Kai von Fintel has a few good questions about the literature on truth in fiction. Roughly, he wants to know which theories lead to the intuitively plausible result that propositions like “In the Sherlock Holmes fictions, Inspector Lestrade has blood type A” are neither true nor false because of presupposition failure. Off the top of my head, I think I recall that Currie’s and Walton’s theories have this feature, but I could be mistaken. Anyway, it’s a little late for me to do actual, er, research on this, but Kai would be happy if anyone who has more info would click over to his site and leave a comment.