The knowability paradox and truths dependent on their epistemic status

Byeong D Lee


On the anti-realist notion of truth, truth is not evidence-transcendent. Thus, it seems, the anti-realist should be committed to the claim that all truths are knowable. But Fitch showed that we can derive a contradiction from this knowability claim and two very plausible principles of knowledge. In this paper I argue that Fitch’s paradox arises because we wrongly assume the validity of the following inference: one is entitled to assert the existence of unknown truths; therefore, one is entitled to assert for some specific proposition that it is an unknown truth. Based on the invalidity of this inference, I show that we can block the derivation of Fitch’s proof without weakening the knowability claim, although I do not endorse the knowability claim itself. I also show that my restriction strategy disarms the paradox in such a way that is free from the criticisms made against Tennant-style restriction strategies.


  • There are currently no refbacks.