Lewis as the philosophically insane.

I’ve been reading van Iwagen lately and I now realize why he is so revered. A topic which has come up several times in his articles is the role propositions play in ones ontology. He critiques Lewis’ possible world assay almost entirely on the fact that his Lewis’ reductionism does violence to what we thought we mean when we express modal propositions. Consider the following:

(1) It’s possible that that JFK died of natural causes.

On Lewis’ account, the proposition expressed in (1) really means this:

(1)’ There is a world spatiotemporally unrelated to ours where JFK died of natural causes.

The ramifications of such a reduction are manifest, for it implies there really is world, full of atoms and space and even has JFK as one of its members that really exists (in the sense that it’s a concrete world), but we are spatiotemporally not related to that world. Other queer entailments include the following: The word ‘actual’ functions as an indexical: When we express the proposition ‘It’s actually the case that JFK was assassinated’ the word ‘actually’ is referring to the world we are spatiotemporally related to. When the people in W2 say ‘it’s actually the case JFK died of natural causes’, by ‘actually’ they are pointing out the world (W2) where they are spatiotemporally related. Does this mean that proposition ‘JFK died of natural causes’ (as well as its contrary) is both true and false? No, because the referent of that proposition is ambiguous- for it does not designate a specific world. HA!
The craziest implication of Lewis’ view, I think, is that there is not just one unique JFK, but possibly millions. Consider the following propositions:

(2) JFK never married.
(3) JFK lived until 1989.
(4) JFK was a Soviet spy.

(2)-(4) are all true in some really existing world and the singular term ‘JFK’ in each proposition picks out the JFK in that world where the proposition is true. What’s worse, every possible proposition that includes the singular term ‘JFK’ picks out a really existing JFK. Ergo, there is an uncountable (if not an infinite) number of JFKs currently in existence.

Are you kidding? What’s the fruit of a reductive analysis if it comes at such a cost? And back to van Inwagen’s point (and Plantinga, and Kripke, et al.), the proposition expressed in the sentence ‘It’s possible JFK died of natural causes’ I am predicating a modal term on a proposition, and to say what I really mean by this is the nonmodal proposition ‘there is world spatiotemporally unrelated to us where JFK died of natural causes’ is to change subjects.