Today I received the Summer issue of Dialogue: Canadian Philosophical Review / Revue canadienne de philosophie, the journal of the Canadian Philosophical Association. It's a bit of a mixed bag, as it always is, but usually there's something interesting, and so it was this time.
Jefrey Grupp has an article, "Problems with the Platonist Exemplification Tie between Located Entities and an Unlocated Entity," which was certainly interesting. I've read it through twice, and read some passages in it several more times, and my conclusion is that the reason why Platonic realists don't really bother much with the problem is that it's a complete non-problem. The article is something of a sequel to another article by Grupp, which I haven't read. There he argues that direct attachments like the Platonist notion of exemplification between located entities (in this case, particulars) and unlocated entities (in this case, universals) are impossible, "since such an attachment requires the unlcoated entity to be in space, at a spatial place, since the located entity is necessarily spatially located" (p. 494). Thus they would imply entities that are located and unlocated at the same time. Now, it's possible that Grupp has a knock-down argument in the article I haven't read, but I've come across similar claims in philosophy of religion, and they never work. And we have, in fact, good reason to think none will. An 'attachment' like a particular's exemplifying a certain Platonic universal is likely to be sui generis, so arguments like Grupp's really do need to argue that such attachments are conceptually impossible. But how would they be? The concept of the terms wouldn't rule it out, and it's hard to see why the concept of exemplification or attachment itself would rule it out. And this is especially true if the attachment is a relation in a fairly ordinary sense, since you can potentially have relations between any sorts of objects you want (even between a located and a non-located entity).
However, this article is not on this argument, but on a further argument. Grupp's further argument is a response to those who claim that perhaps the exemplification tie is simply an ontological primitive that doesn't cross 'realms'. He argues:
1. Suppose we have a wholly spatially located entity, L, e.g., a lion, and a wholly unlocated platonic universal, S, e.g., sublimity, which L exemplifies.
2. Since L is wholly spatially located, L only exemplifes n-adic properties, such as S, at x and nowhere else, because L is nowhere else but at x.
3. Therefore an exemplification not at x is an exemplification that does not have to do with L.
4. Since S is wholly spatially unlocated, it cannot fail to be spatially unlocated (Grupp, oddly, decides to call this 'being at y'), S only involves a direct attachment to the exemplification tie at y, and nowhere else, because S is nowhere else but at y.
5. Therefore an exemplification not at y is an exemplification that does not have to do with S.
6. If L exemplifies n-adic properties only at x, if S involves a direct attachment to the exemplification tie only at y, and if the exemplification tie does not cross realms, since x is not y, L and S apparently cannot have any dealings with each other: for L to tie to S, S, which is wholly at y, must be at x, and thus be both located and unlocated, or L, which is wholly at x, must be at y, and thus both located and unlocated.
7. Therefore L and S cannot be tied by exemplification.
My first thought: the attachment in this case is not between L's locatedness and S's nonlocatedness, but between L, which necessarily has a physical position, and S, which is not the sort of thing to have a physical position at all. Grupp describes the latter as 'being at y'. But this 'at' is not a physical 'at'. Now, exemplification by its nature would be an asymmetric attachment: L exemplifies S at x, but S does not exemplify L at all; its role in the story is just to be exemplified by L at x. Therefore, it isn't obvious that S and L are parallel in the way Grupp's argument actually requires. L's exemplifying S perhaps necessarily requires the exemplification of S to be where L is; but we have no reason whatsoever to think that S's being exemplified by L requires the exemplification of S to be unlocated. S's being exemplified by L requires only that L exemplify S; it does not necessarily require anything about S. If this is so, however, Grupp cannot force a contradiction.
Further: S may be wholly unlocated in itself, as a Platonic universal, but it does not follow from this that S cannot be located in any way; particularly if you think 'being exemplified by' is one way something can be located somewhere. In this case, S would be wholly unlocated in itself, but located in L by L's exemplification of S.
Grupp's argument is very ambitious; if it worked it would destroy any metaphysical position that posited a relation between located and unlocated entities (Plantingan possible worlds metaphysics, most theisms, Platonic philosophies of mathematics, Cartesian dualism, etc.). And, what is more, it would destroy them all as a result of identifying a fairly basic contradiction at their heart. We have, however, no reason to think it works; and this is not surprising (what would be surprising would be if all these positions turned out to be subject to such an easily identified contradiction). In fact, arguments like Grupp's have been considered, and rejected (rightly, I think), in discussions of divine omnipresence, on the grounds that location itself is a derivative property depending on more fundamental properties. It therefore cannot conflict with any attachments that are involved in these properties (and while we need not assume that one of these attachments is exemplification, Platonic realists certainly would).
There's also an interesting article by Manuela Ungureanu, called "Reading the Minds of Others: Radical Interpretation and the Empirical Study of Childhood Cognitive Development", defending Davidson from a common criticism. The Davidson stuff is blah, but the cognitive stuff is very interesting.
And Larmer continues his debate with Overall on miracles. This has to be one of the most pointless disputes I've ever seen; I don't know why they keep it up. I don't agree with a lot of what Larmer has said in it; but Overall's last contribution was bad to the point of being an example of what one should not do in a philosophical discussion, consisting as it did almost entirely of things like rhetorical questions (yes, for several pages most of the sentences where either direct rhetorical questions or the sort of indicative sentence that really is more like a rhetorical question than a genuine argument). Very disappointing; and very undergraduate. Her original paper was much better, although the argument in it is just silly: she's arguing that if there are miracles, this would prove that God does not exist. Yes, you read that correctly. It turns out just to be a bizarre variation of the problem of evil.