01 December 2010

Making a Davidson Quote more Available, and a Hegel Paper

I've gestured at this quote before (in posts and in conversation), but Google shows that I've never actually typed it all up: it only showed up on Google Books and then in some guy's dissertation in Portuguese. I think it's revealing enough to deserve more prominence than that.

"I am deeply puzzled by McDowell's alternative account.... I do not see how the (propositional) content one takes in can be evidence for a belief, since it does not, in itself, have any subjective probability (if it did, it would be a belief). How can an attitude that assigns no probability to a proposition convey a probability (positive or negative) of, or provide positive evidence for, a belief?" ("Reply to Roger F. Gibson", p.135 in "Donald Davidson: Truth, meaning, and knowledge").

PS: I will finish recapping McDowell week after I get a draft knocked out for a paper I'm writing on Sebastian Roedl for a class on internalism & externalism. Sorry it's taken so long, been a busy couple of months.

Also, "Hegel on Reality as a Modal Notion" is one of the most stimulating things I've read in a while (and I heard Stekeler was a lot of fun in person). Here's some stuff I wrote on Facebook, which I shamelessly recycle to generate blog-content. The original context was my defending Stekeler's interestingness and intelligibility against a cultured critic; I would go back in and add citations to Hegel if I wasn't a hack, but it's all in the Logic chapters that are explicitly on modal topics ("Reality" in Stekeler's paper is Wirklichkeit, not Realitat -- which makes his title seem silly, since "actuality is a modal notion" is not news to anyone):

I thought he really did make Hegel's views on modality clearer. The role of contingency in Hegel's thought seems to me to be pretty much what Stekeler says on p. 21; or at least if this is Hegel's view, then this makes sense of how Hegel can both insist that there really is contingency (and that it would be foolish to try to reduce away all contingency as merely veiled necessity) while also often sounding like a necessitarian, even affirming some form of the principle of sufficient reason: it always makes sense to ask after the cause of any happening.

Answering this causal question takes the form of characterizing the happening as the result of some ground that it follows from in accord with some "generic law", some general claim that relates the sorts of things that the cause and the effect are such that the effect makes sense as being there, given that the cause is there. But no system of "generic laws" says what particular things there are, and to try to make "generic laws" imply the existence of particulars would be to fall into confusion about how causal explanations function (since they do their work by setting particulars in a broader framework, and using the fact that the relations in the framework are understood to make sense of the relations the particulars stand in; if the relations the particulars stood in needed to be understood to make use of the framework, then positioning them within the terms of the framework wouldn't be explanatory, but would merely bring back out what was already known). So it's foolish to try to pretend there's no contingency in nature, just because everything in nature has a causal explanation: Denial of "contingency" in the sense of things that happen for no reason at all, like the atomists' clinamen, doesn't mean that contingency has been shown to be an illusion or a result of our ignorance (which is how Spinoza thought of it - E1P29). Both of these views misunderstood contingency.

This also fits in with another of his criticisms of Spinoza in the "History of Philosophy" lectures: Spinoza has no way of getting attributes into the picture. He fudges that part: God has infinite attributes (by definition: E1D6), but "attribute" is something Spinoza characterizes only by reference to "what the intellect perceives as constituting the essence of substance" (E1D4 -- he needs to appeal to *us*, as intellects, to get attributes into play). He doesn't actually prove that God/Nature has any multiplicity to itself at all, from definitions; he merely seems to, because he's smuggled in our own viewpoint alongside the definitions. Thus Hegel's regarding Spinoza as an "acosmist", someone who has no place for a world at all. And as Spinoza is the paradigm denier of contingency in nature, it seems natural to connect these points in Hegel: denying contingency in nature happens when you've actually lost nature from the picture.

Before reading this paper, Hegel's stuff about contingency had confused me. Now it seems like good sense.

6 comments:

Evan said...

Thanks for posting your thoughts on the Stekeler paper. I was hoping to be there at the workshop, but my son had the nerve to be born that morning and so I was a bit busy. I'll have to get through the paper and then get back to this post.

Daniel Lindquist said...

Congrats on the kid. Hopefully he only keeps you from boring workshops in the future.

"Thoughts" is too strong -- this is just my reaction to one or two pages of the paper, really. There are a lot of other things in it I have opinions about, it's just that this was the one part I'd already scribbled something about.

N. N. said...

"I do not see how the (propositional) content one takes in can be evidence for a belief, since it does not, in itself, have any subjective probability (if it did, it would be a belief). How can an attitude that assigns no probability to a proposition convey a probability (positive or negative) of, or provide positive evidence for, a belief?"

Thanks for this. I'm not sure I understand "subjective probability." Is the objection that experienced propositional content cannot have a truth-value (otherwise it would be a belief), and therefore, that it cannot ground the truth of a belief?

Daniel Lindquist said...

"I'm not sure I understand "subjective probability." Is the objection that experienced propositional content cannot have a truth-value (otherwise it would be a belief), and therefore, that it cannot ground the truth of a belief?"

Something like that.

Davidson (like other decision theorists) has a notion of "credences" in propositions: measures of how much you're willing to bet on the truth of a claim, basically. Subjective probability deals with that sort of thing. Some decision theorists treat beliefs as just subjective probabilities above a certain bar (say >.9, or =1); I'm pretty sure that Davidson isn't among them, but I'd have to check his reply to Isaac Levi in the Schlipp volume to confirm that. Though it's worth noting that even if you do distinguish credence and "full belief", you can treat full beliefs as credences of 1 when it comes to handling things like how much you should be willing to bet that so-and-so has the Ace of Hearts, given other bets you're willing to make.

N. N. said...

Thanks. As I think about Davidson's point I wonder: why isn't this a devestating objection to McDowell's project? While there may be independent reason for thinking experience is already always conceptual, McDowell's reason -- viz., that if it were not, experience could not be a constraint of belief -- appears to be wrong. Has McDowell addressed Davidson on this score?

Daniel Lindquist said...

"As I think about Davidson's point I wonder: why isn't this a devestating objection to McDowell's project?"

It would be if McDowell based our perceptual knowledge on an inferential or quasi-inferential basis. Then our perceptual knowledge would have to be based on something that could serve as a premise in a materially good inference to the knowledge-claim. But this is not McDowell's picture at all: perceptual knowledge is noninferential. Experience gives us noninferential warrant to know that things are thus-and-so. We don't go from the experience to the knowledge based on it by any course of reasoning (either deductive or probabilistic), but by acting on a warrant to know that things are thus-and-so.

This is part of McDowell's disjunctivism: veridical seeings provide a warrant to know such-and-such by how things look to you. Nonveridical seeings do not provide any such warrant. There's not an epistemic status in common between the two, apart from both warranting you to say how things appear to you to be.