Description:
My claim in this paper is that the notion of linguistic category currently used in anthropology is outdated. This is particularly clear in the study of kinship systems, and the article starts from the classical example of the Aranda people in Central Australia to make this point. It is therefore necessary to elaborate a new linguistic theory if linguistic categories used in kinship terminology are to be understood. The linguistic theory advocated in this paper establishes the specific function of ritual in the construction of linguistic meaning, which is not considered any longer as a mere propositionnal content encapsulated in pragmatic communication. A major consequence is that the notion of totemism, which was once dismissed by the structuralist anthropology theory and which reappears in contemporary anthropological discourse, can be fitted again within this new framework.