Blogosophy

Are fictional characters abstract?It doesn't

Are fictional characters abstract?

It doesn't seem likely. Ben Caplan has a nice discussion of this in his paper Fictional Characters and Other Abstract Objects According to him, David Lewis claims that abstract objects can be characterized in one of two ways: "The Way of Example" (numbers and sets are abstract, rocks and chairs are concrete) and "The Negative Way" (abstract objects aren't located in space or time, and they neither causally affect or are causally affected by anything). Further, Caplan points out that The Negative Way has definite problems when it comes to "creatures of fiction": If Dickens caused Mrs. Gamp to come into existence then she is in a causal relationship with him, and she exists in time since there was a time before which she didn't exist. He suggests that one might try to resolve these problems by arguing that The Negative Way is flat wrong (abstract object can exist in space & time and enter into certain kinds of causal relationships), or partially wrong (abstract objects can be located in time, just not in space), or one could concentrate on the Way of Example instead.

It seems to me, though, that it may just be better to abandon the contention that fictional characters are abstract objects, and grant them their own category. For one thing, on the Kripke-van Inwagen view discussed in Caplan's paper, as abstract objects fictional characters don't have the properties attributed to them (being old, being fat, being a woman), but instead have the property of having properties attributed to them. But this leads directly to wo's observation that on this kind of view, Sherlock Holmes (or Mrs. Gamp) is a ghostly invisible character who lives at no place in particular and never does anything at all. It seems to me that we want to be able to say "Sherlock Holmes was a detective" and have it be both meaningful and true (at least in some sense), and "Sherlock Holmes was a ghostly invisible character who never did anything at all, but the property of being a detective was attributed to him" isn't an adequate substitute. Moreover, John Burgess's Numbers and Ideas  paper offers a pretty plausible argument for why for abstract entities like numbers it makes no sense to ascribe a position in space and time and or to regard them as causally active or acted upon. If that's so, then either fictional characters are not abstract (or are a different kind of abstract), or it's wrong to argue as above that they exist in time, if not in space, and are caused by their authors--but the reasons that Burgess offers for supposing that about numbers (e.g. that it's crucial there are an infinity of them) don't apply to fictional characters.
Posted by joshua at May 16, 2003 05:26 AM