levy-Bruhl, durkheim, and the positivist roots of the sociology of knowledge* Journal of the History of the Behavioral Sciences 32 (1996): 424-40 Warren Schmaus ILLINOIS Institute of technology Sociologists of knowledge continue to debate the degree to which people from different societies have different modes of thought. In the course of these debates, social scientists often cite Edward E. Evans-Pritchard's (1937) study of Azande witchcraft as evidence that people from non-Western societies are indifferent to what Europeans would regard as logical contradictions (e.g. Bloor 1991:138-46; Latour 1987:186-94). It is commonly understood that Evans-Pritchard's interpretation derives from Lucien Lévy-Bruhl's so-called "primitive mentality" thesis, according to which "prelogical" thought, unlike modern or scientific thought, is indifferent to contradictions and impervious to experience. Although for reasons of style I have not put every occurrence of such biased terms as "primitive" and "prelogical" in quotation marks, this is how they should be read., Initially, Lévy-Bruhl's concept of a pre-logical mentality was criticized by cultural relativists on the grounds that indigenous peoples should be evaluated on their own terms without reference to our logical standards. C. Scott Littleton, however, argues that Lévy-Bruhl was at least as relativist as his critics and regards him as the founder of the contemporary cognitive relativist view that different societies have incompatible rules of logic (1985:xvi, xx). Littleton defines cognitive relativism as "the notion that the logic we bring to bear in our descriptions of the world is not universal, but rather a function of our immediate techno-environmental circumstances and our particular linguistic and ideological heritage, and that no one logic is necessarily superior to any other logic" (1985:vi). However, the degree to which Lévy-Bruhl's primitive mentality thesis was inspired by the positivist philosophy of Auguste Comte is less well appreciated. According to Comte's three-stage law, the fundamental principle of his sociology, our earliest explanations of nature and society were in terms of theological and metaphysical causes, which are gradually being replaced by positive laws governing the phenomena. Lévy-Bruhl assumed something like Comte's radical opposition between religion and science in distinguishing between primitive and modern thought. Émile Durkheim, on the other hand, broke with Comte's positivist conception of science in arguing for the opposing claim that modern, scientific thought grew out of primitive religious thought. Hence, although present-day social scientists may be closer to Lévy-Bruhl than to Durkheim in their conception of the relationship between science and religion, as Robin Horton (1973:283) suggests, from an historical point of view Durkheim's hypothesis was bolder and more original than Lévy-Bruhl's. Lévy-Bruhl's Theory of Primitive Mentality In 1910, after a long career as an historian of philosophy, the fifty-three year old Lévy-Bruhl published his first major work in the sociology of knowledge, Les fonctions mentales dans les sociétés inférieures. His goal was to discover the laws that govern collective representations, which for Lévy-Bruhl, Durkheim, and many others are the representations common to the members of a social group that inspire conformity among them. Like Durkheim, Lévy-Bruhl believed that to discover these laws, one needed to study what he regarded as the earliest, simplest societies. However, Lévy-Bruhl thought that we could not explain the institutions, customs, and beliefs of "undeveloped" people on the assumption that they had minds like our own (1910:425 [1985:361]). Primitive thought, Lévy-Bruhl argued, is not an inferior or more childish, ignorant use of principles common to people in other societies, such as those of modern Europe (1910:454 [1985:385]). He regarded as inadequate in particular accounts of primitive mentality expressed in terms of the misuse of the principles of causality and the association of ideas or in terms of the post hoc, ergo propter hoc fallacy (1910:76 [1985:76]). For example, he rejected Tylor's and Frazer's hypothesis that the earliest religions consisted of animist beliefs that arose through the primitives' misapplication of the concept of causality and the principles of the association of ideas. Religion is a social phenomenon, he argued, that could not be explained in terms of individual thought processes (1910:6-14 [1985:17-23]). Although he conceded that primitives may be just as prone to the post hoc fallacy as every one else, he said that explanations of primitive thought in terms of this fallacy "if not completely inaccurate, are surely incomplete" (1910:73 [1985:73]). I provide my own translations from the French, but also give page references in square brackets to English translations where they are available. Lévy-Bruhl believed there to be something more and something different than just a misuse of the principle of causality involved in primitive mentality, since the connections the primitives draw among the phenomena are not merely relationships of succession in time. These relationships of succession suggest to the primitive some sort of mystic relationship in which the antecedent has the power to produce the consequent (1910:73-74 [1985:73-74]). Like Durkheim, Lévy-Bruhl saw the notion of power or force as being of collective origin, or a collective representation. Left to its own devices, the individual mind would perceive only the succession among phenomena. In Les Fonctions mentales, Lévy-Bruhl defended two "working hypotheses" (1910:426 [1985:362]) regarding the collective representations of so-called primitives: 1. The institutions, practices, and beliefs of "primitives" imply a prelogical and mystical mentality, oriented otherwise than our own. 2. The collective representations and the connections among these representations that constitute this mentality are governed by the law of participation, and, as such, indifferent to the logical law of contradiction (1910:425 [1985:361]). Primitive mentality is mystical with regard to the contents of its representations and prelogical with regard to the connections among them. By calling primitive thought "prelogical," Lévy-Bruhl meant only that it takes no pains to avoid contradiction and not that it belongs to some earlier stage of thought prior to logical thought. Have there ever existed groups of human or pre-human beings, of which the collective representations have not yet obeyed the laws of logic? We do not know of any: in any case, it is very little likely. At least, the mentality of societies of inferior type, which I call prelogical, lacking a better name, does not at all present this characteristic. It is not antilogical; it is no more alogical. In calling it prelogical, I want to say only that it does not oblige itself above all, like our thought, to abstain from contradiction. It obeys first of all the law of participation (1910:79 [1985:78]; Lévy-Bruhl's emphasis). Lévy-Bruhl did not say that primitive thought was governed by an alternative logic: not only did the evidence not warrant this conclusion, but the fact that we can understand and communicate with indigenous peoples was evidence against it. He saw no reason to think that the connections among collective representations were governed solely by laws of a logical character and held the very notion of a logic different from our own to be only a "negative and empty concept" (1910:68 [1985:69]). Years later, Lévy-Bruhl described how he had once entertained the hypothesis of an alternative logic, which was suggested to him by his reading about Chinese philosophy, and had turned from philosophy to social anthropology in order to investigate it. The more he learned, however, the more he came to modify his original hypothesis and the idea of an alternative logic gave way to that of a prelogical mentality (Lévy-Bruhl et al. 1923:20-22). Nevertheless, as he conceded in his posthumously published notebooks, the Carnets, in his hypothesis of prelogicality one can recognize in a highly attenuated form his original hypothesis of an alternative logic (1949:61 [1975:48]). For Lévy-Bruhl, the claim that primitive thought is not restrained by logic does not mean that it is wholly arbitrary. On the contrary, it is fixed, invariable, and bound by tradition. The reason is that this mentality, although not subject to a logical mechanism, or rather precisely because it is not subject to it, is not free. Its uniformity is the reflection of the uniformity of the social structure, to which it corresponds, and which it expresses (1910:115 [1985:108]; Lévy-Bruhl's emphasis). It is not clear that Lévy-Bruhl believed that the structure of a culture's thought could be explained by or reduced to its social structure. When Mauss reproached Lévy-Bruhl at a meeting of the Société française de Philosophie in 1929 for failing to explain the categories of primitive thought in terms of their social structure, Lévy-Bruhl replied with Hume's remark about his sounding line not being able to reach such depths (Lévy-Bruhl et al. 1929:124-27). In his second "working hypothesis," Lévy-Bruhl opposed logical thought, based on the principle of contradiction, to what he called the "law of participation." In the following preliminary statement of this law, there appear to be two different senses in which one can represent persons and objects as participating in one another: In the collective representations of primitive mentality, objects, beings, phenomena can be, in a fashion incomprehensible to us, at once themselves and something other than themselves. In a fashion no less incomprehensible, they emit and they receive mystical forces, virtues, qualities, actions, that make themselves felt outside themselves, without ceasing to be where they are (1910:77 [1985:76-77]). The first sense of participation, in which something can be two things at once, is exemplified by the alleged Bororo belief that they are red parakeets (1910:77 [1985:77]). In the second sense of participation, the primitives perceive everything as imbued with mystical properties (1910:30-40 [1985:38-45]). This second sense of mystical participation plays the same role in primitive mentality that the concept of causality plays in modern thought (1910:332 [1985:285]). At times, Lévy-Bruhl appeared to attribute also a third, Platonic sense of participation to indigenous peoples: "it is for us that, since Plato, participation is a metaphysical problem. For the Eskimo, one may say that it stands on its own (va de soi)" (1927:414 [1928:324-25]). In the same work, he attributed to native Americans the Platonic notion that individual animals of the same species are but the "multiple and transient expressions of the same unique and imperishable essence" that they feel to be "even more real than the individuals of which it is composed" (1927:62, 64 [1928:61, 63]). In the Carnets, however, he distinguished the primitive from "our" Platonic notion of participation. Our notion of participation, but not the primitives', assumes that separate objects are first independently represented and then thought to participate in each other. In Lévy-Bruhl's example, we first represent the corpse and imagine the ghost and then conceive them as participating in one another. But the primitive has separate representations of the ghost and the corpse only after he reflects upon the participation, which is experienced at first only emotionally. It is the initial emotional reaction to death that Lévy-Bruhl identified with participation in this case (1949:1-4 [1975:1-3]). Lévy-Bruhl qualified that the law of participation governs only collective representations. "Considered as an individual, to the extent he thinks and acts independently, if it is possible, of these collective representations," he said, "a primitive will sense, judge, and conduct himself most often in the fashion that we would expect." With regard to such mundane practical activities as hunting, avoiding dangerous animals, and getting in out of the rain, "The inferences he will form will be just those that would appear reasonable to us in the given circumstances" (1910:79 [1985:78-79]). In the minds of primitives, then, the law of contradiction must co-exist with the law of participation. However, he qualified, these two laws do not govern distinct domains of thought. Unlike oil and water, they "permeate" one another. According to Lévy-Bruhl, it is "very difficult, not to say impossible, to trace a clear line of demarcation between individual and collective representations." Among primitives, individual sense perceptions are combined with mystical elements of collective origin and are thus influenced by the law of participation. On the other hand, the law of contradiction influences operations that would be impossible without it, such as numeration and reasoning, and also those governed by the law of participation (1910:112-13 [1985:106]). The laws of participation and contradiction continue to co-exist in "advanced" societies, Lévy-Bruhl believed. Prelogical thought never wholly disappears. He cited belief in God and Christian beliefs regarding Lazarus as examples of residuals of thought governed by the law of participation (1910:455, 452-55 [1985:377, 384-86]). The difference between modern and primitive thought is that modern concepts tend to be more carefully defined, reflecting more analysis. Due to this lack of analysis, in primitive representations, unlike ours, affective and even motor elements are blended with the intellectual ones (1910:28 [1985:36]). Lévy-Bruhl cited Théodule Ribot as the source of this idea (1910:3 [1985:14-15]). According to Mousalimas (1990:34f), Lévy-Bruhl may have begun to consider the affective aspects of human thought even earlier, while working on his book on Friedrich Heinrich Jacobi (1894). Lévy-Bruhl does not cite Jacobi in his work on primitive mentality, however. This lack of analysis also explains the primitives' apparent unconcern for logical contradiction and counter-evidence (1910:113-14 [1985:107-8]). Prelogical, mystical thought will evolve only when the connections among collective representations are gradually dissolved by experience and logical exigency. When logic wins out over the law of participation, thought becomes free and progress becomes possible (1910:116 [1985:109]). Primitive yields to modern, logical thought with the development of an educated caste of people who are able to devote time to the analysis of concepts (1910:440-41 [1985:374]). In his more radical moments, Lévy-Bruhl maintained that the translation of primitive beliefs into European languages was "equivalent to treason," since when Europeans use their own concepts to express these beliefs, precisely because these are concepts, they distort what they find (1922:506 [1978:434]). For this reason he found especially unreliable the reports of missionaries who attributed such Western concepts as the distinction between body and soul to indigenous peoples (1927: 128 [1928:111]). Interpreting indigenous cultures presents us with a dilemma: either we can present a faithful account of their beliefs that reveals what we take to be confusions and contradictions but do not appear so to the primitive, or we can impose on their beliefs a clarity and internal coherence for which the primitive has no need. To avoid this dilemma, Lévy-Bruhl said, we must place ourselves at the point of view of primitive mentality (1927:127-28 [1928:110]). To put ourselves at the point of view of primitive mentality we must come to grips with the primitives' ways of representing the fundamental categories of thought. It is precisely the differences between our categories and theirs that create the appearance that they do not follow the same rules of logic that modern Europeans do. In his account of F. H. Cushing's experiences among the Zuñi, Lévy-Bruhl said: Our intellectual habits are very far from those of the Zuñis. Our language (without which we represent nothing and do not reason) implies categories that do not coincide with theirs. . . . Thus the mentality of inferior societies is doubtless not as impenetrable as if it obeyed a logic other than ours, but it is nonetheless not completely intelligible to us. We are led to think that it does not obey exclusively the laws of our logic, nor perhaps laws that may be wholly of a logical nature (1910:70 [1985:70-71]; Lévy-Bruhl's emphasis). Starting with his first book on primitive mentality, Lévy-Bruhl showed how primitive thought does not have our categories of causality (1910:332 [1985:285]; 1922:85, 126-27 [1978:90, 124]); space and time (1922:89-91, 126 [1978:93-95; 123-24]); natural forces (1910:105-6 [1985:100]); the world or cosmos (1935:1 [1983:32]); number, viz., unity, duality, and plurality (1927:188 [1928:155]); individuality or personality (1927:70f [1928:67f]); the soul or spirit (1910:92-93 [1985:90]; 1927 [1928]); or life and death (1910:358 [1985:306]; 1927:291ff [1928:232ff]). These differences in the categories account for indigenous people's inabilities to grasp modern Western European religious concepts such as that of a monotheistic god, the soul, and the afterlife (1927:374, 391-94 [1928:295, 308-10]; 1931:xxix [1973:29]). Western religion was just as foreign as western medicine and science to their way of thought. He also took issue with Durkheim and Marcel Mauss's 1903 paper on primitive classification, questioning whether primitives really have a concept of a hierarchical arrangement of genus and species. Indeed, he argued, it is precisely because such indigenous peoples as the Australians, West Africans, North American Indians, and Chinese do not have a hierarchical arrangement of concepts that all sorts of animals, plants, stars, directions, colors, and inanimate objects generally are included in the same groupings as members of the society (1910:138-39 [1985:127-29]). Primitive categories are governed by the law of participation. As Lévy-Bruhl argued in L'Ame primitive, once we recognize that the native's thought obeys this law, we are no longer limited merely to reporting what he or she feels, but can make sense of his or her thought and behavior: If we never lose sight of the fact that his thought is mystical, and that it obeys without resistance the law of participation, doubtless everything will not become clear, but one will be able to account for many of the apparent confusions and contradictions (1927:208 [1928:170]). Take, for example, Lévy-Bruhl's account of the primitive category of space in La mentalité primitive. He gave this account in the course of explaining primitive practices for determining the direction in which a lost person or hunted quarry is to be found. One such practice from southern Africa is to light a fire and see which way the smoke blows. For Lévy-Bruhl, this practice made sense given the primitives' way of conceiving space. People, animals, plants, and other things and the parts of space in which they are found "participate" in one another. The primitive does not have the homogeneous representation of space made familiar by Euclidean geometry. Rather, each region of space is inseparable from the things that fill it. Hence, not all directions in space are the same. Since nothing happens without a reason for the primitive, the smoke must be drawn in a certain direction by something affecting a region of space in that direction (1922:231-32 [1978:208-9]). Far from regarding primitive thought as illogical, then, Lévy-Bruhl came to see that once it is interpreted in the light of its own categories, it is consistent with itself. As he explained to Evans-Pritchard in 1934, primitive thought if anything is overly-coherent (1952:120). Evans-Pritchard also criticized Lévy-Bruhl for exaggerating the mystical character of primitive thought as well (1934:8), to which he replied that he had done this deliberately for rhetorical effect (1952:118-19). By the time he wrote his fifth book on primitive mentality, La mythologie primitive (1935), he even came to associate the notion of "participation" with similarity relationships governed by our principles of logic. Specifically, these similarity relationships are governed by the principle of transitivity, much in the same way as is the concept of equality: Just as, for our abstract thought, two quantities equal to a third are equal to each other, this self-evident proposition having no need of proof, so, to the sensibility (sentiment) of these primitives, the beings that participate in the nature of the same Dema are all similar to each other, since they are all similar to it (1935:110-11 [1983:109]). The Dema are the mythological ancestors, endowed with supernatural powers, of the Marind tribe of New Guinea. Ultimately, due to the misunderstandings it engendered, Lévy-Bruhl abandoned the hypothesis of prelogicality and stopped opposing the law of participation to the law of contradiction. Nevertheless, he continued to believe that participation characterizes primitive thought and accounts for its lack of intelligibility. In his Carnets, he admitted that in Les Fonctions mentales he was not careful to distinguish logical contradictions from physical impossibilities. Primitives will accept a logical contradiction no more than anyone else. Physical impossibilities are another matter. He came to realize that many of his examples of primitives supposedly failing to recognize a contradiction, such as allowing for a person to be in two places at the same time, in fact illustrated their tolerance for what we would reject as physically impossible. Lévy-Bruhl began to disavow the prelogical mentality hypothesis at least as early as 1929 (Lévy-Bruhl et al. 1929:109). He explicitly abandoned it in his Carnets (1949:60 [1975:47]), where he also explained the reasons he no longer believed that participation is a "law" of thought (1949:77-78, 129-31 [1975:60-61, 99-100]). The Carnets, precisely because they are notebooks that Lévy-Bruhl wrote for himself in preparation for his next book, which he did not live to write, are very repetitious. That is, he kept returning to the same problems, arguments, and distinctions again and again in order to clarify his own thoughts before writing his book. Hence, his thoughts on such topics as logical and physical possibility must be gleaned from scattered passages throughout the Carnets (1949:9, 79, 166 [1975:7, 62, 126]). As Lévy-Bruhl explained in Kantian terms, what we take to be physically impossible are those events that are incompatible with the a priori conditions of possible experience or the categories. Although the primitive may have the same basic categories as everyone else, these constrain only his everyday experiences and neither his myths, his mystical experiences, nor his sense that people and things participate in one another (1949:65-67, 157-59 [1975:51-53, 120-21]). Because the primitive takes everyday life to be replete with mystical events, he allows for exceptions to the implications of his general concepts. For example, since he believes that he is just as likely to meet with a magical jaguar as a real one, his concept of jaguar is of little use in guiding his behavior towards these animals (1949:224-25 [1975:172-73]). Another reason the primitive may tolerate what appears physically impossible to us is that the primitive does not subsume one concept under another in a hierarchy of increasingly generality. For instance, Lévy-Bruhl cited the Bororo belief that the members of the Trumai tribe sleep at the bottom of the river, and explained that, unlike Europeans, the Bororo do not subsume the concept of human being under the general concept of air-breathing mammals (1949:11-13, 178-80 [1975:9-10, 135-36]). Durkheim's Criticisms Durkheim was perhaps the first person to raise the objection that Lévy-Bruhl had not given sufficient attention to the analysis of modern thought to establish that it is in fact different than primitive thought. In his review of Les Fonctions mentales, Durkheim claimed to share Lévy-Bruhl's "fundamental principles" that different kinds of mentality have succeeded one another in the course of history and that primitive mentality is essentially religious. However, he disagreed that primitive, religious mentality is the opposite of modern, scientific mentality. For Durkheim, the categories presupposed in modern scientific thought evolved from primitive religious categories, not in opposition to them. These categories make logical thought possible in the first place (1913a(ii)(6) and (7):35-36). In The Elementary Forms of the Religious Life (1912a), Durkheim defended the analogy between modern and primitive thought. He cited Lévy-Bruhl as having argued that primitives are indifferent to contradictions and argued that this is not the case and that primitives simply make different distinctions than we do (1912a:341 [1995:240]). Durkheim was arguing a position similar to that which Lévy-Bruhl himself maintained in his later works, according to which apparent contradictions in primitive thought can be explained once we are sensitive to the distinction between logical and physical impossibility and to their beliefs regarding what is physically possible. Of course, Durkheim did not live to see any except Lévy-Bruhl's first of six books on primitive mentality. Hence, his criticisms were necessarily directed at the original version of Lévy-Bruhl's primitive mentality hypothesis, according to which it was distinguished from our own by its prelogical and mystical character. However, Durkheim raised another point with which Lévy-Bruhl never would have agreed. According to Durkheim, although contemporary scientific explanations may rest on more methodical observations, they are of the same "nature" as primitive ones: both show how one thing "participates" in another. Émile Meyerson also raised the argument that contemporary science continues to rely on a notion of participation (Lévy-Bruhl et al. 1929:136-37). Hence, primitive mentality is not unrelated to our own; "our logic is born of this logic." He argued that when primitives say that a man is a kangaroo or that the sun is a bird, that is, when they identify very different things with one another, their way of thinking is no different than ours when we say that heat is motion or that light is a vibration in an ether. Durkheim cited Lévy-Bruhl as having argued that because the participations postulated by primitive mythologies violate the principle of contradiction, they are opposed to those implied by scientific explanations. On the pages he cited, however, Lévy-Bruhl did not even characterize scientific explanations as involving participations, a suggestion he surely would have resisted. Durkheim (1912a:340-41 [1995:240]), which contains two citations to Les Fonctions mentales: 1910:77ff and 79. According to Durkheim, to explain the facts in any science "is to establish among them relations that make them appear to us as functions of one another, as vibrating sympathetically according to an internal law, founded in their nature." These relations are not revealed by sense experience but must be hypothesized by the mind. "The great service that religions have rendered to thought," Durkheim said, "is to have constructed a first representation of what these relationships of kinship among things could be." It was important for the primitives to try even if they did not succeed, for it was essential that the mind not remain enslaved to appearances. Once people had the notion that there were internal connections among things, science and philosophy became possible (1912a:339-40 [1995:239]). Although Lévy-Bruhl never published a reply to Durkheim, in a letter to Georges Davy written in 1930, he characterized his own concept of science as more empiricist than that of Durkheim, whose philosophy of science implies a metaphysics that he did not accept. Letter dated March 4, 1930 (Davy 1957:471). Presumably, he was unhappy with Durkheim's grounding of physical laws in the natures or essences of things. Nevertheless, at least in his later works, he agreed with Durkheim that primitive thought was a necessary first step in human intellectual development. In L'Expérience mystique et les symboles chez les primitifs (1938), he compared the way that primitives experience nature with the way in which animals do. Unlike animals, the primitive lives in a dual reality: one of sense experience that he shares with the animals, and the other of mystical experience. Just as Durkheim had argued that religion emancipated us from our enslavement to appearances, Lévy-Bruhl assigned a similar role to mystical thought: If experience had never revealed to man another reality than that of the sensible world in which he is plunged, undoubtedly his mental activity would have remained fundamentally like that of the higher animals: more varied, more ample, more rich, perhaps, but like it glued to its object and unable to turn away from it. It would remain literal, so to speak. Nothing would incite it to raise itself above the immediately sensed and perceived reality, to dominate it, to imagine another that would not be given in the same manner. That things can be other than they are, this idea could be born only in mystical experience, itself made possible by articulate language, social institutions, and the mental progress of homo faber (1938:95; Lévy-Bruhl's emphasis). Lévy-Bruhl went on to say that the absurdities, contradictions, and confusions in primitive thought mattered little, as long as the "decisive step" of opposing an imagined world to the world of sense experience was taken (ibid.). Again, Durkheim could not have known of these later developments in Lévy-Bruhl's thought. Durkheim regarded it as a strength of his own sociology of knowledge that he did not postulate the existence of an "abyss" between primitive religious and modern scientific thought (1912a:342 [1995:240]). In speaking of an "abyss," Durkheim may have associated Lévy-Bruhl's with Comte's distinction between science and religion. According to Comte, theology and physics are so different from each other, "so profoundly incompatible, their conceptions have a character so radically opposed," that a metaphysical stage was a necessary transition (1975 I:24). For Durkheim to associate Lévy-Bruhl with Comte, I will argue, was not wholly unwarranted. Lévy-Bruhl and Comtean Positivism Lévy-Bruhl's reputation was closely bound up with his work on Comte. He taught a course on Comte first in 1895 at the École Normale Supérieure (Käsler 1991:124) and then at the Sorbonne (Merllié 1989b:498). In 1898 he published an article on Comte and John Stuart Mill in the Revue philosophique and then edited and published their correspondence the following year (Nandan 1977:334). When he published his book La Philosophie d'Auguste Comte in 1900, he sent Durkheim a copy (Durkheim 1969b). At a meeting of the Société française de philosophie in 1902, Lévy-Bruhl along with Émile Boutroux defended Comtism as a living philosophy, one that they would like to develop in the Kantian direction of an investigation into the conditions of knowledge (Boutroux et al. 1903). In 1903, he published La morale et la science des moeurs, an attack on traditional philosophical ethics that cites Comte in the first paragraph and argues that the time has come for a positive science of morals (1905:1). The longest chapter in Lévy-Bruhl's History of Modern Philosophy in France (1899) is devoted to Comte. In fact, not only is this chapter highly sympathetic to Comte, but the whole history has a Comtean flavor. His choice of philosophers and philosophical texts to write about and his estimation of their worth is strongly influenced by Comte. However, this book did not contribute to Lévy-Bruhl's reputation in France, as it was published in English in the United States. Lévy-Bruhl maintained an interest in Comte up until nearly the end of his life, having planned a conference on Comte at the University of Prague for 1935 (Lévy-Bruhl 1989). Lévy-Bruhl embraced a positivist view of science. His book on Comte does not criticize his philosophy of mathematics and the natural sciences, but defends it against various objections raised by Émile Littré, Herbert Spencer, Charles Renouvier, and Max Müller. Merllié quotes a 1927 interview in which Lévy-Bruhl explained what he has taken from Comte: Although I may not be a positivist, I have no less undergone strongly enough the influence of the spirit of Comte. He has made me lose the taste for all philosophy that is not closely linked to the history of the sciences and to the present state of research and scientific speculation (Merllié 1989a:444). In qualifying his remark about Comte's influence by saying that he may not be a positivist, Lévy-Bruhl may have been referring to the fact that he was never a member of the Société Positiviste that was headquartered in Comte's former home at 10 Rue Monsieur-le-Prince in Paris, from which it re-issued his publications. In any event, Lévy-Bruhl was often perceived as a positivist (Merllié 1989a:444). The Comtean influence Lévy-Bruhl described is evident in his treatment of Descartes's philosophy as concerned with foundational issues in the sciences (1899, ch. I). Although this may be a standard interpretation today, in Lévy-Bruhl's time it was more common among philosophers to divorce Descartes's metaphysics from his mathematics and science (Cavaillé 1989:458). In general, Lévy-Bruhl appears to have adopted from Comte a kind of historical relativism with regard to science and philosophy. In the letter to Davy cited above, Lévy-Bruhl characterized his own philosophy of science as not only more empiricist but "plus relativiste" than that of Durkheim. See note 12. "Relativist" is a term used by both Lévy-Bruhl and Comte to characterize positivism. Comte employed the term in at least three different senses: first and foremost, he characterized positive science as aiming at knowledge relative to human needs, while metaphysics and theology aimed at "absolute" knowledge (1929 I:452; cf. Lévy-Bruhl 1900:35 [1903:31]). Knowledge is also relative to our sensory physiology; if we were blind, he argued, there would be no astronomy (Lévy-Bruhl 1900:83-84 [1903:74-75]). Finally, Comte held that our knowledge is historically relative. Even theological theories "constituted . . . the best system compatible with the corresponding age of human development" (1975 II:730; see also 1929 IV:220). In true Comtean fashion, Lévy-Bruhl also militated for replacing the study of ancient languages with that of the sciences as a condition for students to pursue graduate work in philosophy (Merllié 1989a:441n66). Finally, he defended Comte's view that sociology embraces the philosophy of science. For Comte, the sciences are themselves social facts and the analysis of their methods belongs to sociology. Positivism transforms the philosophy of the theological and metaphysical stages into the science of sociology (1900:401-3 [1903:349-50]). However, there is a sense in which Lévy-Bruhl may not have been fully a positivist, and that has to do with his attitude towards the postulation of unobservable entities in the sciences. Specifically, it is not clear how a positivist could embrace the concept of collective representations, which are a type of real but unobservable mental entity for Durkheim. I argue this point at length in chapter 3 of my book on Durkheim (Schmaus 1994). One could argue that it is not inconsistent with positivism to recognize the existence of mental states. Indeed, Lévy-Bruhl, citing William James, distinguished between acknowledging that one has a subjective awareness of one's own mental states and having a metaphysical commitment to some sort of mental entity, even to the existence of oneself as the subject of these mental states (1927:2 [1928:15]). A positivist may accept certain kinds of mental states as simply the appearances or phenomena by which he obtains a knowledge of the world. However, to theorize about collective representations is to go beyond the appearances and to refer to mental states other than one's own. In keeping with an empiricist or positivist methodology, Lévy-Bruhl attempted to define collective representations in terms of their observable effects: The representations called collective, to define them only roughly and without going deeply, may be recognized by the following signs: they are common to the members of a given social group; they are transmitted from generation to generation; they are imposed on individuals and they arouse in them, according to the case, sentiments of respect, fear, adoration, etc. for objects (1910:1 [1985:13]). In L'Ame primitive, he characterized in similar terms the collective representations in primitive mentality of different kinds of animals. First, he explained that the primitive's mental representation of an individual animal is not clearly distinct from "a more general image, which, though not a concept, comprises all similar beings." The representation of it, he added: is characterized at the same time by the objective qualities that the primitive perceives in beings, and by the emotions they arouse in him. It is . . . a sort of essence or type, too general to be an image, and too emotional to be a concept. It seems to be nevertheless clearly defined, above all by the sentiments that the sight of an individual of the species excites, and the reactions it provokes (1927:59 [1928:59]). If we can assume that Lévy-Bruhl identified these sentiments with their expression or the "reactions" to which they give rise, we can interpret him as having suggested that the primitive's concept of a type of animal can be behaviorally defined. In his next book, Le surnaturel et la nature dans la mentalité primitive (1931), Lévy-Bruhl characterized the category of the supernatural in terms of emotional responses, and argued that these were the most important elements of the supernatural for the primitives themselves. In explaining what he meant by the "affective category of the supernatural," he described a process of abstraction by which the mind focuses not on properties or features common to a group of objects, but on an emotional coloration or tonality common to a group of mental images or representations. Those that elicit the same emotional response belong to the same "affective category." The affective element that defines the supernatural is the reaction of fear and anxiety (1931: xxxiv-xxxv [1973:32-33]). As he made clear in the Carnets, the affective category of the supernatural is thus not a category in the Kantian sense of an a priori condition of possible experience, but only in the classificatory sense of a set of things sharing some common characteristic (1949:35-36, 75, 117 [1975:27, 59, 90-91]). Generalizing from the examples of animal concepts and the affective category of the supernatural, we can suppose that Lévy-Bruhl would reconcile his positivism with his postulation of collective representations by proposing behavioral or operational definitions of the latter. Among the behaviors associated with collective representations would be the verbal behavior of primitives. In fact, although he cautioned that one could not simply infer the mentality of a people from its language due to the effects on it of migrations and the intermingling and absorption of groups (1910:151 [1985:139]), Lévy-Bruhl often appealed to linguistic evidence for collective representations. For instance, as evidence that people originally conceived causality in terms of spatial contiguity rather than temporal sequence, he cited the fact that in nearly all Indo-European languages the prepositions that signify causal relationships originally referred to spatial relationships (1910:333 [1985:286]). Elsewhere, he cited the manner in which certain languages add possessive suffixes to substantive nouns as evidence of the way the people who speak these languages conceive their relationships to each other and to things (1927:76f [1928:72f]). For Lévy-Bruhl, the reality assigned to collective representations was no more mysterious than that assigned to linguistic facts: "a language, although it exists, properly speaking, only in the minds of the individuals who speak it, is no less an indubitable social reality, founded upon an ensemble of collective representations . . . " (1910:1 [1985:13]). In the Carnets, Lévy-Bruhl resolved to speculate no longer on mental representations and their connections, but not so much because they are unobservable entities as because they belong to a tradition in philosophy and psychology that he feels has led him astray and to ask the wrong questions: The idea of representations, sorts of entities that are separated or at least always separable for which it is necessary to find a satisfactory means for the mind to reconnect (say a word about Gestalt psychologie) forms a part of a whole of superannuated psychological and logical conceptions, evidently proceeding from the associationist school and its English and French predecessors of the nineteenth and eighteenth centuries. Their way of posing problems does not resist a more exact knowledge of the facts, and more often they troubled themselves only with pseudo-problems (1949:78 [1975:61]). Lévy-Bruhl said no more about Gestalt psychology or about how it could dispense with talk about mental representations. Nor did he make it clear whether he thought that the social sciences could dispense with mental representations altogether, or only with representations as they are conceived within associationist psychology and the "way of ideas" tradition in philosophy. Comte and Lévy-Bruhl on Primitive Mentality Comte's account of primitive mentality is contained in his defense of the three-stage law. Lévy-Bruhl's critique of this defense stands out in his otherwise sympathetic book on Comte's philosophy of science. Lévy-Bruhl found fault with two of Comte's assumptions. First, like other philosophers before what Lévy-Bruhl called "the recent progress of anthropology," Comte simply "constructed" his concepts of primitive humans and society. Second, instead of considering the history of the whole of humanity, Comte took into account only that of western civilization, which according to Lévy-Bruhl he had "no right" to do (1900:320-21 [1903:276-77]). He repeated these criticisms of Comte's social science methodology in his first book on primitive thought (1910:4-5 [1985:15-16]). Nevertheless, he at least initially accepted certain elements of Comte's characterization of primitive thought, especially his radical opposition between modern and primitive modes of thought. Lévy-Bruhl's project, it seems, was to advance Comte's sociology by correcting its ethnocentrism and revising its account of primitive thought in the light of more recent work in anthropology. At a meeting of the Société française de Philosophie in 1923, Paul Fauconnet pointed out the analogy between Comte's theological stage and Lévy-Bruhl's primitive mentality (Lévy-Bruhl et al. 1923:45). It appears that in Lévy-Bruhl's opinion, Comte had already analyzed modern, scientific thought well enough and that the principal part of the sociology of knowledge that remained to be done was the analysis of non-scientific modes of thought. Comte's defense of the three-stage law begins with the philosophical argument that hypotheses are necessary to organize our perceptions (1975 I:23). Without an hypothesis, he said, an observer would not even know what to look at in the phenomena taking place before his eyes (1975 II:139). Primitives, however, have few observations on which to base positive hypotheses. Pressed between the need to have some hypotheses and the lack of a sufficient basis for positive hypotheses, they form "fetishistic" theological hypotheses. These hypotheses arise from their making analogies between external objects and themselves, ascribing human sentiments to what we would regard as inanimate objects (1975 I:23, 666; 1825:138-39; 1929 II:84; 1856:8). Comte described primitive intellectual life as involving a "logic of the heart, that is, the combination of ideas according to the connection of sentiments." The combination of ideas according to sentiments leads the primitive to attribute these sentiments to the objects that cause the impressions that give rise to these ideas (1929 II:88). It appears that Comte's account of our need to begin with theological hypotheses may have been the source of Durkheim's and Lévy-Bruhl's theory that human intellectual life began with primitive mystical or religious thought. Also, by having claimed that primitive thought is more emotional where ours is more conceptual, Lévy-Bruhl seems to have accepted something like Comte's notion of a "logic of the heart." However, there are some important differences. First of all, where Comte held that theological hypotheses are necessary even for perceptual experience, Durkheim and Lévy-Bruhl held that religion was necessary to rise above such experience. Perhaps more importantly, Durkheim and Lévy-Bruhl emphasized that religious thought is of collective and not individual origin. Neither would have accepted Comte's belief that primitive theology arose solely through the misuse of analogy. Similarly, although Lévy-Bruhl would have agreed with Comte that the primitive attributes emotions to the objects of perception, he would have insisted that the affective elements in primitive perception were of collective origin. However, Lévy-Bruhl and Comte agreed that the modern mode of thought developed in opposition to the primitive. Both held that with regard to practical matters such as avoiding danger, a primitive will respond in a way that appears reasonable to a modern person. Gustave Belot brought out this analogy between Comte and Lévy-Bruhl at the 1923 meeting of the Société française de Philosophie (Lévy-Bruhl et al. 1923:32). According to Lévy-Bruhl's interpretation of Comte, the positive method develops through the following dialectic: observation, made possible by theological hypotheses, reveals regularities that compromise these hypotheses and give rise to the positive philosophy (1900:48 [1903:43]). Like the common wisdom from which it arises, science abstains "from searching for the causes, the ends, the substances, and that which escapes verification by experience" (1900:68 [1903:61]). Science and common sense are concerned exclusively with observing the constant relations or laws that govern the phenomena. According to Comte, the germs of the positive method were present from the earliest stages of human development (1975 II:222-23). The positive method was used for solving everyday problems, while the theological method was reserved for "higher speculation" (1975 II:730). Everyday life depends on the ability to predict, which requires that phenomena be conceived as subject to invariable laws (1975 I:45, II:733; 1929 I:10-11). In La Mentalité primitive, Lévy-Bruhl contrasted primitive with modern modes of explaining natural phenomena in a way that assumes that the latter is characterized by a positivist lack of concern for more fundamental causal explanations: When we want to explain [a phenomenon], we search for the necessary and sufficient conditions in the series of phenomena itself. If we succeed in determining them, we demand nothing more. The knowledge of the law satisfies us (1922:512 [1978:438]). The primitive, on the other hand, Lévy-Bruhl asserted, is not satisfied with a knowledge of the invariable antecedent of the phenomenon in question, but seeks the "true cause" in a "metaphysical" realm of "unseen powers" (ibid.). In a later work, Le surnaturel et la nature dans la mentalité primitive, Lévy-Bruhl drew the contrast between primitive and modern thought in similar terms (1931:165-66 [1973:153-54]). Here he was even more explicit about his assumption that modern thought was identical with positivism, admitting that he was talking about the "difference between our positive manner of thinking and the mystic mentality of the primitive" (1931:31 [1973:57]). He described "our" positive mode of thought as follows: In our relations with the various physical forces that act upon us, and upon which we feel ourselves dependent, we have the inveterate habit to rest first of all upon the most extensive and most exact knowledge possible of the laws of nature. "To know in order to foresee in order to do [Savoir pour prévoir afin de pouvoir]:" we regulate ourselves by this maxim, without even thinking about it, so trained are we by our education to put our confidence in science, and to profit in all security by the advantages that its applications procure for us (1931:xvi [1973:20]). The words "foresee [prévoir]" and "foresight [prévoyance]" are technical terms in positivist philosophy. For Comte, what distinguishes positive from metaphysical hypotheses is that only the former allow us to make predictions that guide us in our practical lives: See Laudan (1971) for a concise introduction to Comte's positive philosophy. In sum, science, from which foresight; foresight, from which action: such is the very simple formula that expresses, in an exact manner, the general relation between science and art, in taking these two expressions in their total acceptation (1975 I:45 [Comte's emphasis]). According to Comte, where the positive method seeks a knowledge of the laws governing phenomena as a basis for making predictions in our practical lives, the theological philosophy conceives phenomena as controlled by arbitrary wills and thus as eminently variable (1975 I:454-55; 1929 I:10). Thus, the theological philosophy was never able to embrace the practical side of life. This "scission" between everyday life and the theological philosophy eventually brought about the destruction of the latter (1929 I:9-10). Only the positive method could simultaneously extend our knowledge and link its parts. The theological and metaphysical philosophies, on the other hand, could not extend our knowledge and could connect it only in a vague, unstable sort of way (1975 II:730). Thus, the first state of our intelligence allowed for no "durable harmony" (1929 II:81) or "logical equilibrium" (1929 I:452) between practical and theoretical reason. The picture of primitive thought that Comte gave us is one of an unstable mix of positive and religious elements. Lévy-Bruhl's exegesis of Comte's views on this tension in primitive thought is uncritical. He used the term "contradiction" (1900:31 [1903:28]) to characterize the opposition between the positive method of explanation by law and the theological and metaphysical method of explanation by cause and essence, but did not explain where the contradiction lies. Although positivism may be inconsistent with metaphysics, this is not the same thing as saying that aiming for laws in science is inconsistent with the search for "deeper explanations." One makes this slide only if one accepts something like Comte's philosophy of science, which Lévy-Bruhl seems to have done. Durkheim, on the other hand, grounded causal laws in essences. For a detailed defense of this claim, see chapter 4 of my book (Schmaus 1994). Like Comte, Lévy-Bruhl also characterized primitive thought as an unstable mixture of the religious or mystical and the positive. He even described it in just these terms in L'Expérience mystique (1938:50, 56, 131). Just as Comte thought that the theological method had governed higher speculation, Lévy-Bruhl held that the law of participation governed collective representations. Comte would have accepted Lévy-Bruhl's characterization of primitive thought as less analytical than modern thought, as he identified analysis with the methods of positive science (e.g., 1929 I:581; 1856:59). However, this notion of a tension in primitive thought between the religious and the modern seems to have troubled Lévy-Bruhl. In the Carnets, he complained that in Les Fonctions mentales, he had merely juxtaposed these two modes of thought without explaining how the primitive mentality could "co-exist with the logical exercise of our mental activity." He then corrected what he had said in his earlier work by specifying that instead of a primitive mentality distinguished by its mystical and prelogical character, there is only a mystical mentality that is more easily observed in primitives but shared by everyone (1949:131 [1975:100-1]). Horton has expressed surprise that, for someone so interested in contrasting primitive with modern thought, Lévy-Bruhl had so little to say about the latter (1973:253). Both Mauss (Lévy-Bruhl et al. 1923:25) and Evans-Pritchard criticized Lévy-Bruhl for not having sufficiently studied modern thought in order to establish that it is in fact different than primitive thought. At a meeting of the Société française de philosophie, Lévy-Bruhl admitted that he did not really know how the average person of his day thought (Lévy-Bruhl et al. 1929:117). In response to Evans-Pritchard, Lévy-Bruhl simply asserted that modern European thought was "sufficiently known" (1952:120). Evans-Pritchard reports that in his personal conversations with him, Lévy-Bruhl revealed that with regard to the way in which contemporary Europeans think, "he felt himself to be in a quandary." Although he believed that contemporary Western religions were also mystical, he deliberately avoided talking about the mystical aspects of Western cultures in order to avoid giving offense. Evans-Pritchard argues that in so doing, Lévy-Bruhl "vitiates his argument" (1981:130). Mousalimas argues that Lévy-Bruhl was contrasting primitive thought not with western mentality in general but only with the positivist mindset of certain people like himself (1990:36). Lévy-Bruhl in effect conceded this point in a later work, La Mythologie primitive (1935). Here he took the fact that needed to be explained to be not so much that the primitive accepts as true his myths about people and animals changing into one another as the fact that Europeans no longer believe in their own stories of this sort. The answer that he provided was expressed in terms of the historical development of a positivist emphasis on verifiability: The reason for it is, without a doubt, at least in part, in the rational character of the civilization that classical antiquity has established and bequeathed to us. From the experience held as valid, the givens that could not be checked and verified (les données incontrôlables et invérifiables) -- that is to say those of mystical experience, by which the action of invisible and supernatural powers are revealed -- gradually found themselves excluded (1935:317 [1983:255]). The emphasis on verifiability led to the development of a concept of reality or nature as governed by physical law. With the exception of religious experience, everything that did not fit within this system of law was rejected as impossible. However, he pointed out, even in the West this new intellectual mindset was adopted not universally but "in certain communities only; in those, moreover, only after centuries of resistance" (ibid.). In the Carnets, he was careful to distinguish the primitive's indifference to physical impossibility from the thinking of lay people as well as scientists (1949:63 [1975:49-50]). It often appears that Lévy-Bruhl was contrasting primitive thought not so much with modern thought in general as with that of ethnologists and other scholars studying indigenous peoples. For instance, he remarked that the primitive feels no need to delimit or define his concept of witchcraft in any way: he simply knows it when he sees it, or rather, he reacts emotionally to the presence of witchcraft. On the other hand, "we," with our logical method of thought, attempt to analyze its essential characteristics (1931:225 [1973:195-96]). Philosophers at least since Wittgenstein, however, have argued that so-called "modern" people do not work with clearly defined concepts in their everyday lives either. An ethnologist attempting to formulate a theory of witchcraft, on the other hand, would feel it incumbent upon him- or herself to clarify and define his or her terms. Conclusion Lévy-Bruhl's notion of an opposition between primitive and modern modes of thought has its roots in Comte's philosophy. Both Lévy-Bruhl and Comte held that primitive thought is more emotional and modern thought is more analytic. For both, science aims at invariable empirical laws for the purpose of making predictions to guide our practical affairs. Hence, they thought, the modern mode of thought arose from practical life at the very beginning stages of human development. Both thinkers also associated this practical mode of thought with the individual and associated the religious with the collectivity. They both perceived a tension between the practical and the religious elements of primitive thought, with the religious gradually giving way to the practical. To be sure, Lévy-Bruhl and Comte constructed the tension between modern scientific and primitive religious thought by contrasting religious thought with their ideal of what science should be. To the extent Lévy-Bruhl went beyond Comte by grounding a theory of primitive religious thought in ethnographic data, he created an empirical social psychology of the primitive which he then contrasted with a normative, philosophical ideal. Durkheim, at least, cannot be accused of this. In defense of the positivist position, one could argue that Durkheim's essentialist conception of scientific law reflected the sciences of his day only to the extent that they were still in a transitional, metaphysical stage. This defense, however, would still concede that the positivist conception of law was but an imperfectly realized ideal. Although Comte and Lévy-Bruhl both considered themselves to be relativists, their relativism differs from contemporary cognitive relativism. The positivists' relativism allowed for progress, where contemporary relativists refuse to allow the past to be judged by present standards. This new relativism borrows elements from Durkheim and Franz Boas as well as from positivism. It takes the opposition between science and primitive religious thought from Comte and Lévy-Bruhl, but also accepts Boas's denial of a linear development in the direction of science. From Durkheim and especially Mauss it adopts the view that a primitive society's collective representations together with its practical and economic life form a coherent whole. Mauss explained how the social and economic cohere as parts of a "total social fact" in The Gift (1990). Putting all three elements together, we obtain the contemporary relativists' notion of coherent social systems, each with its own sets of standards or values, in opposition to one another. Mary Douglas's early work, such as her Purity and Danger (1966), provides the best contemporary example of a work in the social sciences that treats a society's beliefs and practices as forming a coherent whole. Although she has subsequently moved away from this position, her early work was an important formative influence on Bloor's relativist sociology of scientific knowledge (1982, 1991). In describing a society's beliefs and practices as forming a coherent whole, however, the Durkheimians tell perhaps too good a tale. Without the sort of tension that Comte described, how would intellectual change come about? Comte may have held the logical consistency of all thought, including the practical and the religious, as an ideal to be achieved in the final, positive stage, through establishing the Religion of Humanity. However, he had no qualms about attributing intellectual tensions to past and present thought. It is disappointing that Lévy-Bruhl at the end of his career felt that postulating the juxtaposition of competing modes of thought created problems that should be avoided. Why could he not simply accept that people may have conflicting beliefs and practices? Social anthropologists should avoid seeing too much coherence in a society's beliefs and practices and seek out the tensions in its thought. At the same time, they should avoid importing their own ideals into the interpretations of other societies, as Comte and Lévy-Bruhl did in supposing that primitives assume a positivist conception of law in their practical affairs. notes * An earlier version of this paper was presented in a session on Durkheim at the First History of Philosophy of Science (HOPOS) Conference in Roanoke, Viriginia on April 19, 1996. I would like to thank John Brooks, Terry Godlove, and Bob Jones for what I learned from their participation in that session, as well as Neil Gross for his helpful comments on a later draft.  bibliography Bloor, David. 1982. "Durkheim and Mauss Revisited: Classification and the Sociology of Knowledge." Studies in History and Philosophy of Science 13:267-297. Bloor, David. 1991. Knowledge and Social Imagery, 2nd. ed. Chicago: U. of Chicago Press. Boutroux, Émile, et al. 1903. "Comte et la métaphysique." Bulletin de la Société française de Philosophie 3:1-24. Cavaillé, Jean-Pierre. 1989. "Notes et documents sur 'Le Descartes de L. Lévy-Bruhl.'" Revue philosophique de la France et de l'étranger 114:453-463. Comte, Auguste. 1825. "Considérations philosophiques sur les sciences et les savants." Repr. in 1929 vol. IV, appendix pp. 137-75. Comte, Auguste. 1856. Synthèse subjective. Paris: Victor Dalmont, Libraire. Comte, Auguste. 1929 [1851-54]. Système de politique positive. Paris: Société Positiviste. Comte, Auguste. 1975 [1830-42]. Cours de philosophie positive. 2 volume edition ed. Michel Serres et al. Paris: Hermann. Davy, Georges. 1957. "Pour le centième anniversaire de la naissance de Lucien Lévy-Bruhl." Revue philosophique 82:467-93. Douglas, Mary. 1966. Purity and Danger. London: Routledge & Kegan Paul. Durkheim, Émile. 1912a. Les Formes élémentaires de la vie religieuse. Paris: Alcan. Durkheim, Émile. 1913a(ii)(6) and (7). Combined review of "Lévy-Bruhl, L., Les Fonctions mentales dans les sociétés inférieures," and "Durkheim, Émile, Les Formes élémentaires de la vie religieuse." L'Année sociologique 12:33-37. Repr. in Durkheim 1975a I:405-7. Trans. in Durkheim 1972:246-49, 1975b:169-73, 1978:145-49. Durkheim, Émile. 1969b. "Une Lettre inédite d'Émile Durkheim." L'Année sociologique, 3rd series (1967), p. 14. Repr. in Revue philosophique 94 (1969):381; 1975a, I:177. Durkheim, Émile. 1972. Emile Durkheim: Selected Writings. Ed. Anthony Giddens. Cambridge, UK: Cambridge U. Press. Durkheim, Émile. 1975a. Textes. Ed. Victor Karady. Paris: Les Éditions de Minuit. Durkheim, Émile. 1975b. Durkheim on Religion. Ed. W. S. F. Pickering. London, UK: Routledge & Kegan Paul. Durkheim, Émile. 1978. Emile Durkheim on Institutional Analysis. Ed. and trans. Mark Traugott. Chicago: U. of Chicago Press. Durkheim, Émile. 1995. The Elementary Forms of the Religious Life. Trans. of 1912a by Karen Fields. New York: Free Press. Durkheim, Émile, and Marcel Mauss. 1903. "De quelque formes primitives de classification: contribution à l'étude des représentations collectives." L'Année sociologique 6:1-72. Trans. as Primitive Classification by Rodney Needham, Chicago: U. of Chicago Press, 1963. Evans-Pritchard, Edward Evan. 1934. "Lévy-Bruhl's Theory of Primitive Mentality." Bulletin of the Faculty of Arts of the University of Egypt 2:1-36. Evans-Pritchard, Edward Evan. 1937. Witchcraft, Oracles, and Magic among the Azande. Oxford: Clarendon Press. Evans-Pritchard, Edward Evan. 1981. A History of Anthropological Thought. Ed. A. Singer. New York: Basic Books. Horton, Robin. 1973. "Lévy-Bruhl, Durkheim, and the Scientific Revolution." In R. Horton and R. Finnegan, eds. Modes of Thought. London: Faber & Faber. Pp. 249-305. Käsler, Dirk. 1991. Sociological Adventures: Earle Edward Eubank's Visits with European Sociologists. New Brunswick, NJ: Transaction Publishers. Latour, Bruno. 1987. Science in Action. Cambridge, MA: Harvard U. Press. Laudan, Larry. 1971. "Towards a Reassessment of Comte's 'Méthode Positive'." Philosophy of Science 38:35-53. Reprinted in Science and Hypothesis. Dordrecht, Holland: D. Reidel, 1981, ch. 9. Lévy-Bruhl, Lucien. 1894. La philosophie de Jacobi. Paris: Alcan. Lévy-Bruhl, Lucien. 1899. History of Modern Philosophy in France. Chicago: Open Court, 1924. Lévy-Bruhl, Lucien. 1900. La philosophie d'Auguste Comte. Paris: Alcan. Lévy-Bruhl, Lucien. 1903. The Philosophy of Auguste Comte. Authorized trans. of Lévy-Bruhl 1900. New York: G. P. Putnam's Sons. Lévy-Bruhl, Lucien. 1905. Ethics and Moral Science. Trans. of La morale et la science des moeurs (1903) by Elizabeth Lee. London: Archibald Constable & Co., Ltd. Lévy-Bruhl, Lucien. 1910. Les fonctions mentales dans les sociétés inférieures. Paris: Alcan. Lévy-Bruhl, Lucien. 1922. La mentalité primitive. Paris: Alcan. Lévy-Bruhl, Lucien. 1927. L'Ame primitive. Paris: PUF, 1963. Lévy-Bruhl, Lucien. 1928. The Soul of the Primitive. Authorized trans. by Lilian A. Clare of Lévy-Bruhl 1927. New York: MacMillan. Lévy-Bruhl, Lucien. 1931. Le surnaturel et la nature dans la mentalité primitive. Paris: PUF, 1963. Lévy-Bruhl, Lucien. 1935. La mythologie primitive: le monde mythique des Australiens et des Papous. Paris: Alcan. Lévy-Bruhl, Lucien. 1938. L'Expérience mystique et les symboles chez les primitifs. Paris: Alcan. Lévy-Bruhl, Lucien. 1949. Les Carnets de Lucien Lévy-Bruhl. Paris: PUF. Lévy-Bruhl, Lucien. 1952. "A Letter to E. E. Evans-Pritchard." Dated 14 November 1934. British Journal of Sociology 3:117-23. Lévy-Bruhl, Lucien. 1973. Primitives and the Supernatural. Authorized trans. by Lilian A. Clare of Lévy-Bruhl 1931, originally published in 1935. New York: Haskell House Publishers, Ltd. Lévy-Bruhl, Lucien. 1975. The Notebooks on Primitive Mentality. Trans. of 1949 by Peter Rivière. New York: Harper & Row. Lévy-Bruhl, Lucien. 1978. Primitive Mentality. Authorized trans. by Lilian A. Clare of Lévy-Bruhl 1922, originally published in 1923. New York: Macmillan. Lévy-Bruhl, Lucien. 1983. Primitive Mythology: The Mythic World of the Australian and Papuan Natives. Trans. of 1935 by Brian Elliot. St. Lucia, Australia: U. of Queensland Press. Lévy-Bruhl, Lucien. 1985. How Natives Think. Authorized trans. by Lilian A. Clare of Lévy-Bruhl 1910, originally published in 1926. Princeton: Princeton U. Press. Lévy-Bruhl, Lucien. 1989. "Ce qui est vivant, ce qui est morte dans la philosophie d'Auguste Comte." Revue philosophique de la France et de l'étranger 114:479-80. Lévy-Bruhl, Lucien et al. 1923. "La mentalité primitive." Bulletin de la Société française de Philosophie 18:2-48. Lévy-Bruhl, Lucien et al. 1929. "L'Ame primitive." Bulletin de la Société française de Philosophie 29:105-39. Littleton, C. Scott. 1985. "Lucien Lévy-Bruhl and the Concept of Cognitive Relativity," introduction to Lévy-Bruhl 1985. Merllié, Dominique. 1989a. "Présentation: Le Cas Lévy-Bruhl." Revue philosophique de la France et de l'étranger 114:419-48. Merllié, Dominique. 1989b. "Lévy-Bruhl et Durkheim. Notes biographiques en marge d'une correspondence." Revue philosophique de la France et de l'étranger 114:493-514. Mauss, Marcel. 1925. "Essai sur le don: forme et raison de l'échange dans les sociétés archaiques." L'Année sociologique, n.s. 1:30-186. Mauss, Marcel. 1990. The Gift: The Form and Reason for Exchange in Archaic Societies. Trans. of Mauss 1925 by W. D. Halls. New York: Norton. Mousalimas, S. A. 1990. "The Concept of Participation in Lévy-Bruhl's 'Primitive Mentality.'" Journal of the Anthropological Society of Oxford 21:33-46. Nandan, Yash. 1977. The Durkheimian School: A Systematic and Comprehensive Bibliography. Westport, CT: Greenwood Press. Schmaus, Warren. 1994. Durkheim's Philosophy of Science and the Sociology of Knowledge: Creating an Intellectual Niche. Chicago: U. of Chicago Press. PAGE 30 PAGE 1 PAGE 34 PAGE 35