Case Studies in Archaeology and World Religion: The Proceedings of the Cambridge Conference 9780860549567, 9781407350783

Citation preview

BAR S755 1999

INSOLL (Ed)

CASE STUDIES IN ARCHAEOLOGY AND WORLD RELIGION

B A R

Case Studies in Archaeology and World Religion The Proceedings of the Cambridge Conference

Edited by

Timothy Insoll

BAR International Series 755 1999

Published in 2016 by BAR Publishing, Oxford

BAR International Series 755 Case Studies in Archaeology and World Religion

© The editor and contributors severally and the Publisher 1999 The authors' moral rights under the 1988 UK Copyright, Designs and Patents Act are hereby expressly asserted. All rights reserved. No part of this work may be copied, reproduced, stored, sold, distributed, scanned, saved in any form of digital format or transmitted in any form digitally, without the written permission of the Publisher.

ISBN 9780860549567 paperback ISBN 9781407350783 e-format DOI https://doi.org/10.30861/9780860549567 A catalogue record for this book is available from the British Library BAR Publishing is the trading name of British Archaeological Reports (Oxford) Ltd. British Archaeological Reports was first incorporated in 197 4 to publish the BAR Series, International and British. In 1992 Hadrian Books Ltd became part of the BAR group. This volume was originally published by Archaeopress in conjunction with British Archaeological Reports (Oxford) Ltd/ Hadrian Books Ltd, the Series principal publisher, in 1999. This present volume is published by BAR Publishing, 2016.

BAR

PUBLISHING BAR titles are available from:

EMAIL

PHONE FAX

BAR Publishing 122 Banbury Rd, Oxford, OX2 7BP, UK [email protected] +44 (0)1865 310431 +44 (0)1865 316916 www.barpublishing.com

This book is dedicated to Freya

TABLE OF CONTENTS

Foreword and Acknowledgements Introduction. Research Foci in Archaeology and World Religions.

V

Timothy Insoll

1

Introduction The Papers Sacred Space Theoretical Issues Iconography Aspects of a Religious Whole The Development and Understanding of World Religions The Papers in Context

Buddhist landscapes and monastic planning: the elements of intervisibility, surveillance and the protection of relics.

1 1 2 2 2

3

Julia Shaw

6 7 8 10

13 14 14

Emma Loosley

John Binns

27

29 31

Nayanjot Lahiri

Introduction Preliminary Formulations: Patronage, Proprietorship and Colonial Intentions A Mediating Viceroy and his Difficult Customer: Curzon and Khrishna Dayal Gir Conclusion

Ayodhya and the Ethics of Archaeology. Introduction Types of reconstruction

26 26

Introduction. The Great Laura Archaeology of the Desert Communities 'The Desert a City' Conclusions

Bodh-Gaya: an ancient Buddhist Shrine and its Modem History ( 1891-1904 ).

18 18 18 19 20 24

Introduction N.W. Syria and its churches Church-building and Liturgy The Berna, Architecture, Archaeology and function Conclusions

The Concept of Sacred Space in the Monasteries of Byzantine Palestine.

5 5

Introduction Sacred Landscape and archaeological theory The Stupa Inter-site spatial relationships Intra-site spatial relationships Serpent-guardians 'Sacred landscapes': Discussion Conclusion

The Early Syriac Liturgical Drama and its Architectural Setting.

1

Nandini Rao

33 33

33 40 42

44

44 44

Reconstructions of the National Movement Post colonial reconstructions The Creation of a Hindu identity and a Hindu India The Ayodhya Controversy Use of history, archaeology, emotion and religion Reclaiming a sacred space Conclusion

45 45 45

46 46 47 47

Archaeology, dance and religion in Java: the Prambanan complex.

Alessandra Iyer

Introduction Point of departure: the living Javanese dance culture Knowledge of dance in ancient Java: archaeological evidence from the Prambanan temple complex Analysis of the reliefs Prambanan: purpose of the sequence and the politics of South and Southeast Asian interaction Conclusions

A Symbol of the Deity: Artistic Rendition of the 'Hand of God' in Ancient Jewish and early Christian Art. Introduction The Hand of The Hand of The Hand of The Hand of Conclusions

Rachel Hachlili

48 48 48 48 52 55 57

59 59 59

God God in ancient Near Eastern pagan art God image in Jewish art God image in Early Christian art

60 61

66 68

Text, Tradition and Transformation: Mandan 's 'Devata-Murti-Prakarnam '.

RimaHooja

Introduction Iconography, Archaeology and Textual Traditions The Iconography of Religion Mandan's Devata-Murti-Prakamam: Elucidating a Tradition Interpreting a Canon: DMP and the Correlation of texts, Iconography, Archaeology and Religion Transcreation, Syncretism and Changing Realities Conclusions

The Archaeological Visibility of Caste: An Introduction.

Robin Coningham & Ruth Young

71 71 72

75 77 81 82

84 84 84 85

Introduction Definitions Caste in contemporary south Asian communities Caste and the past The archaeological visibility of caste: towards a methodology Occupation Diet Conclusions

Christianity and Islam in the Middle Nile: Towards a study of religion and social change in the long term.

71

87 88 88

90 92

David Edwards

94 94 96

Religious identities Historical background Archaeological approaches Archaeological ceramics in social contexts

97 98

ii

102

Conclusions

Tadmakkat and the Image of Mecca: Epigraphic Records of the Work of the Imagination in I J1hCentury West Africa.

P. F. de Moraes Farias

Introduction Epigraphic Evidence for the identification of :lssuk as Tadmakka (Tadmakkat) The need to perceive towns marginal to the Islamic world as images of Mecca, and to gather local life into Islamic chronology Textual analysis oflnscription :lssuk N° 104 Textual analysis of Inscription :lssuk N° 105 Textual analysis oflnscription :lssuk N° 106 Conclusions: the broader scope of West African epigraphic studies

The Archaeology of Muslim Pilgrimage and Shrines in Palestine.

Andrew Petersen

Introduction Muslim Pilgrimage in Palestine Archaeological Evidence for local Shrines in Palestine Five Case Studies General Conclusions

105 106 107 107 112 112 114

116 116 116 119 121 126

Simon Coleman & Jas' Elsner

Archaeology and Christian Sacred Space at Walsingham

105

128 128 129

Introduction A Tale of Two Traditions

The Medieval Jewish Quarter of Regensburg and its Synagogue: Archaeological Research 1995-1997.

Silvia Codreanu-Windauer

Introduction The Jewish community Archaeological Research 1995-1997. The Romanesque Synagogue The Gothic Synagogue Conclusions

139 139 141 142 146 151

Archaeology, Nonconformist Missions and the 'Colonisation of Consciousness' in Southern Africa, c.1820-1900. Paul Lane Introduction Christianity and Missionary Endeavour in Southern Africa Archaeological Remains of 19th Century Missions Kolobeng (1847-1852) Ntsweng c.1865-1937 Phalatswe (1889-1902) The L.M.S. Mission Station The 'Native' Church Other Christian Elements Summary and Conclusions

The Mahabharata, Archaeological and Literary Evidence.

153 154 155 156 158 158 159 161 161 161

Sandhya Chatterjee Chakrabarti

iii

153

166

166 166 167 170 173

Introduction Origin of the Mahabharata Archaeological Evidence Literary Evidence Summary and Conclusions

Chronos, Saturn, Mithra: Archaeology and the Pagan Origins of Christmas.

Kate Prendergast

175 175 176 178 180 180

Introduction Chronos and the Kronia: the Mythic and Archaeological Evidence Saturn and the Saturnalia: Archaeological and Documentary Evidence Mithra and Mithraism: Archaeological and Documentary Evidence The Development of Christmas Conclusion

Archaeology and Library Work in the Study of Early Christianity.

175

Prof. W.H.C. Frend

182 182 182 183 183 184 185 186 186

Introduction Christianity Arians Gnostics and Manichaeans Donatists Nubia and Roman Britain Italy and Beyond Conclusions

iv

Foreword and Acknowledgements These papers result from a conference entitled, 'The Cambridge Conference on Archaeology and World Religion: The Examples of Judaism, Islam, Christianity, Hinduism, and Buddhism', which was held in St. John's College, Cambridge between the 14th-16th April 1998. Thanks are offered to St. John's College for hosting the Conference, whilst the editor and conference organiser was in receipt of a Research Fellowship there. Gratitude is also extended to the Master and Fellows of St John's College for financially sponsoring the attendance of some of the overseas participants; without this support

their participation would have been impossible. Equally, the contribution of the four anonymous referees who read and commented on all or part of the papers prior to their publication must be acknowledged, and they are gratefully thanked for their hard work. Finally, sincere thanks are offered for the admirable help rendered by the Conference Assistants, Salma Akhter, Mark Brand, Michael Frachetti, Robert Harding and Emma Read, and thanks also to David Edwards for help in preparing the camera-ready copy.

V

vi

Introduction. Research Foci in Archaeology and World Religions. Timothy Insoll, Department of Art History and Archaeology, University of Manchester. Introduction

This volume is in many respects difficult to introduce as it contains such a diversity of material. Having said this, a common focus is provided by the fact that each paper presents a detailed case study which utilises archaeological research, or the results thereof, as a means of furthering our understanding of World Religions, and more specifically (in no particular order), with regard to Judaism, Buddhism, Christianity, Islam, and Hinduism. The 18 papers in this volume can be classified under five broad headings; sacred space; theoretical issues; iconography; aspects of a religious whole; and, archaeology and the development and understanding of World Religions. These reflect the original sessions the papers were presented in at a conference on archaeology and World Religions held in St John's College, Cambridge, between the 14th-16th April 1998.

The Papers Sacred Space

The influence of World Religions upon spatial use goes far beyond merely dictating the nature of places of prayer. Public and private buildings, landscapes, gardens, and city or town plans can all be influenced by religious or sacred considerations. Similarly, continuity in sacred spatial use and construction is often apparent and this issue needs to be explored with an emphasis upon how this might be approachable through archaeology. World Religions can overlap, temporally and spatially, the legacy of older traditions can be utilised for various purposes, issues of obvious relevance to the archaeologist, and often manifest in the use, re-use or re-structuring of sacred space. Such notions are considered by Julia Shaw in her examination of five Buddhist stupas in the central Indian state of Madhya Pradesh. Developments in the general archaeological study of sacred landscapes appear to have had little impact upon the archaeology of Buddhism in India thus far. Importantly, Shaw illustrates that these structures are of much greater significance than mere architectural monuments, in their function as containers for Buddhist relics. They appear to indicate that intersite spatial relationships were important in creating a sacred landscape. Current archaeological theory is here successfully applied to new material. In contrast, Emma Loosley's study appears at first glance to be more particularist, being concerned with details of Early Syriac Christian liturgy. Yet she illustrates that this was by no means fixed and has in fact evolved and changed over the centuries. This is related to material culture and is connected to the architectural form of the church, 'the church as microcosm', with architecture, visual imagery and liturgical furniture used to signal early ritual, which

she reconstructs, and thereby shows how the liturgy developed from one of participation to the more static form evident today. Texts are not allowed to dominate and material culture is 'vocalised'. Similarly, John Binns describes how architecture and monastic foundations themselves could be used to 'civilise' the desert in Byzantine Palestine. Here, monasteries are shown to be integral parts of their environment and rooted in society, with a sacred/secular division largely proving an invalid concept. Theoretical Issues

The second group of papers to be briefly considered are those concerned explicitly with theoretical issues as they relate to archaeology and World Religion. The archaeological study of World Religions cannot be thought of as being divorced from theoretical topics, as the very study of peoples' extant beliefs and of living faiths in itself raises a number of questions, both morally and practically. Primary amongst these are the relationship between religious texts and archaeology. Specifically, can religious texts be treated as historical documents?, or do they merit special treatment?, and hence do the usual strictures regarding the relationship between history and archaeology apply? Does archaeology with its emphasis on material culture dispel notions of the ideal/divine contained in religious texts, through the possible conflict between the two sources of evidence and the resulting differences in interpretation of events? Secondly, the issue of how politics and religious sensitivities can affect the research we undertake should also be considered. To what extent can archaeology be abused for religious propaganda? Can this be avoided or is it inevitable? Is it necessary to be an adherent of the faith the archaeologist is studying? or does this complicate issues? Can a secular 'Religious Archaeology' be said to be developing, especially in certain areas of the Western World, where religion is abstracted or absent from everyday life? Is it morally justifiable to apply 'scientific' procedures such as radiocarbon dating to the study of World Religions and hence people's extant faith and beliefs? All these issues are of relevance to the archaeological study of World Religion. However it should be noted that papers containing a more overt theoretical content, or considering these specific issues are under-represented in this volume as a separate collection focusing upon these will appear (lnsoll in preparation). Nonetheless, almost without exception all the papers presented here illustrate that theoretical concerns cannot be divorced from the individual case studies under discussion. Two papers stand out because of their focus on case studies which emphasise aspects of the theoretical relationship between archaeology and World Religion. Nayanjot Lahiri cogently illustrates how the history and archaeology of one Buddhist shrine, the Bodh Gaya temple complex in

Bihar, India, has been highly politicised since its 'rediscovery' in the nineteenth century, and how it has been used by various competing groups for political ends. Lahiri's study emphasises how the past can be used in the present. These are themes which Nandini Rao also discusses with reference to a more recent controversy, that surrounding the Bahri Masjid/Ram Jammabhumi in Ayodhya. Rao presents an unbiased picture of how Hindu and Muslim politics collided and how archaeology and religion were used and abused in late twentieth century India.

remit of our usual preoccupations, such as economic and political structures. In fact many, if not all areas of life can be influenced by religion - diet and cuisine, dress, pilgrimage, social structures, personal possessions, warfare. To this can be added the influence of World Religions upon death and burial - conceptions, treatment, disposal, commemoration. In sum, there is a possible religious whole, many components of which are approachable by the archaeologist. The fourth group of papers considers how aspects of this religious whole may or may not be recognised by archaeologists.

Iconography

Robin Coningham and Ruth Young attempt an archaeological examination of the visibility of caste, through utilising, predominantly, evidence from the citadel of Anuradhapura in Sri Lanka, a centre of Theravada Buddhism. A number of important conclusions are reached indicating the complexity of the subject and the futility of uniform archaeological assumptions concerning this issue. Likewise in assessing diet, David Edwards, although considering the overall context for the transition from Christian to Muslim Nubia, indicates that a pot is more than a receptacle for storage or serving, but provides a means to reconstruct cultural continuity in the face of religious change through archaeological evidence. Diet, or more specifically the consumption of alcoholic beverages, is shown to have had a social significance beyond the immediately apparent.

Research in archaeology and World Religion is obviously also concerned with the specifics involved in the recognition and development of religious traditions and elements thereof, and the latter three groups of papers are concerned, albeit loosely, with such issues. The third section of papers are those which consider iconography. Art and iconography are a medium which above all are used to express religious belief. Iconography in all its many forms can be used to create, perpetuate and signal religious identity. Within the context of World Religions the use of iconographic codes is far-ranging in purpose and myriad in diversity, from rendering the basics of sacred texts accessible to the illiterate to the intricate complexities of calligraphy perhaps aimed at a select few. How stone, paper, clay, wood can be used for religious iconography ('art', symbolism, epigraphy) is open to discussion, but emphasis needs to be placed upon the 'whys' and 'wherefores', rather than upon typology and reportage.

Also in Africa, Paulo de Moraes Farias is concerned with a further aspect of the 'religious whole' visible though material culture, the creation of religious 'belonging' or identity through, for example, topographic imagery, and its recording by Muslim epigraphy in medieval West Africa. Farias indicates that epigraphic evidence has been under-utilised for interpretative purposes and that it can in fact contain much information of relevance for the study of Islam above and beyond that usually considered, i.e. names and dates. Andrew Petersen looks at pilgrimage, another facet of religious life, with reference to several Muslim shrines in Palestine. He indicates that a hierarchy of pilgrimage exists, that this is in part approachable through material culture, and that the pre-Islamic heritage must be acknowledged. Pilgrimage is also a subject considered by Simon Coleman and Jas' Elsner. The focus of their study is the shrine of Walsingham in Norfolk, England, important to both Catholics and Protestants alike. The history of the shrine and its archaeological investigation are examined especially with regard to the relationship between archaeological research and the various appropriations of the shrine, or rather the two shrines which exist today. Again, how the past can function in the present, and how it can be misinterpreted, is apparent.

Such issues are examined by Alessandra Iyer in her study of the archaeology of Hindu dance in Java. This is achieved through focusing on the temple complex of Prambanam. Iconography, the dance images carved on the temple complex, are utilised to reconstruct specific dance moves which in tum provide information on religious and cultural interaction between South and Southeast Asia over time. Rachel Hachlili likewise considers a specific body of material, depictions of the 'Hand of God' in Ancient Jewish and Early Christian art. Rather than merely cataloguing such images, Hachlili illustrates that this iconographical information indicates both continuity and dissimilarity in this artistic tradition as carried between these two major World Religions. The final 'iconographic' paper returns to Indian material, but Rima Hooja approaches her primary evidence, a text from medieval Rajasthan which discusses Hindu iconographical traditions, the DevataMurti-Prakamam, with a broad brush. The DMP is placed firmly within the context of the archaeology of largely Hindu religious evidence in the region, and important points concerning the material recognition of transient religious practices and changing religious 'tastes' raised.

The Development and Understanding of World Religions

Aspects of a Religious Whole

The final broad grouping of papers is concerned with the contribution archaeology can make to our understanding of the development of specific World Religions or elements thereof. Silvia Codreanu-Windauer presents the results of recent archaeological research looking at the medieval Jewish quarter in Regensburg, Germany.

Archaeologists, whilst perhaps acknowledging that the individuals which made up the past communities they study may have been religious, represented, say, by a mosque, temple, synagogue, or church, often separate out the other spheres of life, and study these under the

2

Good preservation of the structural remains along with accompanying historical sources has allowed a detailed reconstruction of both the residential area and the synagogue. The latter is of especial importance as rare evidence for early synagogue architecture in Germany. Equally, the demise of the community is also testified. Paul Lane considers the impact of the mission stations established in various regions of Southern Africa in the nineteenth century upon the indigenous population. This is in part achieved through examining the material remains of these missions. However, he utilises both the material and historical evidence to assess the concepts of religious syncretism, religious 'indigenization' and religious authenticity in this region.

seemingly disparate range of papers linked only by a varying religious content? This is a valid query. The utility of both the conference and the proceedings, in the main, lay in the fact that it brought together people from different disciplines and from different archaeological 'traditions'. Anthropology, History of Art, History, Religious Studies, Architecture, scholars from all these disciplines were represented, in addition to archaeologists, who are usually labelled as Islamicists, Indianists, Africanists, Medievalists, Prehistorians etc. In other words, groups who go to their own conferences emphasising their own research specialisations were brought together. The inter-disciplinary and interregional emphasis was significant, and as these papers illustrate, although they might be concerned with very different regions and subject areas, it was surprising how many similarities exist, both in problems encountered, categories of material analysed, and research methodologies employed. The conference and this volume brought the cumulative efforts of all these diverse scholarly traditions together in one forum, material may in certain cases have appeared in altered versions elsewhere, but is frequently largely inaccessible to all but the dedicated researcher who steps out of his or her own specialisation. Equally, many of these papers have benefited, by their authors own admissions, from the new perspectives gained from the conference which sought to emphasise the archaeology of World Religion as a whole.

The last of the papers which focuses upon Indian material is provided by Sandhya Chakrabarti, and she is concerned with the development of World Religion far removed in both space and time from Lane's study, with her research concentrating upon evaluating the archaeological and literary evidence for the great Hindu epic, the Mahabharata. Through a detailed analysis of both types of evidence Chakrabarti evaluates the historicity of the epic and its heroes, thus indicating certain factual corroborations. Kate Prendergast similarly focuses upon religious traditions in antiquity, but through material removed one step further from the World Religions under consideration, she evaluates the pagan origins of Christmas by examining its nonChristian roots in pagan mythic mystery cycles, that of Chronos, Saturn, and Mithra. Continuity in religious practice is explored through various sources of evidence including archaeology. Finally, a broad overview of how archaeology and the interpretation of documentary sources recovered through archaeological research has contributed to the study of Early Christianity is provided by William Frend. As befits someone who began working in North Africa on this subject in the late 1930s, this paper presents a personalised view of developments over the course of the past 60 years. A synthesis of the relevant evidence is provided from a traditional viewpoint and this final paper contrasts with, and complements, many of the other studies presented by younger generations of scholars in this volume.

In this respect, in its multi period, geographic and religious focus, this volume differs from many of the existing studies of archaeology and World Religion which concentrate upon one religious tradition, Buddhism for example (see Barnes 1995), or one specific component of religion and archaeology, churches perhaps (see Addyman and Morris 1976; Platt 1987), or pilgrimage (see Graham-Campbell 1994). The focus of the case studies in many of the papers in this volume reflect the changing patterns of research into the archaeology of World Religions, and indeed, of the archaeology of religion as a whole. A more theoretically aware and questioning approach is evident in comparison with the work of earlier generations of scholars, who used or misused archaeology for various religious ends. In the past archaeology was utilised to prove or disprove events connected with sacred texts or sacred figures, two famous instances being the search for fragments of Noahs's Ark and research surrounding the identification of the Turin Shroud (see for example Bailey 1977; Bortin 1980; Gumbel 1998 for interesting summaries). To these could be added attempts to identify the Biblical Flood (see Keller 1965: 49-51).

The Papers in Context

Although this is not the place to review previous research which has been undertaken looking at archaeology and World Religion, a massive undertaking, it would appear that conferences with a similar theme, and resulting volumes are rare. Symposia discussing the · study of archaeology and religion have been held, notably the 'Sacred and Profane' conference (Garwood et al. 1991), or the various sessions which have been held at WAC over the years, including the one which resulted in the 'Sacred Sites, Sacred Places' volume (Carmichael et al. 1994). However, conferences such as the one from which this volume of proceedings results are infrequent (even this is possibly an exaggeration, as none similar were known to the organisers, though this is hard to believe, hence the proviso, 'infrequent').

Similarly, there is an absence of so-called religious checklists, supposed identification manuals to the archaeology of World Religions which can be uniformly applied regardless of cultural context through correlations of presences and absences (these are briefly and critically discussed by Lane in the introduction to his paper below, and are evaluated in greater detail with regard to their modem application by Insoll (1999). The stresses and strains often evident when archaeology and textual sources are utilised in conjunction with each other to further understanding of the history and development of World Religions are not such a factor in

But it could be asked what is the point of such a conference or indeed the drawing together of such a

3

these studies as perhaps they once were (see Bartlett [1997: 10-14] for a consideration of this subject as regards Biblical archaeology). Perhaps, it can be suggested, a certain degree of maturity has been reached, where it is recognised that both sources of evidence can frequently be used in conjunction with each other to achieve a common end in some of the following papers.

Bartlett, J. 1997. What has Archaeology to do with the Bible - Or Vice Versa? (In), Bartlett, J. (ed.), Archaeology and Biblical Interpretation. pp. 1-19. London: Routledge.

In summary, the utility of inter-disciplinary contact for scholars interested in the archaeology of World Religions was indicated by the conference, and it is hoped in the selection of papers presented here. It is also apparent that this will be only the first of a series of such ventures, and that research into the archaeology of World Religions is a thriving area. However, many problems in research foci remain. The emphasis upon sacred structures and grand architecture persists, for example, to the detriment of research into the religion of everyday, of all the population, of popular religion and practice, and its material correlates. Equally, the sensitivities involved in this subject area must also be remembered, as indeed must the possibilities for misinterpretation or misuse of the results which exist. These case studies in archaeology and World Religion indicate the many dimensions to this vast subject.

Carmichael, D., Hubert, J., Reeves, B., Schanche, A. (eds.), 1994. Sacred Sites, Sacred Places. London: Routledge.

Bertin, V. 1980. Science and the Shroud of Turin. Biblical Archaeologist 43: 109-17.

Garwood, P., Jennings, D., Skeates, R. And Toms, J. (eds.), 1991. Sacred and Profane. Oxford: Oxbow Books. Graham-Campbell, J. (ed.) 1994. Archaeology Pilgrimage. World Archaeology 26 (1).

of

Gumbel, A. 1998. Ten Years on, the Debunked Turin Shroud gets a second coming. Independent on Sunday 1st February. Insoll, T. 1999. The Archaeology of Islam. Oxford: Blackwells. Insoll, T. (ed.), (in preparation). The Archaeology of World Religion.

References.

Keller, W. 1965. The Bible as History. Archaeology Confirms the Book of Books. London: Hodder and Stoughton.

Addyman, P. and Morris, R. 1976. The Archaeological Study of Churches. London: The Council for British Archaeology.

Platt, C. (ed.), 1987. Archaeology of Church. World Archaeology 18 (3).

Bailey, L.R. 1977. Wood from 'Mount Ararat': Noah's Ark? Biblical Archaeologist 40: 137-46. Barnes, G. (ed.), 1995. Buddhist Archaeology. World Archaeology 27 (2).

4

the Christian

Buddhist landscapes and monastic planning: the elements of intervisibility, surveillance and the protection of relics. JuliaShaw, Department of Archaeology, University of Cambridge. (1854). Since then Sanchi has attracted the attention of many scholars of architecture, art-history and epigraphy (Maisey 1892; Marshall and Foucher 1947; Dehejia 1996). However, with the exception of Satdhara, which has recently become the focus of renewed excavation and restoration (Agrawala 1997), the other hill-top sites, whose architectural remains are less well preserved, less elaborately carved, and more difficult to reach, have largely been neglected. This seems to be symptomatic of the tendency in Indian archaeology to focus on the architectural or artistic merits of single (and especially religious) structures, whilst those sites with less obvious monumental or aesthetic appeal have been ignored. Secondly, since these have typically been viewed in isolation from any kind of ritual, social or archaeological setting, it is hardly surprising that, apart from a few recent exceptions (e.g. Lahiri 1996), the issue of sacred landscape has received scant attention in Indian archaeology.

Introduction

The subject matter of this paper is a group of five Buddhist stupa and monastery complexes (Sanchi, Sonari, Morel Khurd, Satdhara and Andher) which were established between the 3rd and 1st centuries BC in the hills around the early historic city of Vidisha, MadhyaPradesh, India (Figure 1). I shall discuss the wider ritual significance of the stupa later on, but at this point, a brief definition may be necessary. According to the Buddhist tradition, the earliest stupas were set up at eight places after the demise of the Buddha, as receptacles for his bodily relics, which were subsequently redistributed into 84,000 stupas following the popularisation of the stupa cult by emperor Asoka (273-36BC). The core of the main stupa at Sanchi, the earliest and most well-known of the central Indian Buddhist sites (Figure 2), dates to the Asokan period, when the dissemination of Buddhism was largely facilitated by the westerly expansion of the Maurya empire from its base in Magadha. It was subsequently enlarged and encased in stone around the mid-2nd century BC, which is the period to which much of the building activity at the neighbouring sites dates (Marshall 1920: 23).

One of the aims of the present paper therefore is to begin the process of setting these sites within their wider archaeological and cultural context, and in particular to show how the connection between them may have been reinforced by the mechanism of intervisibility. Not only did this help to propagate the power of the Buddhist dharma across the landscape, but may also indicate a preoccupation with the protection of Buddhist relics and

All five of these sites were explored and documented in the mid-19th century by Sir Alexander Cunningham

0

I

2

3

4

SMiles

Figure 1. Map showing monastic sites and places mentioned in text (after Cunningham 1854)

5

Figure 2. Sanchi, stupa no. 1 archaeology and anthropology to break it down (Asad 1979; Evans 1985; Carmichael et al. 1994), has often been perpetuated in archaeological circles because of a restricted focus on well-known, archaeologically visible sites, resulting in the false impression of static ritual practices which do not extend into other areas of life (see Sherratt 1996: 149; Barrett 1990; Hodder 1984; Parker-Pearson 1982).

the control of the monuments in which they were deposited. On an intra-site level, I shall try to illustrate how the importance of vision, and in particular, surveillance, may have been manifested in spatial terms within the individual compounds themselves. Sacred landscapes and archaeological theory

In India, these issues have remained largely unproblematised. By accepting the boundary-lines drawn around the central Indian monastic sites by 19th century surveyors (Cunningham 1854), a site's continued trajectory of influence into the surrounding landscape is effectively precluded; and by focusing on those remains which have remained in monumental form, largely through the historical particularities of archaeological preservation, other substantial, albeit fragmentary, Buddhist remains, such as those found within the city of Vidisha itself, have received little attention (ASIAR 1913-14; 1914-15; Cunningham 1880). The result is the perpetuation not only of the well-known sacred:profane polarity, but also one based on unhelpfully rigid religious categories such as 'Buddhist' or 'non-Buddhist'.

This reappraisal of the spatial meaning of the central Indian sites can be usefully contextualised within the framework of recent developments in European archaeological theory and method. Of particular relevance is the affirmation that ritual sites should be situated within their broader socio-political setting; the following statement made with reference to the monuments of Wessex, is equally applicable to the study of Sanchi's sacred landscape, 'Landscape archaeology, as it is practised, involves the study of systematic relationships between sites. A time-space perspective, on the other hand, is concerned with the routine movement of people through landscapes, constituted by the locales in which they came into contact' (Barrett et al. 1991: 7-8). Theoretical models which focus on the value of the entire landscape as opposed to the isolated monument, must be seen side by side with developments in surface-survey methodologies and in particular the concern with 'off-site' archaeology (e.g. Foley 1981). In India however, although survey as an alternative to traditional site-based archaeology has recently been embraced (e.g. Erdosy 1988; Lal 1984), these are largely enquiries into settlement and subsistence patterns, which only give a cursory glance at the issue of a co-existent ritual landscape. This traditional polarity between the sacred and the profane, despite the concerted effort in

A further development of considerable importance for the study of these monastic sites has been a reaction to static forms of visual presentation (such as those provided by plans and maps) which can suppress the experiential ways in which people interacted with their environment (Thomas 1991: 30; Ingold 1993; Tilley 1994). Chief among the strands of social theory which have influenced recent interpretative models in landscape archaeology have been Bourdieu's theory of

6

practice and Gidden's structuration theory. Both of these, by being manipulated into a phenomenological framework, have been applied to John Barrett or Richard Bradley's 'archaeology of action' (Bourdieu 1977; Giddens 1979: 66-73; Barrett 1989; 1990; 1994; Bradley 1989; 1991), which have helped to draw on the interactive relationships between people and buildings, as opposed to structuralist spatial theories which stressed the inherent and essential meanings of the built environment (e.g. Yates 1989). Of particular relevance to the present paper are those studies which have examined the ways in which the physical layout of Neolithic ritual sites may have helped to maintain behavioural and ideological regularity, especially through the control and restriction of bodily movement (Barrett ibid; Bradley ibid; Tilley 1994; Richards 1993; Bender 1992). The issue of visual experience, and in particular intervisibility between sites has also been recognised as an important consideration when examining ritual monuments within the landscape (Tilley 1994; Scully 1961). This is not to suggest an unproblematic relationship between architecture and ideology (such as that presented in normative spatial models, e.g. Kent 1990); the issue of multivocality and resistance is something which has been addressed in archaeological theory in general (e.g. Barrett 1989; Ferguson 1991; papers in Miller et al. 1989; Bender 1993).

generally absent is an awareness of the relationship between monastic architecture and somatic or sensory experience (although Bentor 1996 describes Buddhist ritual in detail, stupa ritual does not figure in her account). An example which illustrates this point is the debate over whether or not the stupa, in keeping with many traditions of sacred architecture in India, embodies cosmological symbolism (Ruelius 1980: 267-276; Trainor 1997: 99; Paranavitana 1946: 20-24; Bandaranayake 1974: 48-49). By restricting the enquiry to architectural and textual sources alone, as opposed to placing the stupa within its dynamic ritual setting, this discussion ignores a major part of the evidence. As Ruelius points out, although we may be unable to find direct cosmological associations in Sri Lankan stupas, we can only hope to understand the symbolic meanings of archaeological structures by viewing them within their wider ritual context, i.e. the way they are 'made, used and experienced by man' (Reulius 1980: 272). In one way this problem has been redressed (perhaps coincidentally) by Gregory Schopen and Kevin Trainor, who have explored the nature of the relics deposited in stupas and the rituals that surrounded them. Their work has helped to dispel the idea (to some extent a product of Protestant-influenced historical analysis) that the veneration of relics was the exclusive concern of the Buddhist laity (Trainor 1997: 55; Schopen 1997). More recently, Trainor has discussed the way in which a monument can, through the 'force of ritualization', mould the way in which people move around and 'read' its various parts. However, despite touching on some of the concerns of Barrett's 'archaeology of action', Trainor does not make reference to wider theoretical movements in archaeology (Trainor 1996: 18-35).

All of these theoretical models can be fruitfully deployed at the monastic sites of central India. It has been recognised, however, that the more subjective approaches, whilst introducing an important dynamic element to the archaeological landscape, may be guilty of ahistoricity, and of undermining the meaning of specific structures. They also tend to produce generic disembodied actors rather than convincing historical agents, and discount the importance of the type of deity propitiated at a particular spot. Due largely to the limited nature of the data at many of the prehistoric settings in question, the type of 'historical phenomenology' as propounded by Julian Thomas (1991), remains little more than an optimistic speculation. However, although conclusions are necessarily provisional, pending the results of my ongoing survey work in the area, I would like to propose that even at this stage it may be possible to introduce a convincing spatial dynamic to the central Indian monastic sites, thanks to the relatively complete state of the monuments and the rich body of textual, inscriptional and art-historical evidence which can be drawn from them.

Both Schopen and Trainor, having come from text-based traditions of Buddhist scholarship, have prided themselves on pioneering a break-away from the subject's traditional reliance on texts. To them, archaeology 'offers a perspective on what people actually did, as opposed to what they were supposed to do' (Trainor 1997: 61; citing Schopen 1987: 193-4). This awakening, however laudable, is problematic (and ironic) in that it perpetuates the classic opposition of text as text and archaeological record as 'fossil' (for critique, see Conningham 1998). Generally absent is an awareness of recent theoretical movements within archaeology which acknowledge, among other things, the 'textual' quality of the archaeological record itself (Patrik 1985; Tilley 1989; 1991; for Sri Lanka, see Duncan 1990, aspects of which study Trainor nevertheless discusses in 1997: 116-17). Without discounting the importance and usefulness of the large body of stupa-related scholarship, I would submit that there is a need to reconsider these monuments from a phenomenological angle, applying what can be gleaned from texts and inscriptions to what can be found on the ground at specific places. The enquiry I envisage goes beyond scriptural texts and archaeological remains per se and attempts to make a 'somatic' assessment of the ways in which people in the past responded to and interacted with their surroundings. This attempt to recover the ritual dynamics of specific Buddhist sites and the landscape in which they were set, is largely unprecedented within the context of Indian archaeology.

TheStupa

The stupa has been the focus of a disproportionately large body of scholarly work, and its architectural, artistic and symbolic aspects have all been extensively discussed (Mitra 1971; Brown 1986; Pant 1976; Dallapiccola and Lallemant 1980; Snodgrass 1985; for summary, see Brown 1986). In addition, a variety of donative and reliquary inscriptions have been used to shed light on the wider socio-historical context of the stupa (Dehejia 1992; Singh 1996; Willis forthcoming). However despite such a diverse body of scholarship,

7

coincidental that the sites of Morel Khurd, Andher, Sonari and Satdhara are positioned so as to form a rough semi-circle, with Sanchi in the centre, and given the importance of pradaksina, the sites may have been visited as part of a clock-wise processional circuit.

Inter-site spatial relationships

The information gleaned from the reliquary inscriptions found in several of the stupas at Sanchi and the surrounding sites has provided a useful starting point in the attempt to recover how these sites and the surrounding landscape were perceived in ancient times (Cunningham 1854; Majumdar 1940; Willis forthcoming). An analysis of these has shown that all the Buddhist sites in the area were linked to the Hemavata school of Buddhism, led by a teacher called Gotiputa. The Hemavatas appear to have arrived in Vidisha in the second century BC; they took over the older sites of Sanchi and Satdhara and established new monastic centres at Sonari, Morel Khurd and Andher (Figure 1). The reliquaries show that all the sites are linked and were established (or renovated in the case of Sanchi and Satdhara) in a single campaign (ibid.).

A poignant indication of the 'life-force' of the relics and stupas is provided by an inscription on the west gateway at Sanchi. This warns that 'he who dismantles ..... the stone work from this [stupa], or causes it to be transferred to another house of the teacher [acarya kula ], he shall go to the [same terrible] state as those who commit the five sins that have immediate retribution' (Majumdar1940, I: 342, no.404; also inscription no.396; translation by Schopen 1997, 129; for an alternative interpretation of the term acarya-kula , see Majumdar 1940, I: 341, who argues that the inscription is a warning against removing the relic to another non-Theravada school). The five sins alluded to in this record are those defined in the Vinaya; the fact that four of the five sins involve either murdering or injuring a living being illustrates the degree of 'humanness' which was attached to the stupa. The fact that the stupa was seen as a living being, can therefore be taken as one of the most obvious explanations for the need to protect it from dangerous forces. On a cosmic level, the stupa represents what Eliade would call a hierophany, the breaking through of the sacred into the profane (Eliade 1961: 43). Ritually speaking, any liminal point is vulnerable (Turner 1964) which is why the stupa was protected at its point of emergence from the ground by railings with staggered gates, often in the shape of a svastika. This diverted the gaze of the evil eye, which is traditionally understood to only travel in straight lines.

In addition to being connected through the agency of the Hemavatas, it is my contention that the element of intervisibility between the monastic sites may have played an important role in forming an interlinked Buddhist landscape. For example, from the hilltop at Sanchi, it is possible to see the stupas at Morel Khurd to the east; further in the distance, is the mountain of Andher; and to the south-west is a line of hills, immediately behind which are Sonari and Satdhara. This arrangement was far from a passive network. Texts and inscriptions indicate that stupas were not simply important buildings, but containers of a 'living presence' which projected the power of the Buddha, the Arhats and the Dharma into the surrounding space (Schopen 1997). At Andher, the most isolated of the sites being discussed, the stupas are perched dramatically on a high cliff-top and can be seen from a considerable distance in several directions. These stupas, like those at Sanchi, highlight the importance of 'seeing' (dassana ) at the Buddhist sacred sites. Both Trainor and Schopen have drawn attention to the analogy between 'seeing' and worshipping, and have noted the canonical sanction for this in the Mahavamsa (Digha Nikaya 16: 5: 7-8) which advocates that the four main pilgrimage sites connected with the Buddha's life are places which 'ought to be seen' (dassaniya) (Trainor 1997: 174-77; Falk 1976; Schopen 1997: 117, n.9). The objects of this 'seeing' are not merely sacred places, but 'relics of use' (paribhogika dhatu); the places where the Buddha was born, gained enlightenment, taught and died were transformed into 'relics' because they had been 'used' by the Buddha. Simply seeing these places therefore, was equivalent to seeing and worshipping the Buddha. The parallel concept of darshana, found in varying degrees in other Indian faiths, ensures that through the auspicious sight of the venerated object, a devotee gains spiritual merit (Eck 1981). These factors help to explain the importance of ritual circumambulation (pradakshina) as one of the main aspects of stupa-worship (for canonical sanctions, see Falk 1976: 290; also Trainor 1996). Furthermore, as Nancy Falk has suggested, pilgrimage-circuits around Buddhist sacred sites may have taken the form of a huge pradakshina, similar to the pilgrimage circuits at Hindu tirthas (1976: 88; for modem Buddhist analogy at Bhaktapur, Nepal, see Gutschow 1980). Accordingly, it may not be

Stupas could also be threatened by 'mundane' human action, which may help to explain why Andher, to take the most dramatic example, occupies such a strategic position in the landscape (Figures 3 and 4). In the 19th century, Cunningham was impressed by the view from Andher and commented that it was possible to see across to Lohangi rock (in Vidisha) as well as the stupas at Morel Khurd and Sanchi (op. cit.: 342). Not only did this element of intervisibility help to enforce the continuous presence of the Buddhist Dharma across the landscape, but it also provided strategic advantages befitting the fortress-like location of the site. With a sheer cliff on one side and the long gradual slope on the other, it would have been difficult for anyone to approach the stupas without being noticed. This military aspect suggests that monastic sites were subject to hostile threats at various times in their history. For example, the hill at Sanchi was given a very substantial wall, probably in the eleventh century, and Verardi has noted that the monastery at Uddandapura may have been attacked by Muslims because of its resemblance to a fortress (Verardi 1996: 244). The post-Mauryan king Pusyamitra seems to have been hostile towards heterodoxy, especially Buddhism. Traditions which associate the horse sacrifice with the Sungas and identify them as brahamanas lend support to their position as representatives of the vigorously orthodox (smarta). This has led some archaeologists to suggest that the main stupa at Sanchi could have been intentionally damaged in the post-Mauryan era (Marshall op. cit.: 23-4; Verardi 1996: 230-31). While there is no direct proof

8

Figure 3. View of stupas at Andher from below

Figure 4. View from stupas at Andher

9

that Sanchi was attacked by the Sungas, the circumstances are sufficiently compelling to see the injury of the stupa there as an assault on the Shasana and, quite literally, the 'body' of Buddhism. This was not simply a question of heterodox versus orthodox views. As the story of the War of the Relics show (as depicted on the Sanchi gates), relics were sources of contention from the earliest days of Buddhism. This was because relics lent themselves to use as instruments of political legitimacy, the spread of the Dharma being easily appropriated by kings who sought to draw on analogies between themselves and the Buddha as Dharmaraja and Cakravartin. Although the use of relics as instruments of polity received its fullest elaboration in Sri Lanka, Strong has put forward convincing arguments that this mode of kingship was first developed by Asoka in the third century BC (Strong 1983; for Sri-Lanka, see Trainor 1997; Duncan 1990) Intra-site spatial relationships

While the locations chosen for the hill-top retreats ensured a safe refuge for both the relics and the Buddhist monks, I would also like to suggest that the dynamics of the monastery and relic cult help to explain certain spatial relationships between the structures that survive on the ground within the monastic complexes themselves. This adds considerably to our understanding of the sites by explaining how people moved round the stupas and other monastic structures, thus possibly providing the first viable 'archaeology of action' within the Indian context. Over a century and a half ago, Cunningham noticed that the angle between stupas 1 and 2 at Sonari was remarkably similar to that at Sanchi. He even suggested that 'there must have been some peculiar significance in this particular angle' but failed to pursue this line of investigation any further (op. cit.: 315). One explanation for this arrangement is provided in the Vinaya-rules, which stipulated that a monk's stupa should be positioned according to his rank in life and, more specifically, that the stupas of the Buddha's chief disciples, Sariputa and Mahamogalana should be appropriately situated beside that of the Buddha. The stupas containing the relics of Sariputa and Mahamogalana at Sanchi and Satdhara appear to conform to these regulations (Mahaparinirvanasutra 95.2.7; Roth 1980: 184-5; Willis forthcoming). Also in keeping with the Vinaya rules is the fact that stupa 2 at Sanchi is set outside the main compound wall, seemingly because it contained the relics of the Hemavatas who post-dated the Buddha by several hundred years. It is important to stress, however, that since the main stupa at Sanchi came up during the Asokan period (by which time the Vinaya had been substantially codified) the Hemavatas' stupa which was constructed a hundred years later, was conforming to an earlier power structure. In contrast the entire complex at Sonari appears to have been built by the Hemavatas in the second century BC. Because the site was 'new' the Hemavatas were able to manipulate the rules and position the stupa which contained their relics directly beside the main stupa at the site. This would seem to be a blatant statement of Hemavata eminence.

My contention here is that the positioning of the monuments may also have been part of a strategy aimed at protecting the relics through surveillance and by regulating access to the stupas. In all instances, the position of the stupas seems to bear a relation to the large stone-masonry platforms that are found at all the monastic sites. Cunningham suggested that they were bases for temples, but their true function still puzzled him (op. cit.: 328). His reason for this supposition was that there were ruins of a temple on top of the platform at Morel Khurd. The temple remains at Morel Khurd, however, date from at least the seventh century, i.e. eight hundred years after the platforms were first built. The later addition of a temple, absent at the other sites, cannot be taken as a useful guide to the original function of these structures. All the platforms are covered with accumulations of crumbling red brick, indicating that the upper walls were once brick-built with, most probably, timber superstructures. Narrative reliefs at Sanchi and other early sites leave no doubt that most prominent buildings in ancient India were raised on poles or large platforms with the upper structure made of wood and other perishable materials. Later temples (such as those at Sanchi) continued to be built on large platforms and this probably inspired the medieval additions to Morel Khurd. After about the eighth century, however, the extensive use of wood for sacred structures was mainly restricted to the towering monasteries of Nepal and other parts of the Himalayas. The perpetuation of both of these elements in remote areas leads one to suspect that the vertically orientated Himalayan monasteries may also find their ultimate origin in the stone platforms of India, a subject that warrants further investigation. For example, the Morel Khurd platform reaches almost 10 metres and would have been surmounted by an additional structure of two or three storeys. When seen from Andher or Sanchi, the platform is even now one of the most conspicuous architectural features at Morel Khurd; in the first century it would have been twice the present height and would have dominated the entire site. All the platforms have internal staircases which cut into the body of the structure, (Andher seems to be an exception here. However since the platform is completely ruined, a staircase might be revealed following excavations). This would mean that to enter the main monastic building one would have entered a narrow covered passage. This device, enhanced at Satdhara by a bent entrance, would have been a useful device for monitoring movement in and out of the building. Marshall notes that the second phase of temple 40 at Sanchi has two stairways internally set into the northern wall of its podium, which are similar in style and date to the stairway of the 'temple' (i.e. the platform in question) at Sonari (Marshall 1940: 65). This tradition was not abandoned in later architecture, for example the medieval Jaina temple at Ranakpur. The practice of carving guardian figures (dvarapala) at the doorways of later temples, lends support to the idea that entrances were regarded as vulnerable features, in need of protection and regulation. Although they are not found before the fifth century, the yaksha figures often found on the gateways of earlier Buddhist monuments seem to have served the same purpose.

• •

• •

~

I I I

I I

.

I I

---------

•-----.-♦---------

•.

---------------- - ------;/

,, ,,

Figure 5. Plan of platform and stupas at Sonari

Figure 6. View of Sonari stupas from platform noted that later monastic buildings tended to be oriented with their entrances facing the central stupa, albeit situated at some distance from it, and that later on, the stupa became incorporated into the centre of the monastery compounds, such as the 7th century monastery below temple no. 45 at Sanchi (Schopen

Even without their superstructures, the platforms at the Buddhist sites provide an all-encompassing panorama of the stupas (Figure 6). Going back to our discussion of 'seeing', this view meant that the occupants of the surmounting building were afforded the auspicious sight and presence of the Buddha and the Arhats. It has been

11

the stairs has been drawn to a slightly larger scale to highlight their position. Several factors support my suggestion that the staircase was singled out as the part of the stupa most in need of protection. The stairs took one to the upper balustrade, that is as close as it was possible to get to the relic chamber, where one could circurnnambulate the sacred 'traces' of the Buddha (Falk 1976). On a cosmic level they offered the possibility of ascending to a higher spiritual plane. Adrian Snodgrass discusses this connection with reference to the stairways of south-east Asian terrace stupas, paying particular attention to their serpent-balustrades, as symbols of the 'rainbow bridge' which links heaven and earth (1985: 282-5), but whose role as guardians of the Buddhist relic should not be forgotten, (the link between nagas and protection is discussed later in the paper). That the stairway led to the most important part of the stupa is also corroborated by Trainor's description of the rules stipulating that the stupa should only be approached after removing one's shoes and assuming a proper attitude of respect (Trainor 1997: 154; for ethnographic descriptions of modem Buddhist ritual, see Gombrich 1988:115-6). Furthermore, there are countless canonical references to the ability of relics to 'disappear' or to make their own arrangements to move if not accorded proper worship (Trainor 1997).

1997: 258 ff.; Roth 1980: 186). Despite this recognition, the ritual dynamics at these early sites has not been fully understood, prinicipally because the platforms discussed here have not hitherto been recognised as monasteries (compare Trainor 1996: 32, who states that 'there are no monastery remains clearly datable before the Gupta period'). This perhaps explains why stupas 1, 3 and 5 at Sonari are not in a straight line but slightly staggered, which allows for all three monuments to be kept in clear view from the platform (Figure 5, in which the lines of vision from the platform are shown by arrows). Similar arrangements are found at all of the other sites (for Morel-Khurd and Sanchi, see Shaw forthcoming: 40). As discussed above, this would have ensured that maximum spiritual merit was accrued through the act of 'seeing'. I would also like to propose that this configuration allowed for a further means of surveillance, and that this concern with .surveillance extended into the surrounding area, as suggested by the fact that one of the platforms at Satdhara doubles as a 'look-out' for the river-valley below (Figure 7). A curious fact at all of the sites is that there is no consistency in the positioning of the staircases leading to the upper terraces of the stupas. For example, the stairway of stupa 2 faces east, whilst that of numbers 1, 3, and 5 are all positioned towards the west. Marshall notices the same lack of 'order' at Sanchi, where the entrance of stupa 1 is in the south, whilst that of stupa 2 faces east. (Marshall 1940, II: notes accompanying plate 71). Considering the importance usually attached to the cardinal directions, this is indeed puzzling. However an examination of the spatial organisation at Sonari, for example, shows that the particular positioning of the staircases ensures that all of them are visible from the platform. This is illustrated in Figure 5 where the size of

At Satdhara the broad sweep of the double staircase is very wide and cannot be taken in completely from the platform which overlooks the river. This factor seems to have led to the need for a second platform from which it was possible to keep the other half of the staircase, as well as the adjacent small stupa under surveillance (Figure 7, with arrows again used to show lines of vision).

9...__..___..____.____.____.___,__,___,____.___,1qom

Figure 7. Plan of platform and stupas at Satdhara

12

Serpent-guardians

forthcoming: 41-2).

In considering the protection and worship of relics, it is important to recognise the role played by serpent-kings (nagaraja), ancient snake-deities who were assimilated into both the Buddhist and Brahmanical religious imagination. There are countless accounts of the Buddha challenging and superseding the power of dangerous naga-kings, such as the story of how he overcame a fierce serpent inhabiting the fire-temple of the Kashyapas (the episode is shown in one of the gatereliefs at Sanchi) (Marshall 1940: pl. 52). Demoted from their position as powerful deities in their own right, naga-kings were often subordinated to guardians of the Buddhist relic, a direct adaptation of their role as protectors of subterranean and sub-aquatic treasure. Buddhist texts frequently recount how relics were guarded by nagas until they were made available for human worship (e.g. Dathavamsa 4 : 28-37; see Shaw

In addition there is strong archaeological evidence for naga cults and other pre-Buddhist practices at major Buddhist sites, elements of which often seem to have reemerged following the demise of Buddhism (Cunningham 1892; Kosambi 1962; for similar process in Thailand, see Byrne 1995). Maniyar Math at Rajgir, for example, derives its name from Manikara Naga who, in local tradition, resides in a well as the guardian of treasure and who, in the Mahabharata, is the protector and rain-giver of Rajgir (Bloch 1906: 103). The naga's reputation as relic-protector is also given a very practical dimension through the very real fear for robbers, or archaeologists, that poisonous snakes may take refuge from the sun beneath the cool rocks of the stupa (Willis, pers. comm.). Furthermore local folk-tales are full of references to temples or entire cities being founded from

Figure 8. Nagafigure at Nagauri

13

the proceeds of treasure (Cunningham 1880: 65).

found

in

Gwalior) from ancient Vidisha, for example, underline the importance of the area to faiths other than Buddhism. That Vidisha was a pilgrimage centre of pan-Indian importance is attested by the discovery of a Vasudeva temple, which dates to at least the second century BC, and the associated inscribed pillar, set up by Heliodorus, an envoy of the Indo-Greek king Antialkidas of Taxila (Khare 1967). This level of heterogeneity might have increased the need for security-measures at the stupasites, but we should not conclude that the relationship was always one of conflict. Not only did much in Buddhism grow out of a common pool of Indian religious themes, but as already alluded to, many of India's sacred sites emerged from older folk traditions which centred around place-specific cult spots (Eck 1981; for examples see Shaw forthcoming: 44 ).

snake-holes

The first anthropomorphised depictions of serpent deities, as opposed to earlier polycephaloustheriomorphic, date from the first century AD (Misra 1982)· One such naga stands at Nagouri, a low hill immediately south of Sanchi (Figure 8; Marshall 1940: 244). The naga faces directly towards the main stupa and in keeping with the nag a' s traditional role as protector of relics, can be envisaged as guarding not only the stupas but the whole valley between the two hills. Interestingly, the traditional association between the naga-cult and water may also have been maintained because an ancient embankment suggests that there was once a small lake between Sanchi and Nagouri (ibid.:13)· Other naga-figures have been found at the nearby villages of Gulgaon (ASIAR 1972-3: 59; Williams 1976), Firozpur and Andher, where the naga stands in the valley floor below the stupas (see Shaw forthcoming: 43)

The above discussion is far from definitive, but I think enough material has been given to show parallels with 'archaeologies of action' and other post-processual approaches to landscape archaeology. These may be grouped together because of a shared sensitivity to the wider archaeological and cultural settings of ritual sites: for example, the issue of intervisibility; the relationship to topographical features and elevation; or consistencies in directional orientation (Sherratt 1996). Although a more 'somatic' approach may help to transcend the static models of structuralist or processual archaeology, the main problem (already alluded to in our theoretical preamble) is that this type of analysis often produces ahistorical models which fail to distinguish between different categories of both worshippers and deities. The advantage of doing archaeology at the Buddhist sites in central India is that we do have relevant textual and epigraphic information without which much of the sites' spatial or iconographic significance would be obscured. It is perhaps necessary to state the obvious, namely that the texts do not have a one-to-one relationship with the monuments that survive. There is much in them which can no longer be traced in archaeological terms and there is equally much on the ground which cannot be textually corroborated. More particularly, there is no Vinaya material, as far as I am aware, which contains specific instructions as to how the stupa should be protected and controlled in spatial terms. Furthermore, given the importance of the Buddhist relic, it is indeed puzzling that the Pali Vinaya contains no references to stuparelated ritual, an anomaly which has excited the attention of scholars of Buddhism (Roth 1990; Schopen 1997: ch. 5; Hallisey 1990; Gombrich 1990)·

The fact that naga-figures guard from afar conforms to the Vinaya which list 'serpents and other non-human forms' as one of the classes of beings banned from the monastic complex (Frauwallner 1956: 76-7). Given the frequency of metamorphosis one encounters in ancient Indian texts, this probably refers to both serpent-kings and their devotees. Since naga-kings would have been feared and respected by a large percentage of the population, it was sensible for the Buddhist community to adopt them as guardians. This is particularly apt in the light of Schopen's analysis of the names in the donative inscriptions at Sanchi (1996). He argues that since a high percentage of these have naga associations, they provide some insight into local forms of Buddhism, as opposed to that of the canonical tradition (see also Singh 1996). Of course in the light of the common tactic of setting up new religious centres at sites already considered sacred by the local population it is also possible that the stupa sites were actual naga cult-spots in their own right (see Cunningham 1892; Bloch 1906, Kosambi 1962). Not only do such actions reveal a common regard for what Rudolph Otto calls the numen locii, but it also means that whenever anyone looked at the naga, their eye would also have been drawn to the stupas on the hill above (for numen locii, see Otto 1925; for studies of monuments built around older locales, see Tilley 1996; Bradley 1993; Scully 1961). 'Sacred Landscapes': Discussion

Conclusion Although we can distinguish between the types of deities being propitiated at particular sites, it is inaccurate to speak of a 'Buddhist period' or indeed of an exclusively 'Buddhist landscape' (however, see Gutschow 1980: 138 for a description of how even in Bhaktapur, the Nepali town dominated by the Hindu deity, Durga, there is still a strong sense of a Buddhist landscape). Not only must we recognise the potential for resistance to the devices of control discussed in this section, we must also remember the schisms that already appeared in the Buddhist Sangha as early as the Asokan period. The great yaksha figures and the other early sculptures of Hindu deities (many in the Archaeological Museum,

The details of this particular controversy need not detain us here; the main issue to stress for the present purpose is that although we may not find textual correlations or explanations, there is still enough convincing evidence to support my principle thesis, namely that many aspects of the relic-cult were played out in spatial, and in particular, visual terms. The dynamics of these sites and buildings, so long obscured because of static models of text and monument, can be regained (albeit only in part) by considering how people in the past looked at, moved round and responded to 'sacred presences' in the wider landscape.

14

Acknowledgements

Bourdieu, P. 1977. Outline of a Theory of Practice. Cambridge: Cambridge University Press.

I would like to thank Michael Willis for asking me to contribute to his forthcoming book a section which deals with many of the issues discussed in this chapter (Shaw forthcoming). I have benefited greatly over the last year from his help and support, and from countless illuminating discussions on Buddhist art and architecture. I am also grateful to him for giving permission to reproduce figs. 5 and 7 from the same publication (Willis forthcoming). Thanks are also due to Tim Insoll for inviting me to present a nascent version of this paper at the Cambridge Conference on Archaeology and World Religion. Others who have kindly read earlier drafts of this paper and offered valuable suggestions and critical comments are Ian Hodder, Raymond Allchin, Nicole Boivin, Emma Blake, Paul Lane, and the anonymous readers on the selection panel for this book. I am extremely grateful to all of them. Needless to say, any shortcomings in this paper are my own.

Bradley, R. 1989. Monuments and places. (In), Garwood, P., Jennings, D., Skeates, R., and Toms, J. (eds.), Sacred and Profane. Oxford: Oxford University Committee for Archaeology Monograph 32. Bradley, R. 1991. Altering the Earth: The Origins of Monuments in Britain and Continental Europe. Edinburgh: Society of Antiquaries of Scotland. Brown, R.L. 1986. Recent Stupa Literature: A Review Article. Journal of Asian History 20: 215-32. Byrne, D. 1995. Buddhist Stupa and Thai Social Practice. World Archaeology 27: 266-81. Carmichael, D.L., Hubert, J., Reeves, B., and Schanche, A. 1994. Introduction. (In), Carmichael, D. L., et al. (eds.), Sacred Sites, Sacred Places. pp.1-8. London: Routledge. Coningham, R. 1998. Buddhism 'Rematerialized' and the Archaeology of the Gautama Buddha. Cambridge Archaeological Journal 8: 121-26.

References

Agrawal, R.C. 1997. Stupas and Monasteries: A Recent Discovery from Satdhara, India. (In), South Asian Archaeology 1995. vol. 1. pp. 403-15. Delhi. Munshiram.

Cunningham, A. 1854. Bhilsa Tapes. London: Smith Elder.

Asad, T. 1979. Anthropology and the Analysis of Ideology. Man 14: 607-27.

Cunningham, A. 1880. Archaeological Survey of India Annual Report I 0.

A SIAR - Archaeological Survey of India Annual Report.

Cunningham, A. 1892. (reprint) Mahabodhi, or the great Buddhist temple under the Bodhi tree at Buddha-Gaya. Varanasi: Indological Book House.

Bandaranayake, S. 1974. Sinhalese Monastic Architecture: The Viharas of Anuradhapura. Leiden: E.J. Brill.

Dehejia, V. 1992. The Collective and Popular Basis of Early Buddhist Patronage: Sacred Monuments 100BCAD250. (In), Stoller-Miller, B. (ed.), The Powers of Art: Patronage in Indian Culture. Delhi: Oxford University Press.

Barrett, J. N. 1989. Towards an Archaeology of Ritual. (In), Garwood, P., Jennings, D., Skeates, R., and Toms, J. (eds.), Sacred and Profane. Oxford: Oxford University Committee for Archaeology Monograph 32.

Duncan, J. S. 1990. The City as Text: The Politics of Landscape Interpretation in the Kandyan Kingdom. Cambridge: Cambridge University Press.

Barrett, J.N. 1990. The Monumentality of Death: The Character of Early Bronze Age Mortuary Mounds in Southern Britain. World Archaeology 22.

Eck, D. 1981. Darsan: Seeing the Divine Image in India. Chambersberg: Anima Books.

Barrett, J.N. 1994. Fragments from Antiquity: An Archaeology of Social Life in Britain, 2900 - 1200 BC. Oxford: Blackwell.

Eliade, M. 1961. The Sacred and the Profane. New York: Evanston: Harper and Row.

Bender, B. 1992. Theorising Landscapes, and the Prehistoric Landscapes of Stonehenge. Man 27 (4): 735-55.

Erdosy, G., 1988. Urbanisation in early Historic India. Oxford: BAR.

Bender, B. (ed.), 1993. Landscape: Politics and Perspectives. Providence: Berg.

Evans, C. 1985. Tradition and the Cultural Landscape: An Archaeology of Place. Archaeological Review from Cambridge 4: 80-94.

Bentor, Y. 1996. The Consecration of Images and Stupas in Inda-Tibetan Tantric Buddhism. Leiden: E.J. Brill.

Falk, N. 1976. To Gaze on the Sacred Traces. History of Religions 16: 281-193.

Bloch, T. 1906. Maniyar Math. Archaeological Survey of India Annual Report (1905-6).

Ferguson, L. 1991. Struggling with Pots in Colonial South Carolina. (In), McGuire, R., and Paynter, R.

15

(eds.), The Archaeology London: Unwin Hyman.

of Inequality.

pp.28-39.

Lal, M., 1984. Settlement history and the rise of civilisation in the Ganga-Yamuna Doab (from 1500BC300AD). Delhi: B.R.

Foley, R. 1981. Off-Site Archaeology: An Alternative Approach for the Short-Sited. (In), Hodder, I. et al. (eds.), Pattern of the Past. pp.157-83. Cambridge: Cambridge University Press.

Lane, P. 1986. Past Practices in the Ritual Present: Examples from the Welsh Bronze Age. Archaeological Review from Cambridge. 5(2): 181-92.

Frauwallner, E. 1956. The Earliest Vinaya and the Beginnings of Buddhist Literature. Rome: ISMEO.

Maisey, F.C. 1892. Sanchi and its Remains. London.

Giddens, A. 1984. The Constitution of Society: Outline of the Theory of Structuration. Cambridge: Polity.

Marshall, J. and Foucher, A., with epigraphic notes by N.G. Majumdar. 1940. The Monuments ofSanchi. 3 vols. London.

Gombrich, R. 1988. Buddhism Transformed: Religious Change in Sri Lanka. Princeton: Princeton University Press.

Miller, D., Rowlands, M. and Tilley, C. (eds.), 1989. Domination and Resistance. London: Unwin Hyman Ltd.

Gombrich, R. 1990. Making Mountains without Molehills: The Case of the Missing Stupa. Journal of the Pali Text Society 15: 141-43.

Misra, B.N. 1982. Sculptural Representations of Naga Machalinda. (In), Sharma, R. K. (ed.), Indian Archaeology: New Perspectives. pp.293-302. Delhi: Agam Kala Prakashan.

Gutschow, N. 1980. The Urban Context of the Stupa in Bhaktapur, Nepal. (In), Dallapicolla, A.L. and Lallemant, S.Z. The Stupa: Its Religious, Historical and Architectural Significance. pp.13 7-43. Wiesbaden: Franz Steiner.

Mitra, D. 1971. Buddhist Monuments. Calcutta. Otto, R. 1925. The Idea of the Holy. London: Oxford University Press. Pant, S. 1976. The Origin and Development of Stupa Architecture in India. Varanasi: Bharata Manisha.

Hallisey, C. 1990. Apropos the Pali Vinaya as a historical document: a reply to Gregory Schopen. Journal of the Pali text society 15: 197-208.

Paranavitana, S. 1946. The Stupa in Ceylon. Colombo. Hinuber, O.V. 1990. Khandhakavatta: Loss of Text in the Pali Vinayapitaka. Journal of the Pali Text Society 15: 127-38.

Parker-Pearson, M. 1982. Mortuary Practices, Society and Ideology: An Ethnoarchaeological Study. (In), Hodder, I. (ed.), Symbolic and Structural Archaeology. pp.99-113. Cambridge: Cambridge University Press.

Hodder, I. 1984. Burials, Houses, Women and Men in the European Neolithic. (In), Miller, D., and Tilley, C. (eds.), Ideology, Power and Prehistory. pp.51-68. Cambridge: Cambridge University Press.

Patrik, L. 1985. Is there an Archaeological Record? (In), Schiffer, M. (ed.), Advances in Archaeological Method and Theory. Vol. 8. pp.27-62.

Ingold, T. 1993. The Temporality of Landscape. World Archaeology, 25 (2).

Richards, C. 1993. Monumental Choreography: Architectural and Spatial Representation in Late Neolithic Orkney. (In), Tilley, C. (ed.), Interpretative Archaeology. Pp.143-78. Oxford: Berg.

Kenoyer, M.M., Clark, J.D., Pal, J.N., and Sharma, G.R. 1983. An Upper Palaeolithic Shrine in India? Antiquity 57: 88-94.

Roth, G. 1980. Symbolism of the Buddhist Stupa. (In), Dallapiccola, A.L. The Stupa. pp.183-217. Wiesbaden: Franz Steiner Verlag.

Kent, S. (ed.), 1990. Domestic Architecture and the Use of Space: An Interdisciplinary Cross-Cultural Study. Cambridge: Cambridge University Press.

Ruelius, H. 1980. The Stupa in the Silpasastras of the Sinhalese. (In), Dallapiccola, A.L. The Stupa. pp. 26799 .. Wiesbaden: Franz Steiner Verlag.

Khare, M. D. 1969. Discovery of a Visnu temple near the Heliodorus pillar, Besnagar, Dist. Vidisha (M. P.). La/it Kala 67: 21-7.

Schopen, Presence Study in 193-225.

Kosambi, D. D. 1962. At the Crossroads: A Study of Mother Goddess Cult Sites. (In), Kosambi, D.D. Myth and Reality. pp.82-109. Bombay: Popular Prakashan.

G., 1987. Burial 'ad sanctos' and the Physical of the Buddha in early Indian Buddhism: A the Archaeology of Religions. Religion 17: (Reprinted in Schopen 1997: 114-147).

Schopen, G. 1996. What's in a Name: The Religious Function of Early Donative Inscriptions. (In), Dehejia, V. (ed.), Unseen Presence. pp.58-73. Bombay: Marg.

Lahiri, N. 1996. Archaeological Landscapes and Textual Images: A Study of the Sacred Geography of Late Medieval Ballabgarh. World Archaeology 28: 24464.

16

Schopen, G. 1997. Bones, Stones, and Buddhist Monks:

Tilley, C. 1989. Interpreting Material Culture. (In), Hodder, I. (ed.), The Meaning of Things. pp.185-94. London: Unwin Hyman. Tilley, C. 1991. Material Culture and Text: The Art of Ambiguity. London: Routledge.

Collected Papers on the Archaeology, Epigraphy, and Texts of Monastic Buddhism in India. Honolulu: University of Hawai'i press. Scully, V. 1962. The Earth, the Temple and the Gods. New Haven: Yale University Press.

Tilley, C. 1994. A Phenomenology of Landscape: Places, Paths and Monuments. Oxford: Berg.

Shaw, J. 1999. (forthcoming). Relics and Monasteries in the Sacred Landscape. (In), Willis, M., with contributions from Cribb. J., and Shaw. J., Buddhist Reliquaries from Ancient India. pp.28-45. London: British Museum Press.

Tilley, C. 1996. The Power of Rocks: Topography and Monument Construction on Bodmin Moor. World Archaeology 28: 161-75. Trainor, K. 1996. Constructing a Buddhist Ritual Site: Stupa and Monastery Architecture. (In), Dehejia, V. (ed.), Unseen Presence. pp.18-35. Bombay: Marg.

Sherratt, A. 1996. Settlement Patterns or Landscape Studies? Reconciling Reason and Romance. Archaeological Dialogues 3.

Trainor, K. 1997. Relics, Ritual, and Representation in Buddhism: Rematerializing the Sri Lankan Theravada Tradition. Cambridge: Cambridge University Press.

Singh, U. 1996. Sanchi: The History of the Patronage of an Ancient Buddhist Establishment. Indian Economic and Social History Review 33: 1-35.

Turner, V. 1964. Betwixt and Between: The Liminal Period in Rites de Passage. (In), Helm, J. (ed.),

Snodgrass, A. 1985. Symbolism of the Stupa. Ithaca: Cornell University.

Symposium on New Approaches to the Study of Religion. pp.4-20. Seattle.

Strong, J.S. 1983. The Legend of King Asoka.: A Study and Translation of the Asokavadana. Princeton: Princeton University Press.

Verardi, G. 1996. Religion, Rituals and the Heaviness of Indian History. Annali 56: 215-53.

Thomas, J. 1991. Rethinking the Neolithic. Cambridge: Cambridge University Press.

Willis, M., with contributions from Cribb. J., and Shaw. J. 1999. (forthcoming). Buddhist Reliquaries from Ancient India. London: British Museum Press.

17

The Early Syriac Liturgical Drama and its Architectural Setting. Emma Loosley, School of Oriental and African Studies, University of London. Georgio Arbelensi vulgo adscripta. Over the course of this century academics have disputed the original attribution of the manuscript to the ninth-century author George of Arbela. It now seems likely that it was written in the eighth century by an unknown East-Syrian (Nestorian) theologian. Along with mention of the bema (a straight-sided horseshoe-shaped platform in the nave facing the apse) and its function in a variety of other Syriac and Armenian sources we can use this manuscript with the material remains and knowledge of current liturgical practice to reconstruct the ritual life and world view of the early Christian population. The importance of the Syrian tradition is that it extended the concept of microcosm further than other early Christian traditions by the extensive use of visual imagery coupled with liturgical furniture

Introduction

Spending Easter with the Syrian Orthodox Church is an experience that is in many ways alien to those of us who have been raised in the traditions of western Christianity. From the first day of Lent onwards the community is preparing itself for the momentous events of Passion week. On Good Friday the church is packed hours before the service. Old ladies fight each other for the seats at the front in behaviour not unlike that of teenagers wanting the best view of the latest music sensation. The events of the next three hours are played out in the sanctuary of the church. The people reach fever pitch when the flower filled bier, representing Christ's body taken down from the Cross, leaves the east end of the church and is manhandled around the building by local young men acting as pallbearers. Those standing around the perimeter of the church become almost hysterical as they attempt to take flowers from the bier as a form of blessing. When the bier reaches the safety of the sanctuary again and is symbolically 'buried' behind the altar, the red velvet curtains before the sanctuary swish together, to remain shut until they are opened to celebrate the resurrection late the next day on the eve of Easter Day .

0 0

The impact of the modem Good Friday festival is extremely theatrical and the faithful are merely spectators to the action as it unfolds. There is little evidence of the sacred topography that played an integral part in early Syrian Christianity. The priests and deacons with their (male) pallbearers are the only active participants in the service. The church building itself also plays little part in the proceedings. Apart from the procession all the events centre on the altar, the lectern or Golgotha, as the Syrian Orthodox call it, with the Gospel book and the cross and censers. The Syrian evidence is important because it is the only early major Christian rite that can be studied through its material remains. Antioch was an important centre of early Christianity with Rome and Jerusalem, but whereas in the two latter cases the churches of the area have been destroyed or extensively altered over time obscuring the original plan, in the towns and villages to the east of Antioch the churches have been preserved unchanged allowing us a clear view of their spatial arrangement and liturgical features.

60

120km &O

IRAQ

Figure 1. Location map N. W. Syria and its churches

The limestone massif of north-west Syria is located between Aleppo and Antakya (Antioch) (Figure 1). The area flowered briefly between the fourth and early seventh centuries when olive oil, the spice route and large numbers of pilgrims travelling to visit local holy men, as well as en route to the Holy Land, created prosperity in the hinterland of the great city of Antioch. The communities which prospered there were Christian and spoke Syriac, a form of Aramaic, with the more educated possessing a knowledge of the Greek used in Antioch itself. Thus the inscriptions recorded in this area are mainly Greek, with some Syriac and Latin. The

The Church as microcosm is not a new idea. Anthropologists are familiar with microcosm as an element of ceremonies in traditional societies and architectural historians have suggested that this cosmological view is illustrated by a variety of monuments (see McVey1983; McEwan 1993, 1994). In the case of the Syrian liturgical tradition we have a surviving document from the East Syrian tradition that explains the church interior in cosmological terms. This text is known as the Expositio officorum ecclesiae,

18

peace and prosperity enjoyed by these settlements meant that they had the leisure and funds to undertake a number of building projects. Although on a provincial scale, the limestone buildings were constructed skillfully (so much so that in many cases only the roofs are missing today) with competently carved, but simple, decoration. The ruins are so numerous that locally the people call the area 'the Dead Cities'. The area was surveyed by Georges Tchalenko in the 1950s and his resulting three volume study Villages Antiques de la Syrie du Nord (1953-58) remains the standard text on the sites of the massif.

or mud-brick depending on local resources (Figure 2). The churches of the limestone massif all date from the fourth to the first decade of the seventh century, a span of just over two hundred years. They generally follow the type outlined above, with a few centrally planned martyria providing the exceptions, as at Qal' aat Sem' aan, the well known pilgrimage site that centred on the column that St. Simeon Stylites mounted in the fifth century (Figure 3). The apsed basilica was used for both parish and conventual churches (Figure 4) a fact that in some cases has led to disputes over whether certain churches were part of monastic foundations or instead served the local community with outbuildings for catechism classes and accommodation for the priest. These extra buildings served administrative purposes as church life came to be regulated by diocesan authorities for the first time. It was in the fourth century that Christian ritual was codified for the first time and an organised church hierarchy began to institute a fixed rite according to accepted doctrine.

Very few of the villages have been comprehensively excavated and in most cases only survey work has been carried out. Each village has at least one, and often as many as three, churches. On the hills around the great church of Qal'aat Sem'aan, the area known as the Jebel Sem'aan, there are approximately two hundred churches. There are other 'Jebel' groupings slightly to the south, although the Jebel Sem'aan appears to have a slightly larger number of settlements than the other areas and is better known because of the pilgrimage centre at Qal'aat Sem'aan.

Much attention has been paid to the stationary liturgy that developed in Jerusalem and which is so famously explained by the pilgrim Egeria and later commentators, but for those outside the Holy Land who wished to recreate a sacred topography, symbols had to replace the monuments themselves. This is where the church interior assumed sacred meaning and evolved into a 'holy place' rather than merely a place for the faithful to gather. For those who could not undertake a pilgrimage to Jerusalem the clergy enacted a weekly ceremony that symbolically drew a map of the world for the faithful. The bema played a central part in this evolution. There was no fixed size prescribed for the bema but it was usual for it to sit twelve members of clergy. At the west end was what was called by Tchalenko the 'bema throne'. In actual fact this 'throne' was a lectern to hold the Bible during the Gospel reading (Figure 5). The bema was joined to the sanctuary by means of a sacred pathway known as the bet-sqaqone, which is largely thought to have been purely symbolic, although clearly visible pathways have been documented in Iraqi bema churches.

Amongst these churches around forty-five stand out from the rest. This group is scattered across these hills, with one or two churches further to the south and east and one notable exception (Resafa) located in the desert far to the east. These churches are different because they all possess a bema. Whilst they have been considered from an archaeological perspective (Baccache 1979/1980; Tchalenko 1990) and have been discussed by liturgiologists (Renhart 1995; Taft 1968), surprisingly nobody has yet used first hand knowledge of the buildings coupled with the existing Syriac texts in order to reconstuct the rituals that took place within them. The liturgiologists have not spent time visiting the remains and the archaeologists have not read all the Syriac texts. Together the churches and the texts give a fuller picture of the ritual life of the early church and the world view of those who worshipped within these buildings.

l

II II

Sanc1uary

'----

::S~€-i~~ "'=

1 )

:II: Po.ssible ---'-'-....,..,_.__...I -/-~} site of barrier

+N

11

Figure 2. Syrian 'type' church Church-building and Liturgy

At this stage it is necessary to pause for a moment and discuss church-building in fourth-century Syria. It is widely accepted by both archaeologists and art historians who have concentrated on this area that the Syrian 'type' was an apsed basilica built of limestone, basalt, gypsum

19

The purpose of the bema was to enable the weekly reinactrnent of the crucifixion and resurrection through its use in the liturgy of the word. When it was time for the Gospel reading the clergy would leave the sanctuary, which symbolised the heavenly Jerusalem, and carry the Word, that is Christ represented by His Gospel, along the bet-sqaqone to the bema. At the bema the book would be placed on the lectern, which was, and is still, known as the Golgotha. It would then denote Christ crucified in the earthly Jerusalem. The bema represented Jerusalem in its entirety, on another level it also symbolised the upper room and the events of the last supper. It was no coincidence that the usual number of seats was twelve for the apostles, with Christ presiding over them as the Gospel on the throne. The homily would be conducted from the bema and hymns sung before the clergy would take the Bible and solemnly process back along the bet-sqaqone to the sanctuary to show Christ returning to the heavenly Jerusalem. Therefore the liturgy of the word was conducted in the midst of the people.

Figure 3. Central octagon of the great church of Saint Simeon Stylites ( Qal 'aat Sem 'aan),facing east The Rema, Architecture, Archaeology and function

The clergy, although removed a little as they sat above the faithful on the bema, were not closeted in the east end of the church enacting the rite purely amongst themselves as they are today. By coming out of the east end of the church and bringing the service to the people the process fostered a much closer relationship between the clergy and the laity. They were also able to illustrate the central message of the Gospel in a physical way. By this movement around the interior of the church the clergy were bodily reinforcing the fact that God made his son flesh and sent him to live amongst the people. There was less of a physical barrier between clergy and laity. This ritual of entering the midst of the people and talking from the nave gave an immediacy and visual impact that is lost today as the priest stands high up at the east end of the church, as if on a stage. The earlier intimacy is lost and the feeling is that the priest is apart from, and not standing amongst, his congregation as he prays.

Half the population benefitted particularly from the bema. The women traditionally occupied the western half of the church, entering from a south-western door, whilst the men came in at the south-east and stood before the altar. It was the women that the priest would be addressing directly in his homily, whilst the men stood in front of the bema. This issue of women at the back has long been debated in the Syrian Orthodox Church with the more conservative communities of the Jezira in north-east Syria retaining the tradition until the present day. There has been much debate about the textual sources and whether this seating was followed everywhere. Archaeological evidence has been difficult to interpret because these barriers were apparently wooden and therefore have not survived. However at the church at Kafar Dar' et 'Azze Tchalenko found a notch

20

Figure 4. South-west convent church at Deir Sem 'aan, facing south-east

in a pillar around a metre from the floor (1953-8). He suggested that this was where the wooden barrier that separated the men from the women was attached, and this mark is still visible today forty years after Tchalenko's survey. The pillar had fallen parallel to the bema around a third of the way along the nave, thus the back two thirds of the church would have been taken up by women (Figure 6). This division of the sexes would actually echo the male to female ratio of contemporary Syrian congregations where many men must work on Sundays. Work could have accounted for a similar ratio in late antiquity. Further investigation by the author has revealed similar notches in pillars, in this case still standing, at the nearby bema churches of Burj Reidar and Kharab Shams, in these cases located further east than at Kafar Dar'et 'Azze. This tradition of women at the back is an obvious impediment when the ritual unfolds at the far east of the building and so this division became more of a problem when liturgical changes occurred in the seventh to ninth centuries.

This perhaps supports the view taken by many liturgiologists that liturgical change generally moved from west to east rather than the other way around. Certainly the bema evidence offered from archaeological remains suggests that the earliest bemata were built around Antioch in the mid to late fourth century. The evidence of synagogue bemata and the Manichaean bema festival must be omitted in this case as being outside the central issues of this paper. There is evidence that the churches of Iraq have retained the bema and its associated liturgy (Dr E. Hunter, University of Manchester, pers. com.) which would perhaps suggest that the bema liturgy was connected to a particular rite which died out in western Syria after the early decades of the seventh century. The alternative is that the west Syrian bema evolved into the byzantine ambo, another nave-platform that was moving closer towards what we recognise as the contemporary pulpit today. This disappearence of the bema in Syria coincides with the desertion of the so-called 'dead cities' of the limestone massif, and whilst I do not wish to get involved in the long-running debate as to why the villages were seemingly abandoned at the beginning of the seventh century, this desertion seems to have occurred at much the same time as a reform of the liturgy took place. The Church at Antioch appears to have undergone liturgical change at the same time as lso Yahv III was carrying out his changes further to the east. It has sometimes been argued that there is a link between this apparent return to the cities and the increased influence of the monasteries so, like western Europe in the middle ages, religious change was centred on the monasteries and their attendant schools. This meant that the secular community took less part in religious life.

This interpretation of the bema is not mere conjecture. The Expositio officiorum explains this symbolism but has not been extensively studied yet. This is due to the fact it is written in a particularly opaque form of Syriac and belongs to the East Syrian tradition which has received less attention than the West Syrian tradition because of the old view of the eastern church as heretical. This view is now recognised as stemming from ignorance concerning Assyrian traditions. The text is seemingly eighth century, with extensive references to liturgical changes instituted by the Assyrian (Nestorian) Catholicos lso Yahv III in the seventh century but the liturgy it describes can be related to the existing monuments in Syria.

21

Figure 5. Berna-throne and bema at Qirq Bize, facing east Friday. These processions around the perimeter of the church interior gave, and still give today, the faithful the chance to touch the Bible, the Cross or the vestments of the bishop as he passes, however the earlier intimacy where the clergy sat amidst the laity, albeit at one remove, has now been lost.

This argument goes a long way to helping us in explaining the disappearence of the bema. No bemata have yet been discovered in a monastic institution. They are always found in churches built for the lay community, and with the exception of a village called Ruweiha, said by Tchalenko to have possibly had two bema churches (Tchalenko 1990: 187) no village has more than one bema church, even though these settlements have on average two or three churches each. The case of Ruweiha is not a certain exception, for while one church in the village clearly possesses a bema today, the other church possesses an unmarked flagged floor. Tchalenko bases the evidence for a bema in this church on re-used masonry in the walls of the building and suggests that the bema was removed when the more important 'Church of Bizzos' was built. This would support the hypothesis that only one church at a time in a village possessed a bema.

The contemporary Syrian Orthodox liturgy is static and conducted, with the exception of an Arabic homily, in Syriac. This Aramaic dialect was once the language of the faithful, but now it is not widely studied outside the priesthood and this adds to the distance between clergy and laity. In some cases women now sit on the south side of the church and men to the north, rather than being divided east-west, and in even more liberal communities the sexes are not segregated. Elements of the earlier rituals remain however. The lectern is still called the Golgotha, although it now stands in the sanctuary, rather than at the centre of a symbolic Jerusalem. And a prayer of entrance mentions the bema, although it seems to refer to the sanctuary which is the most commonly known meaning of the word and is used in this sense in Greek and Armenian as well as in Syriac.

This would reinforce the idea that the symbolism of the bema, and in fact the whole concept of the church interior as microcosm, was developed primarily for the education of the laity and was therefore obsolete in a monastic foundation. The implication here is that the monks were educated enough not to need the weekly reinforcement of sacred topography offered by the bema liturgy. This retreat of the learned into the monasteries and the attendant liturgical changes caused a growing remoteness between clergy and laity, and particularly affected the women who remained at the back of the church, far removed from the ritual that was now centred in the sanctuary at the east end of the building. This effect was mitigated in part by the later addition of ritual processions on Holy days, as on Palm Sunday or Good

The clearest sign that vestiges of bema-related liturgical practices remain is on Maundy Thursday when a ritual reinactment of the Last Supper takes place. This is played out on a wooden platform built before the sanctuary. In fact the parallel with the bema is so clear that as soon as I entered the church for this service many friends, most of whom spoke little or no English, but had some knowledge of my work, rushed up and gestured at the platform saying simply 'bema'.

22

As in Europe, where a perceived distance between the priesthood and the faithful led to the Vatican II reforms and other discussions of how a largely mediaeval liturgy could be reformed in order to return to the earlier roots of Christianity, the Syrian Orthodox are attempting

internal reform. This process has begun through the encouragement of choirs and congregational participation in hymns and prayers rather than simply listening to the priests and deacons.

Figure 6. Berna-throne (foreground) and bema at Kajar Dar' et 'Azze, the notched pillar is level with the bag on the left. View facing east

Figure 7. Father Na'aman on the bema at Resafa,facing east

23

If it continues to embrace further contact with the laity we may soon find more interaction, which can only help a younger generation of women who sometimes feel excluded by the current beliefs which prevent them even assisting at the altar. However it is extremely unlikely that the church interior will ever take on such a detailed ritual topography again.

Conclusions

This study illustrates how liturgiologists and archaeologists can benefit by working together. In an area that is famed for the sheer volume of its archaeological sites it is wrong simply to discount the evidence of the buildings in favour of the few texts that remain. In many cases the churches fit exactly with the textual descriptions and fill in the missing gaps. By attempting to fit in the missing pieces and explain why the Syrian Orthodox Church has processions around the church interior on holy days, but otherwise remains essentially static when the earlier texts suggest almost an air of audience participation in the ceremonies, we are not only questioning a change in liturgical practice but also a changing world view. At some time the clergy decided that it was no longer necessary for the whole world to be contained within the four walls of the church. This symbolism was no longer needed in a world that was growing smaller and is now largely forgotten and unlikely ever to gain such prominence again.

There is one instance where a bema survived beyond the seventh century. We can trace the history of this particular church because it was an important cultic centre. The church is located far to the east of the limestone massif in the Syrian desert and is the Basilica of the Holy Cross at Resafa, originally called Sergioupolis. The city was believed to have been founded on the very spot where Sergius, a Roman, was martyred for refusing to kill Christians. A martyrium was built in the city to house his relics, but the city was partitioned in the seventh century causing the relics to be translated to the basilica, which already possessed a bema. Tchalenko discovered a crypt beneath the east end of the church and has hypothesised that the relics were kept there and brought out for special festivals. On these feast days the casket containing the bones was placed upon a ciborium situated in the centre of the bema. The bema at Resafa is much larger and more elaborate than usual. It possesses a small vestibule that the clergy entered before going up another level to the benches that sat twenty four instead of twelve. This could have significance as double the amount of the apostles, but it is more likely that my Syriac tutor, an Aleppian priest, was correct when he suggested that in this case they represent the twenty four thrones of Revelation. We have documentary evidence that this site was a popular centre for pilgrims up until the thirteenth century, and the archaeological remains tell us that the bema must have continued in use throughout this period. The bema is still intact except for its ciborium, throne and benches for clerics, making it clear that it was in use for the entire working life of the church. However during the thirteenth century the Caliph ordered the evacuation of the city and so the church was abandoned with everything else at Resafa. Resafa is three to four hours drive east of Aleppo in the midst of the Syrian desert and this isolation has meant that the remains are exceptionally well preserved. However it is this isolation which may have originally led to the continuation of the bema liturgy long after it appears to have disappeared in other parts of Syria. The location of Resafa so far to the east also suggests that the church could have played a key role in the transmission of the bema to the East Syrians.

However in a climate where the laity are demanding change the Church is looking to the past for clarity in liturgical matters. The archaeological evidence provides information on a period before textual evidence was commonplace and gives us a view of life outside the great cathedrals. There is always textual evidence at the centres of power. The churches of the limestone massif enable us to reconstruct the lives of those at a lower level. By looking at these village churches in context we can come to more informed conclusions concerning the extent to which ritual was codified in these early years of Christianity.

Acknowledgements

I would like to thank the Seven Pillars of Wisdom Trust, the British Institute at Amman for Archaeology and History and the Louis H. Jordan bequest for comparative religion for the grants that enabled this fieldwork. Many people have offered their time and advice. I would like to thank my supervisor at SOAS, Dr A. Palmer, also Dr E. Renhart, Dr S. Brock and Dr E. Hunter for taking the time to answer specific queries. There are many people who aided me in Syria but special thanks must go to Metropolitan Grigorios Yohanna Ibrahim of Aleppo for his hospitality and to Abouna Antoine Deliapo, Farida Boulos and Samir Katerji for their continuing assistance in countless ways.

It is also perhaps this isolation that sees the only continuation of the bema liturgy in contemporary Syria. At nearby Raqqa, Father Na'aman, a Roum (Greek) Catholic priest, is the only clergyman between Aleppo and the Jezira. As such he ministers to all Christians from the Armenians to the Assyrians. Every October 7th, the eve of Saint Sergius' day, he takes a bus of the faithful out to Resafa and celebrates the Feast of Saint Sergius from the bema. A symbol perhaps of the continuity of the ancient liturgical traditions, it is fitting that the congregation should be of mixed denomination (Figure 7).

References

Baccache, E. 1979. Eglises de village de la Syrie du nord, vol. 2; Planches. Paris: P. Geuthner. Baccache, E. 1980. Eglises de village de la Syrie du nord, vol.I; Album. Paris: P. Geuthner. Connolly, R.H. (ed.) 1911-15. Anonymi auctoris expositio officiorum ecclesiae Georgia Arbelensi vulgo adscripta. Paris: Corpus Scriptorum Orientalium

24

Scriptores Syri. Ser.2. Tom.91,92.

Renhart, E. 1995. Das syrische Berna: liturgischearchiiologische Untersuchungen. Graz: Grazer Theologische Studien 20.

Lassus, J., Tchalenko, G. 1951. Ambons Syriens. Cahiers Archeologiques 5: 75-122.

Taft, R.F. 1968. Some notes on the Berna in the East and West Syrian Traditions. Orientalia Christiana Periodica 34: 326-359.

McEwen, I.M. 1993. Hadrian's Rhetoric I. The Pantheon. Res: Anthropology and Aesthetics 24: 55-67. Mc Ewen, I.M. 1994. Hadrian's Rhetoric IL Thesaurus Eloquentiae, the villa at Tivoli. Res: Anthropology and Aesthetics 25: 51-60.

Tchalenko, G. 1953-58. Villages antiques de la Syrie du Nord. Paris: P. Geuthner. Tchalenko, G. 1990. Eglises syriennes Geuthner.

McVey, K.E. 1983. The Domed Church as Microcosm: literary roots of an architectural symbol. Dumbarton Oaks Papers 37: 91-121.

25

a bema. Paris: P.

The Concept of Sacred Space in the Monasteries of Byzantine Palestine. John Binns, Great St. Marys, The University Church, Cambridge. A full account of the early history of the Great Laura, or the monastery of Mar Saba as it has come to be known, is to be found in the Life of Sabas by Cyril of Scythopolis. Cyril was a monk of the Judaean desert, who was born in the Galilean capital, Scythopolis, in about 525; had met, as a young child, the monk Sabas whose life he was later to write; had become a monk in the desert east of Jerusalem; and lived briefly in the Great Laura. He wrote a series of seven biographies of monastic holy men, the longest of which is the Life of Sabas. The Life provides a mass of historical and topographical evidence and has been a valuable source of information about the sites for archaeologists.

Introduction. The Great Laura

Jerusalem is recognised as a holy city by three world religions. For Christians it is the place where Jesus Christ died and rose again. For the Jews it is the place where the Temple was built. For Muslims it is where the Prophet Muhammad left on his journey to heaven. It has therefore many holy places. Among the sites which make a deep impression on visitors and pilgrims is the monastery of Mar Saba, or the Great Laura (Figure 1). It is located about 15km east of Bethlehem in the Judaean desert. The monastery was founded in 478, although most of the present construction was carried out after an earthquake which took place in 1834. There are several factors which contribute to the impression of holiness which it conveys. The landscape is rugged and magnificent. The silence is deep and embracing - accentuated rather than diminished by the occasional sound of desert birds or a shepherd's pipe. The monastic buildings cling precariously to the steep side of the ravine of the wadi Kidron. But holiness is more than an aesthetic, subjective personal opinion. Holiness is shared and has a social dimension. It also has a history. The purpose of this paper is to describe the process by which this site came to be recognised as a holy place. It draws on both literary and archaeological evidence, concerning both the monastery of Mar Saba and also similar sites in the locality as well.

Here are some extracts from Cyril's account of the founding of the monastery in about 478, Our holy father Sabas was praying to God through the night and there appeared to him an angelic fonn who showed him a gorge descending down Siloam to the south of the hill and said, 'if you want to make this desert into a city, stay here and climb the east side of the gorge where you will see an intact cave. Make it your home and God will take care of you'. He found the cave and hung a rope to its mouth to use when ascending and descending because of the difficulty of the ascent. He spent five years here alone by himself and then he began to receive those who came to him each of these he provided with a suitable spot consisting of a small cell and a cave. By divine grace his community grew to seventy persons .... He built a tower on the hill on the northern end of the gorge after the bend, in order to take possession of the place while it was unoccupied .... Then because they had no water, he prayed one night in these words, 'Lord God, if it is your will that this place become a city deign to provide us with a little water'. Then he heard the hoof beat of a wild ass and peering down - it was a full moon - he saw it digging deep into the earth with its hoofs, and when it had made a large hole he saw it drinking. He then climbed down and found flowing water .... On another night he was walking in the gorge alone and saw a pillar of fire whose head stretched into the heavens. He pondered on the words of the Patriarch Jacob, 'How fea,ful is this place! This is none other than the house of God'. There he found a large and marvellous cave that had the shape of a church of God. After setting the cave in order he gave instruction for the divine office to take place here on Saturdays and Sundays. (Cyril 1991: 106111)

■

■

Small Coenobium ■

■ ■

■

■

■

A Hostel

O

500m

r

Cyril's narrative describes the construction of the essential elements of the monastery - the cells in which the monks lived; the tower; the water supply; and the church. There were, of course, other buildings which he refers to at other points of his writing in connection with this or other monasteries, such as a guest house, a

Figure 1. The monastery of the Great Laura, or Mar Saba, showing the location of cells built along the valley, and the church and cave of Sabas at the centre of the Laura

26

example is the monastery of Martyrius now surrounded by the new Israeli settlement of Ma' ale Adummim. It covers an area of 70m x 80m, and includes four chapels, a refectory, kitchen, and stables (Hirschfeld 1990: 2021) (Figures 3-4). The laura was more suited to the eastern, drier regions, where monks lived scattered over a wide area, and either collected edible herbs or made objects such as baskets which could be sold in the local markets. The laurite style of living was more suited to the more arid parts of the desert, and is the type described by Cyril in the extracts above. It will therefore be the main focus of this essay. The Great Laura is the best-known of the many laurae of the Palestinian desert. In spite of this difference, the monastic buildings in each were similar, and confirm the descriptions of Cyril of Scythopolis.

bakery, and a hospital. The monastery provided a home for large numbers of people - in this example taken from an early stage of the history of the monastery there are already seventy monks - in an inhospitable desert environment. Its construction was a gradual process and the final result was a complex community.

Cyril makes reference to several essential features; the cells, the tower, the water supply and the church. The cell provided the living accommodation for the monks. The Greek 'kellion' has a primarily economic reference, and describes the small dwelling which was both a place for living and for working. The cells of the Great Laura have been examined by J. Patrich and others. They discovered forty-five dwelling complexes of varying size and arrangement. As Cyril reports, cells consisted either of a cave or a cell building, or both. Some were large enough to house several monks. These were constructed in a 1.5km stretch of the Kidron gorge. (Hirschfeld 1992: 176-190). They were the basic units out of which the laura was constructed. Then the tower was constructed. Towers are a common feature of the monasteries, and their function has been much debated. While they could serve as living quarters or for storage, their primary purpose was defensive. It provided an observation post, and a place of refuge, but of more significance was the fact that it occupied a commanding position, claiming, as Cyril points out, the land as the possession of the monks. (Festugiere 1963: 82-92). In the case of coinobia, the wall surrounding the monastic buildings took the place of the tower.

L L

Jericho

C L

■ c

C

L

C

L

L

C (Martyrius) Jerusalem ■

c/L

C

C C

C

C L

C

C

L

C C L

C

C

Bethlehem ■ C C

C

C

C

L

L (Great Laura) C

Thekoa ■

L

L

C

Dead Sea

C

0

10

20km

The excavations of desert sites emphasise the importance of a water supply. Whether water came from rain water which was conserved and collected, or from a natural spring, a complex system of gutters, pipes, cisterns and reservoirs ensured that this scarce resource was not wasted. At the monastery of Martyrius, water was collected from the drainage of the roofs and courtyards and from the ridge west of the monastery, from there it was taken along canals to the numerous cisterns. It is estimated that 20,000m3 could be stored within the monastery. (Hirschfeld 1992: 155-6). The provision of a church, mentioned by Cyril, scarcely requires comment. A church formed an essential part of all monasteries. In this case, the discovery of a suitable cave is seen as a sign of divine favour for the settlement of the site.

L

Figure 2. The distribution of Laurae (L) and Coenobia (C) in the Palestinian desert Archaeology of the Desert Communities

Archaeological research provides further evidence for the nature of these desert communities. A thorough survey of seventy-three monastic sites was conducted by Yizhar Hirschfeld between the years 1981-1990. There are two distinct types of monastery in the Palestinian desert - the coenobium and the laura (Figure 2). The coenobium - or communal - monastery is more often found to the west of the area, especially in the hilly country around Bethlehem, where the rainfall is higher and cultivation possible. The monastery is enclosed by a wall, and monks live within it. The best preserved

The survey of the sites made it clear that there were a large number of monasteries. Furthermore they were concentrated in two main areas. The high plateau region around the Great Laura and the plain to the south of Jericho. Within these areas an average distance between monasteries is 2-3km, but in some cases they are less than 1 km away from each other. This distance contracts further when it is taken into account that the laurite cells

27

Figure 3. The monastery of Martyrius. The refectory

Refectory

_ _]

0

25m

Figure 4. Plan of the monastery of Martyrius, showing the surrounding wall, with church and refectory, and the cluster of cells and other buildings inside

28

constitutionally and administratively the cities were the cells of which the Empire was composed' (Jones 1964: iii.71 2). Only small areas of the Empire did not benefit - at that time - from the civilising effects of urban society. These were mainly in the frontier areas of Arabia and Palestine, which was precisely the area in which Sabas was building his monasteries. Thus in constructing his communities, Sabas was participating in the process of building up the Empire by colonising areas which had been previously unoccupied. The Great Laura, which is the example which has been used, was only one of seven foundations which he made. It seems from this achievement of city-building that the growth of the Palestinian monasteries cannot be understood only as an ascetic flight from the world, although we are left in no doubt that the monks loved the wildness and emptiness of the desert. it was also a process by which new cities were built, the Empire extended and the frontiers made secure. Antony's monks in Egypt were fleeing from the cities. Sabas' monks were creating them.

could be located several hundred metres distant from the central buildings of the monastery. Some of the cells of the Great Laura are closer to the neighbouring 'Coinobion to the North' than to their own monastic centre. The monasteries were connected by paths, which were often supported by substantial retaining walls. These paths 'changed the face of the desert during the Byzantine period' (Hirschfeld 1992: 205). 'The Desert a City'

We should recognise that the monasteries of the Byzantine desert formed a large and complex network of settlements. They were capable of housing about 3,000 monks. A phrase which Cyril uses twice in the extracts was to 'make the desert into a city'. This has become familiar through its use by Derwas Chitty for the title of his well-known study of monastic history - The Desert a City. It was used in the earlier Life of Antony. Here we are told that 'the desert was made into a city by the monks going out from their own dwellings and enrolling in the heavenly kingdom' (Life of Antony 14). Here the phrase is used metaphorically. The point is that the desert is NOT the city, but nevertheless monks were able to live in it (Figure 5). In the writings of Cyril, the meaning has shifted. The making of the desert into a city is a historical process by which elements essential to urban living are constructed in this environment.

This is a significant step in the history of monasticism. Other early monastic literature emphasises the departure of the saint from his home (anachoresis), the settlement in a deserted spot, then often a kind of chase-game around the desert as he tries to escape from the enthusiastic disciples who will not give him any peace. It is the ascetic discipline of the holy man which is important. But here the place matters. A physical community is constructed which settles in a defined and sacred space. It matters that the community is here and not somewhere else. In the description of the founding of the Great Laura summarised earlier the place is selected by God, not at the whim of the founder.

The urban characteristics of the monasteries, among them the Great Laura, give them a recognised place within an urban society. The historian A.H. Jones wrote, 'The Byzantine Empire was an agglomeration of cities, self-governing communities responsible for the administration of the areas which they occupied

Figure 5. Sabas' monastery. The rocky terrain of the gorge of Siloam

29

An angelic voice, a wild ass digging, a pillar of fire over a cave-church are the signs which lead the saint to initiate the stages of the construction process. Although Sabas founded further commumtles, it was inconceivable that he could desert this place given to him by God. He was a peripatetic figure who travelled often to Jerusalem and even to Constantinople on two occasions, but he always returned to this place and eventually died in it in 532.

points to a greater consciousness of divine guidance in the building of this monastery-city, it is an element which could be appropriately applied to any city, for all cities were under the rule of the deity. Further elements of the sacredness or holiness of the space can be identified. It was within the jurisdiction of the church. The role of Sabas within his communities is shown by the titles which Cyril gives to him. These include 'nomothetes' or lawgiver; 'odegos' or guide; 'prostates' or patron (a title explored by Peter Brown in his study The Rise and Function of the Holy Man in Late Antiquity). Most significantly he was 'archimandrites'. This word, which is transliterated into English as archimandrite, was used in a variety of contexts to refer to someone holding authority in monasteries, but in Palestine it was developed into a quasi-episcopal role. One archimandrite, Passarion, who lived in the 4th century, was spoken of as 'archimandrite and chorepiscop' (or rural bishop), and so clearly had extensive authority. Passarion lived in Jerusalem, while Sabas was a monk of the desert. As abbot of a leading monastery he was chosen to exercise this important role. In the diocesan organisation of the church of this period, episcopal authority was always exercised over a city, and so the use of this title confirms the impression that the monasteries were seen as together constituting a city (Binns 1994: 175-77). It was a city with no civic government but only a religious authority. It could be suggested that in this case the secular elements of the city have been removed to leave only the religious components.

To recognise the urban quality of these monastic communities might suggest that they had developed away from the primitive nature of the simple structure of a group of disciples gathered around a holy man into secular institutions, integrally connected to contemporary urban society. But this would be a misunderstanding. The cities of the classical world had a sacred dimension to them. The contrast between sacred and secular is a feature of modem society. In the period we are discussing, all human society had a religious quality. In a recent study of the classical city, E. J. Owens points out that, 'temples, shrines, and other sacred places were an integral part of the urban framework' (Owens 1991: 4). The founding of the city was often a religious act. The Romans, according to legend, followed Etruscan ritual in founding cities or extending the boundary of existing towns. First, auspices were taken, and then the perimeter of the city was marked out with a bronze plough. The furrow afforded sacred protection to the inhabitants from the gods of the underworld. At the point where the gates were to be located, the plough was lifted in order to permit access without having to cross the sacred line of protection. In the centre of the space thus designated, offerings were made and a sanctuary established (Owens 1991: 4). Temples were built at significant points, such as the high ground, which was of strategic and defensive importance, or the centre of the city. In Cyril's home town of Scythopolis, the capital of Roman Galilee, excavations revealed a temple dedicated to Zeus on the hill above the town, built around 100 CE. At some stage, probably in the fifth century, this temple was demolished and in its place a large church was built. The church is unusual in design, consisting of a circular roofed ambulatory with an apse at the east end. The length of the east-west orientation of the structure is 50.4m. This large and imposing building demonstrated that the religious allegiance of the city had become Christian (Fitzgerald 1931: 18-31). The familiar sight of a temple or a church placed either at a high point or in the central agora of the city indicates that the space is dedicated to the divine being and hence has a holy quality.

The holiness of the monastic site was further confirmed through the burial of the founder, or other holy men (Figure 6). In the case of Euthymius, whose Life was also written by Cyril of Scythopolis, a special burial place was constructed. Three months after the saint's death, a certain Fidus 'built a great and marvellous vaulted chamber. In the middle he constructed the tomb of the saint; on either side he provided tombs for other pious men .... After laying the precious remains to rest a tombstone was set in place with a crucible fixed above the heart. This crucible pours forth every kind of benefit for those who approach with faith' (Cyril 1991: 58). Derwas Chitty excavated the site at Mishor Adummim and discovered the tomb beneath the funerary chapel, at the centre of the monastery. The chamber is vaulted, as described, and is 4m in height. The central burial trough is only 1.6m in length, which concords with literary descriptions of Euthymius as a dwarf, and other troughs are arranged around it. Skeletons and clay lamps for offering were found in the chamber (Hirschfeld 1992: 137-8). Excavations of other monasteries confirm that the tomb of the founder was often located in the centre of the monastery, while other monks are buried in caves, usually to the east of the monastery. Thus the honour and respect in which the holy man was held was transferred to the place, and the tomb became a place of pilgrimage where miracles took place.

The presence of temples - and, later, churches witnesses to the desire of the inhabitants that the city should be under the protection of the deity and should be, in some sense, directed and governed by him. This element is especially clearly present in the case of the founding of the monastic-city described by Cyril of Scythopolis. Each stage of the construction is only undertaken in response to clear divine guidance. An angelic voice directs Sabas where to make his home - a wild ass shows him where to dig for water - a pillar of fire guides him to the cave-church. But although this

A place becomes holy only if people consider it to be holy. So it needs visitors and pilgrims to come with reverence valuing the place and expecting miracles to happen. There was no shortage of pilgrims visiting Jerusalem, and as a result there was growing demand for

30

Figure 6. Sabas' monastery. The tomb of St Sabas suitable locations for pious VlSltors to come to. The network of footpaths shown by archaeological surveys shows that the monastic sites were connected with centres of pilgrimage such as Jerusalem and Bethlehem. The pathway leading along the brook Kidron eastwards from the Mount of Olives was known as the 'road of the Great Laura' (Cyril 1991: 178). The reputation for holiness of the monastic sites of the Judaean desert developed along with the growth in pilgrimage. Pilgrimage was a complex social phenomenon which developed rapidly following the Christianisation of the Empire under Constantine. Al.though even a brief description of this movement is beyond the scope of this paper, it must be noted that it had political and economic - as well as religious - causes. It was related to loss of parts of the Wes tern Roman Empire to Germanic invaders, and economic dislocation which displaced people from their homes, as well as the desire to worship at places which were significant for the growth of Christianity. The result was a huge influx of persons and money into Palestine which enabled the growth of many sites, among the monasteries of the desert.

Conclusions

The growth of the monasteries of the desert shows that holiness has a history. While there may well be natural features which encourage people to value the place such as natural beauty or an awe-inspiring location - a site can only develop as a holy place through people visiting it, and these visits are brought about for many different reasons. The holy place is a part of the society which values it and cannot be separated from it. Holiness of place is brought about by a complex blend of circumstances in which the sacred and secular interact.

References

Binns, J. 1994. Ascetics and Ambassadors of Christ. Oxford: Oxford University Press. Brown, P. 1971. The Rise and Function of the Holy Man in Late Antiquity. Journal of Roman Studies 61: 80-101.

31

Hirschfeld, Y. 1990. List of the Byzantine Monasteries in the Judaean Desert. (In), Bottini, G. et al. (eds.), Christian Archaeology in the Holy Land. New Discoveries. pp.1-90. Jerusalem: Franciscan Printing Press.

Chitty, D.J. 1966. The Desert a City. Oxford: Oxford University Press. Cyril of Scythopolis. 1991. Lives of the Monks of Palestine. Kalamazoo, Michigan: Cistercian Publications.

Hirschfeld, Y. 1992. The Judaean Desert Monasteries in the Byzantine Period. Yale: Yale University Press.

Festugiere, A-J. 1963. La vie de Sabas et les tours de Syrie Palestine. Revue Biblique 70: 82-92.

Jones, A.H.M. 1964. The Later Roman Empire. Oxford: Oxford University Press.

Fitzgerald, G. 1931. Beth-Shan Excavations of 1927. The Arab and Byzantine Levels. Philadelphia: Publications of the Museum of the University of Pennsy 1vania.

Owens, E.J. 1991. The City in the Greek and Roman World. London: Routledge.

32

Bodh-Gaya: an ancient Buddhist Shrine and its Modern History (1891-1904). Nayanjot Lahiri, Department of History, University of Delhi. Introduction

modes that were selective, which excluded 'inconvenient' histories. On the other hand, this coincided with an imperial initiative which, contrary to the principle that the British government did not interfere in matters of religion, invested great prestige and energy in the dispute. This imperial vision also tried to culturally homogenise the shrine by viewing it only in terms of the moment of its birth. Finally, as the paper shall argue, Buddhist relics and shrines were used to publicise what the Indian government thought was a more cultured image of imperial rule.

This paper proposes to highlight some little known aspects about a renowned ancient shrine in order to address and contextualize its recent disputed history. The shrine is the Mahabodhi complex in the Indian state of Bihar and stands close to where Siddhartha Gautama, the historical founder of Buddhism attained enlightenment (Figurel). Since the time when a temple was first built here in commemoration of that sacred incident, with only occasional interruptions, it has remained a place of worship. However, on account of this long history, it became a religious place where a mass of heterogeneous practices and traditions, both Buddhist and Hindu, came to be inscribed.

Preliminary Formulations: Patronage, Proprietorship and Colonial Intentions

Curiously enough, the revival of Buddhist patronage in the latter part of the 19th century coincided with the beginning of a contentious politics that has since surrounded the shrine, politics that involved the reification of Mahabodhi into an exclusively Buddhist place of worship where other ritual practices, correspondingly, came to be viewed as acts of usurpation or defilement. The list of episodes, from petitions and litigations to arbitrations and confabulations, culminating in the Bodh Gaya Temple Act of 1949, as also more recent agitations relating to the composition of the temple committee attest, is a long one. This paper takes up a mere thirteen years of this century old dispute (1891-1904). These years, in hindsight, however, seem particularly important. This is because they mark the first phase in the disputed recent history of ancient Mahabodhi and also because, for the first time, the shrine's archaeological and historical elements came to be utilised for pushing forward new religious and political agendas.

Before discussing aspects of the dispute, some preliminary clarification on Indian religious shrines in general and the Mahabodhi complex in particular, may be offered. One may begin with Mahabodhi's archaeological history, in order to get a sense of the shrine's character and patronage. The first temple was built by a powerful royal patron, the Mauryan emperor Ashoka (c.3rd century BC), who also made a pilgrimage to Sambodhi (as he is supposed to have called it). Only a few remnants of that temple survive although his gift, the famous vajrasana throne has endured (Cunningham 1892: 4-7). A few centuries after Ashoka, 'gifts to the king's temple' (rajapasadachetika), inscribed on the old stone railing, were added by more ordinary donors, some of whom were women (Cunningham 1892:15-6). Along with various scenes from the life of the Buddha, this railing was adorned with figures of popular Hindu divinities including one of Siva, represented with a snake hanging down from his wrist and the trident on his left side. In explaining this, Benimadhad Barua suggested that 'the general motive behind these iconic representations was to add to the sanctity, charm and grandeur of the erection in the eyes of the people' (1934: 94). As for the temple itself, if the Chinese pilgrim Hiuen Tsang (c.7th century AD) who visited Mahabodhi, is to be believed, it went through a major phase of reconstruction which was undertaken by a high caste Hindu Brahman (Beal 1884, reprint 1969, Volume II: 119-20), a detail which is also mentioned in the 14th century AD Burmese inscription which records that, 'in In lapse of time, having fallen into disrepair, it was rebuilt by a priest named Naik-Mahanta' (Cunningham 1892: 76). What seems equally striking, as Theodore Bloch perceptively pointed out, is that neither the Buddhists who told Hiuen Tsang this story, nor the Chinese pilgrim himself, considered such a tradition either incredible or absurd (1912:149). Over time, the manner in which Buddhist patrons venerated Mahabodhi underwent change. While in the 1st century BC, Kurangi, the wife of king Indragnimitra, donated several

This connection, however, has not been suggested in the impressive literature on the shrine complex. The British initiative for resolving the dispute is also practically unknown. In fact, the explosive situation that developed in those years has most often been understood within a fairly straightforward framework, in which primary emphasis has been given to the umbrage experienced and expressed by devout Buddhists, reacting to incidents of 'desecration' by a Saiva priest and a Hindu pilgrim traffic at the shrine. My intention here is simply to suggest that the BodhGaya question is not capable of being read from such a simple perspective. Instead, it is better perceived in terms of wider, specifically 19th century, social and political situations. The paper will try to demonstrate that on the one hand, there were groups, themselves products of a Buddhist revival, who used a new panBuddhist axis to push forward an agenda in which the historical evolution of the religious complex was ignored. In this, demands were articulated through

33

MAHA-BODHt.

~-· ~~ . .i.,.,,.

@

.

...,,.,,

•

-

I

I

I

t

1

I

Figure 1. Alexander Cunningham's map of Mahabodhi and its surroundings Along with intense Buddhist veneration, the medieval centuries also witnessed other kinds of worship. Vaishnava worship is suggested both by the addition of the Gaya-mahatmya to the Vayu Purana prior to the 14th century AD ( 1940: 17), in which the Bodhi-druma

pillars, in the 7th century AD, hundreds of cows were given for ensuring that ghee (oil) lamps were regularly lit at the shrine and in the monastery (Ahir 1994: 22, 30).

34

(Bodhi tree) was mentioned as part of the Gaya pilgrim circuit (as it still is), and by the presence of Vishnu images of the 9th century AD at Bodh-Gaya (Huntington 1984: 97). An inscription of the 9th century also recorded, among other things, the setting up of a Chaumukha Mahadeva (a four-faced phallic emblem of the god Siva) within the boundaries of the temple complex (Cunningham 1892: 3, 63-4). This was set up for the benefit of snatakas (Saiva Brahmin scholars) who were 'the inhabitants of Mahabodhi' (Mahabodhinivasinam) (Barua 1934: 231). These details imply, as also does the site's sculptural wealth encompassing Buddhist images and Hindu ones (Figures 2-3) that there are multiple lineages which inform Bodh-Gaya's cultural history. This is a cultural history involving patrons and pilgrims of several types and faiths.

requires emphasis, that there was nothing unique about the fact that the legal proprietor of Mahabodhi was not a Buddhist by faith. Thirdly, the proprietorship of a Saiva mahant at Bodh Gaya did not create any problems for the conduct of Buddhist worship. As Justice Macpherson of the Calcutta High Court in 1895 put it (Macpherson and Banerjee 1895: 746), It does not appear that any hindrance was ever offered to them or that any complaints were ever made by them, and, before the occurrence in question, there is no instance of any disturbance between the Buddhist worshippers and the Hindu Mahants or their subordinates in regard to their respective rights. This fact is of some importance in the present case where each party charges the other with being the aggressor.

Secondly, Mahabodhi in the 19th century was just one of the many places enjoying the worship of people of various faiths. If here there were Buddhist worshippers paying obeisance and Hindu pilgrims offering oblations, this must have been similar to what was encountered at Dayalpur near Delhi where Siva lingas (phallic emblems of the god Siva) were worshipped in the verandah of a Sikh gurudwara (temple) (Figure 4). Hinglaj in Baluchistan is another instance and was dedicated to a goddess known as Bibi Nani by Muslims and Parbati or Kali by Hindus (Minchin 1907: 35-8). Most importantly, in the late 19th century, the proprietress of Hinglaj was a Muslim woman. This also shows and this

Even in September 1903, many years after the dispute began, Denzil Ibbetson, the Home member of Lord Curzon's Council had pointed out to the viceroy that there were no complaints on the part of Buddhist worshippers (Indian Archaeology 1899-1905: 298; IA hereafter). In fact, Anagarika Dharmapala himself, the man who was central to the Bodh-Gaya dispute, had worshipped many times before the large medieval image of the Buddha in the main shrine and, on his own admission, without oppos1t1on (Macpherson and Banerjee 1895: 744; also see Figure 5).

Figure 2. Buddha surrounded by Bodhisattvas

35

Figure 3. Images of Brahmanical deities at the Mahabodhi complex

Figure 4. Siva linga on a platfonn in the verandah of a Sikh Gurudwara (temple)

36

ruled and managed' (Said 1994: 158). In case 'natives' refused to be easily 'managed' then the state, in the name of salvaging India's archaeological heritage, could consider using legislation for taking over the management of religious shrines in order to ensure the 'proper conduct of worship'. That in this case 'proper conduct of worship' was defined not in terms of customary forms of worship but with reference to textual notions of Buddhism will be discussed subsequently.

Finally, a word about India's rulers. Many of their statements underline that a great deal of money was spent by them for the purpose of architectural restorations at Bodh-Gaya. The Bengal government is said to have spent 2 lakhs of rupees for this purpose between 1876 and 1884 (IA : 219). Thus, they believed, as Ibbetson (IA: 298) put it, that, 'the present temple owed its existence to us'. Equally important, moreover, is the fact that they were not the first rulers, to have done so, royal missions from the Siwalik area, Ceylon and Burma, the latest in 1877, had also financed restoration there (Ahir 1994: 30-1, 83-94). What was, however, new and unprecedented in the British government's initiative, was their view that by repairing Mahabodhi, they also acquired certain rights over it.

Buddhist revival and the beginnings of the dispute

Moving on to the dispute itself which began in the last decade of the last century, one of the questions which may be posed is -why these particular years? Different types of pilgrims, after all, had been coming to BodhGaya for well over a thousand years while the legal proprietorship of the Taradih village and the shrine complex was granted to a mahant in 1727. So, what is it about the 19th century which will help us to understand why the multi-religious character of Mahabodhi, with its ancient antecedents, came to be regarded as evidence of the shrine's desecration?

Lord Curzon stated this with more candour than most when he said, that the acts of restoration at Mahabodhi, 'seem to involve the gradual assertion of a co-ordinate authority, with power, if not to dispose of the shrine or to expropriate, the Mahanth, at least to superintend his superintendence and to control his control' (IA:230). So, the architectural conservation of Mahabodhi was permeated with the same ideological rationale as imperial rule - 'reconstitute the native as someone to be

Figure 5. The main image of the Buddha inside the Mahabodhi temple as it appears today. That it is clothed and gilded by Buddhists is ironic since, among other things the Maha Bodhi Society in the late 19th century actively campaigned against the mahant of Bodh-Gaya for undertaking such practices

37

In order to answer this question it is necessary to draw attention to the fact that the dispute was born in the midst of a general Buddhist revival in Asia, a revival which encompassed several regions including Burma, Siam and Ceylon. This does not mean that prior to this, Buddhism was either inactive or on the decline. 'Revival' is only used here to describe a process of change, in many cases marked by attempts to bring Buddhism into conformity with its early doctrines and also establishing links among groups of believers and sympathizers in Asia and in the West.

patron and promoter of Pali orthodoxy, and Trevor Ling has pointed out the similarities between the reforms there and those initiated in Thailand (1979: 65). In Ceylon as well, there were various Buddhist fraternities that prescribed adherence to the religion's early principles and doctrines. The point that really emerges from all this is that Dharmapala' s perception of Mahabodhi seems to spring from the general background of movements for reestablishing a 'pure' Buddhism to which can also be traced the links that he established with Buddhists of similar leanings in other Asiatic countries.

Before expanding on the character of this revival, it is perhaps necessary to point out that 19th century Buddhism at no point, not even after the impact of reform movements, could be described as a religion that was defined exclusively by an ancient canon. Thai Buddhism, for instance, was marked by the cult of the phi and the chao (the spirits of the dead), cults centering around Brahmanical gods and demons and practices relating to spirit doctors, magicians, cabbalists etc. (Ling 1979). Burmese Buddhism also encompassed a variety of religious observances, many of which enjoyed royal patronage. King Bodawpaya (1782-1819) is known to have sent two persons qualified in witchcraft, to perform magical rites at the Maha Muni temple before he attacked Arakan (Ling 1979: 49). In many other northern Asiatic countries as well, Buddhism was marked by various tantric practices as also the worship of a variety of gods and goddesses that were strikingly similar to their Brahmanical counterparts (Getty 1914). While some scholars, Austin Waddell for example, tried to suggest that this was due to the influence of Hindu tantrism, others have argued (see Govinda 1956: 362), that the Buddhist tantric system had crystallized by the end of the 3rd century AD and, thus, was earlier than its Hindu counterpart.

Additionally, the Protestant nature of reforms in these countries owed a great deal to the presence of Christian missions and to the challenge that they offered to Buddhism through education, preaching, the printing press etc. The situation in Ceylon was particularly confrontationist. Kitsiri Malalgoda has demonstrated that the Buddhist response to Christianity assumed a coherent form only in the latter half of the 19th century and that the outcome of the confrontation was the emulation of the model of Protestant Christianity by Buddhists (1976:260), Noteworthy among the resultant changes were the imitation by the Buddhists of the organizational forms, tactics and expedients of Christian missionary bodies, as also the acquisition of a new militancy, which had more in common with the spirit of Christian missionaries of the time than with their own more traditional attitudes towards other religions. In keeping with the same developments were also the process of laicization, or the involvement of laymen in roles of religious leadership, and its concomitant, the displacement of monks from some of their traditional positions of leadership.

At the same time, colonial contact here, as elsewhere, disrupted tradition patterns of belief and worship. In this specific case, the impact (or challenge?) of Christian missionaries was of special significance and led to important changes in the definition and practice of Buddhism among influential circles of believers. These developments have generally been described as 'revival' movements.

Interestingly, the most important religious leader in the Bodh-Gaya dispute, Anagarika Dharmapala, was a Ceylonese layperson, not a monk, and many of the elements outlined above - active intolerance towards unorthodox Buddhist practices and other religions, propagation of a religious agenda through publication and distribution of pamphlets and periodicals, preaching - were very clearly articulated in his person and programme.

While the details of such reforms across South and Southeast Asia cannot be summarized here, some common patterns in these revivals are worth highlighting. For one thing, there was a systematic emphasis on reordering Buddhism. This reordering involved attempts to return to orthodox practices and attitudes as contained in Pali sacred texts. Simultaneously, this also involved explaining away the historical evolution of Buddhism in the intervening centuries as the 'degeneration' of the religion. An instance in point is Thailand where Mongkut who subsequently became the king of Thailand, set up a school of monks known as the Dhammayut monks (those who adhere to the Dhamma or law of the Buddha). The purpose of this sect was mainly to purge the Sangha of what they understood as the 'nonBuddhist' and folk accretions which had come to be a part of Thai Buddhism. In Burma, during the reign of King Mindon (1852-1878), royal Buddhism became the

Finally, strides in communication, in many ways the linkage effects of the British imperial impact, greatly facilitated long distance travel and the dissemination of all kinds of information. On the one hand, these helped in expanding and intensifying the range of influence of various Buddhist groups and, on the other hand, in forging links among them and with non-Buddhist organizations who were known to be sympathetic to the religion. It is this which helps in contextualizing the support of the Theosophical Society and its founders, Colonel Henry Steel Olcott and Madame Helena Petrovna Blavatsky to the organizational work of Buddhist societies in Ceylon. It is such communication innovations which also allowed Dharmapala, in his effort to restore Mahabodhi to the Buddhist community, to address with relative ease and within a couple of years (1891-93), gatherings in Ceylon, India, Rangoon, London, Chicago and Tokyo (Barua 1981: 84-95). And,

38

it was the same railways and steamships which had facilitated Dharmapala's travels that brought a much larger influx of Buddhist pilgrims and travellers from Asian countries which had experienced revival movements to Sarnath, Bodh-Gaya and other Indian places which were associated with early Buddhism and the life of Gautama Buddha.

A reading of Dharmapala's diary also underlines this. An entry of June 30, 1893 reads (cf. Valisinha 1950: 351), Got up early morning and meditated. Yes, the land must be got. The Temple without the land is head without the body. If we get the land the Temple being public property it will always be ours.

The revival across Asia and its implications, in fact, deserve to be the subject of a separate study but, from the perspective of the present paper, it is necessary to move on to the 20th January 1891 when Anagarika Dharmapala, the scion of a wealthy Ceylonese family, first visited Bodh-Gaya. Dharmapala was no ordinary devotee worshipping at the shrine but came as a religious crusader in the mould that had become emblematic of Buddhist revivalism (see his biographical sketch in Sangharakshita 1952: 9-65). Since his twenties, he had been deeply influenced both by Megettuvatte Gunananda (1823-90), the leading Buddhist figure at Colombo and by Olcott and Blavatsky of the Theosophical Society. By the time he came to Bodh-Gaya, he had preached over large parts of Ceylon, raised funds for his faith and had been the moving force behind The Buddhist (established in 1888), an English weekly concerned with the propagation of Buddhism and which enjoyed a wide circulation.

Legal control was attempted through various ways. Initially, Dharmapala hoped to acquire the custody of Mahabodhi through a dialogue with Hem Narayan Giri, the mahant of Bodh-Gaya. This continued well into 1893 when Dharmapala was joined by Colonel Olcott (by then an honorary director of the Maha Bodhi Society) in his attempt to persuade Krishna Dayal Giri, the successor of Hem Narayan. The mahant, however, refused to either sell or give a lease. Simultaneously, other sorts of pressures were brought to bear on Mahabodhi's proprietor ranging from public appeals, to the organization by Dharmapala of an International Buddhist conference in October 1891, attended by delegates from Sri Lanka, Japan, China and Chittagong. This conference also resolved to approach the mahant with a Japanese proposal for the purchase of the temple (Barua 1981: 91). Since these attempts failed, Dharmapala decided to enshrine an image of the Buddha in the upper floor chamber of the temple. This image was procured from Japan and was 'to be for many years the storm-centre of the battle for Buddha-Gaya' (Sangharakshita 1952: 48). In 1894, Dharmapala announced that this would be placed in the temple on 19th May in the presence of the collector of Gaya. As the Calcutta High Court Proceedings record (Macpherson and Banerjee 1895: 744),

Dharmapala came to Mahabodhi inspired by Sir Edwin Arnold as well. Arnold was the author of Light of Asia, a blank-verse life of the Buddha in eight books. This, after it was published in 1879, went through at least a hundred editions and, more than any other work of the day, drew the attention of the English-speaking people to Buddhism (Almond 1988: 2). Arnold's disgust, which came to be subsequently articulated frequently by others, at what were clearly familiar sights at Mahabodhi is also barely disguised in his writings (quoted in Ahir 1994: 100): ' .... when I sojourned in Buddha Gaya a few years ago, I was grieved to see Mahratta peasants performing 'Shraddh' in such a place'. The sight of a Brahman priest of the Saivite sect, near the Bodhi tree, intoning mantras relating to the oblations offered by Hindu pilgrims to their ancestors, was equally distressing for him (Ahir 1994: 99). More importantly, in order to end such 'defiling' practices and make Mahabodhi 'a living and learned centre of purified Buddhism' (quoted in Ahir 1994: 102), Arnold had been the first to suggest to Sinhalese Buddhists and to the British authorities that the temple might be, through the intervention of the state, placed in the hands of a representative committee of the Buddhist nations (Barua 1981: 81). Right from the beginning, the aim was to transfer the custody of the main shrine of Mahabodhi from an Indian legal proprietor to Buddhists from Ceylon, Thailand and Japan.

There was no authority for the latter part of the announcement and the Mahant had not been consulted. The latter, learning of the intended installation by Dharmapala, objected to it and closed the temple down to prevent it. This he afterwards opened in compliance with an order of the District Magistrate made under section 144 of the Criminal Procedure Code. It was mentioned in the order that no image would be set up in the temple that night without the Mahant's consent, and the Magistrate at the same time sent a demi-official letter to Dharmapala, in consequence of which he desisted from placing the image.

Notwithstanding this, on 25th February 1895, Dharmapala made an attempt to enshrine the image. This was resisted; fist fights were followed by a criminal litigation which was eventually decided against Dharmapala in August 1895. According to the Calcutta High Court, following his unsuccessful endeavour to obtain a lease on the temple, Dharmapala's attempt to install an image was a covert way of doing what he had failed to do directly and openly, namely, to bring the temple under the control of Buddhist priests (Macpherson and Banerjee 1895: 751).

Inspired by this message, Dharmapala visited Bodh Gaya and later in that year, he founded in Colombo the 'Maha Bodhi Society'. One of the objects of this society was to reclaim the Mahabodhi temple for the Buddhists. While Mahabodhi had been in the possession of mahants since 1727, it had never been converted into a Hindu temple. Buddhist pilgrims had free access and liberty to worship in it. So at least one fact appears to be certain, the question was not of worship, it was of legal control.

In the meantime, from July 1894, Krishna Dayal Gir began carrying out a systematic Hindu worship of the

39

main image of the Buddha which understandably was regarded as repugnant by the Maha Bodhi Society. This was, however, an act of retaliation, a stratagem on the mahant's part to strengthen his position against the threat of Dharmapala. As J. A. Bourdillon wrote (IA: 307),

Archaeology 1899-1905, covering 120 closely printed pages. A comprehensive account of even the salient points of such a detailed correspondence seems impossible here. Therefore, only a few important issues that emerge from this viceregal endeavor will be discussed which will help in perceiving it as part of a wider pattern. The resistance offered by Krishna Dayal Gir will also be briefly touched upon.

It also appeared that since the misguided attempts of certain members of the Buddhist faith to place by force an image of Buddha in the temple and thus assert and establish a right to the building, efforts had been made by way of retaliation to push and extend the Hindu worship, thus of necessity giving pain and offence to the Buddhists.

First of all, Curzon, like the Maha Bodhi Society, regarded the Bodh-Gaya temple complex as only a Buddhist shrine. This view of Mahabodhi was premised on a classification of religious complexes in which these were perceived as belonging to only one religion and its community of believers -the religion of the founders of the shrine. The problems that this could and did create for places of worship that had been 'living' shrines for many centuries with different types of histories encoded in them, is evident from the case of Mahabodhi itself. At the same time, this is only one example of the various classificatory systems spawned by colonial regimes and which were meant to engender and solidify 'ideologies of difference' (see Loomba 1998: 68-9, 107ff.). In such classifications, in addition to the necessity of segregating the rulers and the ruled, there were detailed delineations, in terms of race and culture. These were painstakingly codified and various types of decisions came to be based on such codifications. For example, the division of Indian places of worship into 'Hindu', 'Buddhist', 'Muslim' and so on was consolidated into a reasonably watertight one. Any hybridity, heterogeneity or blurring of distinctions among these categories, in terms of worship and ritual, was generally seen as evidence of debasement and/or degeneration. Thus, the treatment of the Buddha in Hinduism as an avatara (incarnation) of Vishnu was described as 'one of the most extraordinary moral perversions of which the Brahmin mind has been guilty' (Munro in IA: 245). This kind of reading was especially common in treatises relating to ancient Indian monuments. In a masterly 'tour de force' of the historiography on the Durga temple at Aihole, Gary Michael Tartakov (1997) draws attention to the ways in which the temple's most unique features, because of such conceptual schemes, came to be perceived as somehow wrong. A relevant example is James Fergusson, for whom the apsidal plan could only be a Buddhist form. Because the Durga temple had a semicircular west end, it was, therefore, understood by him as 'an inglorious, structural version of a Buddhist caitya hall, appropriated by Brahmanical Hindus' (Tartakov 1997: 6).

Incidentally, what is practically unknown is that while the agenda of the Maha Bodhi Society since its inception was opposed to artefacts of Hindu worship in the temple complex, a worship which, in the case of the Bodhi tree, certainly goes back to the 14th century AD, if not earlier, it did formally express its desire to install there some British monarchical symbols. In 1904, Dharmapala had approached Curzon and King Edward VII with a proposal for perpetuating the memory of Queen Victoria through material additions to Maha Bodhi (Dharmapala 1904). These included erection of permanent lights at the central shrine at Bodh Gaya, the construction of a marble altar for her under the sacred Bodhi tree and, additionally, a hall for Buddhist monks was to be called 'Victoria Memorial Hall'!

A Mediating Viceroy and his Difficult Customer: Curzon and Khrishna Dayal Gir From this narrative of legal proprietors and claiming to represent the pious, let us move to the third element - the imperial government formally entered the picture when, in 1902 undertook to solve the dispute.

people look at which Curzon

The Viceroy and Mahabodhi should, logically, have stood in a significant relationship to each other since Curzon believed that the conservation of India's monuments was an elementary obligation of the British government and Mahabodhi was one of the oldest structural temples of India to have survived. Initially, when the examination of the archival papers which this research entailed was started, it was this relationship that it was hoped, would be recovered. Certainly, in the correspondence there are sections which suggest some archaeological concern on Curzon's part. He is known to have extracted from the mahant a promise to remove the Mauryan pillars from inside his monastery and get the Vajrasana throne installed on the main shrine (IA: 335-6). However, it soon became evident that Curzon's interest in and visit to Bodh-Gaya were not connected with his ambitious programme of conservation. Rather, it had everything to do with the shrine dispute and its resolution which he hoped to bring about. The inputs made by the imperial and governments can be gauged from the large letters, draft resolutions and notes that were on the Bodh-Gaya question. Of these, 72 enclosures were included by Curzon

It is this which helps in understanding Curzon's vision. Bodh-Gaya, for him, was only a Buddhist shrine, 'intended for the accommodation of Buddhist images and Buddhist worship' (IA: 228). While the oblations of Hindu devotees at the Bo tree, one of the 45 sacred spots in the Gaya tour of pilgrimage, was grudgingly acknowledged not as a traditional practice but as an 'interest', it was viewed disparagingly (/A: 228): 'this interest has developed into something approaching a right, owing to the proprietary advantages enjoyed by the Mahant'. For him, this was a site of Buddhist origin which 'since passed into Hindu hands', because of which a Hindu priest 'appropriated Buddhist offerings to a Buddhist shrine' (Curzon 1902: 3). Nowhere does he

provincial number of exchanged letters and in Indian

40

choose to refer to features connected with Bodh-Gaya's archaeology and history or to the fact that these clearly attested to the antiquity of its mixed pilgrim traffic. This is interesting and important, especially for those who have followed and admired Curzon's deep interest in Indian monuments. Clearly, this particular monument's past had to be marginalized in order to resolve its current dispute, especially since this past did not fit in with the rigid division of religious shrines as conceived by the imperial state.

gathered at the shamiana (tent) where the Piprahwa relics were formally handed over to the Royal Commissioner (Hoey 1899b: 12-14). More than anything else, in view of the extraordinary significance of the event, the failure of the British government to invite any Buddhists is worth nothing. The ashes of the bones of the Buddha for the British Raj were only instruments for diplomatic goodwill. From a heritage perspective, one could argue that these should have been preserved in or around Piprahwa. Even if the government had wished to hand them over, they could have been offered to Buddhists in British India - for instance, to the nine million Buddhists in Burma. But the king of Siam was chosen and this choice was not incidental. After all, British colonial expansion had ensured that the Buddhist kingdom of Burma was ruthlessly overpowered in 1885, King Thibaw deported and Upper Burma annexed formally to British India in January 1886. That meant that the only remaining orthodox Buddhist monarch of Southeast Asia with whom the British Raj could deal with on an equal, ruler to ruler, basis was the Thai king. So, precious archaeological relics were given away as a 'proof of the good will of Great Britain, and as a practical illustration of the sympathy of the British government towards the faiths of others.

The viceroy's understanding of the temple complex may also have been shaped by the manner in which Buddhism was recreated in Great Britain in the latter half of the last century. In his important work The British Discovery of Buddhism, Philip Almond traced the various phases through which Buddhism and 19th century discourses about it were constructed and the ways in which these were determined by Victorian culture. What is especially relevant is that by the 1850's Buddhism came to be understood as a religion that was best viewed from an analysis of its texts rather than its current practices. This textual past, incidentally, was progressively collected, translated and published by Western scholars and institutes. So, as Almond (1988:13) puts it, 'by the middle of the century, the Buddhism that existed 'out there' was beginning to be judged by a West that alone knew what Buddhism was, is and ought to be'. From this perspective of an ideal textual Buddhism, contemporary Buddhism in the East was seen as being in a general state of decay (Almond 1988:37). Curzon's views on Mahabodhi, based as they were on what it was originally, rather than on how it had evolved, fits in with this Victorian conception of the religion.

One imagines that in addressing the issue of Bodh-Gaya, Curzon was also moved by similar considerations. J. A. Bourdillon, the Lieutenant Governor of Bengal, in fact suggested to Curzon that at a time when British India's relations with China and Tibet were strained, a spontaneous concession to the Buddhists at Mahabodhi, would have a tremendous beneficial effect (IA: 303-4). In another context, shortly after Curzon's visit to BodhGaya, the political mileage that was to be gained from compromise between the viceroy and the mahant had also been highlighted by Oldham, the collector of Gaya. In suggesting that the temple be handed over by the mahant to the government, Oldham (IA: 236) wrote, 'I then pointed out what an unique occasion this was for doing an act of grace and justice. It was the month in which the proclamation of their Imperial Majesties had been celebrated with such enthusiasm, and then there was the auspicious occasion of the visit of His Majesty's representative, His Excellency the Viceroy. No fitter occasion could present itself.' That the mahant was not moved by such a consideration is another matter.

Finally, Curzon's initiative has to be understood with reference to the 'cultured' image of the British Raj that was sought to be propagated from around the turn of the century. For instance, around 1899, following the discovery of the stupa at Piprahwa on the border of Nepal, the relics of Gautama Buddha which were enshrined there were offered by the Indian government to the king of Siam, and received on his behalf with due ceremonial by Phya Sukhum, Royal Siamese Commissioner of the Ligor Circle at Gorakhpur. That this was an event which carried all the trappings of a diplomatic visit is vividly captured in the records of the Home Department. The address of W. Hoey, Commissioner of the Gorakhpur Division, underlines the diplomatic significance of the occasion ( 1899a: 14),

In any case, in his resolve to find a solution, a number of possibilities were explored - initially, as has been mentioned earlier, Curzon suggested an agreement by which the shrine would be held in trust by the government, 'The shrine should not be handed over to the Buddhists as their property, since the Government has no desire to oust one proprietor merely in order to install another, but would be held in trust by Government, who would issue regulations for the proper conduct of Buddhist worship there' (IA: 232). This scheme was premised on the creation of exclusive areas of worship in the temple complex -visitation rights but no worship by Hindus inside the temple, the western Bodhi tree reserved for Buddhists, the second Bodhi tree reserved for Hindus (IA: 232). Subsequently because Krishna Dayal Gir backed out, after promising

As an Englishman I feel that the nation to which I belong must view with satisfaction the fact that its past relations with His Majesty the King of Siam have always been friendly, and the ceremony which has brought us together today is a proof of good will and of a desire that those relations should continue to be friendly.

A special railway saloon carriage was placed at the disposal of Phya Sukhum who also stayed at Hoey' s private residence during his stay and is said to have enjoyed the annual fancy dress ball at the Gorakhpur Club with him. Most significantly, on 16th February, leading government officials and heads of departments

41

acquiescence, the Viceroy set up an informal commission, made up of Justice S. C. Mitra and Hara Prashad Shastri, to inquire into the nature of Hindu worship there and 'whether such worship as is carried on is genuine, orthodox and worthy of commendation' (IA: 308). The findings of the commissioners did not match up to the Viceroy's expectations. While they did not find any evidence of regular and continuous worship of the image of the Buddha inside the temple prior to 1894, they discovered that local people had been in the habit of making votive offerings in the main shrine much prior to the dispute and also that Hindu offerings were made at both the trees, i.e. the Mahabodhi tree at the back of the temple and at the pipal tree to the north (IA: 257-8).

create certain rights over the shrine and thereby to cause a friction and disturbance.

Conclusion In conclusion, one acknowledges with sadness, that whereas Bodh-Gaya could have been a living example of shared religious tolerances and worship, which is what its archaeological history also pointed to, its distinctive character was destroyed because of the larger politics of the parties involved -a group committed to a programme of pan-Buddhist revival, a viceroy obsessed with an imperial agenda and the retaliatory tactics of a Saiva priest. Even more sobering is the knowledge that the shrine's multiple and frequently overlapping historical strands were, at best, considered irrelevant, and at worst, were deliberately marginalized in religious agendas and political programmes.

Notwithstanding their recommendations and various draft agreements and also the threat of Curzon to take over Mahabodhi through legislation, Krishna Dayal Gir was obdurate and refused to seriously negotiate. There is no doubt that this, however, came at the end of a long process during which he did give his qualified consent on many occasions, which was quickly withdrawn every time, almost simultaneously. That this was a serious rebuff to Curzon is also not in doubt. As he put it (IA: 273-74), 'my absence from the scene did not, however, entitle the Mahanth to make a series of promises and then one after the other to break them. Such a line of action was not only discreditable to himself, it was also disrespectful and insulting to me'. There is, in fact, every reason to believe that the mahant of Bodh-Gaya did not in any honesty, notwithstanding his frequent reiteration of loyalty to the British government, attempt to seriously effect a compromise on the issue.

A case study like this also shows that disputes around religious monuments in the Indian subcontinent have a long history and consequently, these have to be understood in terms of their own particular contexts. From the particularities of some of these emerges the knowledge, and the hope, that the resolution of such disputes does not always destroy composite patronage. In another part of India, around the time that the Bodh Gaya dispute escalated, a long- standing quarrel among the Jainas and Hindu Vaishanavas was amicably resolved. The Bawangaza mandir (temple) dispute was resolved not through the judicial machinery of the British Raj but by the diwan of the local princely ruler. His solution was one in which, unlike Mahabodhi, the multiple strands of Bawangaza were recognized and strengthened.

But at the end of this story about an important phase in the recent history of Mahabodhi, what seems striking is that Krishna Dayal Gir was the only one who put forward a resistance premised on the multi-religious character of this shrine and similar shrines in India. Apart from his rhetorical statement, 'the temple and its vicinity were all Hindu, and the inmates of the place were also Hindu', he eloquently stated his reasons for being so uncompromising. As he put it, while there were other Indian temples, like Mahabodhi, where worship was carried on by men of different sects and religious faiths, nowhere had the British government desired to exercise control or superintendence as they were trying to in the case of Mahabodhi. His statement, citing examples where people of various faiths worshipped, often in their own ways, is wrote quoting (IA: 319-20),

Achnowledgements My student, Baldev Purusharth, at my request, crosschecked a number of details regarding the shrine's history and took photographs for which I am grateful. lndrani Chatterjee and Sumit Guha took the trouble of reading the paper and offering important suggestions. Karan Lahiri offered valuable advice on how to simplify the style of the paper.

References The proceedings of various government departments mentioned below were consulted in the National Archives of India, New Delhi.

... the renowned Durgah, or tomb, of Ata Saheb in Ajmer Rajputana, is daily visited by numberless Hindu and Muhammadan pilgrims from different parts of the country simultaneously; the shrine on the Observatory Hills at Darjeeling is regarded as a very sacred spot by the Hindus (of all classes) and the Buddhists, each performing their worship in accordance with their own faith; the Hindus offer sacrifices of animals too; the temple of Bodh in Nepal is a place of great sanctity to the Hindus and Buddhists, where both the sects perform their worship without the least friction ..... But the object of so-called religious crusaders is quite different from that of bona fide pilgrims, viz. to

Ahir, D. C. 1994. Buddha Gaya Through the Ages. Delhi: Sri Satguru Publications. Almond, P. C. 1988. The British Discovery of Buddhism. Cambridge: Cambridge University Press. Barua, B. M. 1934. Gaya and Buddha Gaya, Volume II. Calcutta: Indian Research Institute . Barua, D. K. n.d. Buddha Gaya Temple. Its History.

42

Buddha Gaya: Buddha Gaya Temple Management Committee.

(In), Proceedings of the Home Department. Nos. 3-20 Public -A: 12-14.

Beal, S. 1961. (reprint) Si-Yu-Ki Buddhist Records of the Western World. Delhi: Oriental Book Reprint Corporation.

Huntington, S.L. 1984. The 'Pala-Sena' Schools of Sculpture. Leiden: E.J. Brill.

Indian Archaeology 1899-1905. Printed copy in the Curzon collection in the India Office Records of the India Office Library, London.

Bloch, T. 1912. Notes on Bodh Gaya. (In), Annual Report 1908-9 of the Archaeological Survey of India. pp.139-58. Calcutta: Superintendent, Government Printing.

Ling, T. 1979. Buddhism, Imperialism and War. Burma and Thailand in Modern History. London: George Allen and Unwin.

Cunningham, A. 1892. (reprint) Mahabodhi or the Great Buddhist Temple under the Bodhi Tree at Buddha Gaya. Varanasi: Indological Book House.

Loomba, A. 1998. London: Routledge.

Curzon, G. N. 1902. Note dated 27th April. (In), Proceedings of the Home Department. 7.·- Public Deposit 3.

Colonialism/Postcolonialism.

Macpherson, Justice and Banerjee, Justice. 1895-96. [60] Criminal Revision .. Jaipal Gir and others - Petitioners versus H. Dharmapala - Opposite Party. Indian Law Reporter. [Calcutta] 22-23: 739-751.

Dharmapala, A. 1904. Letter dated 7 February to the Home Department. (In), Proceedings of the Home Department. Nos. 466/69, 1901. Public -B.

Malalgoda, K. 1976. Buddhism in Sinhalese Society 1750-1900. A Study of Religious Revival and Change. Berkeley: University of California Press.

Getty, A. 1914. The Gods of Northern Buddhism. Oxford: Clarendon Press.

Minchin, C. F. 1907. Las Bela. Baluchistan District Gazetteer Series, Volume VIII. Allahabad.

Govinda, A. 1956. Principles of Tantric Buddhism. (In), Bapat, P.V. (ed.), 2500 years of Buddhism. pp.358-75. Delhi: Publications Division.

Said, E. 1994. Culture and Imperialism. Vintage: London.

Hazra, R.C. 1940. Studies in the Puranic Records on Hindu Rites and Customs. Dacca: The University of Dacca.

Sangharakshita, B. 1952. Anagarika Dharmapala. A Biographical Sketch. (In), Maha Bodhi Society of India. Diamond Jubilee Souvenir 1891-1951. Calcutta: Maha Bodhi Society.

Hoey, W. 1899a. Address by the Commissioner, Gorakhpur Division, to the Royal Commissioner of Siam. (In), Proceedings of the Home Department. Nos. 3-20. Public -A: 14-15.

Tartakov,G. M. 1997. The Durga Temple at Aihole. A Historiographical Study. Delhi: Oxford University Press.

Hoey, W. 1899b. Extract from the Commissioner, Gorakhpur Division, to the Chief Secretary to the Government of the North-Western Provinces and Oudh.

Valasinha, D. 1950. Diary Leaves of the late Ven. Anagarika Dharmapala. The Mahabodhi 58 (10):34-51.

43

Ayodhya and the Ethics of Archaeology. Nandini Rao, Madras. Thus we see the birth of the 'spiritual east', the 'bigoted Muslim', the 'tolerant Hindu' and so on. There were, of course, echoes of this same exercise among some groups in India, but I will not be discussing those here.

Introduction

Although scholars in the West are increasingly accepting that the socio-political context of the practice of a discipline affects it, there are many in India who do not. Nor do they accept that the past is created from and in the image of the present although Sudipto Kaviraj said over ten years ago with regard to history that,

Once and for all, certain temperaments, behaviours, customs, places and monuments, became identified with certain categories. The earlier porous boundaries that existed between groups and communities became defined and static.

[It] was not a purely academic but a pragmatic science; its task was not to leave the world as it found it. The past was an image created in the interest of the present. History [and the past] was not what happened, but what people could be made to believe [had] happened. (1988: 4)

These classifications and categories were the building blocks of British colonial policy - 'Divide and Rule', the basis of its imperial prowess. The categories provided colonial officers with easily digestible capsules on the colony and its inhabitants and helped them understand and control the subject people through the reiteration of long-standing but until then, shifting zones of conflict. The strong boundaries between the categories were to create divisions that would go far deeper than the British could ever have hoped.

In this paper I hope to discuss the impact of colonialism on the Indian past and on the practice of history and archaeology, as well as on how these have created deep divisions (religious and other) between people and groups, leading to violence and bloodshed. Like anthropology, archaeology developed in the West during the period of colonial expansion, when colonisers found it necessary to make sense of the 'other' people they encountered, and their apparently incomprehensible cultures. In India too, mainstream archaeology, anthropology and history as they are practiced by and large even today, are the outcome of the ·colonial encounter.

Types of reconstruction

Based on these categories or summaries of the diverse cultures and peoples of the subcontinent, the British proceeded to map the definitive past of their subjects. According to Kaviraj, they 'construct[ed] an image of a subject people whose whole history destined them for British conquest just as by their [i.e., the Europeans] telling their own story of Europe, they created a myth of what they were, destined to rule the world, through a somewhat long preparation since Greek times' (1988:4).

For effective governance, it became essential for colonial governments to map the past and the present of their colonies in social as well as territorial terms. But in order to do so, the British had also to make sense of the apparent chaos that overwhelmed them on their arrival in the subcontinent. And so their quest began to locate the authentic, to discover identities and locate origins and roots and this was to have important consequences. This quest led them to construct catch-all categories, classes and generalisations about virtually everything - places, monuments, cultures, religions, peoples, temperaments and inter-community relationships among hundreds of other things. These categories, though vast and somewhat hazy, served colonial purposes and resulted in the construction of deep divisions between categories and an obliteration of differences within them. Such categorisation also led to the stagnation of the categories and their contents. Dynamism and change were excluded once and for all.

As Thapar (1981) shows, the colonial recreation of our past reflected the concerns of British intellectuals at the time of colonial expansion and consolidation predominantly, of the Orientalists and the Utilitarians. Both contributed to the reconstruction of the Indian past in different ways, although the impact of the reconstructions was surprisingly similar. The Orientalists, on the look-out for Utopias beyond a rapidly industrialising Europe, attributed a 'spiritual' nature to the Hindus. They looked for evidence of 'Aryan' elements and for sites connected with the ambiguous ancient texts, the epics and other sacred and non-sacred literature. The Utilitarians, on the other hand, provided us with a past that 'befitted' a colonised people. Their contempt was scarcely disguised, John Mills, in his 'History of British India' arbitrarily split our past into the 'Hindu', 'Muslim', and significantly, 'British', not 'Christian' periods (Thapar 1981). He was contemptuous of what he called ancient 'Hindu Culture' which he considered retrogressive. He was more sympathetic of 'Muslim culture'. But the Utilitarians also transported the prevailing tension in Europe between Protestants and

The categories had necessarily to be created from generalisations which flattened out ambivalence. They were forcibly made coherent and were then given legitimacy. 'Chaos' had been tamed into apparently orderly, comprehensible compartments. These soon provided the framework within which both the colonials and the colonised were viewed and viewed themselves. These manageable categories made it easier to stereotype people, situations and interactions and so control them.

44

Hindus shared this, they formed a nation (Golwalkar in We or Our Nationhood Defined). Their belief was vindicated to a great extent by the creation of Muslim Pakistan.

Catholics to India and superimposed it on the Hindus and Muslims. Each group was seen as distinct and monolithic; differences within the categories were elided. Zones of conflict were reiterated leaving groups no space to revitalise the zones of cooperation that had existed alongside. Both the Orientalists and the Utilitarians had constructed, subtly, a very communal past.

Post colonial archaeology was (and has remained) clearly focused on the incorporation of science into the discipline, on scientific digging and on perfecting methodologies, complementing the emphasis of the nascent state on science as the avenue to progress. Archaeology played a crucial role in producing scientifically excavated evidence from a 'Pre-Muslim' past for a 'Hindu' identity. The Oriental legacy of locating places mentioned in ancient texts continued in Independent India and led to a curious blend; archaeology was used to provide a scientific basis for myths and legends.

Reconstructions of the National Movement During the struggle for freedom, Indians discovered the need to 'own' their past and began to reconstruct it from a 'nationalist' perspective. But they took as their point of departure, the colonial reconstructions. The framework of the discourse was thus colonial, and so what was produced was a past still distorted in many important ways, but one that was more acceptable to a nation in the making. Mill's arbitrary periodisation was retained and a result was the re-emphasising of the colonial categories and classifications.

It is suggested that the Babri masjid/Ramjanmabhumi

conflict is a direct outcome of these factors: the colonial categorisation and classification of cultures and an obsession with identifying· the authentic. These, it is suggested, led to the writing of revenge histories and the attempt to identify India as Hindu.

The legitimising of these categories led in the realm of politics, to the creation of separate electorates for different communities and the logical conclusion of this was of course, Partition, first into two and then three countries. In the academic world and especially the areas of popular history and archaeology, it led to polarisation and the creation of what have been termed 'revenge histories' (Trigger 1984). In the realm of religion, it has led to a reiteration of caste and sectarian or communal differences that continually simmer below the surface, frequently spilling out into carnage.

The Creation of a Hindu identity and a Hindu India Although many were sceptical of the one religionculture-nation theory put forward by the Hindu Right, the post-Independence revitalising of Indian identity largely become the revitalising of what was seen as a Hindu identity, (It might be pertinent to point out here though that according to Pannikkar [1997] before the eleventh century AD, there were no Hindus but, 'only a large number of sects, none of which were either known as, or called themselves, Hindu'. The creation of a Hindu identity was therefore a conscious effort).

Post colonial reconstructions Since Independence and partition, some of the more progressive historians have attempted to remove the manifestly communal overtones in Mill's periodisation but this did not help very much, especially because the country had already been rent apart into Muslim Pakistan and what is now being increasingly identified as a 'Hindu' India. The creation of Pakistan also obfuscated the differences and fragmentation between Muslims and it appears that, there too, the use of colonial categories has continued.

In rewriting a past more in consonance with new national aspirations, there was an emphasis on the period of history in which a form of Hinduism was established (AD 300-700). Both of these required the using of historical and archaeological data and these data were made available. Archaeology was also used to try to authenticate ancient literature, seen today as 'Hindu' - the epics (the Ramayana and the Mahabharata ), legends, and so on (see paper by Chakrabarti, this volume).

One of the results of this in India, was the reemphasising of categories created in the colonial period. The Orientalist view of a glorious 'Pre-Muslim' or 'Hindu' past was built upon, and gained legitimacy.

Subsequently, the epics, were also serialised several times on television and symbols from them and their televised versions were deployed in the service of the state during elections and for electoral propaganda. Within this broader context, the attempt made by some to break free from the tyranny of colonial categories has unfortunately not been successful. They have been unable to replace the Orientalist construction which remains firmly entrenched in popular imagination, freely interwoven with nationalist symbols including revenge histories. These post colonial attempts to break free have been unable to withstand the onslaught of the stronger fundamentalist reconstructions which have focused on what was called the 'humiliation' of 'all Hindus' and which began to be identified increasingly with the Babri Masjid and its location at Ayodhya.

Numerous reversionisms had sprouted during the freedom movement, revealing the fragmentation inherent within Hinduism. However, only two of them 'succeeded' in any measure, the Gandhian and the Hindu Right. Neither gained much leverage until recently although revenge histories continued to be written by Rightist groups and soon, Muslims were viewed by some Hindus as 'foreigners'. Hindu fundamentalist groups began claiming Hindustan for Hindus, defending this claim by appealing finally to the 'logic' of numbers. The ideologues of the Hindu Right held that culture, religion and nationality were synonymous and since

45

Muslim groups point out that Babar's diary does mention the building of a mosque, but not the destruction of a temple. Surely, they suggest, if Babar was as bigoted as the Hindu groups suggest he was, he ought to have mentioned the destruction of a temple. Other writers, contemporaries of Babar and his grandson, Akbar, do not mention the destruction either.

The Ayodhya Controversy

Briefly, the controversy has arisen because certain Hindu and Muslim groups have both been claiming a particular space as theirs by historic right. The Ayodhya issue, however, is much more than about the destruction of a shrine and the demand to build another of another faith in its place. It is an issue that includes more than archaeology and religion. It is also about the several thousand people killed over it; about the riots and bombs that have exploded across the country; the pogrom unleashed against Muslims in places like Bombay and it is also about the collapse of a Government. But archaeology and history have been implicated in it and therefore, it is necessary to examine this.

An examination of the archaeological evidence is similarly unrewarding. The original excavator of the area near the mosque, B. B. Lal, claimed in his first published report, that he had discovered nothing of significance (with regard to the temple). The earliest Jain figurine had been found, some crude pottery and little else (Lal 1976-77: 44-45). Fourteen years later though, Lal claimed that evidence of a temple had in fact, been found. Evidence of brick pillar bases that are suggested to have provided support to a multi-pillared temple was suddenly revealed. It is not necessary to repeat the discussion of the evidence here, but it is worthwhile stating that the archaeological community has not entirely accepted this 'evidence', but few have denounced Lal's statement and many have said the issue is one of 'faith'.

The shrine is a mosque in the north Indian city of Ayodhya in which Muslims offered prayers from the mid-1500s to 1949, erected during the reign of the first Mughal emperor Babar, which is claimed by some Hindu groups to have been built on the site of the birth of their god and king Rama. They contend that Babar ordered the destruction of a temple that stood on the site, and the construction of the Babri Masjid on its ruins. Hindu and Muslim groups have both used archaeological and textual evidence, 'religion', myths and legends and have appealed to 'emotion' as well, to strengthen, validate and legitimise their claims.

History and archaeology are being used differently by different groups to lay claim to the 'sacred space', they are being re-written to reflect the needs of today's groups. Hindu groups have simultaneously also appealed to religion and miracles insisting that whether or not, archaeological evidence is found, they will not give up their claim to the space as it is to them, a matter of religious faith. To defend this position, they maintain that a 'miracle' occurred at the site in 1949. At the climax of a nine-day vigil held outside the mosque, on the night of 22/23 December, idols of Rama and his companions, Sita, Lakshmana and Hanuman appeared inside the mosque. The Muslim groups perceive the 'miracle' completely differently and use police records to support their view. The report states that a group broke into the mosque and installed the idols within it.

Use of history, archaeology, emotion and religion

The Hindu groups maintain that according to one of the legends in the Ramayana, Rama was born in and ruled from Ayodhya. He is believed by them to be an incarnation of Vishnu, one of the Hindu Trinity. However, it must be mentioned that there are many Ramayanas such as those attributed to Valmiki, Tulsidas, Kritivas, Namdev, Kalidas and Kamban; there are the Buddhist, Thai, Cambodian and Indonesian Ramayanas, as also women's Ramayanas and Ramayanas in the oral traditions of groups referred to as tribal or Adivasi. Not all of these position Rama's Ayodhya where the contemporary one in north India is located today. Each claims a legitimacy. Is there, indeed can there be a legitimate, an authentic one?

In their appeal to religion to justify their claim to the site, some Hindu groups have appointed themselves the spokespersons of Hinduism, arrogating to themselves the right to represent 'all Hindus' and defining what it means to be Hindu with the help of violent and incendiary slogans. They are also simultaneously redefining the character of the god Rama from, as Kapur has pointed out incisively, deity to crusader (1993). In their redefinition, the Hindu groups are once again resorting to the use of simplified categories that permit neither nuances nor differences within them, categories that mask complexities. The many sects that were taken loosely to comprise Hinduism are found to be too full of contradictions for comfort and so, a further flattening out, of complexities and ambivalence is occurring. More disturbingly, a resurgence of militancy is also witnessed.

Hindu groups insist that Babar demolished a temple that existed on Rama's birth place to build the mosque. They bolster their claim with recourse to various colonial documents (specifically Carnegie 1870, Nevill 1905 and Beveridge 1921) and by appealing to the sentiments of 'all Hindus' discussed below. The Muslim groups too, quote textual, historical and archaeological data to prove that no temple existed on the spot. The authors of the three colonial documents that the Hindu groups present as evidence hold the view that since Babar was a good Muslim and as the duty (as they interpreted it) of a good Muslim was or is to build mosques, Babar must have destroyed a temple and built a mosque on its ruins. None of them investigated the validity of the conclusion and, in fact, their documents reveal less about Babar and more about their attitude towards Muslims (for a discussion, see Gopal 1991).

Ambiguous Hinduism permitted within it apparently contradictory philosophies and until its latter day selfappointed guardians took it upon themselves to 'liberate' Rama from the 'prison' of the mosque, permitted also the belief that for prayer, neither temple nor deity were essential.

46

justification for the demolition of a sixteenth century monument as part of a programme to avenge what some perceive as the humiliation of a fragmented group being pulled together and called Hindu. It is important to note that the spokespersons of the 'Hindu' groups making these demands have neither the religious nor moral authority to speak on behalf of all Hindus. Nor are they the legal heirs to the space or a temple, should one be found to have existed. These groups demanded the 'return' of the sacred space and indeed took it by force in 1992. But Ayodhya, like so many other sacred spaces, represents a palimpsest of religious interests, being sacred to Jains, Buddhists, Shaivas, Vaishnavas and Muslims. It cannot belong to one group alone. The space is, and the sixteenth century mosque was, the heritage of us all.

Reclaiming of a sacred space

Perhaps the encounter with unambiguous Islam, with its focal point defined at Mecca required an answer from ambiguous, multivocal Hinduism and in the postcolonial era, Ayodhya increasingly appears to be identified in the minds of many who claim to be Hindu, with the absolute position that Mecca appears to have in the Muslim. Ayodhya as the birthplace of Rama, was until recently, a matter of complete indifference to most Hindus (as it represents no continuity in the various traditions of, for instance, the northeast or south). But it has successfully been converted into an important religious symbol reaching down from the past to the present through tradition. It has become a crucial means of expressing the new Hindu identity and has ensured that although there appears to be no evidence of a temple or of a Rama cult in the region prior to the eighteenth century, the call to destroy the mosque served to unite warring, disparate Hindu factions, albeit superficially.

References

Militant groups used various symbols to unify these factions; of Rama as a child imprisoned in the mosque, of Rama as avenger prepared to do battle for his space and finally, through the collection and consecration of shilas or bricks that were sent from every village in India to be used in the construction of a new temple at the site. The results were catastrophic; riots broke out in several parts of the country, thousands were killed, a government collapsed, and finally, in December 1992, the mosque was demolished by a mob of 200,000 people who had gathered at Ayodhya ostensibly to demand the peaceful 'return' of their sacred space.

Beveridge, A. 1921. (Trans. 1997) Babur-Nama (Memoirs of Babur) Zahirud-din Muhammad Babur Padshah Ghazi, New Delhi: DK Publishers. Carnegie, P. 1870. A Historical Sketch of Tahsil Fyzabad, Zilla Fyzabad including Haveli Oudh and Pachhimrath, with the Old Capitals Ajudhia and Fyzabad, Lucknow: Government of India. Gopal, S. (ed.) 1991. Anatomy of a Confrontation: The Babri Masjicf-.Ram Janmabhumi Issue. New Delhi: Viking.

One of the things that has, it appears, been overlooked, is that Hinduism both has and does not have the tradition of sacred space. While some places are considered sacred, there is the equally legitimate view that those same places are not any more sacred than any other, as all earth is Prithvi and therefore sacred. There is also the view that Rama cannot be reduced to one spot on earth. Locating Rama and restricting him to the Babri Masjid site is the result, I would contend, of a misreading of plural, multivocal, Hinduism. Indeed, there is a wellknown saying, Avadh tahan jahan Ram ivasu, or wherever Rama resides, is Ayodhya.

Kapur, A. 1993. Deity to Crusader: The Changing Iconography of Ram, (In), Pandey, G. (ed.), Hindus and Others: The Question of Identity in India Today. pp.76109. New Delhi: Viking. Kaviraj, S. 1988. Imaginary History, Occasional Papers on History and Society (Second Series). New Delhi: Nehru Memorial Museum and Library. Lal, B.B. 1976-77. Indian Archaeology-A Review. New Delhi: Archaeological Survey of India. Nevill, H.R. 1905. District Gazetteer of Fyzabad. Allahabad: Government of India.

Conclusion

Panikkar, K.K. 1997. Communal Threat, Secular Challenge. Madras: Earthworm Books.

The question of Ayodhya and the sacred space within it is really more about ethics and the practice of archaeology in the contemporary polarised world, than about religion. As archaeologists we must ask ourselves whether the discipline can be allowed to be abused in such issues. It is necessary to ask whether the evidence of a temple if found, would provide sufficient

Thapar, R. 1981. Communalism and the Writing of Ancient Indian History, (In), by Chandra, B., Mukhia, H. and Thapar, R. (eds.), Communalism and the Writing of Indian History. New Delhi: Peoples Publishing House.

47

Archaeology, dance and religion in Java: the Prambanan complex. Alessandra Iyer, Wolfson College, University of Oxford. meaning 'to dance', also means 'uniting with the Supreme Being' or 'achieving Emptiness'. In the 14th century Old Javanese work kakawin Sutasoma it is explained that ~iva and Buddha are the same: 'bhinneka tunggal ika' (Sut 139, 5).

Introduction

In this paper the disciplines of dance, archaeology and religion are brought together and connected, through a case study whose focus is the Hindu temple complex of Prambanan, in Central Java. In so doing, the disciplinary boundaries are dissolved and redefined. At one end of the spectrum we have archaeology as a method of enquiry, as a way of probing into the past, interpreting and reconstructing, trying to recreate the past in the present on the basis of its material culture. At the other end there is religion, the particular systems of faith, worship and beliefs which exist and have existed among various groups of people throughout the ages. In between the two there is dance, the most ephemeral of human activities, obsolete within seconds of being performed and centred on the use of the human body as a means of expression.

Javanese culture is often described as a synchretism of Hinduism, Buddhism, Islam and local animism. This kind of perspective has had the unfortunate result of creating a 'foreign/indigenous' opposition and is representative of an increasingly outmoded way of thinking, as will be highlighted later in the paper. What should be noted at this juncture is the connection which exists between dance and spirituality in the living culture of Java. Mataya in Javanese culture is dance as a batin activity. The division between batin 'inner' and lair 'outer' runs through the whole of the culture. Batin refers to inner spirituality. This aspect of dance as ritual and its spiritual dimension has been extensively discussed by Benedictus Suharto, a Javanese dancer from a kraton or 'court' tradition (1990).

The link between dance and religion is known. In a great many cultures dance finds sustenance in religion and is ancillary to it, it provides a system of discipline and body training aimed at transcending everyday reality to enter a more spiritual realm, it is a form of prayer/worship and is also a means to transmit specific religious concepts, ideas and teachings, without verbalising them.

Suharto points out that within Javanese culture dance functions as entertainment and as individual/collective ritual; the two are not mutually exclusive, even though the performer may not be necessarily aware of the spiritual dimension of his/her dance. The ritual aspect of the dance is not always determined by its function, which differs, but on the whole it is the activity of dancing which is intended as a means to attain a psychophysical union with the divine. In the words of Suharto, the 'Javanese dancer must go through two kinds of transformation ...One is an outer (physical) transformation that includes wirama and wiraga [viz. 'rhythm' and 'body movement']. The other is an inner, spiritual transformation in a conscious state. Both kinds of transformations are not exclusive of the other' (Suharto 1990: 18). Javanese dancers are traditionally taught that to become good performers their personality must not show through: they must achieve a state of emptiness, that is, of taya.

The link between archaeology and dance is less easily perceived but their converging concerns and ways of working will soon be appreciated. Dance is forgotten as soon as it is performed and has to rely on a variety of systems of retrieval from memory for recreation of the dance action in the present. Dance and archaeology focus on the past and its reinterpretation. Dance and archaeology deal with time and sequence, as a way of rearranging the past to present it, and this is in itself performance. When the dancer and the archaeologist come together to engage in an archaeology of dance (or dance archaeology) their concerns merge: reconstructing the past from its surviving material culture becomes a retrieval and reconstitution of dance performance in every sense of the word.

Without dwelling further on living Javanese court dance culture, one may note the presence of certain key concepts which can help to understand the tradition and make sense of its past. It is these concepts that provide a starting point for the investigation presented ·in this paper.

Point of departure: the living Javanese dance culture

Taya is a Javanese noun which means 'dance'. As a word for dance, however, taya is never used in connection with the name of a specific dance or style or movement. It denotes dance at its most abstract and in a non-specific sense. Its active verb form mataya means 'to dance'. Sastra is a noun and can be rendered in English as 'knowledge'. However, taya entered the modem Javanese vocabulary from Old Javanese (Kawi), in which it is a noun used to denote the Supreme Being and also Nothingness or Emptiness. It can refer to the Hindu god Siva but it can also refer to the Buddha (Soeharto: 1990:10-11). Thus mataya, apart from

Knowledge of dance in ancient Java: archaeological evidence from the Prambanan temple complex

On the surface we do not know much about dance in ancient Java. Many people will deny even the possibility of knowing anything at all because 'dance is ephemeral' and 'we have no records'. One wonders what the assumption behind such statements is. What kind of records are we talking about? Films,

48

Dieng

•

l 0

50

100

150 200

Borobodur /j. Merapi • eSolo e•PRAMBANAN

250 JOOlan

Figure 1. The Island of Java dance, such as the sacred ritual dance of the kraton, known as bedhoyo, to which Suharto referred when writing his remarks on taya .

photographs, video-recordings, notated scores? If this is the case, then the answer is definitely that we have no such records. On the other hand, we do have other kinds of records, which might be useful in providing the raw material for an interpretation and a reconstruction. There is not only one possible interpretation, there are many. The criterion used is that of plausibility, for 'the interpretative encounter is inherently non-completable' (Tilley 1993: 3).

It is perhaps with some effort that one accommodates the bedhoyo tradition within the present framework of Islamic mysticism, even though it is within this Islamic culture that bedhoyo has been nurtured as a the supreme form of Javanese palace dancing. Islam has been the predominant religion of Indonesia from the 15th century onwards, but this did not significantly alter the function of dance and music in the culture of the previously Hindu/Buddhist Javanese courts. Becker suggests that the bedhoyo has very strong Hindu /Buddhist Tantric links (Becker 1993: 116-40) but these were dissolved and resolved into Javanese mysticism, the batin-lair culture, which has no specific formal religious affiliation, though it could be linked with Sufism.

There is in fact much data available on dance in ancient Java. Apart from references in Old Javanese literary texts, the walls of candi-s (temples) are covered with images of gods, demi-gods and humans shown in the act of dancing or in a dance-related posture. The reliefs of the 8th century candi Borobudur, a Buddhist temple in Central Java, show different scenes with dancers (Figure 2). Some are court scenes, others show dance in a ritual context e.g. dancers near a Buddhist strlpa. Such images were noted and it was acknowledged that dance had a role to play in ritualistic activities, as a kind of offering, or that it was an entertaining past time of some sort, in all social strata (in a court environ, as a street performance). These are aspects of dance known in many cultures in different parts of the world, including the West and thus they are easily recognisable when encountered in visual representations.

Prambanan is a village, about 15 Kms from Yogyakarta. A large Saivite complex is found here, made up of a series of concentric courts which contained some 224 minor surrounding temples and 8 temples in the inner court. On the outer balustrade of the tallest temple of the inner court, candi Siva - and only on this temple - there are, as said earlier, 62 reliefs showing dancers and musicians (Figures 4 and 5). That the reliefs show dance scenes can be ascertained through a mere glance. The question is what kind of dance is represented and why.

From this perspective it is also understandable why, when faced with the dance images from the Prambanan temple complex (Figure 3), also in Central Java (see Figure 1), thought to have been built in the early part of the 9th century AD, scholars did not deem them to be worthy of much attention, classing them as representations of dance with a decorative function. However, the 62 dance reliefs from the ~iva temple, the largest and tallest of the complex, more popularly known as candi Loro Jonggrang (a legendary local princess) have much to tell us about dance in ancient Java. They highlight the link between dance and spirituality which one encounters in the living traditions of Javanese

The significance of the presence of these reliefs cannot be understood fully unless the mystery of what they represent is first unravelled. This requires a technical analysis of each relief which is extremely laborious and time consuming. Briefly, it is possible to identify each one of the dance reliefs as a representation of a karar;za movement. A karal)a is a small movement sequence which forms the basic unit of an obsolete technique of dance described in the Sanskrit text Natyasllstra, the foremost work on dance, music and drama known from

49

Figure 2. Dance scene from candi Borobudur, central Java

Figure 3. View of the Prambanan complex, central Java India, the date of which is unfortunately not fixed with any certainty, though the general consensus is that it cannot be later than the 6th century AD. This is not the first time it is suggested that the dance reliefs at

Prambanan are karava -s (or kara7Jarelated) but it is the first time that the identification is based on movement analysis. Such an identification would not have been possible without previous research on the kararia-s

50

Figure4. Candi Siva, Prambanan, central Java

Figure 5. View of the dance reliefs, candi Siva, Prambanan, central Java

51

seen on Indian temples carried out by a number of other scholars and their reconstruction of the kara'fla-s based on readings of the Natya§listra.

and one female, representing Devf, originally prepared by Tirujnanam. P2 Prenkholitam (K84)

It was held until recently that the dance reliefs of Prambanan could not be representations of kara1J,a-s because no relief is exactly identical with an existing relief seen in Indian temples with identified kara~a representations. However, those who thought that identity of representation was the criterion that would allow any matching of kara1Ja-s, viewed the karm;a-s as static poses, without realising that each kara'l}a is a movement and therefore any portion of that movement can be visually rendered. This flexibility allows for multiple visual interpretations which can greatly differ from one another because different sections of a particular movement sequence may be focussed upon by the artists. But if the whole movement sequence of each kara'(Ja is known, it is not difficult to relate to each other the different portions seen in different sculptures.

Reconstruction of the karana movement:

The execution of K84 can be reconstructed as follows, using the textual description found in the Natyasastra (NS 4.144-5) and the evidence from the South Indian temples at Chidambaram and Kumbakonam. Tanjore's kara"f},areliefs stop with K81, so no information on this particular kara'(Ja can be obtained from any of the Tanjore reliefs. The textual definition says that the dancer swings one leg in dolapada cari and ends by placing the foot in aficita on the ground. This is a leg movement which literally indicates an oscillation of the entire leg (dolapada cliri) at the end of which the foot is placed on the ground on its heel (aficita). This oscillation gives the dancer momentum and she performs a jump with a tum in mid-air led by the other leg extended in sucii a leg action which involves a full extension of the leg and foot. She lands with the toes in agratala, a position of the foot in which the weight is on the toes, with the tip of big toe in contact with the ground. The hands are first in dola, at their own side in a relaxed position, neither low nor high, then while jumping one is near the chest, the other performs a swift udvestits movement, that is, a rotation of the hand on its wrist led by the index finger. Both hands then stretch up above the head and perform a rotation. At the moment of landing, one arm is stretched across the chest to the other side in dola, the other hand is in mu§{i (a gesture of the hand in which the fingers are brought in contact with the palm, rolled inwards and the thumb placed over them to form a fist) under the elbow of the dola arm/hand

Analysis of the reliefs The following are examples of the analysis carried out in order to arrive at an identification of the movements shown in all the reliefs. Relief P2, P3, PS, have been selected, the P here standing for Prambanan, corresponding to K84, K9,K32. K refers to the karana in the NllJYasastra list. In most reliefs we find animation, shown through three figures which indicate different portions of the same kara"/Ja. Sometimes the animation works from left to right, sometimes from right to left. The most important figure, both iconographically and in terms of movement, is always the one in the centre foreground. The reliefs are being compared to the authoritative kara"IJ,a reconstruction devised by Subrahmanyam (1978), shown through matching drawings, with two figures, one male, representing Siva

Figure 6. K84. (courtesy of Dr. Padma Subrahmanyam)

52

Figure 7. Relief P2, candi Siva, Prambanan, central Java

Comparison of the relief with the reconstructed movement

the hand in mu~[i under the elbow. The drawing shows the action of reeling in the air prior to the landing, which is shown instead by the main figure in the relief. This is how the figures in the reconstruction drawing and the figures in the reliefs show a relationship with each other, in terms of movement sequence.

This relief shows no animation. The subsidiary figures are simply accompanying the main dancer. Figure 1 (left) is playing the drum with his right hand and with his left hand in kha/akamukha hasta he appears to be holding the drum. The kha[akamukha hasta is a gesture performed by bringing together the tip of the index finger and that of the thumb, with the little finger held straight and pointing down wards, the ring finger bent and pointing forward and the middle finger bent to touch the base of the thumb. The plastering of figure 2 leaves much to be desired. Figure 3 (right) has a lifted hand as though ready to beat a drum, the outline of which can be guessed. The thin hand joined to this figure does not seem to be appropriate to the figure's size. Did it belong to another relief? The central figure shows the end of K84, the landing with feet in agratala and in a slight svastika, a position which indicates a crossing of the limbs, one arm stretched across the chest, the other with

P3 Nikuttakam (K9) Reconstruction of the karana movement:

The following reconstruction of the movement is in accordance with NS 4, 69-70 and evidence from the South Indian temples of Tanjore, Kumbakonam and Chidambaram. The dancer performs udgha!(ita of the foot, that is, the heel is moved up and down. The body is in dvibhanga (a 'two -bend' position, in evidence if one looks at the silhouette) and the weight is alternately transferred. One hand does pallava nritta hasta in

53

Figure 8. K9. (courtesy Dr. Padma Subrahmanaya)

Figure 9. Relief P3, candi Siva, Prambanan, central Java

54

Figure 10. K.32. (courtesy of Dr. Padma Subramanyam) sukatu"f)cj.a, the other performs an aviddhavakra movement in lata. The first movement is performed by raising the arms above the head and lowering them with the hands performing an inward and then outward rotation. The index and ring fingers of both hands are bent. The second movement involves slowly lowering the arms on their own sides, in imitation of waves and placing them in a relaxed position in low level at their own sides. A recaka (small movement done for the purpose of embellishing) is recommended Comparison movement

of the

relief

with

the

agratala/kuiicita. The latter position is one in which the foot is perfectly arched and the toes are bent in (Figure 10). Comparison movement

of the

relief

with

the

reconstructed

Figure 1 holds a drum, figure 3 holds cymbals in the lifted hand, figure 2 (centre foreground) is seen while ending the reeling action typical of this kara,;;a. Note the perfect correspondence with the female figure in the drawing. Here she is seen ending the turn on the opposite side. (Figure 11)

reconstructed

Compare the figure in this relief with the male figure in the drawing (Figure 9). The apaviddha hand of the definition of this hand given in the Natya!iistra (NS 9,220) and used for this karaoa is interpreted here as uru hasta (viz. on the thigh), whereas the drawing shows a do/a (usually an apaviddha movement indicates a movement away from the chest). Figures 1 and 3 (background) are badly damaged. No animation can be shown between the three figures of this relief.

Prambanan: purpose of the sequence and the politics of South and Southeast Asian interaction

The question 'what dance is shown in the reliefs?' can thus be answered, in what at present seems to be a very plausible way, in the sense that we now know what dance technique is represented. The process of reconstruction, which took a long time, as each relief had to be studied in detail, is still open to further interpretation and is not posited as final. The reconstruction of the dance technique was not however in itself the overall aim of the work undertaken, only an objective. The initial questions were 'what' and 'why'. The investigation would be incomplete without any attempt to reflect on what the purpose of these reliefs might have been. Is this dance taya, in the sense highlighted at the beginning of this paper? Before committing oneself to any answer there are still a number of issues which need discussing and unfortunately at this stage of the research we cannot do them full justice. First and foremost there is the problem of the sequential order of the reliefs, which implicitly

PS Ghurnitam (K32) Reconstruction of the karana movement:

An examination of NS 4, 92-93 and relevant sculptural evidence from the temples at Tanjore, Chidambaram and Kumbakonam leads to the following reconstruction. The dancer performs a spin (bhramari) which ends in samunnata parsva, which means that one side of the torso is slightly raised, with one hand near the chest doing a recaka movement and the other in lata. The final foot position is in a svastika with the back foot in

55

Figure 11. Relief P 5, candi Siva, Prambanan, central Java

relates to their purpose and function. We know that the present sequence is not the original one. Photographs from the Dutch Archaeological Service in Indonesia, taken between 1893 and 1937, show very clearly that the reliefs were not in place when the restoration of the Siva temple began.

can comment that it would be fitting for the Cosmic Dance of the Lord to be represented on this temple and the presence of this sequence suggests an attempt at depicting the ta1J,rf,ava dance, which, as we are told in the Indian agama texts, is not a single dance but has many variant forms.

Was the sequence meant to be made up of all the 108 kara,;a-s listed in the Na_tyasastra- in which case some are missing - or was it meant to portray a string of karaua-s, technically known as angahara, out of the 32 mentioned in the Sanskrit text? As yet there is no answer to this question. The reliefs show a dance sequence of some sort in very great detail, and although the original sequence is now lost, one can say that their raison d'etre was most unlikely to be mere decoration. The Natyasiistra tells us that the Cosmic Dance of Siva, known as ttiJJ(/ava, is articulated through karal}a-s. The significance of the ta1J,rjavais profound. The Loro Jonggrang temple is a temple dedicated to Siva. One

Understanding what the Loro Jonggrang sequence is about has greater importance than one would usually warrant it. The point that needs to be addressed is that the purpose of the complex is still not known with any certainty, despite one or two attempts at investigating it, such as the recent work by Jordaan (1996). The iconographic programme of a temple has usually much to do with its raison d'etre. It would seem that knowing the purpose of the dance reliefs would allow for a greater understanding of Loro Jonggrang and by extension, the complex.

56

Related to the above is the question of the identity of the performers in the reliefs. Could the dancers - all of them male - be Siva himself dancing his Cosmic Dance? The figures in each relief usually show three dancers, all identical, which could well be a device deployed to show the unfolding of the kararJa movement through time, with the 'performers' being in fact only one. To say that the Loro Jonggrang figures represent a dancing Siva is controversial, because it is believed that there are no representations of a dancing Siva in the whole of Indonesia. A full identification of the performers in the reliefs and the establishing of the purpose of the sequence will remain unsettled issues until the cultural/religious background of ancient Java is investigated in greater depth.

objectivity and authenticity and few of them would deny that contemporary interests shape the material past, something of which the various 'Indianization' and 'localisation of Indian influence' theorisations of the Southeast Asian past are a classic example. If the search for origin and purity and all that this entails is unmasked as a rather futile exercise, which posits the past as an objective entity, a more suitable approach to Southeast Asian archaeology and history would seem to be the search for nexus, as an alternative strategy to make sense of the past, acknowledging its dynamism and heterogeneity. Such a shift in emphasis in the way the events of the past are studied and reconstructed is sorely needed, in order to free Southeast Asian archaeology from the burden of creating a mythical past to legitimise present political concerns.

The Prambanan kararya series is about 100 to 200 years older than the first known kara1:a series in India, which is at the Cola Brhadisvara temple in Tanjore, South India. There does not seem to have been any specific prototype or sample for the kara1:a-sof Prambanan - the images can be regarded as a local concept. Could the idea of representing kara1:a-s in a serial order in Cola times have been taken back to India after Prambanan was carved? This is not to say that the concept of karana was not known in India before Prambanan nor that the Natyasastra is of Javanese origin. What is being suggested is that the idea of putting kara1:a-s in a sculpted series of some sort may have developed in Java, generating a fashion, regardless of where individual karana representations were first found. If that were the case, it would be what one could call an instance of recycling, of 'going back and forth'. This begs the question of whether there were other instances of recycling such as this, which is an invitation to rethink South and Southeast Asian interaction throughout history.

Conclusions

Through the specific case study of Prambanan, a large Saiva complex in Central Java, it has been possible, in this paper, to establish a relationship between archaeology, religion and dance. The approach has allowed redefinition of the scope and concern of the archaeological, religious and dance enquiry, highlighting a common ground and creating a new perspective on the way the past, performance and the present interact and are perceived and interpreted. Taking as a point of departure the Javanese concepts of taya and mataya, which belong to the living dance culture of the court, it has been suggested that the Prambanan temple complex, where the main god in worship was Siva, (known in Javanese as Taya, the Supreme Being but also the Dance or, as known from Indian sources, the 'doer' of the Cosmic Dance as well as the Cosmic Dance itself) can be archaeologically investigated as a nexus of dance and religion across regions and across centuries. This has led to questioning some approaches to the Southeast Asian past which still have wide currency and which do not make explicit the mediating role of archaeology in the interpretation of this past.

Earlier this century and until about the early 1950s it was usual to view Southeast Asia (and therefore Indonesia and Java in particular) as an ancient Indian colony, influenced through and through by Indian culture, indeed civilised and moulded by India. The parallel with the civilising mission of European colonial powers is transparent and one need not dwell on it. Later, the emphasis was on the 'local genius' theory and the appropriation and transformation of Indian culture and its accoutrements, so that everything was distinctly indigenous to Southeast Asia (in our case Java), the 'Indianisation' being an easily discarded surface layer. Somehow in this pretty and orderly picture of the Southeast Asian past the awareness that the past did not unfold less dynamically than the present seems to be totally lacking. The travelling of people - and of ideas between South and Southeast Asia must have been a two-way traffic, as Verwey already pointed out in 1962 (Jordaan 1996: 43), yet this is never seriously considered in all its implications.

References

Becker, J. 1993. Game/an Stories, Tantrism and Aesthetics in Central Java. Arizona: Arizona State University Iyer, A. 1998. Prambanan: Sculpture and Dance in Ancient Java. A Study in Dance Iconography. Bangkok: White Lotus.

The point is that different though the 'Indianisation' and 'localisation' theories might be, their common denominator remains a search for origin and a definition of authenticity based on origin and purity. However, a great number of archaeologists today do feel ill at ease with notions of a single and abstract principle of

Nagar, R.S. (ed.) 1988-1989. Natyasastra of Bharatamuni with the commentaty Abhinavabharati of Abhinavaguptacarya. New Delhi: Parimal Publications. Jordaan, E. R. 1996. In Praise of Prambanan. Leiden: KITLV Press.

57

Unpublished PhD thesis, Annamalai University.

Soeharto, B. 1990. Dance Power: The concept of Mataya in Yogyakarta Dance. Unpublished MA thesis, University of California, Los Angeles.

Tilley, C. 1993. Interpretative Archaeology. Oxford: Berg.

Subrahmanyam, P. 1978. Kara!}a-s in Indian Sculpture.

58

A Symbol of the Deity: Artistic Rendition of the 'Hand of God' in Ancient Jewish and early Christian Art. Rachel Hachlili, University of Haifa. Introduction

that pavements are trodden upon would ensure that no sanctity could be attached to any scene depicted there. Such a depiction could not be considered a 'graven image' prohibited by law. The theory accepted by most scholars is that motifs used in Jewish representational art become void of their original, symbolic, idolatrous significance, and evolve into mere ornamental motifs (Avigad 1976: 282, 285). Certain pagan mythological and symbolic motifs are acquired by the Jews through the influence of Jewish legends and Midrashic literature, however, the vast majority of the appropriated pagan motifs are used solely as ornamental designs, considered as a means of decoration, in order to add beauty and ornamentation to a structure.

This study endeavours to inspect a symbol of the deity in the form of the Hand of God. It appears already in ancient Israel and is favoured later in ancient Jewish art, from the 3rd century CE onwards, as well as in early Christian art. The question of the visual appearance of the deity merits examination and will be dealt with in this article. The development of Jewish figurative art and archaeology in Late Antiquity is all the more surprising in the light of the previous attitude towards figurative art. The art and archaeology of the late Second Temple period (1st century BCE until 1st century CE) is purely aniconic and no figurative designs were depicted (Hachlili 1988: 65-83; 103-19), probably due to the prohibition of the second of the Ten Commandments:

The Hand of God probably has its origin in the ancient pagan art of the Near East, where it is part of a formula of symbols of deities. In ancient Near East art, the hand is carved on stone reliefs and appears as an isolated symbol in several forms. However, whereas the Near Eastern examples symbolize the deity by a single carved hand (or a pair), the Hand of God in Jewish and Christian art is usually rendered in narrative biblical episodes where God is rendered in a distinctive role of intervention in the scene.

Thou shalt not make unto thee any graven image, or any likeness of anything that is in heaven above, or that is in the earth beneath, or that is in the water under the earth. Thou shalt not bow down thyself to them, nor serve them. (Exodus 20:4-5; Deuteronomy 5:8-9)

Following the destruction of the Temple in Jerusalem (70 CE), rabbinical attitudes begin to change to one of greater tolerance. Such changes, reflected in Talmudic literature, are the result of political, economic and social circumstances. Now, the rabbis emphasize the latter part of the commandment 'Thou shalt not bow down thyself to them' (Kraeling 1979: 343-45). The Jews of this period were indeed unafraid of idolatry. The leading rabbis tend to enforce the prohibition only where danger of idolatry is present. It seems that by the second to third centuries, Jewish religious leaders permit iconic, representational decoration, and the sources testify to a policy of religious pragmatism and avoid the formulation of binding teachings. Rabbinical evidence suggests that figurative art was tolerated if it did not encourage cultic worship (Urbach 1959). A further reason for the lenient attitude towards figurative art was that no Jewish law forbids the depiction of religious subjects. On the contrary, they were allowed. The archaeological data from the Land of Israel and the diaspora in late Antiquity (3rd until the 6th centuries CE), proves the extensive use of figurative motifs and themes on architectural ornamentation, mosaic pavements, wall paintings and other arts and objects from synagogue structures (Hachlili 1988: 285-316, 321-46; 1998: 23762). No sanctity was ascribed to a painting and, therefore, there was no reason to prevent the depiction of representations on wall paintings such as at Dura Europos (Hachlili 1998: 135-55, 178-82, 193-97), and on mosaic pavements such as at Beth 'Alpha, Na'aran, Hammath Tiberias and Sepphoris in the Land of Israel (Hachlili 1988: 287-316). Furthermore, the very fact

The Hand of God

Three formulas of the Hand of God are rendered in ancient Near Eastern, Jewish, and Early Christian art: 1) The Hand of God appears as an isolated symbol in ancient Near Eastern pagan art in three forms. 2) The Hand of God is depicted as part of biblical scenes in several forms in Jewish art. It appears in the Dura Europos synagogue wall paintings, on the Beth 'Alpha synagogue mosaic pavement, and on a clay lamp. The Dura Europos synagogue (Syria) was constructed in the mid-3rd century CE, in a renovated dwelling house in a residential area (in the western half of block L7) with access from the wall street. The building was not a detached, freestanding structure. The synagogue structure was secluded, inconspicuous and indistinguishable from the neighbouring houses. The early and later buildings, including the Assembly house with annexed House H, consists of a large complex. The more important main complex, consists of the synagogue assembly hall, a forecourt and rooms, some of which probably served as the synagogue guest house and residence. The hall has a semi-circular niche used as a Torah shrine in the centre of the west wall opposite the main, wider entrance. The walls are plastered and are covered with wall paintings. The Dura Europos synagogue wall paintings, the visual expression of a community's religious philosophy, are considered some

59

of the most important and unique in the ancient world. The synagogue hall walls are covered by biblical narrative compositions and ornamentation that from the very beginning form an integral part of the synagogue design concept, and remain so throughout all three building stages. At each phase of the building, wall paintings were executed as part of the decoration of the synagogue (Kraeling 1979; Hachlili 1998: 98-197).

palm outward, possibly in the act of blessing; the other hand holds and draws a bow, symbolizing victory (Figure 1).

The ancient synagogue of Beth 'Alpha is situated at the foot of the slope of the Gilboa mountains in Israel. It consists of a courtyard, a vestibule and a main hall, the south wall of the nave ends with an apse, oriented toward Jerusalem, which housed the Ark of the Scrolls and possibly also two menoroth. The synagogue was decorated with mosaic pavements, the most important of them ornamenting the nave mosaic. It is divided into three main panels: the southern panel in front of the apse, shows the Ark of the Scrolls in the centre, flanked by two seven-branched menoroth, and four ritual objects (as well as a pair of lions and a pair of birds). This probably resembles the actual position of these ritual vessels in the apse. The central panel depicts the zodiac in a stylized pattern of an outer and an inner circle within a square; the inner circle shows the sungod driving a four-horsed chariot against a background of stars and a half-moon. The outer circle depicts the twelve signs of the zodiac with their respective names in Hebrew. The four comers of the square indicate the four seasons. The north panel depicts the Biblical scene of the Sacrifice of Isaac, with Abraham holding Isaac next to the altar, a ram tied to a tree with the Hand of God above it, and the boys with the ass. At the entrances a lion and a bull flank two inscriptions. The Greek inscription commemorates the craftsman who laid the mosaics. The Aramaic is a donation inscription giving the date of the synagogue construction in the reign of Justin, probably Justin I (518-527 CE) (Sukenik 1932).

Figure 1. 'Obelisk' ofTiglathpileser (after Keel 1978: fig.297) On Tomb II, one of a group of Iron Age tombs discovered at the village Khirbet el-Korn, west of Hebron, a hand is carved below a grafitto inscription on the east pillar between chambers 1 and 2, dated to the 8th-7th century BCE (Dever 1970: 159-69, Pl.VI:B). The carefully carved right hand, with spread palm (Figure 2), seems to be part of the original carving of the tomb as it does not eliminate any part of the inscription. This open palm shape seems similar to the so called 'Hand of Fatimah', or hamsa, which later became a popular symbol on amulets with a magical function as protection against the evil eye almost exclusively in Islamic countries (Wallis-Budge 1961: 467-71). Line 2 of the inscription relates 'Blessed be 'Uriahu by Yahweh'. The hand was probably carved purposely, and symbolizes the Hand of God (Yahweh) mentioned in this inscription, indicating an early model of an outstretched hand representing divine blessing. This is the earliest example of a Hand of God represented as a spread palm depicted in ancient Hebrew art (Dever 1970: 169).

3) The Hand of God in Early Christian art is rendered in certain biblical scenes, almost exclusively in those of the Sacrifice of Isaac, of Moses on Mount Sinai, and Moses Receiving the Law; only seldom does the Hand of God appear in a scene depicting Christ. These representations appear on wall paintings and catacomb sarcophagi, ivory boxes and plaques, and in wall mosaics. The Hand of God in ancient Near Eastern pagan art

The Hand of God appears as a sole symbol in ancient Near Eastern pagan art in three forms: A) An open palm hand depicted on a 12th century BCE obelisk. B) A pair of upright hands, on stelae or reliefs. C) A hand holding a thunderbolt mostly on Dura and Palmyra reliefs. A) The Hand of God as an open palm hand is on the upper part of an incomplete 'obelisk' of Tiglathpileser I (12th century BCE) from Nineveh (Parrot 1961: 35, fig. 40; Keel 1978: 215, 217, fig. 297). A winged sun-disk is rendered with two hands issuing from it, one with the

Figure 2. A grafitto inscription on Tomb II from Khirbet el-Korn (after Dever 1970)

60

limestone stelae from Carthage (Yadin 1970: 218-222, Pis. 22-3; Moscati 1968: fig.37; 1988: 311-12), and might represent the continuation of a tradition; they usually accompany the 'sign of Tanit' representing the goddess Tanit herself. Other scholars have still no answer to the meaning and origin of this symbol and consider these hands as having a protective capacity (Schrire 1966: 8; Moscati 1968:160-61). Gonen (1994) maintains that the depiction of hands accompanying the goddess Tanit is the iconographic source for the much later North African hamsa (the amulet with the downward hand). It seems that the Berbers, the autochthonous population of North Africa, preserved anthropomorphic representations which can be found in their pottery and tomb markings, sometimes echoing the 'sign of Tanit'; it seems the hand preserved its original meaning as a magical sign and as a symbol of God's power (Gonen 1994: 51, fig. 10). She further argues (ibid: 55) that all elements of the hamsa, the upsidedown hand attached to a loop for hanging, resembles the basic shape of the 'sign of Tanit', its designation as 'the hand of Fatima' plausibly indicates its descent from Tanit; both Tanit and the hamsa amulet were called upon in situations of stress. On a Palmyrean altar (2nd-3rd century CE), two pairs of upraised hands with their palms outwards are presented (Tanabe 1986: Pl. 152).

Figure 3. Two hands carved on a basalt stelafrom a temple at Hazor (after Yadin 1970: jig.11) B) Another example of the symbol is depicted as a pair of upright hands: Two hands raised up towards a crescent with disk are carved on a basalt stele from a Late Bronze Canaanite (14th-13th centuries BCE) temple at Razor (Figure 3). The upward-stretching hands appear connected to the Razor temple chief moon-god and his consort (Yadin 1970: 216, 222, fig. 11). Keel (1978: 321-23) maintains the Razor raised hands are a symbol of fervent prayer addressed to an astral symbol, the crescent moon with the sun disk, symbolizing the god Shamash. Similar outstretched hands (Figure 4) appear on a marble relief from Enkomi, Cyprus (Keel 1978: fig. 432). Yadin noted that similar detached upraised hands, one or two (Figure 5), appear on 3rd-2nd century BCE votive Punic

C) A third form, with a hand holding a thunderbolt, is

depicted on 2nd-3rd century CE Durene and Palmyrean art. An example appears on a gypsum relief found in a private residence at Dura Europos (Downey 1977:14647, no.179; Moon 1992: 608); it is suggested that this is an abstract symbol probably of Baalshamin or Hadad meant to protect the house from evil. A similar hand holding a winged thunderbolt is depicted on an altar from Palmyra (Tanabe 1986: Pl. 151). The Hand of God in pagan art is frequently the right hand, probably because it is considered the 'hand of honor', as the left hand is used for acts regarded as unclean (Wallis-Budge 1961: 469). The Hand of God in these examples is almost always interpreted as a symbol of power, in a protective role, or in an act or sign of blessing (Wallis-Budge 1961:467-69; Schrire 1966: 8), whether it appears as a hand with an open palm, as a pair of hands raised up, or a hand holding a thunderbolt. The Hand of God image in Jewish art

In Jewish art the Hand of God image appears from the mid-3rd century CE onwards, on wall paintings, on a mosaic pavement, and on a clay lamp. The Hand of God appears in three forms: A) The right palm opened outwards; once the hand appears surrounded by a cloud B) The back of the hand. C) A closed hand grasping the hair. The Dura Europos synagogue wall paintings (245-256/7 CE), covering the hall walls in narrative compositions, show the Hand of God in five biblical episodes (Figure 6), and in all these renditions the hand is represented as a

Figure 4. Two hands carved on a marble relief from Enkomi, Cyprus (after Keel 1978: fig.432)

61

Figure 5. Examples of the Hand connected to the Tanit goddess, on Punic stelae, Carthage (after Moscati 1988: 312)

a

~

b

d

Figure 6. Hands of God on the wall paintings of the Dura Europos synagogue separate entity and is not connected to any other element (Hachlili 1998: 144-46).

the scene in the upper right corner: a conical tent in which stands a small figure. This figure is identified as different individuals. The tent is interpreted in various ways such as the booth Abraham made for the feast of Tabernacles, or as the tent with Sara standing in it. The Sacrifice of Isaac above the Torah shrine at Dura Europos belonging to the first stage of the wall paintings, is thus the earliest of the episode paintings. The convention of the Hand of God in this earlier painting is different from every other Hand of God depicted in the scenes of the second stage at Dura. The formula used by the artist to portray the Sacrifice scene was apparently different from all the other narrative scenes.

The story of the Sacrifice of Isaac (Genesis 22: 1-19) is portrayed at the right of the arcuated panel of the niche (Figures 7-8), belonging to the first stage of the later building (mid-3rd century) of Dura Europos synagogue (Hachlili 1998: 100). The tall figure of Abraham is seen from the rear, clad in white himation, chiton and brown boots, holding in his right hand a white knife; the small figure of Isaac lies atop of a large white altar to the right above Abraham; a ram and tree are depicted at the bottom of the scene; the Hand of God appears surrounded by a kind of cloud above the altar. The artist has added a further detail to

62

Figure 8. Dura Europos, Sacrifice of Isaac (after Sukenik 1947: fig.18) depicted as a left open palm hand above the figure of Moses and a right open palm hand above the group crossing the sea The scene developing from left to right on the extreme left panel of WCI (Kraeling 1979: 143150; Weitzmann and Kessler 1990:108) portrays a tall woman holding a limp child in her arms handing him to a man reclining on a couch in the centre, who also holds a child in his arms (Figure 10). To the right the woman is shown again, this time holding a clothed child in her left arm and gesturing to the man on the couch with her right hand. The Hand of God is depicted above, stretching towards the reclining man as a symbol of lifegiving power. The scene is generally identified (Hachlili 1998: 119) as Elijah reviving the son of the widow of Zarephath (I Kings 17: 17-24), and is part of the Elijah cycle of scenes depicted on the south wall (SC1-SC4).

Figure 7. Dura Europos, the central niche, with the Sacrifice of Isaac scene to the right of the niche The other four scenes in which the Hand of God is rendered as an open palm are painted in the second stage of the later building: Moses and the burning bush depicted on Wing Panel I, Register A, to the right of the upper centre panel Kraeling 1979: 228; Hachlili 1998: 111-12). A figure stands frontally; the head is depicted in a frame, he is bearded and has curly black hair, and is dressed in chiton and himation. His right hand is extended and points towards a bush on the left, his shoes placed beside him next to the bush; a Hand of God is shown above the figure to the left. Between the figure's legs a cast shadow is rendered as an inverted V. A dipinto, two lines to the right of the figure's shoulder, reads in Aramaic 'Moses, son of Levi'. Scholars agree in identifying this scene as Exodus 3 showing Moses and the Burning Bush.

The longest panel in the synagogue paintings NCI, Register C, portrays the Ezekiel cycle (Figure 11), Ezekiel and his prophecy about the Dry Bones. It is subdivided by Kraeling into three sections from left to right, A, B and C, by a change in the background colour (Kraeling 1979: 178-202; Hachlili 1998:123-27. But see Sukenik 1947: 128ff). Kraeling suggests the three sections (A-C) are each divided into several scenes. Section A shows three scenes: on the left in scene Al, three large figures in Iranian dress stretch out their hands (Ezekiel 37:1-2). On the left, a Hand of God grasps the hair of the first figure's head and another extending Hand of God, is placed above the figure's outstretched hand. Human heads, hands and feet are scattered on the ground in the lower left of the scene.

A long composition of three scenes of the Exodus (Figure 9) is depicted on panel WA3 (Kraeling 1979: 74-86; Hachlili 1998: 114-16), which develop from right to left (Exodus 9:22-26; 12-14): Scene 1 portrays a city wall with an open gate and free standing columns, representing Egypt. The large bearded figure of Moses dressed in chiton, himation and sandals, with a raised right hand holding a staff about to strike the water, leads a host of civilians and soldiers from the city towards the water. In scene 2, the drowning of the Egyptians is presented by figures swimming and drowning in the water. Moses with the staff in his right arm stands to their left. Scene 3 shows Moses holding the staff to a narrow strip of black, fish-inhabited water. To his left two companies cross the Red Sea. A red pair of Hands of God appear out of the cloud in scenes 2 and 3,

Scene A2 shows a double range of mountains on the right are crowned with a tree. A Hand of God, extended to the left, is placed along the upper border. Human heads, hands, and feet are scattered on the lower slopes of the mountain. In scene A3, three corpses lie ranged one above the other on the right under a mountain crowned with a tree. Above them on the mountain is

63

Figure 9. Dura Europos, panel WA2, Exodus (after Sukenik 1947:fig.29)

Figure 10. Dura Europos, panel WCI, Elijah reviving the son of the widow of'Z.arephath (after Hachili 1998: fig.lll-14) indicated by his gesture and the Hand of God above (Ezekiel 37: 10). These scenes in sections A and B are interpreted by scholars as showing the prophetic visions of Ezekiel, the events of the Valley of the Dry Bones and the destruction and restoration of national life.

rendered a chest (Ezekiel 37: 4-8). Section B shows three scenes that should be read from left to right, each scene represented by one large figure. In Scene Bl, on the left, a man in Iranian dress with his right hand stretched out, is touched from above by the Hand of God. Three winged figures, Psyches, fly from left to right at the top of the panel. A fourth, larger Psyche moving to the left touches the upper of three corpses that lie before her (Ezekiel 37: 9-10). This scene shows the assembling of the bodies and the revivification by the Psyche. Scene B2 shows a tall man standing in the middle of the section at the centre of the panel, identical to the man on the right in scene B3 (Ezekiel 37: 21-24). Scene B3 shows at the right a large man standing beside a range of mountains crowned with a tree; a head, hands and feet appear below his extended right hand. To his left, ten smaller men (the elders) are arranged in three rows (Ezekiel 37: 10-14). The revived dead are already standing, with the figure of Ezekiel participating,

The prophetic visions of Ezekiel contained the action and events of the Valley of Dry Bones, 'are a parable describing the process by which the dry bones are transformed into living human beings' (Ezekiel 37: 1-14; 21-28). "Those people revived represent the whole House of Israel which in its present exiled state is without hope, and is described that its "bones are dried up" (Kraeling 1979: 186, 193-4). The vision further states that although 'the people are already in their graves God promises to open their graves, he will place his spirit in them so that they will truly live again, and will bring them back to their own land, quite as he revived the dead in the Valley of Dry Bones'. God

64

!~

~

,

::.. ...

-~.

7'

90·

-~~

'"'¾'""' If

,•

Figure 11. Dura Europos, panel NCI, The Ezekiel cycle (after Sukeniak 1947) issued instruction to Ezekiel 'as to what he shall do and what he shall say when the children of his people speak to him asking the meaning of his acts. What he is to say not only reaffirms the promise of the return to the Land of Israel, but tells of the coming of God's servant David as their king, of the establishment of the new covenant of peace, and of the restoration of God's sanctuary in their midst'. All the painted episodes are presumably based on the biblical account in Ezekiel 37:1-28 as well as on rabbinical literature. All six figures are assumed to be the prophet Ezekiel, showing him first being inspired (the first four Iranian-clad figures), then (the two Greekclad figures) making his prophecies, and finally surveying the fulfillment of his prophecies.

theme painted on the Dura Europos synagogue (see above), and on a medallion of a 4th century lead plate (Rosenthal-Heginbottom 1996:55-6, fig. 10). Based on the above discussed comparative material, it seems that this lamp was manufactured in an eastern Mediterranean workshop for Jewish clients. Subjects from the Old Testament were not common in the Eastern Mediterranean in the period before Constantine, and Christian themes first appear in the middle of the 4th century CE (Rosenthal-Heginbottom 1996: 52-8). Themes selected for Jewish lamps were used to convey a specific message; thus, the Sacrifice of Isaac on the lamp was a visual symbol representing God's protection and salvation for the individual customer.

The Hand of God in these scenes is usually the right palm opening outwards, with fingers spread (Figure 6a, hands 2, 3; 6b, hand 1; 6c, hand 1); in two cases the back of the hand appears (Figure 6a, hands 4, 5); in one rendition a closed hand is shown (Figure 6a, hand 1); the left hand with palm outwards is depicted in two instances (Figure 6b, hand 2; 6d, hand 2). Only once is the Hand of God portrayed partly closed, grasping the hair of Ezekiel on the far left (NCI; Figure 6a, hand 1). The form of the Hand of God depicted in the Sacrifice of Isaac (Figure 6e) differs from all the others by the addition of a cloud, portrayed as a two-lined border.

0

II I \

\

The Hand of God is also depicted on a mould-made clay lamp with a representation of the Sacrifice of Isaac on its discus, dated to the period between 150-350 CE (Figure 12) (it came from the antiquities market in Palestine; now in the collection of the Campo Santo Teutonico in Rome [Rosenthal-Heginbottom 1996]). The scene of the Sacrifice of Isaac shows in the centre a bearded Abraham wearing tunica and pallium, holding a knife in his right hand. His head is turned, looking back towards the Hand of God on the left. A naked Isaac kneels beside Abraham. On the right is a homed altar, above it, possibly a column of fire. On the left a ram lies on the ground and above it a bush or tree, over which is a large Hand of God. The lamp, typologically, falls within the examples of Roman Levantine discus lamps, illustrated with various themes drawn from Greco-Roman mythology, Oriental beliefs and daily life. The discus lamp depiction of the Sacrifice of Isaac is similar to the

\

Figure 12. A Jewish discus clay lamp (after RosenthalHeginbottom 1996:fig.l) Another example of the Hand of God in ancient Jewish art is portrayed in the scene of the Sacrifice of Isaac on the third panel of the Beth 'Alpha synagogue mosaic floor (Figure 13), dated to the mid-sixth century CE (Sukenik 1932: 40; Hachlili 1988: 288-92). The composition of the narrative is divided into three events: on the left it shows the donkey and the two lads. In the centre are depicted the ram tied to the thicket with the Hand of God above. On the right Abraham holds Isaac next to the burning altar. The Hand is represented with

65

Bush, 'And the angel of the God appeared to him in a flame of fire out of the midst of a bush' (Exodus 3:2).

spread palm in a cloud (?), with rays of light issuing from it. This hand differs from all other hands discussed before. The Sacrifice of Isaac is an event which held a deep religious implications in Judaism and later came to symbolize the covenant between God and the Jewish people.

The Hand of God image in Early Christian art

In Early Christian art the Hand of God image appear from the 4th century CE onwards, on wall paintings, sarcophagi, lamps and glass dishes from the catacombs in Rome as well as on church mosaic walls and ivory objects (see below). The image appears mostly on certain scenes: the Sacrifice of Isaac, Moses and the Burning Bush, Moses receiving the Law, one episode of Moses being stoned, and one scene of the Sacrifice of Abel and Melchisidech; in several later examples the Hand is shown above the cross or above the image of Christ. The usual convention in Christian art shows the Hand of God above the scene in a sort of cloud. The Hand of God appears in several forms: A) A right or left hand with open palm outward B) A right back hand C) The hand with two fingers together separated from the other three fingers D) A forearm and hand E) A hand holding a scroll

Figure 13. Hand of God on the mosaic pavement at the Beth 'Alpha synagogue (after Sukenik 1947) The idea of representing God by a hand in Jewish art possibly developed from the biblical metaphors 'the Hand of God,' 'the mighty hand', or 'the outstretched arm' (Deuteronomy 7:19), which are conventional formulae. Other examples are Job saying he was touched by the hand of God (Job 19:21) and Daniel who was also 'touched' by a hand (Daniel 10:10).

These can also be considered in greater detail. A) A right or left hand with palm open outward. In scenes of the Sacrifice of Isaac on wall-paintings in the catacombs of Rome, on sarcophagi, and on a lamp (Figure 14), the Hand appears above, with Abraham turning his head towards it (Hachlili 1998: 242-46; figs. V-3b-d, V-4). It is interesting that the sarcophagus of Junius Bassus (dated to 359 CE) has many other biblical scenes (Tronzo 1986: 67, fig. 98), but the Hand of God appears only in the Sacrifice of Isaac scene (Figure 14.4). A simpler rendition is the Hand of God hanging down on the upper left part of the Sacrifice scene (Figure 15) on the painted frieze in the cupola of a mausoleum at El Bagawat (Egypt) dated to the 5th century (Fakhry 1951: 44f).

The Dura representations of Exodus and Ezekiel follow the wording of the Hebrew biblical text mentioning the Hand of God. In the Exodus story it is written, 'I will stretch out my hand and smite Egypt' (Exodus 3: 20); 'for with a strong hand has God brought you out of Egypt' (Exodus 13: 3, 9, 14, 16). Following the drowning of the Egyptians, the Bible concludes: 'and Israel saw the large hand which God did against the Egyptians' (Exodus 14:31). It should be also noted, that a midrash (Exodus Rabbah XXII,2; Kraeling 1979: 83, n. 251) mentions two hands of God involved in the Exodus story: one saved Israel from the Sea, and one overthrew the Egyptians, although both were interpreted as right hands.

Among the many biblical themes on the Via Latina catacomb wall-paintings in Rome (dated to the 4th century CE) only the scenes of Moses Receiving the Tablets and the Sacrifice of Isaac appear with a Hand of God (Ferrua 1991: Pls.90, 125). On the upper panel of the right side of an ivory casket from Brescia (dated to third quarter of the 4th century), on the left of Moses on Mount Sinai appears a right Hand of God with palm opened outward (Volbach 1961: 328, Pls.87,88; 1976: 104-5, nos.161,162). On the mosaic wall frieze in the central nave of the 5th century Santa Maria Maggiore, on the lower register, the scene shows the Stoning of Moses. A right Hand of God with palm opened outward is rendered coming out of the clouds above the figure of Moses (Volbach 1961:336, Pl.128).

In Ezekiel's vision, the image clearly indicates the intervention of God: 'And he put forth the form of an hand and took me by a lock of mine head' (Ezekiel 8:3). The biblical text of Elijah reviving the widow's son mentions only God, not his hand (I Kings 17:22), although in another scene 'the Hand of the God was on Elijah' (I Kings 18:46). In the bible, therefore, the Hand of God image is used metaphorically and factually. In the Dura paintings, in the Exodus narrative as well as in the Ezekiel cycle (and possibly in Elijah and the widow), the paintings follow the words of the biblical story, and represent the Hand of God intervening in the scene. However, note that the Hand of God represents the angel in the episodes of the Sacrifice of - 'But the angel of God called to him from heaven, and said ... ' (Genesis 22:11), and in that of Moses and the Burning

B) The back of the right Hand of God. On two 5th century ivory pyxis from Trier and from the area of Koblenz the back of the right Hand of God with

66

Figure 14. Hand of God in Sacrifice of Isaac scenes on wall paintings and sarcophagi at catacombs of Rome (after Ehernstein 1923: Kapitel IX, 5-8, 11, 13, 18)

Figure 15. A Hand of God in the Sacrifice of Isaac scenes at the El Bagawat wall paintings (after Goodenough 1964)

67

separated fingers appears in the scenes of the Sacrifice of Isaac (Smith 1922: no.114, Figure 7; Reusch 1965: nos. 54, 106a; Volbach 1976: 104-5, nos.161,162). An ivory pyxis (now at the Hermitage in Peterburg; Volbach 1976:117, no.190) shows the Hand of God coming out of a cloud above the figure of Moses.

example on the presbyterium apse mosaic of Sant' Appolinare in Classe, Ravenna, shows the Hand of God, holding a scroll, stretched out towards Moses (Deichman 1958: Pl. 319). It also appears on the apse wall mosaic of the 6th century Monastery of St. Catherine at Mt. Sinai (Weitzmann 1982: 13, figs. 9, 14; Pl.3). The 11thcentury Cosmos Indicopleustes painted miniature shows Moses receiving the Law in the form of a scroll from the Hand of God (Sinai Cod. 1186 Fol. 101; Weitzmann 1982: 13, fig. 14).

C) This form shows the Hand with two fingers together, separated from the other three fingers, expressing a Christian blessing (Gutmann 1984: 119); all the examples date to the 6th century. This convention appears in the scenes of Moses and the Burning Bush, the Sacrifice of Isaac, and the Sacrifice of Abel and of Melchisedech on wall mosaics on the lunettes of the north and south walls of the presbyterium of San Vitale, Ravenna (Deichman 1958: Pis. 312, 313, 315, 318). The same Hand convention appears on the Cosmos lndicopleustes painted miniatures (Vatican library Hs. 699, dated to the 9th century, originating in the 6th century) showing the sacrifice of Isaac and Moses on Mount Sinai (Ehernstein 1923: Kapitel IX,21; Kapitel XVII,6). The back of the right Hand of God on the presbyterium apse mosaic of Sant' Appolinare in Classe, Ravenna, is shown differently in that it hangs down in the centre above the cross (Volbach 1961:344, Pl.173). On amulets depicting the Sacrifice of Isaac which Goodenough (1953, II: 224; III: figs.1039-1041, following Bonner) considered to be Jewish, the divine Hand is shown with the thumb and first two fingers extended as in the 'Sabazius hand'. However, as there is no example of this form in Jewish art it is more plausible that these amulets are Christian.

Figure 16. A hand of God in the Sacrifice of Isaac, on a glass dish from Bologna (after Milburn 1988: fig.177) Only a few examples depict the image of Christ with a Hand of God. On the central panel of the back side of an ivory casket from Brescia, the right palm Hand of God, in the separated fingers convention, appears between Christ and two youths (Volbach 1961: 328, Pls.87,88). On a 4th century ivory panel from Bamberg, Christ is depicted ascending to heaven, grasping the Hand of God stretched out to him (Volbach 1961: 329, Pl.93).

D) The Hand of God as a dressed left forearm and hand. This form is represented on two glass dishes depicting the Sacrifice of Isaac, from Bologna (Figure16) (Milburn 1988: 269-270, fig. 177), and from Trier dated to the 4th century (Smith 1922: no. 88; Reusch 1965: no. 56). Another example of a hand and arm of God appears as part of a schematic depiction of the Sacrifice of Isaac on a medallion in the centre of a 4th-century lead plate (Rosenthal-Heginbottom 1996: 56, Figure 10).

Conclusions E) The Hand of God presenting Moses with a closed scroll.

The hand as a symbol of a deity presented as an open palm seems to originate from pagan art of the ancient Near East. The hand appears in a few examples in isolation from the 14th century BCE onwards. The Hand of God, whether it appears as a hand with an open palm (as on the grafitto from Khirbet el-Korn [Figure 2]), as a pair of hands raised up (as on a stele from Hazor, a relief from Enkomi and stelae from Carthage [Figures 3-5]), or a hand holding a thunderbolt, represents either the deity itself (as in the example of the goddess Tanit), as a symbol of power in a protective role, or it symbolizes God in the act of blessing. Although the Hand of God was periodically used in the art and archaeology of the ancient Near East, it seems that only by the mid-3rd century CE on the paintings at the Dura Europos synagogue - the Sacrifice of Isaac, the exodus scene, Moses at the burning bush, the scene of Elijah reviving the son of the widow of Zarphath (Figures 8-13) - was the Hand of God, established as an artistic formula and convention in ancient Jewish art. It represented God's intervention in significant episodes in

The scene in which Moses receives the Law in the form of a closed scroll representing the Tablets of the Law, occurs already in early times and is frequent in Moses scenes in Christian art (Weizmann 1982: 400). It was explained as an influence of Jewish legend and as an interpretation of this scene as Moses on Mount Horeb. Another explanation is the typological tradition of early Christian art in which Moses was presented as a type of Christ and connected to his New Law ( St. Clair 1984: 16, and notes 15-23). It might also strengthen the interpretation of the scene on Dura Europos Wing panel III that the figure represents Moses holding the Law, which is depicted as a scroll. Moses receiving the Tablets of Law in the form of a scroll is depicted in a 5th century sarcophagus from Tarragona, a silver reliquary from Thessaloniki, and the 6th century stone relief 'stele of Joannes' from Constantinopole (St. Clair 1984: 16, figs.8-10). Another

68

in significant episodes, in a gesture of salvation, of life giving, and of blessing.

Jewish history, and the commemoration of special events. In the Elijah scene painted at the Dura Europos synagogue (Figure 12), the Hand of God may represent a lifegiving power (Weitzmann 1990: 108), as well as intervention by the Angel of God. The Hand of God images at Dura Europos synagogue, although seeming to be within an already established convention, are the earliest appearances of this image in ancient Jewish art. It seems reasonable to assume that this formula influenced the rather later artistic convention of Early Christian biblical scenes in paintings, reliefs, and mosaic ornamentations.

References

Avigad, N. 1976. Beth She'arim III, Catacombs 12-23. Jerusalem: Israel Exploration Society. Deichman, F. W. 1958. Fruhchristliche Bauten und Mosaiken von Ravenna. Wiesbaden. Dever, W.G. 1970. Iron Age Epigraphic Material from the Area of Khirbet el-Korn. Hebrew Union College Annual 40-41:139-204.

Comparing the depiction of the Hand of God in Jewish and Christian art shows some important similarities and differences. In Jewish art, the Sacrifice of Isaac appears several times; the Dura paintings show the Hand of God also in the episodes of Moses on Mount Sinai as well as in other episodes. In Early Christian art the Hand of God appears almost exclusively in scenes of the Sacrifice of Isaac and Moses at the Burning Bush and Receiving the Tablets; frequently the two scenes are paired (St. Clair 1984: 16). Note that even when many other scenes appear, as in the examples of the paintings of the Via Latina catacomb and the Junius Bassus sarcophagus relief, only these two scenes render the Hand of God.

Downey, S.B. 1977. The Stone and Plaster Sculpture. New Haven: Yale University Press. Ehrenstein, T. 1923. Das Alte Testament im Bilde. Wien: A. Kende. Fakhry, A. 1951. The Necropolis of El-Bagawat in the Khaga Oasis. Cairo. Ferrua, A. 1991. The Unknown Catacomb. Florence: Geddes and Grosset.

The Hand of God holding a closed scroll (as the representation of the Tablets of Law) stretched out towards the figure of Moses is prevalent in Early Christian art. The form of the hand in the Jewish scenes is usually a spread palm, whereas the hand in Christian art has several forms such as a dressed arm, and is frequently shown as a hand with two fingers together raised in the act of benediction (form 3). The size of the hand in the Jewish examples (Figure 7) is usually conspicuous, slightly or much larger proportionally than the figures in the scenes, thus possibly stressing the difference between God and humans. In the Christian catacomb scenes, as well as in the later Ravenna wall mosaic examples, the hand is the same size as the other hands of the figures, and so it is a personification of the human by the hand. The Dura paintings show the hand depicted without any addition, on the border of the panels; only in the Sacrifice of Isaac at Dura and Beth 'Alpha mosaic does a type of cloud appear. In the Christian examples the hand appears in a small or large cloud.

Goodenough, E. R. 1953-68. Jewish Symbols in the Greco-Roman Period, vols. 1-13. New York: Panteon Books. Gonen, R. 1994. The Open Hand on the North African Hamsa and Its Sources. The Israel Museum Journal XII: 47-56. Grabar, A. 1968. Christian Iconography. A Study of Its Origin. Princeton: Princeton University Press. Gutmann, J. 1984. The Sacrifice of Isaac: Variations on a Theme in Early Jewish and Christian Art. (In), Ahrens, D. (ed.), Festschrift far Josef Fink. pp. 115-122. KolnWien. Hachlili, R. 1988. Ancient Jewish Art and Archaeology in the Land of Israel. Leiden: Brill. Hachlili, R. 1998. Ancient Jewish Art and Archaeology in the Diaspora. Leiden: Brill

The difference between the way the formula of the Hand of God is represented in Jewish and in Christian art may be the result of two different sources, one serving Jewish art and the other used in Christian art; or it may be because the meaning conveyed by each of the arts was dissimilar and served different purposes. Whereas the pagan Near Eastern examples of the Hand of God are symbolic in form and concept, in Jewish and Early Christian art in almost all the renditions known, the Hand of God is shown as an important part of a biblical scene in which an activity is depicted, and the Hand has several symbolic forms. The archaeological material presented above indicates the continuing use of the image both in form and in its symbolic meaning. The Hand is portrayed as an open palm, and is interpreted as a symbol for a deity representing the intervention of God

Keel, 0. 1978. The Symbolism of the Biblical World. New York: The Seabury Press. Krealing, C. 1979. The Synagogue: The Excavations at Dura Europos, Final Report VIII, part I. New HavenNew York: Yale University Press. Milburn, R. 1988. Early Christian Art and Architecture. Chico, California: University of California Press. Moon, W.G. 1992. Nudity and Narrative: Observations on the Frescoes from the Dura Synagogue. Journal of the American Academy of Religion LX: 587-658. Moscati, S. 1968. The World of the Phoenicians.

69

Tanabe, K. 1986. Sculptures of Palmyra I. Tokyo: The Ancient Orient Museum.

London: Weidenfeld and Nicolson. Moscati, S. (ed.), 1988. The Phoenicians. New York: Abbeville Press.

Tronzo, W. 1986. The Via Latina Catacomb. Imitation and Discontinuity in Fourth-Century Roman Painting. University Park and London: The Pennsylvania State University Press.

Parrot, A. 1961a. Sumer, The Arts of Art. New York: Golden Press. Parrot, A. 1961b. The Arts of Assyria. New York: Golden Press.

Urbach, E.E. 1959. The Rabbinical Laws of idolatry in the Second and Third Centuries in the Light of Archaeological and Historical Facts. IEJ 9: 149-165; 229-245.

Reusch, W. 1965. (ed.), Fruhchristliche 'Zeugnisse im Einzugsgebiet von Rhein und Mosel. Trier: UnitasBuchhandlung.

Volbach, W.F. 1961. Early Christian Art. New York: Harry N. Abrams.

Rosenthal-Heginbottom, R. 1996. A Jewish Lamp Depicting the Sacrifice of Isaac. Eretz Israel 25: 52-60.

Wallis-Budge, E.A. 1961. Amulets and Talismans. New York.

Schrire, T. 1966. Hebrew Magic Amulets. New York: Behrman House Inc.

Weitzmann, K. 1982. The Mosaic in St. Catherine's Monastery on Mount Sinai. (In), Weitzmann, K. (ed.), Studies in the Arts at Sinai. pp 5-18. Princeton: Princeton University Press.

Smith, A.M. 1922. The Iconography of the Sacrifice of Isaac in Early Christian Art. American Journal of Archaeology 26: 159-73.

Weitzmann, K. and Kessler, H. L. 1990. The Frescoes of the Dura Synagogue and Christian Art. Washington: Dumbarton Oaks.

St. Clair, A. 1984. The Basilewsky Pyxis: Typology and Topography in the Exodus Tradition. Cahiers Archeologiques 32: 15-30.

Yadin, Y. 1970. Symbols of Deities at Zinjirili, Carthage, and Hazor. (In), Sanders, J. A. (ed.), Near Eastern Archaeology in the Twentieth Century. Essays in Honor of Nelson Glueck. pp.199-231. New York: Doubleday and Company.

Sukenik, E. L. 1932. The Ancient Synagogue of Beth 'Alpha. London: University Press. Sukenik, E. L. 1947. The Synagogue of Dura-Europos and its paintings. (in Hebrew). Jerusalem: Mosad Bialik.

70

Text, Tradition and Transformation: Mandan's 'Devata-Murti-Prakarnam'. Rima Hooja, Institute of Rajasthan Studies, Jaipur. calibration). Five 'fire-altars' were unearthed from the mound located east of the 'Lower City', while another set of seven 'fire-altars' in a row was excavated from mud-brick platforms found in the southern half of the 'Citadel' complex. One of these contained animal bones. Consequently, archaeologists have conjectured that the people of Kalibangan Period II practiced fireworship and animal sacrifice. A few have even viewed this as a precursor to the Vedic (Indo-Iranian) tradition of rituals performed around a sacred fire-altar, in which a sanctified fire became the medium for conveying the prayers and offerings of humans to the gods. Others regard this as further proof that the Harappans were identical to the Vedic people. In both cases, the projection of a known form of religious practice and rituals onto past phenomena can be noted.

Introduction

The Indian city of Jaipur has several claims to fame; one being the marble statues carved in the 'Murti Mohalla' ('Sculpture Quarter') of the old Walled city. Interestingly, a question prospective buyers often hear (whether articulated in Hindi or English), is - 'Do you wish to buy a 'murti' or a statue?' The question seems innocuous - even ludicrous - since a 'murti' means a statue. And yet the question encompasses an entire tradition and belief system; for an idol that is intended for worship must be created from the very start according to prescribed norms, while a statue required for a decorative purpose is free from constraints of iconographical precision. This fine distinction between what is sacred or has religious connotations, and what is an artistic object of creativity and beauty is a problem archaeologists often face when looking at material remains vis-a-vis the issue of religion and archaeology. What gives a material object a religious sanctity? How accurate can we be in ascribing a religious meaning to an artefact? How does archaeology walk the tight-rope between recognising phenomena that apparently reflects religious beliefs at an archaeological site, and over-zealous scholarly attribution?

One may add here that no 'mother-goddess' figurine, so ubiquitous to the Indus Valley Harappan sites, have been found at Kalibangan. Does that mean the religious beliefs of the Harappans varied from area to area and site to site - rather like the religious life of the early Greek city-states and certain proto-historic I early historic cities of the 'Fertile Crescent'? Iconography, Archaeology and Textual Traditions

In the case of South Asian archaeology, one well-known depiction on seals of the Indus Valley (or Harappan) culture sites shows a cross-legged seated figure surrounded by animals. Archaeologists tend to view this figure as a 'Lord of the Animals' (Pashupati) Proto-Siva, given the fact that such an aspect of Siva is recognised in later Indian tradition. Similarly, the terracotta female figurines found at many of the excavated and explored sites of the Harappan Culture (besides in other areas and contexts), are usually classified as '(so-called) mothergoddesses'. How verifiable are the hypotheses? How does archaeology develop a reliable methodology for examining such aspects of past cognitive behaviour?

Some of the questions concerning archaeology and religion are lessened when the religious tradition being looked at is a 'living one'. In the case of Hinduism (the development of which is not being discussed due to lack of space), iconographical traditions have been generally described, even prescribed, in minute detail in various Indian treatises. These developed from descriptions given in early religio-philosophical literature. Once written down in the canons, it is generally believed that these have been (and are) strictly and dutifully adhered to. However, upon a comparative study over different centuries, it becomes obvious that there is no such phenomenon as a permanent and constant 'Hindu' iconography.

Are the terracotta bull-figurines found in large numbers in the Ahar Culture of south-eastern Rajasthan and other c.3rd-2nd millennia BC chalcolithic sites ritual objects, toys, or items reflecting human artistic creativity? Is there any religious meaning behind the moulding of these bull-figurines, or alternately, do they have 'symbolic' (promotion of cattle-wealth?), or 'functional' (keeping track of the number of herds?), or other meaning?

For example, Radha is popularly accepted as Krishna's consort - being regarded from the medieval period onwards as a manifestation of Goddess Lakshmi, the consort of Vishnu, who took the human-born form of Radha to co-exist on the earth while Vishnu's incarnation Krishna did so. Radha and Krishna are unhesitatingly regarded as two halves of a composite whole by the average Indian today (who may, on the other hand, need to think for a moment before naming any of the wives of Krishna mentioned in the epic Mahabharata).

To take another example, 'fire-altars' containing burnt offerings were noted during the 1961-1969 excavations at the site of Kalibangan in northern Rajasthan, from the Mature Harappan Kalibangan Period II levels, which date from between c.2500-2000BC (MASCA

While idols of Radha and Krishna are made and jointly worshipped in large numbers now, none of the very

71

Archaeological evidence for a stone enclosure occurs at Nagari (ancient Madhyamika-nagari), near Chittorgarh in south-eastern Rajasthan, which the Ghosundi-Nagari Inscription, datable to c.200-160 BC, associates with 'Narayan-Vatika' ('Narayan Woods'), connecting it with the worship of Vasudeva-Sarnkarshna. Besides throwing light on the evolution of Vaishnavism, this Ghosundi inscription also provides one of the earliest recorded references to an image of Vasudeva-Sarnkarshna. Unfortunately the appearance of the image remains unknown (a later inscription dated AD 424, found in situ at Nagari, mentions a brick-built temple to Vishnu standing on that site. So does another inscription dating to c.7th century AD).

ancient canons on iconography provide any instructions whatsoever for sculpting an image of Radha. In fact, though early literary references to Radha include the c.2nd century AD Prakrit composition Gatha-Saptashati by Hala, as well as texts like the Padma Purana, the present iconographic position of Radha seemingly evolved from around the 12th century onwards. This accompanied the joint worship of Radha-Krishna popularised by Nimbarka and various other Bhakti poetsaints, and was accentuated through religious-cumliterary works. Thereafter, Radha became the subject of paintings, literary compositions, etc., a position that continues in the late 20th century. Similarly, Buddha's incorporation as one of Vishnu's ten incarnations was a transformation in the iconographical tradition that came about in the post c.7th-8th century AD period. This was a consequence of the successful attempts by resurgent Brahmanical Hinduism to bridge the schism with Buddhism. This altered the scheme provided in older texts, which delineated Balarama and Krishna as the 8th and 9th incarnations respectively. Henceforth, different iconographical treatises listed either Krishna or Balarama as the 8th incarnation of Vishnu, while invariably placing Buddha firmly as the 9th. An example from Rajasthan is the c. lOth century AD panel from Kumher in the Bharatpur Museum, which shows Buddha among the Ten Incarnations.

Thus, by about the 2nd-1st centuries BC the synthesis of Vasudeva-Krishna with Narayan-Vishnu was established. This is confirmed by the Besnagar Inscription from Central India in honour of the God Vasudeva by the Greek envoy Heliodorus; south-eastern Rajasthan's Amaleshwar Inscription from Pratapgarh (district Chittorgarh); and the Ghosundi Inscription (referred to above); all three dating to the 2nd-1st centuries BC. However, archaeologically reconstructing the process of this syncretism, and the consequent changes in iconography, philosophy and religious practices that may have occurred, remains a problematic issue. This synthesis was apparently complete, with Vishnu (and the equally important Siva) attaining the 'classical Hindu' form expressed in traditional Purana and similar texts and iconography by c. AD 350-400, the period when the Gupta dynasty expanded its empire across much of South Asia. Many of the Gupta dynasty emperors declared their devotion to 'Vasudeva-Vishnu' on coinage, epigraphs etc. (as did other contemporary ruling houses). Consequently, there was a proliferation of temples and sculptures in honour of Vishnu and his incarnations like Vaman, Varah, Narsingh, Balarama and Krishna over different parts of the Sub-continent, including Rajasthan. (The situation regarding Siva temples and idols was similar).

Iconography, thus, reflects existing beliefs and is often coloured by a time-space-specificity. Over the ages, changes in the core-body of the relevant religious tradition has resulted in the creation of a revised iconography, which in tum becomes codified, either in writing or as part of an oral tradition. The archaeological recognition of this process is not always easy. Let us take the cult of Vishnu (Vaishnavism) - the 'Preserver' element of the Hindu Trinity (comprising Brahma denoting the Creator, Vishnu the Preserver, and Siva the Annihilator-cumReproducer), for instance. A definite identifiable iconography of Vishnu evolved by about the 5th century AD, to which additions continued to be made in ensuing centuries. Lack of adequate archaeological data means, though, that there is still speculation, of necessity, about the exact period and process by which an initial synthesis occurred (probably sometime in the 1st millennium BC) between the Vishnu who is a lesser solar deity in Rig Veda, the cosmic god Narayan (with his cult of Pancharatra) whom the Satpatha Brahmana acknowledges as the Supreme deity, and the cult of Vasudeva-Krishna and the 'Five Heroes' of Krishna's Vrishni-Yadava clan.

This proliferation, no doubt, established and perpetuated the prevailing accepted iconography for depicting Vishnu (and Siva) and resulted in its further codification for future generations. In the case of Vishnu, it is from the Gupta period that the image of a four-armed Vishnu bearing a mace, discus, lotus and conch-shell became firmly established in the tradition (placing these attributes in the hands of Vishnu statues in prescribed permutations determined that specific idol). Other 'innovations' were the Vaikunth and Vishvarupa idols of Vishnu. In the former additional heads - usually of Vishnu's Narsingh and Varah incarnations - were depicted on a Vishnu image. The latter was a 20 armed idol, with a halo incorporating various other portrayals of Vishnu images, to represent Vishnu's all-pervasive aspect.

The scanty literary and archaeological evidence available indicates that these three cults evolved and developed independently. As a result of religious syncretism that occurred around the 2nd century BC the deities of the three became identical. As such, on the one hand, the solar deity Vishnu was identified with the cosmic Narayan, and on the other the Vasudeva-Krishna cult was incorporated.

The Iconography of Religion

It may be relevant to briefly summarise the iconographical traditions of Rajasthan here. While rock

72

art, 'votive objects', and terracotta figurines etc. are known from different pre-, proto- and early historic contexts, and early (c.?1000-600 BC) literary evidence refers to the area comprising Rajasthan, firm evidence for religious iconography and architecture in the context of Rajasthan dates from around the 3rd Century BC. This was the period when the Mauryan Empire stretched across much of South Asia.

Hinduism in Rajasthan. The Mauryan-Sunga iconographical tradition was thus a mixed one. The ensuing Kushan period was marked by Yaksha images, Siva-lingas (among them the one from Nand, near the ancient pilgrimage site of Pushkar), various Buddhist and Brahmanical sculptures in the 'Mathura Art' style, as well as terracotta plaques. The terracotta plaques found at northern Rajasthan sites like Rang Mahal, Badopal, Munda, Pir Sultan-ri-Theri, which are attributed to the late Kushan-early Gupta period, are noteworthy as they reflect the influence of Gandharan art. Many themes are based on Krishna's life. Sivarelated subjects also occur. It is from this point forward that the fusion of various religious cults and sects becomes iconographically recognisable. By the 3rd century AD more terracotta plaques, stone images and, a little later, temples of the Classical Gupta form begin to appear.

In fact, it would be well to bear in mind that various large empires or kingdoms - e.g. the Mauryan (c.3rd-2nd century BC), Sunga (c.2nd century BC), Kushan (c. lst centuries BC-AD), Gupta (c.4th-6th centuries AD), Harsh Vardhan (7th century), Gurjara-Pratihara (c.8th10th), among others, enabled a uniformity in iconography over larger areas than an archaeologist would normally consider geographically probable. This, along with later canonical injunctions may be a factor influencing the overall 'pan-Indian' nature of South Asian iconography; of course with regional variations.

The Gupta period iconography is well known. Statues from Jagat, Amjhara, etc. are notable among the examples from Rajasthan. The Gupta period iconographical tradition became the basis for later iconography. The religious sculptural tradition of Rajasthan is marked henceforth by Hindu and Jain statues in the 'Classical' Gupta and Post-Gupta style, with various sub-regional stylistic traits under the patronage of different political ruling dynasties. These include the Gurjara-Pratiharas of Rajasthan who established a large empire across much of northern India during the 8th-11th century AD. They and their local contemporaries raised temples to Vishnu, Siva, Shakti and the other deities in Rajasthan in the iconographical tradition that was by then well delineated. Significantly, the overall iconographical tradition is markedly 'panIndian'.

There is archaeological evidence from the site of Bairat (eastern Rajasthan) for a circular Buddhist temple, monastery and stupa-remains, along with two Rock Edicts of the Mauryan Emperor, Ashoka, and other artefacts datable to the 3rd-2nd centuries BC. Buddhist stupas and other remains also occur at sites like Nagari, Chittorgarh and Lalsot. The Mauryan period, which saw an increase in terracotta and stone sculptures, was marked by the appearance of 'Yaksha' figures, both colossal and small-sized, which were placed on platforms or under trees and ritually worshipped. Yakshas continue in the succeeding centuries too, albeit with changing fortunes. One famous 2.5m high Yaksha image - probably belonging to the 1st centuries BC-AD Kushan period, comes from Noh (in eastern Rajasthan, and geographically within the sphere of the Mathura 'school' of art). Archaeological evidence suggests that, apparently, there were no temples or other buildings used to house these Yaksha figures.

Meanwhile, Buddhism gradually lost its hold in Rajasthan, while Jainism continued to grow in strength. From this period onwards, the Jain usage and adaptation of 'Hindu' iconography and vice-versa is also noticeable. One example is the idol of 'Sacchiya Mata', that attracts thousands of Jain pilgrims every year to her shrine in Osian (western Rajasthan). This was originally a 'Mahishasura-Mardini' idol of Hinduism (Agrawala 1996). Jain texts state that she was a fierce goddess to whom animal sacrifices and wine were offered as oblations. The revered Jain teacher, Ratnaprabh Suri, converted her from her fierce to her present benign form, however, and to vegetarianism and Jainism (ibid). Jain adaptations have used other 'Hindu' deities in a transformed manner, including as door-keepers and attendants of the Tirthankars.

As early as the 2nd-1st centuries BC Yakshas, and their female Yakshini counterparts, were adapted into Buddhism, and depicted accordingly. The 'Yaksha cult' appears to have been freely incorporated in other traditions. Examples are the Kushan period Siva-lingas from Gamri (near Noh), and from the nearby Chauma Bhandpura, which depict the image of a yaksha as well as of Siva. By the Gupta period the Yaksha figure had lost its earlier position, for it features as a demi-god and guardian figure in 'Classical' Hindu iconographical tradition. It is in this latter form that Yakshas find mention in Hindu and Jain iconography in Mandan's Devata-Murti-Prakamam. It may be this Yaksha legacy that is still reflected in the local folk-tradition of erecting a free-standing image or symbolic stone as the village 'guardian' deity.

Around this time composite iconographic forms became established as part of the iconographic tradition in different parts of Western India (e.g. Osian, where the trend of idols like Harl-Hara - combining Vishnu and Siva, can be noted). The period between the 9th century and the end of the 12th century saw the development from the uni-cella temple architecture (housing a single deity) to bi- and tri- and eventually penta-cella templecomplexes firmly established, both in Rajasthan and in neighbouring Gujarat and Central India - areas geographically contiguous and culturally linked with

Despite Mauryan patronage to Buddhism (and Jainism), Pushyamitra Sunga, who overthrew the last Mauryan Emperor, encouraged a Brahmanical revival and in Rajasthan this revival is vouchsafed by the discovery of pillars known as Yupas as at Nandsa (225 AD), Barnala (227 AD), etc., dating from a century or so afterwards and which point to the resurgence of Brahmanical

73

and the human were continuously and variously intermixed, many of the sacred images created by this art tended to become diagramrnatical' (Bussagli n.d.:7). This iconography - whether Buddhist, Hindu or Jain was, admittedly, simply a means to focus and concentrate the mind of the devotee beyond the earthly plane. None the less, rules for portraying the Divine soon emerged.

southern Rajasthan (Rangarajan 1990:11-12). Both developments seem linked to deliberate attempts at effecting philosophical harmony and religious accord between various 'Hindu' cults. The bi-cellae temples were invariably dedicated to Siva and Vishnu, the tri-cellular ones to Siva, Brahma and Vishnu, and the penta-cellular ones (or Panch-ayatana) to Siva, Vishnu, Ganesh, Shakti and Surya. Early examples of the penta-cellae temples from Rajasthan come from two Hari-Hara and one Surya panch-ayatana from Osian dating to the 8th-9th centuries. It is this collective tradition that became the legacy for medieval Rajasthan's iconography.

The establishment of figurative formulae did not necessarily lead to a stilting or impoverishment of art, however, for Indian art responded creatively to contact with groups like the Iranians, Greeks, Scythians, and others, and the formulas too 'evolved with the changing taste, techniques and social structures' (ibid: 24).

Notable works in the ensuing centuries include the Delwara temples to Jain Tirthankars built in the 11th and 12th centuries. It may be pertinent to note the 13th century Jain 'Tower of Glory' at Chittorgarh too, because Kumbha, in his reign, had a 9 storey, 40m high Victory Tower erected (Mandan was not its architect), which bears an inscription of AD1460, and is covered with sculpture on its internal and external surfaces. Such towers are not recognised parts of any canonical tradition that one can recognise for Rajasthan. The iconography described in Mandan's DMP, however, is well-illustrated by the images found here.

Once the iconographical tradition was established (whether written down, or passed on orally from one generation to the next in an unaltered form - as was the case with much of the religious and philosophical body of knowledge), adherence, rather than artistic freedom for continually fresh interpretation, was apparently encouraged by the canons. These canons seem to have become the norm for pan-Indian iconography, with occasional acceptable revisions by devotee~scholars. Among such traditional treatises on architecture and Hindu iconography are texts like the Brihat-Sarnhita, Matsya-Purana, Agni-Purana, Vishnudharmottar-Purana, Mansar, Maymtam, Shilpa-Ratna, Vastu-shastra, etc. These originated and were written down in different regions of India at different times. Of significance, though, is the strand of commonality that runs through these different works.

Much of the iconography that developed had its roots in poetic imagery provided in very early oral and textual references like the Vedas and similar literature. Indian iconography tried to give shape to 'abstruse metaphysical concepts on which, in the cultural and social atmosphere of India, all other human values depended. Based on a vision of life in which the divine

Figure 1. Kumbhalgarh Fort ( 15th century AD)

74

It is against this backdrop that my paper looks at a 15th Century AD Sanskrit text on iconography from the erstwhile kingdom of Mewar (now part of the modemday State of Rajasthan) in India - the Devata-MurtiPrakarnam - by Mandan (a manuscript I have translated from Sanskrit to English [Hooja, in press 1998]).

Chaturvarga-Chintamani by Hemadri (surprisingly, Sankhyatirtha [1936: 5] believed that Mandan's DevataMurti-Prakarnam is more South Indian rather than North Indian in character). The Devata-Murti-Prakamam (hereafter referred to as DMP), is a work in eight chapters which graphically discusses various aspects of Hindu iconography. A brief summary of the DMP is relevant here. The text takes the artist from the initial step of choosing a suitable working medium, to fabricating it into an idol according to the established conventions of iconography and in the prescribed scale, and finally installing it in an auspicious position within a shrine.

This is a text which was composed at a stage when Hindu iconographical traditions were firmly set, and features like composite figures part of the established working tradition. As such, it provides some insight, and also enables the framing of certain questions regarding issues of religion, iconography and archaeology. At the same time a comparison of this 15th century AD work with contemporaneous iconographical traditions, as well as with present-day practices can help us to understand the processes by which traditions and practices are both established and transformed over the course of time.

The first three chapters discuss aspects like the different heights and sizes for idols and shrines which are both acceptable and auspicious from the religious point of view; the interpretation of divine omens and unnatural phenomena; the misfortunes which will occur if the statues are disproportionate or made incorrectly; 'correct' proportions and measurements for statues of deities, demi-gods, demons, humans, birds and animals; the correct placements for various idols, the 'auspicious' and 'inauspicious' lines of vision of a statue, and so forth. Taken together, these three chapters underscore the fact that a sculpture had to be prepared in accordance with established conventions for it to be sanctified as an idol.

Mandan's Devata-Murti-Prakarnam: Elucidating a Tradition. Mandan, also known as Sutradhar (or 'Artisan') Mandan, was one of the architect-artisans at the court of Mewar's renowned ruler, Maharana Kumbha (r.14331468 AD). Kumbha was a scholar and writer, patron of the arts, and an acclaimed warrior who was successful both in protecting his ancient ancestral kingdom from the frequent attacks and invasions of neighbouring states, as well as in its territorial expansion. Tradition holds that 32 out of the 84 famed forts and fortresses of Mewar were constructed or renovated during Kumbha's reign.

For an archaeologist, such chapters may appear as mere words to be read through rapidly, before getting to the 'informative' part of a manuscript, where (one hopes) iconography is detailed, or other 'useful' information made accessible. There is a cautionary aspect that needs to be underlined here. These first three chapters of the DMP contain much in the form of ideology and beliefsystems that is not usually easily available to an archaeologist. The specific nature, in this particular case, that ideology or religio-philosophical beliefs is less important here as compared to the salutary lesson it underlines; namely, that as archaeologists our information-base necessarily chases an elusive element locked in the hearts and minds of past societies. As such, while archaeology needs caution in reading too much or too little into the religious aspect, we also need to explore alternative ways of tapping into that information-base, and of keeping our options open for doing so.

Included in the architectural legacy left behind by this ruler are the fortresses of Kumbhalgarh (Figure 1), Achalgarh, Chittorgarh, the Ranakpur and Eklingji temples, the Victory Tower at Chittorgarh, and numerous other forts, palaces, temple-complexes and structures across his domain, as well as in areas conquered by warfare. To Mandan goes a considerable part of the credit for designing and controlling construction at some of the most notable of these including the forts (containing public buildings, waterreservoirs, temples and palaces within them) of Kumbhalgarh and Achalgarh, and the Vishnu temple at Eklingji (Figures 2-3). The fame of Mandan has survived for more reasons than his simply being a master architect-artisan, though. For, he also set down in written form contemporary prevalent rules for iconography, town-planning and architecture in several treatises. These include his Vastu-Mandan, Rupa-Mandan, Prasad-Mandan, Raj-Vallabh, Vastusaar, Vastushastra, Apa-tattva, Rupavtar and Devata-MurtiPrakamam.

The remaining five chapters of Mandan's DMP deal with the portrayal of specific deities. Thus, chapter four describes the iconography of Surya the Sun, Brahma, the four Vedas, etc. Many of the images described here - for instance the 'Nine Planets' - are no longer idols of primary importance to a 20th century Hindu worshipper. However, several of these remain easily recognisable to a devotee well-versed in traditional mythology and imagery, whether she/he be a theologist, or non-literate.

Mandan' s manuscripts were based on earlier established canons (like the ones mentioned above), as well as the contemporary working tradition of medieval Rajasthan and western India in general. The latter is described in works like the c.10-11 th century AD Aparajit-Prachha of Bhuvan-Dev, the Samarangana Sutradhara by the Parmar dynasty King Bhoj of Dhar, and the 13th century

Chapter five focusses on different idols of Vishnu and companion statues. These include the 24 emanations (vyuhus), and forms like Trivikram, Anant etc. Shalagram stones (ammonites) venerated by Vaishnavites are also discussed. Surprisingly, the famous 'Ten Incarnations' of Vishnu are merely referred

75

Figure 2. Ruins of a Jain Temple (Kumbhalgarh, c.15th century AD)

Figure 3. Remains of the so-called 'Siva Temple' (Kumbhalgarh, c.15th century AD) The iconography of the 24 Jain Tirthankars features in the seventh chapter. The continued importance of Jainism in the 15th century (unlike Buddhism, which had ceased to be an independently recognised beliefsystem in Rajasthan nearly a millennium earlier), is reflected in Mandan' s statement in the DMP that Vishnu, Brahma and the Jain Tirthankars are of the same nabhi or origin. This declaration is far from traditional as far as the older texts are concerned, and appears rather in the nature of an attempt to consolidate and foster the established relationship between Hinduism and Jainism in Western India (Figures 6-7).

to in passing in the DMP, though their iconography is not described (which it is in Mandan's Rupa-Mandan, where Buddha is the ninth of the ten incarnations). The sixth and longest chapter of the Devata-MurtiPrakarnam mainly describes the iconography of Siva. In addition, this chapter discusses composite or syncretic statues combining the attributes of two or more different deities in one single idol. Among such composite statues listed in the DMP are the idols of 'SivaNarayan', 'Krishna-Karttikeya', 'Krishna-Shankar', and 'Surya-Hari-Hara-Pitamah'. Such images combine the iconography, variously, of deities like Siva and Parvati, Siva and Vishnu, or Vishnu, Siva, Brahma and Surya, among others. Different types of Siva-lingas are also discussed (Figures 4-5).

The chapter also discusses each of the respective male and female attendants, known as yakshas and yakshinis of the 24 Tirthankars. Yakshas and yakshinis were well-

76

Figure 4. Karttikeya, the son of Siva (Kumbhalgarh, c.15th century AD) established as subsidiary deities within Jainism by this period. Sometimes a Hindu goddess was transformed into a Jain yakshini, retaining her original name and iconographical specialities, but with a different mythology unconnected with her Hindu origins (Banerjea 1955:299).

from it which have general implications for how archaeologists consider the 'religious' material they encounter. A text like the DMP, together with contemporaneous iconography, is useful for viewing the increased or decreased importance of cults over time, for instance. The DMP contains details for depicting different idols of Brahma and Surya. Brahma remained important in Rajasthan and Gujarat up to c.1300 or so, despite ceasing to have a cult-following in other parts of South Asia by about the 7th-8th centuries, (in fact, the only 'practising/ living' temple of Brahma in northern India today is located at Pushkar in Rajasthan). In the case of Surya too, the solar cult remained popular since about the 1st centuries BC-AD, when early idols of Surya appeared in the Kushan period, though the familiar standing image (wearing calf-length boots) became common in north Indian iconography from the Gupta period onwards. The solar cult was extremely popular for more than a millennium, with a profusion of images and temples raised to Surya, particularly between c. AD 600-1400. Neither Brahma nor Surya enjoys preeminence as a major deity today, however, even though no religious text has advocated a decline in their importance.

This penultimate chapter also mentions the constellations and zodiacal signs associated with the birth of each Tirthankar. This is the only chapter in which this kind of part-astronomical and partastrological information is provided, possibly because the Jain Tirthankars (like Buddha and the Bodhisattvas) are viewed as great path-finders or teachers who were born as mere mortals, and subsequently achieved salvation and became venerated. The eighth and final chapter of the DMP mainly describes the idols of various forms of the 'Great Goddess'. .Idols of Ganesh, Karttikeya, Kshetrapala, etc. are also detailed. Mandan's Devata-Murti-Prakarnam thus reflects both the established iconographical tradition of the region, as well as the related religious and philosophical transitions that were further made within and amongst different cults during the 8th to 15th centuries.

Interpreting a Canon: DMP and the Correlation of Texts, Iconography, Archaeology and Religion

Similarly, the DMP describes idols like Vishvarupa, Trivikram, Varah, Narsingh, the Eleven Rudras, Indra, Harsiddi, and the Ashta-Matrikas. Sculptural and architectural remains of Mewar shows that in the case of some of these, there existed separate temples where their

Mandan's DMP provides us with a vivid picture of 15th century regional beliefs. As such, several issues emerge

77

Figure 5. Ambika (a mother-goddess) holding an infant (Kumbhalgarh, c.15th century AD) idols were installed as the main deity of that shrine, but their worship later declined. Their detailed treatment in the DMP and contemporaneous sculptures from Chittorgarh, Achalgarh, Basantgarh, Kumbhalgarh, Eklingji and other parts of Mewar indicates, however, that they remained venerated during Mandan's period.

Such stanzas underline the importance of the sacred or religious aspect in the physical lay-out of towns and cities of medieval Rajasthan. If this was - as implied an important element in the making of a social/ religious landscape, and therefore in the planning and building of cities and towns, it is a factor which archaeology needs to consciously focus upon, and identify tools to handle.

The DMP also provides time-specific information about some contemporaneous ritual activities. This includes festivities like Devata-yatra - or processions of idols, and ritual-practices accompanying the consecration ceremony of a statue or a Shiva-linga. These find only passing reference, though, and are alluded to without any accompanying delineation of the actual activity. There are precise directions too about aspects like the directions shrines of specific deities should face, placements of idols, of other deities in relation to these images, and so on. Wrong or injudicious placement of an idol or shrine, the DMP emphasises, deprives devotees from receiving that divinity's benediction and protective care.

The DMP underlines the fact that adhering to the established - or textual - sculptural tradition was what was desired from an artisan rather than creative individualism. The very first chapter of DMP warns a sculptor to conform to the prescribed size, scale, proportions, and so forth, since deviation would result in calamity (ch. 1, st. 27-39). Mandan warns, for instance, that ' ...an idol with smaller limbs and a large head will bring destruction of its donor and one with disproportionately large limbs will cause the destruction of the sculptor' (ch. 1, st. 30). Such injunctions probably ensured that artists paid due attention to scale and form.

78

Figure 6. 'Parikar' (canopy) for a Jain deity (Kumbhalgarh, c.15th century AD)

Figure 7. Four-armed Jain 'Shasan-Devi' (Kumbhalgarh, c.15th century AD)

79

like the 'sacred' areas of Kurukshetra, Varanasi, Prayag, the confluence of rivers like the Ganga, Yamuna, Saraswati, Narmada etc., Kedar and Prabhas. All of these form part of the Hindu pilgrimage circuit today. Mandan mentions rituals performed at such pilgrimage sites, thereby providing further insights to medieval beliefs.

Such strictures help illuminate established convictions held by 15th century AD Rajasthan. The DMP states that famine, pestilence and other calamities can result both as a consequence of an artist making a mistake while sculpting an idol, and also when the populace commits evil deeds, thereby displeasing the gods (ch. 1, st. 27-38, 40-61). This, declared Mandan, ' ...results in the angered gods sending different kinds of catastrophes and misfortunes, which are manifested through omens and· changes in terrestrial-related (bhoum), sky and cosmos-related (antariksha) or supernatural (divya) phenomenon' (ch. 1, st. 41-42). Referring to 'the past collective knowledge of the learned', Mandan then lists various omens which are harbingers of divine wrath, and needed to be propitiated by a ruler through pacificatory prayers and other rituals, (not described in the DMP), besides the donation of gold, cattle, land etc (ch. 1, st.45).

One aspect of medieval iconography which becomes apparent from Mandan's treatise is the extent to which detailed directions are provided for the user/ reader of the DMP. These range from determining the nature and even gender - of a particular stone to ensure whether it was appropriate for sculpting the idol of a god or goddess; to providing the 'proper' attributes to an idol in the 'correct' hands and prescribed manner. These instructions were intended for someone already familiar with, and working in, the established tradition. Thus, very often, while the ornamentation of a particular idol is detailed, something as crucial as whether the statue is in a seated or standing pose is not referred to in the text. To my mind, this raises certain questions regarding why and for whom texts on iconography were written down, and who utilised them?

Such stanzas, while providing yet another aspect of beliefs prevalent in medieval Rajasthan, also underline a major problem faced by archaeologists in our attempt to understand past religious phenomena and its manifestation(s). For, while the literary mention of such beliefs in texts like the DMP act as an aid for our understanding a particular set of religious beliefs being referred to, these are not issues that would naturally occur to an archaeologist or art historian, analysing (and admiring or disparaging, as the case may be), the proportions, or angle, or scale, or iconic peculiarities, of statues and temples. Certain facets of the belief-system, unfortunately, need not be overtly apparent in the archaeological record - leading, again to a possible distortion in interpretation through over or underattribution of motives. And as such, our interpretation can be influenced in directions other than those originally intended. Thus, a text like DMP, used judiciously alongside archaeological work, can become a medium for investigating aspects of religious change or shifts in iconographic representation and importance.

It may be important to stress here that we do not come across iconographical treatises written in Rajasthan prior to the period of Mandan and his patron, Kumbha. Mandan's treatises appear to have been the first such texts composed in Rajasthan. That does not mean that the art and architecture of preceding centuries did not follow the basic iconographical canons of India. Given that in Mandan's age there existed a working tradition, common with contiguous areas of western, central and northern India, and that Mandan merely carried forward this established working tradition, the question of 'why?' he wrote DMP and the other texts becomes a moot point. Moreover, most artists worked from acquired knowledge, rather than from a Sanskrit text (a language not every artist either knew, or was expected to know). Why, then, did Mandan need to write down an existing iconographical tradition that was in practice, particularly as iconographic innovation was neither expected nor encouraged?

Texts like· the DMP throw light on issues like socioeconomic conditions too, much of which has relevance for the archaeology of religions. For instance, the DMP provides details regarding which specific images of Vishnu and ammonites could be worshipped by people belonging to the four castes of Brahmin, Kshatriya, Vaisya, and Shudra; by groups called Bhils, Medas and Kirats (all regarded as forest-dwellers or people beyond the pale of society, for whom the four-caste social system did not apply, but who were not 'foreigners' or 'aliens'); by distillers and vendors of liquor, clothlaunderers, actors, itinerant performers and dancers, etc.; or by Varuts (outsiders, 'barbarian foreigners'), etc., respectively (ch. 5).

There are a couple of points that may be pertinent to note here. One is that a sculptor visualised and meditated upon the image to be sculpted, at least in theory, so that the true essence of the concerned deity could be reflected through the finished idol (traditional artisans say they still follow this procedure). The other is that (as mentioned towards the beginning of this paper) Kumbha's reign saw a tremendous amount of construction and renovation, including of temples, in his domain and new areas occupied by Mewar. This task probably required the deployment of a greater number of artisans and master-sculptors than had hitherto been necessary over the past few centuries. However, all of them had to work within the religiously sanctioned or acceptable iconographical tradition.

Besides describing aspects like society, economy, and so forth, the DMP also serves as a window on the beliefs and ritual-practices of medieval Rajasthan. The details about constellations, planets and the zodiacal system, for example, serve as a possible pointer for archaeology regarding this, peripheral, aspect of belief-systems.

Mandan's texts may have been meant, therefore, to provide artists working in different areas, and under a number of different architects and supervisors, with a descriptive iconographical guide that enabled all the artists engaged on a particular temple to visualise the

Other information one can gain from the DMP are perhaps references to various sects and cults (e.g. of Lakulish and 'Eka-Dandins'), and places of pilgrimage

80

Kumbhalgarh. The DMP and other texts by Mandan help in carrying forward the concept of this type of temple-complexes, since a lot of construction was carried out in accordance with Mandan' s instructions.

image to be made, while at the same time ensuring that an accepted or correct scale, imagery and iconographical convention was applied throughout the land. Probably texts like Mandan's also served to collate and give legitimacy to newer trends that had entered the working tradition since the previous canons had been written.

This was an age when syncretistic images and multishrined temples were being manufactured. Numerous sects were also flourishing. In addition, there were many who revered locally deified folk-heroes and saints like Goga-ji, Ramdeo-ji, etc. who had lived between the 12th and 15th centuries AD (though their iconography did not figure in traditional canons, idols were gradually made of such deified figures, based on existing iconographical norms). The influence of the Bhakti and Sufi tradition was pronounced all over Rajasthan (and India) too around this period. The end of the 14th and beginning of the 15th century had seen several Bhakti poet-saints repudiate the ritualism and caste-rigidity of Brahmanical Hinduism. Rajasthan also had several centres where Sufi mystic saints - like Khwaja Shaikh Muin-ud-din Chishti and Shaikh Hameed-ud-din Nagauri - had settled from the late 12th century onwards.

Significantly, texts written by Mandan were known outside Mewar (and Rajasthan, in general). Within two centuries of their being set down in writing over half a dozen copies of his works were in the library of one Kavindra-acharya, a Deccan Brahmin of Benares, from where they spread westward and eastward. Manuscripts of Mandan's texts also spread across Central India to reach parts of South India (Mitra 1936:3-4).

Transcreation, Syncreticism and Changing Realities The blending of different cults and religious traditions by this period, leading to the development of composite images of Vishnu and Siva, or Surya and Siva, or Surya, Vishnu, Siva and Brahma; and to the occurrence of combined temple-complexes like the Panch-ayatana form is reflected in the iconographical descriptions available in the DMP. Mandan himself noted in the DMP that 32 types of idols can be made by combining the attributes of Vishnu and Siva, adding that he was not detailing all of them to prevent his text from becoming too long, referring the 'wise' users of his book to learn about such composite forms from certain other named texts (ch.6, st.59). It may be worth noting that as far as the iconographic tradition of northern and western India is concerned, at least one syncretic image - that of 'Surya-Hari-Hara-Pitamah' (the Sun, Vishnu, Siva and Brahma) - occurs in Rajasthan and some parts of Central India and Gujarat only (Rangarajan 1990:154).

Religion (and society) was, thus, in a state of flux and fusion. Maharana Kumbha himself ruled as the 'Steward' of Lord Eklingji - i.e. Shiva, but had a special place in his heart for the Vishnu cult. He also revived the declining (though scripturally lauded) ancient practice of Vedic style oblations (yagnas) performed through sacred fire-altars, by having such an altar built in Kumbhalgarh, where he regularly performed yagnas (the altar and building is extant, though added to and altered over time). Traditions hold that Kumbha also favoured Jainism. These multiple belief systems, and the patronage extended to them by rulers, may have had some impact on traditional iconography and architectartisans like Mandan.

Alongside occasional appropriation of the sanctuary of one religious cult by another and partly altering the earlier religious symbol, there also developed a syncretistic attitude in religion reflected in the appearances of composite images and multi-shrines temples (Bhattacharya 1955:328). Thus, the panchayatana shrine, a ' ...modified monolatry, which maintained the supremacy of the particular supreme deity of the community while at the same time admitting the existence and right to worship of other gods ...made its appearance ...harmonising monotheism with polytheism and allaying communal bitterness and religious quarrel' (ibid:328-329).

From the point of view of an archaeological understanding of the issues of religion, the DMP helps in framing questions to examine the process of syncreticism. This has relevance for understanding contexts other than Hinduism. For example, the processes between Christianity and local religious traditions; the Sufi saints and their interpretation of Islam; Buddhism and the assimilation of local practices outside of India; Roman religious beliefs and the incorporation of local beliefs during the expansion of the Roman Empire. The text also throws light on notions of 'sacred space' and religious activity. Indirectly, and certainly without meaning to, it leads one to ponder the building of literally thousands of shrines and temples during Kumbha's reign across his kingdom and the areas conquered by him. Contemporary sources speak of Kumbha as being a dramatist and writer, a linguist, a great builder - including of public works, and wellversed in Indian philosophy, metaphysics, logic, mathematics, grammar, music etc., whose court was the centre for scholarship and the arts. He gave liberally to various religious cults, and remains famous for his battles against the Muslim Sultans of Gujarat and Malwa, and fellow-Rajput rulers within Rajasthan. There is every indication that besides a religious 'renaissance' of Hinduism under Kumbha, religious

The Panch-ayatana, as the name indicates, is a specific type of temple-complex in which five (panch) shrines (ayatanas) of Siva, Vishnu, Shakti, Surya and Ganesh are provided within a complex. This type of building has one central shrine dedicated to one particular deity, with four other shrines, dedicated to the remaining four deities, around it. The idol placed within the central shrine denotes the main deity of that particular panchayatana. The statues of Siva, Vishnu, Shakti, Surya and Ganesh have a predetermined position within the panchayatana in relation to whichever of the five deities is placed in the central shrine. A fine example of the panch-ayatana temple-complex, which was designed by Mandan and built following his directions, is at

81

Figure 8. Transformation of a Tradition: Reconsecration of a disused idol as 'Bhairon-ji' (Kumbhalgarh, 1995) act1V1tyand building was also used as a device for emphasising the ruler's political power. Thus, it leads one to review the awe with which religion has been regarded by archaeologists, and our effort to separate what is sacred from what is 'secular' in the past.

The DMP discusses aspects of religious beliefs not easily recoverable from material remains, for example auspicious times and methods for selecting and extracting stones worthy of making into idols; which desecrated idols may or may not be re-consecrated. For an archaeologist, such references underline a major problem area; namely the baffling task of attempting to interpret past belief-systems from an inadequate database.

Conclusions

To sum up, Mandan's Devata-Murti-Prakarnam highlights possible problems that archaeologists face while looking at religion. At the same time, it provides a time-specific information base for comprehending some of the religious beliefs and practices of medieval Rajasthan. Besides providing the 20th century researcher with a 'traditional' text, it also demonstrates how traditions are/can be transformed and take a newer, slightly modified form, while retaining their 'sacred' character.

A connected issue is that of recovering and interpreting archaeological data regarding transient religious practices and rituals. This again has universal application. Examples from India include religious processions; ritual bonfires on occasions like Holi, which are lit at different sites each time; open-air worship on occasions like pre-construction groundbreaking rites to propitiate deities; the public burning of effigies in the annual autumnal Dassehra re-enactment of an episode from the epic Ramayana symbolising the

82

victory of good over evil. In addition, there are various impermanent idols made from perishable material for a specific purpose (Gangaur, Durga-Puja etc.), after which they are ritually immersed in water or similarly set aside. Similarly, shifts in cult popularity may be influenced by considerations like deliberate mending of religioncentred breaches, political necessity, royal/priestly/elite patronage, accumulation of wealth and prestige by a particular group leading to its enhanced popular attraction (Figure 8).

around him too, which find expression in the sculptural and architectural tradition that evolved in response. As such, a text like the DMP not only throws up a number of issues and questions pertaining to the archaeology of religion, it also reflects how a text on iconography may encapsulate both the existing religious traditions at a specific time in human history as well as the transformations in those traditions. Acknowledgements

Many of the major religions known to India have been practiced in Rajasthan; sometimes concurrently. While there exists a range of archaeological, epigraphical, literary, and architectural data indicating when and where various religions and/or cults were practiced, the question regarding why one religious practice gives way to another, and the process of transcreation and change has been inadequately tackled through archaeology. Similarly, certain deities mentioned in texts like DMP are not in favour in the late 1990's, though their divinity is still acknowledged. Once again the material record does not help in understanding why some deities gained and others lost out in popularity.

I would like to acknowledge the role of the British Council and St. John's College, Cambridge, in facilitating my participation at the Cambridge Conference. I would also like to thank the Archaeological Survey of India for permisssion to photograph at Kumhalgarh.

References:

Agrawala, R.C. 1996. Rajasthan ke Jain-Devi Sacchika. (In), Dube, S.N. (ed.), Religious Movements in Rajasthan: Ideas and Antiquities. pp.154-158. Jaipur: Centre for Rajasthan Studies, University ofRajasthan.

For much of the above, the material remains may not be easily retrievable and understood through present archaeological methodology. How can such information be recovered? Can texts (like DMP) help archaeologists in arriving at better targeted queries and methodologies? The DMP acts as a time-specific information-base too which may be used judiciously. One may, for instance, look at the 'continuity' in tradition exhibited between it and a late 20th century example, the recently built Swami Narayan temple in London. The temple adheres to approved religious iconography and architecture, and contains many statues that would have been instantly recognisable to Mandan. Many of these images could even serve to illustrate the DMP! At the same time, there are some statues, for example of Rama's brother Lakshman, and of Sita with Rama, that are not described by Mandan in DMP. These are, however, very much a part of the contemporary iconographical tradition. A text like DMP can be useful, therefore, for shedding light on traditions as well as the transformations within it leading to the establishment of modified iconographic traits, which in tum can prove a helpful tool for archaeology.

Banerjea, J.N. 1955. Iconography. (In), Majumdar, R.C. (ed.), The Age of Imperial Kanauj. Vol JV. pp.296-300. Bombay: Bharatiya Vidya Bhawan. Bhattacharya, H.D. 1955. Minor Religious Sects. (In), Majumdar, R.C. (ed.), The Age of Imperial Kanauj. Vol IV. pp.327-352. Bombay: Bharatiya Vidya Bhawan. Bussagli, M. n.d. Introduction. (In), Bussagli, M. and Sivaramamurti, C. (eds.), 5000 Years of the Art of India. pp.7-41. Bombay: Tulsi Shah Enterprises. Hooja, R. 1998. Mandan 's 'Devata-Murti-Prakarnam '. Jaipur: Prakrit Bharati Sansthan. Mitra, H. 1936. Introduction. (In), Sankhyatirtha U.M. (ed.), Devatamurtiprakaranam and Rupamandanam. pp. 1-4. Calcutta: Metropolitan Printing and Publishing House Ltd. Rangarajan, H. 1990. Spread of Vaisnavism in Gujarat Up To 1600 AD (A Study with Special Reference to the Iconic Forms of Visnu). Bombay: Somaiya Publications Pvt. Ltd.

In conclusion, it is important to emphasize that the DMP is not (and was never intended) as a text of revolutionary innovation that deliberately and calculatedly sought to overset previous sculptural traditions. Rather, Mandan drew upon and carried forward earlier, established, iconographical traditions. The DMP (and Mandan's other treatises) mirror the transformations in the world

Sankhyatirtha, U. M. (ed.), 1936. Devatamurtiprakaranam and Rupamandanam (Manuals of Indian Iconography and Jconometry). Calcutta: Metropolitan Printing and Publishing House Ltd.

83

The Archaeological Visibility of Caste: An Introduction. Robin Coningham and Ruth Young Department of Archaeological Sciences, University of Bradford. Arthasastra of Kautiliya (Kangle 1965). Traditionally accepted as a practical manual of information for a ruler written in the third century BC, it advocated the spatial separation of groups within cities (Figure 1). Brahmans were to be in the north, Ksatriya in the east, Vaisya in the south, Sudra in the west and outcastes outside the city (Arthasastra 2.4.7-23). Further codif-ication of the regulations of access to dress, food, habitation and social and sexual relationships for each group was recorded in the Manavadharmasastra or 'Laws of Manu' which was compiled between the second century BC and second century AD (Buhler 1886).

Introduction

Caste is one of the most powerful dynamics ordering social space within settlements in south Asia. It is a dynamic which controls marriage partners, ritual purity, diet, occupation and the location of habitation. Present in rural and urban settings, it is a dynamic which is found in Hindu, Buddhist and Islamic communities throughout the subcontinent, from Sri Lanka to Nepal, and from Bangladesh to Pakistan. A recognisable modem phenomenon, attempts have been made to demonstrate that it forms a very real and pervasive link with the oral and written histories of south Asia. Such a powerful dynamic has not been ignored by archaeologists and historians; a number of such scholars have attempted to trace and identify it in architectural and artefactual distributions, ranging in date from the Early Historic to the Prehistoric (Auboyer 1965; Jarrige and Santoni 1979; Ratnagar 1991). This paper will evaluate the success of these attempts, frame a methodology for testing two archaeological indicators occupation and diet - and their application to archaeological evidence, and examine the archaeological visibility of caste.

It is also noteworthy that whilst some scholars, such as Leach (1960), have argued for a definition of caste as a purely south Asian social phenomenon, others, most notably Barth (1960), have argued that it is merely a local term given to a universal form of social stratification. These problematic and well-discussed definitions aside, one of the most clearly presented working definitions of caste was produced by Hutton (1946: 49), (i) A caste is endogamous. (ii) There are restrictions on commensality between members of different castes.

Definitions

The very definition of caste is problematic as the word is derived from a sixteenth century AD Portuguese term used to describe trading communities on the west coast of India. In this, its first application to south Asia, it meant nothing more than species or breeds amongst animals or plants; and tribes, races or lineages amongst humans. This simple meaning has become increasingly distorted and now is often used, as demonstrated by Quigley (1993), as an interchangeable translation of two very different south Asian terms varna and jati. The former term, vama, refers to a four-fold division of society into Brahmana (priests), Ksatriya (warriors), Vaisya (merchants and craftsmen) and Sudra (labourers) which ensures social harmony and cosmic stability; the latter, jati, refers to a shared common origin through birth (ibid.). A clearly recognisable phenomenon in contemporary south Asian communities, a number of attempts have been made to demonstrate that caste forms a very real and pervasive link with the oral and written histories of the region. The earliest reference to this four-fold division is found in the Purusa-Sukta or 'Hymn of Man' of the Rig Veda (O'Flaherty 1981) which is traditionally dated to the first millennia BC:

(iii) There is a hierarchical grading of castes, the best-recognised position being that of the Brahman at the top. (iv) In various kinds of context, especially those connected with food, sex and ritual, a member of a 'high' caste is liable to be 'polluted' by either direct or indirect contact with a member of a 'low' caste. (v) Castes are very commonly associated with traditional occupations. (vi) An individual's caste status is finally determined by the circumstances of his birth, unless he comes to be expelled from his caste for some ritual offence. (vii) The system as a whole is always focused around the prestige accorded to the Brahmans.

His mouth became the Brahman; his anns were made into the warrior; his thighs the people, and from his feet the servants were born. (Rig Veda 10.90.12)

Whilst such clear-cut attributions have been demonstrated to be more flexible in a number of cases (Banks 1960; Yalman 1960; Quigley 1993), archaeologists and historians still rely heavily upon such definitions, and we shall continue to use the term caste.

However, the earliest evidence to suggest that this division was spatial as well as social is only found in the

84

Crematorium for higher varnas

11 ___

___

Brahma Gate

Jewellers, workers in metal

~-------.--------,1 Ii------.---Treasury

BRAHMINS

Hospitals

11 Royal Temple

Storehouses i_.._

s

u D R A

s ■

•

Senapati gate

__

_,__

_

__.

Chariots and Carriages

11

'-----'------'

Royal Residence (Antahpura)

Stables for camels, donkeys. Palace Workshop

City Ce~tre

ARMOURY Storehouse for forest produce

Horses, Cows

K Teachers Purohitas Councillors Ritual Areas Water Storage Kitchen Granary Elephant Stables

Commodity Warehouse Records/Audit Office Palace Workers Qrs.

'

I

VAISHYAS

s

H A T R I

y A

Perfumes Flowers Liquor Store Toilet Articles Many Articles

s gate

l

l

11

Ir

Guilds (Washermen. Tailors, etc.) Foreign Merchants

Grain Merchants Other Merchants Factory Officials Control Officials

Guilds

Bars

Restaurants Entertainments

Courtesans Yama gate Crematorium for lower varnas

Figure 1. An idealised plan from the Arthasastra (after Rangarajan 1992) Caste in contemporary south Asian communities

i Aristocratic managers and administrators, officer and courtiers ii Tenant farmers and village labourers iii Craftsmen and traders.

Archaeologists are often tempted to use the present to make sense of the past, and this is certainly the case for the archaeology of caste. However, before presenting a number of archaeological attempts to identify caste in the past, it is useful to illustrate two 'classic' contemporary communities in which, in Leach's words, caste 'defines the structural role of every sector in a total organic system' (Leach 1960: 10). The first example is that of the village of Kumbapettai in Tanjore District of southern India as recorded by Kathleen Gough in the early 1950s (1960; 1981). This village of 962 people in 199 households was divided into three main groups:

C. Adi Dravida or 'Original Dravidians'. Gough also recorded seven main ways in which these three groups were differentiated:

A. Brahmans

(i) Habitation location - Brahmans lived in the north of the village (1), non-Brahmins in the centre (2 & 3) and Adi Dravida on the western and southern outskirts separated by paddy land, streams or cremation grounds (4 & 5) (Figure 2).

B. Non-Brahmans

(ii) Building materials - Brahmans lived in

85

Inter-dining between these groups was impossible.

brick and tiled houses, non-Brahmans in large mud and thatch houses with door and windows, and Adi Dravida in small mud and thatch shacks with no windows.

(vi) Intimate interaction - With the exception of Barbers and Midwives no group is allowed to touch persons of a higher category. Sex between Brahman men and Pallas women is strongly condemned whilst a case of a nonBrahman man who had sex with a Brahman woman resulted in his castration and murder.

(iii) Access to space - Adi-Dravida were not allowed into the Brahman street, 'polluting' non-Brahmans are allowed into the Brahman street but not the houses and 'clean' non-Brahmans were allowed into Brahman houses but not the kitchens. Adi Dravida may enter the streets of nonBrahmans but not the houses and viceversa.

(vii) Cremation ground location Brahmans are cremated at (6) and nonBrahmins at (7), whilst the Adi Dravida are buried at (8).

(iv) Occupation - despite the appearance of the modern economy, Gough recorded that, 'The spatial organisation of castes corresponds in large measure to their occupational specialisation and mutual ritual rank' (ibid: 18).

Similar social and spatial differentiation has been found within non-Hindu rural settlements elsewhere in south Asia as illustrated by Yalman amongst a Buddhist community in Sri Lanka (1960) and by Barth within an Islamic community in northern Pakistan (1960).

(v) Diet - Brahmans were vegetarian, nonBrahmins and Adi Dravida were meat eaters, although only the Pallas ate beef.

..., ... "6 ' .,. ,,.,. .,.

Sr,1hrn.n ,,..111;11t1011 '~

,'

,'

,✓•

:na,tld

Palbn w,1t.:/1mo1n

0----------------------------------

I

/ O Pallan watchm,1n I

018

-1-'

\

;Nc,n.=.n.nman

... _ " cnmat,on ground

Paddy f,eid:, I I

'I I

----,•,-~=~=~~!--1E~--------~Upp,ar """C. Widdla I l~Sertton ~Sm.-.S~ I

1

I I

: -.....,.-_

0

- - ---

!rr,tit.on dl.;..,,.C'I

5.-thlnl pool i=ootp,at.h

fr

I

•

,-..,

\....

.,,_,. Stnwc

cmmmJIIIIID(?=

CIIIIIIIIIJ CitIID CIIDlIIIIl Oc,.,endr:ol:.hllan nre■ t.:.

cmmmJ

l ....,,_

u

\. _ _..)Uppu- S~t ._ -

~M;aini-o..:I

:'

rmDlllll

crcm11CMJftcrownd

Shr-,n~

Figure 2. Plan of Kumbapettai (after Gough 1960)

86

,--~LowerS~t \- ✓ c,.,...,.~n

0 "'°"'net

10

20

30

Figure 3. Regimented grid-iron street pattern at Mohenjo-daro (after Marshall 1931) this social organisation, presumably caste, closely with the arrival of Indo-European languages and Vedic Hinduism. 'It is not due to chance that these important transformations which take us from the Chalcolithic cultures of the third millennium to the rural society of the historical periods are set against the spread of a language that was to remain the chief vehicle in India and a religion that was to serve as a framework for later religious developments' (ibid.).

Caste and the past

Such a visible social structure has, of course, been used by archaeologists and historians as a model for the functioning of past societies as well as by social anthropologists and sociologists in attempts to explain present behaviour. As early as the 1920s the Bronze Age Harappan, or Indus, civilisation was noted for the absence of royal palaces and tombs and the presence of an extremely strong conformity of material culture through time and space which Wheeler referred to as 'disciplined and even regimented' (1953: 40) (Figure 3). Whilst he did not actually suggest that this was a result of the caste system, Kolenda asked the following question in 1978: 'There were occupational specialists in Harappan society, but nothing is known of their social organisation ... were they closed endogamous descentgroups like the later jatis?'. This line of approach has also been followed by Malik (1968) and Gupta (1974) who have suggested that the stability and continuity of material culture within the Mature Harappan was a result of the presence of the caste system.

The final archaeological example considered here is set in the Early Historic period of south Asia which dates between the beginning of the first millennium BC and the fifth century AD. As the Arthasastra and Manavadharmasastra were compiled during this period, many historians and archaeologists have assumed that such social and spatial planning were physically realised, regardless of the absence of actual excavated evidence. Thus Rowe stated that: 'from the very earliest times, the Indian city has provided a symbolic representation of the social order, both in its spatial arrangement and in its social structure'. Gough provides a similar statement, 'from about the tenth century BC, the kingdoms of India all possessed five main orders ... For three thousand years, these orders and their attached specialists formed the backbone of the caste system and underlay the basic principles of religious ranking. Only with the rise of capitalism are they being eroded' (1981: 26). The historian, Auboyer, writing in the 1960s stated that, 'it does seem, though, that the capital was in fact divided into a certain number of well-defined districts ... In such a scheme, each caste is supposed to have occupied a particular district' (1965: 120). More recently, the archaeologist Erdosy, on the basis of a study of caste within modem village plans, has further suggested that,

Jarrige and Santoni have made a further case for the presence of caste at the Chalcolithic/lron Age site of Pirak in western Pakistan. Here the excavators identified a craft zone within the settlement dating to between the eleventh and ninth century BC. They stated that, 'the division into quarters, the existence of a block of buildings where craftsmen seem to have lived in a community, the stereotyped character of the layout of the houses ... and the conservatism of the material culture throughout almost ten centuries are factors that suggest an organisation into social classes ...just as in the present day villages' (Jarrige and Santoni 1979: 411). They link

87

'as the social organisation operated at both village and city levels and as the idealised plan of the ancient texts themselves were meant for both types of settlements, we may expect to find our hypothesis drawn on the basis of village plans to be confirmed in cities' (1986: 156). •2

The archaeological visibility of caste: towards a methodology

•l

Whilst many archaeologists, historians, sociologists and anthropologists have thus argued for the presence of caste in the past, there have been very few suggestions of how one would actually identify it. Indeed, most seem prepared to side-step the issue as illustrated by Ratnagar who stated that 'there may have been a caste system in Harappan times ... But even if this were detectable from archaeological remains, it would leave unanswered the question of political structure and economic system' (1991: 19). This is indeed a question of how one would identify caste from archaeological remains. Perhaps the solution is provided by Thapar who suggested that 'material culture and cultural differences occurring consistently in distinctive sections of the town may suggest a variation of social grouping' (1978: 205). Surely such a varied urban community, if spatially and socially differentiated, should be archaeologically visible!

N

6--~56oKM

A

Figure 4. The location of Anuradhapura and other sites mentioned in the text. I. Mohenjo-daro, 2. Pirak, 3. Kumbapettai, 4. Anuradhapura

For the purposes of this paper the spatial variability of two categories of material culture - craft waste and faunal remains - from a single Early Historic urban site will be examined. The former is interpreted as being indicative of occupation and the latter of diet, although clearly taboo and other symbolic pressures often distort the deposition of artefacts within the archaeological record (Hodder 1986; Moore 1982). Both selected categories are taken from a single site - the Citadel of Anuradhapura in Sri Lanka, which was excavated between 1969 and 1994 by the Archaeological Survey Department of Sri Lanka and between 1989 and 1994 by a joint Sri Lankan - British project (Figure 4).

Occupation

As illustrated above, traditional occupations are often attributed to caste. Such occupations may be identified in contemporary communities through a combination of actual physical activities, specialist tools, specialist architectural elements and waste products. For example, a Potter may make pottery using a wheel before firing it in a kiln and discarding wasters. Whilst some occupations, such as potting, may thus be easy to identify in the past, others, such a leather-working or shaving, are more difficult due to differential survival within the archaeological record. Furthermore, activities carried out 'off-site', such as washing, are even more difficult to find and identify. For the purposes of the present paper we shall limit our study to two main activities:

One might query the selection of Anuradhapura as a test site as it is located at the southern periphery of the south Asian region, and secondly, because the city was historically a centre of Theravada Buddhism. However, as noted elsewhere, there is clear evidence that the island was not peripheral as it played a pivotal role in trade networks during the Early Historic period (Coningham et al. 1996) and there is also evidence that caste has played an important role in the ancient and modern social organisation of the island (Ryan 1953; Leach 1961; Yalman 1960; Coningham 1994). Finally, Anuradhapura represents one of the most detailed archaeological samples from an Early Historic urban centre in south Asia (Coningham in press). These samples are furthermore important as they are drawn from a total number of 13 trenches excavated at different locations throughout the walled site itself (Figure 5). Trench ASW2 in particular, excavated close to the city's centre, has provided an extremely well-dated structural and artefactual sequence from the city's origin as a small Iron Age settlement at the beginning of the first millennium BC to its place as an imperial capital in the tenth and eleventh centuries AD.

1) Metal-working - as represented by moulds, furnaces, crucibles and slag. 2) Semi-precious stone-working - as represented by blanks and debitage. As discussed in a recent paper (Coningham 1997), we share with Pracchia et al. (1985) the belief that whilst finds of these categories may not represent the precise boundaries of processing, certainly the waste products, such as slag, debitage and cores, are unlikely to have been moved far. Traditional models of Early Historic urban planning (Auboyer 1965; Rowe 1973) indicate that activities were located within prescribed zones of the city, as described by the Arthasastra, or divided into less geographically prescriptive blocks in contemporary rural and urban settlements. We would expect therefore either of two patterns within the distribution of evidence from the thirteen trenches at Anuradhapura:

88

The pattern of the archaeological evidence is, however, very different to either of the proposed models. Whilst the sequence at Anuradhapura stretches from c.900 BC to c.1100 AD, for the purposes of this paper we shall only examine the data from two successive periods within the site, as they are well preserved.

1) Workers of metal and jewels in the north (Arthasastra). 2) A separate zone of craft activities within one area, not necessarily north.

N

t

■ AMP

Figure 5. Plan showing the location of trenches within the Citadel of Anuradhapura

89

Period I dates to between c.350 and 275 BC and Period G/H to between c.275 BC and 200 AD. Of the thirteen localities sampled at the first period, I, a total of seven trenches (ADB, ASWl, ABW3, AGW, ARW, AG and AMP) had evidence of metal-working and semi-precious stone-working waste - giving a very diverse pattern for the distribution of similar crafts within the site. Our larger trench at ASW2, adjacent to ASWl, also showed evidence of metal-working waste, in addition to yielding evidence of bone, shell and antler-working from its more detailed data base belonging to a single structure (Figure 6). Both sets of data suggest a very diverse pattern of the location of craft-working activities within the city. During the succeeding period, G/H, a similarly diverse pattern was found with a total of seven trenches (ADB, ASWl, AG, ABW2, ABW3, AMP and ABWl) providing evidence of metal-working and semi-precious stone-working waste. Again, the largest trench, ASW2, yielded waste products suggesting the presence of metalworking, semi-precious stone working, glass-working, antler and even ivory-working in the same structure (Figure 7). The patterns thus recovered from both periods are clearly at odds with the models forwarded for the Early Historic social organisation of craft activities within urban forms (Coningham 1997) and is similarly at odds with the contemporary ethnographic examples cited above.

Diet

The fauna!' remains from Anuradhapura will also be examined as they are interpreted as being indicative of diet, although as has already been mentioned, taboo and other symbolic pressures often distort the deposition of artefacts within the archaeological record. Our data set is more limited than that for occupation, being reliant upon our own sample from ASW2 and that from a single trench, AG. We are fortunate, however, that whilst the former site is located within the centre of the city, the latter is located within the north-east of the city, thus spatial variation within diet should be clear. Although Deraniyagala interpreted all the bone recovered from trench AG as food waste on the basis of cut marks and the associated ash and charcoal (1972: 155), certain species were evidently also used for traction, haulage, general transport, milk, hides, horn and breeding purposes. We hoped to identify spatial variation within diet between the two trenches, conforming to patterns of modem south Asian communities where selected diets are restricted to caste. Our sample consisted of 1946 identified bones from trench ASW2 and a further 2356 identified bones from trench AG (ibid.). As noted above, further codification of the regulations of diet are recorded in the Manavadharmasastra or 'Laws of Manu' which was compiled between the second century BC and second century AD (Buhler 1886).

debl1age

■ hell

corea deblt~g•

antler

cor••

■ hell

.-~ ••• .. .••:

_..

•

• :·

•

•: '

•:

'

:1

..-~ .,

.....--·--·------------------... .......... ....-----·---··-•·'"'···---

- -- --- - -:

...

·-------------------...., ,;

i. •:

'

I

■ hell core ■

debltao•

:,,: '

crucible

'

,.

I I

I I

I

:

•;

•

•·

•

I

•:

'

e;

•

•: '

■ lag

carbonized

rice

• • 1m

Figure 6. Plan of exposed period I structure at trench ASW2

90

f

,I

, 1'

......

- ... - -

'

:

- - - • -- - -- -.

,- - - ... .,

:

6-~~o~o,)~o &;i.' G§~\s~

\_-----.·

0

--~-=- ____ . _.. :?------_5'.)___ ?:J_o_?~~:'Q@o~~ a ~~- _ -~o ~o~~~ moften

olass

_ ... _ .. - __ J _

-

molten

_

brick~

....

glaa:-~,D

-

- - - -

slao shell

slao

'

,,r----.... : ,

,'

crucible

Ivory

0

,--------"

~

debltag•

_•u0~;;:0a

I

: :: - ...! ,,' ,' ________ ,'

I

- - - tao•

cores

I

antl~LJ

I

-------,.. - - - ....... I

'I' •Jag

-.

''

I

I

slag

:

.......---.......

,

, I I

I '

.

debltage

I

•

CJ

------------·--

..____

'

0

debUage

--------J

,:.

•

.._- .......--------

• ---------- ----

.

alag

N

I

\

•

I

debltage

...

-------..'

~

debltage

: 1m

Figure 7. Plan of exposed period G/H structure at trench ASW2

In comparison with this list it is useful to record which species, allowed and forbidden, which have been recorded at trenches ASW2 and AG:

Reference to the Laws will further assist us in understanding spatial variation within the archaeological faunal record. The laws of Manu, for example, state that any twice born man who knowingly eats village pig or cock will become an outcast (ibid: 172) and that any man who kills a horse, deer, elephant, fish or snake will be degraded to a mixed caste (ibid: 444). It also lists the following as forbidden:

TrenchASW2

1) All carnivorous and one-hoofed animals (ibid: 170). 2) All solitary and unknown beasts (ibid: 172). 3) All five toed animals (ibid.). 4) Animals which eat fish (ibid: 171). 5) Village pigs (ibid.)

91

Forbidden

Permitted

cattle

buffalo

horse

porcupine

pig

hare

deer

terrapin

attempts to reverse this process of Sanskritisation by backward class movements in India as an active attempt to reach a lower grade in order to gain access to reserved government jobs (ibid.). In conclusion, we may thus suggest that this first attempt to test the archaeological visibility of caste indicates that its artefactual manifestations may be more flexible and complex than traditionally assumed and that if we are to study it further we need to modify our methods and guiding assumptions.

Trench AG

Forbidden

Permitted

horse

terrapin

cattle deer pig

Acknowledgements

Whilst it should also be noted that certain forbidden animals can be made acceptable depending on the hunter and the rituals undergone, there is a c.lear pattern of the presence in both localities of species forbidden by the Laws of Manu - moreover, many of those from trench AG are noted as having cut marks. The presence of cattle bones in both areas is particularly surprising considering the major beef-eating taboo in the island which is thought to have occurred after 100 AD (ibid: 157). In summary, a comparison of the major identified faunal species from trenches ASW2 and AG show the same general trends in both the species present and the periods in which they occur regardless of the fact that these mainly consist of forbidden species.

As much of the primary data presented in this paper was collected during the joint British - Sri Lanka excavations at Anuradhapura between 1989 and 1994, the first author would like to thank Dr Siran Deraniyagala, Director-General of the Department of Archaeology, Government of Sri Lanka for his help and advice; and most especially for allowing access to his excavated data and site notes from Anuradhapura. The following bodies are also acknowledged for their financial sponsorship of the fieldwork: the Ancient India and Iran Trust, the Archaeological Survey Department of Sri Lanka, the British Academy, the McDonald Institute for Archaeological Research, the Society of Antiquaries and the Society for South Asian Studies. This fieldwork was carried out whilst the first author consecutively held a King's College Research Studentship, an Ancient India and Iran Trust Junior Research Fellowship and a Lectureship in South Asian Archaeology at the University of Bradford.

Conclusions

We may thus conclude that Thapar's suggestion that a study of the spatial variability of selected material culture within an urban form might confirm the presence of distinctive social groupings (1978: 205) has not been achieved. Indeed, rather than finding clear patterns of spatial variation within the distribution of craft waste and faunal remains we have found broadly similar patterns in terms of the overall distribution of craft activities and the discard of the remains of animal species within a single urban site. In effect, this unexpected result suggests that the formal caste-based models proposed by the Laws of Manu and the Arthasastra were not realized physically. It is suggested therefore that the spatial differentiation which is observed today is a relatively recent phenomenon - the enforcement of a largely symbolic division.

References

Auboyer, J., 1965. Daily Life in Ancient India. London: Weidenfeld and Nicolson. Banks, M. 1960. Caste in Jaffna. (In), Leach, E.R. (ed.), Aspects of Caste in South India, Ceylon and North-Eest Pakistan. pp.61-77. Cambridge: Cambridge University Press. Barth, F. 1960. The System of Social Stratification in Swat, north Pakistan. (In), Leach, E.R. (ed.), Aspects of caste in south India, Ceylon and north-west Pakistan. pp.113-146. Cambridge: Cambridge University Press.

Caste is, of course, a far more flexible system of social organisation than suggested by the Arthasastra or the Laws of Manu, as illustrated by numerous modern examples of communities rising or falling within its rankings. Yalman (1960), for example, cited the example of a group of high caste families moving into a locality within central Sri Lanka which had been recently vacated by lower caste inhabitants. Through this spatial association alone their overall status was lowered amongst the surrounding high caste families. Similarly, there are a number of examples of lower caste groups who underwent a process of Sanskritisation as a strategy to legitimate a claim of higher caste. As Jayaram states, 'the flexible dimension of the caste system in the background of the open-endedness of Hinduism is reflected in the claims of higher caste status by the lower castes following their sanskritisation' (1996: 78). In contrast, there have been a number of

Buhler, G. 1886. The Laws of Manu. Oxford: Oxford University Press. Coningham, R.A.E. 1994. 'Urban Texts': An Interpretation of the Architectural, Textual and Artefactual Records of a Sri Lankan Early Historic city. Unpublished doctoral dissertation. University of Cambridge. Coningham, R.A.E. 1997. The spatial distribution of craft activities in Early Historic cities and their social implications. (In), Allchin, F.R. and B. (eds.), South Asian Archaeology 1995. pp. 351-364. Delhi: Oxford and IBH.

92

Coningham, R.A.E. in press. Anuradhapura, Volume 1: The Site. London: Society for South Asian Studies.

Leach, E.R. 1961. Pul Eliya: a Village in Ceylon. Cambridge: Cambridge University Press.

Coningham, R.A.E., Allchin, F.R., Batt, C.M. and Lucy, D. 1996. Passage to India? Anuradhapura and the early use of the Brahmi Script. Cambridge Archaeological Journal 6.1: 73-97

Malik, S.C. 1968. Indian Civilisation: the Formative Period. Simla: Indian Institute of Advanced Study. Marshall, J.H. 1931. Mohenjo-daro and the Indus Civilisation. London: Arthur Probsthain.

Deraniyagala, P.E.P. 1972. Faunal Remains. Ancient Ceylon 2: 155-158

Moore, H.L. 1982. The Interpretation of Spatial Patterning in Settlement Residues. (In), Hodder, I. (ed.), Symbolic and Structural Archaeology. pp.74-79. Cambridge: Cambridge University Press.

Erdosy, G. 1986. Social ranking and spatial structure: examples from India. Archaeological Review from Cambridge 5.2: 154-166.

O'Flaherty, W.D. 1981. The Rig Veda. Harmondsworth: Penguin Books

Gough, K. 1960. Caste in a Tanjore village. (In), Leach, E.R. (ed.), Aspects of Caste in South India, Ceylon and North-West Pakistan. pp.11-60. Cambridge: Cambridge University Press. Gough, K. 1981. Rural society in southeast India. Cambridge: Cambridge University Press.

Pracchia, S., Tosi, M. and Vidale, M. 1985. On the Type, Distribution and Extent of Craft industries at Mohenjo-daro. (In), Schotsman, J. and Taddei, M. (eds.), South Asian Archaeology, 1983. pp.207-247. Naples: Instituto Universitario Orientale.

Gupta, S.P. 1974. Two urbanisations in India. Puratattva 7: 53-60

Quigley, D. 1993. The Interpretation of Caste. Oxford: Oxford University Press.

Hodder, I. 1986. Reading the Past: Current Approaches to Intrepretation in Archaeology. Cambridge: Cambridge University Press.

Rangarajan, L.N. 1992. Kautilya - The Arthashastra. Delhi: Penguin Books.

Hutton, 1946. Caste in India: its Nature, Function and Origins. Bombay: Oxford University Press.

Ratnagar, S. 1991. Enquiries into the Political Organisation of Harappan Society. Pune: Ravish Publishers.

Jayaram, N. 1996. Caste and Hinduism: Changing Protean Relationships. (In), Srinivas, M.N. (ed.), Caste: Its Twentieth Century Avatar. pp. 69-86. Delhi: Viking Penguin.

Rowe, W.L. 1973. Caste, Kinship and Association in Urban India. (In), Southall, A. (ed.), Urban Anthropology. pp.211-250. Oxford: Oxford University Press.

Jarrige, J-F. and Santoni, M. 1979. Fouilles de Pirak. Paris: Commission des Fouilles Archeologiques.

Ryan, B. 1953. Caste in Modem Brunswick: Rutgers University Press.

Kangle, R.P. 1965. The Kautiliya Arthasastra. Bombay: University of Bombay.

Thapar, R. 1978. Ancient Indian Social History. New Delhi: Orient Longman.

Kolenda, P. 1978. Caste in Contemporary India. Malo Park (Ca.): Benjamin and Cummings.

Wheeler, R.E.M. 1953. The Indus Cambridge: Cambridge University Press.

Leach, E.R. 1960. What should we mean by Caste? (In), Leach, E.R. (ed.), Aspects of Caste in South India, Ceylon and North-West Pakistan. pp.1-10. Cambridge: Cambridge University Press.

Yalman, N. 1960. The Flexibility of Caste Principles in a Kandyan Community. (In), Leach, E.R. (ed.), Aspects of Caste in South India, Ceylon and North-West Pakistan. pp.78-112. Cambridge: Cambridge University Press.

93

Ceylon. New

Civilisation.

Christianity and Islam in the Middle Nile: Towards a study of religion and social change in the long term. David Edwards, McDonald Institute for Archaeological Research, Cambridge. a coherent culture-historical framework has yet to emerge.

Religious Identities

For the study of the history of World Religions, Nubia and the Middle Nile (Figure 1) is an area of considerable interest in representing a region dominated successively by two of the great World Religions; by Christianity for much of the medieval period (c.550-1400 AD), and with its replacement during the post-medieval period by Islam. This latter transformation is also of interest in having taken place not through conquest, but as a more gradual and complex process, associated with both political change, changes in 'state religion' as well as through popular conversion. It is also a transformation that is of a more than historical interest. Islam is an essential element in the modem Sudanese identity as well as a dominant and controversial issue in contemporary politics. As such, we would do well to recognise that any historical study of Islam in the region has relevance to current issues surrounding ongoing processes of Islamicisation, while a more balanced perspective on the historical roots of current conflicts seems a necessary basis for more satisfying analyses of modem Sudanese history (Johnson 1997). These are undoubtedly closely linked with changes in both religious and more general social identities.

our period-based religious labels still prove convenient, research has not yet begun to address more fundamental questions concerning the long-term social effects of religious change in the Middle Nile. As elsewhere in sub-Saharan Africa, studies looking to a more social archaeology remain poorly developed while the archaeological contribution to the study of religion in Africa as a whole remains very limited (de Maret 1994). However, developing more social and comparative approaches requires a broadening of our approach outside the traditional research foci, concerned as they are, almost exclusively with those aspects of the material world which most closely embody these religions. Studies of the religious arts and architecture, for example, have yet to begin to explore their social context or indeed their relevance to the world around them - they are largely defined and selected precisely for their religious associations and, to a large extent decontextualised. If religion has provided convenient period labels, from a social perspective, the usefulness of privileging religion in this way would seem to be a matter to be explored and assessed rather than imposed a priori. If

The conventional archaeological discourse in the region has of course found religious labels very convenient; medieval Nubia is commonly portrayed as 'Christian' Nubia, in contrast to its 'pagan' Kushite/Meroitic antecedents and an 'Islamic' post-medieval period. Medieval Nubian Studies still remains closely allied to medieval and Coptic studies in Egypt, with a particularly strong interest in texts, religious wall painting, iconography and ecclesiastical architecture, while the highly elaborate medieval pottery of northern Nubia is conveniently subsumed under a 'Christian' label (Adams 1977). Superficially at least, medieval Nubia looks very Christian - or at least the archaeology that is done tends to look very 'Christian'.

Very fundamental questions remain to be addressed. When we step outside the churches, the questions we should perhaps be asking are: how 'Christian' was 'Christian Nubia'? Equally, we may ask: how 'Islamic' was Funj Sennar? To what extent did these religions transform societies of the Middle Nile (O'Fahey 1993: 22)? Did they fundamentally change the ways in which social identities were constructed and what kind of mark did they leave on culture more generally? O'Fahey and Spaulding have suggested that both Christianity and Islam rested quite lightly on the indigenous cultures of the Middle Nile (1974), but archaeological studies which seek to address such questions are still lacking. We must also recognise that religious change is a long-term and dynamic process. At a comparative level we may also begin to explore possible qualitative differences between the impact of Christianity and Islam. In the modem world there can be little doubt that the extension of Islam is associated with very fundamental social changes. After several centuries, the processes of Islamicisation are still ongoing, not only among 'nonbelievers' of the southern periphery, but also in redefining Islamic and Sudanese identities for longestablished Muslim populations in the core regions of the state.

The status of post-medieval Islamic archaeology is more problematic, such work still remaining in its infancy (Adams 1987). For the core riverine areas, perceptions of this period still tend to be framed as much around the disappearance of 'Christian' material culture as by the development of distinctive new material markers. Postmedieval castles and domed qubba tombs may provide points in the landscape which we may identify with the history and traditions of this period, particularly that of the Funj Sultanate of Sennar (Spaulding 1985a), but our knowledge of them has provided little more than footnotes for the historical narrative. To the west, in areas which remained outside the sphere of Christian influence during the medieval period, we have virtually no reliable archaeological information concerning the material culture of the 'pagan' medieval period. If 'Islamic' still provides a convenient label for this period,

These latter points can only emphasise that exploring religious change within its wider social context is by no means straightforward, presenting both theoretical and methodological problems. In the first instance, it is necessary to clearly distinguish between religious

94

Aswan 1

3

Meroe Musawwarat

cataracts II

0

400km

Figure 1. Nubia and the Middle Nile supposedly 'Christian culture', it is necessary to acknowledge the likely existence of multi-layered social networks, fluid in time and space - bound by complex webs of power relations, networks in which the influence of religion may be very unevenly distributed.

conversion and more general cultural conversion, a distinction which, in this region, has rarely been welldefined in the past. As is clear in more recent African history, traditional belief systems and forms of social organisation may on occasions be relatively untouched by the introduction of Christianity; much to the dismay of many Europeans who believed Christianity to be intrinsically bound to Western civilisation and its values (Kirkby 1994). Furthermore, rather than maintain implicit assumptions that there existed a medieval 'Christian' Nubian culture, we need to come to terms with a potential multiplicity of parallel, overlapping and possibly conflicting identities within the Middle Nile. Having traditionally focused on the historical narrative of kings, the institutions of the Church and its arts, and a

In such terms, rather than looking simply at Christianity and Islam as manifested in Nubia, we may begin to attempt to explore the dynamic relationship between Nubian society(s) and these religions, as manifest at many different levels and in many different ways. Much will depend on the scale of analysis we choose to adopt. If research has, to date, tended to focus particularly on the institutions, arts and architecture of the Church, and more generally in supplementing the narrative of

95

political history, different levels of analysis may be expected to throw light on other, potentially numerous identities, which may range from the state-level to the domestic. In this respect, with a recognition of the importance of material culture not only in reflecting but creating society and social identities (Appadurai 1986), archaeology has perhaps a special potential to explore the complexities of social worlds in areas where the scant historical record can throw little light.

area seems to have been divided into at least two and initially three kingdoms during the medieval period. In that the ritual roles of the Kushite kings were probably one of their most important and extensive sources of power (Edwards 1996a), we may suggest that the ideological and institutional support for the Nubian monarchy provided by the Church was of a different order. Indeed, as Adams has argued, existing evidence suggests that the Church, dependant for the appointment of bishops from Alexandria and on royal patronage, failed to develop an independent organic local basis or a specifically Nubian character (1977: 541).

In beginning to extend our understanding of the possible social implications of religious change, studies of, for example, ceramic or culinary culture may relate to, and inform us about rather different arenas of social activity, both in domestic and community life, and in the fundamental relations of production and reproduction. It is in such areas that we undoubtedly need to look if we are to begin to assess the wider social implications of the Christianisation or Islamicisation of the Middle Nile. From an archaeological perspective, ceramic studies, in their broadest sense, offer one particularly promising avenue; not just the pots themselves, but also a wide range of social and economic practices with which they may be linked.

As a consequence of this, with the collapse of the Christian kingdoms as political units, the Church lacked the organisational infrastructures to survive without political support; gradually losing priests and churches, the faith withered away. Whether this process was a longer-term phenomenon remains unclear, but it has also be noted that in the later Christian centuries, churches become markedly smaller and organised in a way to suggest they were increasingly sanctuaries for the clergy, excluding the wider populace (Adams 1977: 477). That this reflects a decline in popular religious participation in the later medieval period is probably too simplistic an explanation, but it is likely that further research, attempting to better-contextualize this change, may be profitable.

Historical background

Before approaching archaeological issues, a few introductory comments are required concerning the historical context. This is particularly important in relation to the post-medieval period where archaeological information is still so sparse, and where to a considerable extent we are still only in a position to offer hypotheses and rather speculative interpretations. The progress of the initial Christianisation of Nubia during the sixth century, apparently largely the result of formal evangelising missions by Justinian and the Empress Theodora, bears comparison with other Christian missions of the same period, including those to Anglo-Saxon England. Aspects of its impact on the political world of Nubia have been discussed by Trigger (1978), looking particularly at the institutions of the Church, at monasteries, literacy, new administrative and legal regimes, and a broader integration into larger world systems. On the basis of this and other syntheses (Adams 1977) there seems little doubt that such institutional support will have contributed to the apparent political stability of much of the medieval period. Archaeologically, this is most clearly seen, especially in the north, in a marked extension of permanent rural settlements, substantial mudbrick architecture, and a relatively rich and diverse material culture with links to the Mediterranean world and Near East.

If the Church does not seem to have transformed the

deep structures of political power, we also find few indications of any major innovations in socio-economic organisation during the medieval period which might be associated with it. Distinctive forms of land-ownership and agriculture, closely associated with water-wheel irrigation are documented in Lower Nubia, but these seem as likely as not to reflect the peculiar environmental constraints of this highly marginal area, rather than any bureaucratic and/or legal innovations which may be linked with the Church or its institutions. More generally there is no evidence to suggest that here, the traditionally limited role of the state in controlling and exploiting the subsistence resources of the kingdom, a common characteristic of Sudanic kingdoms (Edwards 1996a), was substantially altered. In the light of later developments it is noteworthy that ancient pre-Christian principles of royal succession were maintained through the medieval period: the right to rule being preserved within the female line of the royal family (Vantini 1975: 163). If we tum to other areas of material culture, the Christian contribution is more ambiguous. Medieval pottery represents certainly one of the defining decorative 'arts' for much of the period, very highly elaborated, in a way that no other domestic material seem to have been. However, the extent to which this phenomenon was something to do with Christianity, whether it can be seen as a Christian art, remains far less certain. As will be further discussed below, this may tell us more about the social construction of Nubian society.

It must be remembered that within a century of the conversion, Egypt was drawn into the Arab and Islamic world, removing any more immediate material benefits which might have been gained through association with Christian Byzantine Egypt. It is also noteworthy that the scale of political integration achieved during the medieval period, never matched that of the earlier Kushite state, which, at its peak had probably controlled some lO00kms of the Nile valley, from the Egyptian frontier to south of Khartoum. By contrast, the same

Many lacunae remain in our knowledge of the history of the post-medieval period, which saw the appearance of an Islamic Fur dynasty in the west and the domination of the central Sudan by the Funj Sultans of Sennar, who at their strongest dominated the river valley from around the Third Cataract to far south of Sennar (Spaulding

96

1985a). However, if we look at histories of Islamic Sennar, and indeed nineteenth and twentieth century history, a number of themes emerge which suggest that, albeit slowly, a new form of state was emerging, increasingly at odds with established traditions of power relations and the forms of socio-economic organisation and social networks they had co-existed with.

Archaeological approaches

This brief reading of the historical outline suggests some ways in that Christianity and Islam effected the Middle Nile in very different ways. While many of the social transformations underway in the post-medieval period were not directly caused by religious change, it is apparent that it played a part in the often radical redefinitions of existing social organisations and social identities during this period. Exploring such transformations in terms of material culture presents a challenge, but one we might expect to be productive. The access which archaeological data may also provide, especially with the benefit of long-term perspectives, to many social arenas not documented in the historical record, may give us new and varied insights.

Certain changes had a specifically religious basis. As Spaulding has argued, a highly symbolic watershed in the nature of kingship and royal legitimacy can be identified in the mid-eighteenth century (1985b). By the reign of the Sultan, Badi IV, the dominance of orthodox Islamic legal principles was assured. The introduction of Islamic patrilineal principles of inheritance removed the role of the royal women, through marriage, in legitimating the sultan and his vassals. This also removed a key network of political alliances which had previously existed between the crown and provincial governors, permitting the establishment of autonomous patrilineages, which in tum seems to have been an major factor in the undermining of central royal authority, and the disintegration of the centralised state in the late Funj period (op.cit: 21-3). Ultimately disastrous for the Funj Sultans, this may be highlighted as a fundamental shift, turning away from very long-established traditions of Nubian kingship.

The choice of ceramics as a research focus is grounded on two general observations, which have a particular resonance with regard to the themes under discussion here. Firstly, for many millennia, ceramics are one of the most prominent aspects of the material culture of this region. From its first appearance as some of the earliest pottery in the world, currently dateable to the tenth millennium b.p. (Close 1997), it is remarkable for a consistent and sometimes spectacular elaboration - a tradition which survived the Christianisation of Nubia and persisted until the very late medieval period. If textiles, gourds, basketry or beadwork may be the chosen media for other cultural traditions, pottery is marked out as a material of particular significance in this region.

Transformations in the traditional pattern of kingship were also accompanied by more widespread social transformations. New elites developed, particularly a religious elite, identifying with Islam and the Arab world, standing outside the traditional structures of palace and provincial vassals. Concurrently the growing influence of a monetary economy and foreign merchants, undermining royal monopolies on long-distance trade and political benefits of controlling prestigious imports (O'Fahey and Spaulding 1974: 74).

Secondly, that in more recent times, within riverine areas of central Sudan, significant potting traditions, and certainly traditions of ceramic elaboration have entirely disappeared. We may also observe that where more diverse and elaborate potting traditions have survived in Sudan, it has been in areas which lay outside the control of the Funj Sultans and the Egyptian state of the nineteenth century, areas into which Islam has penetrated only relatively recently This pattern then raises many questions concerning pottery and its use. At a general level, we may consider why pottery emerged as such a prominent feature of material culture in this region, while more specifically, we may also raise questions as to whether its disappearance may in any way be related to the coming of Islam? While we may suspect that simple mono-causal explanations of this sort will not be very satisfactory, a focus on this area suggests new ways of looking at socio-cultural changes associated with Islamicisation, which in tum may invite comparisons with the earlier impact of Christianity.

There were new ways of defining identities - where status and legitimacy were proclaimed in dubious genealogies claiming Arab ancestors - part of a package linking the penetration of Islamic law - and 'the ways of the Arabs'. In the nineteenth century, the dislocations caused by the development of a far stronger state, following the Egyptian colonisation of the country, saw an acceleration of these processes. State raiding of the peripheries increased on expanding frontiers, but these increasingly predatory activities also developed religious connotations, establishing new criteria for those who took slaves, and those who became slaves (Johnson 1997). We may also discern changes in access to land and the redefinition of land-ownership, particularly with the erosion of institutions and customs of common land ownership. In the Funj period this was often embodied in the institutionalising of slavery at the state level, a widespread practice, creating a new class stripped of their traditional identities. The legal and economic shifts continue, with the expropriation of lands and their cultivation by a new class of wage labourer. In the modem Sudan, such ex-slaves, the original 'Sudani', have emerged as a significant new social grouping.

As a point of departure, it is suggested that in this region, pottery may be intimately connected with, and play an active role in, a number of social arenas. This wider social significance, and importance as a material culture element in sub-Saharan Africa, and many other parts of the world, is now increasingly well-recognised (Barley 1994), if only recently is it beginning to be explored by archaeologists (Barnett and Hoopes 1995; Cumberpatch and Blinkhom 1996). In the Middle Nile, we have a number of very clear pointers to the ways in which pottery may, directly or indirectly, be associated with many aspects of social life, in ritual and ceremonial,

97

in activities structuring intra-community relations, gender-based social divisions as well as in the ordering of the social relations of production more generally. It may at times represent much more than a practical technology. If approached in this way, it clearly has considerable potential for social analysis, not least in assessing the impact of religious changes on existing social structures and practices.

has survived in larger-scale and politically complex societies (Dietler op.cit: 365-370). In rural Sudan into recent times, anything between 20 to 30% of grain might be devoted to beer-making, most of which is then consumed within circuits of feasts and/or work parties (Barth 1967; Holy 1974; lten 1979; Nadel 1947). The tradition of the beer-work party, has remained strong, despite the appearance of wage labour within an increasingly monetised economy, while we may also note here that such practices survive most prominently on the peripheries of the early Islamic states, although their echoes may still be found even in riverine core zones.

Here, one very fundamental linkage may be proposed, established on the relationship between pottery function and the subsistence economy which further suggest that ceramics may, directly or indirectly occupy a privileged position. This is based on subsistence regimes where sorghum and millet, the agricultural staples, have traditionally been consumed as beers and porridges, not as breads. The particular social implications and characteristics associated with this subsistence base are likely to have been very important: cultivation, harvesting, grinding, malting, cooking and brewing are all likely to have been deeply embedded in the fabric of society and the practice of everyday life (Haaland 1995). The long history of sorghum beers may further be singled out for special attention due to the wider social, economic and indeed political roles of drinking, which, as is now well-recognised, are commonly deeply embedded in the social fabric, serving a wide range of social functions (Dietler 1990). In this region, such beer consumption is clearly of some antiquity, being welldocumented in medieval Arabic and earlier sources (Vantini 1975). During the fourteenth century, AlUmari recorded the Nubians' inclination to get drunk on mizr. Al-Idrisi, writing before 1170 noted preparation of a mizr from the millet in the northern kingdom of Makuria, and Al-Aswani, a century earlier, recorded sorghum bread and beer in central Sudan within the kingdom of Alodia (Figure 1). Such beers are traceable still earlier, to at least the Meroitic period during the late first millennium BC, according to Strabo's account of this region.

As I have suggested elsewhere (Edwards 1996b), a possible interpretation accounting for many features of the elaborated Nubian potting traditions may be found in the social significance accorded such consumption of beers and other alcoholic beverages over millennia. At many levels then, pottery may be intricately bound up with the creation and reproduction of social identities, with gender relations in the production and presentation of beer, in community relations through its status associations, and in more overt displays of power associated with its prov1s1on and conspicuous consumption. As such, its changing history may be a valuable indicator of the more fundamental forms of social transformations associated with religious changes which may have affected these practices. Evidence for both changes and continuities in ceramic culture and associated practices may thus provide a valuable index of the degree of penetration, or otherwise, of Christianity or Islam on Sudan. In a study of the impact of Islam on the Nuba this century (Stevenson 1963), it is apparent that after some centuries of its presence, Islam still had had only a very limited cultural impact on, for example, rites of passage, while barely touching the organisation of agricultural life and its rites, the most fundamental bases of society.

In the Middle Nile and adjoining areas, ethnographic and historical accounts over the last 200 years record an often intense ritual and ceremonial importance for such beers (Edwards 1996b). Beer and its preparation has commonly been a focus of status differentiation and power relations at many levels. These are seen in gender-based distinctions, interweaving themes associated with female roles in its production, and distinctive patterns of generally male consumption, associated with, for example, age-related status groups, as well as with displays of hospitality and conspicuous consumption. The ubiquity and social importance of beer cannot be overstressed; it is not coincidental that among the Ingessana of the Blue Nile, the whole annual cycle of festivals are called sak, literally meaning 'beers' (Jedrej 1995).

Archaeological ceramics in social contexts

Developing systematic studies of ceramic culture along these lines will clearly be a major undertaking and is beyond the scope of this paper, not least in requiring much better data than are currently available. However, we may at least examine some areas of the available record which already indicate the potential usefulness of such lines of enquiry. These may provide some pointers to ways in which we may provide some social context for the more overt manifestations of Christianity and Islam which currently dominate our narratives. Archaeological manifestations of the social, and indeed ritual significance of beers are already very apparent in the pre-Christian period - most obviously in mortuary culture. The most distinctive and ubiquitous features of both Meroitic (Figure 2) and post-Meroitic mortuary assemblages (Figures 3-4) are the pottery groups, at their most basic comprising drinking cups/bowls, and large jars. At times very elaborately decorated, their use in funerary rites, rather than as 'grave offerings', seems clear (Lenoble 1994), more specifically associated with alcohol consumption as part of 'funerary banquets' (Edwards 1996a). 'Beer jars' appear interchangeable

In addition to its focal role in ritual and ceremonial life, beer has also been inextricably linked with work and labour, a pattern also well-recognised in many parts of the world (Dietler op.cit; Morris 1979), a linkage established through the use of beer parties as work parties. For acephalous societies such parties are commonly the primary means for mobilising labour beyond the primary productive unit, while the practice

98

with wine amphora both in elite and royal Meroitic graves, as well as Lower Nubian post-Meroitic graves. Figure 4 represents material from post-Meroitic grave groups from northern Sudan, probably nearly contemporary with the appearance of Christianity in the region. Elite graves of a similar period in central Sudan (Figure 3) may include large groups of jars, representing hundreds of litres of drink, leaving little doubt that such beers were not only ritually significant but also linked to displays of status. More overtly ritual and ceremonial associations of pottery may also be seen in a range of highly decorated Meroitic white finewares, largely restricted to small drinking cups and bowls, sometimes incorporated into burial groups but otherwise only found in significant quantities in elite or cultic contexts,

notably at Meroe the capital (Torok 1997) and at Musawwarat es Sofra, a major pilgrimage centre (Edwards 1998). In assessing the impact of Christianity on these traditions, following the conversion of the post-Meroitic (c.350-550 AD) Nubian kingdoms, we are faced with a number of problems due to both major changes in ceramic production as well as in the nature of the material available for study, changes which, in the past have tended to obscure underlying continuities. Bestdocumented in more northerly regions, we can identify two major transformations of ceramic production during the pagan post-Meroitic period, and another generally associated with the arrival of Christianity in the region.

Figure 2. Meroitic pottery group, c.2nd century AD

\I)

rn Figure 3. Post-Meroitic handmade pottery group (Upper Nubia), c.4th century AD

99

YW U-W i3 w

Figure 4. Post-Meroitic wheelmade pottery group (Lower Nubia), c.6th century AD Following the collapse of the Meroitic kingdom, and the disappearance of distinctive products of Meroitic workshops, Lower Nubia came under the influence of Romano-Egyptian products. During the fifth century however, we see a marked divergence from the Egyptian prototypes - a 'Nubianisation' of production, with the appearance of new forms and a marked increase in their elaboration. In terms of form, the most obvious addition to the wheelmade repertoire are large globular 'beer jars' (Figure 4), with a distinctive range of Lower Nubian types occurring in funerary assemblages as far south as the Third Cataract (Edwards 1994). Such wheelmade pottery of the north compares well with a range of handmade types (Figure 3), one of the defining features of assemblages of this period (if occurring in a number of regional styles), at least as far south as the Blue Nile region. The particular interest of these northern groups

however, lies in the demonstration of a 'Nubianising' of an alien pottery tradition, both in terms of specific vessel forms and in the degree of elaboration of vessel finish (Adams 1986). Changes of form were required to accommodate the very different subsistence/culinary traditions of Nubia, while the development of more elaborate vessel finishes and decoration met the requirements for the social and ritual uses of beers, most clearly seen in mortuary contexts. A further disjunction is found in the early Christian period with another major change in vessel forms, again marking external influences, particularly with the appearance of Byzantine-Egyptian forms of dishes and bowls. Monastic potteries may have played a role in the spread of such influences, although their exact role in the initial spread of new pottery repertoires remains

100

uncertain. It is this appearance of distinctive 'Early Christian' wares which has traditionally served as a convenient cultural marker, distinguishing 'pagan' from 'Christian' Nubia. However, several features of the way in which medieval pottery production developed would tend to suggest that following the conversion of Nubia, pottery still retained much of its social significance, and moreover points to the survival of socially significant drinking, deeply embedded in social practice. Despite its practical archaeological use as a chronological/ horizon marker, defining the beginning of the 'Christian' medieval period, such possible continuities demand further attention.

colour in pottery decoration may well be suggested, with white wares throughout the medieval period tending to the most elaborate decoration, with the possible apotropaic use ofred (Adams 1986: 222). In view of the common structuring of African colour systems around the use of Red, White and Black (Turner 1968), further work in this area may be productive. More specifically, however, the unusual elaboration of the functionally ambiguous 'vases' and drinking cups, becomes more understandable if they are associated: as elements of drinking sets. Possible matching sets of such vessels would tend to support this association (Adams 1998: 26). Such a link may be further supported by the consistent decoration of 'larger vessels' (i.e. beer jars) in the most abundant of the earlier W2 white ware (Adams 1986: 102).

The most striking development of the early medieval period is the shift away from the relatively plain, predominantly red wares of the 'Early Christian' period and the development of a new and peculiarly Nubian style in the highly elaborate 'Classic Christian' (c.800 AD) and later periods; a combination of new vessel forms, colour and decoration. A predominance of white wares, echoing the favoured colour of the most elaborate ritual/ceremonial Meroitic pottery, and the development of highly elaborate forms of painted decoration are particularly characteristic. Equally important however, is the prominence of two ranges of vessel forms, the rather enigmatically-termed 'vases' and drinking cups/goblets, both apparently rare among the earliest 'Christian' wares (Figure 5). Both of these forms are also consistently highly-decorated and continue to be so through the later medieval period (Adams 1986:101-2).

As in the pre-Christian period, we may again propose this association of highly elaborate pottery and the social importance of drink-related activities. At this stage, locating the exact contexts within which such practices were maintained remains difficult, with few domestic assemblages being reported in the necessary detail required for such analysis. The adoption of relatively orthodox Christian burial rites also saw the almost complete disappearance of grave furnishings, including beer jars, from burials, removing potentially one of the most obvious manifestations of such practices from the archaeological record. This abandonment of 'gravegoods', has traditionally been seen as evidence for a supposedly swift and universal conversion to Christianity. However, even here, we have some indications that what we may be seeing is no more than a shift in the ritual focus away from the actual burial, now carried-out following Christian rites; effectively an accommodation between the new religion and existing rites of passage.

Several potentially significant features of these developments may be highlighted. As noted above, the revival of white wares may in itself be significant in view of the possible symbolic associations of colours. The existence of clear structuring principles in the use of

Figure 5. Late medieval pottery, decorated goblets, cups/bowls and 'vases' (Lower Nubia)

101

elaboration of 'Christian' medieval pottery is perhaps best interpreted not in isolation, as a 'Christian' art, but rather as a new manifestation and reworking of longestablished and highly elaborated cultural traditions. These were deeply embedded in the social fabric, reflecting what pots were used for. Many questions of course remain unanswered, in particular concerning the changing contexts of pottery production, and the development of the distinctive medieval decorative styles. Some of these may certainly be linked with other Christian arts, such as manuscript illumination, but whether specifically Christian symbolic associations were important remains to be determined.

Recent work has identified early medieval burials where deposits of drinking vessels are still present, not in the graves themselves, but buried on the surface beneath covering tumuli or deposited elsewhere in cemeteries (Edwards 1994; Edwards and Rose 1997). Examples from central Sudan at a cemetery at Gabati can be reliably dated to the late seventh century, at least a century after the official 'conversion' of the region, while at the same site, the earliest dated orthodox 'Christian' graves are even later. Whether these occurrences represent only a transitional phase has yet to be determined, but they certainly indicate, as we should perhaps expect, that the progress of religious conversion as well as its wider impact, was more complex and variable than conventional narratives suggest.

In the post-Christian period, we are left with many more questions than answers. The disappearance of wheelmade production is likely as not to be a symptom of political fragmentation in the late medieval period. Interestingly, however, medieval decorative traditions appear to have survived on late handmade wares for at least a limited period. On the other hand, we have yet to identify a distinctive Funj tradition of pottery, while what little is known of post-medieval pottery points to a restricted and unelaborated domestic repertoire.

More generally, both the ceramic and mortuary evidence may be interpreted within the context of the continued prominence of alcohol through the medieval period. It remains a prominent feature of the medieval material record, not only with beer as a subsistence and social staple, but, at least in northern Nubia, with an abundance of imported and locally produced wine amphorae, and apparently wine production (Adams 1966). It is not surprising that the strong Christian associations of wine would have found a ready audience in a region which already had a long exposure to alcohol in both local and imported forms stretching back over several millennia. The extent to which ritual uses of wine will have impinged on existing custom is certainly an interesting avenue to explore. That wine continued to represent the chosen drink of the elite is very likely, continuing a tradition already well-established during the Meroitic period. Accounts of the famous Baqt treaty between the Nubian king and the Arab conquerors of Egypt in the 650s, record that in the exchange of goods associated with the delivery of slaves to Egypt, the king was to receive 1000 kanyr of wine and his delegates 300 kanyr. A further hint of its social use may also be found in records of the provision of drink (and food) for those witnessing legal transactions, as recorded in a number of Old Nubian documents (Browne 1991).

However, if the basic groundwork has yet to be done, we may be in a position to at least recognise how the social contexts in which pottery was used, were changing. Such observations may suggest fruitful avenues for future research. If beer was becoming increasingly ambiguous ideologically with the spread of Islam due to prohibitions on alcohol, this is likely to have had a significant effect on material objects associated with its preparation and serving - most obviously in the decoration of pottery. However, we might also propose, that the new economic structures associated with the Funj state, if adopting an increasingly Islamic character in many areas, were equally responsible. During the nineteenth century, beer was widely available and publicly consumed (Burckhardt 1819), but significantly, by this date it appears to have entered the market economy. Such open availability stands in sharp contrast to social prescriptions on its availability which are necessary elements in maintaining its often intense social and ritual roles (Haaland n.d.). African social histories of more recent periods are rich in data concerning the social consequences of just such shifts in beer production from being a social to a marketable commodity, undermining a wide range of fundamental social relationships, not least between generations and genders (e.g. Wilson 1977). Here at least, we may suggest that transformations of production, and the social institutions that organised production, were already well under way. By implication, we may expect that traditional forms of display associated with beer will also have been losing their significance by this date.

Conclusions

From a long-term perspective, I have tried to suggest that, on the material evidence, many deeply embedded social structures and their ritual and ceremonial expression appear to have survived the arrival, and departure, of Christianity in the region. A long and continuous interest in alcoholic beverages, in part associated with rites of passage, but also very probably structuring fundamental economic relations, survived both the introduction of Christianity, and, at least on the margins of the Funj state, several centuries of an Islamic presence. While doubtless dynamic in changing meanings and practice we have at least hints of underlying social continuities behind the facades of official religious labels which have traditionally structured the history of this region.

We also find new forms of land ownership and inheritance appearing, the institutionalisation of large scale slavery, at least defined along religious lines if not explicitly Islamic, and the erosion of communal landholdings and working practices. All of these contributed to the erosion of traditional social structures, the structures of productive relations, and the dissolution of communal identities within the state's core, again exactly those areas in which pottery enjoyed a particular

The suggested social significance of sorghum beers may in itself go some way to explain the peculiar Nubian adaptations and elaboration of imported medieval potting technologies. As such, the remarkable

102

Burckhardt, J.L. 1819. Travels in Nubia. (reprint) London: Darf Publishers.

social significance. Under modem circumstances, such transformations may have explicitly religious origins: the role of drink in 'work' parties is increasingly marginalised with the growing influence of stricter interpretations of Islamic law in all areas of social life. We find explicit accounts today from the Blue Nile region, that these essential institutions of communal labour are under threat following the criminalisation of brewing and beer consumption (Medani 1994: 132). Exploring the potentially traumatic social consequences of such changes lies more in the domain of development anthropologists than archaeologists, but we may suggest that tracking earlier manifestations of such transformations through the post-medieval period offers a rich and challenging research agenda.

Dietler, M. 1990. Driven by drink: the Role of drinking in the Political Economy and the Case of early Iron Age France. Journal of Anthropological Archaeology 9: 352406. Edwards, D.N. 1994. Post-Meroitic ('X-Group') and Christian burials at Sesibi, Northern Sudan. The Excavations of 1937. Journal of Egyptian Archaeology 80: 159-78. Edwards, D.N. 1996a. The Archaeology of the Meroitic State: new perspectives on its social and political organisation. (Cambridge Monographs in African Archaeology 38). Oxford: BAR

Over the long term, it may be hoped that archaeologists can begin to contribute to debates concerning the wider social manifestations of religious change, to begin to understand some of the syncretic tolerances, tensions and complexities of medieval and later Nubian societies, and indeed the potential contrasts between what may be characterised as the 'Nubianising of Christianity' and the 'Islamicising of Sudan'. A recognition of such potential tensions and conflicts again has a particular resonance in this region where the failure of post-Independence Sudan to create and maintain a viable national identity, stems in great part from a failure to resolve, or even acknowledge conflicts between religious and other preexisting identities.

Edwards, D.N. 1996b. Sorghum, Beer and Kushite Society. Norwegian Archaeological Review 29/2: 65-77. Edwards, D.N. 1998. Report on the Musawwarat Pottery, 1997. MittSAG 8: 62-67. Edwards, D.N., and Rose, P. 1997. The Meroitic, postMeroitic and Christian Cemetery at Gabati, Central Sudan. Kush 17: 69-84. Haaland, G. (forthcoming) Beer, Blood and Mother's Milk. Sudan Notes and Records.

References

Haaland, R. 1995. Sedentism, Cultivation, and Plant Domestication in the Holocene Middle Nile Region. Journal of Field Archaeology 22: 157-73.

Adams, W.Y. 1966. The Vintage of Nubia. Kush 14: 262-83.

Holy, L. 1974. Neighbours and Kinsmen. A Study of the Berti People of Darfur. London: Hurst and Co.

Adams, W.Y. 1977. Nubia. Corridor to Africa. London: Allen Lane.

Jedrej, M.C. 1995. lngessana. The Religious Institutions of a people of the Sudan Ethiopian Borderland. Leiden: E.J. Brill.

Adams, W.Y. 1986. Ceramic Industries of Medieval Nubia. 2 vols. Lexington: University of Kentucky Press.

Johnson, D. 1997. 'The Sudan is sui generis.' Past failures and future prospects in Sudanese History. Sudan Studies 20: 1-9.

Adams, W.Y. 1987. Islamic Archaeology in Nubia: An Introductory Survey, (In), Hagg, T. (ed.), Nubian Culture, Past and Present. pp.327-61. (Main Papers Presented at the Sixth International Conference for Nubian Studies, Uppsala 1986), Stockholm.

Kennedy, J.G. (ed.), 1978. Nubian Ceremonial Life: Studies in Islamic Syncretism and Cultural Change. Berkeley.

Adams, W.Y. 1998. Kulubnarti II. London: Sudan Archaeological Research Society.

Kirkby, J.P. 1994. Cultural Change and Religious Conversion in West Africa. (In), Blakely, T., van Beek, W.and Thomson, D. (eds.), Religion in Africa. pp. 5671. London: James Currey.

Appadurai, A. (ed.), 1986. The Social Life of Things. Cambridge: Cambridge University Press.

Lenoble, P. 1994. Du Meroitique au Post-Meroitique dans la Region Meridionale du Royaume de Meroe. Recherches sur la periode de transition. Unpublished Ph.D. Paris IV. Sorbonne.

Barnett, W.K. and Hoopes, J.W. (eds.), 1995. The Emergence of Pottery. Technology and Innovation in Ancient Societies. Washington: Smithsonian. Brenner, L. (ed.), 1993. Muslim identity and social change in sub-Saharan Africa. London: Hurst.

de Maret, P. 1994. Archaeological and other prehistoric evidence of traditional African religious expression. (In), Blakely, T., van Beek, W. And Thomson, D. (eds.), Religion in Africa. pp.182-95. London: James Currey.

Browne, G.M. 1991. Old Nubian Texts from Qasr /brim ///. London: Egypt Exploration Society.

El-Medani, K. A. 1994. Some aspects of Indigenous

103

Sudan, 1720-1762. (In), Daly, M.W. (ed.), Modernization in the Sudan. pp.17-27. New York: Barber Press.

Farming Knowledge in the Blue Nile Area: the case of Abu Gumi village. (In), Ahmed, M.M.M. (ed.), Indigenous Farming Systems, Knowledge and Practices in the Sudan. pp.95-135. Khartoum: Khartoum University Press.

Stevenson, R. 1963. Some aspects of the spread of Islam in the Nuba Mountains. Sudan Notes and Records 44: 920.

Morris, C. 1979. Maize beer in the Economics, Politics, and Religion of the Inca Empire. (In), Gastineau, C.F., Darby, W.J. and Turner, T.B. (eds.), Fermented Food Beverages in Nutrition. pp.21-34. New York: Academic Press.

Torok, L. 1997. Meroe City. An Ancient African Capital. London: Egypt Exploration Society. Trigger, B. 1978. The Inter-Societal Transfer of Institutions. (In), Time and Traditions. Essays in Archaeological Interpretation. Edinburgh: Edinburgh University Press.

Nadel, S.F. 1947. The Nuba. London: Oxford University Press.

Turner, V.W. 1968. Colour Classification in Ndembu Ritual. (In), Banton, M. (ed.), Anthropological Approaches to the Study of Religions. pp.4 7-84. London: Tavistock.

O'Fahey, R. 1993. Islamic Hegemonies in the Sudan. (In), L. Brenner (ed.) Muslim identity and social change in sub-Saharan Africa. pp.21-35. London: Hurst. O'Fahey, R. and Spaulding, J. 1974. Kingdoms of the Sudan. London: Methuen.

Vantini, G. 1975. Oriental Sources concerning Nubia. Heidelberg and Warsaw: Polish Academy of Sciences and Heidelburger Akademie der Wissenschaften.

Spaulding, J. 1985a. The Heroic Age in Sinnar. East Lansing: Madison State University.

Wilson, M. 1977. For Men and Elders. London: International African Institute.

Spaulding, J. 1985b. The End of Nubian Kingship in the

104

Tadmakkat and the Image of Mecca: Epigraphic Records of the Work of the Imagination in 11th Century West Africa. P. F. de Moraes Farias, Centre of West African Studies, University of Birmingham. of routes and markets, and the practical time-keeping, directly required by trans-Saharan and Sahelian traders.

Introduction

In the fifth century AH/ eleventh century AD, expatriate Muslims in West Africa, and West African converts to Islam, were engaged in a joint effort of the imagination. Their task was the reciprocal domestication of otherness. It involved mapping West Africa onto the world at large while bringing international references to bear on West African localities. For this purpose, topographical images and chronological frames were used in special ways, which far transcended the pragmatic descriptions

At the :lssuk archaeological site in the Republic of Mali (Figure 1), medieval Arabic inscriptions document that work of the imagination. In so doing, one of them clinches the identification of :lssuk with a medieval town which, because of its name and perceived topography, was presented by contemporary sources as a mirror of Mecca - a concrete local reflection of the distant centre of Islamic culture. 8°E

I

-/.-i - ~:I ~--~·-,__ : \.

I I ■ lfaiwa~·

\.

;~~\

I

\ ·----)

I I

·~

Ay,r

:--,

.,;-~-; I

V