Armchair knowledge of possible worlds?
I’ve been reading Martin Davies’ ‘The Problem of Armchair Knowledge’. Among other things he discusses the following argument:
(P1) I am thinking that water is wet.
(P2) If I am thinking that water is wet then I am (or have been) embedded in an environment that contains samples of water.
(C) I am (or have been) embedded in an environment that contains samples of water.
The problem is that (P1) looks like it is known from the armchair (I think we can read this as ‘known a priori’). (P2) is the kind of thing that externalists claim to know from the armchair. Grant that for the sake of argument. It now looks like an instance of this argument would take us by a valid inference from things we know from the armchair to something that, even if true, is an empirical matter. Davies discussion of the problem is fascinating and takes in both transmission of warrant and the language of thought hypothesis.
I’ve been thinking about another example:
(P1) I believe that p.
(P2) A necessary condition for believing that p is standing in some relation to some proposition.
(P3) A proposition is a set of possible worlds.
(C1) Propositions exist.
(C2) Possible worlds exist.
I think this has the same structure. (P2) and (P3) are things that, for example, Stalnakerians think they know from the armchair. Some instances of (P1) look like things we know from the armchair. With the addition of a few extra principles, i.e. that if X is a relatum then X exists and that if there are sets of Ys then there are Ys, we get at least one conclusion that looks like it shouldn’t be a case of armchair knowledge (C2). Of course, the more realist we are about possible worlds the worse this will be, because the more they will look like something that shouldn’t be discovered from the armchair.
RSS - Posts
Hi Tom,
I might not be following your argument. Concernng the first (p1), i.e., I am thinking that water is wet, I don’t think this can be a priori if (p2) is true. In worlds where there is no water, I do not have the thought in (p1). So, having the thought is contingent on being in the right sorts of worlds. What’s more, it seems the recieved wisdom that water might not have had any of the criterial properties, so it might not have been wet. There are worlds in wihch the watery stuff is not water, but something else. Just other reasons not to believe that ‘water is wet’ describes an a priori truth.
On the other hand, I guess you could argue that it is contingent apriori that water is wet if it turns out that ‘water’ was introduced by the description ‘the wet stuff over there’.
That may well be right. I think the point is that many people want to maintain that we know a priori what we are thinking. It may be that externalism is a threat to that, as people have indeed argued.
Stephen Yablo discusses this kind of problem in his paper in Boghossian and Peacocke’s ‘New Essays on the A Priori’. You might want to check it out.
Maybe I am being dense here, but frankly how in the world (forgive the terrible pun) are we supposed to have discovered the existence of possible worlds *other than* from the armchair? The Lewisian argument for the existence of possible worlds draws that conclusion from Quinean theses about ontological commitment plus the linguistic data that we believe that there are other ways the world might have been. Even actualists like Stalnaker grant that Lewis’ argument establishes some kind of commitment. Do you mean to suggest that we somehow know of the existence of possible worlds *empirically* or do you simply mean to suggest that there is something problematic about gaining knowledge of the existence of possible worlds via the type of reasoning in the example you sketched?
I have the following naïve worry: The more possible worlds are like distant regions of our actual world, the more puzzling it is to say we have knowledge of them from the armchair. Assuming claims about distant regions of our world are established empirically.