Wolfgang Schwarz

Parsimony and Ontological Dependence

This is part 2 of my comments on Fiction and Metaphysics.

Amie Thomasson argues that fictional objects are not as strange and special as one might have thought because they belong to the same basic ontological category as works of art, governments, chairs and other objects of everyday life. Doing without fictional entities, she says, would merely be "false parsimony" unless one can also do without other entities of this category.

I have three complaints.

First, I don't believe in the distinction between good and false parsimony. In particular, I don't agree that doing without some entities is only worthwile if one can also do without all other entities of the same kind. Otherwise it would be worthless to reject phlogiston and unicorns while still believing in oxygen and horses.

But wouldn't it be false parsimony to reject baseball games while still accepting board games? Or to believe that there are only 127 living people rather than 6 billion? Or to believe that there is only one world rather than proper-class many?

The problem with these examples is that it is quite hard to imagine theories that differ on these questions but are equally good otherwise. Of course we shouldn't believe in an implausible, inelegant or empirically inadequate theory according to which there are only 127 people. But that's simply because we shouldn't believe in an implausible, inelegant or empirically inadequate theory anyway. So what we have to imagine is a theory that in every methodological respect (except the number of people it postulates) is just as good as our best alternative, but says that there are only 127 people. Should we then prefer this theory? I don't know.

In fact, I believe that parsimony is not an independent methodological virtue at all. Unparsimonious theories are usually bad not because they are unparsimonious but because they are needlessly complicated and because they violate our pre-theoretical opinions about what there is. Other things being equal, I think I would prefer the 6-billion-people theory over the 127-people theory and over the no-people theory. And I would prefer the 1-world-theory over the many-worlds-theory and the no-world-theory (if all these were equally good otherwise).

So while I agree with Thomasson that the mere fact that some theory A postulates more entities of a given kind than an alternative theory B is in itself not a good reason to prefer B over A, I do not believe that this has anything to do with kinds and categories. Which theory we should prefer depends on their methodological virtues, and parsimony doesn't belong to these virtues. (or, at any rate, it is only a derived virtue.)

This leads to my second complaint. Saying that fictional characters are not new kinds of entities because they share certain features with other commonly recognized entities like chairs and governments sounds like a very cheap trick: For any entities that you could possibly postulate you'll find some features they share with already recognized entities. Cartesian spirits? Universals? Not new kinds of entity, just like numbers: immaterial; Events? Tropes? Just like tables: particular.

My third complaint concerns the feature Thomasson chooses for grouping together fictional characters with other things: ontological dependence. Some contingent entity A ontologically depends on some contingent entity B iff necessarily, if A exists then B exists. Thomasson offers lots of examples in support of her view that many things are in this sense dependent on other things. I'm afraid that I find virtually none of these examples convincing.

The only uncontroversial case of dependence I see is the dependence of every thing on itself. But this hardly provides the means for ontological categorisation. Here are some other examples from Thomasson's book:

• People are dependent on their parents. -- I don't think so. I can perfectly well imagine that some person, say Kripke, was created by lightning in a swamp.
• Chairs are dependent on the intentions of their manifacturers. -- Again, I have no problems with imagining that this very chair I'm sitting on was created differently, or just grew on a tree in said swamp. (Thomasson argues that we wouldn't call such a naturally grown object a chair. If this is correct, I don't believe that being a chair is an essential property of this chair.)
• The colour of this apple depends on the apple. -- Granted that this thing exists at all, wouldn't it also exist if only the apple's peal existed?
• The property of being a son of Adam depends on Adam. -- Granted that this property exists, I have no intuitions whatsoever about the existence conditions of properties.
• Governments depend on the intentions and behaviour of people. -- Again, I'm very uncertain about the existence conditions of governments. Here I'm willing to follow anyone who offers a good theory of governments. If this theory identifies governments with certain set-theoretic constructions, I'll be happy to say that governments aren't contingent in the first place. If the theory says they can be identified with the people who do the governing, I'll be happy to say that the depence on people is a trivial matter of identity. If the theory says that governments are sui generis dependent entities, I'll be happy to say that they are sui generis dependent entities. But I won't agree to this before I've seen the theory. (In fact, I doubt that the best theory of governments treats them as sui generis dependent entities.)
• Fictional works depend on the intentional acts by which they were created. -- I'm quite confident that Arthur Conan Doyle could have written his Sherlock Holmes stories a month later than he actually did, and in somewhat different mental and physical conditions. Would this still have been the very same intentional act? I doubt it, though I have no clear opinions about the individuation of intentional acts.

Note that I don't claim that there are no entities that depend on other entities. I only say that I find none of Thomasson's examples convincing.

Note also that my three complaints are not meant to be arguments against Thomasson's theory of fictionality. I'm skeptical about this theory mainly because it doesn't solve the problems it's supposed to solve (and even creates some further problems), not because it postulates fictional objects. In fact, I'm somewhat inclined to agree that we need fictional objects to make sense of fictional discourse. But these fictional objects will need to have more or less the properties attributed to them in the fiction. ("More or less" to allow for incoherent fictions.) Well, I know, it's always easy to vaguely hint at an approach, and much more difficult to carry it out.