Durkheim argued that the significance of each *totem as a †symbol stemmed, not from any intrinsic attribute, but from its position in the structure of clan totemism. The influence that Durkheim’s theory of the social origin of the meaning of totemic emblems had on †Saussure and the formulation of his structural theory of †semiology is wellknown. Lévi-Strauss later developed the structural theory of totemism, most notably in Chapter 4 of The Savage Mind. He here compares the structural logic of central Australian totemism with that of the Indian caste system. A structural approach is also taken in Stanner’s analyses of Murinbata religion, which gains from its basis in Stanner’s own fieldwork among the Murinbata. Stanner records, in a footnote, that he only learned of Lévi-Strauss’s analysis after he had commenced publication of this series of papers. The semiological approach to art and ritual was brilliantly taken up by Nancy Munn (1973) in her studies of art among the Warlpiri, and by Morphy (1991) in his work on Yolngu art. Both have taken a more generative approach to art and ritual, made possible by Saussure’s development of the Durkheimian theory. In their work the artistic tradition is seen to provide a grammar as well as a vocabulary of visual signs, allowing artists opportunities to create new works rather than simply to reproduce totemic emblems whose form is fixed by tradition.
A similar approach has been taken in the study of ceremony. It is questionable how many performances of the major ceremonies which Warner describes in his ethnography he actually observed, but Warner appears to commit the Durkheimian fallacy of assuming that each performance of a ritual is identical and only amenable to one level of interpretation. †Ronald Berndt restudied the two major Arnhem Land cults documented by Warner and, while Warner’s ‘native’ exegesis appeared to support a functionalist interpretation, Berndt’s informants opted for a more Freudian reading of the rituals (Berndt 1951). More recently, Morphy has shown how Yolngu ceremonies are to a certain extent constructed to suit the occasion, while Keen (1994) has demonstrated that the Gunabibi and Wawilak cults are merely two among many in the region which interpret common elements in different ways. Special mention should also be made of †Kaberry’s pioneering work in the Kimberleys, which showed that Aboriginal women had their own rituals, of which male anthropologists had been unaware