One of the chief objections to relativism runs like this: The relativist claims that all theories are relative to the theorist's (culture, race, religion, class, role, etc.). But, the objector asks, is the theory just stated above also relative in the same way? If it is, then the relativist cannot assert its truth. If it is not, then it denies the relativist thesis.
What follows is not an argument for or against relativism -- I simply wish to note that, at least in regards to relativism conceived in the particular fashion I outline below, the above argument does not work. Here's why:
The relativist is talking about the relationship of what we might call "first-level world views" when she states that "hteories are relative." That can be defended by noting that each such world view postulates certain entities, relationships, and so on, that it cannot demonstrate and does not interroagte, but that are used to demonstrate all other conclusions within the view. (E.g., physics postulates the idea of a "physical measurement.") Because systems arising from different postulates are incommensurable, there is no way to impartially abjudicate between these world views.
Let us call the level of theorizing of such views L1. Now, what the relativist is doing is moving to a level of theorizing we can call L2. L2 does not assume postulates about our immediate or scientific experience of the world. Instead, at L2, the theorist is concerned with examining the postulates of any such theory. Seeing that all of them depend crucially on their postulates, the L2-level theorist might reasonably conclude that all L1-level theories are dependent upon the cultural, social, economic, ideological, etc. committments of those who hold them. However, as someone theorizing at level L2, that statement need not be subject to the same caveat.
Of course, this in no way guarantees that some L3-level theorist will not come along and subsume the insights of the relativist into some higher synthesis. But that would in no way negate the insight achieved at L2.
Let me illustrate with an example. R.G. Collingwood, in The Idea of History, claims that all theories of human nature are historically contingent views based on human nature as understood by certain people at a certain time. Can Collingwood's claim be defeated by contending that it, too must be historically contingent? Not if he adopts the defence outlined above, where theories of human nature are L1 theories, while he is setting forth a theory about theories of human nature, an L2 theory.
The above view is deeply indebted to the consideration of theorizing appearing in the first 30 or so pages of Michael Oakeshott's On Human Conduct.