ABSTRACT
In Edwin Ardener's influential paper 'Belief and the problem of women' (1972), the 'problem', that of finding out how women see the world of which they are a part when our only informants are men, is shown to be twofold. There is a 'technical' problem: women are less likely to 'speak', to act as our sources. There is also an 'analytical', or conceptual, dimension for, even when our informants are women, the model of the world and of their place in it which they give may be less acceptable to the observer than the neat, bounded categories given by the male informant (1972, pp. 1-3).
