He was born in 1908 and trained in philosophy. In 1934, he was teaching at a secondary school in France when he was offered a post as Professor of Sociology at the University of Sao Paulo in Brazil, on the provision that he give his answer within three hours. He spent the next five years in Brazil, during which he made extensive field trips into remote areas of the country.
Much later, he explained his decision by saying, "I became an anthropologist, as a matter of fact not because I was interested in anthropology, but because I was trying to get out of philosophy." In considering philosophy, he felt he was stopping half-way--confining himself to the established Western tradition--while anthropology seemed to offer a chance to explore the fullest range of human thought.
But what immediately struck him in the anthropological literature was a baffling welter of marriage customs. They couldn't be entirely meaningless, because versions of the same custom appeared all over the world, and yet their variety made the idea of consistent purpose seem absurd. His effort at structural analysis was an attempt to get these data in order. This theoretical work, combined with his field research, became the basis for the papers which brought him a reputation in the American branch of the discipline.
He returned to France for military service but, as a Jew, was forced to leave again when Germany occupied the country. He spent three years teaching at the New School for Social Research in New York City and, from 1946 to 1947, was cultural attache to the French embassy in Washington.
He returned to France in 1948 to become Professor of Ethnology at the University of Paris and held a series of other academic posts until his death in 1990.
His Brazilian field work is reflected in THE ELEMENTARY STRUCTURES OF KINSHIP (1949) and TRISTES TROPIQUES (1955), which is a more contemplative and literary work. The major works of the 1960s and 1970s, when his influence was greatest, are the four volumes called MYTHOLOGIES.
His theoretical views are set forth in STRUCTURAL ANTHROPOLOGY (1958). Briefly, he considers culture a system of symbolic communication, to be investigated with methods that others have used more narrowly in the discussion of novels, political speeches, sports, and movies.
His reasoning makes best sense against the background of an earlier generation's social theory. He wrote about this relationship for decades.
A preference for "functionalist" explanations dominated the social sciences from the turn of the century through the 1950s, which is to say that anthropologists and sociologists tried to state what a social act or institution was for. The existence of a thing was explained if it fulfilled a function. The only strong alternative to that kind of analysis was historical explanation, accounting for the existence of a social fact by saying how it came to be.
However, the idea of social function developed in two different ways. In the English school of Alfred Radcliffe-Brown, the goal of anthropological research was to find the collective function, what a religious creed or a set of rules about marriage did for the social order as a whole. At back of this approach was an old idea, the view that civilization developed through a series of phases from the primitive to the modern, everywhere the same. All of the activities in a given kind of society would partake of the same character; some sort of internal logic would cause one level of culture to evolve into the next. On this view, a society can easily be thought of as an organism, the parts functioning together like parts of a body.
The more influential functionalism of Bronislaw Malinowski described the satisfaction of individual needs, what a person got out of participating in a custom.
In the United States, where the shape of anthropology was set by the German-educated Franz Boas, the preference was for historical accounts. This approach had obvious problems, which Levi-Strauss praises Boas for facing squarely.
Historical information is seldom available for nonliterate cultures. The anthropologist fills in with comparisons to other cultures and is forced to rely on theories that have no evidential basis whatever, the old notion of universal stages of development or the claim that cultural resemblances are based on some untraced past contact between groups. Boas came to believe that no overall pattern in social development could be proven; for him, there was no history, only histories.
There are three broad choices involved in the divergence of these schools--each had to decide what kind of evidence to use; whether to emphasize the particulars of a single culture or look for patterns underlying all societies; and what the source of any underlying patterns might be, the definition of a common humanity.
Social scientists in all traditions relied on cross-cultural studies. It was always necessary to supplement information about a society with information about others. So some idea of a common human nature was implicit in each approach.
The critical distinction, then, remained: does a social fact exist because it is functional for the social order or because it is functional for the person? Do uniformities across cultures occur because of organizational needs that must be met everywhere or because of the uniform needs of human personality?
For Levi-Strauss, the choice was for the demands of the social order. He had no difficulty bringing out the inconsistencies and triviality of individualistic accounts. Malinowski said, for example, that magic beliefs come into being when people need to feel a sense of control over events where the outcome was uncertain. In the Trobriand Islands, he found the proof of this claim in the rites surrounding abortions and weaving skirts. But in the same tribes, there is no magic attached to making clay pots even though it is no more certain a business than weaving. So the explanation is not consistent. Furthermore, these explanations tend to be used in an ad hoc, superficial way--you just postulate a trait of personality when you need it.
But the accepted way of discussing organizational function didn't work either. Different societies might have institutions that were similar in many obvious ways and yet served different functions. Many tribal cultures divide the tribe into two groups and have elaborate rules about how the two groups can interact. But exactly what they can do--trade, intermarry--is different in different tribes; for that matter, so are the criteria for distinguishing the groups.
Nor will it do to say that dividing-in-two is a universal need of organizations, because there are a lot of tribes that thrive without it.
"A truly scientific analysis must be real, simplifying, and explanatory," he says (in STRUCTURAL ANTHROPOLOGY). Phonemic analysis reveals features that are real, in the sense that users of the language can recognize and respond to them. At the same time, a phoneme is an abstraction from language--not a sound, but a category of sound defined by the way it is distinguished from other categories through rules unique to the language. The entire sound-structure of a language can be generated from a relatively small number of rules.
In the study of the kinship systems that first concerned him, this ideal of explanation allowed a comprehensive organization of data that had been partly ordered by other researchers. The overall goal was to find out why family relations differed in different South American cultures. The father might have great authority over the son in one group, for example, with the relationship rigidly restricted by taboos. In another group, the mother's brother would have that kind of relationship with the son, while the father's relationship was relaxed and playful.
A number of partial patterns had been noted. Relations between the mother and father, for example, had some sort of reciprocity with those of father and son--if the mother had a dominant social status and was formal with the father, for example, then the father usually had close relations with the son. But these smaller patterns joined together in inconsistent ways.
One possible way of finding a master order was to rate all the positions in a kinship system along several dimensions. For example, the father was older than the son, the father produced the son, the father had the same sex as the son, and so on; the matrilineal uncle was older and of the same sex but did not produce the son, and so on. An exhaustive collection of such observations might cause an overall pattern to emerge.
But for Levi-Strauss, this kind of work was "analytical in appearance only." It results in a chart that is far harder to understand than the original data and is based on arbitrary abstractions (empirically, fathers are older than sons, but it is only the researcher who declares that this feature explains their relations). Furthermore, it doesn't explain anything. The explanation it offers is tautological--if age is crucial, then age explains a relationship. And it does not offer the possibility of inferring the origins of the structure.
A proper solution to the puzzle is to find a basic unit of kinship which can explain all the variations. It is a cluster of four roles--brother, sister, father, son. These are the roles that must be involved in any society that has an incest taboo requiring a man to obtain a wife from some man outside his own hereditary line. A brother can give away his sister, for example, whose son might reciprocate in the next generation by allowing his own sister to marry exogenously. The underlying demand is a continued circulation of women to keep various clans peacefully related.
Right or wrong, this solution displays the qualities of structural thinking. Even though Levi-Strauss frequently speaks of treating culture as the product of the axioms and corollaries that underlie it, or the phonemic differences that constitute it, he is concerned with the objective data of field research. He notes that it is logically possible for a different atom of kinship structure to exist--sister, sister's brother, brother's wife, daughter--but there are no real-world examples of relationships that can be derived from that grouping.
The purpose of structuralist explanation is to organize real data in the simplest effective way. All science, he says, is either structuralist or reductionist. In confronting such matters as the incest taboo, one is facing an objective limit of what the human mind has so far accepted. One could hypothesize some biological imperative underlying it, but so far as social order is concerned, the taboo has the effect of an irreducible fact. The social scientist can only work with the structures of human thought that arise from it.
And structural explanations can be tested and refuted. A mere analytic scheme that wishes causal relations into existence is not structuralist in this sense.
Levi-Strauss' later works are more controversial, in part because they impinge on the subject matter of other scholars. He believed that modern life and all history was founded on the same categories and transformations that he had discovered in the Brazilian back country--the raw and the cooked, from honey to ashes, the naked man (to borrow some titles from the MYTHOLOGIES). He also pointed out that the modern view of primitive cultures was simplistic in denying them a history. The categories of myth did not persist among them because nothing had happened--it was easy to find the evidence of defeat, migration, exile, repeated displacements of all the kinds known to recorded history. Instead, the mythic categories had encompassed these changes.
In sum, he argued for a view of human life as existing in two timelines simultaneously, the eventful one of history and the long cycles in which one set of fundamental mythic patterns dominates and then perhaps another. In this respect, his work resembles that of Fernand Braudel, the historian of the Mediterranean and 'la longue duree,' the cultural outlook and forms of social organization that persisted for centuries around that sea.