In ‘The Future of Kinship Studies’ (1965) and in the new preface to the second French edition of Les Structures élémentaires de la parenté,1 Lévi-Strauss’s statements about ‘elementary structures’ pose with new precision the problem whether his typology can still be considered a useful analytical scheme.
As we have already seen, ‘elementary’ and ‘complex’ structures were originally defined by Lévi-Strauss as follows:
Elementary structures of kinship are those systems . . . which prescribe marriage with a certain type of relative. . . . The term ‘complex structures’ is reserved for systems which . . . leave the determination of the spouse to other mechanisms (1949: ix).
Even though his own definition states literally that ‘elementary structure’ implies prescription, Lévi-Strauss devotes the main part of the new preface to making quite clear his disagreement with Needham’s (1962a) interpretation of ‘elementary structures’ as prescriptive systems. Lévi-Strauss insists that the rather indiscriminate use of words ‘prescriptive’ and ‘preferential’ in his book is not an unfortunately vague way of writing but has to do with a theoretical issue.
I intend to analyse here the logical consistency of Lévi-Strauss’s typology by reference to the analytical criteria he employs, namely: elementary, complex; prescriptive, preferential; prohibition, choice; mechanical, statistical.
The above quotation on the definition of ‘elementary’ and ‘complex’ structures, and the subsequent line of analysis developed in the book, make it clear that the term ‘elementary’ denotes the existence of a positive rule concerning a certain category of individuals. Later, however, Lévi-Strauss adds:
All systems of kinship and marriage contain an ‘elementary’ core which manifests itself in the incest prohibition (1965: 18).
Therefore either prescriptions or prohibitions affecting marriage with an individual of a particular social category are the ‘elementary’ core of a system.
‘Complex structures’, on the other hand, are concerned with ‘choice’. This factor is introduced by the existence of various ‘mechanisms’ other than a prescriptive rule. Lévi-Strauss asserts that:
all systems have a ‘complex’ aspect, deriving from the fact that more than one individual can usually meet the requirements of even the most prescriptive systems, thus allowing for a certain freedom of choice (1965: 18).
Therefore, the term ‘complex’ undoubtedly implies ‘choice’. The first equations that can be established, then, are:
prohibition = elementary (1)
choice = complex. (2)
Concerning models, Lévi-Strauss states that there are:
societies which (as even our own) have a mechanical model to determine prohibited marriages and rely on a statistical model for those which are permissible (1958: 311).
Thus the representation of prohibited and possible marriages implies a code whereby:
prohibition = mechanical (3)
choice = statistical. (4)
From (1) and (3) we can deduce that an ideal ‘elementary structure’ is to be represented by a ‘mechanical model’, and from (2) and (4) that an ideal ‘complex structure’ is to be represented by a ‘statistical model’. In fact, Lévi-Strauss says:
In primitive societies these laws [marriage rules] can be expressed in models calling for actual grouping of the individuals according to kin or clan; these are mechanical models. . . . In our own society . . . it would be a statistical model (1958: 311).
So, as it is quite sure that his ‘own society’ would be classified as a ‘complex structure,’ we find indeed that:
elementary structure = mechanical model (5)
complex structure = statistical model. (6)
When dealing with the criteria of prescription and preference, Lévi-Strauss says that:
The difference between ‘prescriptive’ and ‘preferential’ does not appertain to the systems themselves, but to the way in which these systems are conceptualised, according to what I called elsewhere (1958) a ‘mechanical’ or a ‘statistical’ model (1965: 18; 1967: xxiii).
Whence we can deduce:
prescriptive = mechanical (7)
preferential = statistical. (8)
From (5) and (7), and from (6) and (8), we have:
elementary = prescriptive (9)
complex = preferential. (10)
But, since Lévi-Strauss also states that ‘an elementary structure can be equally preferential or prescriptive’ (1967: xxi), we can further say that:
elementary = preferential. (11)
From (10) and (11), and from (9) and (11), we arrive in the end at the conclusions:
prescriptive = preferential (12)
elementary = complex. (13)
1 The preface has also been published separately under the title ‘Vingt ans après’, in Les Temps Modernes, CCLXV, 1967: 385-406.