1From Parsons on (cf. Parsons 1937), exegetes and critics have questioned whether there might be continuity or rather a break in Durkheim’s intellectual biography. Even today it is not an easy task to provide a definitive answer to this question. Philippe Besnard (2001) recently explored once again the years of the famous «change of orientation» in Durkheim’s thought – which Durkheim himself assigned retrospectively to the year 1895 (cf. Durkheim 1907a: 404) – to suggest that if the private and familiar context were considered, there are a good numbers of elements to confirm the hypothesis of the break; on the other hand, if we take into consideration the polemics with Gabriel Tarde, there is an important feature which links the years before to those after the decisive year of 1895 (cf. also Besnard 1995). According to Besnard, it is possible to condense the events of the private context (that is, those of discontinuity) around a single date:
Cette date est celle du 11 avril 1897, dimanche des Rameaux et premier jour des vacances de Pâques. Ce jour-là, Durkheim déménage; c’est un petit déménagement, pour la distance: il quitte le 179 boulevard de Talence à Bordeaux pour le 218 du même boulevard, dans une jolie maison qui existe toujours. Mais la distance importe peu. Quand un grand esprit déménage, ce n’est pas un petite affaire. Il faut ranger ses papiers, jeter beaucoup de choses. Une page se tourne (Besnard 2001: 29).
2Moving out as an epistemological caesura. A powerful and suggestive image, given its capacity to connect the apparently anecdotal dimension of familiar intimacy to a possible «moment of radical change» of thought that would have influenced the social sciences of the XXth Century. «Setting up one’s own papers», «getting rid of several things», «turning a page» for Durkheim means, according to Besnard, also becoming liberated from the massive collections of juridical statistics on suicides, so as to make room for ethnographical data. It is due to these data, assumes Besnard, that Durkheim will finally find the way to «approach religion sociologically» (cf. Durkheim 1907a: 404). An assumption that may seem at first sight supported by the sudden change of attitude in respect to ethnographical data itself: while in 1895 Durkheim questions its validity (cf. Durkheim 1895a, 1895b), during 1897-98, he will put it on the same ground with the information provided by history (cf. Durkheim 1897a, 1898a).
- 1 For a more detailed demonstration of what I argue in this paper see Zerilli (2001).
3Along with this intriguing Besnardian reading, I would like to suggest another one, rather contextual than textual, which aims to analyze Durkheim’s apparently contradictory attitude with respect to ethnography. I anticipate right away that by taking into analysis the dispersed occasional writings which Durkheim devoted to ethnography along his career, it is possible to elaborate a hypothesis that points instead towards the continuity of his thought1.
4Despite its much more detailed and accurate chronology, Besnard’s reading belongs to an exegetical tradition that finds in Claude Lévi-Strauss an illustrious forerunner. In a well known commemorative article (cf. Lévi-Strauss 1960), in which he aimed to outline to which extent modern ethnology was indebted towards Durkheim, Lévi-Strauss was actually among the first scholars to draw attention to the contradictions existing between a few excerpts of the Rules of sociological method (1895) and some statements of the preface to the Elementary forms of religious life (1912) referring to the importance given to ethnographical data. A time span of almost twenty years, Durkheim’s attitude was thought to have «altered» to the point that Lévi-Strauss considered justified speaking of an actual «conversion» to ethnography. The key element of such a transformation was indicated by Lévi-Strauss through understanding the relations between ethnography and history: while at the time of the Rules the former would be subordinated to the latter, in 1912 «for the first time the two methods have been placed on the same level» (Lévi-Strauss 1960: 58).
5Indeed, before having expressed some doubts with respect to ethnographical sources, Durkheim had underscored many times their basic role for sociology, comparing them to historical documents. To convince oneself on this matter it is enough to take a close look at the first lectures that Durkheim gave at the Faculty of Letters at Bordeaux University, beginning with 1887/1888. In the introductory course of his second Bordeaux lecture (1888/1889) dedicated to the family («La famille: origines, types principaux»), Durkheim claims the necessity to eliminate travel writings and missionary reports, documents «which are not objective enough» (Durkheim 1888b: 18). For Durkheim, in order to be objective, a document should make evidence of those customary and constant practices which can be found «in those ways of acting strengthened by usage denominated customs, law, traditions» (ibid.). Among these law, as far as it possesses «a clearly defined form», continues Durkheim, «represents generally a much more valuable document» while «it is indicated not to make use of traditions but cautiously since they are somehow more indecisive and ephemeral» (ibid.: 20, italics mine). Nevertheless, a few lines later he affirms that when carefully established even traditions «may provide highly useful teachings» (ibid.). And it is on this very aspect that Durkheim focuses and on which he attempts to draw our attention. Not accidentally, he adds that if Morgan had managed to reconstruct «the exogamic clan», it was thanks to the study of the «daily greeting formulas adopted by the members of certain families in Asia, America, and Australia» (ibid.), in other words to ethnography. In short, while he admits that encoded sources such as law are generally more trustworthy, Durkheim wishes to stress that the information resulting from the observation of ordinary practices among primitive peoples may also reveal itself not only extremely useful, but even crucial for the identification of social phenomena.
6Regarding the study of the family, Durkheim underscores in particular the relevance «of the works produced by ethnologists [ethnologistes] and historians on certain specific aspects of the history of the family» (Durkheim 1888b: 21), and at the same time shows himself to be up to date with the utmost recent ethnographical investigations at issue. But if the works conducted by historians and ethnologists on specific aspects are essential, neither the former nor the latter, notices Durkheim, are able to provide the quantitative information necessary for studying a society as a whole. It is for this reason, especially when taking into examination the contemporary family, that one should also rely on the statistical information provided by demography. Hence, we have identified the «triple source» on which Durkheim considers necessary for laying the foundations of the sociological study of the family:
The law, the traditions as we get to know them by means of ethnography and history, and finally the demography of the family, this is the triple source where we should turn to seek the material of our inductions (Durkheim 1888b: 23, italics mine).
7Interestingly, almost ten years before writing the preface to the Suicide, history, ethnography and statistics are already conceived as the three «auxiliary disciplines […] without which sociology cannot do nothing!» (Durkheim 1897a: 45, exclamation mark in the original). It should be stressed that the «traditions» (mœurs) belong both to the field of history and ethnography, and that the latter is opposed to the travel writings. In fact, while travel writings are considered to be misleading, either when elaborated by an external observer or when they stick closely to the indigenous point of view (cf. Durkheim 1888b: 18), ethnography represents – the same way history does – a valid method to collect information, even if it is based on the simple direct observation of primitive societies. As we shall see further on, Durkheim will subsequently match together ethnographical information and travel writings. In this stage, it is the other way round: on one hand, there is ethnography, which like history is assigned the status of science auxiliary to sociology, on the other hand, there are the witness accounts of travelers and missionaries, which are regarded with skepticism.
8During the inaugural lecture of the «Cours de science sociale» of the previous year (that is, the first year of his university career), Durkheim establishes a reconstruction of the history of sociology recalling the names of the scholars who have brought a main contribution to its development, also indicating the directions to be assumed for the future. When it comes to taking into analysis Comte’s contribution, it is of high significance that Durkheim should contend with his forerunner for the insufficiency of the sources used to highlight the functioning of his «three stages» law:
Auguste Comte contented himself with briefly reviewing the main events of German-Latin peoples, without realizing how much controversial would be establishing a law of such importance on such a limited basis (Durkheim 1888a: 89).
- 2 It is worth quoting this text: «The great novelty of the [XIXth]Century […] is the apparition of a (...)
9The reason is that Comte, explains Durkheim, «was encouraged in this perspective by the rudimentary stage characterizing ethnological sciences in his time, and by his lack of interest for these disciplines» (ibid.). This statement leads to think not only that Durkheim was at the time already cultivating a vivid interest for ethnological sciences, but also that he considered that those sciences had made significant progress and hence acquired more credibility, during the time span separated him from Comte. It does not seem exaggerated therefore, to affirm that, for Durkheim, ethnographical sources have represented from the very beginning of his university career a relevant source for the renewal of sociology on the whole. From this point of view, the statements regarding the relations between sociology and social sciences given with Paul Fauconnet in the Revue philosophique (1903), right in the heart of the so-called ethnographical phase, are not to be considered as the expression of a radically new season, but rather as a reformulation of ideas previously elaborated2.
10Let us now return upon the lecture dedicated to the family, and in particular, to the third section where Durkheim addresses explicitly philosophers, historians, and jurists in order to make them aware of the value of the sociological prospective. It is actually here, in a context defined by tough disciplinary competition (cf. Craig 1983), that for the first time he openly approaches the puzzling issue of the relations between history and sociology. For Durkheim, the difference between these two disciplines is reduced substantially to the use each one makes of comparison. While for the historian «the field of possible comparisons is extremely limited» (Durkheim 1888b: 29), since he «confines himself most of the time to the study of one nation alone» (ibid.), for the sociologist this field manifestly extends since he aims explaining particular facts by comparing multiple societies. It is true, Durkheim admits, that authors such as Fustel de Coulanges or Summer Maine have extended the domain of their historical investigations to the study «of analogous institutions examined on the background of different societies», but «these comparisons cannot be fruitful except for those cases which are made on a large scale» (ibid.) This means, explains Durkheim, that
…the lowest [social] species should not be ignored. In this way, the domestic law of Australian or American tribes will let us better understand the ones of the Romans. But the historian has remained by now rather alien to ethnography; he does not even go beyond the classical peoples (Durkheim 1888b: 29, italics mine).
- 3 This is the excerpt referred to by Besnard himself (2001: 30): «Fustel de Coulanges, despite his de (...)
- 4 On the abundant existing literature on Durkheim’s complicated and changing relations with history ( (...)
11It is worth noting that the meaning of these remarks is basically the same of some passages of the introduction to the first volume of L’Année sociologique3. Furthermore, they preannounce the criticism Durkheim will address against historians in the preface to the Forms, namely their lack of interest in ethnography (cf. Durkheim 1912: 9). Compared to the year when he writes these lines (remember we are in 1888) what changes ten years later, in 1898, seems to be more his ideas towards history (and particularly a certain kind of history) rather than his attitude towards ethnography4. In the context of the late1880s, ethnography does not only appear to Durkheim as an integral part of the sociological prospective, but it also assumes the function of overcoming the too narrow limits of historical comparison. Not surprisingly, only two years later in 1890, within the context of another university course dedicated to the «physique des mœurs et du droit», Durkheim clearly reaffirmed the central role of compared ethnography, placing it again side by side with history and statistics:
…the methods employed by the physics of customs and of law are of two types: in the first place there are the compared history and ethnography […]; in the second place the compared statistics (Durkheim 1950: 5, italics mine).
12The acknowledgement of ethnography’s importance, a legacy that Durkheim inherited from Spencer (cf. Borlandi 1993: 103), should then be backdated. According to what has been shown, rather than identifying a pre-ethnographical phase and an ethnographical one, and asking oneself consequently how and why did Durkheim reach to ethnography (cf. Lévi-Strauss 1960; Lukes 1992 ; Vogt 1976; Lacroix 1981; Besnard 1986; Karady 1988), I argue that is more rewarding and interesting to understand how at a certain moment of his career, Durkheim seems to deny the value he had once recognized to ethnography. Some traces of this apparent rethinking may be found in the writings of the early 1890s.
13Since his first book the Division of labour in society (1893), Durkheim dwells upon the difference of status between written sources and sources deriving from the simple observations of customs, underscoring the higher reliability of the former (Durkheim 1893: 109). Two years later in the Rules (1895), he reaffirms that ethnographical information cannot have but a secondary role with regard to historical sources (Durkheim 1895c: 132). Taken as such, these statements have been interpreted as a warning to distrust ethnographical sources, or as an invitation to assign them all the more a complementary function. Once reset in their context, they appear more as a polemical stance against the interpretations of those sociologists who have been «preoccupied to merge documents rather than to criticize and select them» (ibid.). We should ask to which authors did Durkheim make reference in these lines.
14Steven Lukes (cf. Lukes 1992 : 302-313) and subsequently Philippe Besnard himself (cf. Besnard 1995) have pointed out that the scholar absorbing Durkheim’s polemical energies between 1893 and 1897 is especially Gabriel Tarde, and this for various reasons, either theoretical or related to professional considerations. Nonetheless, besides Tarde there are certainly other scholars as well, two of whom are of particular interest here: Edvard A. Westermarck and Charles Letourneau. To them Durkheim addresses critical pages the same year when the Rules was published as a book. In a long review in the Revue philosophique, Durkheim submits Westermarck’s The history of human marriage (London, 1891), (the French translation of which had been recently published as L’origine du mariage dans l’espèce humaine, Paris, 1895) to a detailed critical analysis. The main evidence Durkheim places before Westermarck is of a methodological nature and takes a close look to ethnography: Durkheim criticizes him, in fact, for having employed an «essentially ethnographical and psychological method» and for having neglected the historical sources, while «it is only by connecting ethnographical facts to historical facts that it is possible to eliminate the ambiguity of the former» (Durkheim 1895b: 74). Durkheim addresses similar objections to Letourneau in the article «Lo stato attuale degli studi sociologici in Francia», also published in 1895 in the journal directed by F.S. Nitti and L. Roux La riforma sociale. In this important text originally published in Italian, Durkheim offers an outline of major French sociological trends and identifies three main groups: the anthropological (or ethnographical), the criminological, and the university (cf. Durkheim 1895a). It is the first group that he attributes Letourneau, whom he criticizes, as in the lecture on family (cf. Durkheim 1888b: 20-21), for having always avoided to submit the selection of his own materials to a harsh criticism. On the other hand, affirms Durkheim, «the travelers’ descriptions are the source to which [Letourneau] abundantly resorts» (Durkheim 1895a: 77).
- 5 «Mister Westermack considers that the quantity of information may compensate the mediocre quality [ (...)
15It should be remarked that for the first time, in the criticism against both Westermarck (cf. Durkheim 1895b: 71) and Letourneau, Durkheim seems to settle travel writings and ethnographical data at the same level. While he had posited a clear distinction between them in the past, contesting the value of the former and admitting the interest of the latter, now the two types of sources at least partially overlap. Is it really a reconsideration, a simple confusion, or is it rather a strategic option in order to bring stronger arguments to his dissension regarding the perspectives of these authors that makes Durkheim doubt of the contribution ethnography may bring to sociology? First, it should be recognized that under no circumstance, neither in the Rules, nor in the objections addressed to Westermarck and Letourneau, Durkheim completely denies the utility of ethnographical sources («no fact can be neglected by the scholar», Durkheim 1895c: 132). Second, in this context he underlines rather that they are less useful than history, and that in any case they should not be taken separately from the latter («these two sources of information cannot be consulted separately», Durkheim 1895b: 74). It is clear then that the main objection has a methodological character and refers to the indiscriminate use of ethnography; more precisely, Durkheim contests the absence of a critical analysis of the sources (this refers mostly to Letourneau), and the idea that it is sufficient to multiply the number of examples in order to reach conclusions of scientific interest and value, avoiding misinterpretations (and this refers mostly to Westermarck)5.
16Of course, Durkheim’s criticism is much richer and articulated with respect to what can be mentioned here, and refers to various other specific aspects of the works taken into examination. His criticism actually faces two very different authors, as different are the leading objections which Durkheim casts against them: to Westermarck he contends specifically the thesis of the substantial uniformity of marriage forms in time, and the notion of marriage itself (cf. Durkheim 1895b: 78-80), while in the case of Letourneau he criticizes specifically the ideological character of his scientific perspective, which after all determines and explains his frequent recourse to information which cannot be easily verified (cf. Durkheim 1895a: 78). But actually there is at least one point with respect to which these two authors may be compared according to Durkheim: while preserving their peculiar research interests, they both systematically establish a direct association between primitive peoples and the psycho-organic (biological) level. As for Westermarck, Durkheim considers:
To admit that ethnographical documents in order to be understood, need to be re-conducted to man’s primitive nature, and that the latter may even be reconstructed according to what we know on animal upper species, would mean admitting as an incontestable axiom that our physical constitution, and even our animal nature, namely that part of us which most immediately depends on the organic conditions, is the main source of social life (Durkheim 1895b: 72).
- 6 Principle clearly stated since the first Bordeaux lecture: «social facts cannot be entirely explain (...)
- 7 In his L’évolution du mariage et de la famille, Letourneau wrote: «Human societies have been too mu (...)
17On the other hand says Durkheim, Letourneau, in order to remain «faithful to the spirit of the anthropological school» does not give up establishing correlation between «social institutions» and «various human races, following their evolution in time» (Durkheim 1895a: 77). In other terms, referring in last instance to hypothesis borrowed from natural history and evolutionary theory in order to explain social facts, both Westermarck and Letourneau were undermining the very basis of the durkheimian epistemological project, that claimed autonomy and scientific legitimacy for sociology and affirming the chief «rule» according to which «a social phenomenon cannot be reproduced unless by another social phenomenon» (Durkheim 1895a: 101, italics in the original)6. Relying largely on ethnographical sources and sharing a similar concept of ethnography, Westermark and Letourneau were instead reintroducing organic and psychological factors among the primary causes of social facts7.
18While Durkheim’s criticism of Westermarck is familiar, it is not widely known that Westermarck responded to Durkheim’s objections. Interestingly, his answer appears in a review article published by the Revue internationale de sociologie discussing the work of J. Kohler on the method of prehistorical research (cf. Westermarck 1897). This is an important text for it helps us to better understand how, behind the polemics concerning the nature of sociological explanation, also lays a radically different idea of what ethnography is and how it can be used.
19After having acknowledged to Durkheim that in order to certify the «primitiveness» of any social practice it is necessary to put historical and ethnographical data on the same level, Westermarck indicates where such «primitiveness» should be looked for in case these data were missing:
The problem is to know how can we acquire an amount of information in case history were silent, and in this case it is not only ethnography, but biology and psychology as well that should come and rescue us (Westermarck 1897: 452, italics mine).
20For Westermarck, as for Letourneau, ethnography is hence called to «fill in the gaps of history» (Letourneau [1881-1889]: 1014). Conceptualized as the study of primitive populations, ethnography was seen as a sort of bridge linking history on one hand, and biological and psychological sciences on the other. When history is not able to provide satisfying answers, it is necessary, Westermarck argues, to resort to ethnography, and then to biology and psychology. Moreover, for Westermarck each of these disciplines are not a separate field of inquiry, they are instead related to each other within the natural history of man. This perspective leads Westermarck to bring forth in his answer a crucial point of the polemics, namely Darwin’s theory of evolution. If for Durkheim it represents nothing else but a simple «hypothesis», and as such it may not stand for the foundation of sociology (cf. Durkehim 1895b: 73), for Westermarck it constitutes a scientifically certified fact with powerful implications for sociology as well. Westermarck’s adhesion to Darwin’s theory leads him to mock Durkheim and even to ask whether it is worth having a confrontation with him. Durkheim’s objection, Westermarck writes,
…should sound strange for anyone who is at least a little aware of the huge progress biology have accomplished on the Darwinism base. And I have to admit that I find it difficult to enter into a polemic with an author who thinks ‘contrary to any good method’ the assumption which considers man as descending from an animal of lower species (Westermarck 1897: 452).
21Durkheim’s reply is included in a long and famous (cf. Lowie 1937: 197) review of Koeler’s same work reviewed by Westermarck, which appeared in the first volume of L’Année sociologique, that is almost three years later (cf. Durkheim 1898b). Contrary to Westermarck, Durkheim restates the idea according to which kinship and marriage are facts of an eminently social nature, and should be interpreted as such. Moreover, in this text Durkheim challenges Westermarck’s attacks assimilating his perspective to «common sense»:
Westermarck, voicing Starcke, while refuting Morgan’s thesis he admits that nomenclatures terms correspond to a social kinship relation, which he opposes to natural kinship relation. Yet, each family relation is a social one, since it consists basically of juridical and moral relations, sanctioned by the society. It is either a social relation or nothing else. If Westermarck has made this distinction it is because for him, as for common sense, authentic kinship is consanguinity and just expresses it. But when one get rid of this confusion, there is nothing else but a kinship: the one acknowledged as such by the society» (Durkheim 1898b: 113).
- 8 Invited to collaborate to the Revue internationale de sociologie, in June 1894 Durkheim writes thes (...)
22If one takes into consideration the criticism to which the Rules have been subjected to when they were written (cf. Berthelot 1995: 121-141; Paoletti 1995), one realizes that the principle of the autonomy of social facts – far from being unanimously acknowledged – was contested even at the core of the sociological field itself. Particularly, it was contested by sociologists influenced by organicist theories, a group well represented by René Worms who was at the time one of the most active scholars promoting the institutionalization of sociology (cf. Geiger 1981). Even if he had never been mentioned, it is against Worms that point the initial pages of Durkheim’s article on sociological studies in France that appeared in La riforma sociale. Worms himself, already internationally acknowledged at the time «as a warm apostle of sociology in Europe» (Fiamingo 1896: 341), had collaborated to the Italian journal with an article published in the first volume, dedicated to the teaching of social sciences within French universities. In this article, he was foreseeing the creation of autonomous faculties of social sciences that would have been independent from the faculties of letters or law, and where sociology was making considerable efforts to be acclaimed (cf. Worms 1894). When Durkheim evocates «particularly restless spirits» and raises doubts as for the necessity of establishing sociology chairs in each university (cf. Durkheim 1895a: 74), he is in fact opposing the reformation of the university system suggested by Worms in the columns of La Riforma Sociale. In the same paragraph, the polemical reference regarding the «premature publication» of a sociological journal (ibid.),makes without doubt allusion to the creation of the Revue internationale de sociologie founded (since 1893) and edited by Worms himself, an author for whom Durkheim had, so to say, poor scientific consideration8.
- 9 In the same Italian article, the severe criticism against the Tardian theory of imitation is to be (...)
- 10 The same year of the publishing of the French translation of Westermarck’s The history of human mar (...)
23Gabriel Tarde, who also collaborated periodically to Worms’ journal, is surely the most dreaded antagonist, therefore Durkheim’s favorite target9. Nevertheless, one should consider the fact that in the changing sociological context of the late 1890s there were a good number of other groups and tendencies which Durkheim considers he had to face openly. While at the time of the first Bordeaux lectures he particularly strived so that sociology should be accepted and acknowledged, now he attempts to impose his own sociologicalconception. The scholars who at the beginning of his career could have been perceived as simple interlocutors, even as allies in order to fight a common battle for the legitimacy of a new field of study that, «especially within the context of the university milieu […] was subject to a veritable process of discredit» (Durkheim 1895a: 73), are now seen as potential adversaries, more or less dangerous. Among these, Letourneau is the main representative of a scientific orientation (namely the ethnographical or the anthropological) to which Durkheim is now prepared to firmly oppose. On the other hand, Westermarck is already an internationally well-known scholar whose reputation is sustained in France by the sociological group positioned close to Worms’ circle and institutions10.
- 11 It is worth mentioning that the French translation of the article appeared in La riforma sociale (c (...)
24In order to develop their thesis, we have seen that both Westermarck and Letourneau largely rely on ethnography, a discipline which can be hardly considered a neutral source. Along with archeological, somatic (anthropometric and especially craniometric), and linguistic data, ethnographical data had been traditionally a domain in which the theories of anthropologists confronted each other, either to support or to deny the hypothesis of the multiple origin of human races (cf. Blanckaert 1981). Particularly in the late XIXth Century French scientific context, ethnography constitutes a research field strongly influenced by the naturalist anthropological tradition (cf. Zerilli 1998: 1-25). Durkheim’s article in Italian mostly mentioned, as well as various excerpts and an entire chapter of the Suicide (cf. Durkheim 1897b: 108-132), suggest that he was aware of this intimate and problematic relation between ethnography and anthropology, the latter being conceived as the natural history of man11. At the moment when Durkheim takes position against Letourneau’s «anthropological sociology» and rejects «the considerations that M. Westermarck borrows from natural history» (Durkheim 1895b: 80) he is referring to this historically well established tie. In this context, subordinating the value of ethnographical sources to historical data, specifically means restating the distinctive character of «social facts», and at the same time rejecting organic explanations widespread mostly within the anthropological (i.e. racial) scientific milieu.
25To conclude, in order to understand the criticism against Westermarck and Letourneau, when Durkheim brings forth his utmost severe objections towards ethnographical sources, one should keep in mind the fact that ethnography, far from being an autonomous and independent field of study, is an auxiliary discipline which is used by scholars belonging to various research traditions with respect to their scientific objectives and purposes. In particular, the recourse to ethnography by the representatives of naturalist anthropological tradition has largely contributed in disseminating the idea that primitives identify with human «state of nature», assigning an ambiguous status to the ethnography of the time: that of a field of study located between history and natural sciences. Moreover, research hypothesis elaborated within the naturalist anthropological tradition, like those assigning to ethnography the task of bringing light on human nature from the somatic point of view, were not unfamiliar to Durkheim’s competitors in the sociological domain. This is clear, for instance, if we look at the definition of the term ethnography given by Worms: «ethnography defines the type of race, deals with the study of the population and of the various grouping systems of the individuals» (Worms 1893: 448, italics mine). In 1895, the durkheimian recall to the supremacy of history over ethnography is a way of reaffirming and defending the thesis of the autonomy of social facts. In this context, Durkheim’s apparent skepticism towards ethnographical data should be interpreted as a denial to inquire complicated and unresolved issues to which it symbolically alludes (the organic composition of races and their potential influence on social facts), rather than a reference to the difficulties of studying primitives by way of direct observation, a problem of which Durkheim was also increasingly aware. Furthermore, if we consider that – as with Tarde – Westermarck and Letourneau were assiduous collaborators of Worms’ journal, being at the same time actively involved in the activities of his international sociological association, the criticism addressed against them also evokes the institutional difficulties and weakness of durkheimian sociology during the years preceding the foundation of L’Année sociologique, as outlined by Philippe Besnard himself with sober lucidity (cf. Besnard 1979, 1998).