EN

The deflationist's intuition is that truth is in some sense 'insubstantial' or 'metaphysically thin'. The property of conservativeness comes as a handy explication of this intuition: the deflationist should adopt a theory of truth which is conservative over its base (syntactic) theory. Accepting the conservativeness requirement as given, we discuss a certain objection against deflationism: it was claimed that the deflationist can't explain various 'epistemic obligations', which we should accept once we adopt some base theory S. In particular, anyone who accepts a mathematical base theory S and understands the notion of truth, has a reason to accept the following: (1) Global reflection principle: All theorems of S are true. (2) Local reflection scheme: If a sentence 'phi' is provable in S, then 'phi'. But if our base theory is something like elementary arithmetic, then no deflationary truth theory can prove (1) or (2) on pain of losing its conservative character. In this situation the deflationist could still claim, that it is not our understanding of the notion of truth, but our knowledge of some additional non-semantic facts which explains our readiness to accept reflection principles. We claim however, that the explanation of this sort could perhaps be provided in case of (2), but in case of (1) - that of global reflection - it is out of the question. Since this path is blocked, the deflationist is left with a serious problem.