Charles Siewert's book, The Significance of Consciousness, can be purchased from Amazon.com

Spontaneous Blindsight and Immediate Availability: A Reply to Carruthers

Charles Siewert
Department of Philosophy
University of Miami
PO Box 248054
Coral Gables, FL 33124
U.S.A.

csiewert@miami.edu

Copyright (c) Charles Siewert 2001

PSYCHE, 7(08), April 2001
http://psyche.cs.monash.edu.au/v7/psyche-7-08-siewert.html

KEYWORDS: phenomenal consciousness, spontaneous blindsight, immediate availability, conceivability, recognitional concept, metaphysical possibility.

REPLY TO: Carruthers, P. (2001). Who is blind to blindsight? Psyche, 7(04). URL: http://psyche.cs.monash.edu.au/v7/psyche-7-04-carruthers.html

ABSTRACT: Carruthers' "immediate availability" theory of consciousness is criticized on the grounds that it offers no reasonable alternative to asserting the metaphysical impossibility of spontaneous blindsight. In defense, Carruthers says he can admit a spontaneous blindsight that relies on unconscious behavioral cues, and deny only its possibility without such mechanisms. I argue: (1) This involves him in an unwarranted denial of the possibility that conscious visual discrimination could depend on behavioral cues. (2) We can conceive of blindsight without behavioral cuing; if we can, then we should not accept Carruthers' denial of its metaphysical possibility without good reason. To warrant this denial it is not enough to hold that the concept of consciousness employed is purely "recognitional," and thus of no relevance to modal claims. The concept is not this cognitively primitive.

I thank Peter Carruthers for his stimulating criticisms. While I cannot defend myself with regard to those "many mistakes" he alleges but does not describe, I will respond briefly now to the criticisms he does make explicit, correcting along the way some of his misunderstandings of my position.

His basic complaint, it seems, comes to this. By appeal to thought experiments in which I describe certain forms of blindsight, I explain what I mean by 'phenomenal consciousness' and criticize certain theories as neglectful of consciousness. But this, he thinks, leaves me vulnerable to a dilemma. Carruthers adopts the Milner and Goodale proposal that actual blindsighters' "guesses" are generated by some sort of subtle unconscious cuing from behavior and "motor programming." The question then comes--do I (a): stay close to actual blindsight, and ask us to suppose that the verbalized judgments involved in the "spontaneous" blindsight of my thought experiment would have similar causes? Or, on the other hand, do I (b): "go more imaginary," and ask us to conceive of spontaneous blindsight as operating, not through these behaviorally mediated mechanisms, but only in some more direct way? Either choice is supposed to lead us to see that my discussion is powerless to raise real difficulties for reductive theories of consciousness.

Suppose I choose (a). Then Carruthers says, he (and others, such as Michael Tye) have a ready way of ducking my criticism that their theories commit them to the unwarranted claim that spontaneous blindsight is either inconceivable or metaphysically impossible. On Carruthers' view, when the content of a perceptual--say, a visual--state is "immediately available" to (certain kinds) of "higher order thought," then (by metaphysical necessity) it is a phenomenally conscious visual experience. But what makes a way of being available "immediate"? Presumably this does not exclude some kind of internal mediation between the reception of sensory information and verbalized conceptual response. We are not talking about an absolute immediacy here. But Carruthers seems free to spell out his notion of immediacy in such a way that, if a higher order thought is related to the underlying visual state in a way that depends on subtle behavioral cuing, the availability of the latter to the former is not truly "immediate." And so he can accept the possibility of spontaneous blindsight, by saying that in such a case the informational content of the visual representation of blindfield stimuli would not be available to higher-order thoughts in "the right way" to give one conscious visual experience, according to his theory.

Suppose I then try a different line of criticism, by opting for the dilemma's other horn,(b). Then Carruthers alleges that the scenarios in my thought experiments go way beyond the kind of blindsight that is naturally possible in human beings, and that the use of such scenarios to argue against his theory would be question-begging. Moreover, if I "go imaginary" in this way, my whole strategy of argument can be seen to be illegitimate, once we properly understand the nature of phenomenal concepts.

Let's examine the first horn of the dilemma more closely. If I accept route (a), then I allow that properly "spontaneous" blindsight could occur with something resembling the etiology that Milner and Goodale propose underlies actual blindsight. So here, it seems, we conceive of a case in which, in the absence of "forced choice" questions from someone else, the blindsighter's visual stimulus somehow triggers behavioral responses and motor programming that influence the questions he poses himself about what's in his blind field, and the answers he gives them, so as to make them as accurate as those of some (perhaps only very poorly) consciously sighted subject. I assume we needn't suppose though, that every (behaviorally cued but still) spontaneous blindsight judgment he makes requires he first consciously entertain a question about which of a list of types of stimuli are there, or which kind of a specific category of stimulus (What shape? What color?) is there. For in describing spontaneous blindsight I explicitly ask us to suppose that the judgments can occur unbidden by any such question (Siewert, 1998, pp. 76-78). The essential thing then, would be to allow that calling the judgement "spontaneous" does not preclude its depending partly on some kind of (non-conscious) Milner/Goodale type behavioral influence. The characterization I give of spontaneous judgment in my book does not imply that this kind of influence would keep the judgment from being spontaneous. So let us allow it. Now by accepting no more than this form of (as far as we know non-actual) blindsight, Carruthers could still claim to leave the possibility of spontaneous blindsight open.

However, I don't think this removes the difficulty. For first, I take it that Carruthers also holds that "immediate availability to higher order thought" is necessary for phenomenally conscious sense experience. Now one problem this creates, noted by a number of Carruthers' critics, and mentioned in my book: he is committed to holding that if creatures lack the conceptual wherewithal to entertain thoughts about their own thoughts (as many plausibly do, including young members of our own species), then things must look no way to them at all: they are incapable of feeling pain, and utterly without phenomenal sensory experience. To me it seems pretty clear that this conditional is unacceptable. But here I want to focus on a different worry. For Carruthers' claim that availability-to-higher-order thought is essential to phenomenal consciousness raises new problems when we interpret immediacy as he now proposes. Suppose unconscious behavioral cuing, similar to that alleged to occur in known blindsight subjects, were found to play a role in generating some other subjects' spontaneous verbal responses to visual stimuli. Then on Carruthers view we would have to conclude from this alone that the stimuli look no way at all to them-that they have no phenomenal visual experience of the stimuli--even if they sincerely affirmed otherwise. We'd have to say to them--"No, you must be wrong--actually you are a blindsighter, because your eye and hand movements are unconsciously playing a role in your ability to spontaneously verbally classify your visual stimuli. And it's a metaphysically necessary truth that the availability of conscious visual content to the subject's higher-order thought can never depend partly on unconsciously utilized behavioral cues." But just what reason do we have to believe that there is some such super strong necessity to this effect? None, I think. Thus embracing the first horn of Carruthers' dilemma does not shield his theory, but only lands him in other difficulties.

It is a little surprising that Carruthers even offers me this first horn of the dilemma, predicated on assuming that the form of spontaneous blindsight I describe must be realizable by means similar to the neural mechanisms that enable actual blindsight subjects to make their responses. After all, I explicitly say that, in conceiving of spontaneous amblyopic reflective blindsight, I do not ask the reader to assume that such feats of visual discrimination without visual experience are compatible with the natural organization and functioning of the human nervous system (pp. 78, 94-5). There may indeed be no way for a human nervous system to support this sort of blindsight, given actual human neuropsychology. If that's true, it's certainly interesting, but its relevance here is limited by the fact that my thought experiment does not require that we conceive only of such forms of spontaneous blindsight as are physically realizable in the natural endowment of homo sapiens. And so, even if I'm wrong about the first horn of Carruthers' dilemma, I can still concentrate my criticisms on acceptance of the dilemma's second horn, in which it is granted that I may untether my thought experiment from a concern with its neurophysiological realizability. But Carruthers seems to think that once I go in this direction, my procedure no longer has any bearing on the acceptability of his proposed reductive explanation of phenomenal consciousness.

Perhaps here I should explain why it can be relevant in this context to consider conditions under which certain forms of consciousness are absent or present, independently of a concern for whether these scenarios actually occur, or are realizable in the human nervous system. One reason I think this is worth doing, is that it can sometimes be useful, when opening up a subject of investigation, especially one prone to generate confusion and misunderstanding, to clarify what one means by key terms, and to draw pertinent distinctions. And one way of doing this is to illustrate as explicitly as one can one's application of terms in situations that highlight distinctions liable to be blurred. For such clarificatory purposes, it often is not essential to determine whether the situations in which the term is applied are actual or not, or which scientific laws their actualization would conform to, or contravene. A second reason for "going imaginary" (as Carruthers puts it) is that proposed theories of consciousness are sometimes committed (either explicitly or implicitly) to stronger than nomological claims of necessity and (im)possibility. So those who hope to explain consciousness offer to tell us all we do or should mean by calling a state "conscious," or they claim to tell us what being a conscious state--that property--consists in, or, as with Carruthers, they propose to tell us what "constitutes" phenomenal consciousness. It is reasonable to interpret such claims as involving commitment to the notion that certain situations are inconceivable or conceptually impossible, or else that they are "metaphysically" impossible. (The latter notion is that often illustrated by the claim that even though one can, in some sense, conceive of water that isn't H2O, and one's ordinary concept of water is distinct from the scientific concept of H2O--still, in no possible world may there be any water of a different (non-H2O) chemical nature.) Claims of necessity of both kinds are justified, at least in part, by reflection on what we mean by key terms, sometimes brought out by consideration of scenarios that are not, for all we know, actual. In the case of metaphysical necessity, this has involved, for instance, contemplation of "worlds" in which there is some stuff that behaves to ordinary macro appearances much as our water, but which has a markedly different chemical composition. So thought experiments seem pertinent to the type of necessity and possibility here at issue.

It should also be emphasized that to say that something--"spontaneous blindsight," for example--is either conceptually or metaphysically impossible is a strong claim. And it is to say more than that the occurrence of this is inconsistent with such laws of nature as obtain in the actual world, pertaining to the realization of visual discriminatory abilities. It is certainly to say more than that its occurrence is inconsistent with natural laws governing what human nervous systems can do. For a certain kind of vision may not be naturally, nomologically possible in human nervous systems, but both conceivable and metaphysically possible.

But now we might wonder why a seeker of an explanatory theory of phenomenal consciousness should want to reach for something as strong as either conceptual or metaphysical impossibility. Why not make do with more modest claims of mere nomological impossibility? A full answer to this question would need more time. But we might get some sense of why a self-styled "reductive" explanation of consciousness of the sort Carruthers suggests should seek to establish such ambitious claims of impossibility. For suppose he claimed only that immediate availability to higher order thought is nomologically sufficient for phenomenal vision in human beings. One question that would arise here is why, even if this is true, we should regard the immediate availability of visual content as explanatory of phenomenal vision rather than the other way around. On first examination anyway, it would seem to be the other way around. (Compare: university concert tickets are "immediately available" to registered students--other customers may have to wait awhile. Presumably, the tickets' immediate availability to students is explained by their being university concert tickets; it's not that their being university concert tickets is explained by their immediate availability to students.) Another consideration: we'll think that availability is explained by, rather than explains, phenomenal consciousness, if we think the former is only nomologically sufficient for the latter, and we believe (reasonably enough) that human infants have phenomenal vision before they acquire the disposition to think about their own thoughts.

So perhaps for something like these reasons, if Carruthers wants to claim that immediate availability explains phenomenal consciousness, he needs to make the strong claim (as he does) that the former is not just nomologically sufficient for, but "constitutes," and is metaphysically sufficient for the latter. Now, embracing the second horn of Carruthers' dilemma, I would deny we have reason to accept this claim. So I would say something like this. "Suppose a spontaneous blindsighter's verbalized visual discriminatory judgments are not generated partly by subtle unconscious behavioral cues. If you can conceive of such a blindsighter, then you're conceiving of something that Carruthers is committed to saying is metaphysically impossible. But we just don't have adequate reason to think this."

Just what does Carruthers think is wrong here? First, he says that I ask us to construe "blindsight cases" so broadly "as to cover anything described in any proposed reductive explanation of consciousness." And that is merely to define blindsight so as to rule out any successful reductive explanation of consciousness--which surely is to assume what is to be shown. What's more, he thinks now I am taking my thought experiments as far from actuality as "zombie" scenarios. And a proper theory of phenomenal concepts (as "recognitional") neutralizes any threat from that quarter.

Now I do, as a matter of fact, think I can conceive of a spontaneous form of blindsight that does not depend on any unconscious behavioral cuing similar to that Carruthers takes to be involved in actual blindsight. But to ask you to conceive of this is not ask to you: "Please now conceive of a form of blindsight in which the subject possesses all the visual abilities that any proposed reductive explanation of consciousness deems sufficient for phenomenal vision." No. It is just to ask you to suppose that, however the spontaneous blindsight judgments are generated, they are not generated through unconscious behavioral cuing. Perhaps it turns out some proposed reductive explanations of consciousness leave us no reasonable choice but to maintain that spontaneous blindsight of this kind is metaphysically impossible. If so, we may then argue that since we should reject that claim of impossibility, we should reject the reductive explanations. But to argue in this way is not question-begging.

Next, in response to the criticism that such a blindsight thought experiment propels us into the outer reaches of conceptual space, somewhere near the zombie galaxy, I say this. Again I concede it may well be true that any subjects conforming to the description of such blindsighters would have visual systems organized differently from those of actual natural human beings. And I don't claim to know generally how to measure "distances" between possible worlds with precision. However, surely a world in which such subjects occur could be in many respects much more like ours than Chalmers' zombie world, in which no naturally possible organization of physical particles is sufficient to give anyone in that world the slightest phenomenal tickle. It seems fair to say that world is much more remote from actuality than a world in which some blindsight subjects are wired only as differently from us as they need to be, to make crude spontaneous visual discriminations without the help of the hypothesized Milner/Goodale mechanisms.

But what about the suggestion that thought experiments of the sort I introduce have no force against reductive theories of Carruthers' stripe because they utilize our merely "recognitional" phenomenal concepts? The point seems to be this. I have concepts which I apply "straight off" to my phenomenal experience--the concepts I use when I judge with first-person warrant about, for example, how things look to me, how they feel, whether they feel or look similarly or differently on different occasions. These are "recognitional" concepts. There are, distinct from these, certain theoretical concepts introduced by proposed reductive theories of consciousness to talk about the same states: for example, the concept of being a sensory state whose content is immediately available to higher-order thought. And the fact that these two concepts are distinct allows me to think, without contradicting myself: "Someone might have a visual representation, the content of which is immediately available to higher-order thought, even though, as in blindsight, the object of that visual representation does not look any way to her--she does not have a conscious visual experience of it." Now, one may admit all this, Carruthers tells us, and still hold that being a visual state whose content is immediately available to higher-order thought is what constitutes being a phenomenal visual experience, and it is a metaphysically necessary truth that whatever possesses the former, possesses the latter. And this entitles us to hold that the occurrence of the latter is explained by the occurrence of the former.

But is this sufficient to answer the concern that Carruthers' theory is committed to an unwarranted assertion of metaphysical impossibility? It seems to me that if, on sober consideration, one can conceive of F occurring without G, then one is entitled to interpret 'F' and 'G' in such a way that it's not the case that 'F --> G' is a metaphysical necessity, until relevant positive reasons have been adduced for asserting that they should only be interpreted so that this sentence states a metaphysically necessary truth. The question is whether such reasons have been offered. Of course, one may say that, in general, it is possible for two concepts to be distinct, even if it is metaphysically necessary that everything to which the first concept applies, the second concept also applies. But that does not by itself give us relevant positive reason to think this is the situation in a given case--say the one where F = being a visual representation whose content is immediately available to higher order thought, and G= being phenomenal visual experience (see pp. 161, 354-6).

One might try to argue for the metaphysical impossibility by analogy to some case where such an impossibility is accepted--say, the water/H20 case. There are, as I point out (pp. 159-60), some serious obstacles to such an analogy. However, I do not, as Carruthers suggests, reject his claim of metaphysical necessity on the grounds that it could hold only if we intended to use the concept of phenomenally conscious states to refer to some essence underlying their appearances (as in the water/H2O case). That is not my point. And I do not, as Carruthers claims, reject his metaphysical necessity on the grounds that it could hold only if it were metaphysically necessary that being a phenomenal visual experience is identical to being a visual state whose content is immediately available to higher order thought. Rather, I reject it on these grounds. I can conceive of a case-a spontaneous blindsight case--in which it seems one would have all that one needed to count as the possessor of a visual state whose content is immediately available to higher order thought, in Carruthers' sense, while one lacked the relevant conscious visual experience. So it seems to me that Carruthers-style higher order availability does not entail sensory consciousness. And I can find no good reason to say that, nevertheless, it is metaphysically necessary that Carruthers-availability --> phenomenal consciousness.

Perhaps I will find such reason however, if I focus on this notion that my concept of phenomenal consciousness is "recognitional." But it seems to me that this is relevant only if one claims that the concept in question is "merely" recognitional. That is to say, it is merely an ability to sort similar things together, which does not include the ability to assess accurately what does or doesn't follow from something's being of that sort, and what being of that sort follows from, through consideration of hypothetical situations. Otherwise, how is one to argue that the concept of phenomenal consciousness I employ does not equip me to make legitimate assessments about the lack of necessary connections between occurrences of the sort of state it picks out, and various proposed conditions?

But why should we regard phenomenal concepts as so cognitively impoverished--as merely recognitional? Carruthers sees that I reject this idea, and he views my position here as "weak." However, perhaps this is because he misdescribes the argument I make on p. 162. He thinks I claim the understanding of 'conscious' to which I appeal allows me to determine that it is impossible for sunflowers to have conscious visual experience of the light to which they manifest a phototropic response. But I don't actually say that. What I do hold is that the possession of this understanding of 'conscious' gives one the ability to determine that if a plant responds phototropically to light, that does not entail that it has some sort of conscious visual experience. Someone who understands what I mean by 'a conscious visual experience,' understands that a phototropic response does not necessarily guarantee a conscious visual experience. However, if my concept of phenomenal experience is merely recognitional, it would seem I do not have any right to pronounce on even this much. And that seems to me absurd. Does that make me an "a priori philosopher"? If so, maybe that isn't such a shameful thing to be.

Here then is how I would sum up the situation. On my view, we can know, with first-person warrant, and through considering non-actual situations using concepts that we can apply with such warrant, that our visual experience is conscious, in a sense that would be denied if one said spontaneous blindsight is either inconceivable or metaphysically impossible. In my book, I suggest that Carruthers' theory of phenomenal consciousness is vulnerable to the concern that it leaves him no reasonable alternative to such an assertion of impossibility. And so he is vulnerable to charges of neglecting phenomenal consciousness. In his critical remarks, as I understand them, he proposes the following defense. First, he suggests we say that when sensory content is available to a subject's higher order thought only partly through the unconscious use of behavioral cues, it's not immediately available. Then we can say that immediate availability is sufficient for phenomenal consciousness, even if spontaneous availability is not. But the problem here is that Carruthers also thinks immediate availability to higher order thought is necessary to consciousness. So he needs to say that there couldn't possibly be conscious vision in a case where a subject can spontaneously verbalize and reflect on visual content only partly through unconscious reliance on behavior and motor programming. But that claim seems no more reasonable than a denial of the possibility of spontaneous blindsight. Thus, taking the first horn of Carruthers' dilemma does not help his theory after all.

Carruthers also needs to defend his view against the challenges posed by my taking the second horn of his dilemma. As I understand it, his response to me here rests on the following claims. (1) I generally beg the question against reductive explanations of consciousness. (2) My thought experiments stray as far from actual blindsight as zombie thought experiments. And (3) I try to employ merely recognitional concepts for the job of assessing claims of necessity and possibility-a job for which they are not suited. I have argued briefly why neither (1) nor (2) is correct. As for (3)-again, it's unclear on what grounds we should regard the understanding involved in our grasp of phenomenal consciousness as so impoverished as to be useless to the relevant assessments of necessity and possibility. Further, if we did, it seems we would deprive ourselves of so much as the right to say that it's not necessary to have visual experience of light, to have a phototropic response. (And I wonder on what grounds Carruthers would reject simple behaviorist theories of what property our phenomenal concepts pick out, if these concepts are so intellectually inert.)

In conclusion: it seems Carruthers would defend his theory by claiming that it is reasonable to allow for the possibility of spontaneous blindsight with behavioral cuing, while declaring it metaphysically impossible without this. But Carruthers has not shown that it is any more reasonable to proclaim that metaphysical impossibility than it would be to assert the metaphysical impossibility of spontaneous blindsight generally. And on the test I propose, a theory that leaves us no reasonable alternative to such denials of possibility neglects phenomenal consciousness.

The question may be raised here whether what I have said is sufficient to justify the conclusion that the forms of blindsight that figure in my arguments are metaphysically possible. Maybe not, but that doesn't much matter. The burden of proof is on those who offer theories of consciousness committed to these strong claims of impossibility. As I remark in the book (p. 163-4), to know that my experience is conscious in a sense that would be denied by someone who asserted the metaphysical impossibility of certain forms of blindsight, I do not need to know that these are metaphysically possible. It is enough that the case for impossibility has been found wanting. I may then abstain from pronouncing on just what is and isn't metaphysically possible, either here or quite generally. (However, to the extent that I can see how to apply this notion of possibility to the case at hand, it would seem to me that I have rather more warrant for saying that it is, than that it is not metaphysically possible, for someone to have Belinda-style blindsight.)

In saying all this, I do not accuse Carruthers of having neglected blindsight. Maybe he is implicitly accusing me of this, when he asks in his title "Who is blind to blindsight?" But all he has shown that I previously neglected was to consider how one might use Milner and Goodale's theory of blindsight to defend Carruthers' "immediate availability" theory of consciousness. I hope to have begun to make up for that negligence now, and to have made it clear that, "a priori

References