Mary knows every physical fact about human color vision, but has never experienced color. When she first sees red, it seems she learns what red looks like. The dispute is whether this is a new fact about reality, a new ability, direct acquaintance with an experience, or a new way of knowing something physical she already understood.
The crucial step comes after admitting that Mary’s understanding changes: does that change reveal something nonphysical, or can physicalism explain the new understanding? The competing replies disagree about that inference, not merely about whether seeing red differs from reading about it.
Give Mary the full premise
Frank Jackson introduced Mary in his 1982 paper “Epiphenomenal Qualia”. She studies vision through a black-and-white monitor and has all the relevant physical information. Her education includes the processes connecting light, the nervous system and people’s descriptions of what they see.
This is an idealized philosophical setup. Mary is not a present-day neuroscientist who happens to have missed a few experiments. Explaining that current textbooks leave things out would be an answer to a much easier question. Jackson grants her the completed physical story, then asks whether experiencing color would still teach her something.
Imagine her first encounter is a red ceramic cup. She can already predict the physical events involved in viewing it. Does she nevertheless discover an aspect of the event that prediction could not provide? The cup is our illustration; the question does not depend on which red object she meets.
Separate learning from discovering a nonphysical fact
Jackson’s challenge to physicalism can be put in three steps. Mary has all physical information before release. She learns something about color experience afterward. Therefore, the argument proposes, the physical information was not everything there was to know.
Physicalism holds, broadly, that reality is entirely physical; applied here, conscious experience does not require additional nonphysical properties. Qualia names the felt qualities at issue, such as the appearance of red or the painfulness of pain.
The vulnerable step is what follows from Mary’s learning. Does any new knowledge require a previously unknown fact? Or can someone acquire a new relation or perspective on a fact they already know? The Stanford Encyclopedia’s discussion distinguishes these weaker and stronger conclusions. A change in a person’s knowledge does not, without further premises, establish a nonphysical addition to the world.
This is also why the thought experiment does not establish a ghost inside a brain. Its target is a claim about what physical facts exhaust. It supplies neither an observed disembodied mind nor a working alternative neuroscience.
The ability reply: experience changes what Mary can do
In “What Experience Teaches”, David Lewis argues that knowing what an experience is like involves abilities to remember, imagine and recognize it. On this view, lessons can supply information without supplying those capacities.
Mary’s first red experience changes her practical repertoire. She can later call the appearance to mind or recognize its recurrence. Her gain need not consist of a further fact missing from the physical account.
Consider a different, ordinary case: following instructions for folding a complicated paper bird. Knowing which fold comes next and being able to execute it smoothly can come apart. The example makes room for the distinction between information and ability. It does not prove that Mary’s complete knowledge would leave her unable to imagine red; that is precisely what the philosophical reply must defend.
Lewis also recognizes that information and abilities usually arrive together. Showing that Mary gains capacities is therefore not yet enough. The ability theorist needs the stronger claim that no special nonphysical information is required to explain what she gains.
Acquaintance: meeting what description was already about
Another reply, associated with Earl Conee, treats acquaintance as a distinct kind of knowledge. Mary becomes directly familiar with a quality by experiencing it. The quality itself may still be physical.
Think of researching a particular sculpture and then standing beside it. The later encounter creates a new relation to the object. A complete archive and a visit are different activities, even if there is only one sculpture. This analogy clarifies the proposed relation; a real visitor will probably also discover facts absent from the archive, so it cannot decide Mary’s idealized case.
The challenge returns when Mary says that she has discovered something. Can direct familiarity account for that apparent informativeness without new factual knowledge? Calling the gain acquaintance identifies a candidate answer, but the label alone does not establish its sufficiency.
An old fact under a new concept
A related physicalist approach allows new factual knowledge while rejecting a new nonphysical fact. Mary previously understood a state through physical concepts. After seeing red, she can think about that same state through a first-person concept: this kind of experience.
Consider an invented case in which you know a building thoroughly from survey plans and also recognize it from a street view, then discover the two presentations identify the same building. Your understanding changes without another building appearing. The proposal about Mary similarly distinguishes the fact from how it is grasped.
The objection is whether the analogy really fits complete physical knowledge. If Mary could not make the connection beforehand, has the physical description omitted something, or does using the new concept require a particular physical experience? This reply needs an account of concepts and their acquisition, rather than merely insisting that the underlying state is physical.
Would complete knowledge leave her surprised?
The Internet Encyclopedia of Philosophy discusses another pressure on the argument: perhaps our reaction underestimates the premise. Dennett-style objections ask whether a person with genuinely complete physical knowledge would lack the relevant understanding in the first place.
None of us can inhabit Mary’s stipulated intellectual position. Imagining an exceptionally well-read researcher and finding her surprise obvious does not establish what an ideal knower would do. Conversely, saying she could predict everything does not automatically show that prediction supplies the experience itself. Both sides owe an explanation of that step.
Jackson later rejected his own anti-physicalist conclusion, as the encyclopedia records. That historical change did not settle the argument. Its force depends on the premises and replies, rather than on its originator’s final vote.
When comparing answers, keep two questions separate: what changes for Mary, and what must reality contain for that change to occur? The hard problem of consciousness asks why physical processes have an experiential aspect. Mary’s room sharpens one part of that problem by asking whether a complete description and a first encounter provide the same kind of understanding.
