Mary's Room & the Knowledge Argument

Part 17 separated two refusals of Turing's substitution. Searle's room is about aboutness: following a rulebook is not understanding. Nagel and Chalmers are about experience: a complete account of discrimination and report still leaves out what it is like. Those are different claims. There is a third, and it is routinely folded into the second.

Jackson's 1982 thought experiment looks like a restatement of the hard problem. It is not. The hard problem asks why there is experience at all. Mary asks whether a physically complete story still leaves a kind of knowledge out. You can answer one without answering the other. Keeping them apart is most of the work.

The setup

Mary is a colour scientist. She has lived her entire life in a black-and-white room, learning everything physical there is to learn about colour vision: wavelengths, cone responses, the visual cortex, the reports people give when they see a ripe tomato. Jackson's stipulation is that this knowledge is complete. There is no physical fact about colour she does not already have.

Then she is released, and she sees red.

The intuition the argument needs is that she learns something. If she does, then complete physical knowledge was not complete knowledge. That is the knowledge argument.

Notice what has not been claimed yet. Nothing has been said about whether experience is a non-physical property, or whether physicalism about the mind is false. Those are further moves. The thought experiment, if it works at all, shows an epistemic remainder: a kind of knowing that the physical story did not deliver. Ontology is a separate inference, and it is the one Jackson later took back.

What she is supposed to lack

Before release, Mary can predict any physical fact you like. She can say which wavelength will strike which retina, which neurons will fire, which word a subject will utter. After release she has an experience, and the claim is that this is new knowledge — knowledge of what red is like.

The hard problem would be: why is there something it is like at all, rather than the processing happening in the dark? Mary does not ask that. She already grants that other people have colour experience; that is part of the physical-and-functional story she has mastered. Her question is first-person and epistemic: given all of that, is there still something she does not know?

That is closer to Nagel's bat than it looks, but it is pointed at knowledge rather than at the existence of a subjective point of view. Nagel says the bat's experience is not captured by any objective description. Jackson says: make the objective description complete, give it to someone who has never had the experience, and see whether anything is still missing. Completeness is the extra turn of the screw.

Jackson's leap, and the recantation

From "Mary learns something" Jackson originally inferred that physicalism is false: there are facts about the world that are not physical facts. The inference is not free. It requires that what she learns is a fact about the world, rather than a new way of standing toward facts she already had.

Jackson later accepted physicalism and treated the original intuition as misleading. That recantation is often cited as if it cancelled the thought experiment. It cancelled one conclusion. The setup still isolates a question that physicalism has to answer: if Mary learns something, what kind of something is it — and if she doesn't, why does it so stubbornly seem that she does?

The replies

Three families of reply, and they are not equivalent.

The ability hypothesis (Nemirow, Lewis) says she gains know-how, not a new fact. She learns to recognise red, to imagine it, to remember it. Abilities are not the sort of thing a complete book of facts was supposed to contain, so physicalism about facts is untouched. What this reply owes: an account of why the intuition feels like learning a fact, and why "what it is like" talks like a that-clause rather than like riding a bicycle.

The acquaintance reply says she gains a new mode of knowing the same facts. Before, she knew about red by description; now she knows it by having it. No new property in the world; a new epistemic relation. What this reply owes: a distinction between modes of knowing that does not smuggle a new fact in through the back door. If acquaintance is just more knowledge-that under another name, the argument returns.

The new-fact reading accepts that she learns a fact she did not have, and then has to say what kind of fact it is. If it is a fact about a non-physical property, physicalism is in trouble. If it is a fact about her own representational state — she now has a phenomenal concept she lacked — then the new fact may be old physics under a new mode of presentation. That last move is acquaintance with better vocabulary, and it is where a lot of contemporary physicalists park.

None of these is a knockout. Each relocates the remainder.

Why the third cut matters

These come apart from the other two, and that is the whole reason to keep them separate.

You could hold that Mary learns nothing new — the intuition is a mistake — and still think the hard problem is unsolved, because "why is there experience?" is not settled by denying an epistemic gap. And you could grant Mary a genuine epistemic remainder while denying Searle: her states, once she has seen red, are about redness in whatever way a functionalist needs, and the leftover is only about her knowledge, not about aboutness.

So when someone says a complete physical theory of the mind would still "leave something out," it is worth asking which leftover. A leftover in the inventory of the world is Jackson's original ontological claim. A leftover in what a person in Mary's position would know is the epistemic claim, and it can be true while the inventory stays physical. A leftover in aboutness is Searle, and it does not need colour, or Mary, or a black-and-white room.

The ancient cut underneath

Part 3 already had grades of knowing. Plato's line is not a list of objects; it is a claim that the same territory can be held in different ways — images, things, mathematical objects, the forms — and that moving up the line is not always acquiring a new item so much as a new grip. Mary's room is that claim with the metaphysics stripped out and colour put in.

Aristotle's De Anima hesitated at nous because intellect did not behave like the rest of the living thing's organisation. Jackson's hesitation, before he recanted, was at the same joint from the other side: everything about colour vision can be given as organisation and activity, and then there is the knowing of what red is like, which did not seem to behave like the rest of the story. The later recantation says: it does behave like the rest of the story, once you are careful about what "knowing" was doing. Whether that care is enough is the live dispute.

Twenty-four centuries later the map is still being refined rather than closed. Part 17 named the residue in experience. This part names it in knowledge. Same joint, different instrument.