4—
Characteristics of Religion

As for those wingy mysteries in Divinity and ayery subtilties in Religion, which have unhindg'd the braines of better heads. . . .
Sir Thomas Browne

I

There is no word common to the Indo-European languages which can be translated as "religion." Nor is there any word in either Greek or Latin corresponding exactly to the English word. In Latin alone the etymology of "religion" is disputable: the word is traced from ligare, to tie, or else from legere, to collect. If we are persuaded by Benveniste to opt for the latter construction, we have "recollect," in the sense of to go back on a previous step and to make a new choice; for example, as in being religiosus, checked by scruple in the performance of rites. What is indicated is an inner state, not an objective property of certain things or a set of beliefs and practices.

This doubtful quality of troubled conscience, even if it were thought externally discernible, could hardly make a secure foundation for comparison; and its contingent association with historical, local, and changing forms of action makes it the less reliable as the mark of any distinct and universal institution. The concept of religion is thus easily susceptible to being analyzed away. Similarly, the designa-

tion of a particular religion may, by various use and dissent, become undiscriminating; hence Sir Thomas Browne, in 1643, observed that "the name of a Christian is become too generall to expresse our faith"; and it has grown into a common expectation that religion connotes faction and fission and ceaseless reformulation. There are indeed many grounds on which to conclude that the notion of "religion" is altogether too polysemous, indistinct, and malleable to serve any steady analytical purpose.

Nevertheless, religion has obsessed western consciousness for centuries, and in other civilizations also there are concerns which in one regard or another do appear significantly comparable. If we cannot rely on the concept for any precise task of comparison or interpretation, it still possesses certain odd-job connotations that make it somewhat useful in the preliminary assortment of social facts and in general descriptions. To take the word "religion" very strictly, and to demand of it what we have no reason to think it can afford, would be self-defeating; so let us take it as we use it, and then try to establish what characteristic features give it that use.

To that end I shall resort at the limit to the entire span of world ethnography, though I shall have in mind especially those parts of Southeast Asia that I know better than other regions: these will include the interior of Borneo, the island of Sumba in eastern Indonesia, and more widely the sphere of Malay civilization. Inevitably, on such a scale, my observations will be highly general and impressionistic; I shall be deliberately concerned not to be swayed by any taxonomy of religious phenomena, and at the same time I shall suspend attention to the exceptions that can always be found to challenge generalizations of the kind. From these vantage points, and against the personal background of an intensely

Christian upbringing in the High Church of England, I shall see what I can do toward isolating what can be conceived as ultimate predicates in the comparative study of religion. That is, I shall be trying to determine—in the train of countless other investigators—what we are really talking about when we speak of the religion of others.

This undertaking may at first seem to run the danger of turning into yet another definition of religion; but I have already abjured that aim, and as we proceed it will be seen that the course of the exposition actually leads away from any fixed stance of that nature. The global purview of the comparativist militates against any categorical certainty, and the precept to take nothing for granted conduces instead to skeptical detachment.

II

Perhaps the first presumption to offer itself is that religion, whatever else it may involve, has to do with the metaphysical, that is, with a doctrine of first principles.

We may take this, with Kant, as meaning an isolated speculative science of reason; and all the interests of our reason, he asserts, combine in the following three questions: (1) What can I know? (2) What ought I to do? (3) What may I hope? Without quibbling over alternative definitions of metaphysics, we can at least make a rational start by asking to what extent religion can be expected to provide answers to these questions.

The first question—What can I know?—is purely speculative, and I think it is a common experience of ethnographers to find that exotic religions are not metaphysical in this sense. In nonliterate societies especially the religion may well provide a cosmology, a general scheme of how things are, but it cannot be counted on to provide a critical appa-

ratus for epistemological speculation. Usually, I am sure, there will be individuals who are of an analytical or speculative turn of mind; but this is a temperamental inclination and not a response to intellectual demands of their religion or to problems of knowledge to which the religion can provide the answers. For the outsider, however, there is the temptation to presume that the religion must have this epistemological character, because its tenets can be represented as a set of propositions, and the propositions can be interpreted as answers to questions. No doubt an interrogatory urge did contribute to the formulation of the tenets, and perhaps also these tenets will on occasion provide answers in the minds of latter-day adherents, but these epistemological functions need not characterize the religion in its ordinary practice.

A difficulty in this connection is that the ethnographer is not genuinely a participant in this ordinary practice, but an inquirer; he does not entirely understand what is going on, or how the tenets are systematically connected, so all the time he is asking questions. It is practically certain that many of the questions he asks will never have been posed by the adherents, since the intellectual premises of the latter are not those of the inquirer; and if the adherents can nonetheless supply some answers to such questions, this does not entail that the tenets of the religion—that is, its ideological resources—were contrived as the resolutions to speculative concerns. The likelihood, indeed, is that if the inquirer were not present the religion would go unquestioned; for tenets are to be held to, not queried.

This does not mean, though, that religious tenets are regarded by the adherents as knowledge. I assume that any language contains some means of expression equivalent to the verb "to know," and that all languages provide for the

expression of degrees of certainty in what is known; but what seems very doubtful is that the tenets of religion are generally said by their proponents to be known. On this point, in fact, I cannot cite any evidence, outside certain deliberately theological cultures, that people claim to "know" what their religions teach and to which they give their assent. If this is a correct impression, it is probably a reflection of the nonspeculative character of such religions; and if the tenets are not challenged, there is no need to claim to know them. At the same time, the fact that religions are products of change, and are liable perpetually to further change, does not entail that the modification or the abandonment of their tenets is the result of deliberate cogitation about what is known. There are many ways in which ideas can change without being subjected to cognitive criticism.

Correspondingly, it is not a general characteristic of religions that their tenets are held to be true. Even in India, which has an advanced traditional theory of knowledge, it is said (by Renou, I think) that it is very hard to be heretical; and Dumézil has remarked that India is the one great civilization in which contradiction is not a mortal sin. A dogmatic arrogation of truth is certainly a prominent feature of some religions and sects, but on a global scale I do not think that religions are spontaneously asserted to be true; and if under a special stimulus it is claimed that a religion is true, this need not imply that the tenets of other religions are not true. My own experience with adherents of nonproselytizing religions has been that people readily admit of others that their gods are different (or however the contrast may be phrased), just as their rules of marriage and land tenure may differ. This attitude, once again, is an aspect of an uncritical cast of mind toward what, for the most part, is hardly a

distinct object of thought in any case. And by uncritical, let me quickly be plain, I do not mean unintelligent or deficient in judgment, but merely that by cultural convention certain ideas and practices are not distinguished as matters for questioning.

Among certain Kenyah whom I used to know, it was a striking feature of their idiom that they constantly accused others of lying. Sometimes the allegation was a standard form of disparagement, sometimes it was a joke, and sometimes (it seemed to me) it was no more than one ready means of saying that members of another tribe or longhouse were different. But despite this rhetorical inclination, they never in my hearing—and not even when we were comparing customs—said of the religious tenets of others that they were untrue. (This was before the Kenyah were converted to Christianity by missionaries of rival faiths.) Similarly, among the Penan a word translatable as "true" was commonly resorted to in assessing the accuracy of statements, the genuineness of an instance, or as the intensifier of an attribute; but it would not have made sense to ask if their religious tenets were true or to suggest that those of other peoples were not true. This attitude of mind is not peculiar to such peoples, that is, to nonliterate civilizations lacking a critical intellectual tradition. In Confucian China, so far as I know, it would not have been said of ritual precepts and the spiritual personages to whom they were directed that they were or were not true objects of assent and adherence. It is this abstention from such judgments which permits a Buddhist in Sri Lanka both to resort to the orthodox institutions and practices of Hinayana and also to make sacrifices to individual spirits or seek aid from diviners and exorcists. If what is proclaimed by doctrine is true, as we should assess

it, then ideas that are inconsistent with the doctrine are false; but this logical kind of comparison, by reference to truth values, is not essential to religion.

At this point we come naturally to the problem of belief, that is, to the disposition or attitude of mind which by most western definitions is held to be essential with regard to religious tenets. I have already set out at some length, in a monograph on belief and the language of experience, what I think about this supposed capacity or inner state, so I shall be brief on this score. The crucial point is that there are numerous linguistic traditions which make no provision for the expression of belief and which do not recognize such a condition in their psychological assessments. The notion of belief is a relatively modern linguistic invention, and it does not correspond, under any aspect, to a real, constant, and distinct resource of the self. To adopt Hume's words, it is an artificial contrivance for the convenience and advantage of society, in a particular course of civilization, and as such it does not constitute a natural resemblance among men.

This argument has been found hard by some, scandalous by others, and at least awkward by yet others, which is not surprising, for it strikes against some of the strongest and most fervent assumptions in modern Christianity and in western philosophy. Yet at the same time it is no more than a conclusion to a long line of skeptical considerations stretching from Hume to Russell. Moreover, although my argument was an original construction on new foundations (namely, the empirical resources of comparativism), I have latterly discovered that one of its central contentions had been advanced by I. A. Richards in the columns of an intellectual weekly published in 1934. For Richards, the problem of the nature of belief dissolves into "a cluster of undeveloped problems," and this traditional landmark appears as

merely a fiction. Additional unsuspected support, to a similar effect, is to be had from Waismann's paper "Belief and Knowledge," composed in the 1950s but not published until 1977. And in this latter year a theologian, Wilfred Cantwell Smith, proposed (being unaware of my prior study) the radical thesis that the concept of believing is not to be found in the Bible; "the idea that believing is religiously important," he writes, "turns out to be a modern idea." More generally, Ayer has concluded, in a criticism of Hume, that "the problem of giving an analysis of belief which is neither trivial nor circular still awaits solution."

There is useful backing, therefore, to the idea that belief cannot have the psychological standing that it has been ascribed in the study of religion. My own analysis, at any rate, which I take to be neither trivial nor circular, has not been refuted by counterargument; and with the development of current work on indigenous psychologies it seems that the familiar conceptual landmark of "belief" has at the very least had its reliability much reduced.

Nevertheless, there remain two phenomena of adherence to the tenets or the practices of a religion which seem fundamental, even if these also are hard to account for. One is what we can for the present call "commitment," namely, whatever it is whereby we subscribe, with variable intensity, to an object of thought or imagination. This is in part an emotional or temperamental inclination, and it is implicated in one or another of the numerous senses of the verb "believe," but these facts do little or nothing to make the psychology of commitment comprehensible. There is a curious psychic detail, however, which may help to render it more immediate as an object of speculation and introspection.

When we read a novel that we have read before, or watch

a film that we have already seen, we know in advance what will be the outcome of the climactic points in the evolution of the plot; yet it is common, I think, that our breath is bated at such points, all the same, and that we are tense with all the inner sentiments of uncertainty. When Hamlet comes upon Claudius at his prayers, our prior knowledge that he will not act out his fervent threats does nothing to reduce the drama; and we can actually hope that he will do so, when we know perfectly well that he will not. When we reflect at the time upon the fact that we can indeed foresee the resolution of a fictional climax, even this realization does not subvert that response to imaginary events which is, I suggest, intrinsically indistinguishable from what we distinguish as religious commitment.

The other phenomenon, related but distinct, is that of "conversion," namely, the process of detachment from one object, such as a religious doctrine, and surrender to another. Despite the fact that this change is what proselytizing religions deliberately work for, and notwithstanding the many autobiographical narrations of such change, the psychology of conversion remains (so far as I have been able to discover) almost wholly obscure. No doubt conversion is facilitated in those societies which do not estimate religious tenets as true objects of knowledge, but this easiness does little if anything to clarify what happens in the conversion of persons for whom such conditions do not obtain.

Yet if there are ultimate predicates in the comparative study of religion, presumably commitment and conversion will find their places among them. Only of course they will not be exclusively connected with religion, for they are independent of the object of thought. They may indeed be characteristic features of religion, but they are not specific to it. Perhaps this fact in turn will tell us something about the

justification for delineating religion as a distinct object of attachment.

III

Kant's second metaphysical question—What ought I to do?—is in his view purely practical and hence within the scope of pure reason; but even so it is a moral question and not therefore a proper subject, he says, for a transcendental treatment.

As far as we are concerned, however, the central importance of the question is indeed practical in that we are interested directly in the objects of knowledge at issue; namely, in what we comprehensively allude to as religions, and it is a characteristic feature of these that they prescribe right conduct.

We can take it as given that any language will contain words for "good" and "bad," or equivalent means for expressing judgments of approval and disapproval, but religions differ considerably in the ways in which they promote the good and castigate the bad. Hocart, in an incidental observation, has pointed to a fundamental contrast: among most peoples, he writes, "the idea of evil, pure evil, is completely lacking." He does not elaborate, other than to say that Melanesians do not like death and sickness any more than we do, but his contention seems to be that whatever is classed as bad is not hypostasized into a concept or power of "evil." Certainly in any culture there will be an array of activities forbidden or disparaged as bad, but commonly there is not a generic or essential quality of which the activities are merely instances. This too is a reflection, I suppose, of styles of thought (including attitudes toward thought) different from our own tradition of critical abstraction. The hypostasis of evil is a particular cultural product, the result

in part of the Aristotelian tradition of formal logic whereby attributes are progressively subsumed under higher and higher genera. But a society that does not rely on this method of classification need not subscribe to a conception of the world which the method itself reflects.

It may be helpful, incidentally, to recall that the English word "evil" is derived from IE *up -, connoting up or over, and that it can be interpreted as "extreme," "excessive," or "exceeding due limits"—in other words, going too far. This is a long way from the moral category of quintessential evil or the personification of utter evil in the person of Satan; yet it is close, I think, to the way in which religions characteristically treat that which has to be shunned.

Similarly, in the Christian tradition we have hypostasized "sin" in a way unknown to the Hebrews, and we have thereby departed from a view of religious fault which remains characteristic of other religions. In Austronesian languages, for example, there is an extensive family of words related to *salaq and expressing most generally the idea of an error or mistake. Usually a word such as Malay salah or Penan sala' will be employed to castigate what is religiously forbidden, for example, incest or murder or the transgression of a ritual injunction. In Indonesian, it is only in a deliberately theological or analytical context that the word dosa, introduced from Sanskrit, will be used to express the idea of sin as an offense against religion or divinely sanctioned morality. Ordinarily, though, the response of alien religions to the question "What ought I to do?" is not "shun sin" but "avoid mistakes."

This injunction leads directly to the topic of ritual or, more generally, symbolic action. There are many theories of what ritual itself is about, apart from its special connection with religion, and for the sake of generality I shall adopt

Arthur Waley's stance. "The truth," he writes, "is that there is no 'real reason' for ritual acts"; and where such acts are not linked in the mind of the practitioner to any system of thought they will be explained as "customary" or as "the thing to do." So far as the evidence from ethnography goes, the rites of exotic religions seem to be marked by just this character in the minds of the participants. It may be that the celebrants (priests, or whatever they are to be called) have a clearer idea, or a more explicit rationalization; but those who simply attend need have no precise understanding of the ideology behind the rites, and certainly they need be in no particular inner state such as awe, self-abasement, or anticipation.

I think, also, that our appreciation of exotic ritual is probably misguided by our aesthetics. We are predisposed to picture a religious rite as a sequence of acts distinguished by solemnity, a gravity of demeanor, a careful concern for the meticulous performance of each item in the procedure; silence is important; and when movement from one place to another is called for this is done with measured tread, sometimes in the ordered dignity of a procession. Some of these impressive attributes are indeed found in the rites of certain other religions, for example, those of Bali, but they are not essential to religious performances and they are not characteristic of religion. In my own experience, rites can be muddled, gabbled, enacted out of due order; they can be perfunctory, intermittent, or incomplete; those who attend may or may not pay much attention, or they may pass the time in gossiping, quarreling, making deals or assignations, or in anything but a reverent attention to what is being carried out. So even if a rite is no more than what is done, the way in which it is done is not a characteristic feature of religion.

As for what is actually done, there is room for much dispute about whether such typical activities as "worship," "sacrifice," "prayer," and so on are distinct modes of symbolic action and peculiar to religion. A prudent conclusion is that in each case it can be shown by analysis and comparison that the religious institution is assimilable to some more general form of social action—except for the epithet "religious," and it is precisely this quality that we are trying to comprehend by reference to the rites.

IV

The last of Kant's three questions is What may I hope? This looks more in accord with the view that religion provides a metaphysics, but it is not directly eschatological.

Kant glosses the question as If I do what I ought to do, what may I then hope? All hoping, he writes, is directed to happiness; and he defines happiness as "the satisfaction of all our desires." He has more to say about the nature and conditions of happiness, but let us again take his question, so far explicated, as the starting point of a search for characteristic features of religion.

Hocart has argued at length, in his Kings and Councillors, that rituals and the religions of which they are expressions are communal undertakings in quest of life. Under this wide formulation, it can well be said that religion defines and fosters what is to be hoped for: life, fertility, abundance; freedom from want, illness, and pain; perhaps even immunity to death. In some instances it solves philosophical problems, settles moral quandaries, and is a source of inspiration and a sense of direction in life. In many such ways, it can be said, religion handsomely provides for the satisfaction of desires; moreover, in some cases it first defines the desire, for example, purity or salvation, and then provides for

its satisfaction, an expedient which is both attractive and efficient.

Nevertheless, there does not seem to be much enlightenment to be had from our third metaphysical question. It is not to be thought, after all, that anything we were prepared to call religion would grimly frustrate hope and prevent the satisfaction of decent desires. (That Buddhism advocates the extinction of desire is another matter.) Nor, on the other hand, can it be said that religion can be discriminated by the fact that it proffers hope, since other institutions can do that. And to accord it such a sentimental function would not in any case secure a guideline to the comprehension of religion.

V

The reason for approaching religion through Kant's three questions was that I did not want to risk a petitio principii by importing into my premises any consideration that would depend on a prior conception of what religion was or was not.

Since it seemed likely to be generally conceded that a religion was at least a metaphysics, a set of questions that were claimed to comprise all of man's speculative interests seemed to answer the purpose of the investigation. But of course it will be countered that religion is not entirely speculative, and that distinctive attributes have been left out of my account so far.

Certainly some attributes have been omitted, namely, the grounds of those media and connections that qualify a system of representations as symbolic. These attributes are not speculative, and indeed it has been taken as a peculiar characteristic of symbolism that although it is regular it is not rational. Under this aspect, then, it can be said that a religion

is at least a symbolic classification, and that in its iconic and evocative resources it is of a kind with other employments of symbolism.

In the first place, it is remarkable how generally in world ethnography we find, in what can be described as religions, a common and quite limited stock of symbolic resources: for example, the cross, water, blood, lateral values, cardinal points, elevation, scents, colors, and so on. In the second place, however, it is apparent that in these respects a religion is indeed a symbolic classification, and as such formally indiscriminable from other systems of the kind which are not readily describable as religions.

If the values of right and left are religiously important, as in circumambulation or ecclesiastical architecture or blessing, they are important also in black magic, the prediction of the sex of an unborn child, reaping a harvest, or conferring honor on a guest. Similarly, if certain prestations are invested with religious significance, as in sacrifice or the placation of ancestral spirits or as tokens of gratitude for harvest home, they figure also in marriage payments, fees to a medicine man, and fines to a village council. In fact, religious symbolism is, by a conventional definition, merely a subclass of symbolic classification, and it is to be interpreted in the same way as any other symbolic system such as that of prescriptive alliance or Confucian etiquette.

What might distinguish the symbolism of religion is the valuation placed on the symbols, as expressed through religious personages, rites, and intentions; but it is precisely the peculiar quality of religious values that we are trying to determine, and on present understanding they are not manifested in a distinct mode of symbolic classification. At the same time, there is no argument to make the hypothetical case that symbolism in general is itself a derivative, histori-

cally or intrinsically, of the symbolism of religion, however this may be separately conceived.

For that matter, it could instead be argued historically, where there is the evidence, that religion is a residual sphere of values within which are to be found symbolic resources which were originally far more pervasive, and that the employment of these resources did not depend on the discrimination of religion as a separate focus of social interest. In this regard, religion is closely comparable with, for example, the institution of monarchy. In both cases, we come to terms with the symbolism of the institution by means of a more extensive acquaintance with the vehicles and principles of symbolism in general, not by way of the particular values and purposes (religious or monarchical) in the service of which the resources of symbolism happen to have been contingently deployed.

Among the symbolic vehicles that constitute the limited repertory that I have just mentioned, there are some that have been claimed to have an archetypal character. More widely, it could be claimed that the entire repertory, such as can be established by means of universal comparison, is by this very fact to be considered composed of archetypes. The distribution and the constancy of these features testify to the spontaneous production and recognition of certain primordial images and subliminal connections ("secret sympathies," as these latter used to be called) which represent fundamental inclinations of the psyche. An alternative way to put the matter is to ascribe the archetypes to cerebrational vectors, that is, to locate their source in normal operations of the brain and thereby to posit a natural ground for collective representations which have a global distribution. Religions have their own explanations for the appearance in them of crosses, mountains, metamorphoses, and the rest;

and part of what gives a distinct character to religion is that the adherents are committed to just these explanations.

Borges writes that "perhaps universal history is the history of the diverse interpretation of a few metaphors." (The example that he traces is the image of God as "an intelligible sphere, whose centre is everywhere and whose circumference is nowhere.") Likewise, in the domain of practical reason, Kant sought what he called "principles of the possibility of experience." Perhaps, also, there are (as is claimed by depth psychologists) certain complexes which are paradigms of psychic experience; and it might be these to which the schemes of symbolism, religion, and psychology itself made their several responses.

At the level of events, there are incidents which make an awesome and poignant impact, as though they conveyed a mystagogical significance about life.

In the Gospels, Christ predicts that the cock shall not crow before Peter shall deny him thrice. After Christ has been arrested and taken to the house of the high priest, Peter is three times challenged as one of his followers, and three times he denies knowing him. "At the third time, while he yet spake, the cock crew." Then Luke (22:61) adds the terrible words: "And the Lord turned, and looked upon Peter." How little we need to be told that the disciple, going outside, "wept bitterly."

This kind of profound effect can, in another context, be an achievement of supreme art. In Anna Karenina , for instance, there is the scene in which Koznyshev and Varenka, on a mushroom-picking expedition in the woods, make themselves quite alone. He has already rehearsed in his mind the phrases in which he will declare his love and ask her to marry him; she has divined what it is he wants to say, and her heart grows faint with joy and panic. In a moment her

fate is to be decided. But some perverse reflection causes him to ask instead "What is the difference between a white boletus and a birch mushroom?" She trembles as she answers that there is hardly any difference. "As soon as these words were out of her mouth, both he and she understood that it was over, that what was to have been said would not be said."

In Kurosawa's film Seven Samurai there is a sequence in which the master swordsman, watched with trembling admiration by the neophyte warrior, sits calmly contemplating a flower as he waits for the bandits to return to their horses tethered nearby. One of them blunders onto the scene. The swordsman lightly springs to his feet, sets his legs firmly apart, and with his hand on the hilt of his sword sternly confronts the hapless villain. He does not draw his weapon in a hasty attack: he simply stands there, in a posture of tense challenge and dreadful composure. For a still moment he looks at his petrified opponent with the level appraisal of a superb technician. Then swiftly with perfect movements he draws and strikes.

These affecting depictions, far from being merely dramatic or aesthetic, are direct testimony to deep-flowing concerns which are to be found expressed in religion. They can be seen, in that connection, as corresponding to what Eliade calls "hierophanies," that is, manifestations of the sacred. But the impact of such paradigmatic scenes does not stem from the contestable notion of the sacred; they are exemplary in a far more immediate manner, and they are not to be explained by the idiocratic abstractions of religious terminology. In their archetypal nature they are recalcitrant to definition; this very feature is part of their immediacy, and it is consistent with the supposition that they have their grounds in the unconscious.

VI

If it is difficult to draw up a substantive account of what ultimately we are talking about when we speak of religion, it is easier to learn from global comparison what adjustments we need to make to our expectations.

I am suggesting that there are three main conceptual predispositions, among others, which tend to mislead us in coming to terms with exotic religions. One is our Aristotelian tradition of classification, which inclines us to seek a common essence among social phenomena ("religions") which may stand one to another in relations of only indirect and sporadic resemblance. Another is our stress on inner states, which inclines us to think that the adherents to a religion should be in a peculiar cast of mind. A third is our ceremonial aesthetics, which provides an inappropriate gauge of the external forms of religion.

This is a formidable apparatus of distortions, and unless we take deliberate stock of them it is unlikely that we shall attain a perspicuous view of characteristic features of any social facts that are designated as religions.

On the other hand, if we concentrate positively on symbolic proclivities and affective paradigms, whether in rites or art or inner scrutiny, it is hard to see that there is any need—under these aspects, and on the universal scale of comparativism—to isolate the religious from among the archetypal forms of human experience.