Navigation – Plan du site

Book Reviews - Methods and Principles

American Anthropologist (Washington, octobre-décembre 1915)
Alexander Aleksandrovich Goldenweiser

Note de la rédaction

Source primaire
Goldenweiser (Alexander Aleksandrovich), « Book Reviews - Methods and Principles - Les Formes élémentaires de la vie religieuse. Le système totémique en Australie. Émile Durkheim. Felix Alcan, Paris, 1912. 647 pp. and tribal map of Australia », American Anthropologist (Washington) New Series, 17 (4), Oct.-Dec., 1915, p. 719-735

Source(s) numérique(s) identifiée(s) :

Texte intégral

Les Formes élémentaires de la vie religieuse. Le système totémique en Australie. Emile Durkheim. Felix Alcan, Paris, 1912. 647 pp. and tribal map of Australia.

A contribution by Emile Durkheim always commands attention. His Les règles de la méthode sociologique, De la division du travail social, and Le Suicide have exercised an appreciable influence on sociological theory and are still remembered and read. As editor of L’Année sociologique, Durkheim deserves credit for a methodical and extensive survey of anthropological and sociological literature. In this task he was ably assisted by his disciples and sympathizers, Hubert, Mauss and others. It is to be regretted that this excellent annual has now gone out of existence, its place having been taken by a triennial publication supplemented by occasional monographs constituting a series of Travaux de L’Année sociologique, of which La vie religieuse is the fourth volume.

  • 1 [Émile Durkheim, « La prohibition de l’inceste et ses origines », L’Année sociologique, 1 (1896-189 (...)

As the title indicates, the work deals with Australian totemism, but is also meant as a general theoretical inquiry into the principles of religious experience. Durkheim is a veteran in Australian ethnology. It will be remembered that the first volume of L’Année sociologique (1896-1897) contained a study from his pen devoted to “La prohibition de l’inceste et ses origines.” Volume V (1900-1901) of the Annual contains another study, “Sur le totémisme”; and volume VIII (1903-1904) one on “L’organisation matrimoniale australienne.”[1] One need not therefore be surprised to find Durkheim’s latest work replete with abundant and carefully analyzed data. In this respect the volume compares most favorably with much of the hazy theorizing called forth in such profusion by Spencer and Gillen’s descriptive monographs. But Durkheim’s work contains, of course, much more than a merely descriptive study. He had a vision and he brings a message. To these we must now turn.

While a comprehensive analysis of all of Durkheim’s propositions is entirely beyond the scope of a review, his cardinal doctrines may be discussed under the headings of five theories: a theory of religion, a theory [719] of totemism, a theory of social control, a theory of ritual, and a theory of thought.

Theory of Religion.—Durkheim vigorously objects to the theories of religion which identify it with belief in God or in the supernatural. A belief in the supernatural presupposes the conception of a natural order. The savage has no such conception nor does he know of the supernatural. He does not wonder nor inquire, but accepts the events of life as a matter of course. The attempts to derive religion from dreams, reflections, echoes, shadows, etc., find as little favor with Durkheim. Is it conceivable, he exclaims, that religion, so powerful in its appeal, so weighty in its social consequences, should in the last analysis prove to be nothing but an illusion, a naive aberration of the primitive mind? Surely, that cannot be. At the root of religion there must lie some fact of nature or of experience, as powerful in its human appeal and as universal as religion itself. Durkheim sets out in search of that fact. Presently, the field of inquiry is limited by the reflection that the beings, objects, and events in nature cannot, by virtue of their intrinsic qualities, give rise to religion, for there is nothing in their make-up which could, in itself, explain the religious thrill. This, indeed, is quite obvious, for do not the least significant beings and things in nature often become the objects of profound religious regard? Thus the source of religion may not be sought in natural experience but must in some significant way be interwoven with the conditions of human existence. Now the most fundamental and patent fact in all religion is the classification of all things, beings, events in experience into sacred and profane. This dichotomy of the universe is coextensive with religion; what will explain the one will explain the other. The next important fact to be noted is that the content of religion is not exhausted by its emotional side. Emotional experience is but one aspect of religion, the other aspects being constituted by a system of concepts and a set of activities. There is no religion without a church.

The fundamental propositions thus advanced by Durkheim do not impress one as convincing. In claiming that primitive man knows no supernatural, the author fundamentally misunderstands savage mentality. Without in the least suspecting the savage of harboring the conception of a natural order, we nevertheless find him discriminating between that which falls within the circle of everyday occurrence and that which is strange, extraordinary, requiring explanation, full of power, mystery. To be sure, the line of demarcation between the two sets of phenomena is not drawn by the savage where we should draw it, but surely [721] we should not thereby be prevented from becoming aware of the existence of the line and of the conceptual differentiation of phenomena which it denotes. If that is so, Durkheim commits his initial error, fatal in its consequences, in refusing to grant the savage the discriminating attitude towards nature and his own experience which he actually possesses. The error is fatal indeed, for the realm of the supernatural, of which Durkheim would deprive the savage, is precisely that domain of his experience which harbors infinite potentialities of emotional thrill and religious ecstasy.

  • 2 [Orig.] « un sytème solidaire de croyances et de pratiques ». Cf. « Introduction », Durkheim 1912, (...)

Durkheim’s objection to the derivation of the first religious impulses from what he calls illusions, strikes one as peculiar. For what, after all, is truth and what is illusion? Are not the highest religions, of undisputed significance and worldwide appeal, also based on illusions? Are not ideals, in more than one sense, illusions? Should one therefore be shocked if religion were shown to have its primal roots in an illusion? Thus Durkheim’s search for a reality underlying religion does not seem to rest on a firm logical basis. The author’s definition of religion, finally, represents a conceptual hybrid, the application of which could not but have the gravest consequences for his study. A religion, says Durkheim, is an integral system of beliefs and practices referring to sacred things, things that are separated, prohibited; of beliefs and practices which unite into a moral community called the church all those who participate in them. This apparently innocent definition involves a series of hypotheses. While all will concede that religion has a subjective as well as an objective side, that belief is wedded to ritual, the equating of the two factors in one definition arouses the suspicion of an attempt to derive one from the other, a suspicion justified by a further perusal of the work. Closely related, moreover, as are belief and ritual, they belong to different domains of culture, their relations to tradition, for instance, and to individual experience, are quite different, and the methodology of research in the two domains must be radically different. Unless this standpoint is taken at the outset, inextricable situations are bound to arise. That the body of believers constitutes a moral community is another proposition which one may set out to prove but which should not be taken for granted in an initial definition. The proposition further prejudices the investigator in favor of the social elements in religion and at the expense of the individual elements. The introduction of the term “church,” finally, as well as the designation of the religious complex as an “integral system[2], brings in an element of standardization and of unification, which should be a matter to be proved not assumed. [722]

Theory of Totemism.—Durkheim takes pains to set forth his reasons for discarding the comparative method of inquiry. The pitfalls of this mode of approaching cultural problems being familiar to ethnologists, we may pass over the author’s careful argumentation. As a substitute for the antiquated method Durkheim proposes the intensive study of a single area; for, he urges, the superficial comparison of half-authenticated facts separated from their cultural setting is pregnant with potentialities of error, while the thoroughgoing analysis of one instance may reveal a law. Australia is the author’s choice; for from that continent come detailed and comprehensive descriptive monographs; moreover, there, if anywhere, are we likely to discover the prime sources of religion: the social organization of the Australians being based on the clan, the most primitive form of social grouping, their religions must needs be of the lowest type. The author thus takes as his starting-point the Australian clan, which he conceives as an undifferentiated primitive horde. Each horde takes its name from the animal or plant most common in the locality where the group habitually congregates. The assumption of the name is a natural process, a spontaneous expression of group solidarity which craves for an objective symbol. To the totemic design or carving must be ascribed an analogous origin. Of this type of symbolism tattooing is the earliest form; not finding much evidence on that point .in Australia, the author borrows some American examples. The paintings and carvings of the Australian being very crude and almost entirely unrealistic, the author is again tempted to refer to the American Indian, while ascribing the character of Australian totemic art to the low degree of their technical advancement. The theory of social control will show us how the concept of power, mana, the totemic principle, originates in the clan. Here we take it for granted. Thus, on ceremonial occasions the individual is aware of the presence of a mysterious power; through the vertigo of his emotional ecstasy he sees himself surrounded by totemic symbols, churingas, nurtunjas, and to them he transfers his intuition of power; henceforth, they become for him the source from which that power flows. Thus it comes that the totemic representations stand in the very center of the sacred totemic cycle of participation; the totemic animal or plant, and the human members of the totemic clan become sacred by reflection. When so much is granted, the other peculiarities of totemism follow as a matter of course. Totemism is not restricted to the clans, their members, animals, carvings, but spreads over the entire mental universe of the Australian. The whole of nature is divided and apportioned between the clans, and all the beings, objects, phenomena of nature partake, to a [723] greater or less degree, of the sacredness of the totemic animal or plant or thing with which they are classified. This is the cosmogony of the totemic religion. Individual totemism, the worship of the guardian spirit, is a later derivative of clan totemism, for whereas clan totemism often appears alone, individual totemism occurs only in conjunction with clan totemism. Every religion has its individual as well as its social aspect. The guardian-spirit cult is the individual aspect of totemism. The subjective embodiment, finally, of the totemic principle is the individual soul. But whence the totemic principle? Before passing to the theory of social control which brings an answer to the query, we must pause to examine the theory of totemism as here outlined.

  • 3 [Durkheim mobilise les travaux de Lorimer Fison, Alfred William Howitt, Kamilaroi and Kurnai. Group (...)

While the author’s rejection of the comparative method deserves hearty endorsement, the motivation of his resolve to present an intensive study of one culture arouses misgivings. For thus, he says, he might discover a law. Applicable as this concept may be in the physical sciences, the hope itself of discovering a law in the study no matter how intensive of one historical complex, must be regarded as hazardous. And presently one finds that there is more to the story, for Australia is selected for the primitiveness of its social organization (it is based on the clan!) with which a primitive form of religion may be expected to occur. That at this stage of ethnological knowledge one as competent as Emile Durkheim should regard the mere presence of a clan organization as a sign of primitiveness is strange indeed. For, quite apart from the fact that no form of clan system may be regarded as primitive, in the true sense of the word, clan systems may represent relatively high and low stages of social development. Moreover, even were the social organization of the Australian to be regarded as primitive, that would not guarantee the primitiveness of his religion; just as his in reality complex and highly developed form of social organization appears side by side with a markedly low type of industrial achievement. Also from the point of view of the available data must the selection of Australia be regarded as unfortunate, for, in point of ethnography, Australia shares with South America the distinction of being our dark continent. A most instructive study in ethnographic method could be written based on the errors committed by Howitt, and Spencer and Gillen, as well as Strehlow, our only modern authorities on the tribes from which Durkheim derives all his data[3]. The fact itself that the author felt justified in selecting the Australian area for his intensive analysis, shows plainly enough how far from realization still is the goal which his own life-work has at least made feasible, the rapprochement of ethnology and of sociology. [724]

  • 4 [« Le totem individuel et le totem sexuel », Durkheim 1912, livre 2, chap. 4, p. 223-237]

But let us pass to the concrete points. The conception of a clan name being assumed as an expression of clan solidarity is suggestive enough. On the other hand, one must not be forgetful of the fact that a name serves to differentiate group from group, and that at all times names must have been given by group to group rather than assumed by each group for itself. Not that names were never assumed by groups –such names as, “we, the people” or “men,” etc., bespeak the contrary-but this process must be regarded as the exception rather than the rule. Moreover, groups of distinct solidarity such as phratries or the Iroquois maternal families, often appear without names (in the instance of the maternal family this is indeed always the case), so that the consciousness of solidarity in a group may not be regarded as inevitably leading to expression in the form of a name. As to the objective totemic symbol, the totemic carvings or drawings, it is discussed most loosely by our author. Not finding the totemic tattoo in Australia, he appeals to American examples, but this device, of course, does not strengthen his case except by showing that totemic tattoo occurs in America. Also, he completely neglects the cardinal differences between the totemic art of the Northwest Coast and that of the Aranda–to both of which he refers–in failing to note that whereas among the Tlingit or Haida the carved crests are positively associated with the totemic ideas, among the Aranda the churinga or ground and rock designs are at best but passive carriers of momentary (although recurrent) totemic associations. It is, in fact, quite obvious that the geometrical art of the area has neither originated in nor been differentiated through totemic ideas, but being of an extra-totemic origin, has been subsequently drawn into the totemic cycle of associations without, however, ever becoming actively representative of them. Similarly, with the so-called totemic cosmogony, the fact that social organization tends to be reflected in mythology cannot indeed be disputed; this fact, however, altogether transcends, in its bearing, the problem of totemism. Hence, when we find a sociological classification of the universe coexisting with a totemic complex, we are fully justified in regarding the two phenomena as genetically distinct and secondarily associated. The burden of proof, at any rate, falls upon those who would assert the contrary. Durkheim’s treatment of these as of other aspects of the Australian totemic complex reflects his failure to consider that view of totemism which was designed to show, at the hand of relevant data, that totemic complexes must be regarded as aggregates of various cultural features of heterogeneous psychological and historical derivation. Needless to add, [725] the adoption of that view would strike at the very core of Durkheim’s argument necessitating a complete recasting of the fundamental principles of La vie religieuse. Nor does Durkheim’s discussion of the relative priority of clan totemism carry conviction. Here his facts are strangely inaccurate, for far from it being the case that “individual totemism”[4] never occurs unaccompanied by clan totemism, the facts in North America, the happy hunting-ground of the guardian spirit, bespeak the contrary. Whereas that belief must be regarded as an all but universal aspect of the religion of the American Indian, it has nowhere developed more prolifically than among the tribes of the Plateau area who worship not at the totemic shrine. To regard the belief in guardian spirits, “individual totemism,” as an outgrowth of clan totemism is, therefore, an altogether gratuitous hypothesis! Having satisfied himself that all the elements which, according to his conception of religion, constitute a true religion, are present in totemism, Durkheim declares totemism to represent the earliest form of a religion which, while primitive, lacks none of those aspects which a true religion must have. Thus is reached the culminating point of a series of misconceptions of which the first is Durkheim’s initial view and definition of religion. For had he given proper weight to the emotional and individual aspects in religion, the aspect which unites religious experiences of all times and places into one psychological continuum, he could never have committed the patent blunder of “discovering” the root of religion is an institution which is relatively limited in its distribution and is moreover, distinguished by the relatively slight intensity of the religious values comprised in it. In this latter respect totemism cannot compare with either animal worship, or ancestor worship, or idolatry, or fetichism, or any of the multifarious forms of worship of nature, spirit, ghost and god. Several of these forms of religious belief are also more widely diffused than totemism and must be regarded as more primitive, differing from totemism in their independence from any definite form of social organization. Resuming the author’s argument, we now return to the" totemic principle," the origin of which must be accounted for.

  • 5 [« La notion de principe ou mana totémique et l’idée de force », Durkheim 1912, livre 2, chap.  6, (...)

The Theory of Social Control.—Analysis shows that society has the qualities necessary to arouse the sense of the divine. Social standards, ideals, moods, impose themselves upon the individual with such categorical force as to arouse the consciousness of external pressure emanating from a force transcending the powers of the individual. Through the action of this social force the individual on certain occasions behaves, feels, and thinks in a way which differs from the psychic activities of his [726] daily experience. The psychic situation of the orator and his audience, on the one hand, and, on the other, the actions and psychic experiences of individuals in the crusades or during revolutions, may serve as examples. Now the social unit with which the Australian is most intimately allied is the clan. The life of the clan mates consists of periods of non-eventful daily activities alternating with periods of violent emotional disturbances accompanying ceremonial occasions. While “the secrets” hold sway, to speak with the Kwakiutl, the individual lives on an exalted plane, manifesting qualities which altogether transcend those he possesses under ordinary conditions. The periodic recurrence of these two sets of ideas, emotions, acts, cannot but evoke in the individual the tendency to classify the totality of his experience into profane and sacred. The former embraces all that is strictly individual, the latter all that is social, The sense of external power which acts through the individual on social occasions will tend to crystallize into a concept of an undifferentiated, powerful, mysterious force, which pervades nature and absorbs the individual who feels himself external to that power and yet part of it. This power, as it appears to the Australian clansman, may be called the totemic principle[5]. It is not the clan emblem, the totemic design, which is worshiped, nor the totemic animal, nor the various beings and things which form part of the totemic cycle of participation; but the totemic principle, the mysterious substance which pervades them all and constitutes their holiness. It was shown in the preceding section how this sense of power, craving for objective, expression, attaches itself to the totemic symbols which surround the individual on ceremonial occasions and thus gives the initial stimulus to the formation of a sacred totemic world. Comparison with American data shows that the totemic principle is a forerunner of the wakan, the orenda as well as of the Melanesian mana. The concept is the same, the only difference being that the totemic principle, originating as it does within the clan, reflects the clan differentiation of the tribe, whereas, the wakan, the orenda, etc., belonging to a higher stage of development, have freed themselves from the constraint of the clan limit, and transcending it, have acquired that character of generality and homogeneity which distinguishes these concepts.

Thus a solution is reached not alone of the totemic problem, but of the problem of religion. The reality which underlies religion is society itself. In the Australian situation society appears in its most primitive form-the clan. The totemic principle, the nucleus of the most primitive religion, is the clan itself reflected in the psyche of the individual. Not aware of the real source of his subjective sense of power, the Australian [727] objectifies the latter in the form of religious symbolism, thus giving rise to the infinitely varied world of the concrete carriers of religious values. Thus, while here also there is illusion, it extends only to the content not to the existence of the ultimate reality, which is eternal.

We may first consider the minor issue raised in this section, namely the identification of the totemic principle with mana. On reading the pages devoted to this discussion the unprejudiced student soon perceives that the facts supporting Durkheim’s contention are altogether wanting. There is no indication that the beliefs underlying totemic religion are generically the same as those designated by the terms mana or orenda; and that the wakan and orenda concepts should represent later stages of religious evolution, having superseded a stage in which the totemic principle reigned, is an imaginary construction which cannot be described otherwise than aus der Luft gegriffen. The main issue of the section, however, is the derivation of the totemic principle. This, in fact, is Durkheim’s theory of religion, which is represented as a symbol of social control. Durkheim’s theory has the charm of originality, for no one else before him has, to my knowledge, held such a view, nor has the author himself, in his former writings, ever gone so far in his social interpretations of psychic phenomena. Our first objection to the derivation of the sacred from an inner sense of social pressure is a psychological one. That a crowd-psychological situation should have aroused the religious thrill in the constituent individuals, who–nota bene–were hitherto unacquainted with religious emotion, does not seem in the least plausible. Neither in primitive nor in modern times do such experiences, per se, arouse religious emotions, even though the participating individuals are no longer novices in religion. And, if on occasion such sentiments do arise, they lack the intensity and permanence required to justify Durkheim’s hypothesis. If a corroborree differs from an intichiuma, or the social dances of the North American Indians from their religious dances, the difference is not in the social composition but in the presence or absence or pre-existing religious associations. A series of corroborrees does not make an intichiuma; at least, we have no evidence to that effect, and human psychology, as we know it, speaks against it. Durkheim’s main error, however, seems to our mind to lie in a misconception of the relation of the individual to the social, as implied in his theory of social control. The theory errs in making the scope of the social on the one hand, too wide, on the other, too narrow. Too wide in so far as the theory permits individual factors to become altogether obscured, too narrow in so far as the society which figures in the theory is identified with a [728] crowd, and not with a cultural, historic group. The experience of all times and places teaches that the rapport of the individual, as such, with the religious object is of prime importance in religious situations. While, on the one hand, religious emotions are stimulated (not created) by the social setting, the leaders of religious thought, prophets, reformers, individuals whose lives must be conceived as protracted communions with the divine, do not require the social stimulant, they shun the crowd, the church, the world; their god is within them, and their emotional constitution is a guarantee of an interminable succession of religious thrills. The lives of saints are one great argument against Durkheim’s theory. The psychic cast of many a savage medicineman, magician, shaman, is another. If the social pressure, the ceremonial whirl is so indispensable a factor in the religious thrill, how is it that the world over the novice, in anticipation of the most significant, if not initial religious experience of his life, withdraws from human companionship, spends days, nay months in isolation, fasts and purifies himself, dreams dreams and sees visions? If phenomena of this type are so important in religion at all times, can one with impunity brush them aside in his search for a plausible origin of religion? Or would Durkheim claim that the religious thrill, socially produced, did then in some way become part of the psychic constitution of man in the form of a hereditary predisposition? But our author has not advanced this theory, and it would perhaps be unfair to attribute it to him.

On the other hand, the scope of the social in the author’s theory is too narrow. For, significant as are the functions ascribed to it, the content of the social setting, in Durkheim’s religious laboratory, is curiously restricted. Religion, he says, is society, but society, we find, is but a sublimated crowd. The only aspect of the relation of the individual to the social drawn upon in Durkheim’s theory is the crowd-psychological situation, the effect on the individual of the presence of other individuals who, for the time being, think, and above all, feel and act as he does. We hear nothing of the effect on the individual of the cultural type of the group of the tribal or national or class patterns of thought and action, and even emotion, patterns developed by history and fixed by tradition. Of all this we hear nothing. The only factor called upon to do such far-reaching service is that whimsical psycho-sociological phenomenon which equates a crowd of sages to a flock of sheep. Strange fact, indeed, that one who expects so much from the social should see in it so little!

  • 6 [« Les principales attitudes rituelles », Durkheim 1912, livre 3, p. 427-592]
  • 7 [« Le culte positif. I. Les éléments du sacrifice », Durkheim 1912, livre 3, chap. 2, p. 494]

Theory of Ritual.—It will be impossible to fully discuss in these pages [729] Durkheim’s suggestive analysis of rituals, negative and positive, mimetic, representative, and piacular[6]. We shall restrict our remarks to the types of ritual which bear directly on the theories here discussed. Ritual is essential for belief. Nature goes through certain periodic changes; evidently, thinks the Australian, the divinities controlling nature must go through similar transformations. To this spectacle man may not remain indifferent; he must assist the divinities with all the powers at his command. The divinities, totems, etc., derive their sacred character from man, hence, the sacredness will decline unless revived. The group gathers intent on relieving the situation. But presently they feel comforted: “They find the remedy because they look for it together.[7] On such occasions society becomes rejuvenated, and with it the soul of the individual, for is it not derived from society?

In the mimetic dances of the intichiuma the performers believe that they are the animals whose multiplication they crave, hence they imitate them in cries and actions. This identification of man and animal exists only to the extent to which it is believed, and the rite feeds the belief. The ceremony is beneficent for it constitutes a moral re-making of the participants. Hence the feeling that the ceremony has been successful. But it was intended to further the multiplication of the totemic animal, and now the belief that such multiplication has actually been achieved arises as a correlate of the feeling that the ceremony was successful. Such is ritualistic mentality.

In this case as in others the real justification of a religious rite is in the rite itself, that is, in the effect it produces on the social consciousness. The economic or other uses to which a rite is put are secondary, they vary and the same rite often does service for different purposes.

Another aspect of the ritualistic situation is what one might call an overproduction of thought, emotion, and activity. The elaboration of these processes is accompanied by pleasurable emotion, it becomes an end in itself. This is the threshold of Art.

  • 8 [« Les rites piaculaires et l’ambiguïté de la notion de sacré », Durkheim 1912, livre 3, chap.  5, (...)

A striking example of Durkheim’s conception of ritual and of its effect on belief, is presented in his interpretation of mourning. When an individual dies, the social solidarity of his family is shaken. Driven by the shock of their loss, they unite. At first this leads to an intensification of sorrowful emotion: a “panic of grief[8] sets in, in the course of which the individuals sob, howl and lacerate themselves. But presently the effect of this exhibition of solidarity in sorrow begins to be felt. The individuals feel comforted, reassured. The mourning is brought to an end through the agency of the mourning itself. [730]

But the individual remains perplexed. He must account for the strange exhibitions of mourning. Of social forces he knows nothing. All he is aware of is his suffering, and he seeks the cause for it in an external will. Now, the body of the deceased can surely not be held accountable, but his soul is there and it must be vitally concerned in the processes of the mourning rite; but these processes are highly disagreeable, hence the soul must be evil. When the mourning frenzy subsides, and a pleasurable calm ensues, the soul is again held responsible for the change, but now it appears as a benevolent agency. Not only the properties, but the survival itself of the soul, may, according to Durkheim, be an afterthought, introduced to account for the mourning rites.

Thus the ritual in this and similar cases appears as a spontaneous response of the group to an emotional situation. The beliefs, on the other hand, arise out of speculative attempts designed to interpret the phenomena of the ritualistic performance.

Durkheim’s psychological interpretation of ritual, must, on the whole, be regarded as the most satisfactory part of his analysis. Nevertheless here, as elsewhere, he permits himself to lapse into a rationalistic and behavioristic attitude. While it is, of course, true that divinities exist only to the extent to which they are believed in and that belief is stimulated by ritual, this dependence of the gods on belief is certainly a fact which never enters the mind of the native. He, for one, is profoundly convinced of the externality and objectivity of his spiritual enemies or protectors, nor does he believe in the waning and waxing of their powers, to keep pace with the periodic changes in nature. Moreover, while the rite may properly be regarded as a battery by means of which the participants are periodically re-charged with belief, this function of ritual may easily be exaggerated, nor should other sources be disregarded which tend to preserve accepted belief, such as the forces of tradition, teaching and more strictly individual, as contrasted with social, experience. It must be remembered that ritualism on an extensive scale is, while a common, by no means a constant nor even a predominant characteristic of primitive society. An analysis, from this point of view, of the North American area, for instance, reveals the suggestive fact that ritual en masse occurs mainly in the Southwest, Southeast, Northwest, Plains area, and part of the Woodland area, whereas among the Eskimo, in the Mackenzie and Plateau areas and in California, ritual is, speaking generally, an individual or family function. In other words, ritual en masse is associated with tribes of a complex social type, where the group is differentiated into many definite social units some of which [731] appear as the carriers of ceremonial functions; while the tribes with a relatively simple social structure, based on the individual family and the local community, are on the whole foreign to ritualism of the above type. This generalization cannot be accepted without certain reservations. The situation is really more complex, and other factors, such, for instance, as diffusion of rituals, would have to be taken into account; such tribes, moreover, as those of the Western Plains or the Nootka combine with a relatively simple type of social organization a relatively complex type of ritualism. Within certain limits, however, the generalization holds. Now, it becomes at once obvious that the intensity of religious belief is not correlated with complex ceremonialism. Among tribes devoid of complex ritualism, other factors must be operative to strengthen and perpetuate the existing belief; and, if that is so, we are also cautioned against the exclusive emphasis on ritual as a generator of belief even where it does occur on a large scale. The gods live not by ritual alone.

  • 9 [Orig.] « mentalité rituelle ». Cf. « Les rites représentatifs ou commémoratifs », Durkheim 1912, l (...)

As a most glaring instance of an extreme behaviorist position we must regard Durkheim’s attempt to account for the qualities nay, in part, even for the survival of the soil, by means of the “ritualistic mentality.”[9] Elaborate criticisms of hypotheses such as this are futile, for it obviously represents a deliberate effort to disregard the many emotional and conceptual factors which go to the making of the soul-belief in all its aspects, in favor of a simplicist behaviorist explanation. When Durkheim interprets the belief in the efficacy of the intichiuma as a reflection of the rise in social consciousness brought on by the ceremony he commits a similar error. It seems unjustifiable for instance, to disregard as a contributing factor in furthering the belief, the observation often made by the natives that the totemic animals and plants actually do multiply soon after the performance of the ceremonies. Durkheim does, indeed, note the fact, but he fails to utilize it in his theory.

  • 10 [« Les rites mimétiques et le principe de causalité », Durkheim 1912, livre 3, chap. 3, p. 527]

Theory of Thought.—Whereas the prime object of the author’s work is to trace the origin of religious beliefs and notions, he turns repeatedly to the more general problem of thought, of intellectual categories. While the author’s remarks on that subject are not extensive nor systematic, enough is said before the volume draws to a close, to make his position stand out in bold relief. No less than the categories of religion the categories of thought are of social origin. The importance of individual experience and of tentative generalizations derived therefrom should not be underestimated, but isolated individual experience lacks the elements necessary to give the notions which thus arise that character of [732] generality and imperativeness which distinguishes the mental categories. Mana, the totemic principle, that objectified intuition of society, is the first religious force, but also the prototype of the notion of force in general; just as the concept of soul, the active element in man, is, as shown, of social derivation. Similarly with the category of causality. The “will to believe” aspect of ritualistic mentality, as manifested, for instance, in the intichiuma ceremonies, has been dwelt on at length. But the belief alone is not sufficient; it would, at best, result in a state of expectancy. The rites must be repeated whenever need is felt of them, and the emotional attitude must be supplemented by a concept, if the intichiuma as a method of constraining or assisting nature is to be counted on. The concept that like produces like becomes a fixed mental category, and behind it is a social mandate. “The imperatives of thought seem to constitute but another aspect of the imperatives of Will.[10]

The notion that the qualities of objects can be communicated to their surroundings by a process of propagation, cannot be derived from daily experience, for the phenomenon in question does not occur within the domain of such experience, but constitutes a peculiarity of the religious world. Religious forces, qualities, being themselves but sublimated and transformed aspects of society, are not derived from objects but superadded upon them. The intrinsic virtues of the carriers of religious forces are thus indifferent, and the most insignificant things may become objects of greatest religious import. It is not strange that sacredness can be communicated by contagion from object to object for it is by contagion that sacredness becomes primarily fixed upon objects. Nor is this contagiousness of --the religious irrational, for it creates bonds and relations between objects, beings, actions, otherwise disparate, and thus paves the way for future scientific explanations. What was heretofore called the cosmogony of totemism, the classificatory aspect of the most primitive religion, thus becomes the prototype of classification in general, the first source of the notions of genus, subordination, coordination.

  • 11 [« Conclusion », Durkheim 1912, p. 630]]

The mental categories, concludes Durkheim, are not merely instituted by society, but they are, in their origin, but different aspects of society. The category of genus finds its beginning in the concept of the human group; the rhythm of social life is at the basis of the category of time; the space occupied by society is the source of the category of space; the first efficient force is the collective force of society, bringing in its wake the category of causality. The category of totality, finally, can only be of social origin. Society alone completely transcends the individual, rises above all particulars. “The concept of Totality is but the abstract [733] form of the concept of society: Society is the whole which comprises all things, the ultimate class which embraces all other classes.[11]

The author’s attempt to derive all mental categories from specific phases of social life which have become conceptualized, is so obviously artificial and one-sided that one finds it hard to take his view seriously, but the self-consistency of the argument and, in part, its brilliancy compel one to do so. In criticism we must repeat the argument advanced in another connection in the preceding section: in so far as Durkheim’s socially determined categories presuppose a complex and definite social system, his explanatory attempts will fail, wherever such a system is not available. The Eskimo, for example, have no clans nor phratries nor a totemic cosmogony (for they have no totems); how then did their mental categories originate, or is the concept of classification foreign to the Eskimo mind? Obviously, there must be other sources in experience or the psychological constitution of man which may engender mental categories; and, if that is so, we may no longer derive such categories from the social setting, even when the necessary complexity and definiteness are at hand.

In this connection it is well to remember that the origin of mental categories is an eternally recurring event; categories come into being within the mental world of every single individual. We may thus observe that the categories of space, time, force, causality, arise in the mind of the child far ahead of any possible influence from their adult surroundings by way of conscious or even deliberate suggestion. To be sure, these categories are, in the mind of the child, not strictly conceptualized nor even fully within the light of consciousness, but their presence is only too apparent: the individual experience of the child rapidly supplements the congenital predisposition of the mind. Instructive conclusions, bearing on these and other questions of epistemology, could be drawn from a systematic analysis of the grammars of primitive languages. Grammar is but a conceptual shorthand for experience and the means by which a relatively unlimited experience is squeezed into the frame of a strictly limited grammar is classification. Now, while the psychic processes underlying grammatical categories fall notoriously below the level of consciousness, they do nevertheless represent the deepest and most fundamental tendencies of the mind which, without doubt, provide the foundation for later, more conscious mental efforts, in similar directions. While no intensive study of primitive grammars, from the above point of view, has as yet been made, enough is known to foresee that but a fraction of the categories thus revealed will prove of specifically social derivation. [734]

There remains another equally fundamental criticism to be made of Durkheim’s doctrine. As we have seen, the author maintains that infectiousness is a specifically religious phenomenon. It does not seem that even the infectiousness of the sacred has been satisfactorily accounted for by the author. For, granting that sacredness is not inherent in objects but projected into them, that fact would not, per se, explain why sacredness should be so readily communicable from object to object. The Australian is not aware of the extraneous character of the sanctity of things, and surely it would be impossible for him to believe that his consciousness is if not the ultimate, yet the proximate source of that sanctity. Hence, the infectiousness of the sacred remains, from that standpoint, inexplicable. Another instance of the psychologist’s fallacy! This, however, is but a minor point. But can we follow the author in his assertion that infectiousness is peculiar to the sacred and that the quality is foreign to experience outside of the religious realm? Assuredly not. Daily observation brings before the mind of the savage numerous instances of the communicability of qualities. Wet comes from wet, and cold from cold; red ochre makes things red and so does blood, while dirt makes them dirty; touching rough surfaces brings roughness of skin and soreness; intimate contact with strongly smelling substances communicates the smell; heat, finally, produces heat-and pain. If the sacred is infectious, so is profane nature, and the mind which learns from the one its first lesson in categorizing can learn it from the other as well. I t will be seen that the above criticism is based on a special instance. It must now be generalized. The exclusive emphasis on the religious and ultimately on the social as the source of the fundamental categories of thought is unjustifiable in view of the rich variety of profane experience which is amenable to like conceptualization. While the point, when made in this general form, is fairly obvious, much interesting research work in this neglected field of primitive mentality remains to be done. The magico-religious aspect of primitive life and thought has for years monopolized our attention to such an extent that the less picturesque but no less real concrete experience of the savage has remained almost completely in the background. What does the savage know? should be the question. A vast store of data is available, on which to base our answer, and more can be procured.

The principal criticisms here passed on Durkheim’s work may now be summarized as follows:–

The selection of Australia as the practically exclusive source of information must be regarded as unfortunate, in view of the imperfection of [735] the data. The charge is aggravated through the circumstance that the author regards the case of Australia as typical and tends to generalize from it.

The Theory of Religion is deficient in so far as it involves the commingling in one definition of disparate aspects of the religious complex Many of the special points made in the course of the work are thus prejudged; the individual and subjective aspect of religion, in particular. Thus fails to receive proper attention.

The Theory of Totemism suffers from the disregard of the ethnological point of view which forces upon us the conviction that the institution must be regarded as highly complex historically and psychologically. The resulting interpretation of the totemic complex, while giving evidence of Durkheim’s superior psychological insight and often brilliant argumentation, recalls by its one-sidedness and artificiality the contributions to the subject on the part of the classical anthropologists.

The Theory of Social Control must be rejected on account of its underestimation as well as overestimation of the social, involving a fundamental misconception of the relation of the individual to society. For, on the one hand, the individual becomes, in Durkheim’s presentation, completely absorbed in the social; society itself, on the other hand, is not conceived as a historical complex but as a sublimated crowd.

The Theory of Ritual, while involving much true insight, is narrowly behavioristic and rationalistic and fails to do justice to the direct effect of experience upon the mind. The conception of the subjective side of religion as an after-thought, consequent upon and explanatory of action, must be vigorously rejected.

The Theory of Thought, finally, suffers from an exclusive emphasis on socio-religious experiences as the sources of mental categories, to the all but complete exclusion of the profane experience of the savage and the resulting knowledge of the concrete facts and processes in Nature.

Thus the central thesis of the book that the fundamental reality underlying religion is society, must be regarded as unproved.

Haut de page


Durkheim (Émile), « La prohibition de l’inceste et ses origines », L’Année sociologique, 1 (1896-1897), Mémoires originaux, 1898, p. 1-70

Durkheim (Émile), « Sur le totémisme », L’Année sociologique, 5 (1900-1901), Mémoires originaux, 1902, p. 82-121

Durkheim (Émile), « Sur l’organisation matrimoniale des sociétés australiennes », L’Année sociologique, 8 (1903-1904), Mémoires originaux, 1905, p. 118-147

Haut de page


1 [Émile Durkheim, « La prohibition de l’inceste et ses origines », L’Année sociologique, 1 (1896-1897), Mémoires originaux, 1898, p.1-70 ; id., « Sur le totémisme », L’Année sociologique, 5 (1900-1901), Mémoires originaux, 1902, p.82-121 et id., « Sur l’organisation matrimoniale des sociétés australiennes », L’Année sociologique, 8 (1903-1904), Mémoires originaux, 1905, p.118-147]

2 [Orig.] « un sytème solidaire de croyances et de pratiques ». Cf. « Introduction », Durkheim 1912, p. 65. Joseph Ward Swain, le premier traducteur des Formes en anglais, avait opté pour : « a unified system of beliefs and practices ». Cf. « Introduction », Durkheim 1915, p. 47

3 [Durkheim mobilise les travaux de Lorimer Fison, Alfred William Howitt, Kamilaroi and Kurnai. Group-Marriage and Relationship, and Marriage by Elopement. Drawn chiefly from the usage of the Australian Aborigines, also The Kurnai Tribe. Their customs in Peace and War, Melbourne/Sydney/Adelaide/Brisbane, George Robertson, 1880, 372 p. ; id., The Natives Tribes of South-East Australia, London, Macmillan, 1904, xx-819 p. ; Walter Baldwin Spencer, Francis James Gillen, The Native Tribes of Central Australia, London/New York, Macmillan and Co / The Macmillan Company, 1899. xx-670 p. ; id., The Northern Tribes of Central Australia, London, Macmillan, 1904, xxxvi-784 p. et Carl Friedrich Theodor Strehlow, Die Aranda- und Loritja-Stämme in Zentral-Australien, Frankfurt am Main, Josef Baer & Co, « Veröffentlichungen aus dem städtischen Völker-Museum Frankfurt am Main » (Teil 1, Mythen, Sagen und Märchen des Aranda-Stammes, 1907, 104 p. ; Teil 2, Mythen, Sagen und Märchen des Loritja-Stammes: Die totemistischen Vorstellungen und die Tjurunga der Aranda und Loritja, 1908, 140 p. ; Teil 3, Allgemeine Einleitung und die Totemistischen Kulte des Aranda-Stammes, 1910-1911, 75 p.)]

4 [« Le totem individuel et le totem sexuel », Durkheim 1912, livre 2, chap. 4, p. 223-237]

5 [« La notion de principe ou mana totémique et l’idée de force », Durkheim 1912, livre 2, chap.  6, p. 268-292]

6 [« Les principales attitudes rituelles », Durkheim 1912, livre 3, p. 427-592]

7 [« Le culte positif. I. Les éléments du sacrifice », Durkheim 1912, livre 3, chap. 2, p. 494]

8 [« Les rites piaculaires et l’ambiguïté de la notion de sacré », Durkheim 1912, livre 3, chap.  5, p. 572]

9 [Orig.] « mentalité rituelle ». Cf. « Les rites représentatifs ou commémoratifs », Durkheim 1912, livre 3, chap. 4, p. 541

10 [« Les rites mimétiques et le principe de causalité », Durkheim 1912, livre 3, chap. 3, p. 527]

11 [« Conclusion », Durkheim 1912, p. 630]]

Haut de page

Pour citer cet article

Référence papier

Goldenweiser Alexander Aleksandrovich, « Book Reviews - Methods and Principles », American Anthropologist (Washington), n.s. 17 (4), October-December 1915, p. 719-735.

Référence électronique

Alexander Aleksandrovich Goldenweiser, « Book Reviews - Methods and Principles », Archives de sciences sociales des religions [En ligne], La première réception des Formes (1912-1917) (S. Baciocchi, F. Théron, eds.), I, mis en ligne le 27 février 2013, consulté le 29 mars 2017. URL :

Haut de page

Droits d’auteur

Archives de sciences sociales des religions

Haut de page
  • Logo Éditions de l’EHESS