Carnap, the universality of language and extremality axioms

By far the most important, albeit largely unrecognized, background factor in twentieth-century philosophy of language — and to a considerable extent in philosophy in general — is a contrast between two ways of looking at language and at our relation to it. I have called them the assumption of the universality of language (or of language as the universal medium) and the conception of language as calculus. I have studied their role in a number of publications. (See Hintikka 1988, with further references.) Roughly, on the view of language as the universal medium we are stuck with our actual language. We cannot step outside it and study its relation to the world — or, at least, cannot express in language such an enterprise or its results. Hence all systematic, large-scale semantics (model theory) is impossible — or at least inexpressible — according to this view. In contrast, on the view of language as calculus we can systematically vary the interpretation of our actual language, just as we can vary the interpretation of a formal calculus. In other words, we can develop a viable model theory (model-theoretical semantics) for our actual everyday language and/or the language of science.