logicandlanguage.net

You'd rather he starved himself to death?

Nothing wrong with a little fornication and adultery either.

You know -- I'm going from memory here, but I'm pretty sure about this -- there was a letter from Tarski to Morton White on the analytic/synthetic distinction published in the Journal of Philosophy some time in the mid-90s (1995 perhaps?). I know he said in it that he'd come around to Quine's position, but don't recall how specific he was about the reasons why. But since you're curious, I have a feeling that's the only published source on this.

Ha, if your just stuck on why he believed in Quine's 'position' your one step ahead of me. I don't even know what he was believing in when he said he accepted Quine's position.

I mean Quine quite explicitly was not arguing that we can't consistantly divide up sentences into analytic and synthetic categories. He instead said he was showing there was no basis or significance in this distinction. Yet this is horribly vague. I can't figure out what Quine meant by this statement much less what Tarski was believing.

I mean it seems quite wrong to think Quine was arguing for the empirical proposition that people didn't actually differentiate between analytic/synthetic statements or that this diffrentiation was just the distinction between core and peripheral beliefs. Furthermore he also explicitly was not just arguing that the analytic/synthetic distinction was vague. Putnam's response that Quine was right in spirit and very few sentences were clearly in either category was still clearly a refutation of Quine's claim.

I personally tend to think the best interpratation we can give of Quine's claim is something like this. Supposing the meaning of sentences is determined soley by physical experimentation (and without reference to qualatative experiences) then there is no fact of the matter about what sentences are analytic and which are synthetic. That is if we imagine only looking at ourselves from the outside, just from a description of all the atoms in human bodies, we could alter which sentences are analytic by exchanging which patterns of brain activity correspond to what content. So if we allow the pattern of brain activity which in us represents seeing red to represent seeing a rabbit in someone else's brain we can't uniquely specify analytic and synthetic sentences.

I think there is a simple response to this line. Simply insist that there is some fixed mapping between brain function and intension specified by the experiences these brain states create and that we can induct on this mapping just like any other physical law. However, I'm not entierly sure if this is really what Quine meant and it seems pretty clear that many people interpret his point differently so who knows what Tarski thought he was accepting.