Let us accept that all identity is a differential identity. In that case two consequences follow: (1) that, as in a Saussurean system, each identity is what is is only through its differences from all the others; (2) that the context has to be a closed one – if all identities depend on the differential system, unless the latter defines its own limits, no identity would be finally constituted. But nothing is more difficult – from a logical point of view – than defining those limits. If we had a foundational perspective we could appeal to an ultimate ground which would be the source of all differences; but if we are dealing with a true pluralism of differences, if the differences are constitutive, we cannot go, in the search for the systematic limits that define a context, beyond the differences themselves. Now, the only way of defining a context is, as we have said, through its limits, and the only way of defining those limits is to point out what is beyond them. But what is beyond the limits can only be other differences, and in that case – given the constitutive character of all differences – it is impossible to establish whether these new differences are internal or external to the context. The very possibility of a limit and, ergo, a context, is thus jeopardized. (Laclau, Emancipation(s), 52)
This far, Laclau’s argument seems pretty reasonable, and the consequence presumably would be that, indeed, “no identity would be finally constituted,” that is, the differences that establish any identity would be indeterminate, there would be no point at which we could say we could say we “had” the full determination of the identity, and so any discussion of a particular identity would always be open to further questioning and the need for further investigation into the specificities of that identity. Oddly, though, that’s not the conclusion Laclau draws. Having “jeopardized” the possibility of a total context, Laclau nonetheless goes on to assert that a total context is possible, with the limit being provided not by a difference, but by an antagonism, which “poses a threat to (that is, negates) all the differences within that context.”
This is a strange move by Laclau, because it seems to be a (surprisingly un-poststructuralist) attempt to save structuralism, which proceeds via an odd sort of metalepsis in which practice is brought in to save theory. Laclau identifies a theoretical necessity, the need for something to provide a limit to a context; he names the thing that fills this need “antagonism,” which turns out to be the name of a category of political contingency. The examples Laclau gives of antagonisms are political (for example, the antagonism between worker and capitalist), but their role is theoretical, to deal with the problem – “from a logical point of view” – of establishing a Saussurean system of differences. What this means is that theoretical necessity always ends up ruling over the mere contingent details of politics.
Laclau’s position, then, is what we might call a postfoundationalist humblebrag. It looks like a recognition of the limits of theory, a humility before the impossibility of finding an absolute theoretical foundation for knowledge. But it turns out that Laclau’s theory is able to know precisely what these limits are, and, furthermore, to dictate practice on the basis of this knowledge of limits. I’d prefer a more genuine theoretical humility, which emphasizes that the absence of foundations means there’s no point at which we are finally free of further questions, that there’s no way of knowing what the limits are except by trying to answer the questions that arise in the course of trying to find out.