Wiley □ RAI □ Anthropological Conceptions of Religion: Reflections on Geertz Author(s): Talal Asad Source: Man, New Series, Vol. 18, No. 2 (Tun., 1983), pp. 237-259 Published by: Royal Anthropological Institute of Great Britain and Ireland Stable URL: http://www,jstor.org/stable/2801433 Accessed: 12-02-2016 07:54 UTC Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. (Si m 4 mi Wiley and Royal Anthropological Institute of Great Britain and Ireland are collaborating with JSTOR to digitize, preserve and extend access to Man. STOR http://www.jstor.org This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions ANTHROPOLOGICAL CONCEPTIONS OF RELIGION: REFLECTIONS ON GEERTZ Talal Asad University of Hull This article examines Geertz's well-known definition of religion, with its emphasis on meanings, and argues that it omits the crucial dimension of power, that it ignores the varying social conditions for the production of knowledge, and that its initial plausibility derives from the fact that it resembles the privatised forms of religion so characteristic of modern (Christian) society, in which power and knowledge are no longer significantly generated by religious institutions. A critical evaluation of Geertz's text is accompanied by brief explorations of some of the ways in which power and knowledge were connected in medieval Christianity. The article ends with a plea for investigating religion with reference to the historical conditions necessary for the existence of particular practices and discourses. Clifford Geertz's essay 'Religion as a cultural system' is perhaps the most influential, certainly the most accomplished, anthropological definition of religion to have appeared in the last two decades. Originally published in 1966, it was reprinted in his widely-acclaimed The interpretation of cultures (1973). Like most of his writings, it is carefully, almost fastidiously, put together and represents a style of symbolic analysis that has been adopted by many of his students, some now distinguished anthropologists in their own right. It has even found favour with modern existentialist theologians for its concern with culture as symbol and meaning.1 It is an obvious text with which to begin an exploration of the strengths and weaknesses of recent anthropological conceptions of religion. I must stress that although this will be a critical evaluation of that text, what follows is exploration and not refutation. It is concerned less with the soundness or otherwise of individual claims, and more with trying to trace how and why historically specific forms of'religion' have come to be presented, mistakenly, as having a paradigmatic status. It is first and foremost an attempt to identify obstacles in the way of certain types of enquiry. That enquiry, broadly speaking, has to do with the theme of power and religion—not merely in the sense in which political interests have used religion to justify a given social order or to challenge and change it (an important question in itself) but in the sense in which power constructs religious ideology, establishes the preconditions for distinctive kinds of religious personality, authorises specifiable religious practices and utterances, produces religiously defined knowledge. In my attempt to formulate questions relating to power and religion I shall draw on the early history of Western Christianity. This will undoubtedly be Man (N.S ) 18, 237-59. This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions 238 talal asad seen as strange for two reasons: first, because modern anthropology ('the comparative science of human culture') has rarely thought fit to address itself to Christian history, and in matters of this kind most anthropologists are at least as traditional as the societies they usually study. The second is that Geertz makes no explicit reference to Christianity in his general disquisition about religion but adduces examples from Indie and tribal religions, and therefore it may be felt that characteristics of Christianity are not necessarily relevant to other religions. This may be so, but it must be clearly understood that Geertz's enterprise is to formulate a universal, a-historical definition of religion, and therefore a discussion of Christianity cannot be regarded as a priori irrelevant or even marginal to it. I have focused on Christianity for a number of positive reasons. The very dearth of work by anthropologists on Christianity is itself a reason for urging them to turn to the incomparably rich documentary sources which they can draw on to formulate distinctive problems for research on religion. The early history of Christianity provides a particularly splendid field for analysing relations between power and religion. But although Christianity has had its very distinctive preoccupation with power, theorising it and pursuing it in a way that is quite unique, it would be a mistake for anthropologists to think that the problem is therefore of little relevance to the kinds of society they usually study. There is a final reason for focusing on Christian history. It is part of my basic argument that socially identifiable forms, pre-conditions and effects of what was categorised as religion in the medieval epoch were quite different from those so categorised in modern society. Religious power was differently distributed, and had a different thrust. There were different ways in which it created and worked through institutions, different selves which it shaped and responded to, and different categories of knowledge which it authorised and made available. A consequence is that there cannot be a definition of religion which is universally viable because and to the extent that the effects of these processes are historically produced, reproduced and transformed. Anthropologists, however, continue to offer such definitions whose plausibility today is strengthened by the characteristic displacements that have taken place in the distribution of power, religion and knowledge in Western (Christian) society, but which would not have made good sense in the Middle Ages. Geertz's attempt is only the best-known and most elaborate of these. Geertz begins by observing that the anthropological study of religion is stagnant, and suggests that resort to the concept of culture would revive it. Culture is then defined as 'an historically transmitted pattern of meanings embodied in symbols, a system of inherited conceptions expressed in symbolic forms by means of which men communicate, perpetuate, and develop their knowledge about and their attitudes toward life' (1973: 89). This seems to me a questionable formulation of the concept of culture, and one symptomatic of a weakness in much that is to follow. Culture, we are given to understand, enables people to communicate, perpetuate and develop their knowledge about and their attitudes towards life: but there is no concept of the relationship of culture so This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions talal asad 239 conceived to 'life' itself, or to the material conditions and activities for maintaining (or changing) life. Indeed the very expressions 'knowledge about' and 'attitudes towards' life suggest a distanced spectator-role, as compared to 'knowledge from' and 'attitudes in' living. A consequence of this general formulation is that when religion is conceptualised later in terms of communal symbols, it will be isolated from social practices and discourses, and regarded primarily in terms of consciousness. The trouble with this is that it closes off the possibility of examining how 'knowledge' and 'attitudes' are related to material conditions and social activities, and to what extent they are formed by them. It is only because this entire area of investigation has been dismissed from the start that Geertz is able to assert a basic congruence between 'a particular style of life' and 'a specific (if, most often, implicit) metaphysic', which 'religious symbols formulate' (1973: 90). To what degree, if at all, a 'congruence' actually exists, whether it takes place at a purely discursive level (in language) or somehow by 'life' and 'culture' imitating each other—such questions are never considered. For Geertz the existence of a congruence is a 'truism'; all one needs to know is how that congruence can be fully described.2 This conception of culture sets the scene for a definition of religion. A religion, we are told, is '(1) a system of symbols which act to (2) establish powerful, pervasive, and long-lasting moods and motivations in men by (3) formulating conceptions of a general order of existence and (4) clothing these conceptions with such an aura of factuality that (5) the moods and motivations seem uniquely realistic' (1973: 90). This bold formula is intended to articulate a series of disquisitions which will together establish the status of religion as a universal cultural phenomenon. Geertz's first task is to define symbol. 'Symbol' will be taken to denote 'any object, act, event, quality, or relation which serves as a vehicle for a conception —the conception is the symbol's "meaning"' (1973: 91). But this simple, clear, statement—in which symbol (any object, etc.) is differentiated from but linked to conception (its meaning)—is later supplemented by others not entirely consistent with it, for it turns out that the symbol is not an object which serves as a vehicle for a conception, it is itself the conception: 'The number 6, written, imagined, laid out as a row of stones, or even punched into the program tapes of a computer, is a symbol' (1973: 91). Furthermore, Geertz sometimes seems to suggest that even as a conception a symbol has an intrinsic connexion with empirical events from which it is merely 'theoretically' separable: 'the symbolic dimension of social events is, like the psychological, itself theoretically abstractable from those events as empirical totalities' (1973: 91). At other times, however, he stresses the importance of keeping symbols and empirical objects quite separate: 'there is something to be said for not confusing our traffic with symbols with our traffic with objects or human beings, for these latter are not in themselves symbols, however often they may function as such' (1973: 92). Thus 'symbol' is sometimes an aspect of reality, sometimes of its representation. These divergencies are symptoms of the fact that cognitive questions are mixed up in Geertz's account with communicative ones, and this makes it difficult to enquire into the ways in which the two are connected. To begin with This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions 240 talal asad we might say, as a number of writers have done, that 'a symbol' is not an object or event which serves to carry a 'meaning' but a set of relationships between objects or events uniquely brought together as complexes or as concepts,3 having at once an intellectual and an emotional significance.4 Now, if we start by defining symbol along these lines,5 a number of questions can be raised about the conditions which explain how such complexes and concepts come to be formed, and in particular how their formation is related to modes of communication. Half-a-century ago Vygotsky was able to show how the development of children's intellect is dependent on the internalisation of social speech.6 This means that the formation of what we have here called 'symbols' (complexes, concepts) is conditioned, in part at least, by the social relations in which the growing child is involved—by the social activities which he or she is permitted or encouraged or obliged to undertake—in which other 'symbols' (speech and significant movements) are crucial. 'What are the conditions (discursive and non-discursive) which help to explain how symbols come to be constructed, and how some of them are established as natural or authoritative as opposed to others?' then becomes an important element in anthropological enquiry. The danger of Geertz's exposition, where 'symbols' are presented as sui generis, is that it could be ruled out. It must be stressed that this is not a matter of urging the study of the origin of symbols in addition to their function—such a distinction is not relevant here. What is being argued is that the authoritative status of concepts/discourses is dependent on the socially appropriate production of other discourses/activities; the two are intrinsically and not just temporally connected.7 Systems of symbols, says Geertz, are also culture patterns, and they constitute 'extrinsic sources of information' (1973: 92). Extrinsic, because 'they lie outside the boundaries of the individual organism as such in that inter-subjective world of common understandings into which all human individuals are born' (1973: 92). And sources of information in the sense that 'they provide a blueprint or template in terms of which processes external to themselves can be given a definite form' (1973: 92). Thus culture patterns, we are told, may be thought of as 'models for reality' as well as 'models of reality'. While the notion of systems of symbols being 'outside the boundaries of the individual' is a bizarre one (are words we say to ourselves in silence not symbols?), part of the discussion does open up new possibilities by speaking of modelling: the notion of'models for reality' is well worth pursuing. We are not confined now to the view that 'symbols' must only be concerned with knowledge about and attitudes towards life (as though symbols were one thing and life quite another)—there are practices involved (not least of all control of the material means of existence), for symbols as representations are both the constituents and the products of social practice (and so of'life'). Nor does this notion of culture patterns exclude the possibility of conflicting conceptions and discourses —whether in the 'models of sense, or in that of'models for'. However, at the end of this discussion, Geertz unfortunately regresses to his earlier position: 'culture patterns have an intrinsic double aspect: they give meaning, that is objective conceptual form, to social and psychological reality both by shaping themselves to it and by shaping it to themselves' (1973: 93). This alleged This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions talal asad 241 tendency towards isomorphism, incidentally, makes it very difficult to understand how social variation can ever occur—even if we were to attribute (tautologically) a different cultural pattern to every significant difference within society. It is again the metaphor of reflection, with its implication of fit, that seems to lead Geertz into an impasse. By adopting it, he moves away from a notion of symbols which focuses on the relations between socially signifying and psychologically organising practices, and back to a notion of symbols as meaning-carrying objects external to social conditions and mental states. Yet it is evident that Geertz also thinks of symbols as 'doing' something. Occasionally, in his essay, it is people who do things with symbols; more often it is symbols that do things to people. Always a person's confrontation with symbols seems to imply the impossibility of choosing from among them. This certainly seems the case when concrete religious symbols are said both to 'express the world's climate and shape it' (1973: 95). The notion of climate is of course vague enough to include such things as major disputes and conflicts over religious principle, and even, for that matter, progressive loss of faith. But then the notion of 'shaping' (an action so deliberate, so controlled) something as nebulous and volatile as a 'world's climate' is not exactly easy to grasp. Of course this is not his only metaphor for action by religious symbols, but it is an indication of the difficulty Geertz has when he treats religious symbols as active. Geertz has another formulation of their role which is clearer, and also at first sight more persuasive. Religious symbols act 'by inducing in the worshipper a certain distinctive set of dispositions (tendencies, capacities, propensities, skills, habits, liabilities, pronenesses) which lend a chronic character to the flow of his activity and the quality of his experience' (1973:95). And again, 'symbols'are set against mental states. There are hints here of Radcliffe-Brown's theory of the social function of ritual,8 but only hints. Careful thought will show how dubious the propositions are. Can we predict the 'distinctive' set of dispositions for a Christian worshipper in modern, industrial society? Alternatively, can we say of someone with a 'distinctive' set of dispositions that he is or is not a Christian?9 The answer to both questions must surely be no. The reason, of course, is that it is not simply 'worship' but social and economic institutions in general,10 within which individual biographies are lived out, that lend a stable character to the flow of a Christian's activity and to the quality of his experience. If it is objected that modern, industrial society is unique, or at any rate quite unlike the small-scale societies which anthropologists typically study, one can only reply that a general definition of religion must cover all cases or it ceases to be general. These are, of course, an atheist's doubts. A committed Christian might still want to say something a bit like Geertz—to claim, in other words, an internal efficacy for religious symbols in and through worship, evidence for which cannot (almost by definition) be made available to an atheist. But it must be stressed that this is only 'a bit' like Geertz. The dispositions with which Geertz is concerned are primarily social and not metaphysical, and as such, evidence for their existence must be accessible at the level of social behaviour, not of inward This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions 242 TALAL ASAD being. And that brings us to a rather different objection to Geertz's formulation, from a believer's point of view. Religious symbols, Geertz observes, produce two kinds of dispositions, moods and motivations: 'motivations are "made meaningful" with reference to the ends towards which they are conceived to conduce, whereas moods are "made meaningful" with reference to the conditions from which they are conceived to spring' (1973: 97). A modern believer might say that this is not their essence, because religious symbols even in failing to produce moods and motivations are still religious symbols: in other words, they possess a truth independent of their effectiveness. Yet a deeply committed Christian surely cannot be unconcerned at the existence of truthful religious symbols that appear to be largely powerless in modern society. What are the conditions in which religious symbols can actually produce religious dispositions? How does (religious) power create (religious) truth? The relation between power and truth is an ancient theme, and no one has dealt with it more impressively in Christian thought than St Augustine. Augustine developed his views on the creative religious function of power after his experience with the Donatist heresy, insisting that coercion was a condition for the realisation of truth, and discipline essential to its maintenance. For a Donatist, Augustine's attitude to coercion was a blatant denial of Christian teaching: God had made men free to choose good or evil; a policy which forced this choice was plainly irreligious. The Donatist writers quoted the same passages from the Bible in favour of free will, as Pelagius would later quote. In his reply, Augustine already gave them the same answer as he would give to the Pelagians: the final, individual act of choice must be spontaneous; but this act of choice could be prepared by a long process, which men did not necessarily choose for themselves, but which was often imposed on them, against their will, by God. This was a corrective process of 'teaching', eruditio, and warning, admonitio, which might even include fear, constraint, and external inconveniences: 'Let constraint be found outside; it is inside that the will is born'. Augustine had become convinced that men needed such firm handling. He summed up his attitude in one word: discipline!. He thought of this disciplina, not as many of his more traditional Roman contemporaries did, as the static preservation of a 'Roman way of life'. For him it was an essentially active process of corrective punishment, a 'softening-up process', a 'teaching by inconveniences'—a per molestias eruditio. In the Old Testament, God had taught His wayward Chosen People through just such a process of disciplina, checking and punishing their evil tendencies by a whole series of divinely-ordained disasters. The persecution of the Donatists was another 'controlled catastrophe' imposed by God, mediated, on this occasion, by the laws of the Christian Emperors . . . Augustine's view of the Fall of mankind determined his attitude to society. Fallen men had come to need restraint. Even man's greatest achievements had been made possible only by a 'straight-jacket' of unremitting harshness. Augustine was a great intellect, with a healthy respect for the achievements of human reason. Yet he was obsessed by the difficulties of thought, and by the long, coercive processes, reaching back into the horrors of his own schooldays, that had made this intellectual activity possible; so 'ready to lie down' was the fallen human mind. He said he would rather die than become a child again. Nonetheless, the terrors of that time had been strictly necessary; for they were part of the awesome discipline of God, 'from the schoolmasters' canes to the agonies of the martyrs', by which human beings were recalled, by suffering, from their own disastrous inclinations (Brown 1967: 236-8). Clearly Geertz's formula is too simple to accommodate the force of this motif of religious symbolism. Here it is not mere religious symbols that implant true Christian dispositions, but power—ranging all the way from laws (imperial and This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions TALAL ASAD 243 ecclesiastical) and other sanctions (hellfire, death, salvation, good repute, peace, etc.), to the disciplinary activities of social institutions (family, school, city, church, etc.) and of human bodies (fasting, prayer, obedience, penance, etc.). Augustine was quite clear that power, the effect of an entire network of motivated practices, assumes a religious form because of the end to which it is directed, for human events are the instruments of God. It was not the mind that moved spontaneously to religious truth, but power that imposed the conditions for experiencing that truth.11 Particular discourses and practices were to be systematically excluded, forbidden, denounced—made as much as possible unthinkable; others were to be included, allowed, praised, and drawn into the narrative of sacred truth. The configurations of power in this sense have, of course, varied profoundly in Christendom from one epoch to another—from Augustine's time, through the Middle Ages, to the industrial capitalist West of today—but the patterns of religious mood and motivation, the possibilities for religious knowledge and truth, have all varied with them and been conditioned by them. Geertz gives religious symbols a great deal of work to do, and one consequence of the way in which he does this is that important distinctions are obscured, or even explicitly denied. 'That the symbols or symbol systems which induce and define dispositions we set off as religious and those which place those dispositions in a cosmic framework are the same symbols ought to occasion no surprise' (1973: 98). But it does surprise! Let us grant that religious dispositions are crucially dependent on certain religious symbols, that such symbols operate in a way integral to religious motivation and religious activity. Even so, the symbolic process by which the concepts of religious motivation and activity are placed within 'a cosmic framework' is surely quite a different operation, and therefore the symbols involved are different. Put another way, theological discourse is not identical with liturgical utterances—of which, among other things, theology speaks.12 Thoughtful Christians will concede that, although theology has an essential function, theological discourse does not necessarily induce religious dispositions, and that, conversely, having religious dispositions does not necessarily depend on a clear-cut conception of the cosmic framework on the part of a religious actor. Discourse involved in practice is not the same as that involved in speaking about practice. Geertz's reason for merging the two kinds of discursive process seems to spring from a desire to distinguish in general between religious and secular dispositions. The statement quoted above is elaborated as follows: 'For what else do we mean by saying that a particular mood of awe is religious and not secular, except that it springs from entertaining a conception of all-pervading vitality like mana and not from a visit to the Grand Canyon? Or that a particular case of asceticism is an example of a religious motivation, except that it is directed toward the achievement of an unconditioned end like nirvana and not a conditioned one like weight-reduction? If sacred symbols did not at one and the same time induce dispositions in human beings and formulate . . . general ideas of order, then the empirical differentia of religious activity or religious experience would not exist' (1973: 98). The argument is that a particular disposition is religious only because it occupies a conceptual place within a This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions 244 talal asad cosmic framework. But this raises the vital question, which Geertz nowhere considers, as to how the authorising process represents the practice, utterance or disposition so that it can be discursively related to general ideas of order—in short the question regarding the authorising process by which religion is created. Indeed the ways in which authorising discourses, based on a cosmology, systematically redefined religious spaces, have been of profound importance in the history of Western society. In the Middle Ages such discourses ranged over an enormous space, defining and creating religion: rejecting 'pagan' practices or accepting them;13 authenticating particular miracles and relics (the two confirmed each other);14 authorising shrines;15 compiling saints' lives, both as a model of and as a model for the Truth;16 requiring the regular telling of sinful thoughts, words, and deeds to a priestly confessor, and giving absolution to a penitent;17 regularising popular social movements into Rule-following Orders (for example, the Franciscans), or denouncing them for heresy or for verging on the heretical (for example, the beguines, etc.).18 The medieval Church did not attempt to establish absolute uniformity of practice; on the contrary, its authoritative discourse was always concerned to specify differences, gradations, exceptions. What it sought was the subjection of all practice to a unified authority, to a single authentic source which could tell truth from falsehood. It was the early Christian Fathers who established the principle that only a single Church could become the source of authenticating discourse.19 They knew that the 'symbols' embodied in the practice of self-confessed Christians are not always identical with the theory of the 'one true Church', that religion required both authorised practice and authorising doctrine, and that there is always a tension between them—sometimes breaking into heresy, the subversion of Truth—which underlines the creative role of institutional power.20 The medieval Church was always clear about why there was a continuous need to distinguish true religion from falsehood, as well as the sacred from the secular (religion from what was not religion), distinctions for which the authoritative discourses, the teachings and practices of the Church, and not the convictions of the practitioner, were the final test.21 Several times before the Reformation the boundary between the religious and the secular was re-drawn, but always the formal authority of the Church remained pre-eminent. In later centuries, with the triumphant rise of modern science, modern production and the modern state, the Churches would also be clear about the need to distinguish 'the religious' from 'the secular', shifting, as they did so, the weight of religious truth more and more onto the moods and motivations of the believer. Social discipline would, in this period, gradually abandon religious space, letting 'belief, 'faith' and 'conscience' take its place. But theory would still be needed to define religion. It was in the seventeenth century, following the fragmentation of the unity and authority of the Roman Church, that the earliest systematic attempts at producing a universal definition of religion were made. Significant was Herbert's De veritate. 'Lord Herbert', writes Willey, differs from such men as Baxter, Cromwell, or Jeremy Taylor mainly in that, not content with reducing the creed to the minimum possible number of fundamentals, he goes behind Christianity itself, and tries to formulate a belief which shall command the universal assent of all men as men. It This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions TALAL ASAD 245 must be remembered that the old simple situation, in which Christendom pictured itself as the world, with only the foul paynim outside and the semi-toleratedjews within the gates, had passed away for ever. Exploration and commerce had widened the horizon, and in many writers of the century one can see that the religions of the East, however imperfectly known, were beginning to press upon the European consciousness. It was a pioneer-interest in these religions, together with the customary preoccupation of Renaissance scholars with the mythologies of classical antiquity, which led Lord Herbert to seek a common denominator for all religions, and thus to provide, as he hoped, the much-needed eirenicon for seventeenth-century disputes (1934: 114). And so Herbert produced a substantive definition of what later came to be called Natural Religion—in terms of beliefs (about a 'supreme power'), practices (its ordered 'worship'), and ethics (a code of conduct based on 'rewards and punishments after this life')—said to exist in all societies.22 Thus what appears to anthropologists today to be self-evident, namely that 'religion' is essentially a matter of meanings linked to ideas of general order (expressed in either or both rite and doctrine) and that it has universal functions, is in fact a view which has a specific Christian history. From being a concrete set of rules attached to specific processes of power and knowledge, 'religion' has come to be abstracted and universalised.23 In this movement we have not merely an increase in 'religious toleration', certainly not merely a new 'scientific discovery', but the mutation of a concept and a range of social practices which is itself part of a wider change in the modern landscape of power and knowledge. To understand this mutation it is essential to keep clearly distinct what theology tends to obscure: the occurrence of events (utterances, practices, dispositions), and the authorising processes which give those events religious meaning, and embody that meaning in concrete institutions. Not only does Geertz equate two levels of discourse (symbols which induce dispositions and those which place those dispositions in a cosmic framework),24 he also appears, inadvertently, to be taking up the standpoint of theology. This is done by insisting on the primacy of meaning, but without regard to the discursive processes by which meanings are constructed. 'What any particular religion affirms about the fundamental nature of reality may be obscure, shallow, or, all too often, perverse', he writes, 'but it must, if it is not to consist of the mere collection of received practices and conventional sentiments we usually refer to as moralism, affirm something' (1973: 98-9). It appears that to achieve what is truly religion, 'received practices and conventional sentiments' must be joined to discourses which affirm something, which give these practices some cosmological meaning. This is apparently a simple enough requirement, but through it the entire field of evangelism is opened up: early Christians in the late Roman empire; preaching friars in medieval urban centres; European missionaries in Asia, Africa and Latin America. The demand that a practice must 'affirm something', that it should be able to state a meaning, is the first condition for determining what is truly religion. The unevangelised are seen typically either as those who have practices but affirm nothing, in which case meaning can be attributed to their practices (thus making them vulnerable), or as those who do affirm something, probably 'obscure, shallow or perverse' (an affirmation which can therefore be confuted). In the one case religious theory becomes necessary for a correct reading of the mute ritual hieroglyphs of others, for reducing their practices to texts; in the other it is essential for judging their This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions 246 talal asad cosmological utterances. But always, there must be something that exists beyond the observed practices, the heard utterances, the written words, and it is the function of religious theory to reach into, and to bring out, that background by giving them meaning.25 Geertz is thus right to make a connexion between religious theory and practice, but wrong to see it as essentially cognitive, as a means of identifying religion from a neutral place. The connexion between religious theory and practice is fundamentally a matter of power—of disciplines creating religion, interpreting true meanings, forbidding certain utterances and practices and authorising others. Hence the questions that Geertz does not ask: how does religious discourse actually define religion? What are the historical conditions in which it can act effectively as a demand for the imitation, or the prohibition, or the authentication of truthful utterances and practices? How does power create religion? But let us follow Geertz's argument: what kinds of affirmation, of meaning, must be identified with practice in order for it to qualify as religion? Because all human beings have a profound need for a general order of existence, says Geertz, religious symbols function to fulfil that need. It follows that human beings have a deep dread of disorder. 'There are at least three points where chaos—a tumult of events which lack not just interpretations but interpretability —threatens to break in upon man: at the limits of his analytic capacities, at the limits of his powers of endurance, and at the limits of his moral insight' (1973: 100). It is the function of religious symbols to meet perceived threats to order at these three points (intellectual, physical and moral): 'The Problem of Meaning in each of its intergrading aspects ... is a matter of affirming, or at least recognizing, the inescapability of ignorance, pain, and injustice on the human plane while simultaneously denying that these irrationalities are characteristic of the world as a whole. And it is in terms of religious symbolism, a symbolism relating man's sphere of existence to a wider sphere within which it is conceived to rest, that both the affirmation and the denial are made' (1973: 108, emphasis added). Notice how Geertz seems now to have shifted his ground from the claim that religion must affirm something specific about the nature of reality (however obscure, shallow or perverse), to the bland suggestion that religion is ultimately a matter of having a positive attitude to the problem of disorder, of affirming simply that in some sense or other the world as a whole is explicable, justifiable, bearable.26 This modest view of religion (which would have horrified the early Church Fathers or medieval churchmen)27 is a product of Geertz's recurrent desire to define religion in universal terms: the Human Condition is full of ignorance, pain and injustice, and religious symbols are a means of coming positively to terms with that Condition. One consequence is that this view would in principle render every 'philosophy' which performs such a function into religion (to the annoyance of the nineteenth-century Rationalist), or alternatively, make it possible to think of religion as a more 'primitive', a less 'adult' mode of coming to terms with the Human Condition (to the annoyance of the modern Christian). In either case, the suggestion that religion has a universal function is one indication of how marginal religion has become in modern industrial society as the site for producing disciplined knowledge. As This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions TALAL ASAD 247 such it is actually consistent with the conception Marx had of religion as ideology—that is, as a mode of consciousness which is other than consciousness of reality, external to the relations of production, producing no knowledge, but expressing at once the anguish of the oppressed and the cynicism of the oppressor. Geertz has much more to say, however, on the elusive question of religious meaning: not only do religious symbols formulate conceptions of a general order of existence, they also clothe those conceptions with an aura of factuality. This, we are told, is 'the problem of belief, a problem which anthropology has for too long avoided. Religious belief always involves 'the prior acceptance of authority' which transforms experience: 'The existence of bafflement, pain, and moral paradox—of the Problem of Meaning—is one of the things that drives men toward belief in gods, devils, spirits, totemic principles, or the spiritual efficacy of cannibalism, . . . but it is not the basis upon which those beliefs rest, but rather their most important field of application' (1973: 109). Here Geertz seems to be saying that although religious belief has its origin in the existence of bafflement, pain and moral paradox, it does not depend on their existence, but on an authoritative principle which makes it possible to respond adequately to them. Thus he seems to be arguing that religious belief stands independently of the worldly conditions which produce bafflement, pain, and moral paradox, although that belief is primarily a way of coming to terms with them. He seems to be arguing, in other words, that 'belief is independent of its object—and this is surely mistaken, on logical grounds as well as historical. Changes in the object of belief change that belief, and as the world changes so do the objects of belief, and the specific forms of bafflement and moral paradox, which are a part of that world. What the Christian believes today about God, life after death, the universe, is not what he believed 800 years ago—nor is the way he responds to ignorance, pain and injustice the same now as it was then.28 Geertz's treatment of religious belief, which lies at the core of his conception of religion, is a modern, privatised Christian one because and to the extent that it emphasises the priority of belief as a state of mind: 'The basic axiom underlying what we may perhaps call "the religious perspective" is everywhere the same: he who would know must first believe' (1973: 110). In modern society, where 'knowledge' is rooted either in an a-Christian everyday life, or in an a-religious science, the Christian apologist tends not to accept 'belief as the conclusion to a knowledge-process but as its pre-condition. However, the knowledge that he promises will not pass (nor does he claim that it will pass) for knowledge of social life, nor for the systematic knowledge of objects that science provides. His claim is to a particular state of mind, not to a corpus of knowledge. But the reversal of belief and knowledge he demands is not a basic axiom to, say, pious learned Christians of the twelfth century, for whom 'knowledge' and 'belief were not so clearly at odds. On the contrary, Christian 'belief would then have been built on knowledge—knowledge of theological doctrine, of canon law and Church courts, of the detail of clerical liberties, of the powers of ecclesiastical office (over souls, bodies, properties), of the pre-conditions and effects of confession, of the Rules of religious orders, of the locations and virtues of shrines, This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions 248 talal asad of the lives of the saints, etc. Familiarity with all such (religious) knowledge was a pre-condition for normal social life, and belief (embodied in practice and discourse) an orientation for effective activity in it—whether on the part of the religious clergy, the secular clergy or the laity. Because of this the form and texture and function of their 'beliefs' would have been different from the form and texture and function of contemporary 'belief'29—and so too of their doubts and their disbelief. The main point is that 'the basic axiom' underlying what Geertz calls 'the religious perspective' is not everywhere the same. Geertz specifies 'the religious perspective' as merely one among others —common-sense, scientific, aesthetic—and as differing from them as follows. From the common-sense perspective, because 'it moves beyond the realities of everyday life to wider ones which correct and complete them, and [because] its defining concern is not action upon those wider realities but acceptance of them, faith in them'. From the scientific perspective, because 'it questions the realities of everyday life not out of an institutionalized scepticism which dissolves the world's givenness into a swirl of probabilistic hypotheses, but in terms of what it takes to be wider, nonhypothetical truths'. From the aesthetic perspective, because 'instead of effecting a disengagement from the whole question of factuality, deliberately manufacturing an air of semblance and illusion, it deepens the concern with fact and seeks to create an aura of utter actuality' (1973: 112). Geertz has tried to summarise what he thinks common sense, science, and aesthetics are about in contemporary society. It would not be difficult to disagree with him over his characterisations of these perspectives30—and also with the apparent assumption that essentially our 'perspectives' must be the same as those of other times and places.31 However, the most important point to be made here is that the optional flavour conveyed in the term 'perspective' is very misleading when it is applied equally to science and to religion in modern society: religion is optional in a way that science is not. Scientific practices, techniques, knowledges, permeate and create the very fibres of social life in ways that religion no longer does.32 In that sense religion today is a perspective (or an 'attitude', as Geertz sometimes calls it) but science is not. In that sense, too, science is not to be found in every society past and present. We shall see in a moment the difficulties that Geertz's perspectivism gets him into, but before that we need to examine his analysis of the mechanics of reality-maintenance at work in religion. Consistent with previous arguments about the functions of religious symbols is Geertz's remark that 'it is in ritual—that is, consecrated behaviour—that this conviction that religious conceptions are veridical and that religious directives are sound is somehow generated' (1973: 112). The entire long passage from which this is taken swings back and forth between arbitrary speculations about what goes on in the imagination of officiants, and generalised assertions about ritual as imprinting. At first sight this seems a curious combination of introspec-tionist psychology with a behaviourist one—but as Vygotsky33 argued long ago, the two are by no means inconsistent, in so far as both assume that psychological phenomena consist essentially in the consequences of various stimulating environments. In this spirit, Geertz stresses the central importance of ritual to the 'religious perspective', giving extended examples from Hindu This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions TALAL ASAD 249 India and Bali. The function of rituals in generating 'religious conviction' is postulated (thus, 'In these plastic dramas men attain their faith as they portray it', 1973: 114), but how or why this happens is nowhere explained. Indeed, it is conceded that such a religious state is not always achieved in religious ritual: 'Of course, all cultural performances are not religious performances, and the line between those that are, and artistic, or even political, ones is often not so easy to draw in practice, for, like social forms, symbolic forms can serve multiple purposes' (1973: 113). But what is it that ensures the participant's taking the symbolic forms in the way that leads to faith if the line between religious and non-religious perspectives is not so easy to draw? Presumably the ability and the will to adopt a religious standpoint must be there prior to the ritual performance, so that a simple stimulus-response model of how ritual works will not do. And if that is the case, then ritual cannot be the place where 'religious faith' is attained, but where it is (literally) played out and confirmed. Again, if we are to understand how this happens, we must examine not only the ritual itself, but the entire range of available disciplinary activities, of institutional forms of knowledge and practice, within which selves are formed, and the possibilities of 'attaining to faith' are marked out. In other words, for the anthropologist to explain 'faith' must be primarily a matter of describing a dependence on authoritative practices and discourses, and not of intuiting a mental state lying beyond them said to be caused by ritual. We have noted more than once Geertz's concern to identify religious symbols according to universal criteria, and to distinguish the religious perspective clearly from non-religious ones. It would appear that the separation of religion from science, common sense, aesthetics, politics and so on, enables him to defend it against charges of'illusion', 'illogicality' and the like. If religion has a distinctive perspective, which does not in essence compete with others, and which furthermore performs a function always both necessary and unique, it cannot be accused of generating 'false consciousness'. Yet in a way this defence is equivocal. Religious symbols create dispositions, he observes, which 'seem' uniquely realistic. Is this the point of view of a reasonably confident agent (who must always operate within the denseness of historically given probabilities), or that of a sceptical observer (who can 'see through' the representations of reality to the reality itself)? It is never clear. And it is never clear because Geertz never examines whether, and if so to what extent and in what ways, religious 'experience' relates to something in the social world believers live in. This omission is related to his treatment of religious symbols as sui generis, the precondition for religious experience (which, once registered, must by definition be 'genuine'), rather than a condition of social life (facilitating some objectives and making others difficult). Towards the end of his essay, Geertz attempts to connect, instead of merely separating, the religious perspective and the common-sense one—and the result reveals an ambiguity basic to his entire approach. First, drawing on Schutz, Geertz states that the everyday world of common-sense objects and practical acts is common to all human beings because their survival depends on it: 'A man, even large groups of men, may be aesthetically insensitive, religiously This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions 250 TALAL ASAD unconcerned, and unequipped to pursue formal scientific analysis, but he cannot be completely lacking in common sense and survive' (1973: 119). Next, he informs us that individuals move 'back and forth between the religious perspective and the common-sense perspective' (1973: 119). These perspectives are so utterly different, he declares, that only 'Kierkegaardian leaps' (1973: 120) can cover the cultural gaps that separate them. Finally, Geertz concludes that 'Having ritually "lept" . . . into the framework of meaning which religious conceptions define, and the ritual ended, returned again to the common-sense world, a man is—unless, as sometimes happens, the experience fails to register —changed. And as he is changed, so also is the common-sense world, for it is now seen as but the partial form of a wider reality which corrects and completes it' (1973: 122, emphasis added). This curious account of shifting perspectives and changing worlds is puzzling. It is not clear, for example, whether the religious framework and the common-sense world, between which the individual moves, are independent of him or not. Most of what Geertz has said at the beginning of his essay would imply that they are independent (cf. 1973:92), and his remark about common sense being vital to every man's survival also reinforces this reading. Yet it is also suggested that as the believer changes his perspective so he himself changes, and that as he changes so too is his common-sense world changed and corrected. So the latter, at any rate, is not independent of his moves. But it would appear from the account that the religious world is independent, since it is the source of experience for the believer, and through that experience, a source of change in the common-sense world: there is no suggestion anywhere that the religious world (or perspective) is ever affected by experience in the common-sense world. This last point is consistent with Geertz's entire approach, in which religious symbols are sui generis, but in the present context it presents the reader with a paradox: the world of common sense is always common to all human beings, and quite distinct from the religious world, which in turn differs from one group to another, as one culture differs from another; but experience of the religious world affects the common-sense world, and so the distinctiveness of the two kinds of world is modified, and the common-sense world comes to differ, from one group to another, as one culture differs from another.34 The paradox results from Geertz's adherence to a confused phenomenology in which 'reality' is at once the distance of an agent's social perspective from the Truth, measurable only by the privileged observer, and also the substantive knowledge of a socially constructed world available to both agent and observer, but to the latter only through the former. Perhaps we can learn something from this paradox which will help us evaluate Geertz's confident conclusion: 'The anthropological study of religion is therefore a two-stage operation: first, an analysis of the system of meanings embodied in the symbols which make up the religion proper, and, second, the relating of these systems to social-structural and psychological processes' (1973: 125). How sensible this sounds, yet how mistaken, surely, it is. If religious symbols are understood, on the analogy with words, as vehicles for meaning, can such This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions TALAL ASAD 251 meanings be established independently of the form of life in which they are used? If religious symbols are to be taken as the elements of a sacred text, can we know what they mean without regard to the social disciplines by which their correct reading is secured? If religious symbols are to be thought of as the patterns by which experience is organised, can we say much about that experience without considering how it comes to be formed? Even if it be claimed that what is experienced through religious symbols is not, in essence, the social world but the spiritual,35 is it possible to assert that conditions in the social world have nothing to do with making that kind of experience accessible? Is the concept of religious training entirely vacuous? The two stages which Geertz proposes are, I would suggest, one. Religious 'symbols'—whether one thinks of them in terms of communication or of cognition, of guiding action or of expressing emotion—cannot be understood independently of their relations with non-religious 'symbols' or of their articulation of social life in which work and power are always crucial. My argument, I must stress, is not just that religious symbols are intimately linked to social life (and so change with it), or that they usually support dominant power (and occasionally oppose it). It is that social disciplines are intrinsic to the field in which religious representations acquire their force and their truthfulness. From this it does not follow that the meanings of religious practices and utterances are to be sought in social phenomena, but only that their possibility and their authoritative status are to be explained as products of historically distinctive disciplines and forces. The anthropological student of religion should therefore begin from this point, and not, as Geertz does, from a notion of culture as an a priori totality of meanings, divorced from processes of formation and effects of power, hovering above social reality. For far too long the well-known but increasingly unsatisfactory distinction between technical (or instrumental) action and expressive (or symbolic) action has determined the major orientation of anthropological studies of religion.36 In itself religion has usually been conceived of as expressive, and the main concern has been to discover what, in particular and general ways, it signifies. This applies equally to the so-called Intellectualist and Symbolist writers, in so far as both are primarily concerned to enquire into statements about the world which the religious believer allegedly makes, whether directly or indirectly, whether in practices or in exegeses of those practices. Studies of the social functions of religion tend to be either reductionist ways of arriving at its meanings, or (more usefully) ways of describing social consequences which, although brought about by religious institutions, may also be secured by 'secular' ones. Religion itself is rarely approached in terms of 'technical action'—that is to say, the disciplining of the body, of speech,37 which is used to produce religion in its variety. Such disciplines are preconditions for specific forms of thought and action, but they must be taught and learnt, and are therefore themselves dependent on a range of social institutions and material conditions. Geertz's text has the great virtue of stating with subtlety a distinctive position on a difficult conceptual problem. It is part of the movement in anthropology which has for some years now been concerned to take the signifying systems of other cultures more seriously than was the case before—and also to employ a This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions 252 TALAL ASAD more complex model of the workings of language in the analysis of those cultures. Its strength lies in its attempt to bring together a wide range of important questions—having to do with cognition and communication, authority and disposition, practice and its representation—not often treated together in anthropological discussions of religion. But an overall weakness of Geertz's position seems to be the hiatus it accepts between (external) symbols and (internal) dispositions, which parallels the hiatus between 'cultural system' and 'social reality'. Through my exploration of his text I have come to the conclusion that perhaps the fruitful response to this would be not to attempt a bridge between the two, but to move entirely away from that notorious dualism. Instead of approaching religion with questions about the social meaning of doctrines and practices, or even about the psychological effects of symbols and rituals, let us begin by asking what are the historical conditions (movements, classes, institutions, ideologies) necessary for the existence of particular religious practices and discourses. In other words, let us ask: how does power create religion? To ask this question is to seek an answer in terms of the social disciplines and social forces which come together at particular historical moments, to make particular religious discourses, practices and spaces possible.38 What requires systematic investigation therefore are the ways in which, in each society, social disciplines produce and authorise knowledges, the ways in which selves are required to respond to those knowledges, the ways in which knowledges are accumulated and distributed. Universal definitions of religion hinder such investigations because and to the extent that they aim at identifying essences when we should be trying to explore concrete sets of historical relations and processes.39 NOTES I am grateful to John Dixon, Terry Johnson and Sami Zubaida for their critical comments. 1 For example, Morgan: 'If Tillich has defined culture in terms agreeable to good anthropology (of the kind demonstrated by Geertz) and if his definition of religion, though not substantially in anthropological terminology, is not antipathetic to anthropological method, then there is a beginning point for dialogue . . . that dialogic point is the concept of meaning in religion and culture' (1977: 371). 2 Geertz has attempted such a description in his elegantly-written Islam observed (1968), where he contrasts two 'classical' styles of Islam, the one Moroccan and the other Indonesian. In each case a representative historical figure provides the occasion for generalising about the parallelisms between a personality type, a style of faith, a form of culture. In the later chapters an attempt is made to describe more recent changes and conflicts, but this is difficult to do in terms of the original semantic framework, and so resort is had in an ad hoc manner to well-known orientalist commentators on contemporary Islam. 3 The Russian psychologist Vygotsky (1962) makes crucial analytical distinctions in the development of conceptual thought: heaps, complexes, pseudo-concepts and true concepts. Although, according to Vygotsky, these represent stages in the development of children's use of language, the earlier stages persist into adult life. 4 Cf. Collingwood (193 8: book 2) for a discussion of the integral connexion between thought and emotion, where it is argued that there is no such thing as a universal emotional function accompanying all conceptualisation/communication: every distinctive cognitive/communicative activity has its own specific emotional cast. If this view is valid, then the notion of a generalised 'religious emotion' (or 'mood') may be questioned. 5 The argument that symbols organise practice, and thus the structure of cognition, is central to This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions TALAL ASAD 253 Vygotsky's genetic psychology—see especially 'Tool and symbol in child development' in Vygots-ky (1978). A cognitive conception of symbols, as opposed to one concerned with the representation of meanings, has recently been revived by Sperber (1975). A similar view was taken much earlier by Lienhardt (1961). 6 'The history of the process of the internalization of social speech is also the history of the socialization of children's practical intellect' (Vygotsky 1978: 27): See also Luria & Yudovich (1959). 7 Readers familiar with Austin (1962) will recognise the connexions to which I here refer. They will also, I hope, note how unwise is the use of'symbol' which conflates a variety of quite distinct problems. 8 If we set aside Radcliffe-Brown's well-known preoccupation with social cohesion, we may recall that he too was concerned to specify certain kinds of psychological states said to be induced by religious symbols: 'Rites can be seen to be the regulated symbolic expressions of certain sentiments [which control the behaviour of the individual in his relation to others]. Rites can therefore be shown to have a specific social function when, and to the extent that, they have for their effect to regulate, maintain and transmit from one generation to another sentiments on which the constitution of society depends' (1952: 157). 9 Some ways in which 'symbolisation' (discourse) can disguise lack of distinctiveness is well brought out in Maclntyre's trenchant critique of contemporary Christian writers, where he argues that 'Christians behave like everyone else but use a different vocabulary in characterising their behaviour, and also to conceal their lack of distinctiveness' (1971: 24). 10 The phenomenon of declining church attendance in modern industrial society, and its progressive marginalisation (in Europe at least) to those sectors of the population not directly involved in the industrial work process, is an illustration of the argument that if we are looking for causal explanations in this area, then socio-economic conditions in general are clearly the independent variable and formal worship the dependent. See the interesting discussion in Luckman (1967: ch. 2). 11 This was why Augustine eventually came round to the view that insincere conversion was not a problem (Chadwick 1967: 222-4). 12 A modern theologian puts it: 'The difference between the professing, proclaiming and orientating way of speaking on the one hand, and descriptive speech on the other, is sometimes formulated as the difference between "speaking about" and "speaking to". As soon as these two ways of speaking are confused, the original and unique character of religious speech, so it is said, is corrupted so that reality-for-the-believer can no longer "appear" to him as it appears in professing speech' (Luijpen 1973: 90-1). 13 The series of booklets known as Penitential manuals, with the aid of which Christian discipline was imposed on Western Europe from roughly the fifth to the tenth centuries, contain much material on pagan practices penalised as un-Christian. So for example, 'The taking of vows or releasing from them by springs or trees or lattices, anywhere except in a church, and partaking of food or drink in these places sacred to the folk-deities, are offenses condemned' (quoted in McNeill 1933: 456). (For further details see McNeill and Gamer 1938.) At the same time Pope Gregory the Great A.D. 540-604) 'urged that the Church should take over old pagan temples and festivals and give them a Christian meaning' (Chadwick 1967: 254). The apparent inconsistency of these two attitudes (rejection or incorporation of pagan practices) is less important than the systematic exercise of Church authority by which meaning is assigned. 14 'On the one hand, then, bishops complained of crude and too-avid beliefs in unauthorized and unexamined wonders and miracles, while on the other theologians (possibly also these same bishops) tried to come to terms with the matter. Although they attempted to define miracle by appeals to universal natural law, such definitions were not entirely successful, and in specific, individual cases, common sense was a better guide than medieval cosmology. When papal commissioners sa' down to hear testimony about Thomas Cantilupe's miracles at London and Hereford in 1307 '.hey had in front of them a schedule of things to ask about such wondrous events: they wanted to know, for example, how the witness came to learn of the miracle, what words were used by those who prayed for the miracle, whether any herbs, stones, other natural or medicinal preparations or incantations had accompanied the miracle; the witness was expected to say something about the age and social situation of the person experiencing the miracle, where he came from and of what family; whether the witness knew the subject before as well as after the miracle, This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions 254 TALAL ASAD what illness was involved, how many days he had seen the ill person before the cure; whether the cure was complete and how long it took for completion. Of course witnesses were also asked in what year, month, day, place and in whose presence the wonderful event itself occurred' (Finucane 1977: 53)- 15 By being authorised, shrines in turn served to confirm ecclesiastical authority: 'The bishops of Western Europe came to orchestrate the cult of the saints in such a way as to base their power within the old Roman cities on these new "towns outside the town". Yet it was through a studiously articulated relationship with great shrines that lay at some distance from the city—St Peter's, on the Vatican Hill outside Rome, Saint Martin's, a little beyond the walls of Tours—that the bishops of the former cities of the Roman Empire rose to prominence in early medieval Europe' (Brown 1981: 8). 16 The life of St Antony by Athanasius was the model for medieval hagiographies, and the Antonine sequence of early life, crisis and conversion, probation and temptation, privation and renunciation, miraculous power, together with knowledge and authority, was reproduced again and again in that literature (Baker 1972: 41). 17 The Lateran Council of 1215 decreed that annual private confession should be mandatory for all Christians: 'Every fidelis of either sex shall after the attainment of years of discretion separately confess his sins with all fidelity to his own priest at least once in the year: and shall endeavour to fulfil the penance imposed upon him to the best of his ability, reverently receiving the sacrament of the Eucharist at least at Easter: unless it happen that by the counsel of his own priest for some reasonable cause, he hold that he should abstain for a time from the reception of the sacrament: otherwise let him during life be repelled from entering the church, and when dead let him lack Christian burial. Wherefore let this salutary statute be frequently published in the churches, lest any assume a veil of excuse in the blindness of ignorance' (quoted in Watkins 1920: 748-9). In spite of its profound historical importance, anthropologists of religion have not yet studied the role of private confession in the construction of distinctive types of spirituality. It was through this discipline, in the Middle Ages and later, that specific sins were created as experience—not merely in the confessional, but in everyday life—in acts committed or omitted, in words spoken or overheard, in the very contemplations of possibility entertained in everyday life. Thus out of the avoidance or the absolution of particular sins was constructed a large part of the spiritual life available to laymen. Manuals of confession multiplied the categories of sin, its conditions and its consequences with astonishing relentlessness. Confession itself was a continuous process, a pervasive network, in which the power of the Church produced not only the self-knowledge of the subject, but also the systematic knowledge of a Christian society. Manuals for the guidance of priest and penitent, ordinances of councils, decretals, treatises on canon law, summas, sermon literature, mystical writings—all fed into and grew on the practice of confession. (See Lea 1896.) 18 For a brief introduction to the varying reaction of ecclesiastical authorities to the Franciscans and the beguines, see Southern 1970: Chs. 6 & 7. 'Beguines' was the name given to groups of celibate women dedicated to the religious life but not owing obedience to ecclesiastical authority. They flourished in the towns of western Germany and the Low Countries, were criticised, denounced and finally suppressed in the early fifteenth century. 19 Thus Cyprian: 'If a man does not hold this unity of the Church, does he believe himself to hold the faith? If a man withstands and resists the Church, is he confident that he is in the Cfiurch? For the blessed Apostle Paul has the same teaching, and sets forth the sacrament of unity, when he says, "There is one body, one Spirit, one hope of our calling, one Lord, one faith, one baptism, one God". This unity we ought firmly to hold and defend, especially we who preside in the Church as bishops that we may prove the episcopate also to be itself one and undivided. Let no one deceive the brethren by falsehood; let no one corrupt the truth of our faith by faithless transgression' (quoted in Bettenson 1956: 264). 20 The Church always exercised the authority to read Christian practice for its Religious Truth. In this context it is interesting that the word 'heresy' at first designated all kinds of errors, including errors 'unconsciously' involved in some activity (simoniaca haeresis), and it acquired its specific modern meaning (the verbal formulation of denial or doubt of any defined doctrine of the Catholic Church) only in the course of the methodological controversies of the sixteenth century (Chenu 1968: 276). 21 In the early Middle Ages, monastic discipline was the principal basis of religiosity. Knowles This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions TALAL ASAD 255 (1950: 3) observes that from roughly the sixth to the twelfth centuries 'monastic life based on the Rule of St Benedict was everywhere the norm and exercised from time to time a paramount influence on the spiritual, intellectual, liturgical and apostolic life of the Western Church . . . the only type of religious life available in the countries concerned was monastic, and the only monastic code was the Rule of St Benedict'. During this period the very term 'religious' was therefore reserved for those living in monastic communities; with the later emergence of non-monastic orders, the term came to be used for all who had taken life-long vows by which they were set apart from the ordinary members of the Church (Southern 1970: 214). The extension and simultaneous transformation of the religious disciplines to lay sections of society from the twelfth century onwards (Chenu 1968) contributed to the Church's authority becoming more pervasive, more complex and more contradictory than before—and so too the articulation of the concept and practice of lay 'religion'. For an illuminating sketch of the history of the concept and practice of religious discipline in early Christian and medieval periods see Leclercq 1957: 1291-302. 22 When Christian missionaries found themselves in culturally unfamiliar territory, the problem of identifying 'religion' became a matter of considerable theoretical difficulty and practical importance. For example, 'The Jesuits in China contended that the reverence for ancestors was a social, not a religious, act, or that if religious, it was hardly different from Catholic prayers for the dead. They wished the Chinese to regard Christianity, not as a replacement, not as a new religion, but as the highest fulfilment of their finest aspirations. But to their opponents the Jesuits appeared to be merely lax. In 1631 a Franciscan and a Dominican from the Spanish zone of Manila travelled (illegally, from the Portuguese viewpoint) to Peking and found that to translate the word mass, thejesuit catechism used the character tsi, which was the Chinese description of the ceremonies of ancestor-worship. One night they went in disguise to such a ceremony, observed Chinese Christians participating and were scandalized at what they saw. So began the quarrel of "the rites", which plagued the eastern missions for a century and more' (Chadwick 1964: 338). Anthropologists today are less conscious of such difficulties perhaps because not much of practical consequence hangs on their definitions of exotic religion. 23 Phases in the gradual evacuation of specificity from public religious discourse in the eighteenth century are described in Gay 1970. 24 The results of equating these two discursive levels seem to be connected to Geertz's ambiguous conception of symbol discussed earlier, in which an object's existence and its representation are equally identified as the symbol's 'meaning'. Crapanzano has made a similar criticism in his review article on C. Geertz, H. Geertz, and L. Rosen, Meaning and order in Moroccan society: 'Although it is impossible ... to treat adequately the problems raised by Rosen's work and the work of the other contributors . . . , I would like to suggest that the problems find their source in a failure to differentiate carefully the referential and indexical functions of symbols (I use 'symbols' here loosely as a short-hand) and, as a corollary, to distinguish between talking about symbols and talking about the use of symbols' (1981: 854). 25 The way in which representations of occurrences were transformed into meanings by Christian theology is analysed by Auerbach in his classic study of representations of reality in Western literature—briefly summed up in this early passage: 'The total content of the sacred writings was placed in an exegetic context which often removed the thing told very far from its sensory base, in that the reader or listener was forced to turn his attention away from the sensory occurrence and toward its meaning. This implied the danger that the visual element of the occurrences might succumb under the dense texture of meanings. Let one example stand for many: It is a visually dramatic occurrence that God made Eve, the first woman, from Adam's rib while Adam lay asleep; so too is it that a soldier pierced Jesus' side, as he hung dead on the cross, so that blood and water flowed out. But when these two occurrences are exegetically interrelated in the doctrine that Adam's sleep is a figure of Christ's death-sleep; that, as from the wound in Adam's side mankind's primordial mother after the flesh, Eve, was born, so from the wound in Christ's side was born the mother of all men after the spirit, the Church (blood and water are sacramental symbols)—then the sensory occurrence pales before the power of the figural meaning. What is perceived by the hearer or reader ... is weak as a sensory impression, and all one's interest is directed toward the context of meanings. In comparison, the Greco-Roman specimens of realistic presentation are, though less serious and fraught with problems and far more limited in their conception of historical movement, nevertheless perfectly integrated in their sensory substance. This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions 256 TALAL ASAD They do not know the antagonism between sensory appearance and meaning, an antagonism which permeates the early, and indeed the whole, Christian view of reality'(1953:48-9). AsAuerbach goes on to demonstrate, Christian theory in the later Middle Ages invested representations of everyday life with characteristic figural meanings, and so with the possibilities for distinctive kinds of religious experience. Figural interpretation, in Auerbach's usage, is not synonymous with symbolism. The latter is close to allegory, in which the symbol is substituted for the object symbolised. In figural interpretation the representation of an event (Adam's sleep) is made explicit by the representation of another event (Christ's death) which is its meaning. The latter representation fulfils the former (the technical term, Auerbach tells us, was figurám implire)—it is implicit in it. 26 Cf. Douglas (1975: 76): 'The person without religion would be the person content to do without explanations of certain kinds, or content to behave in society without a single unifying principle validating the social order'. 27 When the fifth-century bishop of Javols spread Christianity into the Auvergne, he found the peasants 'celebrating a three-day festival with offerings on the edge of a marsh.. . . "Nulla est religio instagno", he said: There can be no religion in a swamp' (Brown 1981:125). For medieval Christians religion was not a universal phenomenon: religion was the site on which universal Truth was produced, and it was clear to them that Truth was not produced universally. 28 Thus Auerbach (1953: 555) writes: 'The view of reality expressed in the Christian works of late antiquity and the Middle Ages differs completely from that of modern realism'. But as he is careful to point out, this does not mean that there is nothing in common between medieval and modern views, but that what is common is differently arranged. 29 The assumption that 'belief is a distinctive mental state characteristic of all religions has been the subject of recent discussion. Thus Needham (1972) has argued that belief is nowhere a distinct mode of consciousness, nor a necessary institution for the conduct of social life. Southwold (1979) takes an almost diametrically opposed view, asserting that questions of belief do relate to distinctive mental states, and that they are relevant in any and every society, since 'to believe' always designates a relation between a believer and a proposition and through it to reality. Harré (1981: 82), in a criticism of Needham, makes the more interesting observation that 'belief is a mental state, a grounded disposition, but is confined to people who have certain social institutions and practices'. 30 It is startling to see 'the scientific perspective' summed up thus in a line and a half. Philosophical attempts to define 'science' have not reached a firm consensus. In the Anglo-Saxon world recent arguments have been formulated in and around the works of Popper, Kuhn, Lakatos, Feyerabend; in France, that of Bachelard and Canguilhem. One important tendency has been to abandon the attempt at solving what is known in the literature as 'the demarcation problem' which is based on the assumption that there must be a single, essential, scientific method. Geertz's suggestion that the scientist 'dissolves the world's givenness into a swirl of probabilistic hypotheses' is as questionable as the complementary suggestion that in religion there is no scope for experimentation. On this latter point there is massive evidence against Geertz, even if we confine ourselves to the history of Christian asceticism. Equally, the suggestion that 'art' is a matter of 'effecting a disengagement from the whole question of factuality, deliberately manufacturing an air of semblence and illusion' would not be taken as a self-evident truth by all writers and artists. For example, when the art-critic Berger argues, in his brilliant essay 'The moment of cubism', that cubism 'changed the nature of the relationship between the painted image and reality, and by so doing expressed a new relationship between man and reality' (1972: 145), we learn something about cubism's concern to re-define visual factuality. How do Geertz's phrases about 'the aesthetic perspective' relate to such concerns? 31 The ways in which the concept of'art'—and therefore the practice of art—has been progressively transformed since classical times is briefly indicated in the Introduction to Collingwood (1938). Breaks in discursive forms of representation are analysed at length in Auerbach (1953). For a fascinating argument concerning transformations in the nature of knowledge since the seventeenth century, see Hacking (1975). 32 In case some readers are tempted to think that what I am talking about here is not 'science' (knowledge) but 'technology' (practical application), whereas Geertz is concerned only with the former, I would stress that any attempt to make a sharp distinction between the two is based on a naive view of the historical practice of both (cf. Musson & Robertson 1969). My point is that both This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions TALAL ASAD 257 'science' and 'technology' together are basic to the structure of modern lives, individual and collective, and that 'religion' in any but the most vacuous sense is not. 33 Vygotsky 1978: 58-9. 34 A similar criticism was made by Wood (1981) of Horton. 35 Cf. the final chapter in Evans-Pritchard (1956), and also the Conclusion to Evans-Pritchard (1965). 36 One of the earliest formulations of this distinction in modern social anthropology was Radcliffe-Brown's—as for example in this passage (1939): 'The very common tendency to look for the explanation of ritual actions in their purpose is the result of a false assimilation of them to what may be called technical acts. In any technical activity an adequate statement of the purpose of any particular act or series of acts constitutes by itself a sufficient explanation. But ritual acts differ from technical acts in having in all instances some expressive or symbolic element in them' (1952: 143). Beattie (1964: 202) put it this way: 'Now the chief difference between what we call practical, common-sense techniques for doing things, and ritual or "magico-religious" ways of doing them lies basically in the presence or absence of an institutionalised symbolic element in what is done'. More recently, Leach (1976: 9) has included the distinction within a tripartite scheme: 'We can usefully distinguish three aspects of human behaviour: (1) natural biological activities of the human body—breathing, heartbeat, metabolic process and so on; (2) technical actions, which serve to alter the physical state of the world out there—digging a hole in the ground, boiling an egg; (3) expressive actions, which either simply say something about the state of the world as it is, or else purport to alter it by metaphysical means'. It will be clear where religion fits into this scheme, and how it is to be studied. 37 Since Mauss's famous essay on 'Techniques of the body', there has been, until very recently, little interest in culturally acquired techniques of the body. Some valuable work is contained in Blacking 1977 and Polhemus 1978. Most of this work, however, is focused on the body as a medium of communication, and not as a medium of practical discipline. Techniques of the body are not merely culturally determined 'symbols', they are also essential practical means for achieving social and psychological objectives. Particular disciplines of the body (culturally acquired physical potentialities) make it possible to perform a particular kind of dance, military manoeuvre, industrial work process or ascetic exercise which would not otherwise be possible. It is not what they mean but what they enable that is relevant here. This applies also to speech techniques, in so far as they are among the pre-conditions for producing legal, political, religious, aesthetic, educational, effects. In my view anthropologists whose work is directly concerned with 'culture and symbolism' will not be able to appreciate the full implications of this point unless they begin to take the problem of historical transformations seriously. 38 Foucault's pioneering work on power and knowledge is, of course, directly relevant to such a project. 39 Dumont has perhaps come nearest to questioning the feasibility of a universal definition of religion: T shall take for granted that a change in relations entails a change in whatever is related. If throughout our history religion has developed (to a large extent, with some other influences at play) a revolution in social values and has given birth by scissiparity, as it were, to an autonomous world of political institutions and speculations, then surely religion itself will have changed in the process. Of some important and visible changes we are aware, but, I submit, we are not aware of the change in the very nature of religion as lived by any given individual, say a Catholic. Everyone knows that religion was formerly a matter of the group and has become a matter of the individual (in principle, and in practice at least in many environments and situations). But if we go on and assert that this change is correlated with the birth of the modern State, the proposition is not such a commonplace as the previous one. Let us go a little further: medieval religion was a great cloak—I am thinking of the Mantle of Our Lady of Mercy. Once it became an individual affair, it lost its all-embracing capacity and became one among other apparently equal considerations, of which the political was the first born. Each individual may, of course, and perhaps even will, recognise religion (or philosophy), as the same all-embracing consideration as it used to be socially. Yet on the level of social consensus or ideology, the same person will switch to a different configuration of values in which autonomous values (religious, political, etc.) are seemingly juxtaposed, much as individuals are juxtaposed in society' (1970: 32). Dumont has a subtle sense of the mutation of religion in modern Western history, and the implications of this fact for cross-cultural conceptions of religion. But I would This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions 258 TALAL ASAD suggest that the emphasis should be not on 'configurations of values' but on practices, powers, discourses, otherwise there is always the danger that (a) we may assume a consensus at a level where there is not one, (b) we may forget that the modern state is compulsive in a way that modern religion is not, and (c) we may think that an essentialist definition of religion in terms of 'meanings' is a plausible enterprise. Indeed, Dumont at times seems to be proposing that a universal definition, excluding only modern Christianity, is possible in terms of a specifiable configuration of values. REFERENCES Auerbach, E. 1953. Mimesis. Princeton: Univ. Press. Austin, J. L. 1962. How to do things with words. Oxford: Clarendon Press. Baker, D. 1972. VirDei: secular sanctity in the early tenth century. In Popular belief and practice (eds) C. J. Cuming & D. Baker. Cambridge: Univ. Press. Beattie.J. 1964. Other cultures. London: Cohen & West. Berger, J. 1972. Selected essays and articles. Harmondsworth: Penguin. Bettenson, H. (ed.) 1956. The early Christian Fathers. London: Oxford Univ. Press. Blacking, J. 1977. The anthropology of the body. London: Academic Press. Brown, P. 1967. Augustine of Hippo. London: Faber & Faber. -1981. The cult of the saints: its rise andfunction in Latin Christianity. London: SCM Press. Chadwick, H. 1967. The early Church. Harmondsworth: Penguin. Chadwick, O. 1964. The Reformation. Harmondsworth: Penguin. Chenu, M-D. 1968. Nature, man, and society in the twelfth century: essays on new theological perspectives in the Latin West. Chicago: Univ. Press. Collingwood, R. G. 1938. The principles of art. London: Oxford Univ. Press. Crapanzano, V. 1981. Review of C. Geertz, H. Geertz, andL. Rosen, Meaning and order in Moroccan society. Econ. Develop, cult. Change29, 849-60. Douglas, M. 1975. Implicit meanings. London: Routledge & Kegan Paul. Dumont, L. 1971. Religion, politics, and society in the individualistic universe. Proc. R. anthrop. Inst. 1970, 31-41. Evans-Pritchard, E. E. 1956. Nuer religion. Oxford: Clarendon Press. -1965. Theories of primitive religion. Oxford." Clarendon Press. Finucane, R. C. 1977. Miracles and pilgrims: popular beliefs in medieval England. London: Dent. Gay, P. 1970. The Enlightenment. 2 vols. London: Weidenfeld & Nicolson. Geertz, C. 1968. Islam observed. New Haven: Yale Univ. Press. -1973. The interpretation of cultures. New York: Basic Books. Hacking, I. 1975. Why does language matter to philosophy? Cambridge: Univ. Press. Harre, R. 1981. Psychological variety. In Indigenous psychologies (eds) P. Heelas & A. Lock. London: Academic Press. Knowles, D. 1950. The monastic order in England. Cambridge: Univ. Press. Lea, H. C. 1896. A history of auricular confession and indulgences in the Latin Church. 3 vols. Philadelphia. Leach, E. 1976. Culture and communication. Cambridge: Univ. Press. Leclercq, J. 1957. Disciplina. In Dictionnaire de spirituality vol. 3. Paris: Beauchesne. Lienhardt, G. 1961. Divinity and experience. Oxford: Clarendon Press. Luckman, T. 1967. The invisible religion. New York: Macmillan. Luijpen, W. A. 1973. Theology as anthropology. Pittsburgh: Duquesne Univ. Press. Luria, A. R. & F.Ia. Yudovich 1971. Speech and the development of mental processes in the child. Harmondsworth: Penguin Books. Maclntyre, A. 1971. Against the self-images of the age. London: Duckworth. Mauss, M. 1973 (1935). Techniques of the body. Economy & Society, 2, 70-88. McNeill, J. T. 1933. Folk-Paganism in the Penitentials. J. Relig., 13, 450-66. -&H. M. Gamer, (eds) 1938. Medieval handbooks ojpenance. New York: Columbia Univ. Press. Morgan, J. 1977. Religion and culture as meaning systems: a dialogue between Geertz and Tillich. J. Relig. 57- This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions TALAL ASAD 259 Musson, A. E. &E. Robinson 1969. Science and technology in the Industrial Revolution. Manchester: Univ. Press. Needham, R. 1972. Belief, language, and experience. Oxford: Basil Blackwell. Polhemus, T. (ed.) 1978. Social aspects of the human body. Harmondsworth: Penguin Books. Radcliffe-Brown, A. R. 1952. Structure and function in primitive society. London: Cohen and West. Southern, R. W. 1970. Western society and the Church in the Middle Ages. Harmondsworth: Penguin Books. Southwold, M. 1979. Religious belief. Man (N.S.) 14, 628-44. Sperber, D. 1975. Rethinking Symbolism. Cambridge: Univ. Press. Vygotsky, L. S. 1962 (1934). Thought and language, Cambridge, Mass.: MIT Press. -1978. Mind in society. Cambridge, Mass.: Harvard Univ. Press. Watkins, O. D. 1920. A history of penance, 2 vols. London: Longmans. Willey, B. 1934. The seventeenth-century background. London: Chatto and Windus. Wood, G. 1981. Empirical philosophies of mind. Economy & Society, 10, 125-65. This content downloaded from 147.251.102.162 on Fri, 12 Feb 2016 07:54:40 UTC All use subject to JSTOR Terms and Conditions