AbstractThe knowability paradox, or Fitch's paradox, is thought to threaten semantical (Dummettian) antirealism. This chapter suggests that the lesson of the paradox concerns the theoretical location at which to impose the antirealist's epistemic constraints on truth, i.e., on the central notion of the antirealist's meaning theory. In particular, the knowability principle that every truth is knowable is not a successful way of capturing the antirealistic insight that truth is epistemically conditioned.
