Wednesday, May 24, 2006

Considering as Actual

Jessica Owensby-Sandifer (not to be confused with the other Sandefur!) offers an interesting argument against Chalmers' two-dimensionalism. It's a thoughtful piece which nicely (though unintentionally) illustrates the confusion which can arise from the standard picture of modal space.

The core problem, which Jessica recognizes, is that the standard picture leaves no room for the possibility of any other world being actual, "because for each universe there can be only one actual world." (I assume 'universe' here denotes modal space, or what I sometimes call the 'modalverse'.) She accepts this result and in response proposes a meta-modal space, i.e. a space of modalverses, so that each possible world can be actualized in a different modalverse. I guess this would mean that even though no other world can (in fact) possibly be actual, perhaps it's possible that this could have been possible. It's not true in any possible world of our modalverse. But it is true of another possible world in another possible modalverse. (I find such meta-modal claims deeply fascinating, but of dubious coherence.)

I think that's the wrong response, though. We should instead take the above 'narrow fatalism' to be a reductio of the standard picture. Jessica writes, "Central to the intuitive notions underlying a theory of actuality is the idea that something distinguishes the actual world from being a merely possible world." Indeed. I say that this concrete lump of ours isn't any kind of possible world at all ("merely" or otherwise!). If we think of possible worlds as being maximal properties, then our lump instantiates just one of those properties, which we might call the "actualized world-state". But the lump should not be identified with the property. It could have instantiated a different property, after all (whereas the first property has its identity essentially, and so could not have been the second property). In that case, a different possible world-state would have been actualized. And, intuitively, that's exactly what it means to say that another possible world could have been actual. It means that our concrete lump (*thumps table*) could have turned out a different way. It happens to instantiate world-property w1, but it could have had w2 instead. (Note that these are primitive extra-modal facts which cannot be spelled out using the standard "possible worlds" semantics.)

Now, Jessica wants to suggest a special problem for the two-dimensionalist's notion of "considering a world as actual". She writes:
It my contention that, no matter which theory of actuality one assumes, to consider another world as actual is to make a metamodal claim about the possibility that some other world is actual in another universe or cosmos.

But that can't be right. To consider a world as actual is to invoke the epistemic rather than subjunctive mode of possibility. To consider w2 as actual is to entertain the hypothesis that our lump (not some other one) actually instantiates w2. After all, this might really be the case for all we know a priori.

The lump/property distinction is crucial for making sense of this. For suppose all we could refer to as the 'actual world' is our world-property w1. Then to consider the distinct world w2 "as actual" would be to consider w2 as w1. But that's incoherent. Of course, that's not what anyone takes themselves to be doing when they consider worlds as actual. We're not considering one "world" or property to be another world/property. Rather, we're considering that world to be actual (*bangs table*), i.e. for the property to be actualized by the extra-modal lump.

A simple example might help clarify matters. Let w2 be the Twin Earth world, where XYZ fills the lakes and rivers. We can consider this world as actual, and come to the conclusion that 'water is XYZ' is 1-possible. This is because we're told that XYZ plays the water role in w2, and it's a priori that water actually plays the water role, and hence the indicative conditional "if w2 is actual then XYZ is water" is a priori. No problems.

Even if water is actually H2O, that doesn't mean we were entertaining the hypothesis that H2O is XYZ. That's not how epistemic possibility works: 'H2O' and 'water' are distinct concepts, even if they are actually the same substance. It certainly isn't a priori that "if w2 is actual then XYZ is H2O" -- indeed, that's a priori false! So we need to take care to distinguish the general notion of 'actuality' from the specific world-property which happens to be actualized.

(Another way to respond here would be to invoke the de dicto/de re distinction as it applies to the indexical analysis of 'actual'. For more on this, see my post: P iff actually P.)


1 comment:

  1. Hi Richard,

    I've just read your response today. Thanks; I appreciate your criticism. I think you've brought up some interesting issues; I am still working a formal response.

    The main worry I still have is that no theory of actuality (that I know of, maybe you could help me) has been understood in epistemic terms. Not that we couldn't come up with some theory of actuality that hashed actuality out in terms of our belief that some world is actual, but no one has yet done that yet (and it seems that if Chalmers wants to use the heuristic of 'considering as actual', he needs to provide such a theory). We intuitively don't think of actuality as being translated in terms of what we believe to be the case. So, it would seem that supporting some epistemic understanding of a theory of actuality would be no small task.



Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)