Reading Guide to

Reading Guide to: Levi-Strauss, C (1977) 'Structural Analysis in Linguistics and in Anthropology', in Structural Anthropology, Harmondsworth: Peregrine Books

Dave Harris

Linguistics has become the key to social science, and other social scientists need to learn from it. This is especially so in the anthropological study of kinship, where 'The linguist provides the anthropologist with etymologies which permit him to establish between certain kinship terms relationships that were not immediately apparent' (32), while the anthropologist can explain which types of kinship are common and which persist. As another example, linguists can point to the survival in a modern vocabularies of terms for kinship relationships which have long disappeared.

The real revelation arose with structural linguistics which has had a major impact on a number of disciplines. Troubetzkoy, one of the founders, explains the structural method as consisting of four basic operations:

(a) we do not study conscious linguistic phenomena, but their unconscious infrastructure

(b) we study not independent terms but the relations between terms

(d) we attempt to discover general laws of operation -- 'either by induction "or... by logical deduction, which would give them an absolute character"' (Levi-Strauss quoting Troubetzkoy, page 33).

In this way, social science focuses on 'necessary relationships' (33), and this opens up new perspectives in other social sciences. For anthropologists, 'kinship terms are elements of meaning; like phonemes, they acquire meaning only if they are integrated into systems', and the systems are 'built by the mind on the level of unconscious thought' (34). Furthermore, despite their empirical variability , 'the observable phenomena result from the action of laws which are general but implicit' (34). We should be pursuing the basic methods of analysis as outlined above. Apart from anything else, this would help anthropology overcome one of its major problems, which arises from variability of the data and the temptation towards individualistic, local and historical explanations. The result is 'a chaos of discontinuity' (35) even though we know kinship systems are universal.

However, we must guard against too easy an application of linguistic method. Linguists can proceed by analysing phonemes as groups of one or more sets of [vocal] oppositions. Kinship terms can also be broken into components, such as the term 'father', and this could be associated with various 'connotations -- positive or negative... [for] each of the following relationships: generation, collaterality, sex, relative age, affinity etc' (35). We could then generalise from there. However this would be a mistake: 'A truly scientific analysis must be real, simplifying, and explanatory' (35) and although this fits phonemic analysis, it does not fit kinship terms if we treat them in this way, because 'one moves away from [the concrete]... and the definitive system... is only conceptual' (36). Finally, we gain no explanatory value, since we do not understand the nature of the system nor its origins.

Since kinship terms are elements of speech as well as elements of social structure and action, we must treat them literally as linguistic phenomena first. That is, we must break them down into the equivalent of phonemes rather than working with words themselves. Studying relationships between words will not reveal underlying structures.

There are other differences between kinship systems and linguistic systems too. For linguists, the function of language 'was obvious; the system remained unknown. In this respect, the anthropologist finds himself in the opposite situation' (37). Until we do know the function of kinship terms, we are left with the risk of tautology, or analyses that 'demonstrate the obvious and neglect the unknown' (37). However, we can proceed with a clear case which demonstrates the power of the analogy.

A kinship system has 'two quite different orders of reality' (37). There are terms to express family relationships, but there are also expectations and obligations felt by individuals. This gives us both a 'system of terminology', and a 'system of attitudes'. With the latter we can guess at its function, 'to insure group cohesion and equilibrium' (37), but we don't know how the various attitudes are interconnected and how they form a system. The two systems are different. The psychological and social dimensions of attitudes are also different [the latter are those which are socially supported, 'stylised, prescribed, and sanctioned by taboos or privileges' (38)]. In many cases, these attitudes helped to reinforce or solve problems with terminological systems. [An example turning on 'joking privileges' among Australians illustrates his point, page 38 -- if I have understood it correctly, kinship relations between two unmarried men do not themselves technically permit particular important kinds of marriage, but 'joking privileges' do]. Despite this possible functional relationship between two systems, they have to be treated independently, however.

Using linguistic methods can help us to solve a major problem in anthropology -- the role of the maternal uncle and the relationship with nephews. This can play an important part in 'a great many primitive societies' (39). This relationship was once explained as a survival of matrilineal descent, but this did not fit all the examples, and additional early relationships had to be also cited --'Thus, atomism and mechanism triumphed' (39). Further, systematic analysis showed that an important avunculate relationship could also be found with patrilineal descent. What seemed to be at stake were more general tendencies '"to associate definite social relations with definite forms of kinship"' (Levi-Strauss citing Lowie, page 40). Interest also turned to how attitudes were modified. Yet the whole answer had still not been provided, particularly the question 'why are only certain attitudes associated with the avuncular relationship, rather than just any possible attitudes...?' (40).

There are clear parallels with earlier versions of linguistic theory. Given the huge variety of sounds that can be produced by human beings, a question for linguistics was why only certain sounds were selected, and what the relationship between the sounds might be. The same goes for anthropology -- given a huge variety of 'psycho-physiological material at its disposal' (40) why should particular relationships be retained even in the most diverse cultures?

Radcliffe-Brown got close with his study of the avunculate in South Africa, by noticing that two opposite systems of attitudes were condensed in the term. First, the maternal uncle represents family authority, but the nephew also holds 'privileges of familiarity in relation to his uncle and can treat him more or less as his victim' (41). Secondly attitudes towards maternal uncles and fathers are correlated inversely -- if father-and-son relations are familiar, nephew and uncle relations are more distant and more respectful. Thus, structurally speaking, we have two pairs of oppositions. Radcliffe-Brown thought that forms of descent would determine the choice of these oppositions, so that in patrilineal societies, maternal uncles were treated as 'male mothers' (invoking familiarity), and in matrilineal societies it was the other way around. However, some questions were still unanswered, since the avunculate is not associated always with either matrilineal or patrilineal systems. Further, there are actually four terms involved --'brother, sister, brother-in-law, and nephew' (41).

Some further examples illustrate the complexities -- the Trobriand Islanders have matrilineal descent, familiar relations between father and son, and antagonism between maternal uncle and nephew, while the Cherkess 'place the hostility between father and son, while the maternal uncle assists his nephew' (42). So far, this fits with Radcliffe- Brown. However, while husbands and wives have warm and intimate relationships in the Trobriands, brothers and sisters are much cooler and more rigid. Among the Cherkess, brothers and sisters are much more friendly and tender, but the relations between spouses are different -- 'A Cherkess will not appear in public with his wife and visits her only in secret' (42).

So there is a correlation between father/son and uncle/nephew, but we need consider relations between brothers and sisters, husbands and wives, and brothers and brothers-in-law as well. We can then formulate a general law: 'the relation between maternal uncle and nephew is to the relation between brother and sister as the relation between father and son is to that between husband and wife' (42). Thus if we know one pair of relations, we can infer the other. [And this analysis has applied to other examples, such as kinship in Tonga, New Guinea (Lake Kutubu), Bougainville (Siuai), and among the Dobu (Trobriand Islands).. The symmetry between these different systems is illustrated in the famous diagram on page 45. NB it 'is an over-simplification' (44) -- see below].

A diachronic analysis supports this general system too, so that family relationships in the Middle Ages [in Europe?] shows a pattern where as brothers' authority over sisters decreases, husbands' authority over wives increases.

Thus the avunculate is but one relationship within the system, and we understand it only once we have grasped the structure of the system as a whole. There are four basic terms '(brother, sister, father, and son), which are linked by two pairs of correlated oppositions in such a way that in each of the two generations there is always a positive relationship and a negative one' (46). It is this system which is 'the most elementary form of kinship that can exist. It is, properly speaking, the unit of kinship' (46). This is so logically, since there must always be three types of family relations in any kinship system -- 'a relation of consanguinity, a relation of affinity, and a relation of descent -- in other words, a relation between siblings, a relation between spouses, and a relation between parent and child' (46). These relationships are expressed in the basic unit described above, 'in accordance with the scientific principle of parsimony' (46).

The unit also expresses the 'universal presence of an incest taboo'. What this means is that 'a man must obtain a woman from another man who gives him a daughter or a sister' (46) -- this other man (the maternal uncle) thus is already present, and is a necessary precondition, requiring no explanation. The existence of the child provides another indispensable relation, in compensating for the 'initial disequilibrium produced in one generation between the group that gives the woman and the group that receives her' (47), producing stability over time. Incidentally, sisters and aunts could constitute a system, but this 'is immediately eliminated on empirical grounds. In human society, it is the men who exchange the women, and not vice-versa' (47).

Why is the avunculate not universal? For one thing, kinship is not as important in all cultures, as in our own society. Other means of expression and ways of regulating social relationships may arise instead. Thus while language must always signify, social systems may have different levels of expression. Further, elementary structures of kinship can be built into more complex systems: here, the 'avuncular relationship, while present, may be submerged within a differentiated context' (48), as in systems where the father's sisters and their husbands also have a role [Levi-Strauss argues that these systems too must reproduce the logic he has illustrated by showing 'symmetrical and inverse' relations to link the next generations].

Back to that over-simplification in the diagram. Actual relationships are more than just a warm or cool. There are usually four basic attitudes: 'mutuality, reciprocity, rights, and obligations' (49), and these often come in the form of bundles of attitudes.

These elementary structures are not the same as the biological families identified as fundamental by many other anthropologists. Kinship is socio-cultural, and characteristically diverges from nature. Kinship exists only in consciousness, and it is arbitrary rather than 'the spontaneous development of the real situation' (50). The real situation does have an influence, though, as when biological parenthood contradicts and overpowers artificial systems of kinship [and Levi-Strauss gives an Australian example]. But kinship is constituted through marriage, and there is no biological order producing marriage. It is the relationship between terms which produce families, and 'No other interpretation can account for the universality of the incest taboo; and the avuncular relationship, in its most general form, is nothing but a corollary, now covert, now explicit, of this taboo' (51).

Kinship systems are symbolic systems, and we can now allow anthropology to converge with linguistics. Symbolic systems may have a naturalistic base, but they have emergent qualities in their own right -- 'any concession to naturalism might jeopardise the immense progress already made in linguistics... and might drive the sociology of the family towards a sterile empiricism, devoid of inspiration' (51).