Sunday, May 04, 2008

Believing in

Provoked by some remarks of Tim Crane's at the last meeting (where he was drawing on a paper by Zoltan Szabo), I gave a talk on Wednesday at the local Serious Metaphysics Group on whether a distinction between "believing in Fs"/"believing that there are Fs" can do any serious work for us. My answer was "no". This was dashed off the previous Saturday afternoon, and I'm outside my comfort zone here. But the talk survived the discussion: so for what it is worth, here it is.

2 comments:

Zoltan Szabo said...

Thank you for the comments on my paper. I will think more about the criticism, but for now I just wanted to react quickly to four points:

1. The parity claim

This says that when F is a kind term then answers to the questions 'Does A believe that there are Fs?' and 'Does A believe in Fs?' stand or fall together.

One reason why I am inclined to reject the parity claim is that I think we should respect people's sincere assertions in ascribing beliefs to them. So, when someone who is linguistically competent but fundamentally misguided about what Fs are sincerely says 'There are Fs' I think we should ascribe to them the belief that there are Fs. But since there is no similar disquotational principle at stake, I think we are on much more solid ground if we refuse to ascribe to him a belief in Fs.

There is another reason why I think the parity claim is implausible. This is the fact that many propositional attitude verbs can take non-clausal complements, and when they do the results are arguably not paraphrasebale through an existential proposition. For example, to imagine a tree is not to imagine that a tree exists, to recognize some trees is not to recognize that some trees exist, to think of trees is not to think that trees exist, etc. My claim that to believe in trees is not the same as believing that trees exist fits the pattern.

2. Horatio

The Horatio case is irrelevant to the parity claim because 'things Horatio does not believe in' is not a kind term. I myself very much doubt that kind terms and other bare plurals behave differently in the relevant contexts (i.e. that there is a significant semantic difference between 'dogs' and 'brown dogs', or 'brown dogs I like' when it comes to how they feature in existential sentences.) In any case, my claim was that (i) ‘Horatio believes in things he does not believe in’ has a false reading, and (ii) this cannot be captured by construing the sentence relationally. I still think these claims are correct. On the other hand, I agree that this might be but a "superficial quirk of certain double intentional contexts."

3. Qua

Could we replace my (admittedly somewhat stipulative) use of 'A does not believes in Fs' with 'A does not believe in Fs as such'? Perhaps. I know of three proposals about the semantics of as-phrases in English (Landman, Szabo, and Asher). Unfortunately, none of them could handle this particular locution. I suspect that this use of 'as such' is also somewhat stipulative.

4. Numbers and phlogiston

Suppose that our ordinary conception of numbers is deeply platonic and that there is nothing that comes even close to fitting that conception. My proposal was that we could still believe that there are numbers in such a case, even if we suspend our belief in numbers. Setting aside the particular vocabulary, the idea is that the existential proposition that there are numbers can be true even if nothing even remotely fits our ordinary conception of numbers.

This is said to be wrong because (unlike in the sea-serpent case) "the mistake is just too big for someone who accepts the nominalist arguments still to be able to say that there are numbers but not as such. It is the same as with the phlogiston case: the phlogiston theorists mistake is too big for us to be able usefully to say that there is phlogiston but not as such."

I think it is not the same. The arguments against phlogiston are arguments against the truth of phlogiston theory. By contrast, nominalist arguments against numbers are not arguments against the truth of number theory. The part of our conception of numbers that actually plays a role in arithmetic is encoded in the second-order Peano axioms and the nominalist has nothing of substance to say against those axioms. What the nominalist argues against is the part of our ordinary conception of numbers that number theory is completely silent about - the part that entails that the number 2 is not on the Moon or that the number 2 was not the cause of the mortgage crisis in the US. This is why it seems to me entirely reasonable to think that there are numbers but they don't fit our ordinary conception of them. To say the same about phlogiston would be entirely unmotivated.

Peter Smith said...

Many, many thanks for the comments, Zoltan. Obviously (4) is the key point, and I'll think about this and respond in a separate post. So just for the moment ...

Re (1), I just report that the group of (mostly) graduate students who were discussing this also took the parity claim to be intuitively appealing. (And that, of course didn't depend on anyone being confused about the behaviour of non-clausal complements in other contexts: what was in question was the particular claim about particular contexts that you state.) Now of course, this observation does no more than set up the thought that the parity claim is prima facie attractive and needs to be argued against if it is to be rejected. (I don't see, though, that either of your remarks under (1) here give us the needed argument.)

Re (2): I just report again that most of the grad students in the discussion shared my counter-intuitions about the Horatio examples after I gave the other couple of variants sentences as objects of comparison. The case seems too murky and intuitions too easily steered to rest much on.

Re (3): Actually, I don't mind the thought that the intended use of "as such" locution I was using was a bit stipulative (though I didn't have to work very hard to make myself clear enough). If you like, we are both using terminology a bit stipulatively to draw some distinctions. But the question remains what is the useful distinction to draw. You have it as a distinction primarily between two types of doxastic attitudes (believing that and believing in). I have it as a distinction cutting across that, marking cases where the content ascription is unqualified and when it is more or less hedged.

But as I say, (4) is really -- I'm sure we agree -- the important issue, and we'll get back to that!