Symposium · Philosophy of Mind · Pl. 10Frank Jackson · b. 1943
Mary's Room
A scientist who has never seen red
Jackson imagines Mary, a brilliant scientist confined from birth to a room with no color in it at all, black, white, and shades of gray, studying the physics and neuroscience of color perception through black-and-white books and monitors until she has learned, by hypothesis, every physical fact there is to know about how color vision works. She knows the wavelength of light a ripe tomato reflects, which cones it excites, which cortical areas fire, and which words a person will use afterward. There is no gap in her knowledge, by stipulation. Then the door opens. Jackson's whole argument rests on one question about that moment: does she say, oh, so that is what red is like?
The world it came from
Jackson published Epiphenomenal Qualia in 1982 in the Philosophical Quarterly, and a shorter follow-up, What Mary Didn't Know, in 1986. He was at Monash University in Melbourne, in a philosophical culture where physicalism was close to orthodoxy. J. J. C. Smart and David Armstrong, both Australians, had built the identity theory of mind, on which every mental state is a brain state. Thomas Nagel had argued in 1974 that no objective description could capture what it is like to be a bat. Jackson wanted something tighter than Nagel's appeal to the bat's alien point of view: a case where all the physical information is on the table and something is still missing. The room, he said, is stipulated to be complete so that the question cannot be dodged.
The idea
She knows all the physical facts about us and our environment. And yet, when Mary finally leaves the room and sees a ripe tomato for the first time, it seems she learns something new: what it is actually like to see red. If she genuinely learns something new despite already knowing every physical fact, then physical facts alone cannot be the whole story of conscious experience. The point is not that Mary lacked information because her books were poor. She had every fact that physics, chemistry, and neuroscience could ever supply. What she lacked was a kind of fact that none of those sciences state. Jackson called what she gained knowledge of qualia, the felt qualities of experience, and concluded they are not physical.
How the reasoning runs
Premise one: before her release Mary knows every physical fact about human color vision. Premise two: when she sees red she learns something, namely what seeing red is like. Step three: if she learns something, there was a fact she did not know. Step four: that fact was not among the physical facts, because she already had all of those. Conclusion: there are facts that are not physical facts, and physicalism, the view that everything is physical, is false. Jackson drew a further conclusion that gave the paper its title. If these qualia are not physical, they cannot cause anything physical, so they are epiphenomenal: felt but idle, riding on the brain without steering it. Most readers accept the intuition in premise two and fight over step four.
Seen in life
You have read everything about durian. You know the sulfur compounds, the receptor sites they bind to, the brain regions that light up, the reason one person calls it custard and another calls it drains. You could pass any exam on it. Then someone puts a piece in your mouth. Whatever happens next, it is not a review of what you knew. Jackson's Mary is that moment with all the science stipulated in and nothing left to read. The same scene runs with AI. A language model has processed every description of red ever written. Ask whether it knows what red is like and most people hesitate, and the hesitation is the intuition Jackson was pumping.
Why it matters
This is one of the sharpest modern challenges to the claim that a complete physical description of the world would leave nothing further to know. If Mary's new experience really counts as new knowledge, some facts about consciousness, the subjective, felt quality of experience, appear to resist being fully captured by physical science, however complete that science becomes. The argument's strength is its economy. It needs no bats and no zombies, only a scientist and a door. Whoever wants to hold that the mind is nothing over and above the brain must explain what happened when the door opened, and every explanation on offer has a cost. That is why the case is still the standard test of a theory of consciousness forty years on.
“She knows all the physical facts about us and our environment.”
From the source
Jackson sets up the case directly: “She knows all the physical facts about us and our environment.” The question the thought experiment then turns on is whether her first sight of red still teaches her something that all of that physical knowledge had somehow left out. The 1982 paper opens with a different case, Fred, who can see two shades of red where everyone else sees one, and asks what we would learn if we could see as Fred does. Mary is the second example, and the one that stuck. Jackson describes her as investigating the world from a black-and-white room by way of a black-and-white television monitor, and specializing in the neurophysiology of vision. He calls the resulting argument the knowledge argument.
The pushback
Defenders of physicalism argue Mary does not learn a new fact at all, only a new way of representing an old fact she already knew, gaining a new ability, to recognize and imagine red, rather than genuinely new information about the world beyond what her physical knowledge already contained. That is the ability hypothesis, from David Lewis and Laurence Nemirow. Daniel Dennett pressed a harder line: the story only seems obvious because we cannot imagine knowing everything. A Mary who really had every physical fact would recognize red on sight, say so, and not be surprised, and our sense that she must be surprised is a failure of imagination dressed as an intuition.
The reply
Jackson's 1986 reply to the ability hypothesis is that Mary gains more than skills. On release she learns something about other people: that their experience all along had a character she had not grasped, and that her old conception of their minds was incomplete. That is a truth about them, not a knack of hers. Against Dennett the reply is simply that the intuition is very robust; few readers can make themselves believe Mary would shrug. But the tradition's answer is only partial, and Jackson himself is the proof. In the late 1990s he changed his mind, decided that the intuition is generated by how experience represents itself, and became a physicalist. The argument outlived its author's belief in it.
Where it went next
The knowledge argument joined Nagel's bat and, from 1996, David Chalmers's zombies as the three standard arguments against physicalism, and Chalmers's phrase the hard problem of consciousness names what all three point at. Brian Loar's phenomenal concepts strategy, the claim that Mary gains a new concept of an old physical fact, became the leading physicalist reply. Dennett gave his version a robot, RoboMary, who works out what red is like from the inside. The Norwegian vision scientist Knut Nordby, who was born without color vision, wrote about being a real-life Mary and doubted he could ever know. Jackson's own recantation, published in the 2000s, is now part of the case. The 2014 film Ex Machina put the thought experiment in a screenplay.
Where this leads
What is it like to be a bat? supplies the deeper background this thought experiment builds directly on, Nagel's own earlier argument that subjective experience resists full capture by objective, physical description, no matter how complete that description eventually becomes. The Chinese Room is a cousin from the other side of the mind. Searle's man in the room has all the syntax and none of the meaning; Jackson's Mary has all the physics and none of the feel. Both arguments work by stipulating that one kind of thing is complete and asking whether something else is still missing. Read together, they mark the two places where a full description of a system seems to stop short of a mind: understanding, and experience.