The prevailing paradigm of the nineteenth century scholarship was classical evolutionism; at the end of that century, however, this paradigm fell into disfavor. Accordingly, the studies of Durkheim and Mauss were written in the aftermath of evolutionary theories. Although neither was a hard-core evolutionist, evolutionistic schemes were always more or less implicit in their work. However, because they differed from classical evolutionists in how they applied this scheme, a monolithic interpretation of evolutionary theory in late-nineteenth century scholarship is not possible. This article examines both the developments within, and the differences between, Durkheim's and Mauss's use of evolutionary models, and the scholarly reception of their work.

1. Introduction

At the end of the nineteenth century the schemes of classical evolutionism fell into disfavor. This was partly due to the accumulation of ethnographic studies which ensured that the global speculations of the early "armchair" theorists proved untenable. In fact, as early as 1896 Boas had criticized the unilineal character of the evolutionistic paradigm, and in 1918 Laufer described the paradigm of cultural evolution as "inane, sterile and pernicious" (Harris 1968: 293). In Europe interest soon shifted from evolutionary interpretations to the description of social systems as functioning wholes.

Emile Durkheim and Marcel Mauss carried out their research on society in the latter days of classical evolutionary theories. Reading their work today, it is clear that they participated to a high degree in this paradigm. For example, for both Durkheim and Mauss the religion of the Australian aborigines was crucial for studying the origins of religion and prayer. However, despite their obvious dependence on the evolutionary theories of their time, they also criticized these theories. Analyzing their various statements on evolution suggests that there was a lack of coherence to their thought, a lack of coherence that resulted in a certain dichotomy when it came to the reception of their work. Whereas their evolutionistic statements were emphasized and criticized by some, others declared those same statements to be of minor interest when compared to the innovations Durkheim and Mauss made in the study of society. Sadly, neither interpretation leads to a clear assessment of the evolutionistic aspects of their theories.

In this article I will first survey the divergent receptions which met the theories of Durkheim and Mauss. I will then examine the differing ways in which evolutionary theories were actually incorporated into their work. Studying their writings in this less monolithic way might shed more light on the internal oppositions between Durkheim and Mauss in the Annie sociologique group.

2. The reception of Durkheim's and Mauss's evolutionism

Few writers would contest that evolutionary theories are found in Durkheim's work. For example, although Picketing deals with the topic of evolutionism only in the margins of his study on Durkheim's sociology of religion, he nonetheless labels Durkheim an "evolutionist in the broad sense of the word" (Pickering 1984: 106). In support of this he quotes a letter from Durkheim to Richard, dating from 1899, in which Durkheim refers to "a certain number of elementary notions (I do not say of logical simplicity) which dominate all man's moral evolution" (cited in Pickering 1984:107). Picketing concludes: "Here indeed is Durkheim's confession of evolutionism!" (1984: 107). Despite this widely shared conclusion, in the reception of Durkheim's oeuvre we come across two distinct schools of thought: on the one hand, his statements are emphasized and criticized, on the other, they are regarded as largely irrelevant to Durkheim's thesis.

2.1 Durkheim's critics

Already in 1914, Durkheim's evolutionary approach was criticized, as in when Wallis referred to Durkheim's use of an unacceptable "principle of evolution" (1914: 255-257). Schmidt was of the same opinion: "Durkheim revelled in the most orthodox forms of Evolutionism" (Schmidt 1931: 116). Van Gennep criticized Durkheim's identification of the stage of the Australian's material civilization with that of their social organization. His point was that the more one knows of the Australians, the more one realizes that their societies are very complex and are therefore far removed from the simple and primitive; they are highly evolved according to their own directions (van Gennep 1913: 389).

Despite criticism that classed him as an evolutionist, Lowie remarked that Durkheim "at least in theory revolted against the idea of unilinear evolution, which he deprecated as oversimplified" (Lowie 1937: 201). In support of this interpretation, Lowie cites a comment in Durkheim's review of an essay by Steinmetz: "People argue as though the so-called savages or primitives formed a single identical social type" (cited in Lowie 1937: 201). Despite this, Lowie admits that, in practice, "Durkheim remained an evolutionist of the old school, relapsing into the parallelist error of treating Australian and American societies as rungs of one ladder" (203)--a criticism that, he maintains, mainly applies to The Elementary Forms of Religious Life (206).

More recently, Levi-Strauss suggested that Durkheim confused the historical and the logical point of view (1945: 516). Although Durkheim fought against classical unilinear evolutionism, Levi-Strauss argues that he did so "not so much because it is unilinear or because it is an evolution, as because he was not satisfied with the kind of data which Comte and Spencer had arranged serially" (1945: 525). Whereas Comte and Spencer used historical data, Durkheim used social types or species. Those species however were arranged in genetic series.

Related to these comments are those of Evans-Pritchard who suggested that Durkheim "tried to safeguard himself by saying that he did not use 'primitive' in a chronological sense but only in a structural sense; this was just a trick, for he was too much under the influence of Herbert Spencer not to equate in his thought the two senses and to seek in what he regarded as the structurally most primitive the most primitive in time" (1981: 156). Criticizing this use of evolutionary theory, Evans-Pritchard concluded that "Durkheim was an evolutionary fanatic who wished to explain social phenomena in terms of pseudohistorical origins" (1981: 161). Accordingly, he does not regard Durkheim as a scientist, but "at the best as a philosopher, or I would rather say, a metaphysician" (1981: 168). Moreover, he considers Mauss to be part of the same philosophical tradition--one that stretches from Turgot and Condorcet to Comte and then Durkheim--"a tradition in which conclusions were reached by analysis of concepts rather than of facts, the facts being used as illustrations of formulations reached by other than inductive methods" (1981: 190). According to Evans-Pritchard, Mauss, however, was far less a philosopher than was Durkheim.

Accordingly, one tradition understands Durkheim to have been an evolutionary die-hard. Although Durkheim might have protested that he was not, he was nonetheless understood as working within the evolutionary paradigm.

2.2 Durkheim's supporters

Contrary to this critical reception of Durkheim's evolutionism, Nisbet understood Durkheim as a key figure in "the growing recognition of the need for better explanatory procedure" (1975: 66); seen in this way, Durkheim played a significant role in improving on the procedures of evolutionary and diffusionist schemes that were in vogue during the nineteenth century. Although acknowledging that Durkheim regarded Australian religion as the "simplest"--noting that "[i]n this belief Durkheim was, of course, not alone" (1975: 167)-Nisbet observes that, when Durkheim discussed the topic of "origins", he "has something in mind more near logical than anything that might be thought to be the beginning of religion for all mankind" (167). Despite this important observation, Nisbet concedes that Durkheim's mind was indeed oriented "in substantial degree towards the social evolutionism that had flourished in the nineteenth century" (167). However, he concludes that it is not so much this evolutionary flavor that gives lasting distinction to Durkheim's study, "but the purely analytic and structural framework that is overwhelmingly the essence of the book" (168).

Morris postulates that the nineteenth-century approach of such writers as Muller, Spencer, Tylor, and Frazer--a common approach that viewed human culture from an evolutionary perspective--was in fact strongly challenged by Durkheim. "The rejection of evolutionism, however, also involved the repudiation of historical analysis itself in favor of an organic functionalism. This latter mode of sociological analysis is closely associated with Durkheim" (Morris 1987: 106). Although agreeing that Durkheim had an evolutionary approach to social life--going so far as to admit that, "like Tylor and Frazer, he [Durkheim] took an evolutionary perspective and was preoccupied with discerning the origins of religion" (114)--Morris maintains that "it was expressed at a highly abstract level involving an ideal-type dichotomy between what he termed "mechanical" and "organic" forms of social solidarity" (107).

According to Lukes, Durkheim's view of totemism differed from the evolutionists in that it "looked more to the past than to the future" (1973: 454). The main problem he finds with Durkheim's work is an ambiguity in his use of the term "primitive"; Durkheim means by the "most primitive" religion both the simplest religion as well as the earliest. Whereas "[t]he first interpretation represents Durkheim's residual evolutionism ... the second [represents] his inclination to study relatively undifferentiated, small-scale and closed societies. The first points back to the pervasive evolutionism of the nineteenth century; the second points forward to modern social anthropology" (Lukes 1973: 456). Lukes solves this problem by stating that "the enormous interest and importance of The Elementary Forms is largely independent of its evolutionism; and it is better seen as a classic of social anthropology than as a remote exercise in religious history using the data of ethnography--though Durkheim saw it, inseparably, as both." (1973: 456-457). In other places, Lukes is willing to downplay Durkheim's evolutionism, especially if it is considered as a form of simple, unilinear evolutionism (1973: 140, 149 n. 49, and 281 n. 27).

2.3 Mauss compared to Durkheim

While the critical literature on Durkheim is abounding, only a few critical studies have been devoted to Mauss. Often he is simply referred to as the author of the 1925 essay, "Essai sur le Don". When he is mentioned, commentators usually emphasize his notion of the "fait social total" and his understanding of reciprocity, both of which appear in this essay.

To Levi-Strauss this essay is "sans contestation possible, le chef-d'oeuvre de Mauss" (1950: xxiv). Comparing Durkheim and Mauss, Levi-Strauss has a strong predilection for the latter. Above we saw Levi-Strauss criticizing Durkheim for fighting unilinear evolutionism with the wrong arguments. This tendency, however, is "far less conspicuous in recent works (such as Mauss's) than in the earlier ones" (Levi-Strauss 1945: 525). Whereas Durkheim sees in primitive societies the early stages of social evolution, Mauss examines such societies simply "because they exhibit social phenomena under simpler forms." (1945: 527). Accordingly, we could concludes that Mauss did and Durkheim did not escape the temptation to reconstruct synchronically.

A few years after Mauss's death, Leacock surveyed those aspects of Mauss's work which had direct bearing on the development of ethnological theory (1954); this paper is significant because in it we find the ambiguity that characterizes the reception of Mauss's work. On the one hand, Leacock refers to Mauss's use of quotation marks when using the term "primitive" and to his explicit denial of any wish to find laws of progress or of general evolution. He argues that Mauss "held that there were no general, universal laws of social phenomena, but rather a number of laws of unequal generality" (1954: 62). However, he then goes on to argue that, when dealing with the so-called "primitives", Mauss's "thinking was always within an evolutionary framework" (60). Leacock expresses the ambiguity as follows:

Mauss denied that these simple peoples, e.g., the Australians, represented anything like the earliest or most primitive groups of mankind, but he did feel that they represented an earlier stage of cultural development than other, more complex, societies .... Mauss certainly did not make use of an elaborate evolutionary scheme in which all cultures were placed in a unilinear sequence, but he did on occasion trace the development of institutions through a series of societies widely separated in time and space in the classical evolutionary manner. (1954: 60)

Leacock illustrates this evolutionary emphasis by referring to Mauss's essay on classification and what he terms "its bizarre evolutionary sequence" (1954: 64) and to his essay on the gift, in which societies with a simple type of prestation totale are equated with neolithic civilizations and where he employs the concept of "survivals" (66). He finds this evolutionary thinking still apparent in Mauss's Manuel d'ethnographie (1947), where "Mauss places the Eskimo in the Upper Paleolithic and the Australians in the Aurignacien" (1954: 70).

Cazeneuve states that, like Durkheim, Mauss was also attached to the evolutionistic theories about the progress of the human mind--but in a less systematic and doctrinaire way than his uncle (Cazeneuve 1968b: 13). Cazeneuve also refers to Mauss's views that certain economic practices are the remains of an archaic past (1968b: 30). Here, we must recall that in studies on such topics as classification or the evolution of prayer, data from the Australian aboriginal tribes were of central importance to Mauss, for he, just as much as Durkheim, wanted to go back to the elementary forms. Cazeneuve suggests that Mauss, unlike Durkheim, was able to avoid the excesses of "pantotemisme" (1968b: 34). In another work of the same year he discusses Spencerian evolutionism which is, according to Cazeneuve, hardly present in Mauss's work--one must recall that Durkheim defended himself against charges of his own Spencerian evolutionism. He concludes that Mauss seems not to have accepted "in its extreme consequences the rationalist, Durkheimian hypothesis ... according to which everything we come across in our civilization, be it religion or science, emanates from totemic civilization and even more specific, from the collective primitive life by way of a necessary development" (Cazeneuve 1968a: 10-11).

Instead of trying to find the originary point in an evolutionary process, Mauss was in search of fundamental and universal phenomena. Despite this, however, Cazeneuve admits that in his study of prayer, evolutionary models are used; moreover, Mauss, like Durkheim, thought that Australian aborigines represented the most archaic state of humanity (Cazeneuve 1968a: 77). However, this evolutionary approach may be contrasted with his study on sacrifice. Cazeneuve juxtaposes Mauss's theory of sacrifice to that of Robertson Smith; whereas one looked for the "unite generique" of the several forms of sacrifice, the other traced the "unite genetique" in a continuous process of evolution (1968a: 86). Finally, in his discussion of Mauss's study of the gift and reciprocity, Cazeneuve distinguishes between the apparently irrelevant "arriere-plan evolutioniste", analogous to that of Durkheim, and the "simplification evolutioniste implicite"--the important fundamental structures found in antiquity and in primitive society (1968a: 106).

In 1968, Karady presented the first volume of the works of Mauss. In his introduction Karady hints at the difference between Durkheim and Mauss: unlike Mauss, Durkheim developed a "philosophie ... essentiellement evolutionniste" based on a "methode genetique" (Karady 1968: xxiii). Though he praised Durkheim's genetic explanation, Mauss offered a more nuanced interpretation of the elementary phenomena (Karady 1968: xxiii, n. 34). A further distinction between Durkheim and Mauss emerged once the predominance of the notion of "the sacred" disappeared, for function, instead of genesis, became important (Karady 1968: xliv). Finally, he notes that, from the 1920s onwards, Mauss opposed the term "primitive" as a "prejuge evolutionniste" (Karady 1968: xlv).

As we saw above, Evans-Pritchard placed Mauss in the tradition of Montesquieu, Comte, and ultimately Durkheim. However, in his discussion of Mauss, Evans-Pritchard does not mention his evolutionism, though in Mauss's essays he does identify an always implicit "comparison, or contrast, between the archaic institutions he is writing about and our own" (Evans-Pritchard 1981: 191). And in his introduction to Mauss's study of the category of the person, Allen is well aware of the evolutionistic scheme used by Mauss. However, Mauss's "careful evolutionism" can of course be translated "into synchronic terms" (Allen 1985: 42). According to Allen, then, "Mauss avoided the crudity of certain styles of evolutionism" and it would be better to talk "of Mauss's world-historical awareness" (1985: 27).

2.4 Initial conclusions

Those who have commented on Durkheim and Mauss generally fall into one of two groups: the one emphasizes the evolutionary aspects of their work and the other more or less denies or at least downplays the significance of this aspect to their work. In the case of Durkheim, some of the criticisms were severe: Wallis, Schmidt, and van Gennep denounced Durkheim's evolutionism; Lowie and Levi-Strauss did the same, but in guarded terms. The judgement of Evans-Pritchard was very harsh.(n1)

In accord with these differing judgments, Nisbet identifies a clear ambiguity in Durkheim's works. On the one hand, according to Nisbet, he was indeed thinking within the evolutionary paradigm, whereas on the other hand, he can be labeled as anti-developmentalist. For instance, while his books on the division of labor and on the elementary forms of religious life are clearly evolutionary, in The Rules of Sociological Method evolutionism is criticized and in his study on suicide it is completely absent. Observing this ambiguity, Nisbet asks himself: "How far was Durkheim himself aware of the fact that he was, on the evidence of his writing, both developmentalist and antidevelopmentalist?". His answer? "It is impossible to say" (Nisbet 1975: 249). However, he does go on to suggest that "[t]here is always the possibility, by no means a unique one in Durkheim's writing, of simple contradiction. It is a rare book of major philosophic importance that does not have an inconsistency or two in it" (1975: 250).

As we have seen, the evolutionary aspects in the work of Mauss have been emphasized less than those of Durkheim. Levi-Strauss, Cazeneuve, Karady, Evans-Pritchard and Allen all tone down the importance of evolutionary concepts in his theories. According to their reading of Mauss, evolutionism is only a secondary aspect to his work. In fact, Leacock was the only one to criticize Mauss in this respect.

3. Durkheim and classical evolutionism: History and analysis

Durkheim used the evolutionary model in two ways. In several of his studies he focused on modem, Western society. To understand such things as the function of the division of labor, the rise of suicide, and the modern functions of punishment he developed new concepts, concepts that could be applied to primitive and for modern societies. However, because the focus of these studies was social phenomena in modem society, data and insights derived from so-called primitive societies simply functioned as a point of contrast that could elucidate a social event in modern society. In these studies, primitive and modem societies therefore served as opposite, ideal-types; they were not necessarily connected by means of some evolutionary or developmental scheme.

In a later phase, however, Durkheim became more and more interested in primitive societies. Like before, he presumed that a more intensive study of primitive societies would contribute to a better understanding of modern society; however, primitive and modem societies were increasingly seen as two inter-related points in an evolutionary line. It is at this point that he began to rely on a model more in line with classical evolutionism.(n2) Durkheim wanted to create an autonomous field of study, one that was capable of examining more than just the political society of the nation but one which was more limited than studying humanity per se. This field was the study of civilization, defined as "a kind of moral milieu encompassing a certain number of nations, each national culture being only a particular form of the whole" (Vogt 1976: 36). (This distinction between nation and civilization is also important in Mauss's work [Cazeneuve 1968a: 21, 22].)

To better understand Durkheim's evolutionism, it is necessary to examine each of these two periods.

3.1 The early period

The first study in which Durkheim explicitly dealt with evolutionary concepts was his 1893 study of the division of labor and its relation to social solidarity. In this study Durkheim opposed two kinds of solidarity: mechanical and organic. In a society with mechanical solidarity, all individuals are linked directly, and all in the same way, to society. In a society with organic solidarity, every individual is linked to society as organs within an organism--each has its own function. In the first instance, all members feel and believe in exactly the same way and "[t]his solidarity can only increase in inverse proportion to personality" (Durkheim 1902: 99). A high degree of mechanical solidarity means a high degree of "conscience collective", and at this moment "notre individualite est nulle" (1902: 100). On the contrary, organic solidarity is characterized by differences and a high degree of individuality. The former attains "son maximum de rigueur", "chez les peuples inferieures" (1902: 391). The model for this kind of solidarity is based mainly on classical and biblical antiquity. Only a few pages refer to non-literate civilizations (1902:149-157).

According to Durkheim, one can observe progress from mechanical to organic solidarity. However, in this early period he does not attribute this progress to innate, unilinear evolution. Instead, he postulated a general sociological law: "The division of labor varies in direct proportion to the volume and the density of societies, and if it is progressing in a continual way during the social development, that is because as a rule societies become more dense and very generally more voluminous" (1902: 244). He related both forms of solidarity to juridical codices; mechanical solidarity results in pressure from the collectivity on the individual and in an emphasis on the penal code. He argues that this code is disappearing in modern times while no new laws are as yet formulated--the penal code is, in his opinion, now in an "etat de crise" (Durkheim 1901 a: 95). Organic solidarity results in an emphasis on civil law. Accordingly, Durkheim opposed penal repression to civil restitution (1902: 34, 102, 127-128).

The distinction between the study of nations and the study of civilizations was already present in Durkheim's work on methodology in the social sciences. There, he juxtaposed the historian, who produces descriptive monographs, to the philosopher, who studies humanity. Durkheim then identified the possibility of investigating something else: "il y a des intermediaires: ce sont les especes sociales" (Durkheim 1901b: 77). However, he was reluctant to compare disparate societies, such as comparing ancient Greece to thirteenth-century feudal French society or contemporary primitive societies (1901b: 82). Instead, he postulated a general principle of classification based on the degree to which the members of any given society are ordered into a social structure (1901b: 86). Because he presumed that human social systems began with the horde and moved to more complex societies, Durkheim was surely working within the evolutionary paradigm at this point in his career. However, in this early phase his intention was to formulate social species and not to reconstruct historical phases (1901b: 88, n. 1). In fact, in this study he criticized Comte and Spencer, though he never blamed them for their evolutionary approach. Instead, he directed his criticism at Spencer's definitions (1901b: 21, 38) and at Comte's study of humanity instead of society, (1901b: 77). In fact, in his conclusion Durkheim came up with a rather negative statement about evolutionism: sociology, he suggested, has been too strongly influenced by philosophical doctrines; it has been "successively positivistic, evolutionistic, spiritualist, while it should confine itself to simply being sociology" (1901b: 139).

Durkheim's focus in the last study to be considered as part of his early period was the study suicide; it was not an evolutionary study whatsoever. His aim was to demonstrate the direct relevance of sociology even when studying such seemingly individual acts as suicide (for his other motivations see Lukes 1973: 191-194). In his explanation of this phenomenon, Durkheim distinguished three types of suicide: egoistic, anomic, and altruistic suicide. The first two are found in modern society: egoistic suicide was "clearly linked by Durkheim to the growth of the 'cult of the individual'" (Giddens 1971: 85) and anomic suicide was related to the lack of moral constraint in modem society. Both forms are typical of modern societies characterized by organic solidarity. In societies organized by means of mechanic solidarity and with a strong conscience collective we find altruistic suicide, where individuals sacrifice their lives to further some collective value (the opposite of the cult of the individual). In societies with altruistic suicide it is necessary that "la personnalite individuelle compte alors pour bien peu de chose" (Durkheim 1897: 237). Accordingly, this last form of suicide can also be found in modern armies (1897: 247-261). Although it is surely possible to place altruistic and egoistic/anomic forms of suicide on an evolutionary scale, Durkheim did not intend the concepts for such a purpose.

3.2 The later period

In his later phase, Durkheim began studying civilizations. The contrast of traditional vs. modern no longer functioned as it once had in his work. Durkheim now claimed that in order to understand social institutions it is necessary "to go back as far as possible to its first beginnings", because, between what is and what has been, "il y a ... une etroite solidarite" (1898: 1). Because history and sociology were now considered to go hand in hand, an explanation of such things as, for example, the incest taboo now became possible. Durkheim discerned an evolution from the sacred character of the totem animal, to the taboo on its blood and then the blood of the members of the totem group. He concluded that a result of these identifications was the taboo on women of the same group. Hence, a dichotomy came to light: "le bien et le plaisir, le devoir et la passion, le sacre et le profane" (1898: 61). The presumed evolution from clan to family implied the opposition of the sacred family--with its rule to love one another--to the profane non-family, which comes with the possibility for sexual relations. To arrive at these conclusions, Durkheim leaned heavily on the works on Australian aborigines written by such people as Curr, Fison, and Howitt as well as the compilations of Dawson and Frazer.

In 1903 Durkheim and his nephew Marcel Mauss published their article on primitive classification. In this article, they proposed to study the history of logical thought, which necessitated examining "the most rudimentary classifications made by men" (Durkheim and Mauss 1903: 7). These so-called rudimentary systems of classifications were those of the Australian tribes. After describing this primitive form of classification (1903: 7-34), they moved on to describe classificatory systems with what they considered to be higher degrees of complexity, relying on data derived from American Indian (mainly Zuni) cultures (1903: 34-55). The third system of classification they considered originated with Chinese philosophy. The main hypothesis of their article was based on what Durkheim and Mauss called "sociocentrisme" (1903: 70): categories of classification emerge from social categories. "The first logical categories", they wrote, "have been social categories .... The phratries were the first varieties; the clans the first species" (1903: 67). From this type of classification starts the evolution to modern science. The history of scientific classification is therefore the history of the decline of the social, affective element, "leaving more and more room for reflection and thought of individuals" (1903: 72).

When considering Durkheim's evolutionism, his Les formes elementaires de la vie religieuse is by far the most important study--something that is apparent from its title and opening lines. In this work, Durkheim wished to study the most primitive and simple religion. By "most primitive societies" he meant, in the first place, societies which are not surpassed by another one in simplicity; moreover, he also meant that phenomena in these societies have to be explained without reference to an earlier form of religion (Durkheim 1912: 1). The first statement concerns structure, the second concerns chronology. When he used the term primitive, Durkheim had both meanings in mind at the same time. He defended his position of going back "aux debuts de l'histoire" (1912: 2). For him, recent religions can only be understood by tracing back in history the way in which they have been progressively composed (4).

For Durkheim, primitive in time and simple in structure coincided; he opposed both to developed and therefore complex societies (4-5). In the study of complex religions "qui apparaissent dans la suite de l'histoire", nothing can be found about "les etats fondamentaux, caracteristiques de la mentalite en general" (7). In inferior societies, however, differences and variations hardly occur; individuality hardly exists. So, Durkheim remarked, "[en] meme temps que tout est uniforme, tout est simple" (8). Besides, those societies lack luxury and sacerdotal speculation; everything is reduced to the indispensable, and "l'indispensable, c'est aussi l'essentiel" (8). When one goes back to a period "plus proche de ses debuts" (10), one does not find representations which are elaborated and denatured by scientific reflection (9); because primitive religions still bear the imprint of their origins (10), it is much more difficult to make inferences from the more developed religions. Accordingly, Durkheim admitted that he also studied the old problem of the origin of religion. However, he opposed the subjective speculations of older theorists; instead, he attempted to find the most essential forms of religious thinking and practice, emphasizing the structural component to his work and noting that the causes of the most essential forms of religion can be discerned much more easily in less complicated societies (11). The societies that attracted his attention were therefore those characterized by primitive social organization. Durkheim therefore focused on Australian societies because he took them to represent the simplest social organization known (88). "A l'origine" was therefore equated with "en Australie" (95). Moreover, because beliefs which were near universal in Australia could also be found in North American Indian cultures (96), the comparative method could then be used to arrive at a more primitive religion from which these others were derived (101). For Durkheim, this form of religion turned out to be totemism, and especially the totemism practiced in Australia.

Clearly, Robertson Smith, McLennan, and Tylor had already realized the importance of totemism. However their research depended mainly on data derived from North American Indian cultures--cultures which contained, according to Durkheim, "a lot of traces of totemism, but had already passed through the properly totemic phase" (128). Summarizing the reasons for limiting himself to Australian data, Durkheim remarked that these cultures: (1) are homogenous, all belonging to the same type; (2) are now very well documented; (3) are very near to the possible origins of evolution and contain accordingly the most primitive and most simple religion (135). To support his last statement, Durkheim pointed to their rudimentary level of technology (e.g., absence of houses and even of huts) and to the simple social organization of the Australian aborigines ("organization a base de clans" [136]). Apart from Australian material, Durkheim was interested in using data from Indian cultures in North America, but he deemed these tribes more advanced than the Australian ones. To support this conclusion he pointed to their more advanced technology as well as to their social organization of large confederations organized under a central authority. However the basic social structure "c'est toujours l'organisation a base de clans" (136). Certain vague concepts were also more explicitly expressed in American Indian cultures and in America the totemic system had better defined forms (156). Durkheim therefore concluded that the American totemic system could explain aspects of Australian totemism (138).

4. Mauss and classical evolutionism: History and analysis

Much as we distinguished between earlier and later periods in Durkheim's career, in the works of Mauss it is useful to distinguish between the work done before 1914 and after 1918.

4.1 Before 1914

Mauss's earliest essay was an extensive review of a work by Steinmetz. In that essay, he criticized Steinmetz for his lack of attention to the definition of his subject and his arbitrary use of ethnographic material. But between the lines, Mauss made suggestions that are directly relevant for our investigation: for instance, he used notions such as "des l'origine" (Mauss 1896: 40) and "les societes les plus elementaires" (1896: 49). Moreover, he emphasized the evolution of punishment, from its religious and instinctive origins to the rational and social ideal of modern justice systems (59). At this early point in his career, Mauss agreed with the evolutionary approach of Steinmetz; he simply criticized his psychological approach (271).

Mauss co-wrote his 1899 essay on sacrifice with Henri Hubert. This work examined the nature and the social function of the sacrifice (Mauss and Hubert 1899: 31)--in his introduction to the English translation, Evans-Pritchard remarked that, "as a study of the structure, or one might almost say the grammar, of the sacrificial rite the Essay is superb" (1964: viii). This work was based on three assumptions. First, they understood sacrifice to be a ritual of communication that is comprised of three phases: entree, victime, and sortie. Second, sacrifice was studied as a "fait social totale". Third, we find certain evolutionary notions employed. For example, they argue that from early sacrifices which were believed by participants to release a life creating force, all kinds of later religious representations evolved (1899:115-131).

In this work, Mauss and Hubert opposed the English anthropologists' preoccupation with accumulating and classifying documents. Instead, Mauss and Hubert intended to study "des faits typiques" (1899: 34). Although they took their data from Sanskrit texts and the Bible, going back in time to the oldest texts did not make much sense to them. In fact, they emphasized the unreliability of ethnographic material. Their study was not about the history and genesis of sacrifice; wen they used the notion of anteriority, Mauss and Hubert meant "anteriorite logique" and not "anteriorite historique" (1899: 35). Whereas Robertson Smith's study of sacrifice was looking for a "unite genetique dans une evolution", Mauss and Hubert sought for a "unite generique des diverses formes de sacrifice" (Cazeneuve 1968a: 86).

This suspicion of ethnographic material is nicely in line with Vogt's distinction between the nation and civilization periods in Durkheim's oeuvre. However, Mauss and Hubert expressed themselves more rigidly and still did so when Durkheim was already writing his essay on the history of incest. In fact, Mauss had opposed the method of random comparisons as early as 1896: "One single fact, critically investigated, can prove a sociological hypothesis" (Mauss 1896: 37). He therefore criticized the theories of Robertson Smith, Frazer, and Jevons for tracing the various forms of sacrifice back to one arbitrarily chosen principle, which was the universality of totemism. Mauss and Hubert remarked that "the universality of totemism, starting point for the whole theory, is just a postulate" (1899: 32). However, in 1909 Mauss and Hubert published their article on sacrifice again, rectifying some problems in the preface. One of these points was their remark about the postulated generality of totemism. They admitted that totemism existed not only in Australia and America. Since 1898, "ethnographers have multiplied the proofs for its existence and the reasons to believe in its generality" (Mauss and Hubert 1909: iv).

Because the article on sacrifice dealt with a solemn and public ritual, it must have seemed rather obvious that its nature would be a social one. So, in 1904 Mauss and Hubert published a study on a group of rituals which are hardly solemn and public: magic. Sadly, the distinction between magic and religion made by Mauss and Hubert is a rather ambiguous one (Belier 1995a). They attacked the theories of Tylor, Frazer, and Lehmann; their main quarrel with these writers was with their failure to develop an adequate definition. In setting out to define and thereby distinguish magic from religion, Mauss and Hubert did not focus on the form taken by their rites but by the circumstances in which these rites appeared. A magic rite is therefore "every rite which is not part of an organized cult, which is a private, secret and mysterious cult, which has a tendency to slip over into that which is prohibited" (Mauss 1950:16).

In one chapter they dealt with examples of specific magic phenomena, taken from a wide array of cultures: classical and modem India, Europe in the Middle-Ages, ethnographic manuals, classical Greek and Rome, etc. Based on this data, they related magic and religion to beliefs in mana and orenda, conceived as collective, social forces. These terms refer to a "notion de force-milieu magique" (1950: 109). They concluded that from the notion of mana there evolved the category of causality, suggesting that magic was directly related to the concrete. In their view, magic was the basis of science and technology. Where classical evolutionists presupposed an independent development from magic and religion to metaphysics and science, Mauss and Hubert found in magic the formation of a fundamental category of logic, one necessary for scientific thinking.

In 1904, Mauss published more data about magic. In a "province ethnographique", a term which he borrowed from Bastian (Mauss and Hubert 1909: 133), Mauss now studied data from several Australian tribes. Australian data was used, however, explicitly as a "province ethnographique" and not because of its supposedly archaic character. Then, two years later, Mauss published his Baconian experiment, employing data from only one society, the Inuit. Together with Beuchat, Mauss demonstrated the interdependence of social density and religious phenomena. The contrast between summer and winter as regards population density had become one of the main principles of classification and this contrast was well pronounced in Inuit cultures. However, because they concluded that this principle applied to other cultures as well, Mauss and Beuchat proposed a general law: "Social life does not maintain itself at the same level at different moments of the year; it passes through successive and regular phases of increasing and decreasing intensity, of rest and activity" (Mauss 1950: 473). For our present purposes it is crucial to observe that data from Inuit cultures was not used because it derived from a primitive culture, but, rather, because it illustrated the general principle so well.

The last study from this early period that we will discuss is Mauss's research on prayer. Although Mauss had meant to write a thesis about this subject, he only finished the first chapter and nothing of this thesis was officially published. Instead, the first chapter was spread among his intimate friends in dactylography and Karady published it in his collection of Mauss's writings. Referring to Durkheim's Rules, Mauss opens with a definition of prayer: "la priere est un rite religieux, oral, portant sur les choses sacrees" (414). He understands prayer to be a central institution which is linked with belief representations and with the cult. Significantly, he argues that prayer is a good instrument to measure the "avancement d'un religion" (Mauss 1968: 360). He understood this progress to consist of a growth in spirituality and individuality. Accordingly, his approach to studying prayer consisted of three parts: it must deal with the most elementary forms of prayer, with its role in advancing spirituality, and with its role in advancing individualization. For the first part, Mauss intended to study Australian data, for the second he planned to use data from India, and for the third the Judaeo-Christian culture. Australia was chosen not only because he considered it to form a well bordered ethnographical province but also because its culture was very primitive (421).

4.2 After 1918

It is certain that before the First World War Mauss was not very much interested in an evolutionary approach. In a later phase, however, we find certain articles which built on the approach that seems to have emerged in his work on prayer. The most important of these studies was the one on gift-giving. In this study Mauss distinguished between societies "de type arriere ou archaique" (1950: 148) and modern society in which people became an "animal economique" (271). Societies of the first type were those of Polynesia, Melanasia, and the Indians of north-west America. After discussing the data stemming from those societies, Mauss concluded:

ce principe de l'echange-don a du etre celui des societes qui ont depasse la phase de la <<prestation totale>> (de clan a clan, et de famille a famille) et qui cependant ne sont pas encore parvenues au contrat individual pur . . . et surtout a la notion du prix estime en monnaie pesee et titree. (227)

Accordingly those societies were considered to be somewhere between really primitive ones and modern society. They were seen by Mauss as fine examples of "la grande civilisation neolithique" (266). Mauss even investigated whether there were any survivals left in the ancient laws of Rome, India, and Germany (228).

In this essay there are other points of interest than just this evolutionistic scheme. One of the more important sidelines is Mauss's moralistic tenor. Modern society is described in terms such as "la froide raison du marchand, du banquier et du capitaliste" (270). Citing a Maori verb, "Give as well as take and all will be right" (translation: Rev. Taylor, cited in Mauss 1950: 265, n. 2), Mauss stated that "d'un bout a l'autre de l'evolution humaine, il n'y a pas deux sagesses. Qu'on adopte donc comme principe de notre vie ce qui a toujours ete un principe et le sera toujours: sortir de soi, donner, librement et obligatoirement; on ne risque pas de se tromper" (265).

The same evolutionistic approach, but without the moralistic tenor, can be found in his essay on the physical effects of the idea of death. For his theory Mauss looked for data stemming from Australia, "la plus inferieure [societe] possible ou plutot la plus inferieure connue" (314). These data were then compared with those from Polynesia and New Zealand and, finally, these primitive societies were compared with modern societies (329). The same comparative structure was the basis for Mauss's analysis of another category, the notion of the person, or a moi. This also was a notion which has been "lentement nee et grandie au cours de long siecles et a travers de nombreuses vicissitudes" (333). Just like his study of the notion of death, this investigation led Mauss from Australia to Europe (334).

A contrast can be seen between these last studies (even including his study about prayer) and Mauss's earlier works. Australian data came to be used explicitly because of what was seen to be its archaic character. Mauss was at this moment well in line with the assumptions of classical evolutionism. Although his earlier work contained certain evolutionist explanations, during that period Mauss was mainly interested in the social conditions and not in the evolution of notions about such things as contact between sacred and profane, the rise of the category of causality, or the relation between social density and religion. Although evolutionary schemes are more or less present in his work from both periods, by stressing the logical instead of the chronological character of "anteriority", Mauss--more than Durkheim --seems somewhat distanced from social evolutionary theories.

5. Conclusion

In the opinion of some sociologists and historians of science, the Annie sociologique group was knit together very closely: for example, it was difficult for outsiders to join this group and Durkheim rigidly fixed the scientific rules. I have argued elsewhere that this vision is incorrect (Belier 1994: 154-155). Between Durkheim and Mauss (and also Hubert and Hertz [Belier 1995b: 133-144]) there were many differences--one of which was the role played by social evolutionary assumptions. In fact, they did not really participate in what we would today term a research-group. They shared nothing more than some loose philosophical principles and some general notions about research-methods. The differing roles played by evolutionary theories exemplify this statement.

If we compare the members of the Annie sociologiue group with the classical evolutionists we can distinguish four groups arranged on two axes: nineteenth-century evolutionists vs. twentieth-century nonevolutionists would be found on one axis, and those who locate religion in the mind of the individual vs. those who take a socio-functionalist approach would be on the other (Service 1985: 157). Durkheim, for instance, participated in the evolutionary paradigm of the nineteenth-century comparative theorists. However, because he emphasized a non-individualistic explanation of social phenomena, he also participated in the twentieth-century socio-functionalist paradigm. Moreover, we can distinguish between evolutionists who concentrated on one or on a small group of societies and those who concentrated on humanity. For example, Spencer studied humanity--something that was severely criticized by Mauss. But Mauss did not confine himself to societies which could be compared using the Durkheimian parameters. Even his study on the winter and summer regime in Inuit societies resulted in a comparative conclusion that placed the Inuit duality on a global scale.

For Durkheim the main aim was to demonstrate the scientific character of the new science of society. In light of the individualistic evolutionism of his days, Durkheim opposed psychology as an explanation for social phenomena, but the evolutionistic strategy he took for granted. Mauss, however, was less fixated on method. Although he did presume that human history was an evolutionary process, his explanations were synchronic--about structure--and not diachronic --about origins. Durkheim was interested in elementary forms, Mauss in humanity. Despite this difference, both adhered to sociological methods, offering explanations in terms of social functions and not in terms of individual needs. But where Durkheim understood his Australian aboriginal data to represent the very beginnings of humanity, to Mauss it--like Inuit cultures--simply comprised a field of ethnological study. Accordingly Mauss conceived of these not as primitive cultures. From a "historicist" point of view however, it must be conceded that both were more (Durkheim) or less (Mauss) immersed in the older paradigm of classical evolutionism.

While in his "civilization" phase, Durkheim's use of evolution was ambiguous. In his "nation" phase, however, primitive society was opposed to modern society, using this conceptual dichotomy to elucidate certain aspects of modern society. By way of contrast with traditional societies it was possible to understand problems in the modem one. It was only in his later phase that the importance of primitive societies grew. It was only then that Australian tribes became a reference point for the "germe de 'institution" (Durkheim 1950: 190), "la base de l'institution" (195) or the "point de depart" (195) of social institutions.

It should be clear that neither Durkheim nor Mauss were hardcore evolutionists. But, despite criticizing evolutionistic schemes, social evolutionary assumptions were always more or less implicit to their work. Because these schemes were not always used in the same way, a monolithic interpretation of their work is not possible. An interesting point of difference between Durkheim and Mauss, on the one hand, and classical evolutionists, on the other, is their assessment of the evolutionary process. To the classical evolutionist the gain of evolution was an increase in rationality. Evolution was judged positively. To Durkheim and Mauss, evolution also entailed a loss or decrease in morality. They were constantly aware of problems in the contemporary society. In the case of Durkheim this awareness determined the choice of the theme's under study, e.g., problems having to do with moral solidarity in a society without religion. In the case of Mauss this awareness even showed itself in a romantic vision oil primitive society. Mauss regretted the "froide raison" of modern individualistic society. Levi-Strauss and Dumont in this respect also followed in the footsteps of their teacher.

(n1) As for his remark about the tricky nature of Durkheim's use of the term "primitive", this was not to the point. Evolutionists such as McLennan commonly distinguished between cultures, religions, and societies that were old in terms of their chronology and those that were old in terms of their structure (Burrow 1966: 12-13; Evans-Pritchard 1981: 62); Robertson Smith, when discussing origin, distinguished between (i) first form, (ii) simplest form and (iii) most significant form (EvansPritchard 1981: 78).

(n2) Vogt (1976: 35) relates this change of opinion to internal factors (having to do with the state of the discipline), to socio-professional factors (the result of conflicts with philosophers and historians), and to external factors (the general ideological views in France).

References

Allen, N.J. (1985). The category of the person: A reading of Mauss's last essay. The Category of the Person: Anthropology, Philosophy, History, 26-45. M. Carrithers, S. Collins, and S. Lukes (eds.). Cambridge: Cambridge University Press.

Belier, W. W. (1994). Arnold van Gennep and the rise of French sociology of religion. Numen 41: 141-162.

--- (1995a). Religion and magic: Durkheim and the Annee sociologique group. Method & Theory in the Study of Religion 7/2: 163-184.

--- (1995b). De Sacrale Samenleving: Theorievorming Over Religie in het Discours van Durkheim, Maus, Hubert en Hertz. Maarssen: De Ploeg.

Carrithers, M., S. Collins, and S. Lukes (eds.) (1985). The Category of the Person: Anthropology, Philosophy, History. Cambridge: Cambridge University Press.

Cazeneuve, J. (1968a). Sociologie de Marcel Mauss. Paris: Presses Universitaires de France.

--- (1968b). Mauss. Paris: Presses Universitaires de France. Durkheim, E. (1897). Le suicide: Etude de sociologie. Paris: Alcan.

--- (1898). La prohibition de l'inceste et ses origines. Annee sociologique 1: 1-70.

--- (1901 a). Deux lois de l'evolution penale. Annie sociologique 4: 65-95.

--- (190lb). Les regles de la mithode sociologique. (Revue et augmentee d'une preface nouvelle). 2nd ed. Paris: Alcan

--- (1902). De la division du travail sociale: etude sur l'organisation des societes superieures. 2nd ed., augmented. Paris: Alcan.

--- (1912). Les formes elimentaires de la vie religieuse. Paris: Alcan.

--- (1950). Lecons de sociologie: Physique des moeurs et du droit. Paris: Presses Universitaires de France.

Durkheim, E. and M. Mauss (1903). De quelques formes primitives de classifications. Contribution a l'etude des representations collectives. Annie sociologique 6: 1-72.

Evans-Pritchard, E. (1964). Introduction. In M. Mauss and H. Hubert, Sacrifice: Its Nature and Function. Chicago: University of Chicago Press.

--- (1981). A History of Anthropological Thought. London: Faber and Faber.

Gennep, A. van (1913). Review of Durkheim, Les formes elementaires de la vie religieuse. Mercure de France 101: 389-391.

Giddens, A. (1971). Capitalism and Modem Social Theory: An Analysis of the Writings of Marx, Durkheim and Max Weber. Cambridge: Cambridge University Press.

Harris, M. (1968). The Rise of Anthropological Theory. New York: Cromwell.

Jones, R. A. (1977). On understanding a sociological classic. American Journal of Sociology 83/2: 279-319.

Karady, V. (1968). Presentation de l'edition". Oeuvres 1. Les fonctions sociales du sacre, iliii. M. Mauss. Paris: Les editions de minuit.

Leacock, S. (1954). The ethnological theory of Marcel Mauss. American Anthropologist 56: 58-73.

Levi-Strauss, C. (1945). French sociology. Twentieth Century Sociology, 503-537. G. Gurvitch and W. E. Moore (eds.). New York: The Philosophical Library.

--- (1950). Introduction a l'oeuvre de Marcel Mauss. Sociologie et anthropologie, ix-lii. M. Mauss. Paris: Presses Universitaires de France.

Lowie, R. H. (1937). The History of Ethnological Theory. New York: Holt, Rinehart and Winston.

Lukes, S. (1973). Emile Durkheim. His Life and World' A Historical and Critical Study. London: Allen Lane.

Mauss, M. (1896). La religion et les origines du droit penal d'apres un livre recent. Revue de l'histoire des religions 34: 269-295; 35:31-60.

--- (1950). Sociologie et anthropologie. Paris: Presses Universitaires de France.

--- (1968). Oeuvres 1. Les fonctions sociales du sacre. Paris: Les editions de minuit.

Mauss, M. and H. Hubert (1899). Essai sur la nature et la fonction du sacrifice. Annee sociologique 2: 29-138.

--- (1909). Melanges d'Histoire des Religions. Paris: Alcan

Morris, B. (1987). Anthropological Studies of religion. An introductory text. Cambridge: Cambridge University Press.

Nisbet, R. A. (1975). The Sociology of Emile Durkheim. London: Heinemann.

Picketing, W. S. F. (1984). Durkheim's Sociology of Religion: Themes and Theories. London: Routledge & Kegan. Paul.

Schmidt, W. (1931). The Origin and Growth of Religion: Facts and Theories. H.J. Rose (trans.). London: Methuen.

Service, E. R. (1985). A Century of Controversy. Ethnological Issues from 1860 to 1960. Orlando: Academic Press, Inc.

Vogt, P. W. (1976). The uses of studying primitives: a note on the Durkheimians, 1890-1940. History and Theory. Studies in the Philosophy of History 15: 33-44.

Wallis, W. D. (1914). Durkheim's view of religion, Journal of Religious Psychology 7: 252267.

White, L. (1959). The Evolution of Culture. New York: McGraw-Hill.

~~~~~~~~

By Wouter W. Belier, Leiden, The Netherlands