McDowell on the Ideality of Space and Time

Reading "Hegel's Idealism as a Radicalization of Kant". For the most part, I like it.

I'm not sure how McDowell's argument that the Deduction fails is supposed to work. McDowell claims that sliding from "spatio-temporally-formed sensible intuitions" (which the Aesthetic is to have shown as being a condition on how objects can be given to us) to "sensible intuitions which are formed (in general)" is supposed to undercut Kant's claim for the objective validity of the categories. I'm not sure how this objection is supposed to work. As Kant presents it (in McDowell's interpretation), that our sensible intuitions have some form or other is a prerequisite for objects being presented to us. That the form our intuitions take is spatiotemporal is supposed to be a problem:

If we allow ourselves — as Kant encourages — to play with the idea of sensibilities formed otherwise than ours, we can suppose they would generate formal intuitions that reflect their specific ways of being formed as space and time reflect ours. And we can perhaps imagine that beings endowed with such sensibilities might construct their own transcendental deductions of the objective validity of categorial thinking, each exploiting — as Kant’s Deduction does — the thought that the unity of the mode in which empirical intuition is given in their sensibility is no other than that which the category prescribes to the manifold of a given intuition in general (compare B144-5). But such a fancy is no help in our task of vindicating the objective validity of categorial thinking for ourselves. What that requires is averting the threat that categorial requirements are merely subjectively imposed on objects as they are given to our senses. That was the threat that indeed seemed to be averted when Kant noted that the unity of the formal intuitions, space and time, is itself a case of the objective unity of apperception....

If there are conditions for it to be knowable by us how things are, it should be a truism that things are knowable by us only in so far as they conform to those conditions. And Kant wants it to look as if any hankering after an objectivity that goes beyond pertaining to things as they are given to our senses is a hankering after what could only be a mirage, a violation of that truism. But it is equally truistic that a condition for things to be knowable by us must be a condition for a possibility of our knowing how things are. And if some putative general form for cases of how things are is represented as a mere reflection of a fact about us, as the spatial and temporal organization of the world we experience is by transcendental idealism, that makes it impossible to see the relevant fact about us as grounding a condition for our knowing that things (really) are some way or other within that form. Transcendental idealism ensures that Kant cannot succeed in depicting the formedness of our sensibility as the source of a condition for things to be knowable by us.

I don't see how this objection is supposed to work at all. It seems to me that even if Kant does simply leave the spatio-temporal nature of the forms of our intuition as a "brute fact", this doesn't preclude those intuitions from reaching all the way out to the world; it merely limits the objects they can reach out to to those which are spatio-temporally extended. If there are non-spatio-temporal objects, or if the objects we know of have non-spatio-temporal properties, then these "can be nothing for us", but I don't see how this hurts the Deduction. That there are conditions under which alone we can know objects entails that these conditions must obtain in the case of any objects we know. I don't see why it should be a problem if there are (possibly) some objects which we don't know to which those conditions don't apply -- which we can't know, though they might be perfectly real for all that. If the Deduction goes through, then we can't apply any of the schematized categories to such (hypothetical) unknowable objects, so there is no interesting sense in which we can imagine them to be interacting with objects which we do know, or in which we can imagine them as they might be in themselves. It would mean that the spatio-temporal objects which we do know might have non-spatio-temporal properties which we cannot know, but I'm not seeing why this is an objectionable sort of ignorance. Beings with other forms of intuition than ours might (hypothetically) be able to notice some things about the objects we know which we are unable to know. But we can still notice plenty about those objects. (And indeed we do know plenty about them, as shown in the "Refutation of Idealism".) Kant spoils this by making space and time transcendentally ideal (along with all the objects in it, and our knowledge of them), but that doesn't seem to be quite where McDowell is lodging his complaint (at least in some parts of the paper).

My interpretation also strikes me as more likely to be one Kant held (or at least aspired to); Kant claims the reason we are unable to imagine a "noumenal" causation (such as freedom) is because we lack a schematized category of causation other than the spatio-temporal-sensible one we have, and a locus of freedom (such as the noumenal Will) cannot be spatio-temporal in nature (for to exist in space and time is to be subject to the analogies of experience, and thus to nature's laws, and a free will cannot (Kant thinks) be bound by laws it did not impose upon itself). Kant hypothesizes that the Deity might have an "intellectual intuition" (and so a form of intuition unlike our own spatio-temporal-sensible one) which might enable it to know (and so judge) the free will. As for us, we can only apply the logical form of relational judgement (the idea of one thing standing in some relation to another another in general) to non-spatio-temporal objects, and this gives us only problematic knowledge: Something analogous to causation might apply to noumenal objects. Kant holds that there are objects which we cannot know, because we cannot intuit them. McDowell seems to view this as a failure on Kant's part; I don't see how it wasn't something Kant positively intended to be a part of his system. Kant wanted an empirical realism, but not one that extended the conditions of our knowledge of objects to condition of the existence of objects in general -- not a realism sans term. "I denied reason to make room for faith" and all that.

I have a related quibble with this passage:

According to transcendental idealism, our capacities to know things reach only so far, and beyond that boundary there is something we cannot know — whether things themselves are really spatially and temporally organized. And the thought that we cannot know whether things themselves are really spatially and temporally organized undermines the possibility of recognizing as knowledge the supposed phenomenal knowledge, that things are spatially and temporally organized, that we are supposed to be able to achieve within the boundary.

I think that Kant does hold that we know whether "things themselves" are spatially and temporally organized: They are not. The Aesthetic is supposed to show that objects can only be given to us in space and time if we assume that space and time are transcendentally ideal, that they are forms imposed by the mind on matter given ab extra and do not adhere in the things themselves. (This is something Kant had held since at least the Inaugural Dissertation.) This is why Kant is only an empirical realist. Phenomenal knowledge is knowledge which does justice to our experience, both inner and outer, but not to objects as they would be outside of experience (as in a Divine cognition via "intellectual intuition").

I should probably reread the Aesthetic to make sure I'm not just in error here, but this is how I recall it working when I first read it, and it's how Kenneth Westphal reads it in Kant's Transcendental Proof of Realism. Which is mostly about Kant's (failed) proof of Transcendental Idealism, despite the title. It probably would have been asking too much to want McDowell to address both the B Deduction and the Aesthetic in one article, but I think he should have. He refers to the problem with the Deduction being "imported" from the Aesthetic and spreading from there to poison the whole stew, but I'm not sure I get how he's pinning down the problem in the Aesthetic. (Really, this is a minor quibble: If the problem with Kant is his doctrine of the transcendental ideality of space & time, then the precise sense in which this doctrine was understood isn't a huge deal, so long as we can agree that it is unsatisfying for reason X.)

Despite the above quibbles, I mainly like the article. McDowell comes close to isolating Kant's problem where I think it is, in the Aesthetic's arguments for the transcendental ideality of space and time, and he ends up at a position I find amenable. Everything McDowell says about Hegel I find entirely salutary. I think it was McDowell's discussion of Hegel in "Reading McDowell" that first lead me to read "the Concept" as "the conceptual", and this made Hegel much more comprehensible than he had been previously; McDowell makes the point explicit in this article, in a footnote, since Pippin appears to have read "the Concept" as "the Concepts". Which is just a weird error, and leads Pippin to make some utterly baffling comments ("table of notions"??). McDowell also does a nice job at bringing complaints Hegel makes against Kant closer to Kant himself; Hegel can occasionally seem to be arguing against a strawman-Kant, as when he merely gestures to criticisms by complaining about "formalism" in the Critical philosophy. I think McDowell makes a very plausible suggestion at the end: It's easier to just go from Kant to Hegel than to try to approach Hegel head-on. So we'd do better to look at Kant, look to Hegel for suggestions for how Kant goes wrong, and then use this amended Kantian understanding to try to work out how Hegel argues for his own positions.

There are also some nice smaller points. The suggestion that we should "make sense of the objective purport of intuitions and the objective purport of judgements together" is a good one. It can seem mysterious that what are merely our judgements can "reach out" to the world (rather than falling short of doing justice to what is given us), and it can seem mysterious that objects can be "given" to us (rather than our purported intuitions of objects being dumb & mute happenings). But when the objects which are "given" to us are taken to be already suitable candidates for inclusion in a judgement, then it is no longer mysterious how judgement can "reach out" to the world; and when an object being "given" to us is taken to be a suggestion about how to judge things to be, it is no longer mysterious how a purported event of "intuition" can be an intuition of an object. Both mysteries are dissolved by dissolving them both.

Back to complaints: I am not quite sure what McDowell takes Hegel to be rejecting in Kant. McDowell seems to say both that Hegel rejects Kant's transcendental idealism, which is true, and that Hegel rejects the idea that space and time are forms of our sensible intuition, which is not quite right. Hegel rejects the Kantian articulation of the idea that space and time are forms of our sensible intuition, but I think it's misleading to say he rejects it tout court. For Hegel, space and time are the forms of the objects given in intuition, and the forms in which those objects are given to us. I suppose this is a fairly minor quibble, since Hegel (unlike Kant) doesn't think other forms of intuition are possible; for Hegel extension in space and endurance in time are requirements for objects to have the sort of distinctiveness that is a requirement of their being possibly given in intuition at all, so the fact that the form of our intuition is spatio-temporal says nothing; all forms of intuition would be like that. Which is why Hegel doesn't talk about "forms of intuition" but just of "space" and "time". (In any case, the rejection that seems to be doing the heavy lifting in McDowell's picture of Hegel is the rejection of transcendental idealism. And that's the right move to make.)

Hegel's Idealism

I just finished reading Robert Stern's "Hegel's Idealism" (PDF). Very nice piece.

It's a response to an article of Karl Ameriks's from a while back. In it, he asked if there was any good reason to call Hegel an "idealist" -- Ameriks was skeptical. He does a fine job of showing that Hegel was not a "mentalist", or Berkeleyian, or anything of that sort, and ends up just saying that Hegel was using "idealism" to basically mean "systematic philosophy." Which is not how anyone would understand the term today. So Ameriks said that we should drop the "idealist" moniker when talking about Hegel. Despite the fact that Hegel emphatically embraced the term: "Every genuine philosophy is an idealism".

Stern responds by giving an excellent exposition of some of the main parts of Hegel's system, including Hegel's use of "idealism". Stern's paper is one of the best broad treatments of Hegel's thought I've ever read; I'll have to bump Stern's stuff up a bit in the stack. Already read his bit on the British Idealists too, "The Curious Case of the Concrete Universal"; I'm still not inclined to read any of the British Idealists (beyond Josiah Royce; I've liked what little of his I've read -- I suppose he's not British anyways), but the article does a nice job handling one of the ways in which Hegel handles the "universal". (I sometimes wonder how different 20th-century Anglo-American philosophy would've been if Russell & Moore hadn't been learned about the German Idealists through a bunch of British Idealists....)

Stern's paper inspired me to write up some Hegelian Meditations, so to speak. I figured I'd through it behind a "jump", since it's a bit long. And rambling. And entirely about Hegel. And it ends up discussing theology. You have been warned. Read at your own risk.

Click on Yuki to see the post.


(she's reading the Phenomenology of Spirit, incidentally. You can make out most of the title on the DVD version; haruhi.tv's SOS-Dan webpage listed "G.W.F. Hegeru's [Phenomenology of Spirit]" as one of Yuki's favorite books, along with some other books which are clearly visible during the series, so I went back and did some freeze-framing. It paid off: When Yuki refuses to leave the library in episode ten ("Melancholy of Haruhi Suzumiya IV"), she's reading Hegel. Made my day when I saw that.)

Before he gets to Ameriks, Stern offers some criticisms of Robert Pippin's version of Hegel, whom Pippin presents as a fundamentally Kantian thinker. Pippin presents Kant's "Copernican Revolution" as reframing traditional metaphysical pursuits by replacing inquiry into "how things are" with inquiry into "how things must be taken to be by any 'I'" -- how things must be thought to be. The objective validity of the categories, their necessary and licit use in application to all possible experience, is argued for by showing that this is a necesary precondition of self-consciousness; I can only be self-aware if I regard myself as existing in a world of enduring substances of various sorts engaged in causal relations according to sundry natural laws. Hegel is said to have agreed with Kant in this basic reorientation of metaphysics; Hegel's categories, too, are said to be validated by appeal to the requirements of self-conscious thought. The rejection of a phenomena-noumena dualism in Hegel's account of what is necessary for self-consciousness leads him to speak of this basically Kantian project in terms of "thought's self-determination" (since thought is not (as in Kant) "bound" by a ding an sich) but the form of Hegel's thought remains Kantian in structure: Start from self-consciousness and see what else you must have thought. Pippin notes that Hegel seems to be not terribly clear that this is how he's arguing, and also that he doesn't merely limit his claims to "what must be thought to be" but speaks of "what must be" full-stop. Pippin claims that both of these are "slips", and do not tell us anything of Hegel's true views.

Pippin seems to underestimate what Hegel's lack of a noumena-phenomena distinction leads to. For Kant, what had to be the case if self-conscious thought was to be possible was that various broad facts had to be the case for all phenomena -- there are spatio-temporal substances engaged in causal relations in recognizable ways, etc. Noumena were not held liable to any such constraints, as they were not (and could not be) subject to the categories. Hegel sees no need to hang on to the ding an sich, and so he doesn't have any need to distinguish phenomena from noumena. So where Kant's requirements for self-conscious thought could end up with the form "If self-conscious thought is to be possible, then things must be so for phenomena", Hegel's end up as "If self-conscious thought is to be possible, then things must be so." Thus Hegel speaks of how things must be, full-stop; for self-conscious thought is possible. Hegel does not end up limiting the demand to be subject to his categories to what is imposed by an 'I' on matter given ab extra, but has the demand applying generally: all must be subject to the categories if it is to be at all. The 'I' as a category thus plays no privileged role for Hegel; it is one category among many. There is no pressure to "ground" the other categories in the 'I' anymore than in any other category; they all flow from one another equally, the whole system of the Logic being presupposed if any of the categories is to have a use. Thus Hegel can begin his Logic with Being for purely aesthetic reasons, and has no need to start with self-consciousness, or with any category in particular; "The only genuinely necessary presupposition can be said to be the decision to consider thought at all, which can be taken to be arbitrary." (This is not to say that Hegel's system is "presuppositionless" in a sense in which it is sometimes made out to be; one sometimes is given the image of the entire Logic being convincing even to the most dedicated skeptic -- the standpoint of "Absolute Knowledge" is supposedly reachable even if one denies having all prior knowledge. Hegel's point is rather that any point of thought may serve as a starting-place; a mind which can think at all can come to see the validity of Hegel's system of categories, his Logic.)

Hegel's Logic begins with Being, since it seems like a nice vague, general notion, and Hegel wants to be sure to avoid any appearance of a "weighty" set of presuppositions in his logic. "Being" as such is seen to have no use in thought; if one tries to think solely of "Being", eschewing all other notions, one is put in the same place one would be if one was thinking of "Nothing" alone -- thinking of what isn't "Being". Both "Being" and "Nothing" have a use because they can both be used -- they are used to track shifts in "Becoming", in that which comes into being and receeds into nothingness. But shifts in "Becoming" as such cannot be tracked solely by means of "Being" and "Nothing", since the two are so far distinguished only in that they are different from one another; there must be variety in becomings if they are to be distinguished from one another, and thus their shiftings marked. Thus "Becoming" leads to "Determinate Being" -- that "Being" has a use in thinking thus requires that there are many beings which are distinguishable one from another. "Being" finds its usefulness in noting the presence of some qualities in certain beings; "Nothing" finds its usefulness in noting the absence of such qualities in other beings. This is (in compressed form) the manner in which Hegel develops his system of categories: the new categories prevent the earlier ones from proving useless for thinking with, and are themselves useful because of the use they make of the earlier categories. Hegel's system of categories eventually grows to include notions like "Objectivity", "Judgement", "Cause"; that these categories are licitly employed is shown by their being inextricably bound up with earlier categories. When a category would be used contrary to the role it plays in the logical system of categories, as when the category of mechanical causation is applied to the world as such rather than some particular determinate object, it can thus be seen to be misused: and this is what Hegel holds to be the way in which classical metaphysics operated. Take a concept like "cause", use it in ways it's not fit to be used, end up being deeply confused by the results. (In this case, using "mechanical causation" outside a context in which one is keeping track of various shifts in external relations between objects by means of recognizing measurable regularities in their interactions. Such is the case with the cosmological argument for the existence of God; one ends up with an "unmoved mover" because "every event must have a cause" is thought to hold outside of objects in recpirocal relations with one another, and so it is made to appear that the physical world as a whole must need a cause, too -- a transcendent one, so as to end the regress.)

Hegel's anti-Cartesianism comes from the fact that his system introduces categories like "representation" such that representations are treated as the way in which a subject comes in contact with the world, not "private" mental objects. Hegel's logic handles topics like "the world" and "reality" only insofar as they are useful for thinking, but this is not to restrict his claims any. "That we must think of things thus" just is to have to think of them in the terms we generally use to think of things: as holding of the world, reality, Being, etc -- which is to think of them in terms of the world, reality, Being, etc. If we must think of things thus, then things are thus for us, and so we hold: Things are thus. This is Hegel's "identity of thought and being": the ways in which we must think of things are the way things are.

This "identity of thought and being" is the rhetorically toned-down version of Hegel's claim that his Logic gives us "God Himself as He is in His Eternity, before the creation of a world and of a finite mind" -- what is "in eternity" is just the connections between the abstract categories. And since the categories can only be used for thought by a finite mind, a mind which is thinking in time and is in relations with various objects and with other minds, Hegel is able to speak of his "God as He is in His Eternity" as "becoming man" while remaining God (for the essence of man is to think, and to think is to employ the system of the categories, which are what is "eternal"), thus in finite thought "the eternal" is present in time. But a thought in a finite mind need not itself be finite; when (as in understanding a free action as free) we approach "Absolute Knowledge", the knowledge of the Idea as Idea, God Himself could have no deeper understanding: in Absolute Knowledge we think the thoughts of the Divine Mind. Thus once the standpoint of Absolute Knowledge is reached, God becomes self-conscious in a finite mind; a finite mind thinks eternal thoughts. This is Hegel's odd version of the ontological argument, which is explicitly trinitarian in nature: The God revealed by "the identity of thought and being" is a God who includes finitude as a moment in the Divine Life, a moment which is overcome by the raising of finitude into eternity-in-finitude -- the "true infinite" which is at home with itself in its other.

Hegel's ontological argument is clearly unsuitable for any sort of foundationalist purpose in religion; the human subject figures in Hegel's description of the trinitarian Life as merely an abstract bit of finitude. But religion involves subjective feelings, personal relationships, an individual's moral status, historical symbols, rituals, etc. This is not a failing on Hegel's part, but a positive feature: Hegel regards cultus as an ineliminable aspect of religion, and the heart of religion is picture-thinking -- the contemplation of symbols, to use terminology more in line with contemporary theology. The particular cultus one participates in, the particular symbols one comes to think religiously with, is a contingent matter. Whichever it is, it will be, qua religion, self-standing: there is no need for philosophy to "underwrite" religion. Nor could philosophy do the job if it wanted to; insofar as it involves comprehending the Absolute, Hegel says, philosophy takes its content from religion, and not vice-versa; philosophy comprehends the myriad features of life as a unified whole, and so life must already be a unified whole for thought to become philosophy. Hegel says that this comes, in part, only with religion; the subjective feeling brought about in the mythic religions played the part in Ancient Greece, and Hegel holds that (Protestant) Christianity plays the part in the modern world. Where religion worked with picture-thoughts, vorstellungen, philosophy thinks the self-development of the Concept, the Idea. Hegel's notion of "God" is frequently labeled "heretical" or "unorthodox" or flat-out claimed to be atheistic (or pantheistic); I don't think any such claim has merit. On the level of picture-thinking, of religion, Hegel was a Lutheran. He was reported to be not terribly devout, but he was a Lutheran confessionally; a lack of personal devotion simply fit with his personality. (His wife was pious to the point of enthusiasm.) He rebuts the charge of atheism (or pantheism) brilliantly in subsection 573 of the encyclopedia, where he is (purportedly) defending philosophy as such from the charge of "pantheism, or the All-One doctrine". As far as the matter of orthodoxy goes, Hegel's God is explicitly Triune and Incarnational; his picture of salvation is a matter of a subject coming to know itself as having always implicitly been God's actual presence in the world (and then coming to do God's actual work in the world) by means of recognition of a (singular) God-man. In a broad sense, Hegel's notion of religion, and so of God, is obviously Christian. And any more specific charge of heresy will be made in terms which are simply answerable by repeating Lutheran confessional language; there is nothing in Hegel's thought which would say one way or the other how many "natures" Christ should be said to have, and so he cannot be guilty of having been lead by philosophy into monophysitism; in general, the guiding notion behind orthodox Christology is that Christ is "wholly human and wholly divine", and this is entirely consistent with what Hegel's "revealed religion" claims for its God-man. The only perhaps-credible charge I have seen levied against Hegel's orthodoxy as a Christian philosopher is that Hegel does not allow "sufficient transcendence" for God; if any sense can be made of this charge, then there will certainly be problems for Hegel. Not because it will show he is unorthodox (though it would), but because it would show the entire System to be riddled with problems -- the notion that there is a transcendent "Beyond Thought" is just what Hegel has set out to dispel as a confusion in many places; if it is not a confusion, but a patent fact, Hegel's System is derelict.

In his treatment of religion, Hegel holds to his general method in philosophy: ""Philosophical thinking proceeds analytically in that it simply takes up its object, the Idea, and lets it go its own way, while it simply watches the movement and development of it, so to speak. To this extent philosophizing is something wholly passive. But philosophical thinking is equally synthetic as well, and it proves to be the activity of the Concept itself. But this requires the effort to beware of our own inventions and particular opinions which are forever wanting to push themselves forward." (ss238, Addition)

Audio Links

Currence has passed along some audio links (via e-mail). I am sharing the love.

Two interviews with Davidson -- one by Dummett, and one by McDowell. I am a bit anxious about finding out what Davidson sounds like; for some reason I always read him as speaking with a British accent. Ironically, I was surprised to find that McDowell does have a British accent, since I had never read him as having that sort of voice.

A few Quine panels. I notice that one is labeled as involving Burt Dreben; given that it's impossible (seriously) to find anything Dreben's written, I look forward to finding out if he's as attractive a figure as he's been rumored to be.

And a huge amount of Brandom material, not all of which is audio. I noticed that the Locke Lectures shawn linked to are part of this site; it didn't occur to me to check out what else was hosted in the neighborhood.

Hopefully this does not end up killing some poor fellow's bandwidth; it occurs to me that I've grabbed half a gig of material from his servers. That could add up rather quickly.

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."

McDowell Papers Online

A nice fellow named Tadayasu Murai, who is a philosophy PnD candidate studying McDowell, Sellars, and Kant in Japan, has alerted me to the fact that Charles Travis's website has a pair of McDowell papers uploaded on it, including "What Myth?". The "Berlin" link is actually McDowell's "Conceptual Capacities in Perception". He also points out that there is a McDowell video lecture available; sadly, it's in RealVideo format, but it might be worth reinstalling RealPlayer for this one.

"What Myth?" is McDowell's response to Dreyfus's APA address, which is available here.

Tadayasu-san has also offered me several McDowell articles which are hard to find nowadays; it'll be nice to finally have "Hegel's Idealism as a Radicalization of Kant" as something more than a footnote entry. I'll pass them on to anyone who's interested once I receive them, since I'm sure Duck will want these, too.

Since I am linking to unlimited McDowell works online, it occurs to me that this kolloquuim is probably new to most McDowell fans; McDowell's opening lecture is all I've read of it, but the opening lecture is pretty great. McDowell is particularly interested in defending his Sellarsian credentials in it; I think McDowell is simply right that he's "a better Sellarsian than some people give [him] credit for."

On a side-note, is "McDowell and His Critics" any good? The Rorty volume was fantastic, and I liked what I read of "Dennett and His Critics" (I thought Dennett's concluding essay was very revealing, especially), but I only recognize Bilgrami and Blackburn among the contributors to this volume.

Update!: I have received the papers, in addition to an interview in German titled "Kant ist der Größte." It would really be nice if I could read German.

e-mail me if you want the papers: dmlindquist at gmail

additional update: Tadayasu would probably prefer it if I didn't spell his name wrong! Error corrected.

Another update: I have also acquired "The Apperceptive I and the Empirical Self: Towards a Heterodox Reading of 'Lordship and Bondage' in Hegel's Phenomenology", following a hat tip by Tadayasu.

Artificial Intelligence and Democratic Education

Please note that I have posted some more thoughts on lecture #2 in the comments on the previous post. Also, shawn has had several nice posts recently on "Making It Explicit." Not that I suspect anyone will read this blog who doesn't already read "Words and Other Things" also.

Well, just got done listening to Brandom's third Locke Lecture. I don't know all that much about AI, and I know practically nothing about early childhood education, so I feel like I learned a lot from this. I had no idea that there was a serious problem with teaching fractions as compared to addition; this makes many old "Peanuts" comics even better than they already were. ¦3

I quite liked pretty much everything Brandom had to say about AI, the Turing test, education, and what they have to do with his concern for pragmatics. I'm not sure how much of what he was saying was supposed to be a summary of what everyone "already knew" about the state of the field, and what he was contributing as a novelty, but he acted like he was drawing attention to some neglected topics in computer science; I wouldn't have thought that pragmatics had been so broadly ignored, given the role it plays in the Turing test (we just need an automaton to act indistinguishably from a person to pass the test), but perhaps things are more badly in need of a shake-up than I'd thought.

That one practice was PP-sufficient for another by algorithmic elaboration meant that any creature which could do the first thing could, by following a prescribed algorithm calling for the performance of sub-tasks of the practice already mastered, do the other. It was noted that this is a bit of an idealization if you're trying to apply the model to humans, since there are psychological barriers to responding to certain inputs with certain outputs (even if one is bidden to by an algorithm). Brandom notes that there's a further problem with trying to find a practice which is PP-sufficient by algorithmic elaboration for discursive activity which didn't already presuppose discursive activity, one akin to the "framing problem." When one revises one's beliefs upon a shift somewhere in one's inferential matrix, every element is potentially in need of revision. There needs to be a way of determining what is and what isn't relevant for analysis, but the only way to do this that anyone is aware of requires using language. But, Brandom points out that there is another sort of PP-sufficiency relation we can work with, that of training. Apparently there are actually empirically well-supported algorithms such that anyone who can count can be taught to add -- but not to subtract; still working on that one. I'd be rather interested in reading more about this. This is an addition I didn't see coming; I was not aware that the study of pedagogy had come so far.

I don't have much to say about this lecture; it is simply very solid, enjoyable work. So, to make up for a lack of content: LOLRussell.

Addendum: I have now listened to P. Stekeler-Weithofer's reply, and Brandom's reply-reply, and the following Q&A, and I have to say: I don't have the foggiest idea what P. Stekeler-Weithofer is on about. He seemed to just be going off on tangents for pretty much the entire half-hour (what Hegel's notion of allgemeinheit as opposed to einigkeit, or Heidegger's critique of "humanism" has to do with AI, or with anything Brandom has mentioned, I could not even begin to guess, but P. Stekeler-Weithofer felt compelled to mention both). His criticisms of the Turing test seem entirely to miss the point of the test; it's a way to point out that manufacturing a passable simulacrum is the goal of AI. The claim that you can't know for sure if you're talking to a Turing machine unless you go and find out from the people who made it, and so the Turing test is structurally flawed, strikes me as the stupidest objection to the test that I have ever heard. I will note that Brandom's reply does mention a criticism of the Turing test, but it is not any of the ones P. Stekeler-Weithofer mentioned; I will also note that Brandom's reply was less than ten minutes long, so I am pretty sure I am not alone in thinking that P. Stekeler-Weithofer was a lousy commentator.

On the other hand, I am curious what N.N. will have to say about P. Stekeler-Weithofer's monologue. I know N.N. has some heterodox views on cognitive science, so he might be a more charitable listener than I've been.

A short note on Brandom's second lecture

John McFarlane's dissertation may not be published, but it is available here*. I'm half-inclined to see if I can rush through it this weekend, then listen to Brandom's second lecture again. I know I was not catching some of what Brandom was trying to say about logic there. It does come out (more explicitly) that the reason behind the PV/VP/PP sufficiency notions is to allow Brandom a way to show that anyone possessing a vocabulary sufficient for assertions possesses the necessary skills for employing logical vocabulary (and, I suspect, modal and normative vocabulary, though these topics are going to be handled in detail in a later lecture).

In lecture-form, it's not clear to me how this proof is supposed to go. Brandom notes that he's privileging assertions, and in general the capacity to give and ask for reasons for what one says (does?), but I'm not sure what his justification for this is. I know that Davidson holds that any language-user must have a concept of truth, since radical interpretation requires us to agree with the speaker about most of what is true (hence they must have thoughts that they hold true). I suspect that Brandom has some argument similar in form to this, but I've just missed it. I don't think Brandom is just privileging assertion because that makes for interesting philosophical work, since he speaks of philosophers being "proud of their logocentrism", of having an answer to Derrida, and of Wittgenstein being wrong that "language has no downtown." He seems to secure this point, though, by ruling some things which seem like language-games (such as the "Slab!" bit from PI 2) as "merely verbal." It is hardly revolutionary to find that, if only certain language-games are real language use, then real language has a "downtown" and that it is not arbitrary to privilege it. This does not seem like a satisfactory answer to Wittgenstein; I'll just set Derrida aside.

One reason I can think of for Brandom to be privileging assertion is that this lets him make bolder claims about the universality of logic: Rather than just being another case (for Brandom thinks there are many) where one vocabulary is PP sufficient for another, logic (and some other vocabularies, I think Brandom calls them "LX") is PP sufficient for any vocabulary. But this is just to say that logic is PP sufficient for any vocabulary which includes assertions as a subset. The universality is just the universality of assertion-vocabulary. Given that the universality of logic seems to be one of the things Brandom wants to explain, this can't be the reason for his privileging of assertions.

N.N. should be glad to hear that Brandom spends quite a bit of time in the Q&A clarifying what he means by "vocabulary"... after offering a "clarification" in the main lecture which he's forced to rather heavily amend. But at least the topic is broached.

I am going to give this lecture (and McFarlane's response) another listen before I move on. I am sure I am missing something important at the moment.

*I should mention that Paul Raymont's "An Idle Threat: Epiphenomenalism Exposed" is great, while I'm linking to that site anyway.

So it turns out I have JSTOR access now

I now have a dozen more reviews of "Mind and World" than I had previously. Sadly, some of them are mislabeled; Noam Chomsky did not actually review "Mind and World." I am pretty sure McDowell didn't translate the Thaetetus, either, though I suppose he might have.

I also have McDowell's response to Dreyfus's APA address, along with Dreyfus's response to McDowell's response, McDowell's response to Dreyfus's response to McDowell's response to Dreyfus, and Dreyfus's response to McDowell's response to Dreyfus's response to McDowell's response to Dreyfus. I have no idea why, but for some reason all of these articles were free yesterday. I wasn't signed in, since the "sign in here" field was still visible, and I don't think UT subscribes to "Informaworld"; I know they don't list this "Inquiry" journal as one they have access to. I was actually looking at the Informaworld site yesterday so I could get an ISSN number for placing an interlibrary loan, but for some reason it had little green icons next to all of the articles.

But today the articles are all behind a paywall again. I am confused. I suspect I may actually have gained temporary access while signed on to the UT library site, but said site is currently 404ing on me, so I can't test this. But, if anyone wants access to the articles, feel free to e-mail me: dmlindquist at gmail. It looks like the issue in question might be viewable as a "free sample" if you register an account, but I know I never like having to futz around with journal sites any more than I have to.

So far I've only read McDowell's response, "What Myth?". McDowell answers Dreyfus's charges exactly like I'd thought he would: Dreyfus has an untenable view of language, of rationality, and of the conceptual (in all three cases, the flaw is that they are viewed as essentially situation-independent). These views are also attrubuted to Merleau-Ponty, but McDowell shies away from saying the same of Heidegger. (I wouldn't hesitate to lump Heidegger in with the other two, myself. But McDowell may be a more careful reader of the logos stuff than I am.) All of this is spelled out with reference to Aristotle's notion of phronesis, which is at least a start at elaborating on what McDowell gestured towards in lectures 4 & 6 of "Mind and World"; I know I've wanted to see him be more specific about what he was getting at with the "second nature" talk. I remember someone called him on this in the "Reading McDowell" volume and he demurred, claiming that his invocation of Aristotle was "merely suggestive" and meant to remind us that the notion of "nature" can be broader than "the realm of law". But if nobody can spell out how rationality is supposed to be a "second nature" for a human being, then it remains doubtful that the concept is up to the task demanded of it: in which case it wouldn't help us to see that "'nature' can be second nature" after all. McDowell's project needs a thicker conception of "second nature" than McDowell himself has thus far supplied.

Dreyfus had also made one of the most common charges leveled at McDowell's position, that it has to deny that non-rational animals can (for example) see that there is food in front of them. McDowell had already answered this in lecture 6 of "Mind and World", but I can understand why the complaint has stuck around: Lecture 6 is a mess, with far too much argumentative weight being put on a gesture to a distinction made by Gadamer (following Heidegger), that between "inhabiting an environment" and "being open to the world." McDowell really should have slowed down in that lecture; I know I couldn't figure out how his argument was supposed to work until I'd read a fair amount of McDowell's post-"Mind and World" work. It doesn't help that in "Mind and World" McDowell spells out the Gadamerian insight via reference to the early Marx -- one usually hopes for the explanans to be more readily understood than the explanandum!

In "What Myth?" McDowell revisits the issue again, and I think he does a much more satisfactory job of it. In response to the charge that because we share some of our perceptual capacities with some non-rational animals, our perceptual capacities cannot be "permeated with rationality", McDowell notes that "this sharing comes to no more than that there are descriptions that apply both to our competence and to the competence of other animals. I can acknowledge that, and still claim that there are further descriptions that fit our case only. There is more to our embodied coping than there is to the embodied coping of nonrational animals." Short & sweet.

(Hilary Putnam treats a very similar topic in chapter two of "Renewing Philosophy", part of an extended critique of Fodor. He brings it up again in "Mind, Body & World: The Threefold Cord". His treatment of non-rational animals' perceptual capacities is the best thing of Putnam's I've yet read.)

McDowell closes the essay with a section beginning thusly: "I am all for the project of giving an insightful phenomenology of our embodied coping. But a phenomenology of embodiment should be conceived not as a corrective to the thought that our orientation towards the world is permeated with conceptual rationality, but as a supplementation, filling out the details of something that needs to be presupposed by any acceptable version of that thought." Yes, exactly. Filling out the details in McDowell's story is precisely what I have been looking for in the phenomenologists. It is nice to see that McDowell is also interested in seeing this done; he's done a rather poor job of filling out the details himself, mainly due to sins of omission.

Will Dreyfus come to his senses? I have no idea! The title of his response essay doesn't exactly fill me with hope, but I haven't the slightest idea what Dreyfus would say against McDowell, here. Hopefully the whole thing doesn't devolve into exegesis of Aristotle, Heidegger, and Merleau-Ponty.

hat tip on the articles: spontaneity & receptivity

Brandom's John Locke Lectures Online

Words and Other Things links to Brandom's John Locke Lectures from 2006. So far I've listened to Brandom's first lecture, and McDowell's response. A few scattered thoughts:

Brandom is much easier to listen to than to read. I'd actually seen the John Locke lectures were online before (in text), and got bogged down in the first lecture. Audio's pretty clear, too; I had to crank Winamp to max volume for it to be loud enough, but at least it was intelligible when it was audible. This is often not true of lecture recordings.

I kept wanting Brandom to provide some more examples of his "PV" relations between vocabularies. If he had any examples other than the indexical one (or the merely-alluded-to project of Huw Price, or the Laughing Santa bit), I've missed them. While listening to the lecture I was skeptical of just what McDowell later said he was skeptical of: both the desirability and the possibility of saying "what needs to be done" to count as making use of some particular vocabulary without making use of that very vocabulary (at least in many cases, including all of the interesting ones Brandom mentions). As I recall, this was just what was at issue in McDowell's "In Defense of Modesty" and its sequel, "Another Plea for Modesty", where McDowell was arguing with Dummett about what we should want from a "theory of meaning"; McDowell came out in defense of the adequacy of basically Davidsonian mock-ups (true-in-L and all that) as being perfectly satisfactory for doing what we should want a theory of meaning to do. I am pretty sure these were also some of the essays where McDowell argued with Crispin Wright about the Investigations; Wright seems to be footnoted in every other essay I've read from McDowell's two books. Good stuff.

Another point related to the "Modesty" essays and McDowell's response to Brandom: I can't say I have any idea what Brandom finds attractive in resuscitating "analysis", at least in the handling-one-vocabulary-with-another sense in which he wants to breath new life into the beast. Various attempts at this have certainly produced interesting creatures in analytic philosophy (and in philosophy in general, reduction being not a recent invention), but I can't say I can think of any cases where the undertaking has proved successful. Nor am I sure why we should lust after such a success when the empiricist and naturalist projects are dead (which Brandom takes them to be).

Brandom mentions the AI functionalism program (which he, rightly, thinks is a hugely important part of the trajectory analytic philosophy's taken in the past century). I suspect that one of the big "pay-offs" for Brandom's project is supposed to be that it shows how we can get an automaton to use intentional vocabulary (or at least get an automaton to use some interesting sort of vocabulary). If so, I suspect that seeing why Brandom's attempt fails might be illuminating for seeing why functionalism should join empiricism and naturalism on the scrapheap of the twentieth century. (If Brandom's attempt does not fail, then hey, get some engineers clued in and start cranking out R2 units and positronic brains.)

I am inclined to think that the success of something like Brandom's program is not needed for AI (or whatever the current descendants of AI are calling themselves). If robots aren't "really" capable of using some vocabulary or other (in that they don't do what we do when we use that vocabulary), but can do a fair enough job of passing for doing so, then I don't know what is supposed to go wanting. If a vocabulary which an automaton can understand is not PV sufficient for some vocabulary V, this doesn't mean we can't construct an automaton such that we comport ourselves naturally with it as if it had mastered vocabulary V. Which is good enough for government work. Conversely, I don't think AI work is going to lead to any amazing revelations about human psychology; if we get a robot that can mimic some aspect of human behavior, there's no reason to think it's doing the "same thing" we do when we behave in the relevant manner in the sense that we can use our deeper understanding of the robot we made to understand how humans work. It may very well be "doing the same thing" in the sense that we comport ourselves with the robot as if it were, say, asserting that the weather's turning grim, but there "analysing" what the robot is doing when asserting won't be any easier (or more productive) than analysing what a human does.

Brandom does seem to have some sort of argument to the effect that any discourse which supports assertion (and some other things I'm forgetting) will have to support modal terms, and normative terms, as well. That might be interesting. I still feel the hairs on my neck raise when modal talk comes up; it still looks suspicious to me. Don't trust the stuff more than I have to. I'm fine with "necessarily", "possibly", "maybe" et al as ways to modify the strength of one's doxastic commitments on a particular point; I suspect that something more sinister than this is often afoot. But perhaps I just haven't had cause to need to speak of necessity, and if I do then I'll warm up to those strange little boxes and diamonds.

I've noticed that McDowell quotes Wittgenstein (chapter & verse) rather often when writing (or speaking) about him. I notice that Brandom's longest actual quote from Wittgenstein is "look to the use" or "language is a motley", and I do not believe he gave the citations; generally, the only direct liftings from Wittgenstein are two- or three-word slogans. Brandom also notes that what he gets from Wittgenstein is something Wittgenstein wouldn't have found interesting or worthwhile, and which he would not want to encourage in others. On the plus side, at least Brandom is self-conscious about the fact that his reading has to be seriously, seriously wrong about what Wittgenstein is getting at -- not that he seems to care. I suspect N.N. got annoyed at these bits of the lecture.

I should probably listen to Brandom's reply to McDowell's reply while it is still fresh in my mind, but I should also have started the reading for Property about two hours ago. Hopefully writing a little something helps me to recall what the devil I just spent two hours listening to.

It would certainly be nice for the John Locke lectures to get more attention in general.

I need to get a copy of the current issue of Inquiry. McDowell responding to Dreyfus's APA address sounds awesome and is exactly the sort of thing I want to see happen over and over again.

Huh, I have somehow spent almost an hour and a half tooling around with this post. Hopefully the Property reading's short! ;_;