Tuesday, March 14, 2006

What concept does 'water' express if the actual world is Dry Earth?

In order to answer this question I propose to make the following distinctions (inspired by distinctions made in Bealer "A priori knowledge and the scope of philosophy" and Chalmers "The Foundations of Two-Dimensional Seamntics").

An expression T is semantically neutral iff which concept T expresses does not depend on how the actual world turns out.

An expression T is semantically sensitive iff which concept T expresses depends on how the actual world turns out.

An expression T is semantically narrow iff we could have still possessed the concept actually expressed by T had the world turned out differently.

An expression T is semantically wide iff we could not have possessed the concept actually expressed by T had the world turned out differently.

An expression T is semantically void iff T does not actually express a concept.


What concept does ‘water’ express if the actual world is Dry Earth?

Clearly, 'water' is semantically sensitive; it expresses the concept twater if the XYZ-world is actual, and it expresses the concept water if the H2O-world is actual, and some other concept if a world where the watery stuff is made up of various different chemicals is actual, etc. But what concept does it express if Dry Earth (a world where all the watery stuff is illusiory) is actual? And under this assumption is 'water' wide, narrow or void?

If the actual world is Dry Earth, could we have possessed the concept that is expressed by ‘water’ (i.e. dwater) had the world turned out differently. This involves two questions :

(i) What concept is actually expressed by ‘water’ (assuming actual world is Dry Earth)?
A. dwaterA = watery-stuff concept
B. dwaterB = empty concept; a concept which necessarily applies to nothing
C. no concept

(ii) Could we have possessed that concept had the world turned out differently?
A. Yes (Narrow)
B. Yes (Narrow)
C. No (Void)

Option A: Assuming the actual world is Dry Earth ‘water’ expresses the concept dwaterA, which applies to all and only watery stuff; a concept we could have possessed even if the world turned out differently (i.e. ‘water’ is semantically narrow).

Option B: Assuming the actual world is Dry Earth ‘water’ expresses the concept dwaterB, an empty concept which necessarily applies to nothing; a concept we could have possessed even if the world turned out differently (i.e. ‘water’ is semantically narrow).

Option C: Assuming the actual world is Dry Earth ‘water’ does not express a concept, so we don’t possess a concept.

Option A seems to be the intuitive answer. (Perhaps this is like 'air').

Option B is also seems ok, though not so intuitive. (Perhaps this is like 'phlogiston')

Option C seems bad, in fact I think it is open to a reductio. (Is this like 'mimsies'?)

Assume for reductio that:

(V) If I am a Dry Earthlian, then ‘water’ is semantically void.

Argument Gamma:
(1g) If ‘water’ is not semantically void, then I am not a Dry Earthian.
(2g) If I am thinking that water is wet, then ‘water’ is not semantically void.
(3g) I am thinking that water is wet.
(4g) Therefore, I am not a Dry Earthian.

All the premises can be know a priori. (1g) is the contrapositive of (V). (2g) is obvious. (3g) by priviliged access. So, (V) is incompatible with priviliged access. I think the content externalist should not and need not accept (V), since it combined with priviliged access entails that we can know a priori that we are not Dry Earthians.

So this leaves options A and B. Notice that both result in 'water' being semantically narrow. So although we can know a priori that 'water' is semantically sensitive we can't know a priori whether 'water' is semantically wide or narrow. This seems to tell against the Mckinsey-style reductios which start from the width of 'water' to knowing that we are not Dry Earthians. If what I have said is right then the first premise cannot not be known a priori, thus no reductio. May others have said this in different words (Brueckner, Tye and Mclaughlin, etc.)

In this way, the Dry Earth example should be a weapon the content externalist weilds against its foes, not vice versa.

No comments: