Of all the methodological and theoretical issues raised in the human sciences by structuralism, none is more vexed than its relation to history. Part - but only part - of this conundrum is resolved by analysing the various components of the term 'history'. There remain, nevertheless, certain problems in imagining a truly structuralist history, which we will explore below. However, distinguishing 'history' in at least four senses is helpful at the outset. First, and simplest, is the sense of the word meant by professional historians. To paraphrase Wittgenstein, history is what historians do. In this limited sense, we certainly see grounds for accommodation between history and structuralism, and, indeed, for a structural history of the sort produced by Fernand Braudel. Second is what Lévi-Strauss himself means by the term in at least one key passage: philosophical history, of the universal sort, whether in its Hegelian, Marxist, or evolutionist versions. (That is, historical models that assume a universal model of progression in all times and places (1966c: 257).) This he had little use for (although certainly he employs other aspects of Marx's thought) in the sense that such a schema presupposes the very questions that structuralist anthropology wishes to explore: the way that society is organised functionally and symbolically in relation to its own history and to the material world.