Dispositions, intrinsicality, and the problem of fit

In chapter 3 of The Powers Metaphysic, Neil Williams presents a nice problem for dispositionalists: the "problem of fit".

Dispositionalists hold that there are fundamental dispositional properties. Now consider a particular rock and a particular glass. The rock might have a disposition to break the glass when thrown at it. And the glass might have a disposition to survive impact of the rock. These dispositions are incompatible: if the rock is disposed to break the glass, the glass can't be disposed to survive the impact. But if dispositions are fundamental, then what prevents the rock and the glass from having the incompatible dispositions? The dispositionalist seems to require a mysterious ban on recombination.

In response, Williams suggests that dispositions should be individuated holistically: there is no fundamental disposition to survive impact of a rock; rather, every fundamental disposition contains a "blueprint" settling what would happen under all constellations of all fundamental properties. The problematic glass-rock situation can then be seen as impossible because the envisaged properties would contain incompatible blueprints.

I have some trouble understanding the blueprint metaphor, but the proposal resembles a more familiar picture (to me) according to which fundamental properties are individuated by their place in the world's nomic structure. But on that view, it is tempting to think that what's really fundamental are not the individual nodes in the structure, but the ramsified structure itself. And that goes against the "localist" intuition that how things behave is ultimately grounded in their local, intrinsic properties.

How else could a dispositionalist respond?

A simpler response, I think, is to say that the glass-rock situation is impossible simply in virtue of the essence of the relevant dispositions. If the rock has a fundamental disposition to smash the glass, then the rock has a property whose essence implies being such that the glass would smash when struck. If the glass has a fundamental disposition to survive being struck, then it has a property whose essence implies being such that the glass would survive when struck. These are contrary implications. So the properties are incompatible, not because of some brute ban on recombination, or because the properties contain a blueprint of the world's total nomic structure, but because their (simple) essences are incompatible with one another.

I still think this makes the dispositions look extrinsic. But I never understood how fundamental dispositions are supposed to be intrinsic anyway.

Suppose some thing x has a fundamental, determinate disposition to M under condition C. Presumably it follows that either C does not obtain or M obtains. But both C and M are typically extrinsic. C usually concerns the presence of some external stimulus, and M usually involves the future, and/or the behaviour of things other than x. So the disposition entails a non-trivial disjunction (¬CvM) of highly extrinsic properties. Then how can it be intrinsic?

The problem is exacerbated by the problem of fit. If presence of one disposition in the rock precludes the presence of another disposition in the glass, in what sense can these dispositions be regarded as intrinsic?

Dispositionalists sometimes appeal to a grounding conception of intrinsicality. For example, according to Rosen 2010, a property is intrinsic iff its instantiation by an individual is not grounded in facts about distinct individuals. On this conception, every fundamental property is trivially intrinsic, given that its instantiations are not grounded in anything else.

If that's the relevant concept of intrinsicality, then the dispositionalist's claim that her fundamental dispositional properties are intrinsic is a tautology. Saying that these properties are intrinsic adds nothing to the claim that they are fundamental. It is also, I would complain, highly misleading, since the supposedly fundamental dispositions fail just about every intuitive test for intrinsicality. (On the grounding conception, even the most obviously extrinsic property, like being an uncle, can be called intrinsic if it is declared fundamental.)


No comments yet.

Add a comment

Please leave these fields blank (spam trap):

No HTML please.
You can edit this comment until 30 minutes after posting.