At the recent Bellingham conference, Ted Sider (in discussion of a paper by Kadri Vihvelin) made what I thought was a very good point about dispositions and conditionals. What follows are largely reflections on how Ted’s point affects debates about free will. I don’t think many of my conclusions here are original – what I say ends up being pretty close to what Neil Levy says in “Frankfurt Finked”:http://au.geocities.com/neil_levy/Documents/articles/Frankfurt-finked.pdf (PDF), but perhaps the way I get there will be interestingly different. (And it is an excuse to display my Austinian tendencies in some detail.)

Consider the familiar example of the glass liked by a powerful sorcerer. The glass is a duplicate of my glass that will break if struck. But this glass won’t break if it is struck. The sorcerer will anticipate the strike and at that moment change the intrinsic structure of the glass so that it can handle being struck. Intuitively, the glass is still fragile, but it won’t break if it is struck. So the following identity is false.

* Being fragile = would break if struck

Now here is Ted’s point. From the example we know that not both of the following identities can be true.

* Being fragile = being disposed to break if struck
* Being disposed to break if struck = would break if struck

But the example alone *doesn’t* tell us which of the identities is false, just that one or the other is. Most of the recent literature on dispositions has focussed on the second identity as what is wrong. But once Ted raised it, it seemed to me that we should think again about the first identity.

Now I suspect that once we do that evaluation, we’ll find that we do probably want to keep the identification of fragility with the disposition to break if struck. (Perhaps we’ll want to replace ‘struck’ with something more precise.) But it really isn’t clear, and it certainly isn’t clear that every case that is apparently similar should be treated the same way.

For instance, consider a theorist who follows Ryle in thinking that a lot of mental predicates (like ‘clever’) denote the having of bundles of abilities, and these abilities are in turn to be analysed as dispositions. Thinking hard about finkish cases might well generate an interesting objection to such a theorist. Consider the following, somewhat long, case.

Alex and Cameron are school children. Alex is very good at long multiplication. Ask her to multiply any three digit numbers, and she’ll (after a short while) get the right answers for the right reasons. Alex is clever, and her cleverness is in part constituted (as philosophers might ordinarily put it) by her ability to do long multiplication. In fact this fact about her and multiplication is crucial. If she weren’t able to multiply like this, her other intellectual skills would not be sufficient for her to count as clever, though she’d still be rather smart.

Cameron is an intrinsic duplicate of Alex. But if you ask Cameron to multiply three digit (or larger) numbers she’ll get it wrong. That’s because a nearby sorcerer, who has taken a dislike to Cameron, will muddle Cameron’s brain whenever she tries to multiply three digit numbers, and ensure that she gets the wrong answers. Note that the sorcerer won’t just make Cameron say the wrong thing when answering the question. She’ll temporarily alter Cameron’s brain so she makes simple arithmetic errors along the way, like multiplying 8 by 7 and getting 54. Cameron makes no such errors when multiplying 1 or 2 digit numbers. Note also that Cameron is not right now doing multiplication, she’s playing football, so her brain is not muddled. (The sorcerer, for reasons best kept secret, always restores Cameron’s brain after he muddles it.)

All of the following sentences are true in the story. (I’ll stipulate that in context a multiplication question is large iff it involves numbers with three or more digits, and large numbers are numbers greater than 99.)

(1a) Alex is clever.
(2a) Alex is good at multiplying large numbers.
(3a) Alex is disposed to answer large multiplication questions correctly.
(4a) Alex has the ability to answer large multiplication questions correctly.
(5a) Alex can answer large multiplication questions correctly.
(6a) If Alex were asked a large multiplication question, she’d answer correctly.

What about the following?

(1c) Cameron is clever.
(2c) Cameron is good at multiplying large numbers.
(3c) Cameron is disposed to answer large multiplication questions correctly.
(4c) Cameron has the ability to answer large multiplication questions correctly.
(5c) Cameron can answer large multiplication questions correctly.
(6c) If Cameron were asked a large multiplication question, she’d answer correctly.

I’m pretty sure (1c) is true, and (6c) is false. And I think, though I’m not sure, that as we go down the list from (1c) to (6c), the claims get less plausible. If forced, I’d say that (3c) is true and (4c) is false, but I’m fairly tentative about these answers.

The upshot of all this is that I think it is a mistake to analyse (or conceptualise if like Ryle one doesn’t want to be a reductivist) cleverness in terms of dispositions or abilities. Cameron is clever, but it is far from clear that she has the dispositions or abilities that we’d ordinarily say constitute her cleverness. This is not to say that the cleverness is not connected to those dispositions or abilities. In fact the cleverness might just be the categorical basis of the dispositions and abilities. But it is not identical to them, or (as I misleadingly put it in setting out the problem) constituted by them.