Soon everybody was standing up and talking over the hedge. Tom and Alaric were like suspicious animals, eyeing each other doubtfully. Tom said that his thesis was nearly finished anyway, and that he didn’t think he would be able to use any more material. Alaric hurried to point out that his notes dealt almost entirely with religion and material culture and would therefore be of very little use to anyone writing a thesis on social and political structure.
1 It was not by chance that the expression “social anthropology,” which had been around for some fifty years, came into use in France. At the end of the 1940s, it was the expression “cultural anthropology” that seemed likely to enter general use, as the words “ethnology” and “ethnography,” which had been used since the nineteenth century for the study of peoples who did not use writing, faded into the background. Therefore, when presenting the work Sociologie et anthropologie [Sociology and Anthropology], which brings together the principal studies by Marcel Mauss, Georges Gurvitch wanted to republish texts “which converged on a subject increasingly known by the term cultural anthropology.” And Gurvitch added: “… the title Sociologie et anthropologie came about quite naturally, the term anthropology having been taken in the broad sense of cultural anthropology commonly used in America” (1950, VIII). Claude Lévi-Strauss (1950, XXIX), writing an “Introduction to the work of Marcel Mauss” in the same volume, refers to:
The eminent place of ethnography in the sciences of Man, which explains the role that it already plays in certain countries, under the name of social and cultural anthropology,[ ] as the inspiration for a new humanism, originating in the fact that it presents in experimental and concrete form that unlimited process of the objectification of the subject, which, for the individual, is difficult to achieve.
3 In 1950, anthropology was therefore still both social and cultural for Lévi-Strauss. Ten years later, however, it was a chair of social anthropology that was created at the Collège de France, and it was there that Lévi-Strauss set up a laboratory of social, not cultural, anthropology. What had happened during that decade? One may first suppose that the choice of social as opposed to cultural is neither insignificant nor arbitrary: the terminology indeed reflects a stance on the nature of the discipline of anthropology. However, we must make a detour to grasp the significance of this: it is not in the ordinary sense of the adjectives “cultural” or “social” in French that we will find the key to this shift. If we keep to the normal usage of “cultural” and “social,” it is difficult to understand why Lévi-Strauss, a social anthropologist, essentially devoted himself after 1960 to the structural analysis of myths and then works of art, cultural subjects par excellence. The choice of the expression social anthropology is in fact explained by the relation of Lévi-Strauss to the British and North American academic world. To be more specific, Lévi-Strauss intervened in the polemic which set North American partisans of cultural anthropology against British devotees of social anthropology for some years from 1951 onwards. At the end of the 1950s, he was therefore following the British, or rather the Franco-British, school of thought, which resulted in North American cultural anthropology being marginalized for a long time in France. After describing the details of the anthropological conflict between “culture” and “society,” we will see how Lévi-Strauss came to grips with this in the 1950s. 
Before the Confrontation
4 Until the beginning of the 1930s, the concept of culture was shared by both British and North American anthropology.  On either side of the Atlantic, Edward Tylor’s (1871) definition of culture as “that complex whole which includes knowledge, belief, art, morals, law, custom, and any other capabilities or habits acquired by man as a member of society” is at the origin of the majority of essays attempting to define the discipline (Tylor 1871, 1). In 1930, Franz Boas thus produced the article “Anthropology” in the Encyclopedia of Social Sciences, in which he wrote:
Culture embraces all the manifestations of social habits of a community, the reactions of the individual as affected by the habits of the group in which he lives, and the products of human activities as determined by these habits.
6 For Malinowski, who shortly afterwards wrote the article “Culture” in the same encyclopedia, culture is defined as:
This social heritage is the key concept of cultural anthropology. It is usually called culture. … Culture comprises inherited artifacts, goods, technical processes, ideas, habits and values.
8 These two definitions are extremely close as both Boas and Malinowski consider culture to be everything that is acquired in material form or in the form of customs. The functionalism of Malinowski and the historical particularism of Boas converge therefore on at least one point: they both think that culture is the central concept of anthropology. It is true that the scientific theory of culture proposed by Malinowski some ten years later (1944) emphasizes the satisfaction of needs, as culture was then viewed as a means of satisfying the biological needs of the human being. Culture then seems limited to a set of techniques given purpose by organic human life. The idea nevertheless remains in this theory (a theory which now seems considerably unsophisticated)—in keeping with the preceding definition—that culture is not strictly speaking a biological fact, even if it aims to fulfill food or sexual needs. The fact that culture is at the service of the body and its demands does not imply that approaches and customs are themselves products of nature. Malinowski’s theory is not therefore biological determinism: the cultural serves the biological without being reduced to it.
9 In other words, if Malinowski is opposed to North American anthropology, the difference of opinion lies not in the conception of culture, but rather in the way it is studied. This is quite different from Alfred Radcliffe-Brown, for whom anthropology is seen as a natural science of society. Culture is no longer the central aim of this study of Man; it is society which will henceforth occupy this place in anthropological theory:
We do not observe a “culture” since that word denotes not any concrete reality but an abstraction, and as it is commonly used a vague abstraction. But direct observation does reveal to us that these human beings are connected by a complex network of social relations. I used the term “social structure” to denote this network of actually existing relations. It is this that I regard as my business to study if I am working not as an ethnologist or psychologist, but as a social anthropologist. I do not mean that the study of social structure is the whole of social anthropology, but I do regard it as being in a very important sense the most fundamental part of the science.
11 From a viewpoint which owes much to Durkheim, Radcliffe-Brown sees social structure as a set of institutions whose function is to contribute to the maintenance of society through time. This functionality differs from that of Malinowski, which was developed at the same time: whereas the latter sees function as a response to individual and psycho-biological needs, Radcliffe-Brown considers that institutions play a conservation role in collective totality. As for the word “culture,” it is almost absent from the structural-functional lexicon; or else when Radcliffe-Brown mentions “culture,” it is only to claim its dependent nature in relation to the social structure: “Cultural tradition is a social process of interaction of persons within a social structure.” Culture, he adds, as applied to the:
… process of cultural tradition, is the process by which in a given social group or social class, language, beliefs, ideas, aesthetic tastes, knowledge, skills and usages of many kinds are handed on (“tradition” means “handing on”) from one person and from one generation to another.
13 Culture is therefore merely an effect of the social structure process, that is, of the groups or subgroups which comprise it. From this point of view, it is, for example, the kinship group which is the foundation of children’s education; it is the professional group which organizes the teaching of work techniques; it is the moral community which produces belief, and not the reverse. 
14 In the 1930s, the development of social rather than cultural anthropology in the United Kingdom did not seem to cause any notable reaction in the United States. While followers of Malinowski favored detailed position papers, often based on Oceania, advocates of structural functionalism aimed to find constants from case studies, with a predilection for Africa. The emergence of a “new” functionalism did not go unnoticed in the United States, but it was the conciliation route which prevailed. So, Kroeber, when he wrote a passionate defense of Boas, who had been attacked from a functionalist point of view by Agnès Hoernlé (1933) for his historicism, stated: “He was a functionalist, as his main interest lay in structural interrelations, change, processes, before Radcliffe-Brown or Malinowski had even written a line” (Kroeber 1935, 541). In other words, the English school of anthropology would not have proposed anything original in relation to Boas: for Kroeber, Hoernlé is attacking an imaginary Boas from lack of awareness or understanding of his work, but there are few real differences between British social anthropology and North American cultural anthropology.
15 Until the end of the 1940s, the school of thought known as “culturalist” and the structural-functional group therefore develop in a parallel fashion, with few exchanges or debates. Of course, Radcliffe-Brown and Malinowski taught in the United States during the 1930s, but North American anthropology remained dominated by the followers of Boas. This status quo was not to last beyond 1950.
The War of the Anthropologists
16 How can one be a social anthropologist, if one considers that anthropology is by definition cultural? In other words, is a social anthropologist still an anthropologist? This is the radical question which Murdock asked at the time of the publication of collective works on kinship directed by Radcliffe-Brown and Darryl Forde (1950).
17 George Murdock is famous in the history of anthropology for having undertaken, beginning in 1937, a vast compilation of cultural data entitled Human Relations Area Files (HRAF), first available in 1949. This is a database, organized in files, that deals with kinship, social organization, art, and techniques. But Murdock is also the author of an article (1932) defining the science of culture, in which, taking up the conclusions of Kroeber (1917, 177–78) and Stern (1929, 270), he rejects the idea that the human being is distinguished from other animals exclusively by virtue of his social nature. Many higher animals live in groups, as do certain insects: collective life is not therefore a specifically human trait. For Murdock (1932, 213), Man is an intelligent animal who may acquire customs, living in groups and possessing language. Many species have several of these features, but only the human species has all four. Using a technical metaphor, Murdock compares culture to the stability of a stool: if it only stands on two or three legs, the object is unstable. For example, certain monkeys are capable of inventing things, but they are incapable of passing on this knowledge; they immediately forget, because monkeys cannot form customs. Only Man is therefore capable of raising himself to the cultural level, to the “super-organic” (Kroeber), without falling. Only Man can innovate and communicate his findings to his peers, who can adopt it in perpetuity.
18 One can therefore understand that reducing anthropology to the study of social organization (as was done by Radcliffe-Brown and his students) was not acceptable to Murdock. After several formal compliments (along the lines of: this collective volume will be very useful to specialists in Africa; the professional skills of the authors are of a very high order; and so on), the argument develops around seven polemical points which can be summarized as follows:
Despite these merits, one nevertheless finds a certain number of limitations that numerous professionals outside the United Kingdom find difficult to understand and impossible to defend.
20 Firstly, British social anthropologists are not interested in the whole range of cultural phenomena, but concentrate exclusively on kinship and associated subjects, in particular, marriage, property, and government. Technology, folklore, art, the education of children, and even language are almost completely ignored.
21 Secondly, this research is geographically limited; it is all carried out in British colonial areas. It is therefore research undertaken in Anglophone Africa.
22 Thirdly, only a small number of societies have been studied by “social” anthropologists. Whereas, according to Murdock, two or three thousand “primitive societies” (sic) have been identified throughout the history of anthropology, the British Radcliffe-Brown school of thought limited themselves to some thirty cases. Ethnography produced outside the United Kingdom is neglected, in particular, works published in French and German, which shows a lack of interest in making comparisons.
23 Fourthly, as well as ignoring ethnography produced outside their country, British social anthropologists are indifferent to external developments in theory; only British authors are cited and discussed.
24 Fifthly, there is hardly any interest at all in history. Social structure is only studied at a point in time, in order to see the functional relations. The question of the genesis of institutions is not tackled, or is tackled very rarely.
25 Sixthly, processes of cultural change are neglected. Invention, cultural integration, reinterpretation, selective elimination, integrated modifications, and drift are not included in the field of investigation.
26 Finally, English social anthropology ignores psychology. Murdock remarks that this is all the more surprising given that the approach of Malinowski, whose interest in the psychology of the individual was notable, lies at the origin of numerous studies of the relations between culture and personality in the United States.
27 Murdock pursues what resembles a prosecution case against Radcliffe-Brown. He expresses gratitude to him for having created order in kinship studies, which Morgan and Rivers had left in a sort of tangle. But with the accent on the synchronic structure of a single society, the paths opened up by Tylor (comparison), Malinowski (anthropology open to psychology), and Boas (historical method) are closed. The English scholar is responsible for having locked anthropology, initially into research on universal laws formulated from the study of a small number of societies which are not representative of the whole of humanity, and then into an expression of these laws which does not specify the concomitant behavior of variables. Moreover, Radcliffe-Brown’s theory does not change; it has been repeated in identical form over many years. According to Murdock (1951, 470), the reader will find the same old errors in the introduction to the volume on kinship, “and a few new errors as well.” Radcliffe-Brown’s theory is therefore like an immutable dogma (or one which only deteriorates) around which has formed a “school.”
28 How can we explain the peculiarities of this structural-functionalist school? Murdock arrives at the astonishing conclusion that its supporters are not anthropologists. Indeed, if one considers anthropologists to be scientists who take culture and its ramifications as the subject of their study, then it appears that British social anthropologists do not use this concept and only study a small part of what is collectively acquired. They do not study change, education, or diffusion, phenomena which lead other anthropologists to take an interest in history, geography, and psychology. To sum up, British social anthropologists have nothing in common with cultural anthropologists other than the study of kinship and of societies without writing. And Murdock deals the final blow with:
In their fundamental aims and theoretical orientation, they are more in line with sociologists. Like other sociologists, they are primarily interested in social groups and the structuring of interpersonal relations rather than in culture, in synchronic rather than diachronic correlations. … Our interpretation is, of course, based on the fact that this British school derives from the sociologist Durkheim, via Malinowski and Radcliffe-Brown.
30 With a certain irony, Murdock links the work of Radcliffe-Brown, not to that of the eminent American sociologists of the 1950s, such as Merton and Parsons, but to the “best sociologists of the 1920s, for example Sumner, Pareto and Thomas.” In other words, the British anthropologists of 1951 are sociologists twenty years behind the times…
31 It is therefore a clear and definite rupture in the discipline of anthropology on either side of the Atlantic which is proposed in this article. One may be surprised that the discriminating criterion used by Murdock is not the participant observation method, which now seems to us to be at the heart of the discipline of anthropology. In truth, Murdock hardly ever practices it, and he is even one of the rare devotees of the quantitative method in anthropology (applied to the HRAF), hence his frequent use of statistical notions (“variable,” “correlation”). Researchers in the field are not anthropologists if they are only interested in social solidarity; anthropologists may be statisticians if they go on to make intercultural comparisons. The definition of the discipline therefore centers on notions of culture and society.
32 Faced with the serious risk of a rift between social and cultural anthropologists, the editors of the journal asked Raymond Firth to respond to Murdock in the same volume of the magazine American Anthropologist. Firth succeeded Malinowski at the London School of Economics in 1944. He was therefore in no sense a student of Radcliffe-Brown, who was a professor at Oxford from 1937 to 1946. Besides, Firth is a specialist not in Africa, but in the society of Tikopia, in the Solomon Islands in the Pacific.
33 He therefore notes that Murdock makes exaggerated claims for the unity of British social anthropology, which according to him does not constitute a “school:” there are great differences between Malinowski and Radcliffe-Brown.  Firth nonetheless agrees that British Africanism is not sufficiently comparative in the sense that it uses almost no data originating from other cultural areas, in particular the Americas.
34 As far as the final reproach to British anthropologists is concerned, where they are accused of being sociologists, there is nothing new in this: Malinowski, Firth himself, Radcliffe-Brown, Gluckman, Evans-Pritchard, and Nadel demanded that there should be a necessary rapprochement between anthropologists and sociologists at the theoretical level. What distinguishes sociology from anthropology is not therefore conceptual construction, but the ethnographic method:
The more general theory of the anthropologists, then, is hardly distinguishable in its scope from that of the professed theoretical sociologist, though its different ethnographic base gives it a different illustrative content and a different—sometimes sharper—focus. In field techniques, their work is distinguished by a high degree of intensive first-hand observation of social behavior.
36 Firth’s response is therefore that anthropology is a social science, closely related to sociology, psychology, economics, politics, law, and history when it is problem-oriented. On the other hand, it is only distantly related to biological anthropology, technology, or archaeology. The ethnographic viewpoint gives the anthropologist in the field a more precise picture of social relations, and, if one reads Firth correctly, this is perhaps the only real particularity of social anthropology, if one compares it to the “wide perspectives” characteristic of the sociological point of view.
37 How does Firth respond to the reproach made to the British anthropologists that they neglected culture in favor of society? The two concepts indeed represent for Firth two facets of human existence:
“Society” denotes the human component, people and the relations between them; “culture” refers to the accumulated resources, material and immaterial, which people have acquired, transmitted and modified by social learning. But the study of both must include the study of social relations and values, by examining human behavior.
39 Firth claims that British anthropologists do not therefore feel divided from their American colleagues even if the latter define their discipline in cultural terms. If some British anthropologists refuse to use the notion of culture, this is to distance themselves from the definition given by Malinowski, not from that which is current on the other side of the Atlantic. But, reading the definition above, one can easily understand that anthropology is social before it is cultural, since life in society is a condition of learning, and of the transmission of material and immaterial resources. Even if Firth considers culture to be a facet of human existence associated with learning, anthropology remains fundamentally social. For Murdock, however, human sociality only constitutes one of the carriers which convey culture, along with intelligence, custom, and language. In other words, whereas for Firth culture is an aspect of social life, for Murdock it is an order of facts which are superior (super-organic) to biological and social life.
40 From this perspective, Murdock’s line of argument is close to that of Ruth Benedict (1931; 1934) and Ralph Linton  (1936; 1959 ), for whom culture fashions social life. There are patterns of culture which give a certain shape to relations between individuals in a community. Culture cannot therefore simply be reduced to society; culture is a factor in the organization of social relations, their permanence, and their transformation. From this point of view, anthropology is therefore cultural before being social. It is the science of custom, that is, of all that is learned, as opposed to biological heredity.
41 Less than a year after the publication of the articles by Murdock and Firth, Radcliffe-Brown (1952) responded to Murdock in the same magazine. This text, “Historical Note on British Social Anthropology,” a title which is itself a response to the reproach that the English anthropologist neglected history, begins with a reference to Boas (1940, 633–34), who said that anthropology had two aims: to reconstruct the history of peoples, societies, or particular regions, and to discern the general laws of cultural development through comparison. Radcliffe-Brown considers that social anthropology in the United Kingdom has concentrated on this second objective, the first being reserved for “ethnology” as a historical or museological science. As for ethnography, it is the study of peoples who do not have writing. In 1906, Frazer was the first professor of social anthropology, which was defined as the “sociology of primitive peoples.” Conversely, Westermarck is a professor of sociology, but according to Radcliffe-Brown, his work falls within the field of social anthropology.
42 With the passage of time, ethnography saw itself shift from a dependence on ethnology to a dependence on social anthropology. Although the latter favored the comparative method on a large scale, with no direct link to fieldwork, Malinowski and Radcliffe-Brown established an organic link between theory and empirical enquiry. Ethnography was therefore subject from then on, not to the collection of data aimed at the historical reconstruction of the society studied, but to the principles of comparative study. Ethnography was to concentrate, for example, on kinship or political organization in order to establish a body of data to juxtapose with a set of elements of the same type from the ethnography of another society. Purely descriptive at the outset, ethnography is influenced by concepts of social anthropology.
43 So are British social anthropologists really sociologists with whom other anthropologists no longer have anything in common because the former have no interest in technology, folklore, art, the education of children, or language? The response of Radcliffe-Brown is almost affirmative:
There is a danger of misunderstanding here. I do not suppose that Murdock intends to say that British anthropologists in general do not deal with such subjects as technology, art and folklore, and language. It is only that these studies are not included in that branch of anthropology that is called social anthropology. The study of languages is carried out at the School of Oriental and African Languages, as well as elsewhere. Folklore is dealt with by the Folklore Society, which includes anthropologists in its membership. Technology and art are studied in connection with ethnological museums. In other words, the view that has been taken in England for fifty years is that social anthropology is only one branch of anthropological studies.
45 In other words, Radcliffe-Brown admits that social anthropologists are not interested in the cultural practices mentioned above, but that they concentrate on the study of social structure. And when one reads the arguments of Radcliffe-Brown, one has great difficulty seeing what justifies their attachment to “anthropological studies,” which appear very disparate. Should social anthropologists therefore abandon the term “anthropology?” Radcliffe-Brown ultimately gives a purely historical and institutional justification for keeping it. There are a great number of social anthropology departments in the United Kingdom, and it has become impossible to change this name, even if Radcliffe-Brown admits that it is awkward. Linking structural functionalism to anthropology is in the end based on historical chance, not on the reasoned unity of a discipline.
46 The first stage of this polemic between North American and British anthropologists therefore ends with the recognition, in theory, of a divergence between social and cultural anthropology. Radcliffe-Brown does not contest Murdock’s statement, he simply points out that people other than him are doing cultural anthropology in England. It therefore seems that the only unity of the discipline rests on the use of a word imposed by the history of academic institutions. While Murdock and Radcliffe-Brown are in complete agreement on their divergence, Firth on the other hand proposes a compromise which would include culture and society in the field of anthropology. However, this solution is asymmetrical: culture is one aspect of social life; it is not the principal subject of anthropological science.
47 After the publication of these three articles, the polemic spread on both sides of the Atlantic. If one follows Murdock and Radcliffe-Brown, there is indeed a great risk that anthropology will split into two disciplines: a sociology of peoples without writing on the one hand, and a science of culture favoring the study of art, folklore, religion, and language, on the other.  This split did not, however, take place in the 1950s for the reasons described below.
Anthropologists at Peace?
48 In June 1952, a symposium proposing an overview of anthropology (Anthropology Today) was held in New York, under the auspices of the Wenner-Gren Foundation and presided over by A. L. Kroeber. This conference had been decided upon at the end of 1951, and was organized in sessions based on the discussion texts distributed to the participants several weeks before the meeting. These documents touched on every aspect of the discipline of anthropology (theories, classic methods, new technologies, applications of anthropology, and so on) and its relations with the other sciences, both natural and social (linguistics, biology, archaeology, and physical anthropology).
49 It is not these articles which concern us here,  but rather the debates organized at the time of this symposium, which were transcribed and published by certain participants (Tax et al. [eds.] 1953). The only session organized during the symposium itself had the theme of “Cultural Anthropology/Social Anthropology.” It was therefore not based on any text distributed in advance, but was obviously consequent to the articles by Murdock, Firth, and Radcliffe-Brown published between 1951 and 1952. With forty-four participants, this is the session which involved the greatest number of speakers in this conference. The Americans were in the majority (there were thirty), but among the Europeans were: S. F. Nadel, Austrian by origin, who was a student of Malinowski and a professor in Australia; Darryl Forde, who was a student of Kroeber and Lowie, then a close collaborator of Radcliffe-Brown, with whom he coordinated the collective volume on kinship which originally gave rise to this Anglo-American polemic; and Lévi-Strauss, director of studies at the École pratique des hautes études. 
50 After discussing how to organize the subdivisions of the discipline, the discussion, led by Sol Tax, arrived at the central issue: the pertinence of the distinction between social anthropology and cultural anthropology, in the sense given to these expressions in the United States and the United Kingdom.
51 Robert Lowie, a specialist on the Indians of the North American Plains (in particular the Crow) and one of the oldest participants in the debate,  declared at the outset that “as culture is the whole of social heritage, culture and society are associated (correlated) concepts,” and that ideally, social and cultural anthropology should be united (should be one) (Tax et al. [eds.] 1953, 223). Without in any way excluding the possibility of generalizing or the discovery of “laws” (it is Lowie who puts this word in quotation marks, showing his doubt all the same), he refuses to limit his research to this aim, and defines himself as an ethnographer and cultural historian above all. When and where was the reindeer first domesticated? That question is of much greater interest to Lowie, and of no interest to British anthropology. 
52 Benjamin Paul, a reader at Harvard, proposed clarifying the terms of the debate. The adjective “cultural” has a metonymical meaning, like the word day in English which denotes both a cycle of twenty-four hours and the part of the day during which the sun shines. Anthropology is therefore cultural when it is set in opposition to physical anthropology; and within cultural anthropology, part of the discipline is “social,” the other part “cultural.” This latter orientation favors history, space, and time. In the United States, declaring oneself a cultural anthropologist in reality consists in integrating the two areas of the discipline, culture and society. The distinction between the two sections does not therefore seem very relevant to the Americans. When the British say that they are social anthropologists, they seem, on the contrary, to exclude culture from their study and limit their research to a narrower field.
53 Murdock agreed with this statement and claimed the designation cultural anthropology for the whole of the discipline. Social anthropology is limited to the study of interpersonal relations. In the United Kingdom, distinguishing the social from the cultural has had the detrimental effect of leaving the study of the latter lying fallow . If culture and society continue to be studied equally, it will be possible to consider social structure as a progressive system through time, and to compare these systems in terms of the way they change and adapt. Hence, Murdock implicitly reproaches Radcliffe-Brown for defining the discipline in a manner which does not lead to any understanding of social transformation.
54 Lévi-Strauss interceded by proposing a fundamental distinction: Man may be considered as an animal that makes tools or as a social animal. Cultural anthropology starts from the study of material techniques and then moves to the study of social relations. Social anthropology works in the opposite direction: it moves from social relations, to tools and culture in the wider sense. It is only a difference of viewpoint, and there is no great difference between social anthropology and cultural anthropology.
55 For the English side, Nadel and Forde attempted to minimize the difference of opinion. The former pointed out that in England, Malinowski held a chair of social anthropology, and that he continuously talked about culture, whereas a cultural anthropologist in the United States may quite well talk about social structure all the time. Nadel concluded that there is no difference between these two designations. As for Forde, he stressed the arbitrary history of these words: Malinowski’s chair took the name of “Social Anthropology” simply to distinguish it from that held by Seligman, who was professor of “ethnography” and “ethnology” at the London School of Economics. For both Forde and Nadel, social anthropology and cultural anthropology must make up a single discipline, even if they agree with Lévi-Strauss when he makes a distinction between the two points of view. Tax ended the debate by agreeing with Nadel:
I think the consensus here, with some exceptions, is that we ought to use the words “cultural” and “social” anthropology interchangeably and forget about the question of terminology and deal with the problems involved.
57 It seems, therefore, that we have reached a position of compromise based on minimizing the significance of designations, which are considered more or less arbitrary. However, we must immediately point out that Sol Tax, who moderated this debate, is a North American anthropologist, but that he is in no way representative of “Boasian” anthropology in the United States. Indeed, Sol Tax experienced the influence of Radcliffe-Brown in Chicago where the latter taught in the 1930s (Eriksen and Nielsen 2001, 58); he is therefore no doubt inclined to find common ground between culturalists and functionalists.
58 However, one should not be surprised that Kroeber (1953, 365–66), one of the fathers of culturalism, was less accommodating when he announced the conclusions of the symposium. Unlike Paul, he does not think that the difference between cultural and social anthropology is based on the opposition between research into laws and the description of particular facts situated in time and space. It is a problem of primacy or inclusion. Radcliffe-Brown and the sociologist Talcott Parsons both think that culture is secondary to society, that the first in some way derives from the second: culture seems to be an extension of the social base. Kroeber, on the other hand, in line with Murdock and the majority of American anthropologists, thinks that culture embraces society. One cannot define society narrowly as a “social structure,” and at the same time broadly in a way that includes language and symbolism. How could the vast cultural proliferation be the product of a social structure? For Kroeber, society has no pre-eminence over culture, and it is culture which includes society and not the reverse.
59 The debate continued in similar terms in the columns of the magazine American Anthropologist in 1953. Meyer Fortes, whose text was written before its author was acquainted with the articles by Murdock and Firth, defended the results of British social anthropology in the field of kinship. But in a more general way, Fortes (1953, 21–22), in that article, also developed the idea that if one views culture as a concept subsuming that of social structure, as did Malinowski and Firth, one is led to give equal weight to everything that is produced in a given society. The anthropologist cannot therefore locate the most important institutions—kinship, political institutions, and the legal system—from a functional point of view. Of course, ethnographic observations may be seen as facts of custom, “standardized ways of acting, knowing, thinking, feeling, which are compulsory and universally exploited within a group of people at a given moment.” Here Fortes skillfully combines the definition of the social fact by Durkheim and that of culture by Tylor, but he proposes above all analyzing ethnographic data other than as cultural elements. According to Fortes, we must now concentrate on the social structure, and see customs as the symbolization and expression of social relations. The social structure is not, from this point of view, “part of culture, but the entire culture of a people, handled within the framework of a specific theory.”
60 To read Fortes, one might think that he is defending a perspectivist position similar to that of Lévi-Strauss. However, Fortes moves on to draw up a theory of the relations between culture and society which leaves no doubt as to the nature of the primordial authority:
I would suggest that a culture is a unity in so far as it is tied to a bounded social structure. In this sense I would agree that the social structure is the foundation of the whole social life of any continuing society. … The social structure of a group does not exist without the customary norms and activities which work through it. We might safely conclude that where structure persists there must be some persistence of corresponding custom and where custom survives there must be some structural basis for this.
62 Fortes nevertheless further qualifies this by considering that there are major factors of autonomy in custom: “A house is not just its foundations and custom cannot simply be reduced to a manifestation of social structure.” So the case of migrants like the Chinese and Indians, or that of Black Americans (Fortes cites Herskovits), shows that it is possible for religious and aesthetic customs to be retained in the face of radical changes in social structure. Nevertheless, the metaphor of the “foundation” is significant: social structure supports symbolism, language, and religion, which would not hold up without it.
63 The response to Fortes came swiftly. In the same year (1953), Lowie published an article in the same magazine, in which he gave his opinion on the relations between ethnography, cultural anthropology, and social anthropology. Lowie recognized that the British anthropologists were really ethnographers—the ethnographic method is what anthropologists on both sides of the Atlantic had in common (Lowie 1953, 527–28). They were researchers who, in their best work, dealt with an “important subdivision of culture” (social relations). Lowie even admits that it can be useful to take the social structure as the starting point for the study of a people’s culture. But he rejects out of hand the declaration by Fortes that the social structure should not be “part of culture, but the entire culture of a people, handled within a specific framework.” On the contrary, writes Lowie, “the social structure of a people is one aspect of their culture, in the sense of Tylor,” and it is not because one can link certain craft activities with certain social groups, or certain religious beliefs with certain social sectors that one can dispense with studying the techniques themselves and the content of the beliefs (Lowie 1953, 531–32). The social structure, conceived of as the determining force of culture, in fact leaves a very large cultural residue unexplained. In other words, the social structure only explains a very small part of the cultural whole.
64 Let us end this section with Fred Eggan, who gave the presidential address to the American Anthropological Association in 1953. Like Sol Tax, Eggan was a Chicago-based anthropologist  who had followed the teachings of Radcliffe-Brown. Eggan noted the existence of a “schism” between those who claimed to follow ethnology, mainly Americans, and the new group of followers of social anthropology, who were mainly British. The former are interested in culture, its history, and its transformations (process), the latter are interested in social structure and its functions. This is not just a difference in terminology, and cannot be brushed aside as Nadel suggested: it has a foundation in reality. Eggan was trained in the Boasian school of anthropology, particularly by Edward Sapir, before partly adopting functionalism under the aegis of Radcliffe-Brown himself. Having therefore had a “foot in both camps for twenty years” (1954, 743), Eggan offers to mark out common ground which will satisfy both parties.
65 Of course, one may consider that the British have excellent field observation techniques in certain limited areas, even if their theoretical approach is barren and lifeless compared to the broad focus of studies in American anthropology. Eggan’s proposition, on the other hand, consists of adopting the structural-functionalist approach by integrating it into the American tradition of interest in cultural processes and history. The weaknesses of the British are the strengths of American ethnology (that is, cultural anthropology): Eggan therefore proposes a synthesis of the two approaches.
66 It is necessary to distinguish society from culture, as does Radcliffe-Brown. Social structures (in particular, kinship, political organization, and the legal system) tend to present a limited number of shapes, which enables classification and comparison. On the other hand, cultural data correspond to patterns which one can discern in time and space, even if they do not have the stability of social structures: here one encounters the problem of cultural models formulated by Ruth Benedict (1934). Eggan therefore preserves the theoretical framework of cultural areas, which is typical in American anthropology (Kroeber 1939). Social forms and cultural models may vary independently, while having as the locus the behavior of individuals in social groups.
67 It is in the comparative method that we must be able to find common ground according to Eggan. This method, discredited by Boas for its speculative nature, was sidelined in the United States in favor of the specific history of each culture. Conversely, functionalism proposes finding universal laws which govern every human society. The first approach is too specific, the second too general. Eggan (citing Merton 1949) therefore proposes developing an intermediary comparative method between particularity and generalization; the concept of cultural area which is current in the United States would be an excellent tool, if it does not lapse into diffusionism, and this could be used outside the continent of America with the aim of comparison, in both social structures and cultural models.
68 Eggan’s text—which ends this cycle of polemical publications—in fact constitutes a brilliant asymmetrical synthesis. American cultural anthropology is recognized in this as superior in its ability to propose the total ethnography of a culture (without being limited to the social structure), but it is inferior to social anthropology in the theoretical field. The structural functionalist approach gives a new dimension to American ethnography, which was too “flat” according to Eggan (1954, 746). The common ground therefore “leans” toward the British side: the society/culture distinction, rejected by Murdock, Kroeber, and Lowie, is well validated, even if Eggan integrates the concepts of cultural area, historical process, and social change that the British have sidelined. One may therefore consider that the British managed to enter the world of North American academia, even if American hostility towards social anthropology only subsided definitively in the United States at the beginning of the1980s (Watson 1984, 351–52).
69 We now need to examine the repercussions of this debate in France.
How French Anthropology Became “Social”
70 Lévi-Strauss took an active part in the 1952 symposium in New York, as we saw earlier. He was at that time a devotee of a perspectivist solution in the debate between cultural and social anthropology: the human being is Homo faber (likely to be studied by cultural anthropology) and a social animal (which makes social anthropology possible). In 1954, he returns to this question in a text published by UNESCO (United Nations Educational, Scientific and Cultural Organization) (reproduced with slight modifications in Lévi-Strauss 1958, 414–18).
71 After presenting the terms of the debate as proposed by Murdock, Firth, and the participants in the symposium, Lévi-Strauss claimed that there is “no contradiction, not even any conflict between the two perspectives” (1958, 415). Durkheim, who demands that social facts should be considered as things, supported the cultural anthropology point of view (as this stems from material culture), whereas Mauss, who holds things to be social facts, adopted the typical social anthropology point of view.
72 One may be surprised that Lévi-Strauss considered Durkheim close to cultural anthropology. Firstly, the latter never undertook any research into technology, and the expression “as things” indicates that social facts are imposed on individuals as external facts independent of their will, yet without being of a material nature. Secondly, Lévi-Strauss reduces cultural anthropology to technological studies; the Boasian school of thought, however, in no way limits the concept of culture to the manufacture and use of tools. As for the convergence of Mauss and social anthropology, in its British version this is equally doubtful. The variety of themes tackled by Mauss (sacrifice, magic, gift exchange, the person, the techniques of the body, and so on) gives his work a range which recalls the American conception of culture. We may add that the eclectic nature of his work differs fundamentally from the striving for systematic scientificity displayed by Radcliffe-Brown.
73 It therefore seems to us that the reverse is true: Durkheim is really at the origin of British social anthropology (this origin is supported by Radcliffe-Brown), and Mauss, on the contrary, has a certain affinity with the American school of anthropology. One just needs to read The Gift to understand what Mauss owes to Boas, copiously cited in this text for his work on the potlatch (1950 , 152 sq.; 194 sq.). By wrong-footing the theoretical genealogy claimed by Radcliffe-Brown, the reversal operated by Lévi-Strauss has the effect of blurring the difference between social anthropology and cultural anthropology even more. According to Lévi-Strauss, the first approaches the “whole man” from his representations (which puts him in a close relationship with psychology and sociology), the second from his productions, in particular his tools (which makes cultural anthropology a close relation of geography, technology, and prehistory). But the main thing is to recognize that for the author of The Elementary Structures of Kinship (1949), both culture and society are governed by structures, whose paradigm is linguistic structure. As language is both a cultural fact (as animals do not talk, this is one of the markers of the nature/culture opposition) and a social fact (as speech establishes the link between individuals), we may legitimately, in order to understand culture and society, take inspiration from linguistics conceived as a disciplinary bridge between social anthropology and cultural anthropology.
74 How can we understand, then, the final choice of Lévi-Strauss for the expression of social anthropology? We should perhaps go back to the text of “Social Structure” (translated and adapted in French under the title “La Notion de structure en ethnologie” 1958 , 229–378), which was presented and discussed at the Wenner-Gren symposium, in order to understand its ramifications. This famous text is a sort of manifesto of structuralism where Lévi-Strauss calls “culture,” “any ethnographic whole which, from the investigative point of view, presents significant differences, in comparison to others” (1958 , 352). This formulation no doubt comes from Saussurean linguistics, and Lévi-Strauss had already noted this theoretical influence in 1945 (revisited in 1958 , 43–69). This is far from the classic definitions of culture in the United States, which raises a vehement protest from Margaret Mead: “I do not believe that we will get anywhere if we try to find analogies for morphemes and phonemes in the rest of culture” (Tax et al. [eds.] 1953, 296). Mead thinks that language is inseparable from culture, and that these are therefore categories of cultural anthropology which enable one to think language: linguistics cannot be the epistemological model of anthropology. Let us add that in all these debates, Mead and Lévi-Strauss had several lively exchanges: for example, regarding the existence of national cultures, Mead regrets the lack of cooperation from the French in her project on this subject (Tax et al. [eds.] 1953, 138–39); or on the subject of the use by Lévi-Strauss of the word “garbage” to designate non-Western cultures, which the American anthropologist finds completely unacceptable (Tax et al. [eds.] 1953, 351). In the United States, the propositions of Lévi-Strauss therefore ran up against the resolute opposition of one of the principal figures of culturalism. 
75 On the other hand, Lévi-Strauss conceived of his theoretical propositions as a continuance/extension of the structural functionalism of Radcliffe-Brown, and his success was therefore more notable in the United Kingdom than in the United States.  The first major work by Lévi-Strauss is a study of kinship, a subject favored by British social anthropology, as we have seen.  Here his work is presented in the field of “comparative sociology,” and he himself is presented as a “comparatist sociologist” (1949, XI), just as Radcliffe-Brown before him. He also shares with Radcliffe-Brown the idea that anthropology must become a science like the others, by distancing itself from the humanist impressionism which certain American anthropologists hold to and promote, such as Mead or Lowie, for example (Tax et al. [eds.] 1953, 152). The Lévi-Strauss models are found in structural linguistics, but also in the theory of games and cybernetics, and not at all in philology or the humanities in general. His conception of anthropology as a science of the general properties of social life (1958 , 404) distances him from the historical particularism which left a lot of room for the study of local circumstances and specifics.
76 At the end of the 1950s, the expression “social anthropology” therefore became established, as can be seen in the report by Maurice Merleau-Ponty on the creation of a chair of social anthropology (2008 a ), and in his presentation of the candidature of Lévi-Strauss for this same chair (2008 b ). In the first text, the philosopher writes:
What we today call social anthropology—an expression in common use outside of France, and now becoming more current in France—this is what sociology becomes when it accepts that, like Man himself, the social has two poles or two faces: it is significant, one can understand it from the inside, and at the same time personal invention is generalized and diluted in it, it tends towards the process, it is, according to that famous word, mediatized by things.
78 Starting from Mauss, who had anticipated this “looser sociology,” Merleau-Ponty comes to propose the creation of a chair entirely devoted to the study of social structures. In this text can be found a résumé of the main theses of Lévi-Strauss, although he is never named in it: there is no doubt that this chair, if it were to be created, would be destined for the author of Structural Anthropology, a work which had just been published (1958). Merleau-Ponty does nothing to disguise his intention, as he ends this text writing explicitly “that this sketch of social anthropology” is also “someone’s abstract description.” Once the chair was created, Merleau-Ponty presented the candidate, beginning the introduction with:
Social anthropology, between sociology and ethnography, has recently gained its autonomy. The works of M. Claude Lévi-Strauss are almost alone in France in following this precise direction.
80 At no time in either of these two texts does Merleau-Ponty mention the existence of “cultural” rather than “social” anthropology. In the first document, the adjective “social” is linked to the tradition of Durkheim, in which Merleau-Ponty explicitly locates the creation of the chair. Mauss is presented as the “father” of this discipline, even though he held a chair in “sociology.” As for the reference to general usage outside of France of the expression “social anthropology,” it is incorrect, inasmuch as it is only in the United Kingdom, as we saw earlier, that this wording dominates. In the second text, Merleau-Ponty identifies Lévi-Strauss as the ideal candidate for this chair, and with reason, as it was in fact created as a chair of structural anthropology. 
81 Once elected, Lévi-Strauss gave his inaugural speech; in it he briefly specified his conception of the link between social anthropology and cultural anthropology (1973 , 19). If we consider that anthropology is the study of the signs at the heart of social life, how then should we consider “the tools, the techniques, the means of production and consumption” (1973 , 19)? Are these things signs? No, but they are “impregnated with meaning.” To designate this part of social anthropology, close to geography and technology, Lévi-Strauss proposes the expression “cultural anthropology.” This therefore becomes a subdiscipline, as if culture was reduced to material culture.
82 Beginning from a perspectivist point of view in 1952, with no supremacy of the “social” over the “cultural,” by 1960 Lévi-Strauss has come to favor an inclusivist approach, in which culture is only one aspect of society. To the extent which social anthropology studies every sign in society, its field of action includes the meaning attached to material objects. This is not, however, a return to Durkheim or Radcliffe-Brown. These scholars favored the study of forms of collective solidarity; Lévi-Strauss defines social anthropology as a séméiologie.
83 To understand how social (and not cultural) anthropology became established in France, we have had to take a long detour via the United States and the United Kingdom. The determining player in this genealogy was Lévi-Strauss, of course, but it is only by taking into account his relations with the British and North American scientific areas that one can grasp the significance of the choice of “social” over “cultural.” Lévi-Strauss is seen as a follower of Durkheim and Mauss in the 1950s, as he edited the volume Sociology and Anthropology by Mauss (1950), which certainly went in his favor when he acceded to the Collège de France in 1959. The reference to “social” allows him to manifest this lineage from the French school of sociology, especially as he defined himself at the time as a “comparatist sociologist.” But the choice of “social anthropology” is also explained by the implication of Lévi-Strauss in the Anglo-American debate initiated by Murdock and Firth. By choosing the social over the cultural, Lévi-Strauss distances himself from North American anthropology, although he was trained in the American system. The very broad definition of culture as custom, habit, or tradition, associated with a comprehensive and aestheticized approach that is typical of the anthropology of Boas’ students (such as Lowie, Benedict, or Mead), is the very antithesis of the “scientific” path he intended to follow. Nothing is indeed more alien to structuralism than the psychological anthropology of the school called “culture and personality,” and we have seen that Mead had the greatest reservations, from 1952, with respect to the propositions of Lévi-Strauss: the antipathy was mutual. On the other hand, Radcliffe-Brown’s propositions are compatible in more than one way with the anthropology of Lévi-Strauss: it is a matter of discerning universal (or invariant) laws in the social organization (or the human spirit). Anthropological science must disengage from any impressionism, from any subjectivity, if it is to propose valid results. This is an aim common to both British structural functionalism and French structuralism. This is how social anthropology came to triumph in France, after it had become established in the United Kingdom.
84 Let us however avoid ethnocentricity. If the expression “social anthropology” is perhaps dominant in the Francophone world, this is not the case in the Anglophone area. The conception of anthropology as the study of culture, which thereby includes society without being reduced to just that, remains widespread in the United States. By way of example, neither Clifford Geertz (1973) nor Marshall Sahlins (1999) have ever abandoned it. It enables today’s researchers to be attentive to human diversity rather than to invariants, to favor the study of the (unforeseeable) event over that of structures, and to avoid dogmatism by respecting the meaning that subjects give to their actions. That in no way excludes the collective dimension from the field of cultural anthropology. Significance, when understood at the level of words and actions, and not hypothetical “structures” or “ontologies,”  always develops in interpersonal relations, whether they are peaceful or contentious. If we consider that anthropology aims above all else to understand cultural difference or to interpret cultures (Geertz 1973), that is, the variety of meanings and symbols in humanity, then the future of the discipline perhaps belongs to cultural anthropology.
I would like to thank Marshall Sahlins for his kind remarks on this text; the theories presented in it and any errors are entirely my responsibility.
The quotations from Boas, Malinowski, and Radcliffe-Brown on the concept of culture and successive definitions of it from the nineteenth century are taken from the work by Kroeber and Kluckhohn (ca. 1960 ).
We should note that E. Evans-Pritchard (even if in 1950 he criticized research into “laws” in anthropology and thus opposed Radcliffe-Brown) is in agreement with the latter in defining his discipline as social anthropology, “a branch of sociology which mainly studies primitive societies” (Evans-Pritchard 1951, 11).
Firth (1951, 480) remarks that the work of Malinowski is tainted by romanticism, whereas that of Radcliffe-Brown is imbued with the classical aesthetic. The former has a taste for irregularities and individual variation, the latter for systems ordered by laws.
In 1936, Ralph Linton devoted a chapter of his work The Study of Man to the notion of function, which allowed him to place himself in relation to the emerging British structural functionalism, a scientific movement which he explicitly places in the tradition of the French school of thought, that is, the school of Durkheim. While Linton intended using the expression “social structure” in the limited sense of “forms governing individual interrelations,” the functionalists, according to him, incorrectly extended the meaning of this concept to a broad and undefined sector of “culture” that included economic and religious life (Linton 1936, 401). For Linton, culture is “social heredity,” not “social structure” (ibid., 80). Culture is the condition of possibility of the “social structure,” and not the reverse. He therefore claims (in 1945) that “the structure of a society, that is its system of organization, is itself a matter of culture … Without culture, there could be neither social systems of a human type nor the possibility of adapting it to new members.” (Linton 1959 , 24–25).
The short text from Barbara Pym placed as an epigraph to this article demonstrates this opposition in a most striking way. Barbara Pym, a famous English novelist from the beginning of the 1950s, describes in her fourth novel (Less than Angels 1955) the romantic relations within a little group of anthropologists revisiting or setting off into their “field.” The hedge separating Alaric, a cultural anthropologist who is passionate about African masks (he wears them in his own time, to the stupefaction of his English neighbors), from young Tom, a devotee of structural functionalism, fascinated by the role of the maternal uncle in matrilineages, symbolizes in literary fashion the animosity between the two persuasions. Barbara Pym, who was also editorial secretary of the magazine Africa, well knew the mores of the anthropologists of her time…
These texts were brought together by Kroeber and published under the title Anthropology Today: An Encyclopedic Inventory in 1953.
One may also note the presence in other sessions of the linguist André Martinet, who was then in post at Columbia University; Pierre Teilhard de Chardin, who lived in New York; and Henri Vallois, director of the Musée de l’Homme in Paris.
Lowie was born in Vienna in 1883, Kroeber was born in 1876 in New York into a German-speaking family. We should remember that several founding figures of American anthropology (including Boas, born in Westphalia in 1858) therefore had their origins in the Germanic world. One might therefore interpret this polemic between British and American anthropologists as an opposition between a German tradition, successor to the historic sciences and to diffusionism, and a positivist Franco-English tradition, continuing from Spencer and Durkheim.
Lowie develops his point of view after the Wenner Gren symposium in an article published as a sequel to the Murdock/Firth/Radcliffe-Brown debate in American Anthropologist (Lowie 1953).
Eggan is a specialist in the Indians of North America. One may note that his work of 1937 is explicitly located in the domain of social anthropology.
Mead is cited with positive comments in Les Structures élémentaires de la parenté (1967 , 556, 559) and in texts previous to 1952 of the volume Anthropologie structurale (1958); but the volume of the Pléiade, which contains seven works by Lévi-Strauss (2008) and the four volumes of the Mythologiques, make no reference to this author. Margaret Mead almost completely disappears from the theoretical horizon of Lévi-Strauss after 1952: to my knowledge, the only exception is the article entitled “Reflections on the Atom of Kinship” (1973, 103–35), in which Lévi-Strauss makes abundant use of a comparative work by Mead on kinship (1935) to illustrate his point of view.
In 1953, Talcott Parsons, visiting Paris, offered Lévi-Strauss a professorial post at Harvard, but he turned it down. If I may be permitted to relate this anecdote, Lévi-Strauss declared that it was “because of an American lady who wrote unkindly in a book that [he had] returned to France because [he could not find] a post in the United States.” Kurt Lewin and Kroeber had also unsuccessfully offered a post to Lévi-Strauss a little earlier (Lévi-Strauss and Didier Eribon 1988, 82).The fact remains that the retreat of Lévi-Strauss from the American academic world undoubtedly damaged his influence on the other side of the Atlantic.
The second edition of The Elementary Structures of Kinship (1967) therefore includes a new preface almost entirely devoted to the interpretation of the theses on Lévi-Strauss by Rodney Needham (1962), who was an Oxford professor, introducer, and proponent of structuralism in the United Kingdom. This interpretation is mistaken according to the author, but he remains grateful to Needham for having made the Anglophone readership aware of his work. In 1969, it was in London (at Tavistock publishers), and not in the United States, that the English version of the first book by Lévi-Strauss was published, with Needham contributing to the translation.
Had it been defined more broadly, at least four more candidates could have presented themselves for this chair: Georges Balandier, Roger Bastide, Michel Leiris, and Alfred Métraux. They were all sociologists and ethnographers in the same way as the candidate who was selected, and had published several major works by the date of the election to the Collège de France.
In his research into a table of “ontologies” which would be universally valid, Philippe Descola (2005) takes up the model of social anthropology as it was defined by Lévi-Strauss in the 1950s.