I don’t believe in metaphysical vagueness or, what I think would come to the same thing if it came to anything, metaphysical indeterminacy. I used to think I believed that the metaphysics of representation was vague, but that I think was a mistake. What I should have believed was that there is no metaphysics of representation. What scared me off I guess was the fact that the view most naturally associated with this slogan (there is no metaphysics of representation), Paul Horwich’s minimalism about truth, was thought to lead to epistemicism about vagueness. But Horwich was wrong about that, as John Hawthorne recently argued epistemicism goes best with hyper-steroid-inflated-maximalism about truth. We anti-realists, or at least reductionists, about representation are best off believing some wishy-washy degree-of-truth/supervaluational theory of vagueness. I might develop this more in a later post, but for now I just want to note, I don’t believe in metaphysical vagueness.
The problem is, I do believe, or at least until I realised the difficulty I used to believe, in some things that seemed somehow to imply metaphysical vagueness. Because I believe in fictional characters. I thought the van Inwagen argument for fictional characters was perfectly sound, I thought the Thomasson argument that fictional characters were no more ontologically problematic than other socially constructed abstracta was sound, and I didn’t like any of the alternative theories of fictional characters. Actually, now that I write that all down, I still think I believe all that.
The problem is that it seems any plausible theory of fictional characters will end up with them being somehow vague. It will be vague how many fictional characters there are, and it will be vague whether two fictional characters are identical.
On the first puzzle, I take it that the standard assumption is that when a name is used fictionally, it generates a fictional character. But it can be vague whether or not a certain term is a name. In Ulysses, one might wonder about each of the following terms whether it is genuinely used as a name: M’Intosh, The citizen, I (for the narrator in Cyclops), L. Boom, Blephen, Mhananann Mac Lir, Old Mr Verschoyle, Miss Spruce Conifer, the Archjoker Leopold Rudolph von Schwanzebbad-Hodenthaler, and so on. If there is a one-one correlation between names in fiction and fictional characters, and it is vague which of these are names, then it is vague how many characters there are. So, you might think, we shouldn’t assume a one-one correlation between names and characters. But any other proposal for how to determine when a fictional character exists leads to even worse vagueness. I fear that for any such proposal we can find vague instances even without putting down Ulysses.
Matters are even worse when it comes to identity. The man in the macintosh really is a character in Ulysses I take it. But he might be identical with a more familiar (as such) character. Some lines suggest that he is Mr Sinoco from the story A Painful Case in Dubliners. It seems a live possibility that it really is indeterminate whether Mr Sinoco is the man in the macintosh. If they are (both) real fictional characters, this would be genuine de re indeterminacy of indentity. And I might have mentioned that I don’t believe in any such thing.
So I’m a bit at a loss here. One of the following three things had better happen soon:
- I find a way to acknowledge fictional characters without having them have vague existence or identity conditions
- I give up believing in fictional characters, and find a way to respond to the arguments in their favour; or
- I learn to live with metaphysical vagueness.
The first of these is by far the most preferred, and, not coincidentally, the hardest to see how to make work.