Truth According to a Fiction
This comment by Gideon Rosen in the fascinating thread on IR at TAR made me smile:
Consider two opinionated journalistic essays on the same controversial topic ? say, the morality of Sharon-style extra-judicial killing, and suppose it's clear that both writers agree on the underlying facts. One says that the killings are unjustified because they violate a fundmental moral right to due process. The other says that there is natural rights are nonsense on stilts and that the killings are justified because they maximize utility. [...]
Now suppose that these are not editorials but Borgesian fictions. You can do the same exercise, and I think you get the same results: In the first story ? or perhaps better, according to the first story ? the killings are permissible because they maximize aggregate happiness. According to the second, they are impermissible because they violate a basic moral right.
What pleased me is that Rosen uses "according to the story" instead of "in the story" (even the emphasis is his). That goes very nicely with my recent claim that when we say that impossibilities are true in a story, we invoke a non-standard, hyperintensional notion of truth in fiction. "According to" is clearly a hyperintensional operator, and I'd be surprised if there were limits on its degree of hyperintensionality (short of direct quotation). Moreover, I think whenever S is a non-indexical sentence that literally occurs in a text, there is always a sense in which "according to the text, S" is true. But this is not so for truth in fiction. There are many cases where sentences occur in a story but are definitely not true in the story -- most obviously if the story has a prominent and clearly unreliable fictional narrator, like Huckleberry Finn.
I wonder if it would be better to disentangle the Impossibility Hypothesis from what Chalmers calls Strong Modal Rationalism, the view that (primary) possibility and a priori coherence in principle coincide. After all, most people seem to reject this claim. If it's false, there must be strong necessities in the sense of (primarily) necessary truths whose negation can't be ruled out a priori. In this case it's rather implausible to claim that all impossibilities give rise to IR. I'd have to restrict the Impossibility Hypothesis to impossibilities that are not the negation of strong necessities. I.e. I'd then say that we get IR* iff by the ordinary conventions of truth in fiction, some proposition p is true in the fiction, even though p is ruled out a priori; or better, even though the reader's conditional credence of p given the rest of the fiction is very low.
IR* is the phenomenon that a considerate reader judges a proposition to be false in a fiction even though by the ordinary conventions of truth in fiction, the proposition would have to be true in the fiction. I agree with everyone else that there are lots of other IR phenomena, and I have nothing new to say about most of them. (If anyone complains that my IR* has nothing to do with imagination and hence shouldn't be called "IR", I'm afraid I'd have to concur.)
Wo,
I think you are right that 'according to the story' is probably a healthier idiom than 'in the story', because although the (essentially metaphorical) notion of a fictional 'world' is fine up to a point, there are points past that point where it generates paralogisms concerning asserted impossibilities, Holmes' blood-type/war wound (among other things).
If you want a clear case of an almost book-length story to which your proposed hyperintensional notion of truth applies (if any does), try "The Young Visiters" [which I proposed to Brian as a morally deviant world]. It's linked from this post of mine:
http://examinedlife.typepad.com/johnbelle/2004/03/morally_deviant.html
What is holding the world of this story together, to the extent anything is, is the mind of its 9-year old author. We don't really have any trouble reading and understanding - even imagining - the story, simply because we readily pick up on the psychological and moral and literary causes of its being this odd way. The logic of it is pure psycho-logic (if you will) and childish. So we couldn't really model 'the world' of the story effectively, in isolation from the odd tics of the mind of its author. Truth here is for sure hyperintensional. There are many things true 'according to' the story, but it's hard to make sense of what's true 'in it'. That seems right.
One further thought: you say (in your previous post) that reading fiction hyperintensionally (truth-wise) does not feel like pretense, it feels like supposing, i.e. 'let's suppose X for purposes of reductio'. There is an important exception. In a story like "The Young Visiters" there really is a sort of pretense-like experience, because one gets drawn into the mind of a young child who is rather confused. But part of one's mind floats above the pretense, amusedly noting how every other sentence is the basis for some or other reductio ad absurdum. A lot of nonsense fiction works this way: there is some psychological interest, combined with a purely logical interest in seeing it all fall apart so completely.
Oh, and I like your definition of IR as well. I think you are right to imply that, to the extent there is any core phenomenon here, it is not necessarily anything to do with imagination per se.