1In the legendary narrative of the history of the social sciences runs the theme of a radical opposition between Émile Durkheim and Lucien Lévy-Bruhl with respect to the analysis of “primitive” thinking. Raymond Boudon, for instance (1999; 2010, 65-66), often mentions this opposition and turns it into the founding element of his conception of sociology. Since the study of scientific debates often sheds light on the social history of science, it could be useful to go back to the source of that opposition: the sections of The Elementary Forms of the Religious Life that criticize How Natives Think. [1] I propose here to review and analyze these sections and compare them to Lévy-Bruhl’s work to assess the form, meaning, and impact of that opposition as precisely as possible. I wish to demonstrate that, under a superficial, and commonly overstated, disagreement, lies another one that generally goes unnoticed.

A Common Field of Investigation

2In How Natives Think, [2] Lucien Lévy-Bruhl locates his research on the “path” opened by “Durkheim and his collaborators”, with the aim of formulating “a theory of knowledge, both new and positive, founded upon the comparative method.” As a “preliminary problem,” he intends “to find out exactly what are the guiding principles of primitive mentality,” and thinks he has managed to “show that the mental processes of ‘primitives’ do not coincide with those which we are accustomed to describe in men of our own type” (Lévy-Bruhl 1926, 14). His whole book, like the other five that followed, aims at demonstrating that there exist in “primitive” societies mental mechanisms whose “guiding principles” – or, to use a common term in his work, their “orientation” – differ from those described by philosophers and logicians.

3In The Elementary Forms of the Religious Life, [3] Durkheim does not just intend to account for the social origin of religion, but also for the origin of the knowledge or categories that constitute its “tools” (1915, 20n24). [4] The preliminary chapter, entitled “Religious Sociology and the Theory of Knowledge,” traces the origin of the latter to the former; since “categories of understanding” were “born in religion and of religion”, “they are a product of religious thought.” If “religion is something eminently social,” and “if the categories are of religious origin,” it follows that “they too should be social affairs”, or at least “rich in social elements” (10-11). [5]

4The intention to develop a “theory of knowledge” is thus common to both works, which also share the sociological grounding of this theory and the use of ethnographic data to illustrate it.

Shared Perspectives

5Not only do Lévy-Bruhl’s and Durkheim’s books share a common goal, they also display convergent accounts. Just as Lévy-Bruhl emphasizes the differences in modes of thinking (the object of his research), Durkheim also asserts their variability: “the categories of human thought are never fixed in any one definite form; they are made, unmade, and remade incessantly; they change with places and times” (1915, 15). And just as Lévy-Bruhl is known for the idea that the principle of contradiction does not operate in the same way in different societies, Durkheim similarly wonders whether “the idea of contradiction does not also depend upon social conditions,” since “the empire which the idea has exercised over human thought has varied with times and societies”: the “rules which seem to govern our present logic” are not “engraven through all eternity upon the mental constitution of men, they depend, at least in part, upon factors that are historical and consequently social” (12-13).

6Such a convergence is also reflected in the authors that are criticized: in its introductory chapter, Lévy-Bruhl questions the “animistic hypothesis” of Tylor, Frazer, and the other representatives of the “English school of anthropology.” It is based on a belief “admitted … as a postulate, or rather, an axiom”: “th[eir] belief in the identity of a ‘human mind’ which, from the logical point of view, is always exactly the same at all times and in all places” (Lévy-Bruhl 1926, 18-19). In Chapter 2, Book 1 of Forms, Durkheim introduces Tylor and Spencer’s “animistic theory” and criticizes it for being unable to account for the “origins of religious thought” (Durkheim 1915, 49). Lévy-Bruhl (1926, 18) refers to Durkheim’s article (1909a), which was the first draft of this chapter, for a “critical discussion of the method employed and the results obtained by these experts.”

7Durkheim underlines that similarity, if not in Forms, at least in the review that he devotes to the two works in L’Année: “There is no need to mention that we share some basic principles with Lévy-Bruhl”: the existence of “different types of mentality” and the religious nature of “primitive mentality” (Durkheim 1913, 35).

Continuity versus Discontinuity

8Yet, Lévy-Bruhl quotes Durkheim only in order to locate his own research within a Durkheimian framework, and whereas Durkheim underlines the variability of modes of thinking, he explicitly (and sometimes implicitly [6]) quotes Lévy-Bruhl only in order to distance himself from him. What are the specific points of difference?

9In Forms, Lévy-Bruhl is quoted in three different instances. The first mention concerns the general characteristics of “primitive thinking”; the second, its relation to experience; and the third, concepts. A fourth occurrence can be found in the review of Forms and of Natives that appeared in L’Année. In each of these passages, Lévy-Bruhl is criticized for exaggerating differentiation or discontinuity where, conversely, Durkheim highlights resemblance and continuity. This “disagreement” is formulated most comprehensively in the first and fourth passage:

10

Thus between the logic of religious thought and that of scientific thought there is no abyss.

11

Between these two stages of the intellectual life of humans [“religious and primitive thought” and “scientific and modern thought”] there is no break.

12Therefore, the disagreement pointed out by Durkheim bears on the difference between contrasting types of mentality or the significance that is to be ascribed to that difference.

Specifications

13Let us detail the specific points about which Durkheim chooses to stress continuity, despite differentiation.

141. The principle of contradiction. The first passage is to be found in a section that aims at showing “how logical evolution is closely connected with religious evolution and how it, like this latter, depends upon social conditions” (1915, 234). A footnote specifies that, as these pages were written when Natives appeared, Durkheim confines himself to “adding certain explanations showing in what we differ from M. Lévy-Bruhl in our understanding of the facts” (234n4). A few pages later, in words reminiscent of Comte when he credited the priests of the polytheist stage for having set human thought in motion, once and for all, towards its scientific goal, [7] he notes that

15

“The great service that religions have rendered to thought is that they have constructed a first representation of what these relations of kinship between things may be.… [F]or from the moment when men have an idea that there are internal connections between things, science and philosophy become possible. Religion opened up the way for them.”

16In the following section, Lévy-Bruhl is mentioned (and this time Durkheim might seem to echo Émile Meyerson [8]):

17

So it is far from true that this mentality has no connection with ours. Our logic was born of this logic…. It has been said that the participations of this sort implied by the mythologies violate the principle of contradiction and that they are by that opposed to those implied by scientific explanations.

18But we also unite “heterogeneous terms by an internal bond” (such as heat and movement) and, by so doing, “we forcibly identify contraries.” In fact, “primitive thought” does not have “that sort of general and systematic indifference to contradictions which has been attributed to it [referring to Lévy-Bruhl 1926, 78] … If the primitive confounds things which we distinguish, he also distinguishes things which we connect together” and he does it “in the form of sharp and clear-cut oppositions” (Durkheim 1915, 238).

192. The relation to experience. In Chapter 3 of Book 3, another allusion to Lévy-Bruhl bears on the nature of faith, “impermeable to experience” (referring to Lévy-Bruhl 1926, 59-66), to show that this nature “cannot radically distinguish religious mentality from the other forms of human mentality” and from “that of a scholar,” which “differs from the preceding only in degree” (Durkheim 1915, 360-361). In both cases, it is “a constantly renewed experience” that teaches one not to forsake too quickly a well-tried mode of explanation: “The scholar does not do otherwise; only he introduces more method” (361).

203. Conceptual Thought. Section 3 of the Conclusion ends with a third critique of Lévy-Bruhl’s book. To understand its function, it is useful to recall the general line of argument of that section, which bears on “conceptual thought” and intends to explain how religion could engender “fundamental notions of science” (Durkheim 1915, 431-432). Accordingly, Durkheim subsumes all possible forms of thought under the term “concept.”

21His goal is to show that religious representations “are real concepts” (Durkheim 1915, 432). The nature of the concept is to be (relatively) immutable and universal, or capable of becoming so (433), which is only possible if “it is the product of a collective elaboration” (434). [9] Thus, “each civilization has its organized system of concepts which characterizes it” (435) and scientific concepts, more “methodically controlled” than the others, “are always in the very slight minority” (437-438). “[T]here are only differences of degree” (437) between ordinary concepts, crystallized, fixed, or expressed (433) in the language, and scientific concepts. The former give “guarantees of objectivity” and the latter are also based on a form of faith; their “privileged credit” results from the fact that “we have faith in science” (438). Thus, “conceptual thought is coeval with humanity itself” and “[a] man who did not think with concepts would not be a man.” In order to support “the contrary thesis,” one has to define “concepts … by characteristics which are not essential to them”: “They have been identified with general ideas and with clearly limited and circumscribed general ideas” (438-439). Two footnotes identify this mistake in the understanding of the concept in Lévy-Bruhl’s work (with references to pages 129-136 and 445 of Natives). It is not because “inferior societies … have only rudimentary processes of generalizations” (439) that their members do think “conceptually,” and “since logical thought commences with the concepts, it follows that it has always existed” (ibid.).

22As for that section, it closes with the claim that diversity and continuity must be considered together:

23

To be sure, we cannot insist too much upon the different characteristics which logic presents at different periods in history; it develops like the societies themselves. But howsoever real these differences may be, they should not cause us to neglect the similarities, which are no less essential.

Did Lévy-Bruhl Go Too Far?

24According to Lévy-Bruhl, the three points to which Durkheim objects are interconnected; for example, he notes that “the impossibility of affirming two contradictory statements at the same time” and that “of believing in relations which are incompatible with experience … is felt only” when “collective representations tend toward conceptual form” (Lévy-Bruhl 1926, 378). Since Durkheim thinks that collective representations can only be conceptual (by the very definition of concept), it might indeed well be Lévy-Bruhl’s description as a whole that is put into question. And, for each of these points, what is aimed at, through a two-pronged argument, is the recovery of the resemblance or continuity that exists between religion and science: religion already does what science does (at least in part): establishing connections, drawing on experience, and thinking conceptually. In its own way, science also does what religion has been doing all along: bringing together things of different nature, resisting the invalidation of theories, and drawing on common opinions. In short, Durkheim does not challenge the “facts” that Lévy-Bruhl describes, but his “interpretation” of these facts as indicating too radical a separation between “religious thought” and “scientific thought.” Durkheim sees a necessary continuity between them, as the former represents the source (albeit inarticulate) of the latter: “scientific thought is only a more perfect form of religious thought” (1915, 429). Such would be the price of demonstrating that knowledge has a social origin, once the same has been proven for religion.

25How is one to reconcile the points of agreement noted above and this divergence, single but reasserted thrice? Despite the global nature of the criticism, the point is not to deny what Lévy-Bruhl claims, but to qualify and moderate his analysis. Differences exist and they are significant, but they are not as deep as he depicts them in his zealousness; they are about degree, not substance. By systematically opposing “primitive” and “modern” thinking as two ideal types, Lévy-Bruhl errs as to their homogeneity and the difference between them. Fundamentally, Durkheim seems to be applying to Lévy-Bruhl the language that he uses about the “logic of religious thought,” so as to emphasize its differences from the logic of “scientific thought”:

26

It is voluntarily excessive in each direction. When it connects, it confounds; when it distinguishes, it opposes. It knows no shades and measures, it seeks extremes; it consequently employs logical mechanisms with a certain awkwardness, but it ignores none of them.

27Durkheim acknowledges that the difference here bears on points of view or research objectives: “Yet, our point of view is somewhat different from Lévy-Bruhl’s. He was concerned mainly with differentiating this mentality from ours, to the point of presenting these differences, sometimes, as a real antithesis” (Durkheim 1913, 35). Whereas contrast is inherent to Lévy-Bruhl’s comparative project, continuity is at the basis of Durkheim’s position, because he is concerned above all with tracing science and religion back to the same origin:

28

If, then, human thought has changed across centuries and societies, if it has evolved, its different, successive manifestations have been the source the ones of the others. The higher and most recent forms are not opposed to the most primitive, inferior ones; they were born of the latter.

Did Durkheim Read Lévy-Bruhl Correctly?

29One of the characteristics of this controversy is that it remained one sided. In his subsequent works, Lévy-Bruhl did not respond to Durkheim’s criticisms. He seems to have refused to do so, since he declined the opportunity of reviewing Forms for the Revue Philosophique. [10] Maybe it is because, unlike Durkheim, he did not have a taste for disputes. It is also possible that he did not think that these criticisms ought to affect the specificity of his approach.

30The fact is that Durkheim judges Lévy-Bruhl’s views as overstated or too categorical because he forces or distorts his language and his claims. For example, where Durkheim reads a “sort of general and systematic indifference” to contradiction, (Durkheim 1915, 238), Lévy-Bruhl writes (on the page quoted by Durkheim):

31

Have there ever existed groups of human or prehuman beings whose collective representations have not yet been subject to the laws of logic? We do not know; and, in any case, it seems to be very improbable. [The] prelogical [mentality to be described, in any case,] does not partake of that nature: … it does not expressly delight in what is contradictory (which would make it merely absurd in our eyes), but neither does it take pains to avoid it.

32He further specifies that it is a “mentality in which the logical and the prelogical can be coexistent and make themselves equally perceptible in mental processes.” Similarly, he emphasizes that “primitive mentality” is far from “appear[ing] to us unbridled and unregulated, and just as purely arbitrary” (Lévy-Bruhl 1926, 106, 108). On the contrary, he keeps on claiming that primitive mentality operates according to fixed or established schemes, even though they (or in his words, these “preconnections”) are not called “concepts.”

33Lévy-Bruhl also thinks that “primitive mentality” is based on a form of experience, i.e. a shared experience which is more influential than facts and which he already called “mystical” well before using the term in the title of his 1938 book: “[Primitive thought] has [its] own experience to guide [it], a mystic experience against which, as long as it continues to exist, actual experience is powerless” (Lévy-Bruhl 1926, 378). Yet, by doing so, he distinguishes two types of experience where Durkheim seems to be considering only one. [11]

34The very fact that the “prelogical mind” may contain concepts is another point of convergence: “it, too, is transmitted socially by means of language and concepts without which it could not be exercised” (Lévy-Bruhl 1926, 107; “But,” he adds, “these concepts differ from ours”). Here, Lévy-Bruhl draws a distinction between various forms (“the collective representations of primitives, therefore, differ very profoundly from our ideas or concepts” [Lévy-Bruhl 1926, 37]) that Durkheim tends to conflate.

35In short, Lévy-Bruhl’s views may appear overstated or one sided, but only insofar as Durkheim simplifies his thought, strips it of its nuances, and ignores the distinctions on which it draws.

36Lévy-Bruhl’s thought is further distorted by the very fact that Durkheim retranslates it in the terms of his own perspective. Integrating his criticism of Lévy-Bruhl in the presentation of his theory, he constantly conflates “primitive” and “religious thought” on one hand and “our thought” and science on the other, just as if Lévy-Bruhl himself accepted the conflation of these notions. [12] His only ground for doing so is the use of the term “mystical,” which Lévy-Bruhl had taken great care to distinguish from its modern, religious use. [13]

37Although Durkheim strains or misinterprets his writings, some of the statements he makes are still in line with Lévy-Bruhl’s views. It is for example true of the entire second paragraph of the section that leads to the first criticism of Lévy-Bruhl (Durkheim, 1915, 234-235), which in the table of contents is entitled “The proneness of the primitive to confound the kingdoms and classes which we distinguish.” In some instances, Durkheim goes even further than Lévy-Bruhl in expressing difference. The latter would not write: “As long as men are still making their first steps in the art of expressing their thought, it is not easy for the observer to perceive that which moves them; for there is nothing to translate clearly that which passes in these obscure minds which have only a confused and ephemeral knowledge of themselves” (Durkheim, 1915, 96).

38In short, the Lévy-Bruhl that Durkheim criticizes is largely fantazised or constructed by him.

Did Durkheim Need to Criticize Lévy-Bruhl?

39Why is it the case that Durkheim feels the need to distance himself from the point of view that, with or without reason, he ascribes to Lévy-Bruhl? What is the purpose of claiming a real or alleged divergence three times? Is it a way to respond to objections that some of Lévy-Bruhl’s readers, by using the latter’s arguments, could have leveled against Durkheim? But in order to do so, one should have been able to find in Lévy-Bruhl claims that could warrant such objections. Yet, the divergence between them does not bear on the social nature of primitive thinking. The disagreement about the proper use of the term “concept” seems to boil down to empty talk: [14] whether the “collective representations” of “primitives” are qualified as concepts or not does not affect the specificities of the “preconnections” that Lévy-Bruhl attempts to describe. He indicates that it is a synthetic thought, in the sense that synthesis bypasses analysis: “prelogical mentality is little given to analysis.” According to him, this is where the specificity of “our” concepts manifests itself, as they have thought to draw on analyses that have already occurred, that are already given in language (Lévy-Bruhl, 1926, 106-107).

40It is not clear why the argumentative structure of Forms would be weaker if it did not have any reference to Natives, since the respective goals of the two books do not overlap. Whereas Durkheim’s ambition is primarily theoretical (it amounts, in the very same book, to renewing both the theory of knowledge and the theory of religion – and even philosophical anthropology [15]), Lévy-Bruhl’s goal, which we can legitimately characterize as phenomenological in inspiration (Deprez 2010, 37 and passim) is modestly descriptive, and even presents itself as an introduction to a more complete description. His theoretical ambition is limited to the testing of an explicit alternative hypothesis to the one that “the English school of anthropology” implicitly formulates about the “identity” of the “human spirit,” whose shortcoming consists in superimposing the observer’s categories on those that are to be observed.

41Now, on the one hand, by invoking forms of continuity between “religious thought” and science, Durkheim does not claim that they are “identical,”; on the other hand, Lévy-Bruhl’s working hypothesis never assumes a total difference: “there are features which are common to all aggregates of human beings…. The higher mental operations, therefore, have everywhere a basis of homogeneity” (Lévy-Bruhl 1926, 28). Given this a priori agreement between them, their divergence must be limited to the relative importance ascribed to what is common and what is different. Is it a simple matter of dosage? What is the significant issue for Durkheim?

The Issue of Evolution

42If the divergences created by Durkheim relate, as he says, to a difference of “point of view,” that difference, then, is what should be considered as their source. If one looks for a theoretical foundation for the divergence with Lévy-Bruhl that Durkheim keeps on invoking, it might be found in their more or less evolutionary interpretation of social differentiations. An evolutionary perspective – which minimally amounts to an explanation that must involve the observation of an oriented set of transformations – is inherent to the intellectual framework of Forms, which explains institutions in complex societies as deriving from those of “primitive” ones. “Primitive,” in Durkheim, does not just mean simpler (1915, 5-8), but also closer to origins, because “to study the most primitive and simple religion which it is possible to find,” one must look “to societies as slightly evolved as possible” (95-96).

43Lévy-Bruhl was influenced in his youth by the thought of Herbert Spencer (Lévy-Bruhl 1931, 5-6). However, when he ponders on how such admirable scholars as Tylor and his disciples could take their hypothesis of the invariability of human mind for granted, without questioning it, he thinks that the reason could be identified in the “influence of contemporary English philosophy, and the theory of evolution in particular,” more precisely of “the evolutionism of Herbert Spencer.” “At the present day … somewhat severely judged,” this type of evolutionism could seduce these authors thirty years earlier, as it provided them with “a guarantee for the continuity in man’s mental development which they formulate” and “‘savage’ races” appeared “to be the rudimentary germ of a succeeding state which shall be more differentiated” (1926, 26-27). In that respect, Lévy-Bruhl might have found that Durkheim was very close to the position of Tylor or Frazer. The criticisms of Lévy-Bruhl’s descriptions of discontinuity seem to be a response to analyses that reject an evolutionary view of social institutions, a view that Durkheim shares. In fact, what Lévy-Bruhl seeks in “primitive” thought is not an origin; it is an alterity, the mode of thinking that is most different from “ours.”

44If this interpretation is correct, a paradox emerges. If, in the series of the six books that Lévy-Bruhl dedicated to explaining the specificities of “primitive mentality,” there is indeed a moment when he may appear to endorse the evolutionary theme, it is in the first book (the only one that Durkheim could read and critique). It is only in Natives that he explicitly questions the forms of transition to other types: the ninth and last chapter, representing the fourth part of the book and its general conclusion, is entitled “The transition to the higher mental types.” It contains statements very similar to those Durkheim uses against him:

45

But it is not only the study of inferior races that comprehension of this prelogical, mystic mentality helps. Subsequent mental types derive from it, and cannot avoid reproducing, in forms more or less apparent, some of its features. To understand these, therefore, it is necessary to refer back to a type which is comparatively ‘primitive’. A vast field for positive research into the mental functioning of aggregates of various kinds, as well as into our own laws of thought, is thus laid open to us.

46Undoubtedly, this chapter is quite programmatic and cannot support strictly evolutionary beliefs, but the paradox is even more complex because, in Forms, Durkheim does not provide any comparable indications to outline the transition from “primitive” forms (of religion as well as knowledge) to the contemporary forms within Western societies. In any case, this conclusion shows that, from 1910 onward, Lévy-Bruhl remained convinced that “prelogical” elements do not stop coexisting with “logical” ones. For this reason, the thesis of the structural unity of the human mind, proposed as soon as 1903, [16] is maintained despite the “working hypothesis” (Lévy-Bruhl 1926, 362) of differentiation.

Is Religion Primitive?

47Besides the importance ascribed to evolution as an explanatory model, there is another point of disagreement, which remains hidden and about which Durkheim remains silent, although it directly affects the theory of Forms. It has already emerged through the terminology Durkheim uses to retranslate Lévy-Bruhl, by identifying “primitive” and “religious” mentality: another paradox of this controversy is that Durkheim considers the fact that “primitive mentality is essentially religious” (Durkheim 1913, 35; 1915, 432) as one of the points he shares with Lévy-Bruhl. Not only does Lévy-Bruhl abstain from using the vocabulary Durkheim ascribes to him, but in Chapter 9 of Natives he is very explicit in his reservations about the use of “primitive religion,” since “religion” presupposes a dissociation that is not “primitive,” but is already the product of a transformation:

48

We might say equally well that the mentality which expresses itself in their collective representations is wholly religious, or, in another sense of the word, that it is hardly at all so. In so far as a mystic communion with, and actual participation in, the object of the religious sentiment and ritual practices is of the very essence of religion, primitive mentality must be declared religious because it does realize a communion of such a nature, and indeed to the highest degree it is possible to imagine. But in other respects it does not seem correct to speak of it as ‘religious’, at least to this extent, that by reason of the direct character of this participation it does not recognize as an ideal outside and above itself the beings with whom it feels itself united in mystic and intimate communion. We may recall the definite statements made by Spencer and Gillen on this point…. But to the primitive mind … these objects and these beings become divine only when the participation they guarantee has ceased to be direct. The Arunta who feels that he is both himself and the ancestor whose churinga was entrusted to him at the time of his initiation, knows nothing of ancestor worship. The Bororo does not make the parrots, which are Bororo, the objects of a religious cult…. The ideas which we call really religious are thus a kind of differentiated product resulting from a prior form of mental activity.

49By calling it religious, Lévy-Bruhl fears that one would project something that is external (and inherent to the gaze of the modern observer) on “primitive” mentality, because it corresponds to a situation of maximum overlap between “the individual consciousness of each member of the group” and “collective consciousness.” It is in a different social situation, “when the relations between the social group and the individuals composing it are evolved” (Lévy-Bruhl 1926, 365), that

50

“the communion … will be obtained by means of intermediaries. The Bororo tribe will no longer declare that they are araras. They will say that their ancestors were araras, that they are of the same substance as the araras, that they will become araras after death, that it is forbidden to kill and eat araras, except under conditions which are rigidly defined, such as totemic sacrifice, etc.”

51Therefore, whereas Durkheim considers that modes of thinking derive from religious life, Lévy-Bruhl takes it that religion (or at least what we place under this term) appears to come after “primitive” modes of thinking. It may be this very detail – which Durkheim seems to wilfully ignore, judging from the terms he uses to translate Lévy-Bruhl – that might have convinced Lévy-Bruhl that he did not partake in the theory developed in Forms. For, although he does not respond to the criticisms, leveled in that book, he does indeed return to that question later and note that Durkheim’s use of the term “religion” seems too comprehensive. He proposes the “neologism” of “prereligion,” for which he apologizes, not in order “to contrast them”, but to avoid “projecting characteristics which only appear in more advanced societies on to the quasi-religious facts which are observed in these societies” (Lévy-Bruhl 1935, 217-218; see Merllié 1998).

52“Primitive” thought is not just “prelogical,” but also “prereligious.” Even more paradoxically, Lévy-Bruhl’s argument is very Durkheimian in spirit: its premise is the underdeveloped nature of individuality in these societies and its coalescence into collectivity before the advent of individual differentiation, a process that appears as the impetus of both religions and strictly “conceptual” thinking.

53Readings of Forms generally tend to dissociate the two objects of the book, focusing, depending on the interests of the readers, either on the sociology of religion or on sociology of knowledge. The disagreement with Lévy-Bruhl on which Durkheim draws seems to bear entirely on the latter. Yet, his analysis leads us to the former. In that regard, it should be noted that these two objects, in Durkheim, tend to conflate with one another and to be considered jointly – with religion having also a cognitive function. The merging of these two objects manifests itself in the recurrent use of the terms “religious thought” and “religious mentality.”

54From the comparison of these texts, three main conclusions can be drawn: i. Durkheim’s objections to Lévy-Bruhl concerning three central points of the latter’s description of “primitive thought” are all intended to minimize the gap between “modern” and “primitive” thinking – and, for Durkheim, between science and religion. These criticisms distort what Lévy-Bruhl says on these topics; ii. the underlying theoretical issue seems to relate to Durkheim’s genetic explanation, which distances itself from Lévy-Bruhl’s descriptive or ideal-typical project; iii. at the core of this difference in “point of view” lies a concealed disagreement about the definition of religion and the primacy of this social institution.

55Have commentators grasped these points? To answer such a question would lead to “another story”. Because of the one sided nature of the controversy, they reiterate and strengthen the opposition between separation and continuity, an opposition that Durkheim had already stated and emphasized. For instance – in an account that at least renews the theme of the “abyss” – the term “Manichaeism” is strangely used to describe Durkheim’s criticism of Lévy-Bruhl (Mucchielli 1998, 354). The issue of evolution, identified by Georges Davy, for whom “the notion of progress … does not play … the same role” in the two authors (1930, 284), is forcefully elaborated by Horton, who sees in the opposition between “contrast/inversion schema” and “continuity/evolution schema” (1973, in particular 270) a structure of analysis that can be extended beyond the two authors. The issue of the relation between science and religion, however, is scarcely ever mentioned; it is at least noted by Raymond Lenoir (1922, 217-218). Among more recent commentators, Frédéric Keck (2008, 164) proposes to read Natives “as a response to the analysis of Forms”; whereas Stanislas Deprez (2010, 249) argues that it is Durkheim who diverges from Lévy-Bruhl more than the other way around.