Block describes how in Sperling's (Reference Sperling1960) classic study subjects were briefly shown three rows of four alphanumeric characters. The subjects reported seeing the array, but when asked to identify the characters could only name four or five. Block invokes this study and an impressive number of further considerations to argue that “phenomenology overflows accessibility” (sect. 9) – you can be consciously aware of the alphanumeric array even though you cannot cognitively access the individual characters.
I agree that there is one sense in which this is true. But I'm not sure that it supports Block's larger program of distinguishing “phenomenal consciousness” from “access consciousness.”
Suppose we distinguish “scene phenomenology” from “item phenomenology.” In relation to the Sperling display, you have the former once you are phenomenally conscious of a 3×4 array, whereas you have the latter if you are phenomenally conscious of the individual characters. Now focus on scene phenomenology: Block's arguments seem to me to leave it quite open that scene phenomenology may constitutively require some kind of access. After all, his arguments depend crucially on the fact that normal subjects report that they are visually aware of an alphanumeric array – this is his basic reason for saying that there is more in phenomenology than the four or five characters they can name. At first pass, this certainly suggests that scene phenomenology requires at least “scene access” in some sense that allows normal subjects to report that they are aware of an array, even if they cannot identify all the characters.
No doubt Block would urge that scene phenomenology does not require even scene access. He maintains that the patient G.K. is sometimes consciously aware of a face even though G.K. denies seeing any such thing. Block also suggests that there can be phenomenology in Fodorian modules whose encapsulation presumably precludes cognitive access of any kind. Still, while Block makes these claims, it is not clear to me that he offers any arguments for scene phenomenology without scene access. Rather, his arguments all concern the possibility of scene phenomenology without item access, which is rather different. He shows that I can see the alphanumerical array without being able to report all the characters, but not that I can see it without even knowing that I am seeing an array. (Nor is it is entirely clear to me that Block is consistent in urging the possibility of scene phenomenology without scene access. Consider what he says in section 11 against the suggestion that a state is conscious if it is broadly cognitively accessible in the sense that it can be “amplified if attended to.” Block's objection is that this would let in some “uncontroversially unconscious” states. But what makes a state “uncontroversially unconscious” if it is not that subjects tell us so? Once we allow that a state can be conscious even though normal subjects systematically deny this, then I'm not sure there will be any uncontroversially unconscious states.)
Having got this far, we might well wonder whether Block really does anything at all to separate phenomenology and access. He shows that we can have scene phenomenology without item access, but not that we can have scene phenomenology without scene access, or item phenomenology without item access.
Block argues that we need to let phenomenology without access explain the kind of data displayed in the Sperling experiment: in such cases we have phenomenology (in the back of the head) without cognitive access (in the front). But, for all he says, an equally good explanation would be that we have scene phenomenology/access (in the front of the head) without item phenomenology/access (which would also be in the front).
Does it make sense to posit scene phenomenology without item phenomenology? Can we be conscious of the array without being conscious of the individual characters? I don't see why not. If visual perception involved something like physical photographs in the brain, maybe this would be impossible. You can't photograph an array of characters without photographing the individual characters. But if conscious seeing is always seeing as – always a matter of bringing back-of-the brain activity under stored patterns – then I don't see why you can't consciously recognize something as an array without consciously recognizing the individual characters.
I have been suggesting that Block's arguments fail to drive a wedge between phenomenology and access. However, this is not because I am a “metaphysical correlationist,” as Block suggests in his section 7. I don't think that there is any principled reason for insisting that consciousness must be tied to access. The passage that Block cites (from Papineau Reference Papineau2002, p. 187; see target article, sect. 7, para. 2) is my characterization of the “standard methodology” of consciousness research, not my own view. As is made clear in the later sections of the chapter from which Block quotes me, I am quite open to the possibility that there may be “hidden” conscious states of just the kind Block is interested in – states with phenomenology but to which the subject lacks cognitive access. Still, although I am open to this possibility, I don't think that Block's target article mounts an effective empirical case for such states, for the reasons given earlier.
In my book (Papineau Reference Papineau2002), I conjectured that this issue will always be incapable of resolution, and suggested that this is because the very concept of phenomenal consciousness is too indeterminate to allow serious empirical investigation of its boundaries. But now I am not so sure. Block's richly textured use of the empirical data may not amount to a conclusive case for phenomenology without access, but it certainly teaches us a great deal about both. It also makes me less pessimistic about the possibility that further empirical data may cast even more light on the boundaries of phenomenal consciousness.
