non-discursive other than in contextualizing discursive categories, be they scientific, political or whatever.

Related to this is the question of relativism. It is sometimes assumed that Laclau and Mouffe must be taking up a position of epistemological relativism, but nothing could be further from the case. As may readily be noted, although 'truth' is always theoretically contextual in their frame of reference, there is no shortage of truth claims in their own theoretical discourse. One interesting example here is to look at their treatment of the question of ideology, for so long a stumbling block in terms of the assignation of real interests, correct consciousness, and so on. Laclau and Mouffe's attachment to epistemological security is such that they even take on, within the terms of their own model, the old conundrum about whether people can be said to be 'oppressed' if they themselves do not think they are. This is the subject of a fascinating distinction that they draw between 'subordination' and 'oppression': the former simply marks a set of differential positions between social agents, whereas the latter requires a point exterior to the discourse from which -- for 'oppression' to exist -- the discourse of subordination can be interrupted. And just for those who still see relativism as indexically linked to privileging the discursive, let me quote their definition of I relations of domination': 'those relations . . . which are considered as illegitimate from the perspective, or in the judgement, of a social agent external to them'. ⁵⁶ Far from being 'relativist', these confident formulations, spoken naturally from the position of the judging external agent rather than that of the judges, err on the side of being hard to justify in epistemological terms.

So it is perhaps not surprising to find Laclau and Mouffe offering us a defence of the 'critical', 'epistemological' view of ideology, but of course a fundamentally reformulated one. There are points in the argument of Hegemony and Socialist Strategy where one can say that for Laclau and Mouffe something is 'essentially' of such and such a character, and this is an important recognition. A key point of interaction between epistemology and the general concerns I have indicated about ideology can be found in the conclusion of Laclau's article "'The Impossibility of Society'". Here Laclau clarifies the solid epistemological foundation of their 'anti-essentialism': 'We cannot do without the concept of misrecognition, precisely because the very assertion that the "identity and homogeneity of social agents is an illusion" cannot be formulated without introducing the category of misrecognition.' Hence Laclau concludes that both the category of ideology and that of misrecognition can be retained, but by inverting their traditional content: he suggests that 'the ideological would not consist of the misrecognition of a positive essence [an illusion as to real

-259-