French Studies in South Asian Culture and Society The aim of the series is to provide the English-speaking reader with a relevant selection of contemporary French works in the field of Human Sciences dealing with South Asia. It is widely recognized that French scholarship has produced distinctive theoretical perspectives which differ in many respects from the established Indian and Anglo-Saxon scholarly traditions. Significant studies ranging from Classical Indology to most social science disciplines are thus proposed for inclusion. With this objective in view, the Maison des Sciences de l'Homme in collaboration with the Oxford University Press in India has sponsored the present series under the editorial direction of a joint committee of French and Indian scholars representing the major social science disciplines. Editorial Committee Jean-Claude Galey Rajni Korhari Triloki Nath Madan Charles Malamoud Ja cq ues Pouchepadass Romila Thapar French Stud1:es in South Asian Culture and Society Ill HINDUISM The Anthropology ofa C£·vilization MADELEINE BIARDEAU Translated from the French by RICHARD NICE OXFORD UNJVERSI1"Y PJI.ESS OXFORD YMCA Libra1y Building. Jai Singh Road. New Delhi 110001 Oxford University Press is a department of th.e University of Oxford. lt 1\.trthers the Unive.rsity"s objc{i:ive of ex,·elJen<:e in r<~search. schoiarship. and edtKat.ion by publishing woddwide in Oxford New York Auckland Bangkok Bue!los Aires Cape Town Chennai Dares Salaam Delhi Hong Kong Istanbul Karachi Koli<ata Kuala Lumpur Madrid Melboumc Mexico City Mumbai Nairobi Sao l'aulo Shanghai Singapore Taipei Tokyo Toronto and associated companies in Berlin nndan Oxford is a registered tradC" mark of Oxford University Press in the UK and in n:-1tain other <:ountries Published in India By Oxford University Press. New Delhi The moral rights of the author have been asserted Datab<lse right Oxford Univ<."rsity Press (makeri original edition © Flammarion. Paris, "I 981 English edition © Oxiord Univ~rsity Press. 1989 Oxford India Paperbacks 1994 Sixth impression 2002 The publisher acknowledge~ with thanks the support given by the Maison des Sciences de I'Homme. Paris ISBN 0"19 563389·X All1ights reserved. No part of this publication may be reprod\lced. srored in a retrieval system. or transmitted. in any form or by any means. withO\Jt the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries conceming .reprodu<.tion outside the scope of the above should be sent to the Rights Department. Oxford University Press. at the address above You must not cirlulate this book in any othe-r binding or cover and you must impose this same condition on any ac.quirer Printed ar Baba Barkha Nat.h Printers. New Delhi 110 01 S and published by Manza~ Khan. Oxford University Press YMC.A Libra1y Building.Jai Singh Road, New Delhi 110001 Contents lmroduc;:ion CHAPTER l: MAN AND THE ABSOLUTE 16 The -:A.bsolute and the Brahmans 17 Two paths to salvation 25 The hermit Brahman 30 l·findu man 32 Cl·-IAPTER 2: THE FOUR GOALS OF MAN 41 The levels of dharma 42 Love: women's duty 46 'T'he craft of kingship 53 Portrait of the ideal Brahman 59 Two types of men 64 Renunciation and hurnan love 66 CHAPTER 3: SALVATION TI:--rROUGH DEEDS: THE YOGIN \Y/ ARRIOR 69 An anthropology of desire 70 The cternkt.y of speech and the power of vision 73 Instantaneity 79 From renunciation to bhakti 84 The universe of bhakti 88 God as yogin and creator 91 Fint cosmogony 93 VIShttu and Si'O.Nl 96 Sccuntl cosmogony 98 Cyclic time l 00 Salv:~tion 104 The l"nd of the world and individual salvation 108 Desireless activity I J 3 VI Contents CHAPTER 4: DIVINE LOVES 122 Privileged experiences 123 The goddess 128 The divine game 143 Tantrism 148 Tantrisrn and bhakti !55 CH/i.PTER 5: HINDUISM AND THE FUTURE 159 Notes 169 Glossary 177 Introduction To aim tQ,present a unified account of a culture and a history as vast as that of India, explicitly linking each part to a single totality, is no doubt a somewhat reckless undertaking. Would anyone dare to define :i'unit of the same kind within Western civilization, still less presume to embrace the whole of Western history in a single synthesis governed by a few simple norms that can be traced everywhere? India is, however, a particularly favourable case for an undertak ing of this sort. The texts and archaeological remains which it offers extend over a period of at least three millenia; and yet, in spite of a sizeable collection of inscriptions on stone or on copper plates, which have by no means all been studied or even published, this evidence scarcely enables us ro reconstruct what one would call a history. It is true that we possess a few chronological references partly hypothetical and still disputed-but these rare dates are not always panicuiariy significant. In contrast to what has occurred with some other civilizations, we do not, for example, possess any ancient royal chronicle: even the Rajatarangi~li of Kashmir {twelfth century) cannot be regarded as an exception, composed as it was in that eternal marchland, in contact with the 'Turks' and in an atmosphere in which rlindu norms were constantly flouted. The occasional Greek or Chinese visitors and later Muslim observers provide a few precious landmarks, but these too must be handled with care since they are based as much on hearsay as on direct information. The paucity of strictly historical data might, on the contrary, serve as a challenge to scholars and exercise their sagacity; this it has indeed done, but on the basis of essentially a-historical material, since the enormous Sanskrit literature which extends through the ages up to the present day operates in the realm of the normative or the mythical or in their service.1 The Arthasastra, a treatise on the art of royal governance that is almost unique in its own genre, is 2 Hinduism attributed to Kaurilya, who is thought to have been a minister of rhe emperor Candragupta Maurya (fourth century se). But who would be so bold as to date the text in that period? Estimates range between the fourth century BC and the fourth century AD. The famous Treatise on Love, the Kama Sutra, is equally ill-dated. As for that vast narrative and didactic summum, the epic of the Mahabharata, it is thought to have been composed over a period of eight or ten centuries straddling the beginning of our era, but one would be hard put to find indisputable arguments to support this. Yet we know that the kings had their bards and their panegyrists, but these streams of eloquence have mostly come down to posterity only in mythic form, the person of the king being itself stereotyped or even reabsorbed into the divine model, as the avatara of Vi,<jQU. The other epic, the Rdmaya1Ja, more modest in scale, is not even classified by the Hindu tradition in the same literary genre as the Mahiibharata, although it too is a royal saga: it is a 'poem' and not a history; it was probably composed over a long period, and the arguments in favour of its being eariier than its literary twin are counterbalanced by evidence of later composition. Only when one comes to recent centuries do dates become more precise, attached to works by definite aurhors for which at least a relative· chronology becomes possible.2 But by then, the major mental and social frameworks of 1-iiuduisrn had already been established. Buddhism had been expelled or absorbed, Ja inism had withdrawn into itself to resist Hindu culturai pressure, and the tribes had been driven into the least habitable parts of the territory although they too were not allowed to ignore the powerful society which contained them on all sides. Kingdoms great and small, longlasring or ephemeral, even empires, succeeded one another, but hrahhlai1ic India continued to adhere tO its own norms; its thinkers and authors-all Brahmans have given her a fundamentally timeless image, intended in their minds to live eternally, since she was the cen.tre of the world, the measure of salvation, and spoke the language of the gods. That is indeed the main justification for an attempt at a systematic presentation. This endeavour consists in taking literally the desire of a whole society, as expressed by its scribes but also no doubt with a very broad consensus, to present itself as a well-ordered whole, as the realization of a socio~cosmic order which promises it eternity or Introduction 3 an everlasting renewal. This amounts to saying that change, when it does appear, is only supe~ficial and always refers back to a normative foundation, the one source from which spring the most trimsicnt phenomena. It is then a question of adopting this language as far as one can, of entering into this world view, forgetting one's own categories so as to discover those of the Other, accepting that its lines of cleavage are not ours, that its values are distributed in a way that for us is unexpected. Such a perspeCtive in fact does no more than reconnect with the very modern and very Western preoccupations of the historians of 'mentalities'. There is simply a change of scale when dealing with India, because one cannot 'periodize' its history as one does for other areas, or divide it into territories as restricted as those of European countries. The documentation available does not lend itself to this either in quantity or, above all, in content, since it speaks of something other than 'history'. The psychological, the individual, the momentary collective movement escape us-·-or are they simply absent? A rigorous study of mentalities must in fact take care not to see these lacunae as something negative, as a limit on investigation. On the contrary, they have a heuristic value if they are regarded as indices of what the Hindu will or can say of himself. I do not deny that this is a risky undertaking: how can one be sure that the reduction of the fact to the norm does not conceal the intrusion of a historical contingency the data of which elude us? One cannot hope to make a system of Hindu culture as a whole, without an:y remair1der. In particular, this would be to make light of the centuries of rationalizations which the Brahmans have accumulated in all good faith, deceiving themselves before catching us in their trap. More modestly, therefore, I shall endeavour to show the probability of an interpretation by examining it from several angles, not forgetting that the most solid pieces of evidence may themselves prove to be deceptive, that the insignificant anecdote takes on··the air of a myth the better to charge itself with meaning. In short, l shall place a wager on meaningfulness, while remaining aware that meaningless ness also exists. The twofold demand for an overall system and for a meaning immanent in each of its partial manifestations runs counter to an lndological tradition which has been firmly rooted in the West for decades and which has even caught on among contemporary Indian 4 Hinduism intellectuals. The vast sub-continent of India and its millenia oi history do indeed at first sight give the irnptession of an irreducible diversity, an incoherent proliferation, even on the socio--religious plane where the inviolable norm ought to prevail: an innumerable pantheon, local caste systems, variations in marriage rules or diet, etc. The habit has developed of presenting Indian culture as a mosaic from which the unifying pattern seems to be excluded; some writers asseft that one would be hard put eo find a single belief valid over the whole Indian territory to provide cultural uniry _ What is offered here is, by contrast, a sketch of the main articulations of a single organization of the pantheon, whatever the multiplicity of divine names and cultural forms; of the central features of a caste system underlying the regional or local diversities; and of the principles which govern socio-religious relations. Unity is sought not so much in a synthesis whose harmony derives from the scholar's 'idea', but rather at a deeper level of the analysis of the data, on the basis of the explicit or implicit norms which every Hindu carries in his head. But mosaic-style construction has been only one of the tempta tions of the Indologists, the one which has furnished so many manuals, especially catalogues and indexes which are of great value as exhaustive information. The most profound work, however, that of scrupulous philologists with weH-kept files, has gone on according to quite different principles, in which specialists of classical antiquity would quickly recognize the skeletons in their own cupboards. Meticulous and extraordinarily erudite mono graphs have been published over the last century, and some of them are rightly still regarded as authoritative. Their central concern has been with Vedic literature, in which one might hope to get a grip on things in statu nascendi, or in their relationship to a more distant past, common to the lndo-European peoples who are fast becoming one of the myths of our century_ It is in itself an enormous corpus in which the various states of the language betray an evident temporal thickness, even if relative dating remains problematic. The mosaic effect is then avoided by means of a distribution of supposedly incoherent facts along the course of a development in which history is seen at work, entirely reconstructed according to plausible lines of evolution whose logic is no doubt essentially that of the scholar's mind. Introduction 5 This staggering in time has also been assumed in order to account, firstly, for the break in continuity presented by the. so-called revealed texts-the Veda as a whole--and the later tradition; then, even within classical or more recent Hinduism, temporal evolution seemed the only mental tool capable of bringing order to the proliferation of myths and the multiplicity of apparently irreconcil able versions, as well as the often contradictory diversity of the normativ,e texts. Where cleavages were clearly visible, history came in to shed its own light on the problem. It would be pointless to deny that history had its reasons and that they are irrefutable: the 'Aryan/ entry into India took place from the north-east, along the very route which was that of all the subsequent invasions. But the 'Aryan' tribes-nomadic and pastoral, as they used to be described--did not take possession of a land empty of men. Whatever the ethnic and cultural identity of the so-cailed Indus civilization (two ro three millcnia BC), it seems that largely Dravidian populations occupied the northern plain as well as the peninsula. So one can be sure that the invaders encountered organized societies, with their own beliefs and norms, in a word, their own culture; and since the 'Aryan' advance seems to have progressed towards the east and south, it is also safe to assume that the newcomers were able to conquer the previous occupants. We thus have here a clear principle which will serve to organize the fragmentary information given by the texts: if the Veda oppose noble, pure peoples to human groups whom they cast into outer darkness, this is no doubt because the victors had subjugated the vanquished. But if a god, a notion or a word appear in the texts at a given moment, this is, on the contrary, because the presence of the newly subordinated populations is making itself felt in the mental universe of the victors and symbiosis is producing effects that one would expect. However, it is when one moves from the Vedic corpus to the. enormous mass of archaeological and textual documents of Hindu ism that this explanatory principle proves most effective. The break in level between the two sets of data is such, without it being possible to situate clearly a transitional phase or intermediate states, that the philologist-historian has no words strong enough to characterize this shift: intrusion, irruption, pressure from the base etc., at the origin of which some even go so far as to imagine a