The Crimson Veil Lifted: Reconciling Qualia with Computationalism — Epoche C2
Frank Jackson's Mary is a neuroscientist who has spent her life in a black-and-white room, learning through a monochrome monitor everything physics and neurophysiology have to say about what happens when a human being looks at a ripe tomato; the question of "Epiphenomenal Qualia" (1982) is what happens when she is let out. Jackson's answer is that she learns something — what it is like to see red — and that since she already knew every physical fact, what she learns is a fact that is not physical, so physicalism is false. The verdict defended here is the contrary one: that the phenomenal character of an experience is the content of a representational state of the subject, that a complete physical description therefore fixes it, and that Mary acquires no new fact on release. The defence has to be conducted more carefully than the intuition it opposes, because the version of it that circulates informally leaves out three things — what Jackson's argument actually requires, which premise the reply is attacking, and what the reply costs — and each of them changes the commitments one is taking on. The argument, and what it needs to work Set out as premises, the knowledge argument is short. (1) Before release, Mary knows all the physical facts about human colour vision. (2) On release she learns something about human colour vision, namely what it is like to see red. (3) Therefore not all facts about human colour vision are physical facts, and physicalism is false. Jackson supplements Mary with a second case in the same paper: Fred, who reliably sorts into two classes a range of samples that everyone else sees as uniformly red, and about whom no amount of physiological detail tells us which two experiences he is having. Premise (2) is the one that carries the intuition, and the inference from (2) to (3) is where all the work is hidden. It requires two things that are rarely made explicit. It requires that what Mary acquires be knowledge that something is the case, rather than a capacity or an acquaintance; and it requires that facts be individuated coarsely, by what they are about, rather than finely, by the concepts under which they are represented. If facts are individuated finely — if learning the same worldly fact under a new concept counts as learning a new fact — then (2) is true and (3) does not follow, because a physicalist can hold that every worldly fact is physical while denying that every way of representing a fact is a physical concept. Jackson also took on a commitment that is worth naming because the position defended here rejects it. He held that qualia are epiphenomenal: they are caused by physical states but cause nothing themselves. He needed this because a quale with a causal role in producing behaviour would be a property the physical story has to accommodate. In defence of it he answered the obvious evolutionary objection with the point that a trait can be a by-product of one that is selected for — a polar bear's coat is thick because thickness insulates, and its weight comes along without being any advantage — and added the Humean observation that what causes what cannot be settled in advance of investigation. The account given below denies epiphenomenalism outright: a representational state with a role in recognition, discrimination and the guidance of action is causally efficacious, and that is a feature rather than an embarrassment. One point of dialectical hygiene, omitted from the compressed version of this essay and material to how the argument should be read. Jackson himself abandoned the knowledge argument. In "Mind and Illusion" (2003) he holds that the case for physicalism is strong enough that the argument must go wrong somewhere, and locates the error in the assumption that experience has intrinsic features over and above how it represents the world — arriving, that is, at a representationalist position close to the one defended here. The argument therefore has to be met in the hands of those who still hold it, and the sharpest of those hands is the one that presses the objection set out later in this essay. Which premise the reply attacks Given the structure just described, a physicalist has exactly three exits, and it clarifies the debate to see them as competitors rather than as a single "physicalist reply". Exit Verdict on premise (2) What Mary gains Deny the premise False: she learns nothing Nothing; the stipulation that she knew all the physical facts is one whose consequences we cannot in fact imagine Grant it, deny she learns a fact True in the sense of acquiring know-how Abilities: to imagine, remember and recognise the experience Grant it, deny the fact is new True An old fact under a new concept The first is Dennett's, and it deserves a correction here, because the compressed version of this essay listed Consciousness Explained (1991) among its sources as though it supported the position taken. It does not. Dennett's route is more radical and considerably more expensive: he denies that there are qualia as traditionally conceived at all — items that are ineffable, intrinsic, private, and directly apprehensible in consciousness — and so denies the explanandum rather than explaining it. Anyone who says, as this essay does, that Mary's experience of red is real and has a determinate character has declined that exit. The second is Lewis's ability hypothesis: knowing what an experience is like just is possessing a cluster of abilities, so Mary gains know-how and no information, and (3) fails because knowledge-that was never at issue. The third is Loar's: phenomenal concepts are recognitional concepts, formed on the basis of a sample and applied without the mediation of any description, and they refer to physical or functional properties while remaining conceptually independent of any physical description of them. Mary comes to know a fact she already knew, now under such a concept. The position argued for here is the third, with the second as a consequence