Fictional Realism, Nominalism, and Indeterminate Identity

I’ve posted a new paper: How to be a nominalist and a fictional realist. Here are the Cliff notes.

In my musical works paper, I argued that there are true claims proclaiming the existence of, and properties of, musical works, but that there weren’t really any musical works, because such claims were made true by an ontology that didn’t admit such things. In this paper, I attempt to tell a similar story for fictional characters. It’s literally true that the fictional character Bilbo Baggins exists, and it’s literally that he is a Hobbit according to the fiction The Lords of the Rings. But these claims can be made true without admitting fictional characters, or fictions, into our ontology. What makes them true, I suggest, are our acts of interpreting the fiction. Thus we can account for these truths with a nominalistically acceptable ontology (assuming, as I do, that there is in general a nominalistically acceptable account of the mental).

I also argue that the resulting view solves various puzzle cases concerning fictional characters. The most salient being Anthony Everett’s argument that fictional realism leads to untenable indeterminacy in identity. Everett argues that there are fictions in which it is indeterminate whether A is identical to B. The fictional realist believes in the fictional characters A and B. Whether the fictional characters are in reality identical is determined by whether they are identical according to the fiction to which they belong. So since it’s indeterminate whether they are identical in the fiction, it’s indeterminate in reality whether the fictional characters are identical. Reductio of fictional realist, given Evans’ argument against indeterminate identity.

I attempt to solve this puzzle by locating the source of the indeterminacy to indeterminacy in what fictional character is referred to, thus avoiding conflict with Evans’ conclusion (which is, as Lewis noted, directed only at indeterminate identity de re, not at indeterminacy in identity statements). Roughly, the idea is that when the fiction attempts to make an indeterminate identity, we are forced to interpret the fiction both ways. Given the above account, this results in there being two fictions, and two sets of fictional characters associated with each fiction, and it will as a result be indeterminate which fiction and which characters we refer to. In which case, the statement of identity will be indeterminate, but there will be no indeterminacy of identity de re.

Further details in the paper, of course; comments welcome.