Rereading Durkheim’s Les formes élémentaires de la vie religieuse (PUF, 1960 [1913]), I am struck more than previously by the narrowness of the barrier that separates Durkheim from the originary hypothesis. Durkheim is the first thinker to have ananthropological understanding of the transcendental. This is something that has never been fully grasped even by his disciples and those whom they in turn have influenced. But what is most striking is that it was not fully grasped by Durkheim himself.

1. Origin(s)

Pour bien comprendre un délire et pour pouvoir lui appliquer le traitement le plus approprié, le médecin a besoin de savoir quel en a été le point de départ. Or cet événement est d’autant plus facile à discerner qu’on peut observer ce délire à une période plus proche de ses débuts. . . . Il en est de même pour la pensée religieuse. A mesure qu’elle progresse dans l’histoire, les causes qui l’ont appelée à l’existence, tout en restant toujours agissantes, ne sont plus aperçues qu’à travers un vaste système d’interprétations qui les déforment. . . . On . . . verra [dans cet ouvrage] comment, dans les religions primitives, le fait religieux porte encore visible l’empreinte de ses origines . . . To understand a delirium and to give it the most appropriate treatment, the doctor needs to know its point of departure. Now this event is all the easier to discern as one is able to observe the delirium at a time nearer its beginning. . . . The same is true of religious thought. As one progresses through history, the causes that called it into existence, while remaining active, are perceived only through a vast system of interpretations that deform them. . . . We will see [in this work] how, in primitive religions, the religious phenomenon still visibly bears the impression of its origins . . .
L’étude que nous entreprenons est donc une manière de reprendre, mais dans des conditions nouvelles, le vieux problème de l’origine des religions. Certes, si, par origine, on entend un premier commencement absolu, la question n’a rien de scientifique et doit être résolument écartée. Il n’y a pas un instant radical où la religion ait commencé à exister et il ne s’agit pas de trouver un biais qui nous permette de nous y transporter par la pensée. Comme toute institution humaine, la religion ne commence nulle part. Aussi toutes les spéculations de ce genre sont-elles justement discréditées . . . Ce que nous voudrions, c’est trouver un moyen de discerner les causes, toujours présentes, dont dépendent les formes les plus essentielles de la pensée et de la pratique religieuse. Or, pour les raisons qui viennent d’être exposées, ces causes sont d’autant plus facilement observables que les sociétés où on les observe sont moins compliquées. Voilà pourquoi nous cherchons à nous rapprocher des origines.(1) The study that we are undertaking is thus a way of taking up again, but under new conditions, the old problem of the origin of religion. Of course if by origin we understand a first, absolute beginning, the question is wholly unscientific and must be resolutely put aside. There is no radical instant in which religion began to exist, and we have no intention of seeking a means to transport ourselves to that instant in thought. Like every human institution, religion begins nowhere. Thus all speculations of this sort are justly discredited . . . What we desire is to find a means of discerning the causes, still present, on which the most essential forms of religious thought and practice depend. For the reasons just given, these causes are all the easier to observe as the societies in which we observe them are less complex. That is why we seek to return to the origins.(1)
(1) On voit que nous donnons à ce mot d’origines . . . un sens tout relatif. Nous entendons par là non un commencement absolu, mais l’état social le plus simple qui soit actuellement connu, celui au delà duquel il ne nous est pas présentement possible de remonter. . . . (10-11; emphasis the author’s) (1) One sees that we give this word “origins” . . . a wholly relative meaning. We understand by it not an absolute beginning, but the simplest social state currently known, the state beyond which it is not presently possible to return. . . .

Durkheim’s comparison of religious thought to delirium is less Freudian, let alone Dawkinsian, than it appears; as the rest of the book will show, however “delirious” the manner in which religious thought deals with natural phenomena, this is entirely secondary to the ethical content of this thought and the way in which its norms are reinforced through collective ritual action. Nonetheless, one cannot help but be struck by the contrast between the concrete événement alleged at the origin of the delirium and the outright denial of any specific beginning in the “origins” of religion.

As a social scientist, Durkheim cannot be expected to tolerate unverifiable speculation about these origins. Yet his very use of the word origin is by no means justified by his methodological affirmation that the religions he proposes to study represent l’état social le plus simple qui soit actuellement connu, that is, relative to a necessarily unstable state of ethnological knowledge. However simple these aboriginal societies may be, they are as far removed chronologically as the most advanced societies from the time of the first humans. Even putting aside the fact that their failure to advance very far in material culture in no way guarantees that they have not changed considerably in their religious culture, the equation of simplicity and originarity implies the prior historical existence of a maximally simple, originary religion. This necessity is borne out by Durkheim’s very own “digital” definition of religion as a system of beliefs and practices having at its core the binary pair sacred-profane. As we shall see, Durkheim’s own claim is that these two categories designate two absolutely different ontological spheres, a usage that he sees as unproblematically compatible with his resolute rejection of a premier commencement absolu for religion as a whole. Whatever faith we have in the biological dictum natura non facit salta [nature makes no leaps], such cannot be the case for a binary pair of categories, let alone one that divides the universe into two absolutely separate ontological zones. The implications of Durkheim’s conception of religion lead outside the domain of social science, which they show to be grounded on a historical construct—an originary event—with which it is incompatible.

This is not to say that our task is to “deconstruct” empirical social science as Durkheim shaped it. GA is not a rival to social science, but a way of thinking that complements it, not only by providing it with an ontological foundation, but by suggesting directions for empirical research—in the case at hand, the traces of the originary asymmetry from which the binary symmetries of language were born.

2. Sacred and profane

Durkheim lacked access to a category that would not enter the arsenal of modern thought until Saussure: that of semiosis, the creation of meaning through a differentiated paradigm of signs. Thus when Durkheim speaks of the primary opposition between the sacred and the profane, he ignores the analogy between this binary distinction and those of language. Indeed, he denies this analogy, insisting that the sacred-profane distinction corresponds to an absolute ontological distinction that contrasts with ordinary linguistic dichotomies such as black-white or good-evil, which oppose qualities applicable to the same kind of objects.

This ignorance of Saussurean semiology distinguishes Durkheim (and Mauss) from Lévi-Strauss, for whom structural linguistics supplies the definitive model for paradigms of thought as well as language. Yet, paradoxically, on this point it is Durkheim’s anthropological intuition that is superior. The sacred-profane distinction cannot be understood as a rudimentary system of “differences,” as Saussure describes la langue. The absolute distinction of the sacred creates an absoluteasymmetry between it and the profane, in which all attention and specificity is concentrated on the former, the latter being merely a residue. This asymmetrical distinction, in contrast, corresponds effectively to the notion of language that emerges from the originary hypothesis. If anything, Durkheim’s presentation of the sacred-profane opposition is too symmetrical:

La division du monde en deux domaines comprenant, l’un tout ce qui est sacré, l’autre tout ce qui est profane, tel est le trait distinctif de la pensée religieuse . . . (50-51) The division of the world into two domains comprising, one, all that is sacred, the other, all that is profane, this is the distinctive characteristic of religious thought . . .
Mais, si une distinction purement hiérarchique est un critère à la fois trop général et trop imprécis, il ne reste plus pour définir le sacré par rapport au profane que leur hétérogénéité. Seulement, ce qui fait que cette hétérogénéité suffit à caractériser cette classification des choses et à la distinguer de toute autre, c’est qu’elle est très particulière: elle est absolue. Il n’existe pas dans l’histoire de la pensée humaine un autre exemple de deux catégories de choses aussi profondément différenciées, aussi radicalement opposés l’une à l’autre. L’opposition traditionnelle entre le bien et le mal n’est rien à côté de celle-là: car le bien et le mal sont deux espèces contraires d’un même genre, à savoir le moral . . . tandis que le sacré et le profane ont toujours et partout été conçus par l’esprit humain comme des genres séparés, comme deux mondes entre lesquels il n’y a rien de commun. (53; emphasis the author’s) But if a purely hierarchical distinction is a criterion both too general and too unspecific, only their heterogeneity remains to distinguish the sacred from the profane. However, what makes this heterogeneity sufficient to characterize this classification of things and to distinguish it from any other is that it is very particular: it is absolute. There does not exist in the history of human thought another example of two categories of things so profoundly differentiated, so radically opposed to each other. The traditional opposition between good and evil is insignificant in comparison, for good and evil are two contrary species of the same genus, that is, the moral . . . whereas the sacred and the profane are always and everywhere conceived by the human mind as separate genres, as two worlds that have nothing in common with each other.

What is striking in these passages is that in straining to oppose these two categories, Durkheim in fact makes them more parallel than they should be. The appeal to “the history of human thought” is particularly awkward, since Durkheim was the very thinker who affirmed that the sacred-profane distinction was in fact the point of departure for “the history of human thought.” The idea of a punctual origin, repudiated as “unscientific,” finds its way into Durkheim’s text through the use of hyperbole, of quantitative exaggeration intended to convey an absolute, qualitative distinction.

No doubt the fundamental intuition is more important than the rhetoric with which it is expressed, but the limitations of this rhetoric are also the limitations of the world of social science that Durkheim did so much to found on a rigorous basis. Recasting sacred and profane as a Saussurean paradigm of differences would only weaken the truth that lies behind Durkheim’s words: the sacred is the sole category of significance and it emerges not in opposition to another category that we call the profane but in opposition to the insignificance of the rest of the world. The paradigmatic experience of signification is not the division of the world between up and down, left and right, hot and cold, … but between the sacred that captures and holds our attention and everything else that does not. Thus Durkheim’s fortunate unawareness of his contemporary Saussure and of the “structuralist” paradigm of differences in which the sacred-profane distinction might have found its place allowed Durkheim to grasp and appreciate, expressed in however unsatisfactory a fashion, the asymmetry of the absolute distinction between sacred and profane.

3. Representation

Another passage in the Formes élémentaires lends itself to a similar analysis, but here the missing paradigm leads to a real misstatement, since what is missing here is not the paradigm of differences but the very notion of how the use of signs separates human from animal cognition. That sacred and profane, however understood, are uniquely human categories is never in question; the same is not the case in Durkheim’s use of the more general category of “representations.”

L’homme n’est rien autre chose, au point de vue physique, qu’un système de cellules, au point de vue mental, qu’un système de représentations : sous l’un ou l’autre rapport, il ne diffère qu’en degrés de l’animal. Et pourtant, la société le conçoit et nous oblige à le concevoir comme investi d’un caractère sui generis qui l’isole, qui tient à distance les empiétements téméraires, qui, en un mot, impose le respect. Cette dignité qui le met hors de pair nous apparaît comme un de ses attributs distinctifs, bien qu’il soit impossible de rien trouver dans la nature empirique de l’homme qui la fonde. (325; emphasis mine) Man is nothing other, from a physical standpoint, than a system of cells, orfrom a mental standpoint, than a system of representations; in either case, he differs only by degree from the animals. And yet, society conceives him and obliges us to conceive him as invested with a sui generis character that sets him off, that keeps at a distance any foolhardy attempts to bridge the gap, in a word, that imposes respect. This dignity that puts man in a class by himself appears to us as one of his distinctive attributes, although it is impossible to find anything in man’s empirical nature that justifies it.

As a “system of cells” we may be only quantitatively different from other animals, but as a “system of representations” we are “absolutely” different, because our minds not only contain representations in the sense of perceptive traces but createrepresentations in the sense of words and artificial images. When Durkheim credits society for “conceiving” our mentality as “isolated” from that of other species, he is in fact giving it less credit than it deserves. It is “society,” that is, the human community, that is the source of the distinctly human symbolic representations that are indeed isolated from the signals of other species. Human language is isolated from animal communication not by “society” as an external agent, but by the embodiment of “society” in its very nature.


As these passages make clear, originary thinking, rather than doing violence to Durkheim’s thought, gives it the coherent expression that it lacks in the author’s original formulations. It is paradoxically because Durkheim did not have access to the paradigms of structural linguistics that he retained, in spirit if not in vocabulary, the Kantian notion of transcendence, but now associated not with a disembodied mind but with human “society” as a collectivity unlike any other because defined by a set of shared “representations.” That Durkheim lacked the conceptual vocabulary to characterize these representations as different in kind from those of other animals is a trivial matter in respect of his achievement in valorizing the specificity of the human community that makes possible their sharing.

We should reconsider in the light of these observations the critique, generally considered unanswerable, of Durkheim’s “essentialism” by Lévi-Strauss, who reproaches his precursor for believing that the totem represents “society” rather than being an element in a mental paradigm. This structuralist critique attributes to Durkheim the naïveté of believing that signs “represent” things rather than signifieds in a Saussurian system of differences. But it is precisely this “naïveté” that makes Durkheim able to understand the originary status of the sacred-profane distinction. This distinction precedes and is presupposed by the formation of any paradigm of signification, not because it refers to things rather than ideas, but because for signification to take place at all, it must have a point of departure in the communal agreement to represent rather than to appropriate, to create a vertical relation in the place of a horizontal one. Seen from this perspective, it is rather Lévi-Strauss who is naïve in his insistence that signs can refer only to ideas, as though this process could somehow originate without reference to the things of this world and our relationship to them.

In a previous reference to Durkheim, I claimed that “However loftily he dismisses the ‘speculative’ question of the minimal constitution of society and representation, Durkheim’s sociology of religion is really Generative Anthropology avant la lettre” (Chronicle 198). The point of the present analysis is to make clearer how deeply the intuition of the originary-as-generative is located in the underlying pattern of Durkheim’s thought. I say pattern and not structure because, precisely, it is Durkheim’s ignorance of the semiotic ontology that would come to be called “structuralism” that allows him to combine the proto-structuralist claim that religious classifications are the basis for the categories of human thought with a way of thinking that points to the origin of structure in the singular designation of the sacred, whose binary, predicative opposition to the profane is clearly derivative of the absolute singularity of its ostensive origin.