Disjunctivism, Skepticism, and Why I Shouldn't Start Writing Things at One AM

McDowell has a new article out, in the collection Disjunctivism: Perception, Action, Knowledge. The title of the article is "The Disjunctive Conception of Experience as Material for a Transcendental Argument". It was previously delivered as a lecture in 2006. As best I can tell, there's nothing new in this article; McDowell just covers some of the same material he's discussed in his earlier writings about disjunctivism, skepticism, and transcendental arguments.

McDowell has previously been hesitant to endorse any "transcendental arguments" (since there shouldn't be a need for them if we can therapize away the worries instead), and he's generally just as hesitant here; I'm pretty sure all of the material on Stroud and TAs was already covered in "Reading McDowell". The one thing that did leap out at me is p.382, where McDowell claims that Strawson's reading of the Transcendental Deduction, or perhaps the Deduction itself, might serve as a convincing transcendental argument that any potentially self-conscious state or episode is an episode which claims objective purport. So McDowell is slightly less hesitant to endorse "transcendental arguments" than he has been in times past, but only slightly, and he doesn't actually offer any arguments of his own under that banner, but merely points to the work of others.

The meat of the article is devoted to continuing an argument with Crispin Wright about perceptual skepticism. It all read like stuff I've read before.

I've wasted several hours puzzling over Wright's response, rather than sleeping. I now have a headache. 2400 words worth of dreck about Wright and disjunctivism are hidden behind Rena.


I just don't get Crispin Wright's response. It seems to be as wrong as anything can be; he seems to totally miss how the disjunctive view is supposed to answer perceptual skepticism.

What could make someone want to say something like this, from Wright's response to McDowell: "There is surely no plausibility whatever in the thought that the mere perceptual apprehension of the hand in front of my face is all it takes to put me in position rationally to claim knowledge that it is there." (Emphasis his; I'm not sure what the emphasis is there for. Wright seems to have some pretty crazy ideas about what being "rational" demands, or perhaps he means something remarkable by "claim". In any case I don't know why his emphasis stops short of "knowledge"; it doesn't seem like a publisher's error, since I'm going off of the draft, which was presumably from Wright's own fingers.)

Wright doesn't think it's rational to claim that I know I have hands, when I can see my hands in front of me. He claims that Moore's argument is question-begging, but he seems to hold that it isn't even valid.

Wright repeatedly treats the "disjunctive view" as if it meant that the object of one's perception is the disjunction "p or the-mere-appearance-that-p" -- as if this was the information conveyed by a perception that p, on the disjunctive view. But this is just what the disjunctive view denies. On the disjunctive view, the "highest common factor" is not a possible object of perception; only the disjuncts separately are. I can see a barn, or I can see a fake barn; I can't see either-a-barn-or-a-fake-barn. I might, on reflection, withhold judgement about which of the two I saw, but this would not be to judge that I had seen a third thing, the disjunction of the two. I saw the one, or I saw the other; I am warranted in holding that I saw one of them, and have no grounds for holding that I saw the other. To withhold judgement on which is which may or may not be justifiable, depending on what else one holds true. This is unlike the "highest common factor" view, wherein any inference from what I perceive to what is the case is groundless and without warrant.

If I take myself to have seen a barn, then I'm either entitled (non-defeasibly) to the claim that There's a barn over there, or I am wrong about what I take myself to have seen. And the same holds with the fake barn: If I take myself to have seen an illusory barn, then I am either entitled to the claim that I apprehend the mere shade of a barn there or I'm wrong about what I've taken myself to have seen. In either case, if I saw what I took myself to have seen, then it does follow that what I've seen is the case, and Wright's skeptic is refuted. If what I took to be the case is not, in fact, the case, then I was wrong about what I was entitled to hold. But this is not a case of a defeasible claim being abandoned as more information comes to light; if I was wrong about what it was I saw, then I never had any entitlement to the claim I took myself to be entitled to -- not even a defeasible one. If I take myself to have seen a barn, and what I actually saw was something else (or just a hallucination), then I was flat wrong. I never had even defeasible warrant for my claim, and so there's no warrant that I could be abandoning when it comes to light that I was wrong about the barn. But if the skeptic wants to suppose that all of my perceptual judgements are false (and so any entitlements I ascribe to myself are in error), then this clearly begs the question in favor of the skeptic. If I am wrong about practically everything I take myself to know, then I'll grant that I'm also wrong about what I take to be entailed by what I know. But Wright's skeptic gives me no reason to think that I am wrong about everything I take to be the case by way of perception. Hence if I take myself to have seen various things, and I am right about some of them, the skeptic is defeated. If I am wrong about all of them, then the skeptic's conclusion is true, but he has provided no reason to believe that it is so. If I withhold judgement on all of my perceptions, refusing to either believe that they are veridical or that they are illusory, then Wright's skeptic is correct: I don't know if I've ever perceived anything. But this is trivially true, since knowledge that p requires belief that p; if I refuse to believe all my purportedly perceptual claims, then I can't know anything that way, either. But no argument is given for why I oughtn't to trust my senses, since on the disjunctive view my senses provide me with opportunities to gain warranted knowledge of how things stand, in some cases. Thus, on the disjunctive picture of perception, whatever the position I take with respect to my purported percepts, Wright's skeptical argument does not go through.

Wright writes: "Direct awareness of states of affairs that make P true is one thing; warranted belief that P, for one fully apprised of what it takes for P to be true, is something else. One plausible additional necessary condition for the latter..." and then goes on to some blather about "responsibility" and "rationality". But this just shows that he rejects the disjunctive picture of experience. For the disjunctivist, the only additional "necessary condition" for warranted belief that p, if I have been made directly aware that p, is that I believe that what I intuit is actually how things stand in the world.

Wright considers a hypothetical in which there is a 50-50 chance he is hallucinating that there is a cricket match going on, and claims that in this scenario he is neither justified in claiming that there is a cricket match going on, nor does he know same, and neither can he say the opposite. But there seems to simply be no argument for this conclusion; it is simply called "irrational" to reason otherwise. But I am inclined to say that Wright would be perfectly justified in claiming (even claiming to know) that there is a cricket match going on in such a scenario; it is just that there are even odds that he would be wrong about this. And I'd say the same about his right to claim that he's hallucinating that there is a cricket match going on. If he wants to withhold judgement on the matter, this also strikes me as rational, given that nothing great rides on whether or not there is a cricket match going on -- in this case he will have failed to gain some bit of knowledge which he was entitled to, but he has also avoided gaining a false belief that he might have been lead to hold. None of these three options strike me as more or less rational than the others, given only the details Wright gives us; he seems to be afraid of gambling, lest he possibly be in error. Wright claims that only the third option is rational, but I can't find an argument for this conclusion, unless it's simply supposed to be that perception rests on a defeasible inference, and in this situation it is so defeased, and so the inference would be a bad one. Which, again, would simply be to reject the disjunctivist picture without argument; for the disjunctivist there is no inference involved in perceiving that p.

Wright occasionally states the disjunctive position correctly, but then immediately goes on to act as if the "phenomenal state" common to the two disjuncts was all that was actually perceived. Which is just to go back to a "highest common factor" view. From the close of Wright's paper: "It is the supposedly possible phenomenology of subjective indistinguishability, rather than a supposed "highest common factor", that is at the base of the problem." Wright seems unaware that treating this "possible phenomenology of subjective indistinguishability" as an object of perception is just what the disjunctivist calls the "highest common factor view"! No doubt, the disjunctive view is not a helpful way to dissolve skepticism, if it doesn't exclude the view it was designed to oppose.

On the disjunctivist picture, in perception I am either given that this is how things are or I am not given anything. Wright occasionally claims that to be justified in holding that I have been given that p, I must be able to justify that I am able to be given things in the manner I regard myself as having been given the fact that p -- if I hold myself to have sensed that p, I must be warranted in trusting my senses. But this inference is trivially available to the disjunctivist: if I am given an indefeasible reason to hold that I perceive that p, then the inference to I can perceive things is banal. Hence Wright's skeptical arguments fail on this front, too. They give the disjunctivist no reason to doubt that he has perceived that many things are the case which are the case, and that his so perceiving them justifies his beliefs that that is how things are.


Wright's reply in this volume is the first piece of his I've actually forced myself through, all the way to the end. His style grates on me, and his arguments just seem bad. I've never found "fake barn country" scenarios to lead my intuitions in the way they are intended to; if I see one barn and am fooled a hundred times in addition, then I want to say I saw one barn, and was fooled a hundred times in addition. If I am then told that I passed a hundred fakes, along with a solitary real barn, then I would be inclined to suspend judgement on which of my many perceptual beliefs vis-a-vis barns were true, and which false. But I would know that one of them was true, if I thought I'd seen a hundred and one barns, and there were only a hundred fakes. The moral I am supposed to draw from this story is that I didn't know that I saw any barns, including the one that I know that I saw. But I did know (in one of the 101 cases) that I saw a barn, before I distanced myself from the judgement. If I reverse this distancing, and hold that all of the things I saw were barns, then I will be wrong many times, and right once. The many errors do not detract from the one success, though they might make it harder to take advantage of the knowledge.

Wright applies his "I am in position rationally to claim that" operator incorrectly on page 12. Line I should become "I am either in position rationally to claim knowledge that I am perceiving a hand in front of my face or I am in position rationally to claim knowledge that I am in some kind of delusional state of that phenomenal character", if it is to represent something a disjunctivist would take themselves to be entitled to if they (seem to) see a hand in front of their face. If Wright had applied his invented operator in this manner, then it would be easy to see how II followed from I: Depending on how the world is, one or the other of the disjuncts will be true. If it's the first then II follows, if it isn't then it doesn't. This consequence relation is a matter which is independent of whatever I know or take myself to know, and so there is no question of whether or not I am warranted in believing it to hold; if I have a hand in front of my face and I perceive that this is so, then my rational credentials are impeccable.

I suspect that all of the above may be the wrong tack to take in handling this issue.

I am mystified by Wright's "Cartesian Dreaming Argument" in his paper's appendix. The conclusion of the argument does not strike me as more paradoxical than its first premise, so I don't see the point of the formalisms at all.

"To get empirical evidence you have to carry out some empirical procedure." This strikes me as false. I do not carry out any "empirical procedure" to perceive things; I simply see them. I don't do anything to see them; experience is "forced upon me". Wright uses this claim, along with the claim that "it is plausible that the evidence produced by such a procedure may not rationally be regarded as any stronger than one's independent evidence that the procedure was carried out competently and with due diligence — and hence that it was carried out at all" to argue that I'm not warranted in believing that I'm not dreaming. I don't know what it would mean to perceive "competently and with due diligence". But that doesn't seem to matter, for the passage continues

Dreaming excludes the genuine execution of empirical procedures. Hence the strength of warrant generated by executing such a procedure is rationally limited by the strength of one's independent warrant that one did not dream its execution! Hence no such procedure can generate a warrant that one is not (fully lucidly and coherently) dreaming.

Which is to say, if I want to claim that I saw a hand, I need independent warrant for the claim that I can see. Given that this would apply to every other claim I make to being gained through sight, this strikes me as entirely implausible, unless I can bootstrap my way up by using "I see a chair" to warrant trusting the senses that tell me that I see a desk, and then vice-versa. Which doesn't sound like what Wright is demanding claiming is plausible about warrant. I suspect Wright is trying to rule out responses like Putnam's response to brain-in-a-vat stories: I can see that p; I could not see that p if I were dreaming (or otherwise out of touch with reality); ergo I am not dreaming (or otherwise out of touch with reality). Which strikes me as a perfectly fine way to respond to brain-in-a-vat stories.

I suspect that the only points of substance in this post are my response to "fake barn country" scenarios and my attempt to give a disjunctivist's view of Wright's double blind test. These are the parts of the post I like least.