18 September 2007

The Causal Nexus: a short note on McDowell and Bilgrami

In his reply to Bilgrami's essay in McDowell and His Critics, McDowell notes Bilgrami's suggestion that we should jettison the idea that intentional idioms are "part of the causal nexus that natural science investigates". McDowell says that he is sympathetic to the idea, but objects on the grounds that intentional idioms are (all parties are agreed) part of an explanatory nexus of some sort, and Davidson has made it eminently plausible to regard this sort of explanation as causal: "That we believe and desire what we do makes a difference to what we do" as McDowell puts it. (p. 67)

At this point I wanted to ask: Why should we hang on to the notion that the natural sciences investigate "the causal nexus"? Certainly natural science is often concerned with explaining relations of various sorts causally, and their being concerned with certain sorts of explanations seems to characterize the natural sciences as something distinctive from other varieties of human inquiry. (With the rise of modern science comes the view of nature as a "realm of law", as McDowell puts it in Mind and World, and this way of viewing things has been quite productive in expanding the ways in which we can comprehend and cope with various facets of the world.) But it simply seems false that what is distinctive about the natural sciences is that they are concerned with causal explanation as such, since our use of intentional idioms is not a part of any natural science, yet it is still (often) concerned with explaining various happenings causally. So why should we want to describe the object of the natural sciences as "the causal nexus"? It seems that some regions of "the nexus" are ones which the natural sciences don't devote attention to, but that this is not a failing on the part of the natural sciences. (That "belief that p" is not mentioned in physics, or in chemistry, or in biology, etc. does not indicate that these sciences are in need of revision.)

Then it occurs to me that I should keep reading, and McDowell makes much the same point: Intentional explanations belong to "a causal nexus in their own right" and we should jettison the idea that natural science gives us something we can regard as the causal nexus. (I do not quite like the talk of multiple nexuses; to my ear it suggests that a single nexus, such as those of our intentional idioms or that which natural science investigates, is itself a "complete nexus", yet there are other nexuses which one might be concerned with instead. I should think that this way of talking naturally leaves one susceptible to the worry that a single nexus would already be sufficient to determine "how things hang together", and so there is no room for another nexus to have any say in what follows from what. So if the natural-scientific causal nexus determines how my arms move (via my being an instance of various anatomical structures), then any other causal nexus (say, that of intentional idioms) which wanted to explain why I raise my arm will have to either be "in harmony" with the natural-scientific idiom (perhaps via a psychophysical parallelism, or via strong supervenience, or via the reduction of one type of explanation to the other), or will have to be explaining some other happenings than the ones treated by the natural-scientific causal nexus. But it is surely a confusion to think that the arm I raise when I raise it "because I wanted to stretch my shoulder" and the arm which can be studied by medicine are two different arms; it is because the arm which my doctor studies is the one which I use to satisfy my desires that I visit the doctor in the first place. And the other form of the worry is just the sorts of naturalism which Bilgrami and McDowell want to dissuade us from falling prey to. I think that these sorts of worries are not so forceful if we drop the plural from "nexuses"; then we are only left with the thought that a single happening might have multiple causes, which claim should not trouble anyone. "A causal nexus" appears to have a finality, a completeness, that "a cause" or "a type of cause" doesn't.)

I do not know why there should be a problem with talking about "the causal nexus" tout court. Talk of "all causal relations, in general" doesn't commit us to thinking that there is one sort of explanation which can, in principle, exhaustively treat of all causal relations. And sometimes we want to mention "causal relations" in an arbitrarily-large aggregate: "the causal nexus" seems to me a harmless locution for this purpose.

What counts as a (useful) explanation depends on what we are trying to do at a given moment, and there is no saying in advance what all we might want to do, and so no saying in advance what can count as an explanation or not. But if we cannot list ahead of time all possible explanations that might account for some event, then it would seem that we also can't say what all causal relations involve that event, since some of the explanations we don't think of (because we are not interested in doing anything which would make them useful explanations) might be ones which we would regard as causal if we thought of them at all.

An aside: I notice on page 70 that McDowell notes that his handling of normative talk in a discussion of what is special to free agents is something for which "there is no reason not to classify as metaphysical." This is contrasted to Bilgrami's expansion of Strawson's "Freedom and Resentment" as a way to avoid metaphysics. Given McDowell's oft-mentioned "quietism", this surprised me. I suppose I may have simply assumed that McDowell regarded "metaphysics" as a dirty word (like Wittgenstein did, or like Conant still does), with even his use of Kant and Hegel meaning to keep us from falling into "metaphysical confusions." If McDowell does not actually have any hostility to "metaphysics" as such, then this makes it a good bit easier to see how he might be handling Kant & Hegel's magisterial transcendental works for Wittgensteinian ends: A little metaphysics is called for to dissolve a great deal of befuddlement. A good metaphysical description of, say, perception, might leave us without the temptation to do the sort of metaphysics Wittgenstein condemned, the sort which follows from being bewitched by our own language. "Metaphysics" in the positive sense would then just be one more philosophical method for treating our conceptual maladies.

edit: Kenneth Westphal has an essay on Kant, Hegel, and McDowell that runs something like this: Westphal talks about Kant and Hegel for about ten or fifteen pages, summarizing some of the transcendental arguments Westphal holds them to be making in their major works, mentions McDowell's gestures to several of the things Kant and Hegel were on about, and then goes "McDowell thinks you can play around with these ideas and be a therapeutic quietist? Really? You really think you can adopt a Kantian/Hegelian notion of spontaneity without getting knee-deep in transcendental arguments? I have my doubts. Admit you are not a quietist, McDowell, and join me in my transcendental pleasure-dome!" (I summarize. Westphal does not mention pleasure-domes.)

Westphal's worries strike me as being pretty good ones. Someone needs to point out that "Hegel and Quietism" is an... odd... pairing. One doesn't usually think of "Wissenschaft der Logik" as being opposed to grandiose speculative metaphysical constructions. I think that Westphal's concerns can be answered, but it certainly does merit saying: "Therapeutic Hegelianism" is a pretty odd idea, on the face of it.

(I am inclined to say that one can make "Therapeutic Hegelianism" look perfectly correct either by getting a better view of Hegelianism, or by better appreciating what all can be necessary for philosophical therapy. But that can wait for a later post.)

On a meta-note: I have noticed this problem before: One often gets inspired to write something as a response to something one is reading when one is in the middle of reading it. But if one stops to begin composition immediately, one risks simply predicting how the rest of the work will fall out. But if one finishes reading before one begins composition, one risks losing the sense of inspiration that one had, one might simply no longer have a desire to write after doing the full sum of reading.

On a more minor point, virtus dormativa explanations are not always pointless. If I tell you that opium has made N sleepy because opium possesses a soporific power, then this lets you rule out some possible ideas for why opium made N sleepy: Say, that N has an allergy to opium, or that N's sleepiness and N's consumption of opium are merely coincidentally linked. (This point I take from Paul Raymont's excellent dissertation, (PDF) which I linked to the other day.)

A point which does not merit its own post: Dennett's near-manic focus on prediction has the advantage of making it practically impossible for him to find it plausible that our intentional states are causally inert. It is just the fact that they do have a great deal of sway in determining how things will be that makes them useful to note, on Dennett's picture of what is in view from the "intentional stance."

1 comment:

A said...

Re: your meta-note

This has been the source of many of my margin notes and then, one paragraph later, margin note-erasings. Moreover, the problem created by my tendency to write arguments quickly before reading further was compounded by a distaste for erasing my comments (I preferred to write something like "mis-reading!" or "wrong!", so I could have a mini-genealogy of my developing thoughts, throughout a given work).

Now I am slower to write and quicker to erase.