Sheri Khan Tarakai and Early Village Life in the Borderlands of North-West Pakistan: Bannu Archaeological Project Surveys and Excavations 1985-2001

By Cameron A. Petrie, Farid Khan, J. R. Knox and Kenneth D. Thomas

Between 1985 and 2001, the collaborative research initiative known as the Bannu Archaeological Project conducted archaeological explorations and excavations in the Bannu region, in what was then the North West Frontier Province (NWFP) of Pakistan, now Khyber-Pakhtunkhwa. The Project involves scholars from the Pakistan Heritage Society, the British Museum, the Institute of Archaeology (UCL), Bryn Mawr College and the University of Cambridge. This is the first in a series of volumes that present the final reports of the exploration and excavations carried out by the Bannu Archaeological Project. It marks the first attempt to contextualise the earliest village settlements in northwest Pakistan, along with those situated in other parts of the borderlands zone at the western margins of South Asia. An extensive range of archaeological data from the Bannu Archaeological Project excavations at Sheri Khan Tarakai, including stratigraphic, architectural, ceramic, lithic, small find and bioarchaeological elements, are presented, along with the results of surveys and excavations at several other sites in the Bannu Basin and the adjacent Gomal Plain. The work establishes the nature of the relationships between these sites and other early villages elsewhere in South, central and greater West Asia.

• Language: English
• Publisher: Oxbow Books
• Release date: May 10, 2010
• ISBN: 9781842177358

Cameron A. Petrie

Cameron Petrie is the Senior Lecturer in South Asian and Iranian archaeology at the University of Cambridge (UK). He has extensive field and research experience in India, Pakistan and Iran, and has co-directed collaborative research projects in each of these countries.

Book preview

Preface and Acknowledgements

The topographic region known as the Bannu basin (currently comprised of the administrative Districts of Bannu and Lakki-Marwat) lies in the southern part of Khyber-Pakhtunkhwa, formerly known as the North West Frontier Province (NWFP) of Pakistan (Fig. I). Prior to the 1970s, the archaeology of this region, and the immediately adjacent areas, received relatively limited attention, and most of this interest was focused on a number of major mound and temple sites that have since been dated to the Early Historic and later periods (see Chapter 2.1). In 1904/05 and 1927 Sir M. Aurel Stein made brief forays into both the Bannu basin and the Gomal plain, which lies to the south (Stein 1905, 1929), but he did not observe any prehistoric occupation in these areas and carried out no formal excavations there. After the establishment of the Department of Archaeology at the University of Peshawar, a brief prospection of the southern NWFP led by the late Prof. A.H. Dani took place in 1967 (Dani 1968) as part of a comprehensive reconnaissance of the archaeology of the province. This was followed by further survey led by the late Prof. F.A. Durrani in 1969 (Dani 1970–71: 22) (Chapter 2.1). These surveys led to excavations at Gumla in 1971 (under Dani) and Rehman Dheri in 1976–79 (under Durrani), both of which lie in the north of the Gomal plain. Systematic archaeological surveys of the Bannu basin began in the 1970s, and excavations by a collaborative project involving the Department of Archaeology, University of Peshawar and the Cambridge Archaeological Mission to Pakistan followed. The latter team conducted fieldwork until 1978/1979 and included three of the authors of this volume (see below).

The collaborative research initiative known as the Bannu Archaeological Project conducted explorations and excavations in the Bannu basin between 1985 and 2001. Although fieldwork in the region halted in 2001, the research collaboration has continued and currently involves scholars from the Pakistan Heritage Society (Peshawar), the British Museum, the Institute of Archaeology (University College London), Bryn Mawr College and the University of Cambridge. This volume presents the results of the excavations at Sheri Khan Tarakai in the Jani Khel region of Bannu basin and several other sites. It is the first in a series presenting the final reports of the field surveys and excavations carried out by the Bannu Archaeological Project in the Bannu basin and the neighbouring Gomal plain.

The Bannu Archaeological Project

Archaeological work concentrating on the prehistory of the Bannu basin was begun by Farid Khan in the mid-1970s. He was one of the first students to enrol in the Department of Archaeology at the University of Peshawar and one of the first of its graduates appointed to staff the Department. He is also a graduate of the University of Cambridge. Later he would go on to become Professor and Head of the Department of Archaeology in Peshawar, before retiring from the University in 1994. Farid Khan was born into an old Bannu family in the village of Bazaar Ahmed Khan, which in earlier times was the administrative headquarters of the District. He was the first person to engage in the systematic exploration and recording of ancient sites in Bannu and was uniquely informed as a local man with an extensive knowledge of the region and a clear view of the immense scope for archaeological work there. His early field research there involved extensive exploration, frequently performed under difficult circumstances in insecure areas, and drew extensively on his personal familiarity with Bannu and that of his many friends and contacts in the area.

Figure I. Map showing the location of the Bannu basin and the Gomal plain, which lie at the western edge of both Pakistan and the subcontinent.

In 1976, Farid Khan initiated a collaborative project between the Department of Archaeology, University of Peshawar and the Cambridge Archaeological Mission to Pakistan that focused on the Bannu basin. This enterprise was led by the late Prof. F.A. Durrani, then Head of the Peshawar Department, together with the late Dr F.R. Allchin of the Department of Oriental Studies, University of Cambridge, and Dr B. Allchin of Wolfson (formerly University) College, Cambridge. It was ably supported by Farid Khan, Dr Abdur Rehman and Dr Taj Ali, all of the Department of Archaeology at the University of Peshawar, as well as Prof. Mohammad Said of the Department of Geography and, later, Vice-Chancellor of the University of Peshawar. The group of British researchers taken into the field at that time included Dr Ken Thomas of the then Institute of Archaeology of the University of London (from 2008, Emeritus Professor of Human Palaeoecology at University College London’s Institute of Archaeology). Robert Knox of the then Department of Oriental Antiquities (now the Department of Asia) at the British Museum (until 2006, Keeper of the Department of Asia), and Dr Helen Rendell, then of the University of Sussex (now Professor of Geography, Loughborough University). Ken Thomas and Robert Knox are most grateful to Bridget and Raymond Allchin for providing them with this unique and valuable introduction to the archaeology and to the people of Pakistan. For them it has been a fulfilling experience extending over more than three decades, resulting in strong friendships and complex scholarly activities that have endured to the present day.

The collaborative Peshawar-Cambridge project began work in the Bannu basin, building on Farid Khan’s surveys and carrying out excavations at two sites dating to the third millennium BC: Lewan (1977–1978) and Tarakai Qila (1978–1979). The results of the excavations at Lewan were published in several preliminary reports (B. Allchin 1981; F.R. Allchin and Knox 1981) and an edited monograph (F.R. Allchin et al. 1986). As yet, only a small number of preliminary reports of the excavations at Tarakai Qila have been published (B. Allchin 1981; F.R. Allchin and Knox 1981; Thomas 1983b). The Peshawar-Cambridge collaboration ended after the 1978–1979 season.

Collaborative research in the Bannu region was resumed in 1985 in the shape of a freshly constituted collaboration between researchers from Peshawar and the UK, which was called the Bannu Archaeological Project. This enterprise was established as a joint venture headed by Farid Khan of the University of Peshawar, in collaboration with Robert Knox of the British Museum and Ken Thomas of University College London (UCL) (Fig. II). The Bannu Archaeological Project carried out field research during virtually every year between 1985 and 2001.

With the transfer of the administration of this joint archaeological project to the Pakistan Heritage Society in 1994, the collaborative project assumed a new form, though retaining close links with the University of Peshawar. The Pakistan Heritage Society is based in Peshawar and Prof. Farid Khan continues as its Chairman. In 1996, Prof. Peter Magee, then of the University of Sydney in Australia and now of Bryn Mawr College in the USA, joined the Bannu Archaeological Project as a joint director, though too late for the work at Sheri Khan Tarakai. In 1998, the Project was expanded to include Dr Justin Morris (former curator in the Department of Asia and currently Director of Strategic Planning at the British Museum) and Dr Cameron Petrie (formerly of the University of Sydney and now Lecturer in South Asian Archaeology at the University of Cambridge) (Fig. III). Dr Petrie is General Editor of this volume.

The Bannu Archaeological Project intended to explore as wide an area of the Bannu basin as was both possible and practicable, to record ancient sites of all types, and conduct excavations at important and/or endangered sites. The project initially focused on characterizing the prehistoric sequence in the region, but no chronological constraints were imposed on the surveys and excavations that were carried out. In fact, the surveys identified sites ranging in date from the Upper Palaeolithic to the Islamic period, while excavations were carried out at sites ranging in date from the later prehistoric (or protohistoric) to the Islamic periods.

The Bannu Archaeological Project also aimed to contribute to the development of the study of archaeology in northwest Pakistan and to include in its activities the teaching of students in the field. Each year, graduate (M.A. and M.Phil.) students from the University of Peshawar came to Bannu for periods of training in field techniques and post-excavation analysis. In addition, various students from the Institute of Archaeology (University College London) and the University of Sydney participated in the Project’s excavations. The Project also aimed to broaden the general interest and awareness of the citizens of Bannu District in their past, and with good sense and foresight, land has been set aside in Bannu City for the establishment of a museum to display the material produced by the Project, and collected from other local sources.

Figure II. The founding members of the Bannu Archaeological Project (Ken Thomas, Farid Khan and Robert Knox) in 1988.

Figure III. Bannu Archaeological Project Team – 1986. Back row L to R: Iqbal, Farid Khan, Makin Khan, John Gowlett, Paul Garwood, M. Daud Kamal, front row L to R: Robert Knox, Shehzada Khan, Ken Thomas.

The work carried out by the Project can be divided into several major phases:

1985: an expanded archaeological survey of the district located a significant number of new sites, including the site of Sheri Khan Tarakai, with a hitherto unknown late prehistoric/protohistoric material culture assemblage (Fig. IV);

1986 to 1994: excavations were carried out at Sheri Khan Tarakai and other related sites that were discovered subsequently. Work focused on clarifying the nature of cultural transitions during later prehistory, investigating past environments and subsistence economies, and establishing a radiocarbon dating framework;

1994/5 to 2000: survey and excavation was carried out at the Early Historic period sites of Ter Kala Dheri and Akra, during which a new ceramic horizon, Bannu black-on-red ware was discovered, in association with early iron. This ware has been dated to between 900 and 600 cal BC. Between 1996 and 2000 the major focus was on the excavations of huge mounds at Akra, which are adjacent to Bharat village to the south of Bannu City;

Work on the later prehistoric periods in Bannu was resumed in 2000, with a topographical survey and new excavations at the site of Lewan that continued in 2001;

In 1998 and 2000 archaeological surveys were made in the Gomal plain in the area around Jhandi and Gandi Umar Khan villages, where a complex settlement system was studied.

Figure IV. Robert Knox carrying the complete vessel recovered from Trench II in 1986.

It is important to emphasize that this fieldwork was conducted under quite difficult circumstances. During the 1980s and 1990s, field research in Bannu was only possible with an armed escort, known as khassadars or locally recruited bodyguards, in attendance (Fig. V). These escorts were drawn either from the District constabulary or from local sources, and although this often resulted in unpredictable and often protracted delays, they were not just for show. Disturbances close to the sites that were being investigated meant that surveys and excavations could be interrupted indefinitely at very short notice. As a result, the length of individual field seasons was variable and days that were intended to be spent on-site became devoted to work in the field house or on other activities, such as ethnographic recording. Security limitations also meant that it was impossible to achieve widespread survey coverage of the entire region. This situation had a direct impact on the range, variety and location of the regions and sites investigated, and it also applied tight constraints on the type of sampling that was possible, particularly with regard to sampling raw material sources. At the time, this situation and the precautions being taken might have seemed to an outsider to be a little extreme, but changing political circumstances in recent years have proved that the conditions in Bannu during the 1980s and 90s were relatively benign. Archaeological work in Bannu ceased with the spring 2001 season, and it has not been possible to revive the fieldwork component of this collaborative project due to the serious insecurity in the region that has now turned to extreme violence, particularly in regions adjacent to the settled areas of Bannu District.

Nevertheless, the scope for archaeological work in this part of Pakistan is immense. Although sites ranging in date from the Upper Palaeolithic to the Islamic period have been identified, it is notable that the sequence of occupation phases in the Bannu region does not mirror that seen elsewhere in Pakistan. For example, no trace of the Indus Civilisation has been detected anywhere in the Bannu basin, which stands in marked contrast to the situation in the Gomal plain to the south and the Indus plain to the east. The possibility that such sites might exist cannot be ruled out and surveys of parts of the region that have not yet been investigated in detail should continue when stability returns to the region. There is much still to be learned in this remote place and we hope to be there again when the time is right to begin once more.

Figure V. Excavation team with khassadars and local boys at Sheri Khan Tarakai 1987. Back row, 5th from the left Nasim Khan, 7th from the left Robert Knox, 9th from the left M. Daud Kamal, 10th from the left Ken Thomas.

Acknowledgements

The Bannu Archaeological Project has proved to be an important model for successful collaboration between local and foreign archaeological researchers working in Pakistan. One of the most significant results of this fruitful collaboration has been the development and maintenance of a series of enduring close contacts between its members. It is the belief of the team that great things can be the result of such close, mutually respectful contacts in the context of determined hard work and that the Bannu Archaeological Project group represents one of the best and most fruitful such collaborative arrangements ever developed in post-Independence Pakistan. Its success, however, would not have been possible without essential help from various quarters, as follows.

GOVERNMENT OF PAKISTAN DEPARTMENT OF ARCHAEOLOGY AND MUSEUMS

The Bannu Archaeological Project owes its existence to the encouragement and support of many people. Without the generous support of successive Director Generals of the Government of Pakistan Department of Archaeology and Museums, the late Mr M. Ishtiaq Khan and Dr Ahmed Nabi Khan, the Project could never have resumed its work when it did. Former Director Generals Mr Said-ur-Rehman (himself a native of Bannu) and Dr M. Rafique Mughal, and the current Director General Dr Fazal Dad Kakar, have been of enormous help over the years and remain our very good friends. Mr Khurshid Hassan of the Pakistan Department of Archaeology and Museums, and our close friend and colleague Mr A. Halim, former Director of the National Museum of Pakistan at Karachi, similarly provided encouragement and assistance for which we are deeply grateful. In the field at Sheri Khan Tarakai, members of the staff of the Pakistan Government Department of Archaeology, Mr Makin Khan, Dr Farzand Massih and Mr Azeem Khan acted as official government representatives in the Project. They worked with great dedication for us in the Jani Khel and we thank them most sincerely.

PESHAWAR

The kind encouragement and assistance of successive Vice-Chancellors of Peshawar University, Prof. Khan Tahirkheli and Dr Abdul Matin in particular, in the early days of the Project, were of enormous importance to us. At the Department of Archaeology of the University of Peshawar, the late Prof. F.A. Durrani (former Vice-Chancellor of the University), and Prof. Dr Abdur Rehman were our close associates for many years, and without their constant interest and support this work would never have been undertaken. We are also immensely grateful to Dr Rehman for translating the summary to this volume into Urdu. The late Prof. A.H. Dani of the Quaid-i-Azam University, Islamabad, was enormously helpful, encouraging us especially during the formative stages of the Project and publishing our first article on Sheri Khan Tarakai in his departmental journal. We thank him most sincerely.

Prof. Farooq Swati, Chairman of the Department of Archaeology, and also Provost of Peshawar University, and Prof. Nasim Khan of its Department of Archaeology have been warm friends and superb colleagues for many years, along with their class fellow Prof Ashraf Khan, now of Quaid-i Azam University, Islamabad. We thank them most particularly for their support and encouragement, as well as that of our colleague Prof. Ihsan Ali, now Vice-Chancellor of the Universities of Hazara and Mardan. Mr Assad Ali, photographer in the Department of Archaeology, University of Peshawar, was invaluable in his many services to the project and for his enduring and infectious good humour. Mr Muhammad Naeem from the Peshawar Department assisted at Sheri Khan Tarakai and we are grateful to him for his dedicated service.

Also in Peshawar, old friends at Edwardes College, the late Mr Michael Close OBE and Dr Tim Woolmer, Principal of the College in the very early days of the Project, were always ready with tea and typewriters in the days before portable computers, and for their kindness we shall always be most thankful.

BANNU DISTRICT ADMINISTRATION

At Bannu, successive Deputy Commissioners of the Bannu District: Mr Mohammad Rafique Afridi, Mr A.K. Safi and Mr Shah F. Ahmed, and the Assistant Political Agents for the Frontier Region, Bannu District: Mr Haseeb Athar and later Mr Gul Aslam Khan, were of incalculable assistance to the Project in its first years. The then Political Tahsildar at Bannu, Mr Khan Baksh Marwat, provided invaluable assistance over many seasons. We thank him and remember him most particularly for his contribution to our work and our discoveries. These highly talented gentlemen generously gave us of their valuable time and facilities in arranging matters with local people on our behalf.

This assistance and hospitality was both highly gratifying personally and of fundamental importance to our work in a region of often uncertain safety. The provision of essential security in the field for the team was assumed each season by the District (later Divisional) authorities. We thank those responsible for this generous gesture, in particularly the Superintendents of Police for Bannu District who provided us with security personnel from their own ranks for this purpose. We are grateful also to successive District Commissioners (and later Commissioners of Bannu Division) who lent us the daily protection of armed khassadars provided by Waziri families local to the Sheri Khan Tarakai area, from the Jani Khel itself, the adjoining Sardi Khel and other nearby areas including the Bakka Khel (Fig. IV). We need to thank warmly as well the many unnamed officials at all levels of the Government of Pakistan, and of the Government of the NWFP, who helped the Project to carry out its work in this complex region of the North West Frontier Province (now Khyber-Pakhtunkhwa).

BANNU

We are particularly grateful to Malik Mir Taj Ali Khan of the Wali Noor area on the edge of Tribal Waziristan for taking us by chance one day in 1985 to the site of Sheri Khan Tarakai in the Jani Khel. Without this energetic and helpful man, and his ability to recognize what we were looking for, this important site would not have been discovered. The late Mr Sheri Khan, owner of the ancient mound that now bears his name, was a very old man when we went first to his tiny village perched on one end of the long, low sandbarlike deposit that turned out to be the type-site for the earliest known village occupation in Bannu. The team is grateful to Sheri Khan for allowing us to carry out archaeological work on his land and to his family (in particular to his sons Fazle Manan and Yusuf) for permitting us to continue to work there after his death. We remember as well the many men and boys of the immediate area who worked for us diligently and with great care over the six seasons of our excavations at the site. We are grateful for their hard work and for their traditional Frontier hospitality.

Under the wise and kind leadership of Dr Ruth Coggan OBE, the staff of the Pennell Memorial Hospital at Bannu were our warm friends and staunch supporters. They were always graciously hospitable to the team and we thank them most warmly.

Possibly most important of all, the team wishes to thank Mr Feroz Khan of Bannu City, brother of one of the authors of this volume, who assisted the Project in more ways than can be numbered here. His organizational and diplomatic skills made it possible for the Project to establish itself in Bannu and for the work to proceed smoothly from the start of the enterprise to its final days. We honour him and thank him most profoundly for what he achieved for us over many years.

TECHNICAL COLLABORATION AND ASSISTANCE

The British Museum Research Laboratory was a particularly important supporter of the work of the Bannu Archaeological Project over many years and we are hugely grateful to all our scientific colleagues there for their continuing support. In 2002 Dr Louise Joyner carried out thin-section petrographic and pigment analysis of a small number of ceramic samples from Sheri Khan Tarakai and other sites in the Bannu basin and the results of this analysis are presented in this volume. Dr Caroline Cartwright has provided an invaluable report on the wood charcoals from Sheri Khan Tarakai, which also appears in this volume. Dr Sheridan Bowman and Janet Ambers processed numerous charcoal samples for radiocarbon dates, and we are grateful as well to Dr Mike Tite and Dr Bowman, former Keepers of the Department of Scientific Research, for facilitating these analyses. We would also like to thank Dr David Saunders, current head of the Department, now called Conservation and Scientific Research, who has kindly granted us permission to publish Dr Joyner’s original reports. We would also like to thank Dr Nick Ashton, curator of the Department of Prehistory and Europe at the British Museum, for his assistance in the interpretation of the geomorphology of the Bannu basin and the late prehistoric/protohistoric lithic assemblages, and Mr Karl Lee of Primitive Technology UK for his work on the experimental flint knapping.

We are grateful to Dr Catherine Jarrige of CNRS and the Musée Guimet in Paris for her wise remarks on the Sheri Khan Tarakai terracottas, as well as to Dr M. Rafique Mughal, former Director General of Archaeology and Museums of the Government of Pakistan, for his many helpful comments on the Sheri Khan Tarakai material. We wish to thank Dr J.A.J. Gowlett for his invaluable work with us during the 1986 field season and particularly for taking the first samples for radiocarbon dating from Sheri Khan Tarakai (discussed in Chapters 3 and 8). In 1986 he was on the staff of the Oxford University Radiocarbon Accelerator Unit and is now Professor at the School of Archaeology, Classics and Egyptology of the University of Liverpool. We thank the Oxford University Radiocarbon Accelerator Unit for processing that material. We are grateful as well to Dr Paul Garwood for his assistance in the field in 1986 and to the late Mr Neville Parker who worked with us to great effect in 1990. We wish also to thank Dr Susan Colledge of University College London’s Institute of Archaeology for discussion of the charred seeds and grains from Sheri Khan Tarakai and Dr Louise Martin, also of the Institute of Archaeology, for her advice on criteria for the identification of gazelle bones.

Drawings for this publication were prepared in the field in Pakistan and completed in London by Grahame Reed. Additional drawings were made in Pakistan by Mr Muhammad Daud Kamal, formerly of the University of Peshawar Department of Archaeology, and completed by Ms Annie Searight at the British Museum. Additional lithic illustrations were prepared by Hazel Martingell. Hajji Sultan, whose commercial surveying firm is based in the bazaar in the Bannu Cantonment, completed an accurate survey of the topography of Sheri Khan Tarakai and its archaeological features. The translation of the summary at the beginning of the volume into Urdu was fascilitated by Prof. Nasim Khan.

The original members of the Bannu Archaeological Project team (F. Khan, Knox and Thomas) are grateful to Dr Cameron Petrie for his energetic work as General Editor of this volume and for seeing it through to publication. His efforts on behalf of the Project have been of fundamental importance to the production of this final report. Dr Justin Morris’s major work on lithic cultures of North West Pakistan (Morris 2004) has provided a secure foundation for the analysis of the Sheri Khan Tarakai (informally known as ‘SKT’) stone tool complex and it has a central place in this publication. We are delighted to include these two young scholars in the list of principal authors of this final ‘SKT Volume’.

Funding and support

Since its inception in 1985, the Bannu Archaeological Project has been supported financially, administratively and generally by a large number of institutions, departments, societies, individuals and other bodies. Major financial support came annually from the British Museum. The team is most grateful to the Trustees of the British Museum for their contribution to this Project and for supporting its enterprise in Pakistan for so many years. Without the unstinting support of this distinguished body, the many and varied achievements of the Project could not have taken place. Financial contributions were received in every season from the Society for South Asian Studies (British Academy, London), and the University of London’s Central Research Fund as well as its Hayter Fund, particularly in the area of funding for travel to Pakistan for the British members of the team. The fieldwork conducted in 1998 and 2000 was largely funded by a research grant from the Australian Research Council. The British Museum has provided funds for the publication of these final report volumes, and we would particularly like to thank the Keeper of the Department of Asia, Dr Jan Stuart for her kind and continuing cooperation, and Dr J.D. Hill, the Museum’s Research Manager.

Significant practical assistance to the Bannu Archaeological Project came regularly from the University of Peshawar from the beginning of our collaboration with its Department of Archaeology and later from the Pakistan Heritage Society, both providing support personnel, technical backup, object storage, office space and many other basic amenities. The University of Peshawar and the Pakistan Heritage Society were each, in turn, the senior partners in our collaboration, and the names of the British and other members of the joint project team were attached annually to the excavation licence granted to the local body alone. We are particularly grateful for their logistical support, especially in the area of organization, in dealing with various ministries of the Pakistan Government in getting permits, licences and clearances of all kinds for the Project and acting as principal liaison with the Government of Pakistan Department of Archaeology and Museums.

Successive Directors of the British Museum, Sir David Wilson and Dr Robert Anderson, lent their unqualified backing to the Project from the outset and as it developed. We thank them for this invaluable support for our work over many years. We are grateful also to successive Directors of the Institute of Archaeology, Professors John Evans, David Harris and the late Peter Ucko for their staunch support for the Project. Prof. Grahame Barker, Head of the Department of Archaeology at the University of Cambridge, has supported the preparation of this publication and we are deeply grateful to him for his many kindnesses, which include reading our manuscript in pre-publication format. We thank him most particularly for his enthusiasm and encouragement. We would also like to thank Prof. Dilip Chakrabarti, a good friend and long a supporter of the project, for commenting on a draft manuscript.

FINALLY.

Archaeological fieldwork in Bannu is over for the time being. It seems only yesterday that the Bannu team was actively exploring and field walking, surface collecting and excavating ancient sites in this fascinating and archaeologically highly important area. Although this region has been known for generations for its insecurity, the Project team was always treated with the greatest courtesy. We were received everywhere with kindness, hospitality and consideration. The team was permitted to explore the countryside in relative freedom, to work on people’s lands, often to visit their houses and to share their food. We were always sorry to leave Bannu at the end of each season and invariably glad to return.

Our happy time in Bannu sits in stark contrast to the situation of conflict and warfare that casts a long shadow today over its green farmlands, barren wastes and rich ancient history. Farid Khan has said that, in retrospect, our years working together there seemed like a ‘wonderful dream’. He was right, but the massive results of our archaeological work in Bannu, some of which appear in this volume, show that it was also a substantial and highly successful scholarly enterprise, breaking much new ground and pointing the way to important future work in the border regions when the time is right once again.

Farid Khan

Robert Knox

Ken Thomas

Peshawar and London

2010

A note on terminology

In South Asia, a very specific terminology is used to differentiate the various chronological periods in the past. For example, the term ‘prehistoric archaeology’ typically refers to the Palaeolithic and Mesolithic periods pre-7000 BC (e.g. Sankalia 1962). The term ‘protohistoric’ is typically used to refer to the period beginning in the early farming or Neolithic period, roughly around c. 7000, and continues until c. 700 BC. The Early Historic period is generally perceived to begin in the early to mid-first millennium BC.

Outside South Asia, the protohistoric period is most typically referred to as late prehistory. To ensure that this volume is as broadly relevant as possible, and also ensure due consideration to the terminology used in the subcontinent, both ‘late prehistoric’ and ‘protohistoric’ will be used in conjunction in this volume (i.e. late prehistoric/protohistoric).

The Bannu Archaeological Project has investigated most of the major late prehistoric/protohistoric and historic cultural phases attested in the Bannu basin. By combining these results with those of the Project’s survey work in the Gomal plain, and the exploration that has been carried out by the University of Peshawar and the Department of Archaeology and Museums, Government of the then NWFP, a clearly defined relative chronological sequence for the late prehistoric/protohistoric occupation in these two regions has been delineated (F. Khan et al. 2000b, 2000c), and reference to this sequence will be made throughout this volume.

Each of the major chronological phases that have been identified is characterized by a distinctive assemblage of cultural material, chiefly the ceramic material and figurine types, as well as a number of very specific changes in production technology and distribution strategies (Khan et al. 2000e, 2004; Petrie et al. 2007). There are also examples of transitional deposits at several sites, which suggests that in some instances, the transitions between specific assemblages of material culture was taking place in situ in the Bannu basin and/or the Gomal plain (e.g. Chapter 10.3).

Leaving aside the potentially aceramic deposits at Gumla I, the first clearly defined phase of early village occupation in the Bannu basin and the Gomal plain is represented by the sites of Sheri Khan Tarakai, Girdai, Barrai Khuarra I, Ter Kala Dheri, Jhandi Babar A and Gulgai Kot. When they are discussed as a group, these sites will be referred to here as belonging to the Sheri Khan Tarakai Phase, in honour of the site at which the cultural assemblage was first identified. Although it largely falls outside this volume, the next major phase of occupation in these two regions has been identified at Gumla II, Rehman Dheri I, Lak Largai, Lewan, Maru I, Jhandi Babar A and B, and Ghandi Umar Khan, and has been named the Tochi-Gomal Phase (F. Khan et al. 2000b, 2000c). Some evidence for what appears to be transitional occupation deposits that straddle the transition between the Sheri Khan Tarakai and Tochi-Gomal Phases has been recovered at Ter Kala Dheri, Lewan and also possibly Sheri Khan Tarakai, and will be referred to as being transitional where appropriate. Lastly, and for the sake of completeness, it is important to mention that the Tochi-Gomal Phase is followed by another transitional phase, which appears to be an intermediate phase before the subsequent Kot Diji Phase, and has been identified at Ghandi Umar Khan, and also possibly at Lewan, Rehman Dheri and Gumla. The Kot Diji Phase in the Bannu basin partially correlates with the Kot Diji Phase on the Indus plain, but also appears to be at least partly contemporaneous with the Mature Harappan Phase elsewhere.

Chapter 1

Early village life in the north-western borderlands

Farid Khan, Robert Knox, Ken Thomas, Cameron Petrie & Justin Morris

This volume examines the earliest village settlements in the Bannu and Gomal regions of Pakistan (Fig. I). These archaeological sites make a critical contribution to our understanding of the nature of early settled life in the borderlands zone that lies at the western edge of South Asia, and they are also essential for understanding the relationships between early settled populations in South, Central and greater West Asia. This chapter will introduce the research context and critical research questions related to our understanding of the earliest villages in South Asia, and will also outline the goals and approach of the Bannu Archaeological Project. The chapters that follow will review the history of research in this field, and draw on the excavations at the site of Sheri Khan Tarakai, and several other sites in the Bannu basin and the Gomal plain, to investigate the earliest village settlements in the borderlands zone as a whole. Throughout, evidence of architecture, ceramic and lithic production technologies, small finds made from local and non-local raw material, and plant and animals remains is used to provide insight into the socio-economics, distribution and environmental context of these early settlements, and also the interaction between their inhabitants and those of neighbouring regions.

1.1. Introduction and research context

From the beginnings of interest in the late prehistoric/ protohistoric archaeology of South Asia during the early twentieth-century, there has been a desire to find connections between the early civilisations in the subcontinent and the regions of Western and Central Asia. This aspiration focused attention on sites that lie in the borderland regions at the western edge of the Indus plain, with a specific goal, in the first instance, of establishing the nature of the links between the Indus Civilization and the early civilizations in both Iran and Mesopotamia (see Chapter 2.1). By the late twentieth century, the rationale for investigating such sites and areas had progressed, and there was a specific focus on identifying the earliest settled occupation in the various regions of the subcontinent. However, the desire to understand the broader archaeological and cultural context in which the late prehistoric/ protohistoric populations of South Asia lived has remained. Fundamental questions with regard to the chronology of these sites and the nature of the interactions between their inhabitants persist, and many of these have not received due attention.

In South Asia, as in many other regions, the appearance of the earliest villages marks the critical point of transition between an essentially mobile lifestyle, relying on the hunting of wild animals and the gathering of wild plants, and a primarily sedentary lifestyle, which was increasingly reliant on the cultivation of domesticated plants and the exploitation of domesticated animals. The shift to sedentism and agro-pastoralism is a key socio-economic and cultural transformation that is important for understanding developing complexity in South Asia. These transformations resulted from a range of decisions that have a range of potential causes, including factors such as the need for managing risk (Thomas 2003; Hayden 2009; Winterhalder and Kennett 2009).

Prior to the discovery and excavation of the site of Mehrgarh, our understanding of the earliest village settlements in South Asia was primarily informed by relatively small-scale and/or unsystematic excavations at a range of sites (see Chapter 2.1). Most investigation of the earliest settlements had also focused on the borderland zone in the west of South Asia, particularly the hills and valleys of southern and northern Baluchistan. A great deal of discussion about the earliest sites in these areas has envisaged them as stepping stones to the ultimate appearance of the urbanized Indus Civilisation during the third millennium BC, and this approach has seen the earliest sites in Baluchistan placed at the beginning of a simple linear developmental trajectory that culminated with the formation of the Indus cities. There is, however, a range of evidence which suggests that straightforward reconstructions cannot account for the variation in behaviour seen in different regions, and this is true from the periods of early village occupation onward.

It is inevitable that new research will force scholars to reassess and re-evaluate standing explanatory paradigms. For instance, it was only in the closing decades of the twentieth century that the importance of the archaeology of the southern parts of the then NWFP was recognized, thanks largely to the work of the Bannu Archaeological Project and the Department of Archaeology at the University of Peshawar (see Chapter 2.1). As this volume will show, part of the significance of these regions lies in the fact that they do not conform neatly to linear models of cultural development. For instance, while a degree of regional variation in material culture throughout South Asia during the early periods has been noted previously, variation in early subsistence behaviour is now also evident. Excavations at several sites in the hills and plains of Baluchistan have provided clear evidence that the earliest phases of settled village occupation in the subcontinent relied on an agro-pastoral economy based on wheat, barley, sheep, goat and cattle exploitation (see Chapter 2.1 & 2.4). Research in the Ganges Valley, the Vindhya Hills and South India has, however, shown that the agropastoral economy of early villages in different parts of South Asia was variable. Thus, while the wheat, barley, sheep, goat and cattle based agro-pastoral economy dominated early villages in the west, early villages in other areas engaged in the exploitation of a range of rice, millet and pulse crops that appear to have been domesticated locally (summarized in Fuller 2006, 2008b; Chakrabarti 2009).

This volume is primarily concerned with the early village settlements in a very specific part of the borderlands zone at the western edge of the subcontinent. In the Bannu and Gomal regions in particular, but also in the borderlands as a whole, it is clear that the subsistence practices of early settled populations were dominated by wheat, barley, sheep, goat and cattle based agro-pastoralism. However, in seeking to understand these settlements and to establish their broader context and significance, it is important to acknowledge that throughout late prehistory/protohistory in South Asia, there were variable trajectories of socio-cultural development, and that these trajectories intersected in different regions and at different times.

1.2. Early villages in South Asia: research questions and parameters

Gregory Possehl (2002: 30) and Dilip Chakrabarti (2006: 104ff.), have both noted that at present, the earliest food-producing peoples in the western borderlands of the greater Indus region are not well documented. It could be argued that clearly defined cultural and contextual information is the essential factor for understanding the dynamics that were in operation during the initial phases of early village life in South Asia. However, while a significant number of sites are known, few of these have been investigated systematically, and there are also gaps in our knowledge about the sites that have been excavated (see Chapters 2 & 11). In Baluchistan in particular, but also more broadly across the whole of South Asia, the details of the processes of transition from mobile hunting and gathering to sedentary agro-pastoralism are as yet unclear. This situation is largely due to a lack of detailed research at sites used by mobile populations. As a result, most attention has been focussed on identifying and describing the earliest village sites in various areas.

EARLY VILLAGE SUBSISTENCE: ORIGINS AND DISPERSALS

Mehrgarh stands as the earliest known village settlement in South Asia, and there has been considerable discussion of the nature of socioeconomic development throughout the early sequence at the site, which provides the earliest evidence of a wheat, barley, sheep, goat and cattle based agro-pastoralist subsistence system in the subcontinent (e.g. J.-F. Jarrige et al. 1995; J.-F. Jarrige 2008). With some exceptions (see Chapter 11.3), our understanding of the trajectory of social and economic development at Mehrgarh is clear. There have, however, been relatively few attempts to explain how this agro-pastoralist subsistence strategy dispersed across the hills and plains of South Asia. Those attempts that have been made have tended to focus on putting sites into large scale groups (e.g. P. Singh 2002) or ‘big-picture’ dynamics (e.g. Fuller 2006: 27ff., Figs 10, 13), and have often drawn correlations with the dispersal of language (e.g. Bellwood and Renfrew 2002; Bellwood 2009; Fuller 2003a, 2008c, 2009; reviewed by Chakrabarti 2008). Assessments that have incorporated specific evidence for material and biological remains and dates from individual sites (e.g. Possehl 1999: 443ff.; 2002: 30–40) have tended to concentrate on increases in settled areas and geographical expansion, but have avoided defining the social dynamics that might account for this growth and the concomitant spread of people and/or practices. Meadow has argued (1996: 407) that we should expect the relationships between peoples to be dynamic and fluctuating, and it is also essential to expect that relationships are likely to have changed over time.

There is considerable scope for exploring why, how and when the wheat, barley, sheep, goat and cattle based agro-pastoralist subsistence strategy dispersed across the subcontinent. For instance, there is some evidence for variation in the chronology and trajectory of this dispersal. It would appear that the earliest occupation levels for sites on the plains of the Punjab are considerably later than those from sites in central Baluchistan, such as Mehrgarh and Kili Gul Mohammad (see Shaffer 1992; Possehl 1999). They also appear to be somewhat later than the early levels at the known sites in northern Baluchistan, such as Rana Ghundai, Sur Jangal and Periano Ghundai. The available dates therefore appear to indicate that wheat, barley, sheep, goat and cattle-based agro-pastoralism dispersed through the hills and onto the plains in a particular pattern, but the way in which this happened is not clear. If we are observing the diffusion of practices, how was this manifested? Are we witnessing demic or cultural processes, or combinations of the two? In general, we lack the evidence to assess these questions properly at any of the sites that have been investigated, and explanatory models and theoretical discussions have tended to gloss over local-scale dynamics (e.g. Thapar 1978; Liversage 1989; Saraswat 1992; Agrawal 2002; Fuller 2006). If local-scale dynamics are not considered, however, our understanding of the past is inevitably simplified.

Although wheat, barley, sheep, goat and cattle-based agro-pastoralism has been discussed here as though it were a unified entity, this does not appear to have been a consistent ‘package’ of crops and animals that was adopted widely in South Asia (Fuller 2006). Rather, it appears that across the subcontinent, some populations adopted a wheat, barley, sheep, goat and cattle-based agro-pastoralist subsistence strategy, whereas in certain areas, only some of these domesticated plant and animal species were adopted at any one time, while others were not. It is beyond the bounds of this volume to discuss this process across the subcontinent as a whole, but there is considerable scope for investigating why this might have transpired, and attempts to understand local-scale dynamics in the dispersal of practices will make an important contribution towards our broader comprehension of this process.

THE ROLE OF CLIMATE AND ENVIRONMENT

Evidence for the nature of palaeoclimates continues to be considered in attempting to understand the critical periods of early socio-economic complexity (e.g. Kajale and Deotare 1997; Enzel et al. 1999; Possehl 1999; Staubwasser et al. 2003; Madella and Fuller 2006; Fuller 2006; Agrawal 2007). However, long-term palaeoclimate proxy evidence for western South Asia primarily comes from lakes in the Thar Desert, and the relatively limited evidence for local-scale variation in palaeoclimate throughout the borderlands zone that marks the western edge of the subcontinent largely comes from analysis of samples from archaeological sites (Possehl 1999: 257–268). The specific ecological and geographical contexts of the early village sites in South Asia have the potential to provide key insight into the interaction between settled populations and their local ecosystems. However, the specific environmental and landscape contexts of early sites have received far less attention than they deserve, and a clear understanding of these contexts in most regions is wanting. Meadow (1996: 395) and Thomas (2003: 414) have both noted that early farmers in western South Asia did not employ major irrigation works, rather they used the margins of back-swamps or oxbow lakes for fields, and also made use of areas that were periodically inundated, where they could trap floodwater or run-off, together with its accompanying silts. These specific landscape contexts are clearly visible in satellite images (see Chapters 2.4 Figs 2.4–2.8 & 11.2), but it is important to determine whether early farmers selected particular locations for establishing their settlements.

TECHNOLOGY AND STYLE

During the phase of early village development in the subcontinent, but also more broadly in Western Asia, there were a range of important technological innovations in various craft practices. These are most notable in ceramic and lithic production, and also evident with developments in metallurgical technology. While it is no doubt significant to seek the locus of specific innovations, undertanding the way that they were dispersed and shared between populations is also important. Research on the early ceramic vessels from Mehrgarh has shown that the approaches used there were similar to those used at contemporaneous sites in various parts of Western Asia (Vandiver 1986, 1987, 1995). Similar approaches to production are also evident in various areas in the west of the subcontinent, but as yet there have only been limited attempts to conceptualise how these technologies might have been shared across space and time. The investigation of these processes is, however, likely to provide a specific opportunity to characterise the nature of the relationships between populations living in different regions. Similarly, early approaches to defining the significance of regionally distinct styles of artefacts, particularly ceramic decoration, have been relatively simplistic (e.g. Piggott 1950). While there are clear differences in styles and approach evident at early village sites in the different areas of western South Asia (Kenoyer and Meadow 2000; Petrie et al. 2007), what these differences might represent and the social roles that they might have played are worthy of focused consideration.

PERCEPTION AND BELIEF

A considerable amount of research on the early villages of Western Asia has focused on the behaviour of inhabitants, the perception of their surroundings and their belief systems (e.g. Kuijt 2000; Byrd 2005). There is considerable potential for similar investigations in the subcontinent, particularly in the interpretation of burial practices, the details of decorative styles and the similarities and differences in the distribution of specific categories of material culture. In the latter case, the wide range of figurines that are known from several early village sites have considerable potential.

NETWORKS OF INTERACTION

The routes and means by which ancient South Asian populations interacted with each other are also significant, but are not often considered in detail. The hills and valleys of the Sulaiman Mountains and Baluchistan mark the western borderlands of the subcontinent, but they are an extremely permeable frontier. Hundreds of routes and passes traverse these ranges and many of these link the main valley systems of north and south Baluchistan and the adjacent piedmont and plain areas. Certain routes were undoubtedly favoured in the past, but it is likely that more than one route was used to travel between any two locations. For instance, Loralai, where the site of Rana Ghundai is located, is linked by passes to the southern Gomal plain, while Zhob, where the site of Periano Ghundai is located, is situated in a different valley system that is linked more directly with the Gomal River itself and the north of the Gomal plain. There are a range of possible routes between these two sites alone, let alone any of the other sites whose populations might have been interacting with the inhabitants of these settlements, and in general, there has been limited discussion of how these routes were used. If nothing else, the routes and passes that link northern Baluchistan to the Bannu and Gomal regions were almost certainly being used by pastoralist groups, and they may well have been one of the conduits for the sharing of wheat, barley, sheep, goat and cattle-based agro-pastoral and other practices between the highlands of Baluchistan to the piedmont zones in the Bannu basin, the Gomal plain and beyond (see Chapter 11.3).

RELATIVE AND ABSOLUTE CHRONOLOGIES

One of the major impediments to a clear understanding of early villages throughout the subcontinent is a lack of systematically collected absolute dates. The primary problem has been that large numbers of radiocarbon determinations have been obtained for only a small number of sites, and many of those dates are problematic (see Chapters 2.4, 9 & 11.3). Therefore, for almost all of the known early village sites, a heavy reliance on parallels between elements of material culture to provide relative dates is essential. In most cases, these correlations rely on comparisons between stratified ceramic assemblages and the individual ceramic wares that have been identified. This approach is not without problems, as often the distances over which some parallels are drawn are considerable, similarities tend to be overemphasised, and contemporaneity is assumed rather than demonstrated.

1.3. Goals and approach of the Bannu Archaeological Project

The Bannu Archaeological Project was initiated with a view to reconstructing the history and context of human occupation in the Bannu region and surrounding areas (see Preface). From the outset, it was decided that much of the Project’s attention would be focused on understanding the earliest village settlements of this region, with the clear goal of gaining insight into the social and economic context of these sites. This aim would be achieved by trying to understand both the behaviour of the ancient inhabitants and the nature of their interaction with peoples elsewhere. The decision to focus on the early periods was vindicated by the discovery of the site of Sheri Khan Tarakai in 1985.

The Bannu basin and the Gomal plain both lie in a specific part of the borderlands zone that is in many ways peripheral to Baluchistan and the Indus plains, which are typically perceived as being the core areas of socio-economic development in western South Asia. Nonetheless, these ‘peripheral’ regions are critical to understanding the broader developments that link the early sites in the northwest of the subcontinent during this important phase of social, economic and cultural development.

Given that a large amount of archaeological discourse worldwide focuses on elites, high-status objects, grand urban centres, technological developments, cultic installations, major monuments and the like, there is an inherent tendency for certain types of archaeological data to fall through the proverbial cracks. John Robb (2007: 2) has written of the desire to act in a sense of loyalty to the ordinary past, as in his words ‘throughout human history, most people have not been the scheming political elites, profoundly religious megalith users, or the categories of actors who populate the pages of archaeological theory. If we do not theorise about ordinary people, if we assume that they are mere bricks in the fabric of society, we leave the great bulk of our subject uninvestigated’.

In many respects, the inhabitants of the late prehistoric/protohistoric sites investigated by the Bannu Archaeological Project were ordinary people. Sheri Khan Tarakai, for example, is neither the earliest and nor is it the largest early village settlement site in South Asia. It has no evidence for metal use and it lacks evidence for monumentality. Nevertheless, it is a site of immense importance for understanding early village societies in South Asia generally, and in the northwest in particular. If such sites are not investigated and ultimately situated in their broader context, then we are left with a manifestly skewed perception of the development of prehistoric cultures in the subcontinent as a whole.

Much attention has been given to the early Neolithic occupation at the site of Mehrgarh and the long-term developmental sequence attested in Baluchistan. While this focus is unquestionably deserved, it has meant that there has been a tendency to overlook what took place in other nearby regions. As noted above, we have a relatively limited understanding of how wheat, barley, sheep, goat and cattle-based agro-pastoralism developed and dispersed throughout the western parts of South Asia at a local-scale, and we also have a fairly simplistic awareness of how