ABSTRACT
Contemporary anthropologists, irrespective of whether they take a more 'cultural' or a more 'sociological' perspective, tend to assume that a set of collective representations, or an ideology, or the pattern of symbolic or cultural constructs can provide us with an understanding of important distinctions within a given culture and, more radically, with an understanding of how these important distinctions organise the subject's view of his or her world. This assumption is virtually always implicit rather than explicit, simply because the apparently most significant set of collective representations, or symbolic constructs, or ideology,
may emerge out of analysis as the creation of the anthropologist and, at best, can give us only a partial understanding of the subject's view of his or her own culture or society. And this may be the case even where the anthropologist is specifically concerned to do justice to indigenous categories and indigenous perspectives with respect to that society. I have suggested elsewhere that this problem arises out of the conflation of representation with the processes of thought, an approach that may explain the strong relativist stance taken by anthropologists such as Geertz and Evans-Pritchard in his later work, an approach that has been questioned by a number of anthropologists. 1
Clearly the primary task of the anthropologist is to discover the various principles of order for a given society and this inevitably entails an examination of modes of representation and the interconnections between them. Far from wishing to dismiss such a procedure I have to point out that my own data have largely been gathered and analysed using well-tried anthropological methods. However, as one whose initial training was in cognitive psychology, I have to object to the almost imperceptible slide that makes representation virtually identical with thought. Apart from the fact that representation is not the same as thought, the conflation of the two categories can lead to further systematic errors so that the anthropologist unwittingly privileges certain data, giving them a prominence that they may not necessarily deserve.
