On Fictional Entities

Notes and Fragments ON FICTIONAL ENTITIES by Susan L. Feagin Peter van inwagen has recently defended in the pages of this journad the view that a correct understanding of fictional discourse requires ontological commitment to fictional entities, i.e., fictional persons, places, events, and so forth.1 Applying Quinean criteria for existence, he argues that any fairly sophisticated involvement in fiction must be analyzed in such a way that variables representing fictional entities turn out to be bound by existential quantifiers . Indeed, I find one of the most appealing implications of his approach to be its capacity for distinguishing critical discourse which ainalyzes the internal structure of a fictional work (remarks about characters, characterization, style, etc. — what he calls "grade two" criticism) from critical remarks which treat it as an undifferentiated whole ("grade one" criticism), though I have more doubts dian he does about whether "grade two" remarks apply only to fiction. Notwidistanding its merits, I am still uncomfortable with claims that fictional entities exist, even if they are claimed to exist in the same sense that anything else exists, so I will look more closely at three claims van Inwagen makes in support of his position. His first, most threatening, challenge is diat anyone who denies his view is required to offer an account of how to construe critical remarks which ostensibly carry a commitment to the existence of fictional entities in a way so that they do not make this commitment. This is a daunting task, and in case anyone feels free just to reject the challenge, he says diat this would be analogous to claiming "God created the world" while refusing to admit the existence of God and also refusing to explain how one's claims could mean anything and not require die existence of God. But I fear die analogy is not fair because I, unlike van Inwagen, think that what we need as an alternative to his view is not a theory about the ontological commitments of the language of grade two criticism, but a theory of the language of fiction. Van Inwagen says diat fiction does not entail the existence of fictional entities; it is only because we 240 Susan L. Feagin241 warnt to talk about fictional entities that we are required to say that they exist. But why do we want to talk about them? Because of the way we construe the lauiguage offiction itself. Ifwe had an account of the ontological commitments of fiction which could adequately capture what is said in the fictional work (its "internal structure") without committing us to the existence of characters, etc., we would not have to tadk about fiction in a way which entailed the existence of fictional entities. Grade two critical remarks contain claims about the internal structure of stories — their style, form, characterizations, and so forth — but a critic is no more ontologically committed to the existence of the characters in a novel than I am committed to saying the sun moves when I say "the sun rises." Of course, if I wanted to be scientifically precise I would say something about the rotation of the earth and the paths of light from the sun in relation to my current horizon (do horizons exist?), but I do not because (a) it would be needlessly cumbersome ,2 and (b) like the critic, I do not in fact know exactly what to say in order to avoid reference to entities whose existence is questionable. If we had a philosophy of language which accounted for fictional content without being committed to the existence of fictional entities, there is no logical reason why (true? useful? relevant?) critical remarks cannot also be phrased in ways which avoid ontological commitment to fictional entities. That is, we could rearrange the language of criticism to suit, but whether anyone would want to do so is another question. Whether diis rephrasing would "mean" the same thing seems to me to be a red herring. Does the more scientifically precise rendering of "the sun rose at 6:15 this morning" mean the same thing? Could van Inwagen rule out any competing account a priori on the basis that it did not capture the "meaning...