The term ‘alliance’ as it is used in anthropology refers to those social relations created by marriage. The word derives from the French, where it can be used to describe marriage, or the fact of being in an ‘in-law’ relation.
The reason why a term derived from French is often used in anglophone anthropology, frequently replacing the term ‘affinity’, is because the importance of this type of relationship came to the fore in a controversy between those anthropologists influenced by the French anthropologist Levi-Strauss and those following what has been called descent theory.
Descent theory was a particular formulation of an old view of the history of human society. It concerned only that part of the latter which referred to what held pre-state societies together and ensured limits on anarchy. According to this theory, what led to stability among primitive people were descent groups, often organized in a segmentary fashion. Descent theory became elaborated in the work of British anthropologists in a number of studies of non-state societies in Africa and then became extended to many parts of the world where it fitted rather less well. Because of this, when descent theory was at its height, a number of writers challenged it by referring to Levi-Strauss’s book, which was subsequently translated under the title The Elementary Structures of Kinship (1969 [1949]).
In this topic Levi-Strauss proposed a quite different view of the history of society to that of the descent theorists. What held primitive societies together was a particular type of relation established through marriage, which Levi-Strauss called alliance. Societies so organized had an ‘elementary structure’. Elementary structures were principally found in places such as Aboriginal Australia, certain parts of Southeast Asia, Southern India and Aboriginal South America. These contrasted with ^complex structures, which were to be found in such places as Europe, Africa or among the Inuit (Eskimos). The marriage rules in societies with complex structures were said by Levi-Strauss to be negative because they only specified which kin-person one was not allowed to marry; for example, in such societies one is usually not allowed to marry a full sibling or a parent. In societies with elementary structures, however, the marriage rule is said to be positive because there the rule says what type of relative one must marry.
Levi-Strauss’s book, as its title indicates, only concerns elementary structures and it treats marriage rules as the institution which binds society together. To demonstrate how this occurs, he further distinguishes between two types of elementary structure. In the first type, society is divided into two groups which we can refer to as A and B. In such societies the positive marriage rule states that people of group A must marry people of group B, while people of group B must marry people of group A. Since Levi-Strauss, following the way of talking of the people studied, sees such marriages as the transfer of women from one group to another, such a system is said to be one of direct exchange. In the other type of elementary structure, indirect exchange, the rule is different and the society needs to contain more than two groups because, while the women of group A must marry into group B, the women of group B cannot marry back into group A but must marry into another group, group C; and the women of group C may have to marry into yet another group, D, or, in some cases, into group A. Indirect exchange thus encompasses cases where people marry in a circle and when the chain of groups transferring women is not closed. There is, however, a great difference between the two cases; when people marry in a circle the relationship between the groups is egalitarian, but if the circle is not closed the relationship between the groups is hierarchical.
Levi-Strauss’s book was not the first to discuss the significance of positive marriage rules. Before him, British and especially Dutch anthropologists working in Southeast Asia, such as van Wouden (1968 [1935]) had also stressed the significance of marriage rules for the linking of groups, but Elementary Structures of Kinship was certainly the most wide-ranging of such work.
Since its publication many criticisms have been levelled at it. First of all, the general evolutionary implications of the topic have either been criticized or ignored. Second, "Edmund Leach (1954), in a topic concerning highland Burma, argued that the social implications of marriage rules always needed to be considered in conjunction with other political and economic factors. Third, the correlation between marriage rules and the linking of groups proposed in the topic has been questioned. For example "L. Dumont (1983) has pointed out that although the kind of rule which Levi-Strauss would have no hesitation in taking to indicate an elementary structure exists in much of South India and parts of Sri Lanka, the social implication of such a rule is quite different from what it would be among, for example, the Australian Aborigines. For him affinity does not necessarily lead to the ‘alliance’ of social units.
There has also been much discussion about what exactly is meant by a ‘positive marriage rule’. "R. Needham (1962), who at first enthusiastically welcomed Levi-Strauss’s book, insisted on drawing a sharp distinction between ‘preferential’ and ‘prescriptive’ rules, with only the latter leading to elementary structures. Prescriptive rules were absolute and were normally accompanied with a kinship terminology which equated the term for accompanied spouse with that of the type of kin one had to marry; while systems with preferential rules did not make such an equation and expressed the rule only as a preference. This distinction has proved almost impossible to maintain with clarity, and Levi-Strauss (1965) himself has refused to accept it. This, however, leads to an even greater difficulty for the general theory: namely, that only if the rule is absolute can the social implications suggested by Levi-Strauss possibly exist.
