T antra ('woven together' in Sanskrit) is the Hindu-based religion which originated i ,200 years ago, when the great erotic temples were built. In the West it is now best known for the inspiration of tantric yoga, and its associated ritualistic forms of sex. But is Tantra just about esoteric sexual practice or does it amount to something more? This lively and original book contributes to a more complete understanding of Tantra's mysteries by discussing the idea of the body in Hindu tantric thought and practice in India.
The author argues that within Tantra the body is a vehicle for the spirituality that is fundamental to people's lives. The tantric body cannot be understood outside the traditions and texts that give it form. Through practice (ritual, yoga and 'reading') the body is formed into a pattern determined by tradition, and the practitioner thereby moulds his or her life into the shape of the tradition. While there is a great range of tantric bodies - from ascetics living in cremation grounds, to low-caste people possessed by tantric deities, to sophisticated high-caste Brahmans expounding the ascetic philosophy of Tantra - all share certain common assumptions and processes. Flood argues that while there is a divergence at different social levels and in different levels of tantric metaphysical claims, these levels are united by a process which the author calls 'entextualisation of the body'. The body becomes the text through the tradition being inscribed on it. This general claim is tested against specific ritual and doctrinal examples, and the tantric traditions are linked to wider social and political forces.
The Tantric Body is a fascinating study that makes an important contribution to the study of South Asian religion, and will have strong appeal to students of South Asian societies and cultures as well as to those of comparative philosophy.
Gavin Flood is Professor of Religion at the University of Stirling and Academic Director of the Oxford Centre for Hindu Studies. He is the author of An Introduction to Hinduism (1996) and general editor of The Blackwell Companion to Hinduism (2003).
The Tantric Body
The Secret Tradition of Hindu Religion
GAVIN FLOOD
I.B.TAURIS
LONDON ■ NEW YORK
Published in 2006 by
LB. Tauris & Co. Ltd
6 Salem Rd, London W2 4BU
175 Fifth Avenue, New York NY 10010
www.ibtauris.com
In the United States and Canada distributed by Palgrave Macmillan, a division of St. Martin's Press, 175 Fifth Avenue, New York, NY 10010
Copyright © Gavin Flood 2006
The right of Gavin Flood to be identified as the author of this work
has been asserted by him in accordance with the Copyright, Designs
and Patents Act, 1988
Frontispiece: Cakra man (Wellcome Ms P511). With kind permission of the Wellcome Trust.
All rights reserved. Except for brief quotations in a review, this book, or
any part thereof, may not be reproduced, stored in or introduced into a
retrieval system, or transmitted, in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without the prior
written permission of the publisher.
ISBN 1 84511 on o (Hb) ISBN 1 84511 012 9 (Pb) ean 978 i 84511 on 6 (Hb)
EAN 978 I 8451 I 012 3 (Pb)
A full CIP record for this book is available from the British Library
A full CIP record for this book is available from the Library of Congress
Library of Congress catalog card: available
Typeset in Monotype Ehrhardt by illuminati, Grosmont,
www.illuminatibooks.co.uk
Printed and bound in Great Britain by TJ. International Ltd,
Padstow, Cornwall
Contents
Preface ix
PART I Theory, Text and History
i Introduction: The Body as Text 3
Tantra, Tradition and the Body 7 Reading Strategies: Text 15 Reading Strategies: Body 20 Experience and Asceticism 24 The Argument of the Book 27
2 The Vedic Body 31
The Political and Social Context 32 Legal Discourse 37 Political Discourse 42 The Highest Good 45
3 The Tantric Revelation 48
The Validity of Tantric Revelation 50
The Pdncardtra Revelation 53
The Saiva Revelation 55
Text and Tradition 60
The Tantric Theology of Revelation 62
The Divinisation of the Body as Root Metaphor 74 Tantric Polity 76 The Tantric Temple 81 Tantra and Erotic Sculpture 83 Possession 87
PART II The Body as Text
5 The Pancaratra 99
Emanationist Cosmology 101
The Purification of the Body 106
The Bhutasuddhi in the Tantric Revelation 108
The Divinisation of the Body 113
Inner Worship 116
External Worship 118
6 Saiva Siddhanta 120
Saiva Siddhanta Doctrine 122
The Tattva Hierarchy 126
The Six Paths 129
The Ritual Process: Initiation 131
The Ritual Process: Daily Rites 138
The Ritual Process: Behaviour 144
7 Ecstatic Tantra 146
Absolute Subjectivity and Indexicality 147 The Circle of Deities in the Body 154 Kundalini and the Cakras 157 Two Ritual Systems 162
8 The Tantric Imagination 171
Vision 172
Gesture and Utterance 174
Icon 176
Indexicality 178
Reading 180
Epilogue 185
Abbreviations and Sources 194
Notes 198
Suggested Further Reading 235
Index 236
Preface
This book represents the application of a general theoretical framework to a body of tantric texts that I have been reading, on and off, for a number of years. That theoretical framework develops the theme of the relationship between subjectivity and text. More precisely, the book offers a description and analysis of the idea that subjectivity is textually mediated within a corpus of tantric texts composed in the medieval period. To give an account of this textually mediated subjectivity is also to give an account of the tantric body. A tradition-specific understanding of self and body is constructed, as it were, through the text. The book therefore does not claim to be a work of Indology as such but draws on Indology to present a particular reading of a range of textual material. This is a reading of the body as represented within those texts, along with a tradition-specific subjectivity that the body entails, and a discussion of the implications of that reading in the context of a broader, historical understanding. The specificity of the claim is that in the Hindu tantric traditions focused primarily on the deities Visnu and Siva in the early medieval period, the practitioner becomes divine through the internalisation of the text, through the inscription of the body by the text, and learns to inhabit a tradition specific subjectivity. The text is mapped on to the body. The range of texts I discuss is from the Vaisnava and Saiva tantric traditions, namely the Pancaratra, the
Saiva Siddhanta, and the non-Saiddhantika traditions often referred to as 'Kashmir' Saivism that developed particularly from the ninth to eleventh centuries. While the examples I discuss illustrate my general point, a much wider range of textual material could have been presented but for reasons of space. I do not focus on later tantric traditions and do not deal with the Sri Vidya, although the general framework I develop would be equally applicable there.
Most of this book was written during a wonderful year as a visiting scholar at the University of Virginia and I should like to thank both staff and students for discussion about the project and their astute observations. The two graduate seminars I conducted were especially helpful in testing ideas and I would like to thank all the students in those classes, including Wijitha Bandara, Suzanne Bessenger, Kristen Calgaro, David Divalerio, Andrew Godreau, Julian Green, Chris Hatchell, Gavin Irby, Sara Jacobi, Slava Komarovski, Karen Lemoine, Bianca Pandit, John Paul Patterson, Matt Rose, Carl Yamamoto, Umeyye Yazicioglu and Yongbok Yi. I would especially like to acknowledge conversations with James Gentry, who coined a felicitous phrase 'variable indexicality' to describe some of this work, Andres Montano, Lynna Dhanani and Craig Danielson. I became a student at the stimulating class on Buddhist tantric traditions across Asia conducted by Professors Paul Groner and David Germano, where I learned much (not least the advantages of team teaching). I also gained a lot from the 'Tantra lunches' organised by Peter Ochs, where 'tantric' topics were opened out for discussion within a wider milieu and in the context of other traditions and other thought worlds. These lunches provided an informal yet rigorous forum and, along with professors Groner, Germano and Ochs, I particularly appreciated the contributions of Jeffrey Hopkins and Jamie Ferriera. This was an extremely engaging experience, true to the dialogic nature that should characterise comparative religion.
There are many debts of gratitude in a book such as this. I should also like to acknowledge those teachers who first introduced me to the study of tantric traditions, Andrew Rawlinson and David Smith at Lancaster, a few conversations with the charismatic Agehananda Bharati, and a dept of gratitude to Alexis Sanderson of Oxford, who
has so often responded to my questions with generosity and cordiality and to Andre Padoux, a great scholar who has done so much to further our understanding of the tantric traditions. The Saiva texts were very much brought to life for me at the Centre d'Indologie in Pondicherry some years ago, where I had the good fortune to discuss these topics with Dominic Goodall and to read sections of texts with the deeply knowledgeable Saivasiddhanta Tattvajna R. Subramanian and Dr T. Ganesan. Frits Staal indirectly introduced me to the tantric tradition of Kerala and to my friend and colleague, anthropologist Rich Freeman, who introduced me directly to that tradition. I should like to thank him for his reflections on our shared interest in linguistic anthropology and for his extremely important theory of ritual possession in the tantric context as the paradigm for the divinisation of icon and priest. His theory has been a strong influence on my own thinking. I remember with fondness the somewhat bizarre situation of reading together, late into the night, sections of the Isdnasivagurudeva-paddhati in an old house in a remote Welsh village. Lastly I should like to thank the I.B. Tauris readers for their encouragement. I trust the publisher's title does not detract from the contents. A grant from the AHRB in the UK allowed me relief from teaching to pursue this book during 2003-04. Unless otherwise stated, translations are my own, although Marion Rastelli's work on the Jayakhya-samhita has been a source of guidance at times when the precise meaning of that text has eluded me. An appendix presents the first English translation of one chapter from the published edition of the Jayakhya-samhita. Although I trust the translations are accurate, I have tried to err on the side of readability for the English speaker. I was unable to incorporate an important article that only came to my attention as the book went to press, namely Barbara Holdredge's 'Body Connections: Hindu Discourse of the Body and the Study of Religion' {International Journal of Hindu Studies, 2/3 (1998), pp. 341-86).
Oui, par le corps
Dans la douceur qui est aveugle et ne veut rien
Mais paracheve.
Yves Bonn efoy, 'L'epars, I'indivisible'
PART I
Theory, Text and History
Introduction: The Body as Text
Writing at the end of the nineteenth century, the eminent Indologist Monier Monier-Williams was able to say that the Tantras are 'mere manuals of mystics, magic and superstition of the worst and most silly kind' 1 and that with these texts and their traditions 'we are confronted with the worst results of the worst superstitious ideas that have ever disgraced and degraded the human race'. 2 On this view, the Tantras are a far cry from the nobility of Vedanta or the dignity of the Buddha. In complete contrast, almost a hundred years later at the end of the twentieth century, Bhagavan Shree Rajneesh was able to write 'tantra cannot be understood because tantra is not an intellectual proposition: it is an experience. Unless you are receptive, ready, vulnerable to the experience, it is not going to come to you.' 3 On the one hand we have the critical Indologist writing from within the horizon of the values of his own culture about texts and traditions in clear antipathy to them; on the other we have a modern 'mystic' or experientialist writing from within the horizon of values emerging in late modernity.
This book might be seen as corrective reading of both views in that it seeks to understand the tantric traditions in their historical and doctrinal contexts and to offer constructive readings of the texts that are true to Indology and sympathetic to the internal concerns of the traditions, while at the same time offering a 'third-order' discourse
about them. 4 More specifically, I wish to understand the tantric body, how the body has been conceptualised by tantric traditions and the use of the body in tantric visions of power and liberation. There are complex problems here and we will need to examine the body in terms of technique, in terms of representation, and in terms of formation. Furthermore, we need to ask how techniques of the body and the representation of the body (in metaphor and textual description) interface with Indian scriptural traditions and sociopolitical structures. On the one hand we have techniques of the body, methods or technologies developed within tantric traditions intended to transform body and self; on the other we have representations of the body in philosophy, in ritual and in art. 3 Both of these areas - techniques of experience and representation of body and experience - are intimately linked. Representations (particularly icons of deities) are not simply passive texts but are performative, used in 'life transforming practices', 6 and, conversely, techniques of the body themselves entail representations of it, especially in ritual where the body becomes the deity or icon. Indeed, both representation and technique come together in the divinisation of the body which, as we will see, is the hallmark of tantric culture. We need therefore to explicate the interrelated distinctions of representation/technique and doctrine/ritual, which are encompassed by the text/body distinction. One might even say that as text is to body, so representation and doctrine are to technique and ritual; that is the former is expressed in the latter and the latter is articulated in the former. The text is expressed as body and the body articulated in the text.
I therefore wish to present an argument to support three interrelated views. First, in spite of divergent metaphysical claims and different social locations, the conceptualisation of the tantric body and its expression follows certain principles or processes that might be best expressed in the claim that the tantric body becomes inscribed by the text. What we might call an entextualisation of the body occurs in tantric traditions that is specific yet allows a divergence of views and practices. The body is moulded within the constraints of historical tradition, even in its attempt to transcend those constraints. Second, the body, functioning as the root metaphor or topos of the tantric traditions, operates at different levels
of practice and discourse. The body is the vehicle for imagining and conceptualising tradition and cosmos such that the structure of the cosmos, forms of language, and text and tradition are themselves understood in terms of the body. Representations of the body occur in texts and in the techniques of the body such as ritual and asceticism; the body itself functions as a representation of tradition, text and cosmos. While I think the claim that the body becomes the text or is inscribed as text is true of all scriptural traditions, this book intends to examine the specificity of the claim within Hindu Tantrism. Third, operating within these claims about the body and tradition is the idea of a tradition-dependent subjectivity; that the index of the first-person pronoun, the T, operates within realms of practice and discourse constrained by text and tradition. By 'subjectivity' I do not intend a monad set against the objectivity of the world but rather interiority formed through language and tradition. This linguistic agency is not fixed but in dialogical relationship with others and with social structures and might be called 'variable indexicalityV This is another way of expressing the body as text in that when the body functions within the tradition-specific activities of reading, ritual and asceticism, different notions of the subject come into view. The content of the T is filled out in different ways in these contexts. For example, the tantric practitioner, as we shall see, identifies his body with the cosmos and deity in daily ritual and in yogic practice, identifying himself with something outside of himself that he then becomes.
While my main purpose is to locate the tantric body within the history of ideas, practices and institutions that made up the early formation of medieval India, I would also contend that this reflection raises questions of contemporary cultural and theological relevance. The tantric body is of more than historical interest, as is evident through its mass appropriation in consumerist culture, and raises such challenging cultural questions about the nature of the body, about the relation of the body to language, about human relationships, about the relationship of the human to the wider ecosystem and raises such challenging theological and philosophical questions about the relation of the body to any transcendent reality and about ways traditions construct the self, as to be worth taking seriously as
a resource in our response to such questions. While it is important to maintain discourses within the boundaries of tradition in order for them to retain meaning and relevance for particular communities of readers, it is also germane, enriching and challenging to engage theologically and philosophically with thought systems outside of those discourses. Although I do not directly address questions of theological relevance, my third-order reflection nevertheless goes beyond the description of text and tradition established through the mediating, second-order discourse of philology and history.
In the following pages, the reader will find an argument that the tantric body can only be understood in terms of text and tradition. In my local phone book there is an advertisement for 'cakra balancing' for a reasonable fee (in this respect clearly in accordance with tantric daksina). Implicit here is a Western appropriation of the tantric body that we might see as a reification of it, and a view that the tantric body is something that can be revealed for those with the means to do so. 8 The argument of this book, on the contrary, is that in its medieval Indian context the tantric body is not a given that is discovered but a process that is constructed through dedicated effort over years of practice. The centres of power or cakras within the body that the phone book advert alludes to can be best understood in terms of entextualisation, the body inscribed as the text, which expresses principles at work within the logic of tantric ideology and practice. Any distinctions between knowing and acting, mind and body, are disrupted by the tantric body in the sense that what might be called imagination becomes a kind of action in tantric ritual and the forms that the body takes in ritual are a kind of knowing. Borrowing a phrase from William Blake (and if the adjective 'tantric' can apply outside of Hindu and Buddhist scriptural traditions, then surely he is a good candidate for its application) the tantric body is a 'corporeal understanding'. 9 This corporeal understanding shows itself in the great emphasis on transformative practices in the tantric traditions, ritual inseparable from vision, the body becoming alive with the universe within it, and vibrant with futurity in the anticipation of the goal of the tantric paths.
Understanding the tantric body in its historical locations is no easy task and it is not simply a matter of contrasting an inauthentic
Western view of the tantric body, outside of tradition, with an authentic tantric view, moulded in accordance with tradition. The very category 'Tantra' or 'Tantrism' is contested and itself must be seen in the context of the history of scholarship in the West and colonialism, as some scholars are doing. Understanding the Western tantric body in relation to modernity and postmodernity is a topic in itself, 10 and the only claim I wish to make about that body is that it is modernist in reflecting the reifying tendencies of modernity along with the idea of the practitioner as free-floating individual. By contrast, the traditional tantric body of medieval India is more fluid in terms of its lack of reification and at the same time more conservative in being deeply embedded in traditional understandings and categories. The tantric body is formed in accordance with received tradition, in accordance with scriptural revelation, and in accordance with the somatology of the wider culture. The cultivation of a tantric subjectivity is not the cultivation of individuality (see pp. 12-13).
Tantra, Tradition and the Body
The tantric traditions arose during the early centuries of the common era, developing in Buddhist, Jain and Hindu contexts. The vast body of tantric texts are inseparable from the traditions that gave rise to them. Saiva, Vaisnava and Sakta Tantras were believed by their followers to have been revealed by Visnu, Siva, and the Goddess (Devi), and there were even Tantras revealed by the Sun (Surya), now all lost, whose followers were called Sauras. 11 There were also Jain Tantras believed to be the word of Mahavira and, above all, Buddhist Tantras believed to be the word of the Buddha, which became incorporated into the vast Buddhist canon between c. 400 and 750 CE, to this day integral to the living traditions of Tibetan Buddhism. 12 Using the term 'Hindu' to refer to the Saiva, Sakta Vaisnava and Saura material is anachronistic as the term was used by the Persians simply to denote the peoples of the subcontinent, 13 although there are usages of it as a term of self-description by 'Hindus' as early as the fifteenth century in Kashmir and sixteenth
century in Bengal to distinguish people who shared certain cultural values and practices (such as cremation of the dead, veneration of the cow, styles of cuisine and dress, or shared narratives) from Muslims ('Yavanas'). 14 It was not a common designation until the nineteenth century. But the theistic Tantras and traditions, those of Visnu, Siva and the Goddess, are interrelated and share common structures of practice and belief that can be distinguished from those of the Buddhists and Jains by their proximity to the Vedas, orthodox Brahmanical revelation, and their interpreters. The term 'tantric tradition' refers to those religions, or 'ways of life' to use Inden's apposite phrase, 15 that claimed to develop from textual sources referring to themselves as 'tantras', regarded as revelation, the word of God, by their followers. This diverse tantric revelation must be seen in contrast to the ancient, orthodox Brahmanical revelation of the Veda that the Tantras reject completely or accept as a lower level of scriptural authority. In contrast to the Hindu Tantras, the Buddhist Tantras do not respond to the vedic tradition but rather look to Mahayana Buddhism and see themselves as a development of it, even though much Buddhist tantric material, the YoginI Tantras, was probably derived from Saiva prototypes. 16
Arriving at definitions of 'Tantra' and 'Tantrism' has been notoriously difficult and has varied between presenting external accounts of a phenomenon named 'Tantrism' 17 and internal accounts of what the term tantra refers to. An important indigenous distinction is between tantrika, a follower of the Tantras, and vaidika, a follower of the Vedas. This distinction operates across the sectarian divides of Saivas, Vaisnavas and so on. The former refers to those who follow a system of ritual and teaching found within the Tantras, in contrast to those, especially the Brahman caste, who follow the Veda as primary revelation or sruti (and so called Srautas), or who follow the later texts of secondary revelation called smrti (and so called Smartas). 18 The issue is complicated, however, by some vedic Brahmans, particularly Smartas, observing tantric rites and, as Padoux has observed, some texts in the vedic tradition, namely Upanisads, being clearly tantric in character, 'which tantrika authors (Bhaskararaya, for example) consider as confirming the validity of tantric teachings and practices'. 19
The primary designation of the term tantra is a 'loom' or the 'warp' of a loom, with the metaphorical implication of system or framework. It is derived from the verbal root tan, to extend or stretch and so, perhaps not insignificantly, is related to tanu, 'body'. It came early on to designate a text and there are several examples of the term being used for texts that are clearly not within the tantric tradition, such as the collection of stories the Pancatantra or the famous Mlmamsaka work the Tantravarttika. The term tantra as a noun is a term of self-description that refers to specific texts of revelation and is also a term designating a system of revealed teaching that leads to liberation and power. In this sense the term tantrasastra is used, which, as David White observes, is the closest indigenous category to the English 'Tantrism'. 20 The term agama is used in some Saiva texts as a synonym for tantra with the implication that the text is a disclosure that has come to us. Indeed, Abhinavagupta uses the term to refer to the tantric revelation in general as the 'one revelation' (ekdgama) (see pp. 58-60). The term 'tantra' refers not only to texts but to system and, as Padoux observes, asmin tantre simply means 'in this system'. 21
Some scholars have presented Tantrism in terms of a list of characteristics, such as locating a bipolar energy within the body, 22 while others have offered more precise definitions, which are in fact theories, such as seeing Tantra as a quest for power akin to the king's quest for political power. Drawing on Madeleine Biardeau, Andre Padoux offers the understanding that Tantrism is 'an attempt to place kama, desire, in every sense of the word, in the service of liberation', 23 and David White further develops this in terms of energy. 24 The word 'power' has perhaps a more negative semantic field in English than 'energy', and power relates to the political and historical world in a way that 'energy' does not, although both can be renderings of the Sanskrit s'akti. One interesting thesis presented by Ron Davidson in the context of tantric Buddhism is that the central 'sustaining metaphor' of the Mantrayana, or tantric Buddhism, is that the path of the practitioner is akin to the path of the king on his way to becoming an overlord (rajadhiraja) or universal monarch (cakravartin), expressed through the forms of consecration, self-visualisation, mandalas and 'esoteric acts'. 25 This
focusing on the political dimension of the metaphor of power is clearly important, and power suffuses the concerns of the tantric traditions. The Tantras offer their followers power to achieve world transcendence or magical power over supernatural entities in order to achieve worldly success, such as seduction of a desired woman or the destruction of enemies for a king. Sanderson has pointed out that the tantric traditions of power defined themselves against the vedic tradition of purity and saw their power as lying in the transgression of vedic social norms. 26
Davidson accompanies his claim about the central metaphor of the Mantrayana with a discussion of 'polythetic' categories that function 'to identify prototypical examples that operate as cognitive reference points'. 27 That is, rather than a 'monothetic' understanding of Tantrism, such as Tsong-ka-pa's definition of Tantra as visualisation of oneself as the Buddha or deity, we need to understand Tantrism in 'polythetic' terms. That is, no one thing can be taken to describe a category but, rather, prototypical examples can be identified which may not share all of the traits within the category. As Brooks observes, 'tantric phenomena need not possess all the defining characteristics of the taxon "tantric" and there is no a priori justification for deciding that any single characteristic is the most definitive.' 28 While perhaps the terms 'monthetic' and 'polythetic' are somewhat unnecessary, the now popular use of prototype theory does have force in the understanding of cultural categories. 29 As discussed by Davidson, a robin (both English and American) is a prototypical bird, whereas an emu is not, but is still within the category. A member of a category does not need to share all characteristics to belong: categories have 'fuzzy' edges. 10 Of course, any inclusion in a category as prototypical will involve judgements which need to be based on careful consideration, comparison and scholarship. Due to scholarly endeavour, especially over the last fifty years, we now know enough about tantric traditions to make some claims about them and to make judgements about prototypicality. One such judgement that I would wish to make is that tantric traditions must be understood in terms of pre-modern scriptural traditions, and another is that they involve the divinisation of the body, which is way of saying that the body is inscribed by the text.
The Body as Text u
Davidson's account of Tantrism in terms of power is important and it is surely germane to point to the political dimensions of the tantric practitioner that have been generally neglected or ignored (probably partly due to the clear separation of 'politics' from 'religion' that has, rightly or wrongly, characterised Western scholarship). The practitioner, in Davidson's reading of the texts, seeks to assume kingship and exercise dominion. We could, however, read this in a slightly different way, that the central tantric metaphor is indeed, as Tsong-ka-pa identified, divinisation and that the model of kingship - the king undergoing consecration and so on - is in fact the king becoming divine. The divinisation of the king through ritual consecration is directly akin to the divinisation of the icon in a temple and the divinisation of the practitioner in daily ritual (or even the divinisation in possession). More fundamental than the metaphor of kingship is the metaphor of transformation into a deity. The idea that to worship a god one must become a god is a notable feature of all tantric traditions, even ones which maintain a dualist metaphysics.
The empowering of the body, which means its divinisation, is arguably the most important quality in tantric traditions, but a quality that is only specified within particular traditions and texts. Becoming divine is an ancient trope in Indian civilisation. As Hocart observed long ago with reference to the consecration of the vedic king, it is fundamental 'that the worshipper becomes one with the god to whom the worship is addressed'.' 1 Divinisation in tantric ritual reflects this general idea but is text- and tradition-specific in terms of content and in the explicit focus on the divinisation of the body as the enactment of its revelation, as this book hopes to demonstrate. The practitioner in ritual contexts becomes divine such that his or her limited subjectivity is transcended or expanded and that subjectivity becomes coterminous with the subjectivity of his or her deity, which is to say that the text is internalised and subjectivity becomes text-specific. This is clearly in line with Tsong-ka-pa's understanding in a Buddhist context and also makes sense in a theistic 'Hindu' one. 32 While the idea of liberation as becoming one with the absolute {brahman) has a long history in Brahmanical thinking from the Upanisads, the ritual construction of the body
as the deity through the use of magical phrases or mantras is proto-typically tantric. 13
In a broader sense, the tantric traditions are examples of forms of practice and reflection handed down through generations which locate themselves historically by reference to a foundational text or group of texts, believed to originate in a transcendent source. This is, of course, true of many traditions including Islam, Judaism and Christianity, as well as vedic tradition. But while this is a general point, it is nevertheless an important one, for processes of identification and entextualisation can be identified within wider scriptural traditions that are also typical of tantric traditions. Scriptural traditions all developed before modernity and before the Kantian understanding of the self as an autonomous agent; an idea that connects with the notion of the citizen who has civic responsibilities yet who remains distinct from the social body and an individuality that comes to stand against tradition. In scriptural traditions, such a notion has been alien, and the self is an index of a tradition-specific subjectivity, formed in particular ways in conformity to tradition. 34 In scriptural traditions, the self is constructed through ritual and the development of a tradition-specific interiority or variable indexicality that is not individual in the contemporary, de-traditionalised sense (characterised by fragmentation and alienation). Scripture-sanctioned rituals serve as identity markers for communities in medieval India, and, although these boundaries can be transgressed, 35 such transgression always assumes their existence. The self in such communities is bounded by text and ritual. Such a tradition-specified self, as Maclntyre reminds us, develops philosophy as a craft or techne and needs to develop his or herself into 'a particular kind of person if he or she is to move towards a knowledge of the truth about his or her good and the human good'. 36 Tantra can itself be seen in terms of techne, and the suffix tra expresses the means or instrument of an action expressed by a verbal root. 37 Thus as man-tra might be rendered 'instrument of thought' 38 so tan-tra might literally be taken to mean 'method or instrument of extension', perhaps with the implication that it is the self or body that is extended to become coterminous with the divine body. I do not intend this etymology (nirvacana) to be taken too seriously, but it is nevertheless suggestive.
The specificity of the tantric traditions lies in the ways in which they form a subjectivity, the ways in which the subject of first-person predicates, the T, becomes an index of tradition, and the way the body becomes entextualised. Patterns of text are mapped on to the body in ways particular to Tantrism and in response to other ways of mapping texts on to the body, especially vedic ones.
The theory I wish to present is simply this. The tantric body is encoded in tradition-specific and text-specific ways. The practitioner inscribes the body through ritual and forms of interiority or asceticism, and so writes the tradition on to the body. Such transformative practices are intended to create the body as divine. This inscribing the body is also a reading of text and tradition. Indeed, the act of reading is of central importance in the tantric traditions. The fact that the texts were written is important and has sometimes been underestimated in focusing on orality/aurality in the transmission of texts. But the texts were written in Sanskrit and in doing so their authors were consciously locating them within what Sheldon Pollock has called the 'Sanskrit cosmopolis'. 19 The texts were intended to be read and heard by those with the requisite authority, to be brought to life, and to be performed. The importance of the written word here is evident from the commentaries upon the primary texts by the later tradition. The importance of reading the texts is further suggested by the presence of ritual manuals (paddhatis), 'cookbooks' that served to instruct and remind practitioners about how to undertake particular kinds of performance and about particular tenets of a system. The tantric body, constructed as a public act (even if limited in its public nature through secrecy), is in turn 'read' by traditional practitioners in so far as some tantrikas wore external signs of their cultic affiliation while others disparaged such signs, retaining their tantric affiliation as 'secret'; 40 such secrecy is an overcoding of the body. That is, while some tantric traditions overtly reject vedic tradition and normative, caste and feudal society of medieval India, most must be seen as adding their own writing of the body on to the traditional vedic writing or as reconfiguring the vedic tradition in terms of the tantric. We see this, for example, in the Saiva traditions of Kashmir so eloquently accounted for by Abhinavagupta (c. 975-1025 ce). For him, tantric rites were supererogatory to vedic practice. The body,
the vedic body, is overwritten by the practitioner who constructs a tantric body through a further superimposition of rites and the internalisation of a tantric ideology. Thus, in his famous statement (probably a standard saying), Abhinavagupta writes that externally one follows vedic practice, in the domestic sphere one is an orthodox Saiva, but in one's secret life one is a follower of the extreme anti-nomian cult of the Kula which involves the disruption of the vedic body through ritual transgression of vedic norms and values. 41
In locating the tantric body within an account of text, I intend to discuss a clearly articulated cultural form that has developed well beyond its roots. There is much speculation about the origins of Tantrism. On the one hand the origins have been seen in an autochthonous spirituality or Shamanism that reaches back to pre-Aryan times in the subcontinent, yet textual historical evidence only dates from a more recent period. While certainly there are elements in tantric traditions that may well reach back into prehistory - particularly the use of skulls and the themes of death and possession 42 - we simply do not have sufficient evidence to speculate in this way. As Robert Mayer has shown, there is no evidence for a non-Aryan substratum for Tantrism, which must be understood as a predominantly Brahmanical, Sanskritic tradition with its roots in the Veda. 43 In an important book on the origins of Indian civilisation, Sergent has argued that our main resources for understanding the past are linguistic and archaeological. 44 There is no early archaeological evidence for tantric traditions beyond the common era, and while there is textual evidence for a cremation ground asceticism as far back as the time of the Buddha, 41 as well as tantric-like goddesses in the Veda, 46 the specificity of the tantric revelation appears more recently in the history of South Asia. India clearly inherits its earlier Indus civilisation (as shown, for example, by the persistence of common kinds of measurement) 4 ' but specific tantric elements cannot be located other than in very general ways. Traditions are constantly reconfigured in the light of contemporary situations and there is no reason to think that the tantric traditions are any different. While of course receiving forms of practice and ideas handed down from the past, the Tantras at the time of their composition were a new revelation that transcended the older, vedic
texts. While concerned with body and experience, this tantric body can only be accessed via the texts that form it.
Reading Strategies: Text
The argument I wish to present is not historically neutral in the sense that the Encyclopaedist mind-set of Enlightenment modernity, described by Maclntyre, might understand neutrality as a single framework within which knowledge is presented. 48 Nor does it assume that all knowledge is purely subjectively constructed and that history masks a will to power of particular interest groups. Rather, agreeing with Maclntyre's general argument, I take rational inquiry (such as this) to be enabled by traditions of inquiry, and such inquiry is less a discovery of the past and more the construction of the past from a particular perspective or standpoint. The past is constantly reconfigured in the light of new evidence for a given purpose. That there are degrees of accuracy in such reconfigurations is not in question. Clearly there are positions and readings of the past that contain such prior ideological commitments as to distort the past, as we see in more recent reconfigurings of Indian history seen through the lens of a hindutva ideology. But this very claim can itself only be based on the presentation of evidence in a different vein, drawing from a rationality of historical method that has developed within the Western academy, a rationality which would, of course, claim a methodological superiority to the hindutva reading. But the point is that the presentation and weighing of historical evidence is always within a tradition of inquiry and judgement. Yet this tradition that claims universal truth accessible through an objective, repeatable method needs to acknowledge reflexively that it is itself a tradition of inquiry that never attain its own declared universalist goal. The Encyclopaedist claim to objectivity and neutrality is itself a tradition of presentation and assessment according to criteria developed only within that tradition and not, as that tradition claims, the discovery of a single, neutral narrative. 4 '
To establish the account of the tantric body I have briefly described above, I am bringing together two primary traditions of discourse,
Indology and what might be described as a post-foundational religious studies. At one level the modernist or objectivist assumptions of Indology are fundamentally opposed to a post-foundational understanding of text as infinitely interpretable. Yet any rigorous post-foundational understanding must assume Indology as the discipline that provides the basic materials from which to develop. Indology is the philological study of Sanskrit texts which is the sine qua non for the study of tantric traditions. Without Indology there can be no study of Tantrism. But while one can understand the claim that philology is the eradication of subjectivity in that the objective system of grammar, the language itself, eradicates subjectivist interpretations, there is nevertheless a further level of reading beyond the philological, which intends to place philological readings in a broader context. We might say that philology is indispensable in establishing the plain sense of the texts, yet we must go beyond philology to establish interpreted senses. If philology creates Nietzsche's pathos of distance, it is nevertheless also the case that a text is nothing until it is read and interpreted. 50
I shall defer a discussion of the nature of tantric texts to the next chapter, but suffice it to say for now that these texts are set within the context in which they echo and reflect other texts and in which textual agency is complex because often the texts have multiple authors or were composed over a long period of time. In reading these texts we need to be sensitive to the wider textual field in which they are located. To use Inden's phrase, we need to move from philological texts to dialogical texts. 51 There is a useful distinction within rabbinic Judaism between the plain sense of the text (peshat) and the interpreted sense (derash). The plain sense is the foundation upon which the interpreted sense is built, 52 although even the plain sense is immediately interpreted once read. We might say that the plain sense operates as a constraining force upon the interpreted sense. The interpreted sense should not disrupt the plain sense to the extent that it contradicts it, yet the plain sense is never enough for a particular situation. Interpreted senses are always necessary to bring some meaning to life for some particular community of readers. A post-foundational religious studies develops an interpreted sense of the texts established through Indology, one which takes seriously
the implicit and explicit philosophical claims of the texts but does not share (indeed cannot share) the texts' theological presuppositions. This book is no tantric theology but a dialogical reading that stands outside of the texts while partially entering into them in an act of imagination that allows for their reconstruction and reconfiguring in a new mode. That new mode is the account I present of the tantric body as text.
While Indology and post-foundational developments in religious studies are fundamental to my reading strategy, there is also implicit in the book a theory of reading religious texts that I have developed with my colleague Oliver Davies, which needs briefly to be explicated before we proceed. 53 The way in which the body becomes the text in tantric traditions needs to be understood in terms not only of how the content of texts is imposed upon the body, but in terms of the very nature of the texts and how they are received.
Tantric texts can be divided into those texts of primary revelation, the Tantras themselves believed to be the word of the deity, usually in a dialogical form with the Goddess (Sakti) asking questions of the Lord (Bhagavan), although in some texts the relationship is reversed, and secondary works of commentary expounding the meaning of a text, and works describing practice such as ritual manuals. The Tantras at some point in their history, quite early, were fixed in writing. This is not to say that there were not different versions of texts - the Saiva Siddhanta theologian Ramakantha, for example, had a number of readings of the Kirana-tantra to choose from 54 (see p. 64) - but it is to say that the work achieved some stability through time. In this sense the Tantras can be contrasted with the Vedas, which were not written but nevertheless acquired a high degree of fixity due to methods of conveying them accurately within schools of recitation. While the Tantras seem to have been written, they were often accompanied by oral teachings and commentary, which is corroborated by the sometimes obscure or pithy nature of the material, and closely linked to systems of acceptance or initiation.
Given that the Tantras achieved some stability through time, we can also say that the meaning of the text and its function became determined by the process of transmission. This is not to go against the distinction between plain and interpreted sense, but rather to
say that the text remained alive by being received anew through the generations. The texts of primary tantric revelation probably have multiple authorship and were composed over several generations, which makes agency within the text complex. Indeed, we need to speak of agency within the texts themselves rather than the agency of an individual author. The texts in their intertextuality take on a life of their own. The intentionality of the text, which we might call the 'narrator' and which Bakhtin called the 'author', interfaces with the intentionality of the reader or community of readers who internalise and reconstruct the text in their own lives. As in all texts regarded as revelation, the Tantras were brought to life in the act of reading or reception and in their performance. The receiver of the Tantra, the tdntrika, for whom it is divine word, internalises the text through a process of identification which usually involves ritual enactment. The indexicality of the reader interfaces with the indexicality of the text, and the subject of first-person predicates, the T, becomes an index of tradition (arguably, Greg Urban has suggested, through the function of the floating signifier itself"). The reader also positions himself (and it is usually a he in the tantric traditions) in response to the notional reader assumed by the text, usually an initiate. 56 The reader interprets and internalises the text in the act of understanding and in turn conforms himself to the reader implied within the text. The reader does not simply interpret; the text makes claims upon the initiated reader, which has significant, life-transforming effects.
The sacred text is made 'one's own' through reading and performance, and the 'reader' conforms to the implied reader of the text. This is as true of the tantric traditions as of other scriptural religions. Such a reconstruction of the text in subjectivity is fundamental to the process of textual transmission and religious identity formation. The linguistic anthropologists Greg Urban and Michael Silverstein have identified two processes in textual transmission that they call entextualisation and contextualisation, the taking of a text out of one context and recontextualising it in a new, which are simultaneous. 5 ' The speech agent retrieves the text back into the living matrix of speech through meaningful acts of reading and performance, through encoding the body with the text. Such acts
of reading retrieve a semantic entity from the past, the origin of the revelation, into the present field of meaning. Indeed, commentary upon revealed text is just such a claiming of meaning, the fusion of the world of the text with the reader's own world and the attempted persuading of others of one interpreted sense. Such a reception or reading is communal and tradition-based, only taking shape within communities that have themselves been shaped by prior acts of reading of the same text or group of texts. Radically new or innovative readings might result in new communities being formed and groups questioning the received wisdom of the old tradition. Thus the Tantras of the Saiva Siddhanta have been received by a community of Brahmans who have themselves been formed by the tradition constrained by the text. But monistic Saivas in the ninth and tenth centuries offered corrective readings of the old tradition which helped to form a new community of reading. A community reads its own core texts and acts them out, readings that are themselves already governed by the historical life of the community grounded in successive and often corrective readings of the same text or texts. The plain sense of the text gives rise to new meanings in new contexts. The religious reader or community of readers assumes that the voice within the ancient texts, the voice of God in the case of the Hindu Tantras, has present force. This is a fundamentally important point in the transmission of traditions, for only because of the present force of the text for a reader or community of readers does the text have relevance, a relevance principally enacted through ritual.
For tantric traditions the immediacy of this divine voice can only be accessed through the structures of tradition, involving structures or systems of access, namely initiation, which give privileged access to the text's authenticity. As we will see, the Saiva Siddhanta demands an initiation into the tradition (samaya-diksa) to gain access to its texts. This laying claim by the tradition to the space between the reader and the text is to lay claim to the temporal and spatial structures of the world within which the tradition-constrained act of reading takes place. Thus for the tantric reader there is a strongly cosmological dimension to any act of reading and any enactment of the text in daily ritual. The world of the practitioner who acts out
the text is itself constructed by the text. There is, then, a complex process of enacting an interpreted sense of the text in relation to the plain sense, and of enacting the injunctive claims of the text on its receivers. The Tantras have a unique intentionality that makes claims on its receivers, who have enacted those injunctions through to modernity.
Reading Strategies: Body
Having given some account of religious or revealed text, the modes of approaching such texts, and a theory of scriptural reading, it remains to make some remarks about what I understand by the term 'body' in my title and how I shall 'read' the body. In what ways could the 'tantric' body be distinct from any other kind of body? Clearly the tantric body is a different order category to 'male' or 'female' body, or 'young', 'beautiful', 'lithe', 'sick' or any number of adjectives that could be placed before the noun. The link I wish to establish between body and text more generally, and the tantric body and tantric text specifically, needs to be placed in a broader context of Western academic concern with the body.
The body has become the focus of many disciplines in the academy including anthropology, sociology, cultural studies, philosophy, literary studies, religious studies, and sub-branches of these, particularly medical anthropology and the related enquiry into pain, sexuality, emotion and agency. The development of interest in the body over the last thirty years and the 'somatisation' of social theory 58 might themselves be of interest for the sociology of knowledge as an index of wider cultural values, values that reflect a concern with gender, the post-existential condition in the West after the Second World War, and the recognition that we are embodied beings. Csordas has observed that the turn to the body in the human sciences is linked to the development of the postmodern condition of fragmented meanings and that this turn reflects an attempt to grasp a stable centre, 59 yet this centre remains elusive because the body is not a static, biological given, but has a history. The body changes through time and across cultures.
I refer the interested reader to essays by Turner 60 and Csordas 61 for a coherent account of the development of interest in the body in the Western academy, especially in sociology and anthropology. To describe these developments here would take us too far from our project, but it is worth pointing out that early interest in the body and body symbolism begins in the Durkheimian tradition of French sociology, particularly with an important paper by Marcel Mauss on techniques of the body'' 2 and with Hertz's influential work on right and left symbolism. 6 ' Mary Douglas's Purity and Danger (1966) marks a turning point and in Natural Symbols (1970) Douglas makes an important distinction between the social body and the natural body. This, along with the publication of Blacking's The Anthropology of the Body in 1977, began an interest in the body that developed apace in the human and social sciences, which have demonstrated the diverse ways in which the body is conceptualised and formed. 64
In parallel to this sociological/anthropological concern, the body became the focus of inquiry for philosophical phenomenology, especially the work of Merleau-Ponty and his Phenomenology of Perception (1945), which itself partly draws on the work on body image in the 1930s by a number of psychologists, most notable among them being Paul Schilder. 65 Turner observes that, whereas the French phenomenologists are interested in the 'lived body', Douglas is concerned with the body as a metaphor of socio-cognitive mappings of reality. 66 My concern here is with both, and both are brought together in my argument. On the one hand the tantric body is a metaphor that maps the cosmos, particularly in ritual activity; on the other hand the tantric body is a lived body that performs that mapping, a performance that had and has existential force in the lives of tantric practitioners. The tantric body is both a metaphor of tantric ideas about the cosmos and the human person and the lived body of the practitioner who performs or enacts those ideas.
In arguing for this connection between representation and technique, idea and performance, text and body, the book implicitly and sometimes explicitly draws on the social scientific work on the body carried out over the last century that I have alluded to above. Many writers in the area of cultural theory, such as Donna Haraway,
have highlighted the politics of body representation and argued that attempts to essentialise the body operate in the service of a hegemonic discourse that functions to maintain cultural power interests, particularly a biological discourse that links a proposed givenness of the body to (oppressive) social roles. 67 There can be no uncontested nor unpoliticised definition of the body. 68
This general orientation of theory in favour of a socio-political construction of the body and a suspicion of essentialism is within what can broadly be described as genealogy (which is often subsumed under the - inappropriate - title 'critical theory' 69 ). Indeed, much literature and analysis of the body in culture and society have been undertaken within the genealogical tradition of academic discourse, ultimately stemming from Nietzsche and developed by Foucault, that claims that the body is the locus of contested power. The body is inscribed, both hegemonically by the self and by external relationships, in accordance with the power structures of a given society through time. The laying bare of these relationships and forms of inscription through genealogical analysis is an attempt to dissolve them and thereby to offer liberating social critique. Much of the work of feminist scholarship, for example, has been concerned to uncover foundations of patriarchal power upon which particular, limiting constructions of the female body have been built.' 0 But generally genealogy offers no positive proposal, only critique and a hermeneutics suspicious of all cultural formations as hiding egregious, oppressive power relations between groups.
While being sympathetic to many claims of the genealogists, I do not cohere with this view. The genealogical account of the body would wish to claim that it is culturally or socially constructed and that the construction of the body is its inscription by particular power relationships. The body is not a given but embroiled in a politics that needs to be negotiated throughout life. But while the body is an 'ambiguous space', in Foucault's phrase, 71 it is nevertheless a contained ambiguity, contained (at least until recently) by its genetic code, by its temporal structure and the inevitability of birth and death. Indeed, it is here that we see one of the limits of the genealogist's social constructivist position. While conceptualisations and practices of the body vary, there is a givenness of temporality in
that the body is born, ages and dies. The temporality of being born, aging and dying presents a boundary within which any formation of the body must function. This boundary of temporality therefore means that the body contains inherently within it a narrative structure. To speak of a body is to speak of temporality, and to speak of a body is to speak of narrative. Narrative and the living body are inseparable. The story of a life is the story of a body changing through time, and such a story inevitably entails the stories of others, for, as Maclntyre observes, we are the co-authors of our own narratives. 72 The narrative structure of the body, being born and dying, therefore entails communities of other narrative bodies and the interrelation of bodies through time. Thus the body entails tradition and culture. Furthermore the narrative structure of the body displays a natural affinity with sacred text inasmuch as both are grounded in temporality. The sacred text emerges out of tradition, which constantly reconfigures its narrative through history, and the body in tradition is formed in accordance with this temporality. As I have argued, the narrative of the body is the formation of subjective coherence through the linking of the indexicality of the subject with that of the 'text', an argument that can be fully illustrated as regards the tantric body.
If the first problem with the genealogists' account is their narrative constructionist position and the rejection of a narrative structure inherent in the body, a second related problem is that the only historical telos of the body, on this account, has been the will to power. This view is generally at odds with scriptural traditions which maintain, on the contrary, that the narrative structure of the body is teleological in aspiring to some human good beyond the political inscription of power. With regard to the tantric body, while a genealogical-type analysis might reveal the ways in which the tantric body is linked to traditional power structures, to the cult of the king for example, we need to accept the claims of tradition concerning the transcendent goals of the tantric body as having a legitimacy that can be challenged only on a priori grounds. The goods of tradition are fundamentally opposed to a genealogical analysis of late modernity whose goal is ultimately the analysis of power in tradition in order to dissolve that power.
Experience and Asceticism
When speaking about the body the problem of experience must inevitably be addressed. On the one hand we have the body as representation, as symbolic system that encodes a culture's ideas and practices; yet on the other we have the reality of the lived body, that we 'experience' worlds through the senses and body and that for human beings this is our primary mode of functioning (there may be others such as dream or trance states in which there is no awareness of the physical body). Yet we must be cautious of the term 'experience', especially in relation to religion, for its meaning is extremely opaque, and while the English word has a resonance in contemporary culture in that it legitimises particular ways of thinking and behaving, its universal applicability in an unexamined form must be brought into question.
An important current in modern Western thinking about religion, probably stemming from Schleiermacher, who understood religion as a feeling of absolute dependence, and mediated through Otto, has been to emphasise experience as being at the heart of religion. Indeed, many have claimed that beyond differences of doctrine and practice there is an experience common to diverse cultures and histories, and that if we strip away this overlay we will discover a common core experience, variously expressed as a sense of divinity, a sense of the 'numinous', of merging into an ocean of joy, as becoming one with the divine, and so on. Diverse religions are different paths to the same goal of a unified mystical experience. This has been called the 'common core' theory of mystical experience, or, to use Huxley's phrase, 'the perennial philosophy' view. 71 Others, such as Steven Katz, offered strong refutations of this view, claiming that mystical experiences are tradition-specific, strongly linked to language and the linguistic construction of the world.' 4 There is no space here to review this literature and assess the arguments, but the argument I present is clearly sympathetic to the critique of perennial philosophy, yet would not wish to dismiss all claims to universality. The Katz position, standing at the beginning of the linguistic turn, highlighted the importance of language in the formation of experience. Language and experience are mutually implicated, as there are no
pre-linguistic epistemic givens in this view. All cultural forms are pervaded by language, and we need to know a lot before engaging in the practices of religion, practices that involve sustained learning and internalising of tradition. Inhabiting a religious tradition is more like learning a skill than acknowledging propositions. I think this needs to be complemented by the idea that it is not only language but also somatic patterns of narrative and the enactment of traditions that are deeply formative of experience, and indeed that all human experience is within those boundaries. The anthropologist of Nepal Robert Desjarlais, for example, describes his own trance states as being parallel to those of his Nepali informants, yet these experiences, his own 'shamanic visions', are regarded by those informants as being 'culturally irrelevant'.' 3 Experience is meaningful only within a cultural narrative and the complexity of experience created within the complexity of the interlocking cultural narratives that we inhabit.
If we understand 'experience' not as a timeless mode outside of language and conception, but as a way of speaking about the narrative of a human life, as Oliver Davies does, 76 then the term has relevance, especially when speaking about the body. There is an argument for the resurrection of experience in a new mode. Indeed, experience in this sense is integral to the body as a way of being in the world, what Csordas has usefully called 'embodiment', a central feature of such embodiment being its indeterminacy." The body is the precondition for experience and at this level functions in a precognitive way. The body as experience, as lived body, is arguably a precognitive condition for all cultural and religious expression. Drawing on the work of Merleau-Ponty, Drew Leder argues that the body is experienced as an absence, the disappearance of the body from our awareness while yet functioning as the precondition for awareness. 78 Such disappearance from awareness of the lived body is linked to the body as representation in that representations of the body into which we are habituated become inseparable from our experience. There is a reciprocation between the body as lived and the body as pre-cognitive experience: the experience of the body is fundamentally constrained by the body as symbolic order, and the symbolic order of the body only comes to life because experienced, and this can be at a deep level in a non-cognitive way. To refer
ahead somewhat, the experience of oneself as being identical with the supreme deity Siva is an internalisation of the Saiva symbolic order such that subjectivity is engulfed or overwhelmed by the experience of Siva. The body as symbolic system for ordering the cosmos becomes an existential location for a subjectivity expanded to a tradition-constrained limit. That is, the practitioner achieves a corporeal understanding of the vibrant goal.
In the context of religion, rather than a pristine 'experience' expressed and approached in different ways, we need a much more nuanced argument in which the body is seen in terms of text and the subjective appropriation of tradition. The narrative of the practitioner's life conforms to the narrative of tradition and the body is encoded in text-specific ways. This encoding, this mapping of tradition on to the body, is also the experience of tradition and the fusion of the lived body with the symbolic order of the tradition. Another way of saying this is that the extra-textual subject, what is called the indexical-I, is filled out with tradition and text-specific content and that this is indeed 'experience'.
In an important book on Tibetan Buddhism, Civilized Shamans, Geoffrey Samuel has argued for a distinction between shamanic and clerical Buddhism, where 'shamanic' refers to 'the regulation and transformation of human life and human society through the use (or purported use) of alternate states of consciousness by means of which specialist practitioners are held to communicate with a mode of reality alternative to, and more fundamental than, the world of everyday experience.' 79 On the one hand we have the practitioner focused on somatic experience in contrast to the monk-scholar concerned with monastic discipline and philosophy. In the context of Hindu Tantrism the shamanic practitioner might be seen in the tantrika cremation ground ascetic seeking ecstatic experience through yogic techniques, ecstatic sexuality and intoxicating substances in contrast to the tantric Brahman temple priest or practitioner still within the sphere of orthoprax injunction. This distinction could be reflected in the distinction between the sadhaka, the practitioner desiring pleasure and power in higher worlds (bubhuksu), and the acarya, the teacher desiring liberation (mumuksu). However, the argument of this book is that both ecstatic and formalised Tantrism must be understood
as the encoding of the body with the text. The tradition forms the body of both 'ecstatic' and 'formal' practitioner and neither idea can be understood outside of a textual revelatory tradition.
In the following pages we will see how the entextualisation of the body operates in the tantric traditions in terms of the identification of embodied self with that assumed in the texts, in terms of reading, and above all in ritual and asceticism. In ritual, tradition and text are mapped on to the body through a series of procedures such that the body becomes divinised. In a parallel way this process occurs in what might be called asceticism, where through ascetic practices the practitioner inhabits worlds given in the texts of tradition. Through ritual and ascetic practices, the tantric adept seeks to expand his subjectivity such that he experiences different worlds within the system until he attains liberation, which is understood as the divinisation of self and body. Implicit here is an understanding of ritual as a form through which culture is replicated, that enacts cultural values, and embodies the memory of tradition. Rituals are systems of signs that establish a continuity of identity and through non-identical repetition. 80 The lived body and the symbolic representation of it merge together. This merging of symbolic representation and lived, experienced body is a corporeal understanding of text. A corporeal understanding of the text is a way of inhabiting the text linked to a 'religious reading' rather than a non-corporeal 'consumerist' reading, to draw on a distinction by Paul Griffiths, 81 although in contrast to Griffiths what constitutes religious reading is not the quality of attention but the indwelling of the subject in the text and the text in the subject. This book clearly does not itself represent a corporeal understanding, but does bear witness to such an understanding in the tantric case and claims that such corporeal understanding is always, inimitably, textual.
The Argument of the Book
The book is divided into two parts. Part I, 'Theory, Text and History', outlines the argument and describes the tantric texts and traditions I shall be concerned with in their historical context. In
Part II, I show by some detailed examples how the body is inscribed and the self mapped by the texts within a diversity of metaphysical viewpoints. Some tantric traditions are dualistic in maintaining an eternal distinction between the true self, a transcendent God, and the world, while others are monistic in maintaining their ultimate identity. While some texts are synthetic in claiming that ritual can be done according to a variety of texts or that rituals from one group can be absorbed by another, as Granoff has shown, 82 other texts are clear that ritual must be performed according to the procedures outlined in a specific scripture. The Rauravagama, for example, explicitly says that rites being performed prescribed by one Tantra should not be mixed with rites from another. Mixing texts in ritual is harmful to the king and kingdom. 83
In spite of this diversity, the desire for traditions to distance themselves from each other and their often rigorous argument, the divinisation of the body is a theme and process shared by different traditions. The body becomes the uniting metaphor of these systems and processes at the level of practice and demonstrates a shared substrate of ritual and cosmology in spite of divergent metaphysical claims. In particular, I would wish to identify two processes or fundamental principles (which are also themselves metaphors) that form the tantric body. The first is a hierarchical and emanationist cosmology in which lower levels emerge from higher: a movement from the refined and pure to the coagulated and impure, from refined matter to physicality. In the second, the body recapitulates this hierarchical cosmos; the body becomes a cosmography, a writing of the cosmos. The structure of the body reflects the structure of the cosmos and is itself thought to be an emanation from a higher level. What follows from these two fundamental principles articulated in our texts is: (a) to achieve salvation is to trace a route back through the cosmos to its divine source or the point at which the disembodied self became entangled with matter, which is also conceptualised as a journey through the body; (b) this pathway back to the source is the mapping of the body in tradition-specific and text-specific ways through ritual and interior practice. This is the entextualisation of the body, which we can also speak of in terms of subjectivity having variable linguistic agency in which the boundaries of the subject of
speech change through the internalisation of text. Thus, for example, the tantric tradition focused on the god Visnu (the Pancaratra) envisions the universe in terms of three broad categories: the pure creation, the mixed creation and the impure creation. The mixed is an emanation from the pure, and the impure an emanation from the mixed through God's power or energy. 'Pure' means devoid of physicality and 'pollution', which are features of impure creation. The goal of life, on this account, is to progress through the levels of the cosmos from the impure to the pure, a journey which is reflected in the body; the body becomes an image or icon 84 of the universe and the structure of one is recapitulated in the other. Much of the present book will be an illustration of this fundamental concept.
The consequences of this argument in terms of the history of ideas are first that the tantric body entails an emanationist cosmology which is implicitly (and sometimes explicitly) pluralistic. Rather than monism, often associated with the Indian religion, the tantric traditions inherit the ancient systems of speculative thinking that we can refer to as Samkhya. Second, developed metaphysical interpretations of an idealistic nature about the unity of consciousness are a later, secondary overlay on to the basic ritual and cosmological structure, evidence that supports Sanderson's view about these texts. Indeed, metaphysical speculation sits on top, as it were, of a ritual substrate and we have what Beyer has called, in a Buddhist context, the 'ritualisation of metaphysics'. 8 ' Third, tantric traditions must be seen not as being generated out of a non-dual, spontaneous religious experience which takes on different cultural and doctrinal forms, but as developments of ancient traditions of speculation and practice closely linked to Brahmanical imagination, vedic practice and institutions. Along with ritual, the tantric imaginaire is less concerned with the usual activities of Indian philosophical discourse, namely epistemology and logic, and more concerned with a poetics of imagination and aesthetics articulated in texts and commented upon by a second-order discourse within the tradition. There is thus a devotional or bhakti dimension to many tantric textual traditions. 86 Fourth, the politicising of the tantric body means that traditional power relationships are maintained in the wider social body. The tantric body is a pre-modern, 'conservative' body that conforms to
the structure of tradition and confirms the matrix of social power even in its ritualised flaunting of it.
In the following pages I intend to illustrate and develop the argument of the textual inscribing of the body and its linguistic agency. The tantric body cannot be understood without some account of the vedic body, and the next chapter gives an account of different historical discourses, namely legal, political, and philosophical, that have contributed to formation of the tantric body either positively by being appropriated or negatively by being rejected. Part II develops the argument of body as text with detailed examples from the Pancaratra and Saiva traditions. Here we shall include an account of the breaking of vedic prohibitions in caste-free sexual ritual of ecstatic Tantra intended to realise the goal of uprooting desire. We examine in more general terms the tantric 'imagination', showing how the body becoming divine is the central trope of Tantrism: entextualisation is a topos operative from the king to the village washerwoman. An appendix offers a translation of the divinisation of the body through nyasa from the jfayakhya-samhita, which has provided much illustration of the tantric body in this book.
TWO
The Vedic Body
An important characteristic of scriptural religious traditions is the ability to recognise in the past what could be and what could not be a guide for the future and the ability to identify resources in the past necessary for the construction of the future. 1 Although 'new' in the early centuries of the common era, the tantric traditions nevertheless had a sense of themselves as having a continuity with the past, of being traditions, a sense of receiving something handed down and passed on. Indeed, this heritage is of central importance in the formation of the tantric traditions, which can only be understood in relation to it. The tantric traditions are the inheritors of systems of thought and techniques of the body whose origins lie in the ancient past and which had achieved a high degree of reflexive awareness at the time of the emergence of the tantric systems. In order to understand the tantric body we must offer some account of what we might call the vedic body. These abstractions, the 'vedic body' and the 'tantric body', are intended simply to be a condensed shorthand for representations and techniques of the body in what might broadly be called the vedic and tantric traditions. Both bodies function as symbolic systems and metaphors through which the social world and wider universe are conceptualised; both bodies are the product of deeper cultural forces and structures of power; and both bodies were also the lived bodies of practitioners in the traditions, the existential
modes by which human beings inhabited their world. The vedic body is a vast topic in itself, but for our purposes we need to focus on important dimensions of Brahmanical discourse that affected the practices of being a Brahman, that affected the techniques of the vedic, Brahmanical body. Brahmanical representations of the body are closely related to different realms of value in the Brahmanical universe and different conceptions of the good for an individual and community. These values are articulated in different genres of text and we shall here focus on legal discourse, political discourse, and a philosophical discourse concerned with the highest good. All of these impact upon the tantric conception of the body and practices either through being absorbed by the tantric traditions or through their rejection. Apart from legal, political and philosophical discourse about the body, there are also two further areas of discussion and practice that have a direct bearing, namely medicine and an erotics that interfaces with aesthetics. But first we need to offer a brief description of the political and social context within which vedic and later tantric discourse emerged.
The Political and Social Context
As Sanderson has observed, by the early medieval period Brahmanical traditions of thinking and practice (and such systems were only Brahmanical) were mutually aware of each other 2 and defined their boundaries in response to each other's philosophical positions, often arranging these views in a graded hierarchy. Some schools accepted the Veda as revelation, regarding it to be the source of their tradition; others rejected them. In philosophy we see the development of exegesis with the Mimamsa tradition, various forms of monism in Vedanta, particularly the Advaita tradition developing from Samkara (788-820 Ce), to later Vaisnava forms, the dualism of Samkhya, the heterodox systems of the Buddhists and Jains, along with the development of the tantric traditions. The philosophical positions of many schools also express traditions of practice and the doctrines of wider communities which arose within particular social and political contexts.
The problem with the history of India is that it has so often been difficult to date texts and to place philosophical texts in a social history, but this becomes easier once we enter the first millennium ce. While the early medieval period saw the mutual clarification of philosophical positions, it also witnessed much political turbulence. The comparative political stability of the Gupta and Vakataka empires (c. 320-550 ce) was replaced by a period of some complexity, with different kingdoms and tribal lords coming to political dominance and then passing away. In historically locating the social history of tantric Buddhism, Davidson has laid out the political developments from around 500 to 1200 ce in a meticulous and clear way, making the point that while this period has been neglected by historians - often because the post-Gupta period was associated with decline and decay - the empires of the Gurjara-Pratlharas (c. 725-1018 ce) and the Palas (c. 750-1170) lasted longer than the Guptas. 1 With the destruction of the Guptas and Vakatakas there is increasing decentralisation, with an emphasis on the region and a rise in the status and concept of kingship. Echoing the ideal of a previous age, the early medieval period witnessed the rise of the ideology of the 'universal ruler' (cakravdrtin) and the strengthening of the court as the locus of cultural activity, such as the development of schools of Sanskrit poetry and drama. Alongside the development of the region, Davidson shows how the king becomes divinised and in the new feudal kingdoms divine; royal power is expressed in the regional temples, which 'became showpieces of royal self-representation'. 4
These kingdoms formed a complex network, which Inden has called an 'imperial formation' and Stein has called a 'segmentary state', in which a ritual hegemony operated where a lesser king or tribal lord would pay ritual obeisance and taxes to a more powerful king, as in the case of the Cola state. 5 This model would seem to have been operative at least up to the period of the Vijayanagara empire (1336-sixteenth century) in which the king would on the occasion of the Navaratri Festival receive ritual obeisance, deriving his power from the Goddess herself. 6 It would seem that the model of kingship promoted in Kautilya's Arthasastra, which was composed some time during the first few centuries of the common era, had some currency and reflected the practices of belligerent kings who
waged war on their neighbours while attempting alliances beyond them on the principle that one's enemy's enemy is one's ally.
During this period, different religious groups fared differently at different times. The Buddhists were successful in India up to a point, with large, elaborate monastic institutions such as Nalanda becoming wealthy and attracting royal patronage. The Saiva Pasupata tradition, a renunciate order that rivalled the Buddhists, seems to have become highly successful, attracting royal patronage, as Davidson documents, and becoming associated with royal temple construction in the seventh to tenth centuries.' The Pasupatas were in control of the famous and wealthy temple at Somanath, for example, before its ransacking by Mahmud of Ghazni in 1026. 8 Indeed, Mahmud of Ghazni had previously plundered the Kashmir valley in 1014 CE, presaging a destruction of 'Hindu', especially tantric, discourse in the years that followed with the advent of Muslim rule. As Dyczkowski observes, the consolidation of Muslim rule in north India witnesses, and is partly responsible for, the disappearance of tantric traditions. Agamic Saivism retreated to the south, where it survives in Tamil Nadu, and a tantric tradition also continues in Kerala. Similarly the tantric cult of the Goddess Kubjika retreated and was given royal protection in Nepal. 9 As a result of these historical developments, namely the Muslim conquest, there are very few tantric manuscript sources from northern India, outside of Nepal. Indeed, the tantric tradition in the north more or less completely disappeared, although after the Saiva Tantras or agamas there was a second, later development focused on the Goddess or Sakti rather than Siva, which became especially important in Bengal.
It is not unreasonable to suppose that the political structure that developed had some impact on the conceptual schemes, images and practices of different traditions. As the kings had become gods, the gods became like kings and the royal court became the model for the gods' court in the heavens. Cosmology, so central in the religions of India as a hierarchy of worlds, comes to resemble the social and political hierarchies of the wider social order, and those hierarchies come to resemble the Hindu cosmos: a social order in which everything had its place with a high degree of deference, and which was believed to reflect the natural order. But while the
religious traditions developed in this highly politicised context, and it is important to locate traditions within social and political history where possible, this alone is not enough to explain or understand them. Tradition cannot be reduced to its political environment, and the meaning and significance of a textual history cannot be explicated in terms of social and political history, for the meanings of texts with semiotic density exceed social and cultural particularities and are reconstituted in traditions re-imagined throughout history. The question is open concerning the extent to which the political conditions that favoured Saivism in Kashmir in the late tenth century, royal patronage being a key factor in its dissemination, impacted upon the forms of interiority promoted by the tantras. Indeed, traditions of textual transmission and commentary are fairly oblivious to external political forces, as Halbfass has observed with regard to Brahmanical representation of the 'foreigner', 10 and traditions have often shown remarkable resilience to erosion by external, political forces. The famous Saiva philosopher Abhinavagupta had royal patronage and his non-dual doctrine was highly influential in courtly circles, but one suspects that part of this success was the appeal of the tradition itself and the forms of inwardness is promoted. Abhinavagupta's Tantraloka reflects a rich tradition - or range of traditions - that became successful not simply due to reasons of political patronage, but because the content of the teachings had resonance among an educated Brahmanical elite. Scripture and tradition have an internal coherence or structure of meaning that operates with varying degrees of success outside of particular political and historical circumstances: a coherence which itself partly accounts for the survival potential of any one tradition.
In studying the history of ideas in India we are mainly studying the self-representations of the educated, Brahmanical class who composed the treatises and guarded the transmission of tradition through the generations. Within the vedic tradition the Brahmans were concerned with establishing and maintaining their position as the upholders of moral virtue and social order, namely dharma. Taking our cue once again from Maclntyre (although in a very different context) we might claim that scriptural traditions focus on three areas. 11 First, there is a concern with the body as a marker
of personal identity. Throughout the history of Indian society (and arguably of all human societies) the body is a sign of social location. The subject of first-person predicates not only has but is a body for the traditional Brahmanical community within the sphere of vedic teaching; a body that marks a person as belonging to a particular endogamous social grouping or caste, the property of the body one is born with. 12 While some tantric traditions believed that initiation could eliminate caste, the body nevertheless remains an index of social identity through the marks of one's cult, one's gender and one's practices. Second, there is a concern with agency. Part of belonging to a community and tradition is being able to account for one's thoughts and behaviour to others. Although this is complicated by questions of reincarnation and karma for the broad vedic tradition (which does not unequivocally accept the doctrine of samsara), people have moral and legal responsibility to uphold the values of the social order. For Brahmans this meant above all upholding the rules of ritual purity, but it also meant legal obligations on all strata of society, including the king. Third, the life of the body/self must be seen in terms of a quest.
That life is limited by birth and death is clearly recognised in the vedic tradition with its emphases on the construction of a person through rites of passage (samskaras) and in the sense in the philosophical traditions that there is a continuity beyond life and that what preceded a particular life has a bearing and constraining influence upon it. In Maclntyre's phrasing, 'my life has the continuity and unity of a quest, a quest whose object is to discover that truth about my life as a whole which is an indispensable part of the good of that life.' 13 Although Maclntyre is writing about a very different tradition, his statement holds true for the vedic tradition. Indeed, the Brahmanical tradition thematised the narrative dimension of life and claimed that three and later four goods were crucial to it, namely the values of social responsibility (dharma); material, political and symbolic prosperity or success (artha); and pleasure (kama) within the boundaries of social responsibility. Later the fourth goal of salvation or liberation (moksa) as an ultimate goal or good for a life was added to the list. 14 Of course, the goods of a life in its narrative course are inseparable from the personal identity of the body
and the agency assumed to achieve those goals. The Brahmanical discourses and prescriptions for social-identity-forming practices, for technologies of the body, can be roughly mapped on to these goals. By the early medieval period, rich textual traditions had developed loosely connected to the goals of dharma, artha and kama, namely legal material (dharma-sdstra), political discourse (artha-sdstra) and erotics {kdma-sastro). All these discourses have something to say about the vedic body, although not necessarily in agreement, and the tantric body must be seen in the light of these formations. The tantric traditions are informed by Brahmanical discourse, not least in their rejection of it. The Tantras and their concerns can in many ways be understood as a response to Brahmanical 'legalism' and the sexualised ritual of some tantric traditions as being quite distinct from the erotic discourse of the kdma-sdstra. The tantric traditions, as we shall see, accept the narrative of life as a journey but reinterpret or even reject the vedic configuring of this journey. They often reject that the goal and ultimate good must be determined within the boundaries of vedic social values and break the link between the highest good for a life and an identity determined by brahmanical discourse and power. Rather than a person's highest good being found within the vedic tradition, on the contrary it must be located outside of that tradition in sets of values that are supplementary to the vedic, or, in more extreme traditions, reverse them. Indeed, many tantric representations of the body serve to disrupt that sense of vedic identity, as we shall see. To gain some leverage on representations of the vedic body linked to the scheme of values, we need to examine legal discourse, political discourse, and a philosophical discourse about the self.
Legal Discourse
Brahmanical understandings of bodily identity, agency and goal are articulated in legal texts and commentary upon them. The legal treatises technically known as smrti, 'remembered tradition' that can be responsibly rendered as 'secondary revelation'. The earliest is the famous Laws of Manu (Manusmrti) composed some time
between the second century bce and the second century ce; and the Yajnavalkyasmrti composed probably during the Gupta era are the most important texts in the sense that they both have 'a stream of commentators' 15 and formed the basis of jurisprudence in later colonial India, although they go beyond simply legal concerns. 16 In some ways they might be seen as the very opposite of the Tantras, although a later text, the Mahanirvana-tantra contains legal material derived from British law, making the text a 'juridical hoax' composed no earlier than the mid eighteenth century. 17
Dharmasastra comprises texts that are legal treatises in a very wide sense, for they include material on daily purification practices, rites of passage, atonement for omitted rites and so on. Mainstream tantric texts of the Pancaratra and Saiva Siddhanta maintain a close proximity to the vedic tradition and prescribe a whole way of life that incorporates vedic rites of passage (birth, vedic initiation, marriage and death) along with the supererogatory, tantric rites of their tradition. They supplement vedic ideals with their own accounts of the highest good for a life and while they claim to supersede vedic views, they are nevertheless influenced by the dharmasastra, not only in their incorporation of the general pattern of ritual life but also in the use of terminology. As observed by Buhnemann, for example, in relation to the Kularnava-tantra, impurities that arise at the beginning and end of mantra recitation need to be removed. In analogy to the dharmasastra the text refers to these impurities as jatakasutaka and mrtakasutaka, pollutions that need to be purified in connection with birth and death. 18 The Jayakhyasamhita also refers to this purification in relation to expiation for omitted rites (prayascitta). 19
The general view of the body promoted in dharmic literature is ambivalent. On the one hand great care is taken over the body, a guarding and control of the body's functions in accordance with highest moral duty (parama dharma) for a life; on the other the body is the location of the passions and is inherently impure through its desires, instincts and effluvia. Not only this, in most philosophical systems, which are generally addressed to male adherents, the body's sexuality is itself a distraction from the path of liberating knowledge. As Doniger discusses, the understanding of the body as impure,
along with a distrust of desire, is linked to a radical misogyny in ancient Brahmanical culture and male anxiety in the face of the female body and sexuality. 20 This anti-body rhetoric generally takes the form of listing body parts and functions and drawing the reader's attention to each with a view to highlighting a repulsion generated in this kind of objectification. Roberts insightfully observes that this 'semiotic deconstruction of the body and its organs is the price paid for the tolerable cultural management of sexuality' 21 and it is clearly the case that negative representations of the body are linked to negative views of sexuality and often to a misogyny that identifies women with the body. An example of the objectification of the body in Manu, discussed by Doniger, is as follows:
[A man] should abandon this foul-smelling, tormented, impermanent dwelling-place of living beings, filled with urine and excrement, pervaded by old age and sorrow, infested by illness, and polluted by passion, with bones for beams, sinews for cords, flesh and blood for plaster, and skin for the roof. 22
This passage occurs in the context of a discussion about the circle of reincarnation (samsara), which one who does not have a vision of the supreme self or absolute {paramatma-darsana) will re-enter again and again.
Yet while there are undoubtedly passages such as this that at the plain sense level of the text present an extremely negative attitude towards the body, this cannot be taken as a sign tout court of Brahmanical attitudes. The picture is more nuanced and complex. According to the commentator Bharuci, we must understand the passage in the context of a discussion about the dharma of the renunciate {pravrajita) or ascetic (tdpasa) whose meditation practice gradually allows a detachment (vairagya) from the body to this higher vision. 23 This negative representation of the body occurs in the context, according to Bharuci, of the particular good or value (visesa dharma) of the renunciate who seeks to transcend the social order in the stages on life's way beyond that of student and householder. The renouncer who seeks liberation has gone beyond the world of social transaction and legal responsibility 24 and seeks to go beyond the body in a 'spiritual' (atmaka) liberation. Although renunciation is
excluded from the householder, it is still within the overarching, total scheme of the orthodox, Brahmanical world-view, and in a sense is included by its exclusion. Even the rejection of householder values is incorporated into Brahmanical representations.
While the body of the renunciate is seen by Manu in the negative terms described above, the body of the student and householder is represented not in such stark terms, but in terms of the need for its control and purification. The householder and student operate by a different set of values to the renouncer, those of moral and legal responsibility to the wider social body, which are different in not displaying disgust for particular body parts or functions, but rather displaying a need to control the body through rigorous purification. 2 ' The body is the vehicle for a successful life, but only through its strict control and avoidance of impurity and spontaneous desire. Some of the rhetoric in Manu concerning the restraint of the senses and body is derived from the general yogic discourse that control of the senses leads to a higher knowledge. For example:
A wise man should strive to restrain his organs which run wild among alluring sensual objects, like a charioteer restrains his horses. (Manu 2.88)
Desire is never extinguished by the enjoyment of desired objects; it only grows stronger like a fire [fed] with ghee. (94)
But when one among all the organs slips away [from control] thereby wisdom slips away from him, like the water flows through the one foot of a [water carrier's] skin. (99)
For the dharmasastra the body is not only subjected to rules of ritual purity, but is the subject of legislation; an index of the whole society that reflects social stability and the need to maintain caste boundaries, thereby maintaining power relationships within the community. The vedic body is a controlled body, a control that seeks to keep the body under the sign of Brahmanical authority in formulating the limits of legal transactions, such as inheritance law, and in seeking to control actions from rising in the morning to elaborate rules for cleansing procedures around bodily processes. 26
A large part of this process is the control of women's bodies in legal procedures and in discourse. Although the Hindu legal treatises
were probably the first in human history to recognise women's property rights, by twenty-first-century standards they are inevitably open to critique. Generally women are subject to male authority throughout dharmic literature. For example, there is a debate in the dharmasastra about whether a widow should inherit her husband's property, some texts saying that she should inherit it totally, as the wife is half of her husband's body, and so as long as half of his body lives, how could anyone else gain his property? Another set of texts, however, supports the view that a man's property should go to his male relatives. 27 The eleventh-century Jlmutavahana suggests a compromise, arguing that a widow should inherit if there are no sons, although not be able to dispense with the property. 28 Other examples could be cited to illustrate the general Brahmanical idea that women are subject to male authority, to father as a daughter, to husband as a wife, and to son as a mother. 29 Indeed, according to Manu woman is the field (ksetra) in which the man sows his seed to produce (ideally male) offspring. 10
The vedic body is thus inscribed with vedic values through the ritual processes of rites of passage through which it is constructed (samskdra means 'put together'), controlled through rules of ritual purity, and controlled through legal procedures. Both men and women are subject to these controls in ways which go against contemporary Western values, but which were also challenged at the time of their predominance by both renunciate traditions and by the tantric traditions. The Brahmanical control of the body was rejected in many cases by the Tantras and their followers, sometimes in a mild way through their subversion by overwriting the vedic body with tantric rites, sometimes in an overt way by its complete transgression in ecstatic bodily experience. While the discourse of women's bodies remains ambiguous in the tantric corpus (all texts so far as we know were written by men), there is often an explicit rejection of the Brahmanical control of the body and a reconstruction of it in other ways, even in tantric traditions such as the Saiva Siddhanta that align themselves with the vedic tradition. The body is not simply subject to control by purity laws and is not only the object of legal transactions to maintain the social order, but rather the traditions of the 'left' contain the potential for extreme, ecstatic, experience that
shatters vedic, conformist structures. But the spontaneous rupture of the vedic body in any ecstatic Tantra is a spontaneity nurtured and facilitated only within the specificity of tradition (see pp. 166-9).
Political Discourse
Closely linked to legal (and moral) discourse is a political discourse about the state and the nature of kingship, the raja-dharma. Although integral to the dharmasastra itself (Manu, for example, contains important sections on it) raja-dharma came to be treated in independent treatises, 31 the most famous of which is Kautilya's Arthasastra, the 'science of government' (first-second century ce) concerned with the two aspects of raja-dharma: the development of prosperity (artha) defined as education and riches 12 and government defined as punishment of offenders (dan da) or, more broadly, the exercise of law. Kautilya's work is a theoretical discourse, deeply concerned with the maintenance of power within the segmentary state and the control of populations, not simply as a consequence of brute political force, but because the control of the people by the king is integral to the order of the cosmos, to dharma. Property rights are ruthlessly maintained, including rights over women, through the punishment and torture of thieves, and adulterous liaisons across caste are punished by disfigurement or even death. 33 Kautilya is keen to point out the powers of the king to disfigure, maim and execute for the maintenance of the social body, the upkeep of the segmented hierarchy of the medieval Indian kingdom.
As in medieval Europe, we have in the Indian material a link between the state or the body of the kingdom and the body of the king. 34 According to Kautilya and others, the state (rajya) is made up of seven elements (saptanga): the ruler or sovereign, the minister, the territory of the state itself (rdstra), the fortified capital, wealth in the treasury, the army and friends. 35 These are called constituents (prakrti) or limbs (anga), with the implication that they are the limbs of the social body. While there are very few textual references that directly compare the state to a body, one or two make this explicit connection. Jivananda's Sukranitisara, a digest on governance,
compares the state and specifically these seven constituents to the organs of the body: the king is the head, the ministers are the eyes, the ally is the ear, the treasury the mouth, the army the mind, the capital the hands, and the territory the feet. 36 This idea of society as a body, and by extension the kingdom, is quite ancient in India and is common in modern popular discourse." The Rg-veda contains a famous hymn to the cosmic man from whose sacrifice the cosmos is formed, including the social order, with the Brahmans coming from his mouth as the voice of society, the nobles from his arms as the strength of society, the commoners from his thighs as the support, and the serfs from his feet. 38 For the body to function all elements must work together in harmony, although according to Manu each one is superior to the preceding. Manu compares these limbs to the senses (indriya) restricted to their own domains (visaya), 39 thereby highlighting the conception of the state as a body. Manu's commentator Bharuci observes that a vice (vyasana) in any of this group is likely to destroy the policies of the kingdom, so the king's function is to maintain the health of the social body through the exercise of power in accordance with dharma.
The social body is identified with the body of the king. Kautilya says that the sum total of the constituents of the kingdom is the king and that which he governs. 40 Indeed, because of this link, the moral virtues of the king have a direct impact upon the kingdom and there is a correspondence between the body of the king and his kingdom. Through controlling his senses and behaving like a kingly sage (rajarsi) by eliminating the vices of lust, anger, greed, pride, arrogance and excitability, the king will succeed in a long and prosperous rule. 41 The body of the king reflects the body of the kingdom and vice versa. Furthermore, the king is identified with a deity or deities. Manu, for example, says that the king comprises fragments of the gods 42 and so there is a correspondence between the bodies of the deities, the king and the kingdom. Given this intimate connection, it is no wonder that some thinkers in medieval India, notably Jayantha Bhatta, thought that royal interest in extreme tantric practices would have a detrimental effect on the kingdom. If the king goes against dharma, defined in terms of orthodox, Brahmanical practice, then all the people will suffer because of the connection
between the two bodies, although, in spite of the identification of the king with the divinity, the law books do advocate the forcible removal of a bad king 'like a dog afflicted with madness'. 43
We have seen in the dharmasastra that the ambivalence towards the body lies not in its rejection, but in the need for the body to be controlled and restricted within the value system of dharma. Even, or perhaps especially, the king's body was not exempt. The body is good in so far as it is a means of purifying the self and keeping the dictates of tradition and probity, but bad in that if left uncontrolled it will turn towards vice and the kingdom will suffer. All bodies are interconnected in this world and the higher the status of the body the wider the consequences of action. Marriott is surely correct here in emphasising the transactional nature of personhood. 44 An outcaste (candala) living beyond the cremation ground with 'heretics' (pasanda)^ is far less consequential than the high-caste member of the social body. In one sense, the higher the degree of ritual purity to be maintained, the more the social restriction, and the more damage done to the social body in transgression.
This presentation of the body and its function within the wider culture assumes the validity of the distinction, highlighted by Dumont, between purity and impurity, qualities, and indeed values, reflected in the ritual construction of the body and its gendered role. It is, of course, very difficult to access the social reality of ancient and medieval India other than in its representations, often ideal, such as Kautilya's text. Although Dumont has been criticised, 46 that the purity-impurity (sauca-asauca) distinction is historically valid would seem to be the case from explicit textual references concerning it. While the whole complex web of Indian social history cannot be reduced to this basic division, which itself must be seen in the context of power and social classes vying for position, it is nevertheless of fundamental importance in understanding the vedic body and, as we will see, the tantric body. Other cultural dynamics have been identified in the social field, especially the auspicious and inauspicious by Marglin 47 and the importance and loaded nature of prestations by Raheja, 48 a discussion of which would take us too far from our topic. But it is important to remember that in dealing with the textual history of ideas we are dealing with representations and
the ways in which different groups, mostly of Brahmans, wished to present themselves to their community of readers. An important representation was of the body controlled by purity and impurity and the social body that reflected this distinction. In medieval India Brahmanical men and women were severely constrained by the en-dogamous group they were born into. One way to escape some of these constraints, and to take on new constraints, was through the institution of renunciation, the formal seeking of the highest good, the goal of liberation from the body and social world to a goal defined in various ways by different traditions of renunciation.
The Highest Good
While the rather artificial scheme of the human goals, the purusarthas, has the disadvantage of the oversimplification of competing values available within the social body, it is significant for the very fact of attempting, fairly successfully, to integrate them into a coherent scheme. The world-affirming values of social responsibility, success and pleasure have sometimes been contrasted with the world-denying value of liberation from the world. That these two realms of value exist and are held together does not reflect a contradiction but does reflect a tension in the history of Hindu traditions that is a characteristic of them. It not simply a matter of history that a dominant social group, the Brahmans, that maintained one group of values came to integrate another, contradictory, value. While there is evidence for this in the sense that the three goods of dharma, artha and kama as a coherent group are earlier than the four which adds moksa, the tension between the positive affirmation of social values that emphasises duty, success and pleasure, along with their negation in renunciation, has been there from extremely early on in the tradition. The Upanisads, which reflect this tension, are certainly being composed by 800 bce. 49 We must resist any oversimplification of contrasting a world-affirming arena of vedic values with a world-negating arena of non-vedic values, in favour of a more complex picture of historical development in which the tradition draws life from the tension. On the one hand there are claims that what is
most important in the world is power or pleasure, while on the other there are claims that liberation transcends all worldly values. 50 The tension is seen in Manu, which advocates the importance of dharma and Brahmans fulfilling their social obligations, yet also looks to the transcendent goal of liberation.
The tension between competing goods in the Brahmanical tradition is partially resolved through the institution of the 'stages of life' (asrama) in which the householder can pursue the goods of social obligation, success and pleasure, leaving the world-transcending liberation to the renouncer. This is clearly an affective strategy within the tradition, but one that is not wholly satisfactory to many within it, and some texts, rather than encourage a disjunction between competing goods, try to integrate them. The famous Bhagavad-gitd is an example of this. Here the god Krsna advocates the necessity of doing one's social and moral duty, yet at the same time claims that there can be liberation from the world of action through acting with detachment from its fruits (asakta karma)." The goods of worldly morality and a world-shattering transcendence are placed side by side, and the human condition exemplified by Arjuna is to struggle with the tension.
The vedic body, then, is inscribed by a number of discourses and traditions that the tantric traditions respond to. First, we have the Brahmanical writing of the body in accordance with the highest social good of correct action in accordance with scripture. This is a tradition of ritual that maintains the integrity of the body and the clear differentiation of social and gendered roles that provides the basis for all further speculation. Accompanying this level of ritual action fundamental to the culture, we have a discourse about the nature of ritual action as enjoined by scripture, namely the Mimamsa, which furthermore directly feeds in to a discourse about law, kingship, and the nature of society as a whole. Second, we have at the level of discourse a dualist metaphysics in the Samkhya tradition, which is more concerned with what it sees as the highest value of liberation from the world. Third, we have a monistic metaphysics in the Advaita Vedanta that sees the highest goal as realisation of the self's identity with a featureless, unbounded absolute reality. By the time of the early medieval period and the rise of the tantric traditions,
the picture is more complex, with theistic traditions developing discourses of transcendence, some of which attempt to integrate this with the culture of Brahmanical ritualism. The tantric traditions emerge at a time when the cultural baseline of Brahmanical orthopraxy, with its adherence to the values of caste and stage of life (varn asrama-dharma), were strong yet becoming overlaid with theistic systems of ritual and devotion (to Visnu and Siva). These systems, along with competing discourses about the highest good, are reflected in the tantric traditions and the tantric transformation of the Brahmanical patterns.
THREE
The Tantric Revelation
Writing from his prison cell in Kashmir some time during the closing years of the ninth century, the Nyaya philosopher Jayantha Bhatta defended the authenticity of tantric revelation, but within the boundaries of vedic reason. If the Tantras offer teachings that are acceptable to learned people and if they do not go against dharma, then he can see no reason why they should not be adopted. However, if they proffer immoral teachings then the king should certainly prohibit their continuance. This was indeed the case with the sect of the blue clad (nildmbara), who practised on festival occasions, says Jayantha, unconstrained group sex in public places, simply covered with a blue garment! 1 For Jayantha such behaviour was against the public good and against the scriptures. While Jayantha locates himself within the vedic tradition and espouses its values, he is living in a time when the mainstream, orthodox and orthoprax tradition is being challenged by unorthodox forms of practice and texts that claim to be from a divine source. Jayantha is clearly an intelligent and humorous man, deeply concerned about social values and the possible threat to those values caused by new ideas. He wrote his famous text of philosophy 'The Bouquet of Logic' (Nyayamanjari) in prison to keep himself amused (truly an Indian Boethius!); in it he defends orthodox revelation, the Veda, but is nevertheless open to the possibility of new revelation and
is a realist in understanding that his community needed to adapt to the new challenge. But when that challenge threatened what he saw as the fundamental values of his society, then he strongly defended the old morality. Indeed, after his release from prison he wrote a comic play, the Agamadambara, which Sanderson renders as 'Much Ado About Scripture', highly critical of extreme tantric ascetics in his country 2
Jayantha's writing shows a tension in early medieval Kashmir between Brahmans who regarded the Veda as revelation that should provide and govern values and others who were offering different ways of life and thinking, such as the Buddhists, Jains and those who were propagating different kinds of writing as revealed knowledge, such as the tantrikas. Before proceeding to a fuller account of the body as text in tantric traditions we need some discussion of what the tantric tradition understood by 'scripture' or 'revelation' (tantra, agamd) and how scripture related to other traditions of the time. It is highly significant that tantric traditions are scriptural. Like other Indian religions, they take their doctrine and ritual from scripture and formulate their goals wholly in conformity with the text. If the vedic revelation provides, in Oberhammer's terms, the authority for a tradition passed down the generations (Uberlieferungsautoritcit), 3 then so too do the Tantras. This is often overlooked or underestimated, for to see tantric traditions as scriptural is to emphasise their traditional and conservative nature, even when they fly in the face of orthodox vedic values. Tantric practices are always textually substantiated and the origin of those texts claimed to be beyond the world in a transcendent source. The Tantras of all traditions locate their origin from the mouth of their God (or the Buddha or Mahavlra) and claim that through a process of dilution, simplification and shortening, they have come to the human world via intermediaries, usually sages who have often undergone great penance to gain the scripture. Their purpose is guidance, liberation and pleasure or power for those lost in the ocean of birth, death and rebirth. 4
We need to understand the Tantras in the context of scripture in India. First, text is inseparable from tradition and formed within tradition, although a text can have such consequences as to change tradition completely (as in the case of the New Testament
in Christianity). Second, the tantric traditions are regarded as a revelation from a transcendent source and the texts describe the 'descent of the Tantra' (tantravataraf from a pure, divine origin but becoming eroded in the course of its descent to the human world, where it is sometimes presented as a particular (visesa) or esoteric revelation for the few with the qualification (adhikdra) to receive it, in contrast to the exoteric, vedic scriptures. Third, the Tantras need to be seen in what Inden, following Collingwood, has called a 'scale of texts' in which a text is positioned in relation to others usually in a hierarchy such that '[t]exts at each level in the scale supplement and comment on the levels below.' 6 This is clearly the case with the Tantras, which present themselves in a scale of revelation, relegating other traditions to lower levels of this revelation and reading the earlier traditions through the lens of their own revelation. There is a high degree of intentionality in the scale of Tantras such that if a text does not deal with the details of a particular topic, it is assumed that this is covered elsewhere. Finally, following Inden, we need to understand the anonymous Tantras (and some related texts with named authors) as having a composite authorship,' and so when speaking about the intentionality of a text or 'author' of a text we are not speaking in terms of authorial intention in the usual sense. Thus an account of scripture in Tantrism needs to be placed in an account of the vedic understanding of the scripture and revelation that were current at the time of the rise of the Tantras. There is no space to develop this here, but we can say that according to vedic exegesis, the Mrmamsa, revelation is a system of signs that points to a transcendent meaning. This revelation has no author, and so that transcendent meaning must be understood in terms of its inner intentionality and is therefore self-validating. Nyaya, by contrast, refuted the atheism of Mimamsa and proposed God as the author of the Veda. The Tantras are closer to the Nyaya perspective and are interestingly defended by the Nyaya philosopher Jayantha Bhatta. 8
The Validity of Tantric Revelation
Rigorously defending the Veda against sceptical and Mimamsaka critics, Jayantha offers proofs that the author of the Veda is God
on the grounds that Prajapati, the Lord of creatures, says that he is the author, the Veda is composite like other objects in the world such as cloth, and so, like cloth, must have a maker, and the validity of the Veda is furthermore ensured by their being spoken by trustworthy people. 9 In a parallel way Jayantha defends the authenticity and authority of the tantric revelation. As a theist he accepts the possibility of further revelation from a divine source and as a philosopher maintains criteria for their acceptance or rejection, namely their accordance with received, orthodox scriptural tradition and their wider acceptance by knowledgeable persons. For him scriptural revelation is not a closed canon. There are five criteria of authenticity that Jayantha uses: they must have been accepted by an assemblage of great persons (mahajanasamuhe), by a large number of learned persons (sista); they should not appear unprecedented (napurva ... bhanti) even if only recently composed; they should not be motivated by greed; and they should not cause people agitation (nodvijate). 10 The Saiva Tantras (he uses the term agamas) meet these conditions in that they do not contradict the truths offered in the Veda, being pervaded by Upanisadic teachings about liberation, and do not go against the caste system. Indeed, they only add new rituals. Even the Pancaratra revelation is authentic in Jayantha's eyes, authored by Lord Visnu, who is God, the creator, preserver and destroyer of the world. He cannot imagine that the Saiva and Pancaratra Tantras are composed from motives of greed or delusion, although this is not the case with Buddhist scriptures since they are not affiliated to the Veda and advocate the abandoning of traditional values and the institution of caste. Furthermore, the Buddhists are morally inferior, being indifferent to the world and addicted to animal slaughter. 11 It is not precisely clear which 'wicked Buddhists' Jayantha means, although he is referring to tantric Buddhism and perhaps the more extreme antinomian practices that go against caste and Brahmanical mores, taught in the Yogini Tantras.
Nor is Jayantha completely convinced about the authenticity of all the Saiva Tantras. In very humorous vein in the Agamadambara, Jayantha raises his doubts about the legitimacy of certain groups. In Act 2 of the play, the central character, a Brahman, is astonished (vismaya) to witness a man and woman entwined together in a single
blue garment {nilambara), exclaiming 'Oh alas, such asceticism [I have never seen] before!' 12 They are singing very tenderly (atipesala) but many more come into view, singing songs of their sect (carcari) in the vulgar tongue, drinking spirituous liquor, and behaving in a very excited and dissolute way (ativipluta), their observance (yrata) involving sex that disrupts correct, vedic behaviour with regard to caste and stage of life." The Brahman Sarikarsana observes that this nilambara 'asceticism' is a new practice (nutanamady apravrtta) that is a form of the great vow (mahavrata) that the Lakula Pasupata ascetics followed. He is fearful of pollution and so shocked by such extraordinary ascetic behaviour (tapakaryakaryam) that he resolves to tell the King and to ensure that such people are banished from the land. Jayantha tells us in the play, and in the Nyayamanjari, that King Sarikaravarman (883-902 ce) does indeed ban the nilambara sect. 14 Jayantha is also sceptical of the Kapalikas who beg from a cranial begging bowl and who appear in the Agamadambara as two cremation ground ascetics fearful for their future having heard how the king is cracking down on such sects. 11 Other ascetics are also fearful of the king's wrath, but our hero assures them that sects such as the Saivas, Pasupatas and Kalamukhas have nothing to fear as they are in line with vedic values and practices, as is the Bhagavata sect that reveres the Paficaratra scriptures dealt with in the final act.
That Jayantha is writing about the Tantras probably before 900 ce, a hundred years before the Saiva polymath Abhinavagupta, is significant for it shows that these traditions had achieved a strong degree of development by his time. It also shows that these traditions are indeed still developing with new texts being produced with such appeal that thinkers like Jayatha feel the need to make judgements about them. Indeed, Dyczkowski observes that the Saiva Tantras proliferated at an extremely rapid rate in the centuries before Abhinavagupta. 16 Jayantha specifically mentions Saiva and Paficaratra texts, and it is this broad distinction that we need to give some account of, for this is the body of material that provides us with the ritual foundations that define the tantric traditions and become so influential. This is not the place for a systematic exposition of the main textual developments in the traditions; for that I refer the reader to the excellent essay by Alexis Sanderson. 1 ' The intention is,
rather, to provide some orientation and to give some broad indication of their historical location. The entextualisation of the body in the tantric traditions is the entextualisation of specific texts, written in specific times and places.
The Tantras are dialogues between the main deity of the tradition and his/her spouse or a sage. Tantras focused on Siva are presented as a dialogue between him and his Goddess or Sakti, Tantras of Visnu between him, particularly in his form as Narayana, and his consort LaksmI, or with a sage such as Narada and in some Tantras focused on a form of the Goddess; it is Siva who asks questions of her. These texts are traditionally divided into four sections, knowledge (jndna), yoga, acting (kriya) and behaving (carya). Very few are actually constructed like this and those that are tend to be later, although this nevertheless provides a useful way in which to approach the texts. Most Tantras are primarily concerned with kriya and carya, with daily ritual, with temple construction and the consecration of images. The Tantras themselves are generally little concerned with philosophy in the sense of presenting arguments about the nature of being and knowledge, but they do contain metaphysical speculation about the structure of the cosmos. Indeed, this is fundamental to many texts and, even if not explicitly stated, informs descriptions of ritual.
The Pancaratra Revelation
Along with Jayantha, other orthodox thinkers took up the defence of some of the Tantras. Within the Vaisnava tradition Yamuna (c. 918-1038 Ce), the grand teacher of the famous theologian Ramanuja, wrote the Agamapramanya, a defence of the revelation of the tantric Vaisnava or Pancaratra tradition. The Pancaratra sources provide a large body of texts concerned with the usual tantric topics of cosmology, initiation, daily and occasional ritual, mantras and the construction of temples. Yamuna defends this body of texts as being on a par with the Veda: 'The Pancaratra Tantra is authoritative like the vedic sentences ordaining sacrifice (jyotistoma etc.) on the grounds that it is based on knowledge free from all defects.' 18
Indeed, Yamuna agrees with Jayantha and the Mlmamsakas that scripture is self-validating, that its authority is not questionable because the texts are the utterance of the Lord of the Universe, Vasudeva. According to another Pancaratra defender, Amalananda, the Agamas do not have the same self-authenticating validity as the Veda, but their authenticity is nevertheless assured because the Veda bear witness to the omniscience of Vasudeva." Evidently Yamuna's defence was successful in so far as Ramanuja accepts the authority of the texts (although perhaps with some diffidence) and Pancaratra rites become central in the Sri Vaisnava tradition that became the dominant form of Vaisnavism in the South. 20
Two traditions within Vaisnavism lay claim to the designation tantra: the Vaikhanasa tradition and the Pancaratra. The Vaikhanasa regards itself as wholly orthodox and in line with vedic revelation, although it has its own texts, the fourth-century ce Vaikhdnasa-sutra that described daily worship of Visnu and a collection of Samhitas which describe offerings to the emanations of Visnu or Vasudeva, Purusa, Satya, and Acyuta, that we also know from the Pancaratra Jaydkhya-samhitd (see p. 102). The Vaikhanasa texts, as Colas shows, divide what they call vaisnava tantra into the Vaikhanasa and Pancaratra, where the former is the principal (mukhya) tradition and the latter the complementary (gauna) to protect it. Yet the tradition also claims to be vaidika and of gentle (saumya) quality, in contrast to the Pancaratra, which is tantrika and non-vedic (avaidika). 2] Clearly the Pancaratra must be seen as an independent tradition not subordinated to the Vaikhanasa, but the connections between the two traditions show the complexity and overlapping nature of the terms tantrika and vaidika.
The Pancaratra Samhitas form a massive body of texts which have received comparatively little scholarly attention, although Otto Schrader's Introduction to the Pancaratra (1916) remains an exemplary work. 22 There are three texts regarded as key, namely the Sattvata-samhitd, Pauskara-sam hitd, and the Jaydkhya-samhitd, known as the 'three gems'. 23 These texts are believed to be the revelation of Visnu or Vasudeva, also called Narayana but are clearly within the general category of tantra and dealing with the general topics of cosmology, mantra and ritual. The dating of these texts is difficult to establish,
but the Jaydkhya is quoted by the Saiva thinker Utpaladeva (c. 925-975 ce) and so predates him.
The Saiva Revelation
Orthodox thinkers such as Jayantha Bhatta and Yamuna are keen to establish the legitimacy of much of the tantric revelation, or part of it, by asserting its vedic inheritance and claiming that the teachings of these texts do not contravene vedic injunction. Another strategy, however, was very different, and this was to proclaim boldly the superiority of the tantric revelation over the vedic. The Veda are for an earlier time and for a lower level of understanding, but the tantric revelation is the truth of God opened out in a graded hierarchy for the initiate. This was the view of the non-dualist Saiva thinkers of Kashmir, particularly Abhinavagupta (c. 975—1025 ce), who argued not only for the legitimacy of the tantric revelation but for its superiority, especially in his monumental Illumination of the Tantras (Tantraloka). While theologians of the Saiva Siddhanta, such as Ramakantha, wished to align their scriptures and practices with vedic orthodoxy, theologians of the non-Saiddhantika traditions - commonly referred to as 'Kashmir' Saivism - on the contrary wished to distance their scriptures from what they perceived to be the restrictive and limited nature of the vedic scriptures. While clearly being well versed in the orthodox texts, Abhinavagupta and his followers saw these merely as 'external' scriptures and as inflows into a higher expansion of consciousness articulated through the Saiva revelation.
The structure of the Saiva canon and the traditions that it expresses are complex. 24 Much of the voluminous tantric corpus arose in the context of yogic and visionary practices, particularly the Buddhist 'pure vision' texts and the 'treasure system' or discovering hidden treasure (nidhi, gter-ma) such as sacred texts found in both Buddhist and Hindu tantric traditions. 2 ' While such texts are in one sense new, they are nevertheless part of an ongoing tradition of revelation and a canon that is not fixed in the early medieval period. For now it is important to understand the tantric Saiva view of revelation in order to comprehend the ways in which these texts become internalised by
the practitioner. The body of the practitioner reflects the body of the text. For the non-dual Saiva theologian Abhinavagupta, revelation is divine speech; the making known to human beings the nature of transcendent reality, the processes whereby that reality takes on form as, and in, the world, and the methods for its realisation. Abhinavagupta sees scriptural revelation as the disclosing of divine reality, which for him is pure, universal consciousness {caitanya, cit, samvit), the highest expression of which is articulated in the Tantras of non-Saiddhantika tradition called the Trika and the related tradition of the Krama. Indeed, there are different levels of revelation linked to different levels of understanding, which are further linked to the levels of a hierarchical cosmos. For Abhinavagupta the highest revelation is a text called the Malinivijayottara-tantra, on which he wrote a commentary (slokavdrttika) and on which his Tantraloka is a practical text of exposition or manual {paddhati) along with its summary, the Tantrasdra. 26 While the Ma lint itself appears to follow a dualist metaphysics, as Sanderson has demonstrated, Abhinavagupta projects on to it the monism derived from his Krama sources and from his own lineage in the 'recognition' or Pratyabhijna school. 27 For Abhinavagupta, revelation, consciousness and cosmology entail each other. Thus he saw the texts of the Saiva Siddhanta, the dualist tradition of Saivism that aligned itself with vedic orthodoxy, as being a lower level of divine disclosure than the texts of his own Trika and Krama traditions, which, according to him, revealed the true nature of reality as non-dual; that ultimately there is no distinction between self and Siva, nor between self and world. The truth of scripture, its esoteric heart, reveals the nature of self and world as dynamic, vibrating consciousness (spanda).
Abhinavagupta classifies the tantric revelation into three divisions in his commentary on the Malini: the division of Siva (sivabheda), comprising ten Tantras; of Rudra (rudrabheda), comprising eighteen Tantras; and of Bhairava (bhairavabheda), comprising sixty-four Tantras. 28 These categories of text express the metaphysical positions of dualism, dualism-cum-nondualism, and non-dualism respectively, of which the latter is the superior for Abhinavagupta. Certainly the Saiva Siddhanta accept twenty-eight Tantras as authoritative (the ten Siva and eighteen Rudra), although the lists vary in different texts
The Tantric Revelation
57
and there are also complementary texts or Upagamas associated with them. 29 This fairly simple division is complicated by Abhinavagupta, in that he needs to relate it to the classification found in the Tantras of a division into five streams flowing from the five mouths of Siva in his form as Sadasiva. The form of Sadasiva with five faces is primarily a body of power (sdkta vapus) made up of mantras. On this account the source of the scriptures is the mantra body of Siva, the body of power and sound. Abhinavagupta describes the Sadasiva as having five mantras as his body, namely Isana, Tatpurusa, Aghora, Vamadeva and Sadyojata, each of these facing a direction in which the revealed tantric corpus flows. 30 The mantras of these five, as found, for example, in the Mrgendragama following the Kamikdgama, are as follows: 31
Om hom isanamurdhne namah Om hem tatpurusavaktraya namah Om hum aghorahrdayaya namah Om him vamadevaguhyaya namah Om ham sadyojatamurtaye namah
Each of these is associated with the directions and other pentads in Saiva theology, particularly the five acts of Siva of creation, maintenance, destruction, concealing and revealing and with classes of scripture and teachings. Thus we have the following correspondences detailed by Hanneder: 32
Hanneder explains that the scripture 'consisting of all knowledge' (sarvavidydtmaka) refers to the next set of correspondences, namely the scriptures of the path of mantras. Some later sources complicate
the scheme further by categorising the tantric scriptures into twenty-five streams (five times five faces). 33
Tantric Saivism is therefore the path of mantras which flows from the upper face of Sadasiva. This Isana face is further divided into five currents of groups of Tantras, as follows: 34
Relating this to Abhinavagupta's threefold classification, the sivabheda and rudrabheda flow from the Isana face while the bhairavabheda fuses the northern and southern faces. 35 Abhinavagupta further complicates the scheme by reference to a lower, hidden face turned towards the subterranean worlds (naraka). The Siddhanta Tantras are the twenty-eight dualist texts, and the Bhairava Tantras are those of the non-Saiddhantika tradition that forms the scriptural basis of Abhinavagupta's Trika. Hanneder quotes a text, the Sivatattvaratnakara, that describes the four streams below the Isana face, saying that the Garuda Tantras teach the Tatpurusa mantra as the antidote for snakebites and poisoning; the Bhairava Tantras teach the destruction of enemies; and the Bhuta Tantras teach mantras and herbs for the pacification of ghosts and demons. 36
Abhinavagupta has the highest regard for revelation (agama), which, he says, forms the basis for one's life {upafivyaf 1 and which should be followed in order to reach perfection. This perfection is achieved quickly through pursuing the teachings in the scriptures of the left stream (vdmasasana) and transcending the vedic scriptures, which rest in the 'womb of illusion' (mayodarasthitam). 3S These scriptures lead to the highest perfection of consciousness (samvitsiddham), a perfection to be realised in one's own experience (svanubhavasiddham) beyond the mere ritual action declared in the Veda that should be forsaken. 39 Relying on Saiva scriptures allows us to go beyond apprehension or fear (sankha) characteristic of the
The Tantric Revelation
(based on Sanderson's mapping of the traditions)
Veda Purana Tantra
Mimamsa,
Nyaya
interpretation Smarta worship non-Puranic
of Siva and Visnu worship of Siva
atimarga (Pasupata Sutras)
mantramarga
Saiva Siddhanta (dualist Tantras)
non-Saiddhantika
groups
(Bhairava Tantras)
Kaula Tantras < + Degree of conformity to vedic values >
Veda and orthodox Brahmanical teachings, for the Saiva teachings are their reversal (viparyaya). w Abhinavagupta further subdivides the Bhairava Tantras into four 'seats' or 'thrones' in ascending order of importance, that of mandala (mandalapitha), mudra (mudrdpitha), mantra (mantraptiha) and the throne of vidya {vidydpitha), where vidyd doesn't simply mean 'knowledge' but is a kind of mantra associated with female deities. This is a feature of the distinction, the mantrapitha being connected to male deities, the vidydpitha to female ones. 41 The Svacchandabhairava-tantra, a text popular in the Kashmir valley at the time of Abhinavagupta, is an example of the former, while the Siddhayogesvarimata is in the latter category, with the Mdlinivijayottara-tantra as its essence. 42 It is this text, itself part of the longer scripture (the Siddhayogisvarimata) that forms the basic scriptural authority for Trika Saivism, which Abhinavagupta regarded as the highest revelation of Siva. It describes itself as having been a small part of the much larger scripture but reduced for the understanding of those possessing only weak intellects (alpadhihita). 43 Thus for Abhinavagupta we have a graded hierarchy of revelation,
with the wholly external Veda being transcended by scripture focused on male power (the mantrapitha), being superseded by the most esoteric focused on the power of the feminine divine (the vidyapitha).
Text and Tradition
The precise relationship between the indigenous classification schemes outlined above and the social-historical development of the tantric traditions is not clear, but the schemes do arguably represent forms of self-description that corresponded to specific traditions, although the relationship between text and tradition is complex in the Saiva case. Sanderson has mapped out this relationship in some detail; what follows is a simplified reading of his mapping. 44 If we understand this revelation in terms of proximity to orthodoxy and vedic revelation, then on the one hand we have worshippers of Siva wholly in line with Smarta brahmanical orthodoxy who follow rites of worship expressed in the Puranas, while on the other hand we have non-puranic worship of Siva. These Saivas were ascetics known generally as the Pasupatas, who thought of themselves as following a higher or outer path (atimarga) and fulfilling a fifth stage beyond the four orthodox stages or ways of life (asrama). 4i Although they were ascetics, they became highly successful in terms of control of temples and with a great deal of political influence. Indeed, they displayed martial qualities which aligned them with the later naked ascetics, the Nagas, who defended orthodox dharma. 46 One branch of the Pasupatas, the Lakula, advocated practices threatening to Brahmanical orthodoxy, namely the carrying of a cranium begging bowl and skull-topped staff, and taking the great vow (mahavrata) or penance prescribed in the dharma literature for killing a Brahman of wandering as a mendicant carrying his skull for twelve years. 4 ' In carrying a skull these ascetics imitated Siva, who in myth followed this 12-year penance for decapitating one of Brahma's five heads with the thumb nail of his left hand. In the twelfth year the skull fell from his hand at Kapalamocana in Benares. 48
Technically the puranic followers of Siva were Mahesvaras concerned with ritual purity and following orthodox, puranic worship
of Siva, while those who had undergone an initiation, such as the Pasupatas, were Saivas. So, worship of Siva can be classified into the Mahesvaras and the Saivas. The Saivas themselves can be classified into the higher or outer path (atimarga), flowing from the mouth of Tatpurusa, and path of mantra (mantramarga), flowing from the mouth of Isana, which follows the revelation of the Tantras. This classification scheme further breaks down the mantramarga into the Saiva Siddhanta, whose focus is the deity Sadasiva and whose followers saw their revelation as not disruptive of the Veda and Brahmanical social norms, and non-Saiddhantika groups. The Saiva Siddhanta is normative tantric Saivism, the basic system of the non-Saiddhantika traditions, which sees itself as in line with vedic revelation and the teachings of the orthodox Brahmans. At the other extreme we have non-Saiddhantika Tantras, whose focus is the ferocious form of Siva, Bhairava, and whose followers situated themselves within the culture against Brahmanical orthodoxy. These Bhairava Tantras were the revelation of traditions which propagated practices that went against orthodox values, particularly expressed in making offerings of meat, wine and sexual substances to appease their ferocious gods (see pp. 165, 169). The followers of these texts, and their originators, were the Kapalika ascetics who inherited the Lakula practice of the great vow. They used a skull begging bowl, covered themselves with the ashes from cremation grounds, and carried a skull-topped staff (the forerunners of the modern Aghoris). 49 In the early medieval period, texts produced in their milieu became the main scriptural authority for the monistic Saivism of Kashmir, focused on the ferocious Bhairava or Sakti as Kali in one of her forms. Indeed, Tantras devoted to the Goddess became important especially in the later tradition, and we need to mention here one last classification scheme, that of the revelation of the Kaula Tantras.
While the Bhairava Tantras are an early, prolific and most important development within Saivism, a further group of texts was developing at the same time which saw themselves as being within a tradition that emphasised the Goddess or Sakti, the power of Siva. These traditions called themselves the 'family' (kula) or those traditions related to one of the families of goddesses {kaula). But while there is an emphasis on Sakti, the Kaula Tantras nevertheless
regard themselves as Saiva and worship Bhairava as their supreme deity. In complete contrast to the Tantras of the Saiva Siddhanta, the Kaula Tantras are mostly concerned with private ritual in secluded places and making offerings of meat, wine and sexual substances (kundagolaka) to ferocious Bhairava and his consort Bharavi. An important classification of this group of Tantras is found in texts such as the Yoginihrdaya. This text divides scriptural transmission into four currents: the eastern or primary (purvdmnaya), containing texts of the Kaula tradition and worshipping Siva and the Goddess as Kulesvara and Kulesvarl; the upper transmission of the ferocious Goddess Guhyakali pertaining to the Krama tradition; the Western transmission of the crooked Goddess Kubjika associated with Kundalini; and the southern transmission forming the Srividya tradition focused on the gentle, erotic Goddess Tripurasundari. 50 The Srividya in particular grew and developed in South India, where it exists to the present day. It is often the Srividya which is taken as the standard model of Tantrism, but in the present text it will only be looked at tangentially.' 1
The important Trika ('threefold') based on the non-Saiddhantika Tantras is so called because of the three goddesses Para, Parapara and Apara named in the Mdlim. Abhinavagupta tries to show how these goddesses are themselves emanations of a single, underlying reality of consciousness and he suffuses the text with his idealism partly derived from his initiation into the Krama system, a rigorously idealist system that saw the world only in terms of vibrating consciousness. This text forms the basis of Abhinavagupta's system, and his commentary on the text (vdrtika), along with the independent work Tantrdloka and its summary the Tantrasdra, is exegesis of this scripture. 52
The Tantric Theology of Revelation
While texts of primary revelation, the Tantras, are mostly concerned with cosmology and ritual and not explicitly with philosophical argument, tantric theology, such as the recognition school (Pratyabhijna), tried to maintain the universality of supreme consciousness 53 and
to refute schools such as the Nyaya which maintained a form of dualism in which the body and self can exist without each other. Yet while wishing to maintain the universality and superiority of their doctrines over the vedic schools, and so identifying universality with truth, this identification is not matched at the level of ritual and its textual instantiation. Here, rather than truth being identified with universality, it is identified with particularity; with the particularity of revelation (visesasastra) in contrast with the general revelation (samayasdstra) of the Veda and lower scriptures. On the one hand, in doctrine and argument we have the refutation of other schools and the maintaining of the universality of consciousness; on the other, we have the refutation of other schools by the disparaging of universality and the emphasising of the particular, esoteric revelation of the Tantras in a graded hierarchy, revealed through an initiatory structure through a master (guru, acarya). For the monistic Saivas, the higher up the scale the more particular the revelation and the closer to the truth of pure consciousness; the lower down the scale, the more general the revelation and the further from the truth of pure consciousness. This is not so much a contradiction, because the claims operate at different levels, as an attempt to bring together the universal and the particular, which can be seen, above all, in the tantric ideas of the power, vision and levels of awakening located within the body.
If we can speak of a tantric theology of revelation, then we might say that it is characterised by a belief in a hierarchy of revealed truths and that this hierarchy is liturgically expressed in a hierarchy of initiation. Thus for Abhinavagupta, Saiva Siddhanta initiation revealed in the dualist Tantras is the expression of, and gives access to, the cosmic level from which its revelation originates (namely Sadasiva). By contrast, Trika initiation revealed in the non-dualist Tantras is an expression of and gives access to a higher revelation from the supreme Siva or even from the Goddess (Kalasamkarsim). In all cases we see that the tantric traditions generally regarded their scriptures as transcending those of the vedic tradition. One should, perhaps, speak of tantric theologies of revelation in so far as monistic traditions such as Abhinavagupta's 'recognition' school (Pratyabhijna) must ultimately undermine the very notion of revelation as coming