Blog powered by Typepad

« Guide to the philosophy of mind | Main | A Place for Consciousness »

January 14, 2005


kyan gadac

This sentence doesn't seem to make sense:-
Given that this world is actual, it turns out that 'water' refers to H2O, and its Kripkean intension picks out Venus in all possible worlds.
Surely you mean 'water, rather than 'Venus' in the last phrase. In the next sentence you write 'Hesperus' might have referred to something quite different (e.g. XYZ),.
In both cases I fail to see the relationship between water and Venus or XYZ and Hesperus. Surely they are two separate examples that need not be conflated together?


Whoops. And Godel might have been the watery stuff while inventing the zip! Thanks -- I've fixed it now.



Could you tell me whether 2-D semantics has this important implication? Assume for reductio that (1) S5 is the logic of necessity (as is widely--though not universally--believed).

2. I know that H2O is water. (Assumption)
3. Necessarily water is H2O (from premise 1)
4. It is possible that I know that H2O is XYZ.(epistemic possibility)
5. It is possibly true that H2O is XYZ. (from 4 and from knowledge entailing truth)
6. It is possibly necessary that H2O is XYZ.(from premise 5)
7. If it is possibly necessary that H2O is XYZ, then it is necessary that H2O is XYZ. (from S5)
8. It is necessary that H2O is XYZ. (#@! contradiction 3 and 7).

But then either (i) it is an epistemic possiblity that H2O is XYZ and it is not possible that I know that H2O is XYZ (that can't be right) or (ii) knowledge does not entail truth or (iii) the assumption is false and S5 is not the logic of necessity (and that's important to know).


Well, premise (4) looks pretty dubious. FIrst, one might think it isn't epistemically possible in the broad sense that H2O is XYZ -- this looks pretty close to being a priori false (though maybe there's some strange story one can tell). To fix this, though, we can substitute 'water' for 'H2O' in premise 4 of your argument and in the later steps; it looks like your argument will be at least as good that way. But now one should still deny premise 4: although it is epistemically possible (in the sense of not being ruled out a priori) that water is XYZ (and that one know that water is XYZ), it is not metaphysically possible that water is XYZ (so it is not metaphysically possible that one know that water is XYZ). I presume that "possible" and "necessary" are interpreted as metaphysical modality throughout the above argument (clearly metaphysical necessity is required in premise 3, so if epistemic possibility is involved in premise 4, then premises 6 and 7 will have a mixed modality and the S5 argument for premise 7 will fail).

The moral is that one should deny that when it's epistemically possible that such-and-such, it's always metaphysically possible that one know that such-and-such. That's just a straightforward consequence of the thesis that epistemic possibility doesn't entail metaphysical possibility. Given the relevant sense of 'epistemic possibility', this thesis is in turn a consequence of the existence of the necessary a posteriori. So the diagnosis here doesn't turn on anything specific to 2D semantics.


Sorry about that. Yes, right, 'water' should be substituted for 'H2O' throughout. But then it is epistemically possible that p does not entail that I could know (or discover or learn) that p (where 'could' refers to a metaphysical possibility). I suppose I should try get over the fact that that sounds puzzling.

The comments to this entry are closed.