Supervenience and Necessity
- Posted by Andrew Bailey on Sunday, December 10, 2006 at 1:11 AM
| 
0 Comments |
It's easy to wonder if all global supervenience theses are necessarily true if true at all. And some of them are. But not all. Consider this formulation of global moral supervenience:
P. For any world w, if w and the actual world are exactly alike descriptively, w and the actual world are exactly alike ethically.
It is a contingent matter which world satisfies the definite description, "the actual world." So whether P is true or not depends on which world turns out to be the actual world. Not so if we had used a rigid designator to pick out the actual world (say, "alpha"). For that thesis would be necessarily true if true. It would be true no matter which world turned out to be the actual world. Consider this formulation of supervenience:
Q. For all w and w*, if w and w* are exactly alike descriptively then they are exactly alike ethically.
w and w* are names, rigid designators which pick out their respective worlds no matter which world turns out to be actual. Q is thus necessarily true if true. (Q is Frank Jackson's formulation of global moral supervenience, by the way).
Here's a fun thing to consider. One can make supervenience theses like P even "more contingent" by playing with the quantification over worlds. Here is a weaker ("more contingent") version of P:
R. For any world w where the actual laws of nature obtain, if w and the actual world are exactly alike descriptively, w and the actual world are exactly alike ethically.
R allows that there are worlds exactly alike descriptively (but not ethically) with the actual world. These are far away worlds, though, worlds that don't share our laws of nature. P quantifies over all metaphysically possible worlds, while R quantifies only over nomologically possible worlds. Only on some conceptions of laws will this work, I note (where the laws of nature at a world aren't bundled into its description).
