Friday, September 28, 2007

Another Post, Another Paper

This time it's all about explanation. This is the paper I'm planning to give to the Aristotelian Society in November, so quite a different sort of intended audience from the Compass paper, but I'm still aiming for accessibility to philosophers in general. The paper's called Romeo, René and the Reasons Why, though you'll have to read it to find out why.

The paper suggests (tentatively) a kind of functionalism about explanation. The idea is that our concept of explanation is a role concept closely connected to the answering of why-questions, i.e. the provision of information about what is a (non-inferential) consequence of what. The various realizers of the explanation role include causal realizers, nomic realizers and more. So the pragmatic account folks, the causal theorists, the covering-law theorists and so on each got part of the story right. Everybody has won and all must have prizes.

Comments, as always, encouraged.

Update: I have revamped this paper and the link now takes you to the latest draft.


Anonymous said...

Dear Carrie,

It seems one important distinction is missed here. For any explanandum e there are various propositions that, if true, would explain e. (It shouldn’t matter what stand one takes on what explanations are; I’ll use propositions.) Most of these propositions will be false. That rolling balls stop would be satisfactorily explained by the proposition that demons push against them, if it were true. So this a potential, but not an actual explanation.

Once the distinction is made between actual and potential explanations some of the dimensions of variation seem strange. You write: ‘[The] conception [that has it that whether p explains q is a matter of whether p helps us feel like we understand q] seems to be in play when we say there are lots of good explanations of quantum decoherence but only one of them can be true.’ But that seems wrong: saying this needn’t have anything to do with a subjective account of what makes for explanatoriness, it is just to say that there are various potential explanations of which only one is actual.

More importantly, inference to the best explanation (IBE) must be inference to the best potential explanation, since, as you note, the point of IBE is that we don’t to begin with know which is actual. But surely we infer what we judge is the likeliest of the potential explanations; and this isn’t a matter of whether we feel the potential explanation explains.

The actual/potential distinction cuts across the act/fact/sentence distinction. Surely the right conclusion to take from the context-sensitivity of the relative importance of the explanatory virtues you list is not that whether p is a potential explanation of e is context-sensitive, not that whether p is the actual explanation of p is context-sensitive, but merely that whether the description of an actual explanation is successful is context-sensitive. So being a best explanation cannot be context-sensitive in this sense.


Carrie Jenkins said...

Thanks for this Marco - it's very helpful and I'd like to acknowledge your help when I write the final version of the paper. If you want I can just list you as 'Marco', but if you want to tell me your full name for this purpose please send me an email: carrie *-dot-* jenkins *-at-* nottingham *-dot-* ac *-dot-* uk. Cheers!