Two quick points arising out of Wednesday’s discussion.
Ben Russell (a PhD student in linguistics at Brown) noticed that the following sentence seems odd.
(1) ??I know that Granger might be in Prague.
At least, on Wednesday we all thought it sounded odd, though a few philosophy students I’ve spoken to since then disagree. If it is odd, it is hard to explain why except on a view that takes epistemic modals to be speech act modifiers. It’s certainly very hard to see on our view (or on the contextualist views we attack) this should turn out at all problematic. I can easily say that I know that for all I know Granger is in Prague, and it’s not clear why (1) is more problematic.
One case we worry about a lot in our epistemic modals paper is Maze
Tom is stuck in a maze. Sally knows the way out is to the right, and knows she knows this, but doesnt want to tell Tom. Tom asks whether the exit is to the left. Sally says, It might be. It might not be. Sally might be being unhelpful here, but it isnt clear that she is lying.
This is a problem for most theories of epistemic modals, if the modality in each sentence (in particular the first) is really epistemic. Polly Jacobson suggested it might not be. Here’s an argument for that conclusion. If it’s an epistemic modal, it must somehow mean that for all Tom knows, the exit is to the left. It must be related to Tom’s knowledge, because if it was Sally’s knowledge then the sentence would be false, and there’s no one in the story but Tom and Sally.
Change the case so that Tom really does know the way out, but Sally doesn’t know this. It still seems, I think, that Sally is being unhelpful rather than untruthful. But if her sentence is an epistemic modal relating to Tom’s knowledge, she would be being untruthful here. So I suspect Polly is correct, the modal in “It might be” is not epistemic.