Remembering his own intellectual career, Raymond Boudon reminds us that Durkheim was his ‘first great admiration’. He read Le Suicide ‘in detail’ during the year he spent at Columbia University (Boudon and Leroux 2003: 32). In his complementary thesis, À quoi sert la notion de structure?, published in 1968, Boudon criticizes structural functionalism, but he also make an apology for Le Suicide. ‘If there is a revolution in modern sociology, it must undoubtedly be explained, not by structural-functionalism, but from Durkheim's Suicide, which demonstrated that by giving circumscribed objects, the sociologist could place himself in a quasi-experimental position and build theories with a high level of verification’ (Boudon 1968: 216). Following Raymond Aron, who never failed to criticize the great French sociologist, Boudon, as we will see, gives Durkheim a good place as a pioneer of sociology.
It must be emphasized that his admiration for Durkheim is not fleeting; it is durable, from his first to his last works. His Durkheimian readings were a source of primordial influence in most of Boudon's writings. But it must be said that his vision of Durkheim has evolved a lot, to the point of radically changing. In the 1960s until the 1990s, he presented Durkheim as a positivist, fairly close to Auguste Comte, and he considered Les Règles de la méthode sociologique a mediating work that announced works of the great sociologist of Épinal. Boudon is therefore only taking up a thesis taken for granted by historians of sociological thought. Thus in his Traité de sociologie (1992), he notes that
without noticing it, Durkheim endorses a certain number of ideas of Auguste Comte that seem to him to go without saying, as to many of his contemporaries: he borrows from him his ‘totalist’ or ‘holistic social’ vision, as well of his evolutionary vision of history, his classification of the sciences. It is indeed impossible to understand Durkheim's conception of sociology if we do not see that it is rooted in Comte's classification. In any case, Durkheim takes up literally, consolidates and legitimizes the Comte's exclusions: history is not part of the sciences, psychology manifests illegitimate claims to the rank of science, economics is only an incarnation of metaphysics. (Boudon 1992: 24–25)
As a good Comtian, Durkheim considered history and psychology condemned. It is rare that Boudon cares to show the filiation between Comte and Durkheim so explicitly without offering some critical comments.
There are passages everywhere in Boudon's work from the 1970s where he discussed rationality and action. He has not yet started to speak of the principle of methodological individualism. From that moment, which can be dated around the beginning of the 1980s, he made Durkheim one of the main founders of this principle. This unorthodox reading drew considerable criticism. How can we explain it? One can suppose that it was by his reading of Max Weber, which it seemed he did not know of much before the 1980s. It seems clear that Boudon was encouraged to reread Durkheim through the prism of that important German sociologist and comprehensive sociology.
The first readings
In his first work, L'Analyse mathémathique des faits sociaux, while the use of mathematics is still the subject of strong reluctance among sociologists, Boudon presents Les Règles de la méthode sociologique as a complementary book to Le Suicide (Boudon 1967; Grémy 2018: 13–40). The young Boudon, then strongly influenced by the work of Paul Lazarsfeld, Robert Merton and other important US sociologists (Dubois and Mesure 2018: 41–63; Pawson 2009: 317–333) writes: ‘Les Règles de la méthode sociologique is an unintelligible book if one does not take care to read it by referring to the Suicide. Furthermore, the fact that Durkheim tried to fit his own logic into a form prepared to receive it explains why it did not immediately meet the success it deserved’ (Boudon 1967: 41–42). The Durkheimian model was undoubtedly too sophisticated to be welcomed; the statistical method being in its infancy at the time. For Durkheim to analyze a relationship, one must build a model of causal analysis. Thus, argues Boudon, ‘the importance of Durkheimian innovation lies in the proposition that a correlation can only have meaning if it is examined in relation to a causal structure. This notion of causal structure is undoubtedly what best distinguishes the observation situation from the experimental situation’ (Boudon 1967: 76).
About ten years later, when he explicitly claimed methodological individualism, Boudon decided to include Durkheim as one of the main precursors of this methodological principle. Thus in the Dictionnaire critique de la sociologie of which he is co-author with François Bourricaud, he says, concerning Suicide: ‘Thus Durkheim resorts, against his principles, to an individualist interpretation when he tries to explain why the periods of economic boom seem to be frequently accompanied by an increase in suicide rates: when the climate is optimistic, the individual may be encouraged to raise the level of his expectations, and, thus, to expose himself to the risk of disappointment’ (Boudon and Bourricaud 2004: 42).
This judgment may surprise those who are not familiar with Boudon's work. For him, Durkheim practices an individualistic methodology without knowing it. There are no traces of this idea in his work on Les Méthodes en sociologie (1969). Under the influence of Paul Lazarsfeld, Durkheim was presented by Boudon as a pioneer in quantitative methods and causal analysis in sociology. It should be noted that in 2002 a new edition of the book was published (co-authored by Renaud Fillieule). Causal analysis is seen there in a way as outdated, with limited explanatory potential. One could note that the notion of methodological individualism, which was not found in the 1969 edition. The 2002 edition is quite different. The authors wrote: ‘Methodological individualism makes it possible to formulate true explanations of phenomena while the causal explanation is incapable of it’ (Boudon and Fillieule 2002: 42). Between the 1969 and 2002 versions, there is a huge gap. In the original version, Boudon proposed a fairly orthodox reading of Les Règles de la méthode sociologique, which he would later forcefully reject. His work, in the original version, assumes that social facts are external to the individual. The individual is, in fact, devoid of rationality and power of action; he is completely determined. At first glance, this explanation may seem tautological. But in 2002, when his thought reached its full maturity, Boudon analysed Les Règles de la méthode sociologique1 and Le Suicide side by side. The phenomenon of suicide is explained by an effect of aggregation of individual behaviours; it gives meaning to human action.
For Boudon, Durkheim's work sparked two types of interpretation. The first, which is the most widespread, attributes to Durkheim, following Comte, the authorship of holism, of collectivism; the second, the one to which Boudon subscribes, translates the language of Durkheim into that of the sociology of action.
Let's show it in an example. Anomie, says Durkheim, is one of the main causes of suicide. It explains, for example, that economic expansion, by promoting the deregulation of expectations, generates dissatisfaction … In order to demonstrate that this theory can be expressed in the language of sociologists of action, we build a simple model intended to simulate the effects that Durkheim sought to highlight. This model must be considered a kind of extremely simplified model, in a pseudo-experimental context, compared to the real situations considered by Durkheim, some of the fundamental mechanisms that he sought to highlight. The model realises a very basic situation of competition between a certain number of people by studying the variations in the behaviour of the actors which result from variations in the conditions of competition. (Boudon 1979: 40–41)
There is no question here of claiming to be psychoanalysis, but to show what Boudon calls ‘an ordinary psychology’. He thus concludes that the motivations and the individual actions are unfathomable. For Comte as for Durkheim, the social sciences ‘are never interested, by definition, in explaining individual behaviours in their complexity. The individual is not, as such, their object’ (Boudon 1991: 54). That explains concepts like ‘collective consciousness’ or ‘suicidogenic current’ (Boudon 2012: 58).
According to Boudon, Durkheim now appears as an author subscribing to an individualist and rationalist perspective (see Boudon 1992: 204). It is hard to subscribe to this interpretation, and Boudon is fully aware of this. He admits it in the preface to the reissue of La Logique du social: ‘I have the impression today of having pushed the plug a little too far by annexing Durkheim to the methodological perspective that I defend here’ (Boudon quoted in Besnard 1987: 217). In general, Boudon speaks little about La Division du travail social, at least if we compare the work to Le Suicide and Les Formes élémentaires de la vie religieuse and even to Les Règles de la méthode sociologique. But we can retain some passages from La Division where he tends to overinterpret Durkheim's thesis in order to be able to insert it into his system. Implicitly, he says, ‘the deep analysis of La Division du travail social developed a certain kind of cognitive rationality. What Durkheim, using the vocabulary he had at his disposal, called free thought is indeed just another word to designate what the social sciences call cognitive rationality’ (Boudon 2012: 103). The expressions of ‘free thought’ and ‘reason’ therefore appear to be synonyms. For Boudon, Durkheim refuses to use the word ‘reason’ because it is polysemous. In this perspective, he attaches great importance to individualism, which is mentioned in a passage from La Division du travail social. For Durkheim, individualism ‘does not start anywhere’; this means, explains Boudon, that ‘in any society, the individual has a concern for himself, and that he recognizes the right and the capacity to evaluate the institutions in force and to judge them according to this concern’ (Boudon 2011: 228).
Boudon's conclusion is clear: there is nothing in this case, as elsewhere, to oppose Durkheim to Weber. While in his early work Boudon had not yet read Weber's work seriously, he will later build bridges between the German sociologist and Durkheim. He will even, in Les Formes élémentaires de la vie religieuse, interpret the work of Durkheim through the prism of the Weberian perspective (Boudon 1991: 53).
Reading Les Formes élémentaires de la vie religieuse
It is important to note that from the 1990s, Boudon spoke less of suicide to focus his attention on Les Formes élémentaires de la vie religieuse (Boudon 1999: 149–198; 2007: 193–248). He places this work at the heart of his theory of rationality, which he anachronistically describes as ‘very Weberian in character’ (Boudon 2012: 57–59). We find Boudon's interpretation of Weber's work mainly in the last 25 years of his career. He published a classical article on that topic in the L'Année sociologique.
For Boudon, Durkheim's work, even if certain observations are clumsy or out of date, remains lively, because of the power of the theoretical framework he proposes. Collective beliefs are not sui generis as some passages may suggest; rather they are produced by the activity of individuals.
Collective representations are illustrated by the example of scientific truths. They are the product of a complex process of innovation and selection. In this sense, they are indeed the product of society. The same thing could explain language: it is the product of society, a collected way of saying that it is the effect of initiatives leading to selection and dissemination mechanisms, the details of which can only be followed very incompletely. (Boudon 2000: 66)
For Boudon, it cannot therefore be said that Durkheim, who breaks with his own Règles, gives the impression that one has the feeling of a personification of society that exists only through individuals. ‘It is society that determines scientific truths, a new idea is subject to social filtering to which a multitude of minds contribute. But, in reverse, it is the scientist who proposes such a theory: society cannot do without individuals’ (Boudon 2000: 67). Society is therefore based only on ideas that emanate from particular individuals. Boudon proposes here a new way to understand the Durkheimian perspective: He wrote,
energetically refuses to portray the individual as passively accepting the truths that society would impose on him. On the contrary, he considers … that the human mind could be easily subject [itself ] to the illusion as unrealistic and gratuitous, and stresses that the truths and categories that distil society are subjected to a continuous process of settling and criticism. He also insists that, far from accepting institutions mechanically, the individual recognizes himself only if they make sense to him. This is true even with regard to rituals and cults. (Boudon 2000: 67–68)
Boudon's recent interpretation of Durkheimian sociology is mainly summarized in this passage.
The rational explanation of beliefs seems to meet perfectly the requirements of scientific knowledge. At the heart of Les Formes élémentaires de la vie religieuse, we could find a theory of categories. Durkheim, recalls Boudon, is a reader of Kantian philosophy; he declares himself hostile to empiricism. It has the advantage, compared to Kant, of expanding the list of categories of understanding and sensitivity. For Durkheim,
it is a question of explaining the origin, not only of the notions of time and space, or of the notion of causality, but also, of the notions of genre, force, efficiency, personality, etc. Like the notion of time, the notion of personality implies a connection between sensations: these notions suggest that, beyond the singular manifestations of the behaviour of such an individual, we can distinguish a principle of unity. (Boudon 2000: 69)
Durkheim on this basis retains the idea that certain categories impose themselves on individual behaviour. But the idea that Durkheim is inspired by Kant is not that new; the conclusion of Les Formes élémentaires de la vie religieuse clearly indicates this.
One of the most original aspects that emanate from Boudon's reading is the analysis he offers of the idea of rationality that is implicitly found in Durkheim. It leads to a ‘continuist theory’ of human thought, as highlighted by Gérald Bronner (2019: 173–175). ‘For Durkheim, as for Weber, religious beliefs as well as magical beliefs, must be analyzed as rational’. This fundamental idea was found at the beginning of the nineteenth century: philosophy and science were born from religion. Comte, like Durkheim, saw no gap between them. This is why we can say, according to Boudon, that the Durkheimian perspective is based on a form of ‘continuism’.
But there is more. Durkheim develops a sociology of knowledge that takes into account the principle that some categories are universal while others are not. But that does not mean there is a particular psychology for each person. On this basis, Boudon shows that Durkheim does not subscribe to any form of relativism. ‘The observer does not have the same knowledge of the subject that he observed. It does not use the same concepts, etc. … but if it is not possible to literally reconstruct the states of consciousness of social subjects, it is possible to discuss scientifically the feelings, reasons and motivations attributed to them’ (Boudon 2000: 66).
Good reasons to believe
The observer, following Weberian analysis, must put himself in the place of the subject. He must therefore refrain from judging beliefs and actions that may seem odd to him. ‘If the observer must be careful not to project his own knowledge, his own concepts into the mind of the observed, he must on the other hand start from the assumption that the primitive, like himself, will not be convinced unless by that which makes sense to him’ (Boudon 2000: 85). Thus the primitive rites are based on ‘good reasons’. In fact, behaviour that seems irrational to the observer will be rational for the observer perspective. For Boudon, this idea we find in a diffuse state in Les Formes élémentaires de la vie religieuse allows us to underline a possible link between Durkheim and Weber.
To explain a belief is to find the meaning for the actor of his belief, it is to suppose his rational belief or, in more concrete, and more immediately intelligible terms, to find the reasons that the actor is right to believe in. How to find these reasons, when the belief seems at first sight irrational? By ignoring the cognitive data that the observer gets from his own situation, but which he clearly sees that he cannot attribute to the observed. (Boudon 2000: 86)
Boudon explains that, for Durkheim, religious beliefs are not hallucinations or illusions. They are based on ‘good reasons’ even if these reasons might seem irrational. To better understand Durkheim, Boudon read the work of the French sociologist in the light of Weber's analyses. Like Weber, Durkheim proposes ‘consider[ing] that human thought is one. The primitive and the child use the same rules as the modern adult, but their knowledge, their interpretations of the world and some of their categories are different’ (Boudon 2000: 88).
Therefore Boudon takes up, in a certain way, an idea which is central to Auguste Comte's law of the three states, although without quoting it, namely that religious beliefs are explained in the same way as scientific beliefs. They adhere to theories that aim to explain the real, even if, in certain cases, they may seem incomplete or false. Why? Because the primitive has no reason to know the principles of modern sciences. It is thus perfectly legitimate in the context in which primitive people lived to believe that strange practices could explain, for example, agricultural practice. ‘In the case of the societies envisaged by Durkheim, it is the religious doctrines which provide explanations of the world which make it possible to coordinate the data of sensitive experience. These doctrines therefore play in traditional societies the role of science in our own societies … Religions do represent a first form of explanation of the world’ (Boudon 2000: 107). The idea is interesting, but it is not as new as Boudon suggests, because that is what Comte said, in another vocabulary, at the beginning of the nineteenth century.
How is it that the primitive does not abandon theories which are contradicted by reality? Boudon freely interprets Durkheim's thought by demonstrating that even the modern scientist does not abandon the theory when it turns out to be incomplete; on the contrary, the scientist will try to correct it in the hope that a minor modification will make it compatible with the facts (see Boudon 2013: 185). ‘It is often more rational to keep a theory contradicted by the facts and to try to amend it than to reject it’ (Boudon 2000: 109). How to explain this phenomenon?
It should be added here that reality can confirm false beliefs. It is an argument of remarkable subtlety that Durkheim develops here: the rituals to make the rain fall (or to facilitate the reproduction of the herds) are performed at the time when it is most likely to fall (or at the time when animals mate). Thus, belief in a false causal link can be confirmed by the existence of correlations which, although fallacious, are real. (Boudon 2000: 110)
It is clear to Boudon that moderns, or scientists, do not act differently from primitives. They are guided by their own reason. That explains for Boudon why Durkheim's theory is superior to the theory of his contemporary Lucien Lévy-Bruhl who sees irrationality in the beliefs of primitive societies.
There is no doubt that Boudon offers an original, even provocative, reading of Durkheim's Les Formes élémentaires de la vie religieuse. But Boudon often retains only pieces that are congruent with his own vision of sociology and rationality. We are surprised, for example, that there is no analysis on totemism, which is nevertheless an essential element of the book. This is a whole symbolic life thus concealed. For Durkheim, god and society are one, but, above all, god is something fundamentally superior to man; god is coercive, constraining, god imposes on humans ways of acting and thinking. God is therefore a social creation in the sense understood by Durkheim in Les Règles de la méthode sociologique; by the same token we can read in Les Formes de la vie religieuse that religion is ‘only the figurative expression of society’ (Durkheim 1990: 323).
Boudon never considered himself a specialist of a particular author, but he sought through the writings of the classical sociologists materials to better equip his own sociology. This made the great specialist of Durkheimian sociology, Philippe Besnard, suggest that Boudon proposes a very ‘generous, creative, sometimes distant, even unfaithful in detail’ vision of Durkheim (Besnard 2003: 187). Besnard's commentary leads Boudon to reflect and, ultimately, to justify his own approach. ‘I have often wondered, he writes, what principles govern the interpretation of texts. At first glance, the case of scientific texts seems simpler than literary texts. But I am not so sure and I sometimes feel taken aback by the assurance the commentators give that their interpretations present the thought of such an author’ (Boudon 2011: 225). But what really matters for Boudon, as we can see in most of his works, has been to insist on ideas that could promote an authentically scientific sociology (see Marcel and Valade 2019: 239–251). One of the last sentences of Boudon's work reminds us of this: ‘he is the founder of sociology as a science in France’ (Boudon 2013: 190). Durkheim and Boudon therefore reach out through the century, they have ‘always recognized themselves in a scientific conception of sociology’ (Boudon 2001: 3–50)
Boudon, R. 1968. À quoi sert la notion de structure? Essai sur la signification de la notion de structure de la notion de structure dans les sciences humaines. Paris: Gallimard.
Boudon, R. 2001. ‘Pourquoi devenir sociologue? Réflexions et évocations’. Revue Européenne des Sciences Sociales 34 (120): 5–30.
Dubois, M. and S. Mesure. 2018. ‘La circulation transatlantique d'un ethos scientifique pour la sociologie: La correspondance de Raymond Boudon’, Revue européenne des sciences sociales 56 (2): 41–63.
Grémy, J.-P. 2018. ‘1967. Une thèse en Sorbonne: L'analyse mathématique des faits sociaux’. Revue européenne des sciences sociales 56 (2): 13–40.
Marcel, J.-C. and B. Valade. 2019. ‘Au comité de rédaction de l'Année sociologique: un héritage inopiné à une entreprise de normalisation’. l'Année sociologique 69 (1): 239–251.
Pawson, R. 2009. ‘On the Shoulders of Merton: Boudon as the Modern Guardian of Middle Range Theory’. In M. Cherkaoui and P. Hamilton, Raymond Boudon: A Life in Sociology, vol. 4, Oxford: The Bardwell Press, 317–333.