

Fig. 11 – Structure n°1 at Farmana.

214

HARAPPAN ARCHAEOLOGY IN THE GHAGGAR BASIN WITH SPECIAL REFERENCE TO THE SITES OF GIRAWAD AND FARMANA

Fig. 12 – Steatite seal from Structure n°1 at Farmana.

Fig. 13 – Structure n°2 at Farmana.

215

Vasant Shinde , Toshiki Osada , Akinori Uesugi & Manmohan Kumar

Fig. 14 – Clay sealing from Structure n°2 at Farmana.

216

AUTOCHTHONOUS LATE BRONZE AGE CULTURES IN CENTRAL ASIA: WHERE ARE THEY? T. Sh. Shirinov, S. R. Baratov

‘Fedorovo’ (Fig. 2-5) and ‘Alekseevo’ (Fig. 6) cultures during Late Bronze Age.

The territory of Middle Asia being rich in such minerals as copper, tin, silver and gold, semi-precious stones and also in narcotic plants needed for ancient civilizations in Front East and India drew attention of enterprising traders established settlements and trading stations (Namazgadepe, Altyn-depe, Geoksjur, Sarazm, Shortugaj Tepe Jakh’ja and Shakhr-I Sokhte). The obtained raw materials were processed at these trading stations and transported into regions of empire centres in Elam, Mesopotamia, Asia Minor and India. The researchers of settledagricultural sites of Middle Asia have acknowledged their alien character linked to ancient Oriental civilizations by basically trade relations (V. M. Masson, V. I. Sarianidi, etc.). With time the trading stations were developed into well-organized urban cultural centres (Gonur, Togolok, Dashly 3, Sapallitepa, Dzharkutan and many other) passing the intellectual and technological achievements of their cultures to cattle-breeding tribes migrating in the northern space of Eurasia.

However, the cultures of Andronovo community were not autochthonous population of Middle Asia. The researchers of these cultures have ascertained their cattlebreeding type of economy, which basic territory was Central Kazakhstan, Southern Siberia, the Ural and the Volga areas, where their main well-studied sites and necropolises were located. Any long-term sites of Andronovo culture in Middle Asia are practically unknown. Their few locations, Tugajnoe site in the Zeravshan Valley (Liyonet, Isakov, Avanesova 1996; Avanesova 2002: 16-18), Jakke-Parsan 2 in Khorezm (Itina 1977: 173-210), and Kajrakkum in the Ferghana Valley (Litvinskij, Okladnikov and Ranov 1962), have no significant cultural layers, successively changing constructive periods and horizons and they are not evidence of long-term and total Andronovo populating in Middle Asia. Consequently, their presence in Middle Asian space might be seasonal, episodic, like, e.g., nomadic population’s presence in later periods.

Completely unique cultural-ethnical formation was developed during the mid – late 4th millennium BCE on enough vast territory of South Central Asia, South Afghanistan and South-East Iran. It was represented by similar material cultures, which analyze allows interpreting this phenomenon as the system of settlements with common political and economical interests. Also all these settlements were a union of kindred autochthonous groups according to archaeological researches. H. Nissen dedicating the series of works to forms and content of Mesopotamian trade relations with adjacent regions arrived at conclusions that the trade relations could be only equal partnership at such political structure existed on its territory. The evidence is translations of texts of later clay tablets (Nissen 1995).

Where was autochthonous population of Middle Asia during the period from the 3rd millennium BCE to 14-13th centuries BCE and whether it existed in general? From our point of view it would be completely inconceivable to assume that the wonderful, fertile oases of Middle Asia, vast rich pastures and area rich in minerals were not populated autochthonously. Perhaps, the reason is that scientists working during the 20th century and relying on traditional Soviet schemes of periodization based on the relative chronology and on the evolutional theory of development of bronze metallurgy (BogdanovaBerezovskaja 1966: 163-167) have wrongly age-dated some archaeological cultures. Let us look through the materials of archaeological researches of past years anew.

The archaeological evidences of such a partnership with autochthonous Neolithic Middle Asian population of Kel’teminar culture were proved in works by A. V. Vinogradov and E. D. Mamedov (1975). They disclosed a big number of semi-ready turquoise products such as turquoise pieces with drilled through-holes (people of Kel’teminar culture did not wear turquoise adornments themselves) prepared for barter with settled-agricultural centres.

‘Sujargan culture’ with distinctive Eneolithic aspect and red-slipped ceramics were studied in Khorezm. Initially, the culture was age-dated by М. А. Itina back to late 3rd millennium BCE. But in several years it was attributed to ‘Kamyshlin culture’ age-dated to 13-9th centuries BCE (Itina 1977: 40-44) (Fig. 7). ‘Zamanbaba’ culture with the ceramic complex continuing traditions of ‘Kel’teminar’ culture ceased to exist in the mid-3rd millennium BCE was researched in the Lower Zeravshan. Among materials there were the imported painted ceramics and adornments of the period of Namazga IV (Fig. 8). Ja. G. Guljamov, U. Islamov and A. Askarov (1956) being discoverers of complexes agedated them back to early 3 rd – 2nd millennium BCE (Guljamov, Islamov and Askarov 1966: 118-186). B. A.

During late 3rd – early 2nd millennia BCE the ‘cultural’ contacts were established with the population of the socalled ‘Andronovo cultural historical community’, when initially small settlements were developed into large complicated organized centres. At early stage it was contacts with tribes of ‘Petrovo culture’ (Fig. 1), then with ‘Alakul’ culture’ (necropolises of Chakka and Muminabad in the Zeravshan Valley) and with 217

T. Sh. Shirinov, S. R. Baratov have been divided into 3 groups attributed to 3 chronological periods:

Latynin synchronized Zamanbaba culture with tribes of the pit, catacomb, Afanasievo and Okunevo cultures. V. M. Masson synchronized it with late stages of Namazga IV, Gissar III and early Namazga V (2500 – 2200 BCE). Е. Е. Kuz’mina synchronized Zamanbaba culture with complexes of Gissar III B-C, Namazga IV and V, Shakhtepe II and III and steppe (pit, catacomb, Poltavkino and Afanasievo) cultures (3000 – 2200 BCE). V. Sarianidi collecting the scattered material analogue to Zamanbaba type in Faruhabad oasis in Northern Afghanistan age-dated the culture back to 12-10th centuries BCE. Nowadays A. A. Askarov age-dates the culture back to early 2nd millennium BCE (Askarov 1981: 99-110; Sarianidi 1979: 23-28; Latynin 1958: 49-53; Kuz’mina 1958: 24-33. Masson 1959: Table. 39.2).

1 Period – they compose the most significant group of graves. The complex is represented by artifacts analogue to Zamanbaba burial ground. The graves of this complex occupied the compact central position in necropolis area and can be age-dated back to late 3rd millennium BCE. (Fig. 9, 10, 11) 2 Period – they compose a small group of enough poor graves. The group was formed, obviously, later than the central group of graves. The materials from grave of this group are analogue to complexes of Gissar IIB – IIIA. Their complex of artifacts can be age-dated back to 3 and 2nd millennia BCE. (Fig. 10, 11, 12).

One more undeserved ‘victim’ is ‘Bishkent’ culture represented by the burial grounds of Early Tulkhar, Tigrovaja Balka, Oj-Kul’, Dzhar-Kul’, Vakhsh I and Makonimor. A. M. Mandel’shtam researching materials of Early Tulkhar burial ground with many artifacts analogue to Mesopotamian and Iranian cultures cited in details all possible chronological parallels from the mid3rd millennium to the 8th century BCE. The chemical analyses of bronze articles unearthed at the burial ground indicated that they had rather big percentage of intentional ligature of tin in copper. In that period the evolutional concept on successively developing metallurgy of bronze alloys from lead in Early Bronze Age to tinny in the late period was predominating. That was the reason why A. M. Mandel’shtam summarizing the total sum of analogues and chronological parallels and referring to I. V. Bogdanova-Berezovskaja’s weighty opinion was inclined to age-date Early Tulkhar burial ground back to late date 14-8th centuries BCE (Mandel’shtam 1966).

3 Period – they compose a group of graves consisting materials of Molali stage of Sapalli and Fedorovo cultures, which can be age-dated back to 1500 – 1400 BCE (Baratov 2001: 161-179) (Fig. 13, 14 15, 16) considering the calibrated dates С14. Possible synchronization of archaeological complexes of Early Tulkhar burial ground with the archeological sites of Late Bronze Age in Middle Asia is given in Fig. 17. Thus, we assume that Sujargan, Zamanbaba and Bishkent cultures left by Middle Asian autochthonous population on the territory of Middle Asia existed, at least, since 3 rd and 2nd millennia BCE. Evidently, Zamanbaba culture at early stage of its formation was spread on the territory of the Zeravshan Valley and Bactria. The archaeological complexes of Zamanbaba and Bishkent complexes are good illustrations of tight contacts with settledagricultural population representing the urban cultures and they allow assuming their significant economical and cultural integration with the latter ones.

L. T. Pjankova researched a group of analogue sites such as Tigrovaja Balka, Oj-Kul’, Dzhar-Kul’, Vakhsh I and Makonimor. The culture was age-dated to 8-7th centuries BCE (P’jankova 1989). The burial ground of Tigrovaja Balka, for instance, had a series of calibrated dates С14: GIN 2524 к.8 (3920 100) 1800, GIN 2525 к.8 (4510 340) 3000, GIN 2526 к.8 (4020 80) between 3rd and 2nd millennia BCE (P’jankova 1989: 101).

It is quite obvious that the cultural components of Zamanbaba and Bishkent cultures played an active role in development of cultures of Namazga VI type in Bactria, Margiana and Sogd, which since early 2nd millennium BCE had a predominating position in the area. And, perhaps, exactly these cultural components significantly determined peculiarity of Bactrian-Margianian archaeological complex and differed it from the urban centres of Iran being, apparently, initial at its appearance in Middle Asia.

H.-P. Francfort researching Shortugaj site in Northern Afghanistan during the 70-s obtained the stratified data also including ‘Bishkent’ archaeological complex. Results of excavations are important, because the studied levels had the series of calibrated dates С14. The ceramics of Bishkent type at Shortugaj is corresponded to the levels of 2-4 II period. Dates of layers are: 20602180; 2600-2900; 2900-3340; 2180- 2540 ВСE (Francfort 1989). Thus, it is possible to age-date ‘Bishkent’ complex of early type to late 3rd millennium BCE.

Bibliographical References Askarov, A. A. (1981) ‘K peredatirovke kul’tury Zamanbaba ’. Kul’tura i iskusstvo drevnego Khorezma. М: 99-110. Avanesova, N. A. (2002) ‘Mezhkul’turnye vzaimodejstvija stepnogo naselenija Evrazii i usrbanizirovanykh zemledel’tsev Srednej Azii’. Tsivilizatsii Tsentral’noj Azii. Zemledel’tsy i skotovody: traditsii i sovremennost. Samarkand: 16-18.

We have made an attempt to systematize and age-date the burial complexes of Early Tulkhar burial ground anew. As a result of carried out work 73 graves of burial ground 218

AUTOCHTHONOUS LATE BRONZE AGE CULTURES IN CENTRAL ASIA: WHERE ARE THEY? Vinogradov, A. V., Mamedov, E. D. (1975) Pervobytnyj Ljavljakan. М.

Baratov, S. (2001) ‘Fergana und das Syr-Dar´ya-Gebiet im spätem 2. und frühen 1. Jahrtausend v. Chr.’. Migration und Kulturtransfer. Der Wandel vorder- und zentralasiatischer Kulturen im Umbruch vom 2. zum 1. vorchristlichen Jahrtausend. Akten des Internationalen Kolloquiums. Berlin, 23. bis 26. November 1999. Bonn: 161-179. Baratov, S. R., Smagulov, E. (2001) ‘Archeologitscheskie issledovanija na mogilnike Scherbay v Tschimkentskoj oblasti’. Izvestija. Serija obschestvennyh nauk. № 1 (232). Alma-Aty. Janvar’-Fevral’: 209-214. Bernstam, A. N. (1952) ‘Istoriko-arheologiche’skie ocherki Centralnogo Tyan-Shanya i PamiroAlaya. MIA, 26: 19-22. Bogdanova-Berezovskaja, I. V. (1966) ‘Khimicheskij sostav metallicheskikh predmetov iz mogil’nikov epokhi bronzy v Bishkentskoj doline’. MIA, 145: 163-167. Francfort, H.-P. (1989) Foilles de Shortugai. Recherches sur l´Asie Central protohistorique. Vol. I, II. Paris. Guljamov, Ja. G., Islamov, U., Askarov, A. (1966) Pervobytnaja kul’tura i vozniknovenie oroshaemogo zemledelija v nizov’jakh Zarafshana. Tashkent:118-186. Itina M. A. (1977) ‘Istorija stepnykh plemen Juzhnogo Priaral’ja (II – I)’. TKhAE. Х. М Ivanov G.P. (1988) ‘Kashkartshinskiy mogilnik – novyi pamyatnik epohi bronzy v Fergane’. Obshestvennye nauki v Uzbekistane. 10. Tashkent: 44-47. Kuz’mina, Е. Е. (1958) ‘Mogil’nk Zamanbaba’. SE, 2: 24-33. Latynin, B. A. (1958) ‘O juzhnykh granitsakh ojkumeny stepnykh kul’tur epokhi bronzy’. SА, 3: 49-53. Litvinskij, B. A., Okladnikov, A. P., Ranov, V. A. (1962) ‘Drevnosti Kajrak-Kumov’. ТАN ТSSR, 33. Dushanbe. Liyonet B., Isakov A., Avanesova N. (1996) Sarazm (Tadjikistan). Céramiques (Chalcolithique et Bronze Ancien). Mémoires de la mission archéologique Française en Asie Centrale. T. VII. Paris. Mandel’shtam, A. M. (1966) ‘Pamjatniki epokhi bronzy v Juzhnom Tadzhikistane’. MIA,145. М. Masson, V. M. (1959) ‘Drevnezemledel’cheskaja kul’tura Margiany’. MIA, 73. М.-L. Nissen, H. J. (1995) Kulturelle und Politische Vernetzungen in Vorderen Orient des 4. Und 3. Vorchristlichen Jahrtausend. Beiträge zur Kulturgeschichte Vorderasiens. Mainz. P’jankova, L. T. (1989) Drevnie skotovody Juzhnogo Tadzhikistana. Dushanbe. Sarianidi, V. I. (1979) ‘K voprosu o kul’ture Zamanbaba’. Etnografija i arkheologija Srednej Azii. М: 23-28. 219

T. Sh. Shirinov, S. R. Baratov

Fig. 1 - Petrovo Culture in Zaravshan Valley (after N. A. Avanesova 1996)

220

AUTOCHTHONOUS LATE BRONZE AGE CULTURES IN CENTRAL ASIA: WHERE ARE THEY?

Fig. 2 - Fedorovo Culture in Fergana Valley (after A. N. Bernstam 1952)

221

T. Sh. Shirinov, S. R. Baratov

Fig. 3 - Fedorovo Culture in South Kazakhstan (after S. Baratov, E. Smagulov 2001)

222

AUTOCHTHONOUS LATE BRONZE AGE CULTURES IN CENTRAL ASIA: WHERE ARE THEY?

Fig. 4 - Fedorovo Culture in South Kazakhstan (after S. Baratov, E. Smagulov 2001)

223

T. Sh. Shirinov, S. R. Baratov

Fig. 5 - Fedorovo Culture in South Kazakhstan (after S. Baratov, E. Smagulov 2001)

224

AUTOCHTHONOUS LATE BRONZE AGE CULTURES IN CENTRAL ASIA: WHERE ARE THEY?

Fig. 6 - Alexeevo Culture in Fergana Valley (after G. P. Ivanov 1988)

225

T. Sh. Shirinov, S. R. Baratov

Fig. 7 - Suyargan Culture in Horezm (after M. A. Itina 1977)

Fig. 8 - Zamanbaba Culture (after A. A. Askarov 1981)

226

AUTOCHTHONOUS LATE BRONZE AGE CULTURES IN CENTRAL ASIA: WHERE ARE THEY?

Fig. 9 - Early Tulkhar Burial. 1 Period.

227

T. Sh. Shirinov, S. R. Baratov

Fig. 10 - Early Tulkhar Burial. 1 - 2 Period.

Fig. 11 - Early Tulkhar Burial. 1 - 2 Period.

228

AUTOCHTHONOUS LATE BRONZE AGE CULTURES IN CENTRAL ASIA: WHERE ARE THEY?

Fig. 12 - Early Tulkhar Burial. 2 Period.

229

T. Sh. Shirinov, S. R. Baratov

Fig. 13 - Early Tulkhar Burial. 3 Period.

Fig. 14 - Early Tulkhar Burial. 3 Period.

230

AUTOCHTHONOUS LATE BRONZE AGE CULTURES IN CENTRAL ASIA: WHERE ARE THEY?

Fig. 15 - Early Tulkhar Burial. 3 Period.

Fig. 16 - Early Tulkhar Burial. 3 Period.

231

232

2400

2300

2200

2100

2000

1900

1800

1700

Togolok

Gonur 1

?

Tahirbay 3

1500

1600

Yaz I

1400

1300

1200

South Turkmeniya

Sapally

Dashly 3

I

Dzarkutan

II

Dzarkutan

IIIа

Dzarkutan

IIIb

Dzarkutan

1

Kuchuk

South Uzbekistan

Zamanbaba

Chakka, Muminobad

Tugay

Siabcha,

Zaravshan

Fig. 17 – Synchronous Table [*ETB – Early Tulkhar Burial]

ETB 1

ETB 2

ETB 3*

Tandyryul

Kangurttut

Tadzikistan

Chust

Weight from Sokh

Shagym

Chek, Yapagi, Buadil, Dahana, Arpa, Arsif

Kashkarch,

Dalverzin

Fergana

Osh

T. Sh. Shirinov, S. R. Baratov

EXPLORING CERAMIC VARIABILITY IN IRON AGE SOUTH INDIA: SOCIAL AND POLITICAL IMPLICATIONS Carla M. Sinopoli

doubt complex – involving both the positive interactions of trade, exchange, and kinship and the more negative ones of conflict and competition.

Abstract Along with the emergence of iron technologies and new forms of mortuary treatment, the South Indian Iron Age is also marked by the appearance of new and widespread ceramic wares, technologies, and vessel forms. Elegant and finely made black and red wares, polished black wares and polished red wares largely replace the simple plain wares of the South Indian Neolithic; and at least some portion of vessels are now produced using wheelforming technologies. In addition, the localized stylistic traditions of the Neolithic are supplanted by region-wide forms, particularly in finely made serving bowls, which are recovered in both settlement and ritual contexts. In this paper, I report on ongoing analyses of ceramics from excavations at the Iron Age settlement and mortuary site of Kadebakele, located in northern Karnataka (co-directed by the Karnataka Department of Archaeology and Museum, Kathleen D. Morrison, and Carla M. Sinopoli). I review the nature and distributions of ceramic forms recovered in 9th-3rd century BCE levels at Kadebakele, and draw inferences to large patterns in ceramics forms and their role in diverse ritual and consumption activities across peninsular India during a period of dramatic sociopolitical change.

Even as the Late Prehistoric political landscape was divided among multiple interacting political communities, the second trend that characterized South India’s late prehistoric societies was an increasing material homogenization. Thus, there is a striking material uniformity across much of this broad region. This is seen in many aspects of Late Prehistoric material culture – most notably, the well-known megalithic mortuary features (see Sundara 1975; Moorti 1994; Mohanty and Selvakumar 2002), iron weapons and tools (Chakrabarti 1992; Tripathi 2002), personal ornaments (particularly beads; e.g., Francis 2002; Basa 2002), and distinctive and widely recognized ceramic forms and wares (including Black and Red Ware, polished red and black wares, and toward the end of the period, Russet Coated Painted Ware (RCPW); Gurumurthy 1981). Despite well over a century of research on the South Indian Iron Age, many questions remain about the development, growth, and sociopolitical organization of these late prehistoric societies. We have at best rudimentary knowledge of local and regional chronological sequences for both mortuary and habitation sites. However, available evidence indicates that many of the developments I have noted began earliest in inland northern Karnataka and Andhra Pradesh in the late centuries of the second millennium BCE (see Fuller et al. 2007: 771-772 for discussion of recent dating efforts of the Neolithic-Iron Age transition) and expanded south into Tamil Nadu and coastal regions sometime after c. 500 BCE.

Introduction The first millennium BCE was a time of significant social, political, and economic change across peninsular India. Known variously as the ‘Iron Age’ or ‘megalithic period’, this late prehistoric period was marked by widespread technological innovations and economic intensification, the adoption of new and widespread mortuary and commemorative practices, and expanding long-distance economic, ideological, and sociopolitical interactions. These and other changes contributed to two important and divergent features of the South Indian Late Prehistoric period. The first concerns political organization and the development and consolidation of a large number of (more or less) autonomous polities of varying scales and complexity. Both archaeological data and limited and chronologically late textual evidence (e.g., Sangam literary sources; Asokan inscriptions) indicate that at least some of these polities were organized along principles of social inequality and incorporated multiple settlements. Throughout the first millennium BC, some of these settlements grew quite large and became centers of hierarchically organized settlement systems – associated with smaller villages and perhaps pastoral settlements, with cemeteries, and with complex agricultural features. Many were also loci of specialized craft production and long-distance exchange. Relations among these emergent political centers were no

In this paper, I focus on a small region of northern Karnataka and on one category of late prehistoric South Indian material culture – ceramics – and largely limit my discussion to materials recently recovered in the ongoing excavations of the Late Prehistoric/Early Historic Landscapes of the Tungabhadra Corridor project (LP/EHLTC), a collaborative project of the Karnataka Department of Archaeology and Museums, the University of Chicago, and the University of Michigan. To some extent, the diagnostic ceramics of the South Indian Iron Age are well known. They include characteristic wares – slipped and polished black ware, slipped and polished red wares, and the famous black and red ware. And they include a fairly narrow range of forms – small simply-formed coil-built bowls in all three wares and large wheel-made jars, typically of slipped and polished red ware. These wares and forms are widely 233

Carla M. Sinopoli We do not at present have a good understanding of the contemporaneity of occupation in different areas of the site, though it is unlikely that the entire area was ever occupied simultaneously.

recognized as diagnostic of the Iron Age, so much so that site reports from across South India are replete with such phrases as ‘typical megalithic ware’ or ‘absolutely identical with the megalithic ware of Brahmagiri’ as Thapar (1952) described ceramics recovered at the site of Porkalam located in the Trichur district of the west coast of Kerala.

Current data suggest that the earliest occupation at Kadebakele was situated on the hill’s uppermost terrace: a protected and relatively inaccessible area situated more than 100 meters above the river (Fig. 1). This relatively flat open area is roughly 220 m north-south by 100 meters east-west, and it is there where we have focused our major excavation efforts. Test excavations in six areas of the upper terrace conducted in 2003 and 2005 have yielded evidence for residential structures, stratified midden deposits, and ritual activities, including the construction of megaliths and deposition of debris from large-scale consumption events, possibly collective feasts. Excavations are conducted on a two by two meter grid, initially following arbitrary ten centimeter levels and then, where possible, following natural stratigraphy. All deposits are screened through three mm mesh to recover cultural materials and sediment samples are collected for flotation and phytolith, pollen, and sediment analysis. Twenty-eight calibrated radiocarbon assays from stratigraphic contexts date the main portion of the upper terrace occupation from the late ninth through midfifth centuries BC, with evidence for subsequent less intensive use from c. 400-200 BC, when the major occupation of the site appears to have shifted down slope to the river plain (Sinopoli 2009).

Despite our widespread recognition of broad ceramic categories and intuitive acknowledgment of ceramic similarities across large regions, there has been relatively little effort at detailed systematic characterization of vessel morphology and other attributes and few attempts to document local and regional variability within these broad ceramic categories (for an important exception, see Wessels-Mevissen 1991). In this paper, I report on some preliminary steps we are taking to characterize a subset of ceramics recovered in our project. My interests are typological, chronological, and sociopolitical – to take some initial steps toward refining the broad ceramic categories currently available and to consider the uses and meanings associated with the adoption of these widespread and highly standardized ceramic wares and vessel forms across vast areas of the Indian peninsula. I begin by providing a bit of background on our fieldwork. THE LP/EHLTC Project Initiated in 2003, the Late Prehistoric/Early Historic Landscapes of the Tungabhadra Corridor project focuses on a c. 38.5 sq km area along the Tungabhadra River in Koppal and Bellary Districts of northern Karnataka, on the northeastern edge of the famous medieval site of Vijayanagara. The project area contains more than 20 late prehistoric sites, including two large megalithic mortuary sites and nine sites containing isolated megaliths, four settlements and an inhabited rock shelter, as well as several rock art sites. The sites were identified as part of the project directors’ earlier Vijayanagara Metropolitan Survey project conducted in the region (see Sinopoli and Morrison 2007). Most are located on the slopes and hilltops of the rugged granitic inselbergs that dominate the landscape of this region – with the settlements found on relatively flat natural terraces distributed at various elevations across the massive granite hills that rise above low-lying areas of irrigated fields (extensive Vijayanagara-era and more recent irrigation has likely destroyed many prehistoric sites that once existed in these lower areas).

The excavations have yielded evidence for both household and ritual (i.e., megalithic) architecture and associated deposits. Recovered materials include abundant fauna (see Bauer 2007) and botanical remains, as well as a range of artifacts – including characteristic iron artifacts and semi-precious stone beads. Here, I focus on the most ubiquitous cultural materials recovered – earthenware ceramics. Ceramic Analysis Given the vast quantities of ceramics recovered in the excavations, we are taking a hierarchical approach to our analysis. All recovered ceramics are sorted, counted, and weighed into one of 19 possible ware categories (see Table 1; as well as ‘other,’ ‘indeterminable/eroded,’ and ‘burned’) and subdivided into bowl, jar, and unknown/other morphological categories. A subset of body sherds is subject to more detailed attribute analysis, recording information on sherd thickness and interior and exterior surface treatment and color – variables that we hope will prove chronologically sensitive. An even smaller group of diagnostic sherds are measured and drawn, with detailed information recorded on a range of metric and qualitative variables (e.g., diameters, heights, angles, thicknesses, and non plastic inclusions, surface treatments, production techniques, paste, etc.). Finally, project collaborator Dr. Kajal Shah has conducted petrographic analysis of a small subset of the diagnostic sherds to examine mineralogy and ceramic technology (Shah n.d.).

Given the relatively limited attention that has been paid to settlement sites in Iron Age research (see Moorti 1994), the four documented settlements in the project area have been our major focus, with the majority of our efforts thus far focused on the site of Kadebakele (see Morrison et al. n.d.; Sinopoli et al. n.d.; Sinopoli 2009). Kadebakele is the largest settlement in the project area, with dense surface scatters of late prehistoric and early historic ceramics and other materials found around the base and along multiple terraces on the slopes of a large granitic hill. Late prehistoric and early historic remains extend over an enormous area – well over 60 hectares. 234

EXPLORING CERAMIC VARIABILITY IN IRON AGE SOUTH INDIA: SOCIAL AND POLITICAL IMPLICATIONS serving vessels used in the consumption of food and liquids. The widespread distribution of similar bowl forms across much of South India during the late prehistoric period might point to a common set of commensal practices, involving at the very least individual consumption of food and drink from these often extremely elegant serving vessels.

To date we have documented more than a quarter million sherds to the level of ware (Table 1.a). Detailed attribute information has been recorded on roughly 18,000 sherds, and we have coded up to 20 metric and categorical variables on approximately 1900 rims, bases, decorated sherds, and other diagnostics Although there are a number of rare ware categories, more than 90% of the identifiable sherds fall into one of three ware categories (see Table 1.a; note, frequencies discussed here are computed after removing the 62,277 sherds (27.1%) that were coded as indeterminate or severely eroded, for which ware could not be determined). Slipped and polished (s/p) black wares and slipped and polished red wares occur in roughly equal frequencies, with 63,584 s/p black ware sherds (37.95% of identifiable sherds) and 63,432 (37.86%) s/p red ware sherds. Black and Red Ware (brw) sherds number 24,310 or 14.51% of the identifiable assemblage. Brown slipped and polished ware are the next most numerous category (n=9166 sherds, 5.47%). However, most of these vessels are reddish-brown in color and the vast majority were likely mis- or over-fired red wares (or partially reduced red-ware cooking vessels). We are provisionally retaining the brown ware category, but it is probable that is not terribly meaningful and these vessels should ultimately be added to the red ware totals. In addition, it is important to acknowledge that BRW vessels are inevitably undercounted, since vessels that break high on the vessel wall will appear to be black ware. Nonetheless, it also seems to be the case that Black and Red Ware is somewhat less common than we might expect given its prominence in the archaeological literature on the South Indian Iron Age. Similar ware frequencies were identified by Johansen (2008) in his work on ceramics from two smaller settlements in the LP/EHLTC project region.

To further subdivide the bowls, we have identified a number of provisional rim form categories, based on minor variations in lip and rim form and orientation. While some of these categories are quite distinctive and seem to be associated with fairly restricted vessel shapes, others appear to be relatively minor variants that result from how coils are added and smoothed in the final stage of vessel forming. In future analyses, we will examine how these variants are distributed over space and time, in the hope of being able to refine understandings of ceramic chronologies and ceramic use. Ceramic use: comparison of ritual and domestic spaces at Kadebakele Ceramics were abundant in all of the archaeological contexts excavated at Kadebakele – in residential and trash deposits as well as in ritual contexts. To further parse out ceramic use and variability, I examine below relative vessel form frequencies from two distinctive contexts: one ritual and the other primarily or exclusively associated with domestic debris. The ritual deposit is located on the southern edge of the upper terrace, in an area of dense surface deposits and circular stone alignments. Excavations in this area (Unit -21.6E/-204N) yielded evidence for repeated depositional events involving fauna, ceramics, and other goods, in what is best interpreted as a ritual context associated with large scale commensal or ‘feasting’ activities. The upper levels of this same unit consist of midden deposits that post-date the main occupation of the upper terrace (a calibrated radiocarbon date from the center of these deposits dates from 404-207 BC). The deposits of concern here lie beneath these late deposits and date from 800-500 BCE, the period of major occupation of the upper terrace. In a deposit some 70 cm thick, we identified 11 superimposed strata of alternating lime plaster and dark earth sediment, cut by at least four pits (Fig. 3.a). Extremely high densities of fauna were found in these deposits and have been analyzed by Dr. Radhika Bauer (2007). The faunal remains in these deposits are dominated by domestic cattle (47% of identifiable specimens), typically with high degrees of skeletal completeness and evidence for burning. These patterns are quite distinct from those associated with residential contexts at Kadebakele. Typical domestic deposits (both midden and household contexts) contain a much broader diversity of species, including much higher average percentages of wild fauna (averaging 73% NISP) and lower percentage of domestic cattle (averaging 17.5% NISP). Fauna in domestic deposits also exhibit much lower evidence of burning, which Bauer associates with roasting. Thus, Bauer (2007) has convincingly argued

Wares and vessel form Ware category is closely associated with vessel form (Table 1.b). Thus, the vast majority of jar sherds (73%) are of red or reddish brown ware, a pattern also recognized by Wessels-Mevissen (1991: 15) at Adichchanallur. These are generally large wheel-made vessels (Fig. 2.a-f), slipped and polished on the exterior and plain on the interior, with wheel marks still visible. They were likely used for a range of activities, including storage, transport and cooking. Unlike the predominately red ware jars, nearly 92% of the small finely shaped bowls that characterize the Kadebakele ceramics were of either polished black ware or black and red ware (Fig. 2.g-m). They were hand-built, using the coiling technique, and many exhibit clear breaks along coil joins. The bowls are very thin, with average wall thickness of 42 cm and most are slipped and finely polished. They are generally quite small, with an average rim diameter of 15.7 cm (standard deviation 5.47 cm) on the 964 bowls I have measured. They can be grouped into four tentative size classes (≤ 8 cm, n=49; 9-18 cm, n=685; 19-23 cm, n=141; ≥ 24 cm, n=49); with the vast majority of bowls ranging between 9 and 18 cm in diameter. The most likely function of the majority of bowls is as individual 235

Carla M. Sinopoli that the complex stratified deposits in this first unit (located within a now-much disturbed megalithic stone circle) are the remains of repeated events involving large scale food consumption, which occurred over what was probably a fairly restricted time span sometime between c. 800 and 500 BCE.

Graffiti marks in ritual and domestic spaces The addition of post-firing ‘graffiti’ marks to ceramics is another characteristic shared among numerous Iron Age sites, and Kadebakele is no exception. These marks take a variety of forms, many documented at other sites throughout South India in both burial and domestic contexts (see Yazdani 1916, Boivin et al. 2003; Rajan 1998; Thapar 1957: 86-88). In general, ceramics with graffiti have tended to occur in greatest frequency in burial features; though they are also known from domestic contexts (see Thapar 1957). Although we have excavated several megalithic features at Kadebakele (see Morrison et al. n.d.), including the ritual deposits discussed above, we have not yet excavated any formal burials. It is thus interesting to explore if ceramics with graffiti occur in significantly different frequencies in nonburial ritual deposits versus domestic deposits.

The second unit I consider here is a domestic deposit – a deep stratified midden located on the northern edge of the upper terrace, adjacent to an area where we have excavated the remains of several residential structures and associated features. The deposits in this unit comprise stratified lenses of sediments containing large quantities of ceramics, fauna, and botanical debris, and a wide range of other artifacts (Fig. 3.b). The faunal material, with a range of wild and domestic resources, closely resembles the proportions associated with nearby residential areas. Domestic cattle constitute only 15% of the assemblage and wild fauna and commensals account for 70% of the faunal remains (with domestic sheep/goat, and pig constituting the remainder). Four radiocarbon dates range from c. 800-400 BC, indicating that this unit is broadly contemporaneous with the ritual deposits at the opposite end of the upper terrace.

Overall, graffiti is rare among the Kadebakele ceramics – with only a total of 64 sherds recorded among the nearly 1900 diagnostic sherds coded (3.45%). Most of the sherds are fairly small and complete motifs are rare. Nonetheless, they exhibit considerable variability, with more than two dozen individual motifs represented (most common being Yazdani’s element 52 – an ‘x’ shaped motif, and variants of Yazdani’s motif 67, a checkerboard motif; Yazdani 1916, 56, with 7 examples each). Vessels with graffiti were most common in the excavated residential area (Block B, an area of superimposed household architecture), with 36 sherds of 612 coded (5.88%) diagnostic sherds. They were, however, were surprisingly rare in the associated northeast midden deposit (only 4 of 601 coded diagnostics; 0.67%). Graffiti sherds also occurred in somewhat higher than expected frequencies in the southern slope midden/ritual feasting area discussed above (16 of 369 diagnostics; 4.34%). While these differences are statistically significant, the sample sizes are so small that at this point is probably safest to just confirm that sherds with graffiti occur in a variety of domestic and ritual contexts at Kadebakele. Their paucity in midden contexts may point to a higher rate of curation or different use of marked vessels such that they break and are deposited in domestic middens in lower frequencies than non-marked vessels. Further analysis with larger sample sizes is warranted.

In both areas, diagnostic bowls significantly outnumber jars (Fig. 3.c), attesting to the importance of these small drinking and dining vessels in late prehistoric activities. However, bowls seem to be far more important in the activities that produced the ritual deposits than in the midden area, which contains the deposits of more wideranging domestic activities. In the midden, I recorded some 28% (1.28:1) more diagnostic bowls sherds than those from jars; while in the ritual area, bowls outnumber jars by nearly 2:1. The midden deposits also exhibit a greater diversity of bowl forms than recovered from the ritual deposit, including higher frequencies of larger vessels, and a somewhat greater diversity of ware categories. These differences may provide some insights into the social significance and use of the fine bowls that characterize the South Indian Iron Age. Their abundance in the ritual deposits helps to confirm Bauer’s faunaderived interpretations of the importance of food consumption in a context of large scale commensal activities or feasts. The use of these small bowls in consumption activities associated with ritual feasting in the context of megalithic features further suggests a possible mechanism for their spread and widespread adoption across South India in the late prehistoric period. That is, I would like to suggest that these new ceramic forms – quite distinct from the simple Neolithic cooking pots that precede them – emerged in a context of a new set of ritual behaviors and practices, associated with megalithic features and with large-scale food consumption events or feast. These ceramics then were part of a larger material ‘package’ of change, related to the changing ideological and sociopolitical dynamics of the late prehistoric period – a time, I have argued, of emerging social inequalities and political consolidation.

Discussion The results presented in this paper are quite preliminary, as the analysis of the ceramics of Kadebakele is just beginning. Much work remains to be done both to adequately describe the variability in the ceramic assemblage and to analyze and interpret its social, political, and ideological significance in the dramatically changing world of late prehistoric South Indian communities. Refinement of the ceramic chronology is a critical first step in this work and efforts in that direction are ongoing. In addition, our work seeks to explore how Iron Age communities used ceramics in a broad array of activities at Kadebakele and at other sites in the project area, and how this small area of northern Karnataka fits 236

EXPLORING CERAMIC VARIABILITY IN IRON AGE SOUTH INDIA: SOCIAL AND POLITICAL IMPLICATIONS into the larger world of late prehistoric and early historic South India.

Gurumurthy, S. (1981) Ceramic Traditions in South India. Madras. Johansen, Peter G. (2008) ‘A Political Economy of Space: Spatial Production and Social Organization in Iron Age Period South India’. Unpublished PhD Dissertation, University of Chicago. Mohanty, R. K. and Selvakumar V. (2002) ‘The archaeology of the megaliths in India: 1947-1997’. In Settar, S. and Korisettar, R. (eds) Indian Archaeology in Retrospect: Volume 1: Prehistory. New Delhi: 313-352. Moorti, U. S. (1994) Megalithic Culture of South India: Socio-Economic Perspectives.Varanasi. Morrison, K., Sinopoli C. M., and Gopal R. (n.d.) ‘Early Historic Landscapes of the Tungabhadra Corridor: preliminary report of the 2005 season’, report submitted to the Archaeological Survey of India, New Delhi. Rajan, K. (1998) ‘Further excavations at Kodumanal, Tamil Nadu’. Man and Environment, 23(2): 6576. Shah, Kajal (n.d.) ‘Preliminary report on thin-section analysis of ceramics from VMS-530 (Kadebakele): Unit -21.6E/-204 N, EHLT 2005’, unpublished ms. in possession of the author. Sinopoli, C. M. (2009) ‘Late prehistoric and Early Historic landscapes of the Tungabhadra Corridor’. In Paddayya, K. et alii (eds) Recent Research Trends in South Asian Archaeology. Proceedings of the Professor H.D. Sankalia Birth Centenary Seminar, Pune: 279-294. Sinopoli, C. M., Bauer A., Bauer R. S., Johansen P., Sugandhi N. (n.d.) ‘Early Historic Landscapes of the Tungabhadra Corridor: 2003 field season preliminary report’, report submitted to the Archaeological Survey of India, New Delhi. Sundara, A. (1975) The Early Chamber Tombs of South India: A Study of the Iron Age Megalithic Monuments of North Karnataka. New Delhi. Thapar, B. K. (1952) ‘Porkalam 1948: Excavations of a megalithic urn-burial’. Ancient India, 8: 3-16. Tripathi, V. (2002) ‘The Iron Age in India’. In Settar, S. and Korisettar, R. (eds), Indian Archaeology in Retrospect: Volume 1: Prehistory. New Delhi: 287-311. Wessel-Mevissens, C. (1991) ‘Adichchanallur reconsidered – a typological study of South Indian “Megalithic Pottery” ’. Beiträge Zur Allgemeinen und Vergleichenden Archäologie,11:13-56. Yazdani, G (1916) ‘Megalithic remains of the Deccan – a new feature of them’. Journal of the Hyderabad Archaeological Society 1916: 56-57.

Acknowledgments On behalf of all of the LP/EHLTC project directors, I gratefully acknowledge the support of the Archaeological Survey of India and the Ministry of Human Resource Development for their approval of this multiyear research project (No. F-221/2000-U.4). Mr. T. Gangadhar and the late Dr. Manjunathaiah of the Karnataka Directorate of Archaeology and Museums have been especially helpful, and we acknowledge their kind and generous support of the research and assistance in field logistics. Thanks also to the American Institute of Indian Studies, with which we are affiliated scholars, and especially Dr. Pradeep Mehendiratta and Ms. Purnima Mehta for overseeing administrative paperwork and for their sage advice and guidance. I also thank Mr. U. Srinivas for his dedication and many contributions to my research over more than two decades. Dr. Kajal Shah has conducted petrographic analysis of some of the Kadebakele ceramics, and I thank her for her contributions. LP/EHLTC has been supported by grants from the National Geographic Society (Grant 7110-01), the Wenner Gren Foundation for Anthropological Research, and the National Science Foundation (Grant BCS-0350803), as well as by support from the University of Chicago and University of Michigan. I would especially like to acknowledge a generous gift from Ranvir and Arkush Trehan to the University of Michigan Center for South Asian studies, which has supported aspects of this research. Thanks too to the many Indian and US-based students who have assisted in fieldwork and analysis, and particularly Peter Johansen for his important work on ceramics from surface work on sites in the project area.

Bibliographical References Basa, K. K. (2002) ‘Small is useful: importance of bead studies in South Asian archaeology’. In Settar, S. and Korisettar, R. (eds) Indian Archaeology in Retrospect: Volume 1: Prehistory. New Delhi: 389-418. Bauer, Radhika (2007) Animals in Social Life: Animal Use in Iron Age South India. Berlin. Boivin, N., Korisettar R., and Venkatasubbaiah P. C. (2003) ‘Megalithic markings in context: graffiti marks on burial pots from Kudatini, Karnataka’. South Asian Studies, 19: 21-33. Chakrabarti, D. K. (1992) The Early Use of Iron in India. New Delhi. Francis, Peter, Jr. (2002) Asia’s Maritime Bead Trade, 300 B.C. to the Present. Honolulu. Fuller, D.Q., Boivin N., and Korisettar R. (2007) ‘Dating the Neolithic of South India: new radiometric evidence for key economic, social and ritual transformations’. Antiquity, 81: 755-778. 237

Carla M. Sinopoli Table 1 - LP/EHLTC Ceramic Ware Categories 1a. - Ware frequencies and percentages (most common wares in boldface)

# sherds

% of total

# sherds*

% of total#

Black Plain Ware

608

0.26%

608

0.36%

Red Plain Ware

1395

0.61%

1395

0.83%

Brown Plain Ware

735

0.32%

735

0.44%

63,432

27.60%

63,432

37.86%

755

0.33%

755

0.45%

Slipped/Polished Black Ware

63,584

27.67%

63,584

37.95%

‘Classic’ BRW (2 Colors One Side)

(7142)

(3.11%)

7142

4.26%

BRW Inside/Outside (1 Color Each Side)

(17168)

(7.47%)

17168

10.25%

Black And Red Ware -- ALL

24,310

10.58%

24,310

14.51%

Rcpw

53

0.02%

53

0.03%

Plain Buff Ware

719

0.31%

719

0.43%

Plain Grey Ware

129

0.06%

129

0.08%

Plain Pink Ware

0

0.00%

0

0.00%

White Painted BRW

0

0.00%

0

0.00%

Painted (Decoration) Plain Ware (Any Color)

3

0.00%

3

0.00%

306

0.13%

306

0.18%

7

0.00%

7

0.00%

Reddish-Brown Slipped/Polished Ware **

9166

3.99%

9166

5.47%

Black And Brown Ware (Recorded 2005, 2006)

1068

0.46%

1068

0.64%

Other

1266

0.55%

1266

0.76%

Severely Burned (Recorded 2005, 2006)

10,566

4.60%

0.00%

Indeterminate/Eroded

51,711

22.50%

0.00%

Total

229,813

Slipped/Polished Red Ware Slipped/Polished, Brick-Red, Thick Crackled Slip

Slipped/Polished Buff Ware Black-Painted Red-Slipped Ware

167,536

*excluding burned and unidentifiable; #most likely misfired or over-fired red ware

1b. -Vessel form by ware

Ware

bowls

jars

totals

Red slip/pol (expected)

3760 (19,017)

37,569 (22,312)

41,329

Bl slip/pol (expected)

27797 (19,342)

14239 (22,694)

42,036

BRW all (expected)

16870 (8364)

1307 (9813)

18,177

Brwn sl/pol (expected)

305 (2009)

4062 (2358)

4367

totals

48,732

57,177

105,909

Chi-square=48,220.53; df=3, p 29) are most often-single component (basic)

A complete listing of these results is not possible given the constraints of this paper, however Figure 5 lists the 299

Bryan Wells signs. Of the signs that occur more than 100 times none consist of more than 3 components. Given these distributions one could postulate that the high frequency, single component signs are good candidates for syllabic signs. Clusters of these signs may be syllabic spellings. This postulation is tested in Wells 2006.

graphs (allographic sets) and those variations that are stylistic is a difficult task given that these ancient scripts (proto-Sumerian, proto-Elamite and Indus) are in the process of being deciphered. An important factor is that both Proto-Elamite and the Indus script use the compounding of common signs (especially logograph + number) to create new signs that, while unique, are composed of higher frequency signs. In a detailed sign list it is to be expected that there will be many singletons.

Analysis of the Indus script sign list Constructing a detailed list of Indus signs is not the final goal of this paper, but rather a necessary initial step in the analysis of the character and behavior of Indus signs. Having an adequate sign list makes it possible to answer several important questions: 1.

The frequency of Indus signs can be compared to other ancient scripts in order to examine the relatedness of these systems as can the search for overlaps in sign inventories.

2.

Inter-site comparisons of sign inventories can be analyzed for indications of regional variations in the Indus texts.

3.

Sign inventories can be used to define differences in subject matter between sites, and between artifact types.

4.

Comparison of sign inventories can isolate signs that locate preferentially at a site or on a certain type of artifact.

5.

The analysis of some of the chronological aspects of sign distributions can lead to the identification of changes in the system over time.

6.

The following discussion focuses on these five aspects of the sign list and database.

This discussion has shown that some variations in sign design can be signaling variations sign value with resulting variations in sign use, while some variation are simply graphic variations. In cases where variations cannot be demonstrated to be either graphic or graphemic, it is desirable to keep the signs separate awaiting further analysis. These uncertain signs appear in the sign list as low frequency signs (often as singletons), and have the effect of artificially inflating the count of singletons. High frequency signs can be the result of the abundance of certain artifact types at a site. For example, a comparison of the sign frequencies from Mohenjo-daro and Harappa show variations in rank order and inventory of the most frequent signs. The 10 most common signs from Mohenjo-daro and Harappa are given in Table 1. The rank of each sign at Mohenjo-daro and Harappa is given along with the sign graph, sign number, and total frequency for each entry. The question of why these sequences are different can be answered by a careful examination of the artifact inventories from each site. The presence of large numbers of bas relief tablets (TAB:B) and incised tablet (TAB:I) artifacts at Harappa results in the high frequencies of signs 700, 33, 32, and 400. This is not a coincidence, but rather a recognizable pattern. In fact, TAB:B and especially TAB:I texts very frequently contain the

of both (740) and (2) linked to the large numbers of seals at Mohenjo-daro. It is useful to examine the sign inventories from these sites with signs 2, 740, 700, 400, 31, and 32 removed. This results in the lessening of two effects: 1) the high frequency signs tend to dominate the lower frequency signs in analysis; 2) artifact specific signs tend to exaggerate difference between sites and artifact types. As 82% of the sign occurrences come from intaglio seals (SEAL) and ‘miniature tablets’ (TAB) artifact types, their sign inventories have a pronounced effect on the overall frequencies. If we are to get at the heart of the Indus script it is necessary to look beyond the obvious and examine the details of sign distributions.

There are two striking features of the frequency distribution of Indus signs the large numbers of signs that occur only once (singletons) and the small number of signs that occur extremely frequently. This sort of distribution is not unusual in sign lists from evolving ancient scripts. In Southwest Asia it is very common as Figure 7 shows. The most obvious feature of this graph is that the ProtoElamite script has many more singletons than either of the other two scripts. It can also be seen that the Indus script is most similar to Proto-Sumerian. One reason for the large number of singletons and large numbers of allographic variants is that these scripts had not yet been standardized (Damerow 1989). Another possibility is that minor variations in sign design are signaling a change in , sign 155

(32-700) sequences.

This is the same effect that causes such high frequencies

The comparison of sign frequencies from related scripts

semantic values as with Sign 156

(33-700) or the

The myth of the uniformity of the Indus system of writing can be debunked in several ways. The following discussion of sign distributions examines the differences in the distributions of Indus signs, by artifact type and by site.

, and

sign 158 . Separating sets of signs that are graphically related but with minor variations in the design of their

The distribution of signs in Figure 8 demonstrates that the 300

DEFINING THE INDUS SIGN LIST Indus script is not geographically uniform. Certain signs (176, 240, 220, 3, 590, 405, 156, 413, and 892) occur far more frequently at Harappa, while others (407, 820, and 845) are more frequently found in the inscriptions from Mohenjo-daro. These signs are not the rare signs but rather the more common ones. These differences are minor for the majority of signs, with signs 176, 220 and 240 being notable exceptions. The Indus script is not uniform but with a correlation coefficient of 0.746 (Fig. 8) these differences cannot be characterized as systemic, but rather there are a handful of signs that occur more commonly at one site or the other.

measures. This association is supported by the many examples of paired with , , , and on POT:T:g artifacts, although this relationship is not absolute. It is possible that wages were paid with grains dispersed from a centralized storage facility. There is little archaeological evidence bearing on this issue other than standardized ceramics and the ubiquitous pointed base goblets in later levels. I would suggest that SEAL:S artifacts might have greater proportions of singletons for two reasons: First, their more complex messages and limited space require more compoundin. Second, as the Indus trade relations expanded the Indus script would have to describe a greater number of things and so new signs needed to be invented. New things can be described either by spelling (some of these would be compounded spellings) or with new logographs.

Sign inventories vary between artifact types too. Figure 9 compares sign inventories from rectangular seals (SEAL:R) and square seals (SEAL:S) artifact types from Harappa and Mohenjo-daro. It can be seen from Figure 9 that SEAL:S artifacts from these sites have very similar sign inventories, while SEAL:R artifacts have many differences (r2 = 0.853 and 0.61 respectively). For SEAL:R artifacts there are eight signs that are more common at Mohenjo-daro than at Harappa (1, 820, 220, 407, 798, 920, 415 and 617), with only signs 692, 817, and 154 being more common at Harappa. The similarity of sign inventories from SEAL:S artifacts points to a uniform subject matter and possibly language between these two sites. The differences in the SEAL:R inventories seem to indicate a difference in subject matter. There are at least two possible explanations for this difference. First, SEAL:R artifacts could have different functions at Mohnejo-daro and Harappa. Second, as was shown in Wells 2006, seals were used in the control of trade (as TAG artifacts would seem to indicate). SEAL:R artifacts might be labeling local trade items (thus the site specific inventories), and SEAL:S artifacts might then be labeling Pan-Indus items (a more unified sign inventory). It is also possible that SEAL:R artifacts had a limited set of uses relating to cycle of production and system of exchange, while SEAL:S artifacts had a larger inventory of uses with only a minor overlap in function.

Chronological Relationships There is little data available regarding the chronological distributions of Indus artifacts and inscribed artifacts are no exception. The chronological data published in the early 20th century for Mohenjo-daro and Harappa (Marshall 1931, Mackay 1938, Vats 1940) have been criticized for its crude field methodology and incomplete publication (see Wells 2008). Nevertheless, this corpus remains the largest, and best documented, set of data describing the chronology of the Indus texts. All together the Mohenjo-daro publications list chronological and spatial data for 1270 texts (Marshall 1931, Mackay 1938). By comparison the recent excavations at Harappa (Meadow et al. 1990–2001) list chronological information for 308 artifacts, but only 139 of these can be ascribed to a specific phase (≈200 year period). More than half of the 308 artifacts (157) are fragmentary ceramic texts and of little use in epigraphic studies. Only 15 of the 308 artifacts are SEAL artifacts (SEAL:R = 6; SEAL:S = 9) and 136 are TAB artifacts. With the Vats (Mound F) and Mackay (DK.G) data, texts can be seriated into seven chronologically separate groups, but the relationship between these divisions is not certain.

The occurrences of low frequency signs are not uniformly distributed across artifact types. As shown in Figure 10, regardless of site of origin, seals in general have the highest proportion of singletons. SEAL:S artifacts from Mohenjo-daro have the highest proportion of singletons by an order of magnitude--with ≈45% of all singletons coming from Mohenjo-daro SEAL:S artifacts. The relatively conservative sign inventories of TAB artifacts may be attributed in part to the repetitiveness of the TAB texts and their limited subject matter, but this does not explain the low percentage of singletons in SEAL:S texts from Harappa. Again, this relates to the function of the artifacts, and the fundamental differences in sign inventories.

Singletons At Mohenjo-daro Just as the occurrence of singletons is not uniform for all artifact types, they are likewise unevenly distributed chronologically. An example of this is the sign distributions from Mohenjo-daro’s DK.G area. In Figure 11 the relationship between the total number of signs and the number of singletons by archaeological phase can be seen. The distribution of singletons is not uniform, and in a general way mimics the overall distribution of signs. There are three exceptions to this: the Intermediate III, the Late III, and the Late I phases where the number of singletons is higher than expected. The chronology of the Indus civilization is poorly defined, and so it is difficult link these variations to specific cultural or historical

TAB artifacts may have been used as ration chits or a form of money (Wells 1999:35), with the repetitive use of (700) paired with , , , and (32-35) relating to various values in a possible system of Indus volumetric 301

Bryan Wells influences. The overall increase in the frequency of signs in the Late II phase is a result of the way in which DK.G was excavated. DK.G can be divided into Northern and Southern sections along Central Street. The Northern section was excavated extensively to the Late II phase and in a small area to the Late III phase. The Southern section was excavated through the Intermediate Period and a small area of the Early period deposits were also exposed. So the addition of the Northern section’s texts impacts the Late II period sign frequencies. Note however the proportionally large numbers of singletons in the Late III is an interesting contrast to the decrease in total signs written and texts. In any case the distributions of singletons is complex with many variations spatially, temporally, and among artifact types.

inscriptions. The organization of Indus texts is far too complicated to be fully described by such a simple technique, but sign preferences for certain positions within text can identify important differences between signs with near identical graphs. For example, the case of signs 526 , and 527 (Wells 2006). These signs were identified as two separate graphemes based on the analysis of their contexts, but the separation of 526 527

and

was tentative, and they were maintained as

separate signs awaiting further evidence. Sign 526 occurs in 10 texts for which its position can be determined as follows: initial = 0, medial = 0, and terminal = 10. Sign 527 occurs in 38 texts for which its position can be determined as follows: initial = 1, medial = 5, and terminal = 32. These data support the separation

Allographic Variations Over Time

of signs 526 graphemes.

Sign 740 is the most common Indus sign (n = 1696) and it has six recognizable variations. These variations can be plotted over time and their distributions compared (Fig. 12). Each graph has the same scale for its x-axis. The most obvious feature of this figure is that 740b is the most common variety. Also noticeable is the shift in popularity of 740a and 740c, with 740a being more common in the earlier periods, and 740c in the later periods. This is generally true at both sites, but Mohenjodaro is stylistically more conservative in that 740b monopolizes the sign inventory with other varieties playing only minor roles. At Harappa the sign inventory is less concentrated, and even minor varieties (740a, b and especially e) can make up significant proportion of the inventory during the period of their maximum use. These are the sorts of distributions that are expected from purely graphic variations.

and 527

into two distinct, but related,

A similar problem arose with the analysis of signs 690 and 692

. For these these sign the counts are: sign 690

initial = 8, medial = 55, and terminal = 13; sign 692 initial = 42, medial = 6, and terminal = 3. Sign 690

is

primarily medial (72.4%) while sign 692 is primarily initial (82.4%). They are not only different graphemes, but have very different uses in Indus writing. These sorts of differences can also be seen in signs 920 and 921 . Sign 920

has the following distribution:

initial = 57, medial = 42, and terminal = 4; and 921 has: initial = 4, medial = 8, and terminal = 2. Again the division of these signs into two distinct graphemes seems to have been the right decision.

The variations in the distributions of the graphemes in Figure 13 are very different than those of the graphic variations in Figure 12. The graphemes have distributions that are more mutually exclusive than graphic variants. In other words, the distributions of signs are proportionally constant when compared to the total number of signs, while many variants have distributions that are mutually exclusive (see Fig. 12, signs 740c and 740e).

Conclusion In this paper we have examined in detail a methodology for creating a sign list for the Indus script. This method began with the grouping of signs on the basis of graphic design. Graphic variations were identified through the analysis of their contexts, and the resulting graphemes were analyzed for their chronological and spatial distributions. Additional considerations included artifact type, positioning in texts and pairings with other signs (including numerals). This methodology led to the creation of a sign list with 676 signs, but as has been pointed out earlier not all of these signs are certain to be separate graphemes. Instead, part of the sign list consists of sets of related signs that cannot be defined either as graphic variations or graphemes with certainty. Following a policy of maintaining detail, signs that fall into this category are maintained as separate signs awaiting further analysis. Figure 2 lists the ways in which Indus signs are constructed. These categories are also integrated into the Indus sign list. The number of elements used to construct signs (components) and the frequency of signs are to a high degree inversely proportional (Fig. 6). Complex signs with many components normally have low

The distributions of sign 740 reinforce the results of the contextual analysis that led the categorization of all varieties into a single grapheme represented by a standard graph: . This sort of analysis can be performed for many of the Indus signs, and is another source of data relating to the relationship between sign variants, but again it is restricted by low sign frequencies. Initial, Medial, and Terminal Signs The frequency of sign positions within texts (initial, medial, or terminal) can be easily calculated. There are 1999 texts (11509 signs) that have three or more signs and for which the texts are complete. These data are useful in several ways. First, they can identify signs that occur entirely or primarily (or never) at the start or end of 302

DEFINING THE INDUS SIGN LIST Museums. Meadow, R. H. and Kenoyer, J. M. (2001) Recent Discoveries and Highlights form Excavations at Harappa: 1998-2000. Indo Koko Kenkyu No. 22. Indian Archaeological Society, Tokyo. Meadow, R. H., Kenoyer, J. M. and Wright, R. P. (1994) Harappa Excavation 1994. Director General of Archaeology and Museums. Meadow, R. H., Kenoyer, J. M. and Wright, R. P. (1995) Harappa Excavation 1995. Director General of Archaeology and Museums. Meadow, R. H., Kenoyer, J. M. and Wright, R. P. (1996) Harappa Excavation 1996. Director General of Archaeology and Museums. Meadow, R. H., Kenoyer, J. M. and Wright, R. P. (1997) Harappa Excavation 1997. Director General of Archaeology and Museums. Meadow, R. H., Kenoyer, J. M. and Wright, R. P. (1998) Harappa Excavation 1998. Director General of Archaeology and Museums. Meadow, R. H., Kenoyer, J. M. and Wright, R. P. (1999) Harappa Excavation 1999. Director General of Archaeology and Museums. Meadow, R. H., Kenoyer, J. M. and Wright, R. P. (2001) Harappa Excavation 2000 and 2001. Director General of Archaeology and Museums Parpola, A. (1970) ‘The Indus Script Decipherment: the situation at the end of 1969’. Journal of Tamil Studies II: 89-110. Parpola, A. (1994) Deciphering the Indus Script. Cambridge University Press, Cambridge. Rao, S. R. (1984) ‘New Light on Indus Script and Language’. In Lal, B.B. and Gupta, S. P. (eds) Frontiers Of The Indus Civilization. I.M. Sharma of Books and Books, New Delhi: 193-200. Vats, M. S. (1940) Excavations at Harappa. Munshiram Manoharlal Publishers Pvt. Ltd., Delhi. Wells, B. (1999) An Introduction To Indus Writing. Early Site Research Foundation, Independence. Wells, B. (2006) Epigraphic Approaches to Indus Writing. PhD dissertation submitted to Harvard University.

frequencies. This pattern is also evident in both the protoElamite and proto-Cuneiform scripts, and results in a high proportion of singletons in the sign list. In addition to having large proportions of singletons, these three scripts share other features: a small number of very high frequency signs, many allographic variants, and logographic conflations including numbers. This paper examined the distributions of signs in detail. Far from being uniform these distributions showed that the Indus script is very complex in its regional usage, and that only the inventories from SEAL:S artifacts are significantly correlated between at all major sites. Differences between sites in terms of sign inventories can be attributed to difference in artifact type, and this is also true for different artifact types from the same site. For example, the sign inventories of TAB:B and TAB:I artifacts from Harappa are markedly different. The nature of these differences is also important, as they use similar sign inventories to write their texts, but in different proportions, with some signs being very common on specific artifact types. There are also differences in the signs used at the various sites. About half of all Indus signs are found at only one site as follows: Mohenjo-daro = 196 (134), Harappa = 103 (71), Lothal = 6 (6), Chanhujo-daro = 8 (7), Kalibangan = 11 (10), and Other = 10 (8) signs. The number of singletons is given in brackets. Just as important is that Mohenjo-daro and Harappa share 281 signs, and Mohenjo-daro, Harappa and Lothal share 118 signs. These signs form the majority of the Indus texts. While singletons comprise a large number of signs in the sign list (235 signs or about 35% of the sign list), they represent a small part (0.13%) of the 17423 sign occurrences recorded in the ICIT database. This sign list cannot be considered the final or definitive Indus sign list. Rather it marks just another stage in the ongoing process of analysis of this complex and interesting system of ancient writing.

Bibliographical References Knorozov, Y. (1970) ‘The Formal Analysis of the ProtoIndian Texts’. Journal of Tamil Studies II: 13-28. Mahadevan, I. (1970) ‘Dravidian Parallels in ProtoIndian Script’. Journal of Tamil Studies 2: 157276. Mahadevan, I. (1977) The Indus Script, Texts Concordance and Tables. Archaeological Survey of India, New Delhi. Mackay, E. J. H. (1938) Further Excavations at Mohenjo-daro. Munshiram Manoharlal Publishers Pvt. Ltd., Delhi. Marshall, J. (1931) Mohenjo-daro and the Indus Civilization. Arthur Probsthian, London. Meadow, R. H. and Kenoyer, J. M. (1993) Harappa Archaeological Research project: 1993 Excavations. Director General of Archaeology and 303

Bryan Wells

Fig. 1 – Evidence for perishable writing from Mohenjo-daro, Area DK.G (from Mackay 1938: Plate XC: 17a-c)

304

DEFINING THE INDUS SIGN LIST

1) Simple signs , ,

, and

2) Mirroring (sign reversal):

, etc. a) Allographic b) Graphemic

3) Doubling:

vs. vs.

a) Replication b) Reduplication

4) Inversion:

vs.

5) Elaboration:

a) Full

vs.

b) Graduated

vs.

vs.

c) Design

6) Compound Signs:

a) Attachments

i) Allographic

vs.

ii) Graphemic

vs.

i) Design elements ii) Signs iii) Strokes

b) Infixes

i) Design elements ii) Signs iii) Strokes

c) Conflations

i) Signs ii) Strokes

d) Markings

i) Triangular ii) Rectangular

7) Enclosures

a) Brackets

i) ii) iii)

b) Strokes

i) ii) iii)

c) Cages

i) ii) iii)

8) Strokes

a) Long Linear , , b) Short Linear , ,

etc. , etc.

c) Short Stacked , , 9) Multiple Class 10) Other

,

attached signs infixed,

etc.

conflated signs caged, etc.

, etc. Fig. 2 – Mechanisms used in sign construction.

305

and

Bryan Wells

Fig. 3a – The Indus signs by sign number and Set.

306

DEFINING THE INDUS SIGN LIST

Fig. 3b – The Indus signs by sign number.

307

Bryan Wells

Fig. 3c – The Indus signs by sign number.

308

DEFINING THE INDUS SIGN LIST

Fig. 4 – Variations in the ‘Bearer’ sign.

Fig. 5 – Sign 740 distributions resulting from counting programs.

80 70 60 50 40 30 20 10 1 component 0

2-6 component Complex Signs

2x 3-9 x

Allographic Sets

10-29 x 30-100 x 101+

Fig. 6 – Number of sign components by sign frequency.

309

1x

Bryan Wells

60

250

Sign Distributions For Indus, P-Elamite, and P-Sumerian Writing

Sign Frequencies 225 200

50 175

Indus Script

150

% of All Signs

40

125 100 75

30

50

r2 = 0.982

25 0 0

20

100

200

300

400

500

600

700

P roto-Sumerian

Indus %

10

P -S % P -E %

0 1x

2x

3-9x

10-99x

100-199x

200-399x

400-499x

500-699x

700+x

No. of Occurrences

Fig. 7 – Comparison of sign frequencies from three ancient scripts (data from Demerow 1989, Dahl 2005 and Wells 2006).

Fig. 8 – Comparison of sign inventories from Mohenjo-daro and Harappa.

310

DEFINING THE INDUS SIGN LIST

20

50

SEAL:R

SEAL:S 40

30

Harappa

Harappa

15

10

20 5 10

r2 = 0.610

r2 = 0.853

0 0

10

20 Mohenjo-daro

30

40

0 0

25

50 75 100 Mohenjo-daro

Fig. 9 – Frequency of signs by seal type from Mohenjo-daro and Harappa compared.

Fig. 10 – Singletons by artifacts type.

311

125

150

Bryan Wells

Fig. 11 – Chronological distribution of singletons at Moehenjo-daro, DK.G Area.

Fig. 12 – Sign 740 graphic variants over time from Mohenjo-daro and Harappa.

312

DEFINING THE INDUS SIGN LIST

Fig. 13 – Graphemic variations over time.

Table 1 – Ranked order and frequency by site for the 10 most common Indus signs.

313

SATELLITE REMOTE SENSING IMAGERY: NEW EVIDENCE FOR SITE DISTRIBUTIONS AND ECOLOGIES IN THE UPPER INDUS Rita Wright & Carrie Hritz

Since 1996, we have been conducting a landscape and mapping project along the now-dry bed of the Beas River in the Punjab, Pakistan65. Its focus is on urban and rural interactions between the city of Harappa and smaller settlements within the site’s catchment area. The current project integrates ground collected survey and remote sensing data sets to enhance our picture of landscape and settlement. Its purpose is to develop a more complete picture of ancient riverine environments and settlements beyond what is possible in ground survey. This method has been especially effective in the Near East, where it has proved crucial to the reconstruction of ancient regional landscapes (Wilkinson 2003, Hritz and Wilkinson 2006, Pournelle 2003, Ur 2003).

and Sutlej created an environment of shifting channels and settlement patterns throughout antiquity, resulting in a layered landscape of component features which represent past channel lines and settlements. To begin to untangle the complexity of this landscape, it is essential to first map landscape features and potential archaeological sites not detected on ground survey or outside of the ground survey boundaries. GIS, Remote Sensing and Datasets The GIS and remote sensing datasets proved crucial to broadening our understanding of the regional landscape and drew on methodologies previously developed in the Near East. The commonalities of landscape features and the dynamics of riverine environments suggested that similar methods could be applied to the Beas and the broader Punjab area. With the advent of new technologies such as declassified Corona Satellite photography66, high resolution Digital Globe Imagery and accessible GIS (Geographic Information Systems), a broad geographical and long term temporal view can be obtained and datasets integrated. By using GIS and incorporating diverse datasets for the region, it has been possible to map, integrate and conduct preliminary spatial analysis of the regional ancient landscape, both within and outside of the ground survey boundaries. Using a combination of spatial data collected during the Beas River landscape and settlement survey, and data recorded by digital photography and satellite imagery, site distributions can be traced. Mapping relict archaeological settlements and river channel changes, along with collected geomorphological data, we have begun to reconstruct human-environment interactions in the period of the first emergence of urbanism in the Upper Indus and suggest areas for future ground research.

Background to the Project Similar to many areas of the Near East, the archaeological evidence for early urbanism in the Upper Indus has been confined to limited survey and excavation of major urban centers. The research conducted by the Department of Archaeology of the Government of Pakistan Exploration Branch from 1992-1996 has provided an important baseline for the region in its identification of sites and monuments, including prehistoric and historic settlements (Mughal et al. 1993). The Beas project built on the goals of the earlier survey of the Exploration Branch by conducting a more intensive collection and mapping strategy, documenting visible surface traces, such as streets, kilns, and other activity areas for pre-urban and urban Harappan settlements, where surface traces were visible (Wright et al. 2005a,b). The project also included a geoarchaeological component for environmental reconstructions and the dynamics of the ancient Beas (Schuldenrein et al. 2007, 2004). New data based on archaeo-climate modeling, evidence from ground survey and the Harappa excavations, are resolving aspects of the environmental changes and human responses near the end of the Harappan period (Wright et al. 2008).

As Figure 1 illustrates, there are a number of datasets from past archaeological survey work, satellite imagery dating from the 1960s-1990s and geological and soil maps that were incorporated into a GIS for landscape reconstruction. Publicly available and geo-referenced Landsat images dating from 1987 and 1992 were used as a base for the creation of a GIS. The low resolution of the Landsat images (30 m resolution) prevented their use for small feature interpretation. Nonetheless, large features such as relict Beas channels did appear on these images, and they provided a first round indication of features that might be mapped on more high resolution imagery.

In this project, we utilized techniques not previously applied to this region in order to establish a more comprehensive understanding of the regional landscape context. Despite the systematic survey conducted by the Beas team, landscape reconstruction requires a more broad view of the landscape and settlement pattern in the Punjab area, afforded only by using remotely sensed datasets such as satellite imagery and maps when used in conjunction with ground survey and geomorphological data. The riverine dynamics of the Beas, Ravi, Chenab

66 The availability of declassified Corona imagery is largely due to Robert McC. Adams, who was instrumental in making these images available to the public.

65

The project is being conducted in collaboration with Dr. Joseph Schuldrein and M. Afzal Khan.

315

Rita Wright & Carrie Hritz date from prehistoric to 18th century AD. Detailed site descriptions, particularly from the Punjab survey publication, were selective and are used with caution.

A number of recent archaeological studies has demonstrated that Declassified Corona satellite photographs are a crucial dataset for ancient landscape reconstruction (Pournelle 2003, Hritz 2005, and Ur 2003). In brief, Corona was a US spy satellite taking satellite photographs of areas of strategic interest to US defense from the late 1950s-1979. Because the Coronas date to the late 1960s, they capture a landscape prior to expansion of the region’s agriculture base and other development projects common in the late 1980s.

Using the Corona images, it was possible to locate the 18 Beas sites and 118 of the Punjab surveyed sites in the study area, spanning the range of dates. These surveyed sites ranged in type from small mounded tells to mosques. These visible sites range in size from the smallest at 70(L) x 57(W) x 2.5 (H) m called Tibba Pipli Pathan dating to the 15-18th century AD and the largest 490(L) x 410(W) x 6 (H) m called Tibba Jati Usman dating to 8-12th century AD. On the left side of Figure 2 component features of the landscape are shown along with sites detected on the ground and through remote sensing. There are three types of sites visible that characterize ancient settlements in this area.

In the case of the Beas survey area and the Punjab in general, these images present a landscape which is undergoing increasingly intensive irrigation and agriculture projects. The landscape includes agricultural fields, dominated by large channels running from the Chenab, Ravi and Sutlej rivers and secondary local channels. Dune belts are scattered throughout the area between the rivers. The available Coronas for this project covered the entire Punjab province and dated to 10-101965. Each strip covers 20 x 180 km. These satellite photographs were scanned at 2400 dpi producing a resolution of between 2-5m. Figure 1 integrates GIS and Remote Sensing Datasets, including a baseline Landsat mosaic overlain by soil survey maps and corona images. The white lines mark the hydrology recorded on geological maps showing the Beas, while the grey are newly identified relict channels. White circles are sites identified by the Punjab survey and the dotted circles sites on the Old Beas.

1. 2. 3.

sites that are mounded and cast a shadow to the north; A mounded site with dunes around it in the fields; A site amidst dunes.

Note that the site has a darker color than the dunes and that is a key indicator of an archaeological mound. The color results from the combination of occupational and post-abandonment processes such as the retention of moisture in mudbricks, distinguishing archaeological mounds from dunes. Among the dunes and sites, channels appear as two light lines with a dark space between them indicating the presence of levee edges and a lower elevation of the actual relict channel bed.

While Corona satellite photographs and to a lesser degree the Landsat mosaics, provide a set of base geo-referenced imagery that can be used to detect possible archaeological features, ground truth of remotely sensed landscape features was accomplished by integrating the site and channel locations from the Beas ground survey and Punjab regional survey maps. It was possible to overlay these ground maps onto the satellite photographs to correlate features within a GIS. An additional key source of ground truth was provided by the incorporation of land use/soil maps of the area. These maps are particularly useful for ground truthing of remnant channel beds which were noted on the land use/soil maps. Once these datasets were incorporated into a GIS, it was possible to develop a remote sensing methodology to fill in gaps in the survey record of the Punjab including the detection of archaeological sites and relict channel systems, correlate those sites with channel systems and compare the settlement and distributions that emerge with those noted by archaeological survey and land use/soil maps.

Using the Corona images, we could not detect 114 of the sites that the Punjab survey detected. These sites again ranged in size and date and in most cases it is likely that the site is obscured by modern agricultural activity and modern villages. For example, some of the archaeological features noted by the Punjab survey are historical mosques or forts contained within modern villages. These features cannot be identified and separated from modern features on satellite imagery at the scale of the Coronas. Similar to the conditions of riverine settlement in southern Mesopotamia, these potential tells appear to cluster in a linear fashion along relict channel beds. Relict channel beds in the Punjab province tend to be infiltrated by dunes, again similar to conditions in the central portion of the Mesopotamian alluvium (shown on Figure 2 on the right). Unlike the large wide dune fields in southern Mesopotamia, the dunes that punctuate the agricultural fields in the Punjab are smaller and more distinct. Dunes can present a problem in the visibility of archaeological sites versus dunes. For example, in southern Mesopotamia, dunes are of the Bacharn type which exhibits a star or even crescent-shaped top. Archaeological sites in southern Mesopotamia are differentiated by their conical shape and their darker reflectance which is the result of moisture retained in the mudbrick which comprises the mound feature. A similar principle applies in the Punjab. While the dunes are much

Integration of Results The first landscape feature to be addressed using GIS and remote sensing was the location of ground surveyed archaeological sites and the detection of unsurveyed archaeological sites on satellite datasets. The Punjab survey noted 233 archaeological sites in the area between the Chenab, Ravi and Sutlej rivers, which includes the Beas survey area. The Beas survey focused on 18 preurban and urban Harappan sites. The 233 sites range in 316

SATELLITE REMOTE SENSING IMAGERY: NEW EVIDENCE FOR SITE DISTRIBUTIONS AND ECOLOGIES IN THE UPPER INDUS Beas River emerges.

less defined here, often appearing as bleached out areas on the satellite photographs, the mound or site features do tend to hold more moisture and appear to reflect darkly against the bright background

Examples are the possible relict channels of the Beas to the north and south of the location mapped by the Beas survey team. These layered relict channels are preserved as meanders in the modern agricultural fields. On Figure 3 (right) there are a series of relict meanders in the vicinity of the site Chak 90/121 a pre-urban and urban Harappan site. This possible relict channel is preserved as meanders lined by dunes and makes sense in terms of the site’s location. In general, in the Punjab, it seems that relict channels have been infilled by dunes. This type of infilling is common in other desert and riverine environments. Dunes tend to move to areas of lower elevation as would be the case with relict channel beds. By tracing the dunes and pieces of relict beds which appear within the dunes, it is possible to trace a large channel system running parallel to the Old Beas course mapped by the Beas survey to the north.

To fill in gaps of the surveys, 93 sites within the Punjab area were detected though not restricted to the Beas survey area (Figure 3 on the left side of the image). These sites are dispersed throughout the Punjab region, with seemingly dense clusters in the north-west near the Chenab River and Indus and in the southeast near the Sutlej. The identification of these sites was based on their tonal signature on Corona images. For example, a dark tonal signature on the north side is interpreted as elevation of the site against the flat agriculture fields, indicating the presence of a mound or tell-like feature. All potential sites in this area have been identified as possible mounded features and sherd scatters or small fort-like features, such as those noted by the Punjab survey, are likely not visible on the remote sensing datasets.

Additionally, it was possible to correlate the features appearing on the Corona photographs with those mapped by the land use/soil maps to provide ground truthing for these features. For example, this course sits on the edge of the scalloped interfluve and seems to be related to the Beas course. In a few areas, the land use/soil maps show a trace of river levee in this area. The interpretation of this relict bed is that it represents lateral shifts of the Beas River to the north. By incorporating geomorphological and archaeological survey data in the future, we may be able to correlate the shift of this feature and settlement and therefore address historical changes in settlement patterns.

The results of using remote sensing for possible site identification in this region are two fold. First, we are able to correlate a number of the archaeological sites detected by the ground surveys that are visible on the satellite photographs to determine what an archaeological site in this region should look like on a satellite photograph. Second, we were able to determine a methodology for identifying archaeological sites on remote sensing datasets and to begin to create site categories and plot changes in field patterns and site encroachment over time. This may be used as a future guide for surveys in the region. Unfortunately, we cannot date these sites without ground collection or a clear relationship between sites and dated relict channels, a project we hope to carry forward in the future.

Another area where the ancient landscape can be mapped is in the northwest of the Punjab that may bear relationships to the channel and settlement changes in the Beas. In this area, the remote sensing data not only allows for the mapping of ancient landscape features but also the methods by which those features might be dated (Figure 4 on the left). This Corona image shows the Beas hydrology (in white) and relict channels (light grey). On Figure 4 (on the left) the GIS image includes various data and illustrates relict channels, Beas sites identified in ground survey and others identified from remote sensing imagery. The remains of channel systems in this area demonstrate the complex stratigraphy of ancient landscape features in the Punjab province. In this area (Figure 4 image on the right), large relict meanders are preserved as moisture marks in the agricultural fields. These meanders seem related to the Chenab River and represent a relict course that flowed towards the Old Beas. It is quite clear in the north, but as one moves south towards the Beas the relict meanders become overtaken by large modern channels, roads and intensive agricultural fields.

Channel Systems in the Punjab The location of archaeological sites in the hinterlands of Harappa is useful, but in order to understand the landscape context for the rise of these urban centers, the channel systems which fed these cities must also be mapped and related to the archaeological sites. A number of relict channel remains appear in the landscape in the Punjab region. By mapping a few key remains of channel system in detail, it now seems possible that the dynamics of the Beas River and its tributaries may be associated with settlement pattern shifts. These channels are shown on the left side of the Figure 3. On the left known Beas sites and hydrology are combined with data on relict channels. The Beas survey outlined and mapped the ancient course of the now largely abandoned Beas river channel. Elsewhere we have discussed the morphology of the Beas (Shuldenrein et al. 2004, 2007) and incorporated paleoclimatic data to understand the shift in settlement and the drying of the Beas river channel in particular periods of antiquity (Wright et al. 2008). However, by including Corona images, a more complex picture of the shifting

This feature is lined by archaeological sites detected in the Coronas and also by ground survey by the PAS. The majority of sites located directly along the main branch of this relict channel date to between the 5-15th century AD. Interestingly, a secondary take off branch, contains three 317

Rita Wright & Carrie Hritz Preliminary Results of Explorations: 1992-1996’. Pakistan Archaoelogy, Special Number 29:1-474. Pournelle, J. (2003) Marshland of Cities: Deltaic Landscapes and the evolution of early Mesopotamian Civilization. Unpublished dissertation. Department of Anthropology, University of California San Diego. Schuldenrein, J., Wright R. P., Khan M. A. (2007) ‘Harappan Geoarchaeology Reconsidered: Holocene Landscapes and Environments of the Greater Indus Plain’. In E. Stone (ed.) Settlement and Society, UCLA Cotsen Institute of Archaeology: 83- 116. Schuldenrein, J., Wright R. P., Afzal Khan M., Rafique Mughal M. (2004) ‘Geoarchaeological Explorations on the Upper Beas Drainage: Landscape and Settlement in the Upper Indus Valley, Punjab. Pakistan’. Journal of Archaeological Sciences, 31: 777-792. Ur, Jason (2003) ‘CORONA Satellite Photography and Ancient Road Networks: A Northern Mesopotamian Case Study’. Antiquity, 77:102115. Wilkinson, T. J. (2003) Archaeological Landscapes of the Near East. University of Arizona Press. Tuscon. Wright, R. P., Schuldenrein, J., Khan M. A., Mughal M. R. (2005a) ‘The Emergence of Satellite Communities along the Beas Drainage: Preliminary Results from Lahoma Lal Tibba and Chak Purbane Syal’. In Jarrige, C. and Lefevre, V. (eds) South Asia Archaeology 2001. Paris: Editions Recherce sur les Civilisations - ADPF l: 327-335. Wright, R. P., Schuldenrein, J., Khan M. A. and MalinBoyce, S. (2005b) ‘The Beas River Landscape and Settlement Survey: Preliminary Results from the Site of Vainiwal’. In Franke-Vogt, U. and Weisshaar, H-J. South Asian Archaeology 2003. Aachen: Linden Soft: 101-111. Wright, R. P., Bryson R., Schuldenrein J. (2008) ‘Water Supply and History: Harappa and the Beas Settlement Survey’. Antiquity, 82: 37-48.

dated sites from the Punjab survey, and an additional two sites from remote sensing, that date to the 8-16th century AD. In this case, we can provide a tentative date for this channel system from the Chenab to between the 5-15th century AD, the dates provided by the Punjab survey. We can plot the development of a secondary branch of this system which presumably allowed the expansion of settlement in this area by providing water into a new area to the east, dating to the 8-16th century AD (Figure 4 on the right). This example shows how the stratigraphy of this feature can be mapped on the remote sensing data and correlated with the survey data to begin to describe the relationship between changing channel systems and settlement patterns. Conclusions This paper has demonstrated the utility of remote sensing and GIS in landscape studies in the Punjab province of Pakistan. It has shown that with the incorporation of diverse datasets including satellite photography, geological/soil maps and past archaeological survey data, it is possible to begin to map additional landscape features not visible in ground survey. The features mapped in this landscape provide the context for the development of large cities in the region and present a more nuanced picture of the hinterland of large sites such as Harappa. At its most basic level, this study has shown the presence of a number of potential archaeological sites within the Punjab survey area that escape detection during ground survey. We have provided a methodology for detecting and interpreting ancient landscape features in the Punjab province that may be used as a guide for planning and carrying out future survey projects in this region. Acknowledgments The Beas project was supported by grants from the Nation Endowment for the Humanities (RX2025798), the National Geographic Society (608297 and 642799) and the Wenner-Gren Foundation for Anthropological Research (ICRG11). The Landsat mosaics were provided by the NASA via the Global Land Cover Facility of the University of Maryland. Bibliographical References Bryson, R., Wright R. P., and Schuldenrein J. (2007) ‘Modeling Holocene Climates and Rivers in the Harappa Vicinity’. In Bryson, R. A. and DeWall, K. McE. A Paleoclimatology Workbook: High Reoslution, SiteSpecific, Macrophysical Climate Modeling. The Mammoth Site of Hot Springs, SD, nc:123-128. Hritz C. and Wilkinson, T. J. (2006) ‘Using Shuttle Radar Topography to map ancient water channels in Mesopotamia’. Antiquity, 80: 415-424. Mughal, M. R., Iqbal F., Khan M. A., Hassan M. (1996) ‘Archaeological Sites and Monuments in Punjab,

318

SATELLITE REMOTE SENSING IMAGERY: NEW EVIDENCE FOR SITE DISTRIBUTIONS AND ECOLOGIES IN THE UPPER INDUS

Fig. 1 - This figure shows the integration of diverse datasets geo-referenced into a GIS for spatial analysis. Datasets include a Landsat mosaic, a soil map, Corona satellite photographs, and vectorized ground survey data of archaeological sites, the line of the Old Beas and hydrology in the area. Landsat mosaics courtesy of the Global Land cover Facility; Corona images courtesy of USGS.

319

Rita Wright & Carrie Hritz

Fig. 2 - Types of possible mound sites and components of the landscape on Corona satellite photographs of the Punjab compared to archaeological sites and dunes detected on Corona satellite photographs of southern Mesopotamia. Corona images courtesy of USGS.

Fig. 3 - Series of relict meanders on a Corona satellite photographs around the site of Chak 90/121 an early and mature Harappan site. Corona images courtesy of USGS.

320

SATELLITE REMOTE SENSING IMAGERY: NEW EVIDENCE FOR SITE DISTRIBUTIONS AND ECOLOGIES IN THE UPPER INDUS

Fig. 4 - One the left the GIS shows mapped relict channels and identified archaeological sites. One the right, a Corona satellite photograph shows secondary branch development identified by Remote Sensing. Corona images courtesy of USGS.

321