Skip to content
Coordinates of ThoughtHome of Philosophy
Join

Knowledge argument

Updated 19 November 2026

On this page
  1. The argument at a glance
  2. What it means
  3. The questions it belongs to
  4. Thinkers
  5. Traditions
  6. To read
  7. Sources and further reading
  8. About this page
  9. Cite this page

Frank Jackson's argument that physicalism is incomplete: a scientist who knows every physical fact about colour vision, but has never seen colour, seems to learn something new when she first sees red.

Also: Mary's room, Mary the colour scientist, Mary argument

The argument at a glance

A guide to the entry below, quoting its own words. Skip to the full entry and its sources.

Question
Can someone who knows every physical fact about colour vision still fail to know something about colour experience, and, if so, does it follow that some facts are not physical? Provenance: the argument originated with Frank Jackson, in 'Epiphenomenal Qualia', The Philosophical Quarterly 32, no. 127 (1982), pp. 127-136, and he defended it in 'What Mary Didn't Know', The Journal of Philosophy 83, no. 5 (1986), pp. 291-295. Jackson himself later abandoned its anti-physicalist conclusion (see the paragraph on his change of view below), so this entry describes the argument, not its author's present position.
In the text’s own form , Philosopher’s own formulation, as reported
Show the passage (227 words)

Jackson's 1982 paper offers two cases, not one, and presents them as appeals to intuition rather than as a numbered proof. As Nida-Rümelin and O Conaill report, he first describes Fred, a person who sees a colour unknown to normal human perceivers; no amount of knowledge about what happens in Fred's brain, the case suggests, would tell us what that colour is like for him. Mary, confined to a black-and-white room and specialising in the neurophysiology of vision, makes the same point for an ordinary colour: on her release she appears to learn something that her complete physical information did not contain. In 1982 Jackson joined this to epiphenomenalism, the view that such qualities of experience have no physical effects, as the paper's title indicates; the Stanford entry notes that he was for a time one of the very few philosophers to embrace that view. In 'What Mary Didn't Know' he defended the argument against critics. Brie Gertler reports Jackson's answer to Paul Churchland, who had taken what Mary gains to be a kind of acquaintance with what it is like to see red and had argued that a parallel argument would tell equally against dualism: one could know all the physical facts about seeing red without knowing what it is like, Jackson replied, but one could not know all the facts, physical and non-physical, without knowing it.

Reasons , Reconstruction by the scholar named in the entry

In 'Epiphenomenal Qualia' (1982) Frank Jackson imagined Mary, a brilliant neuroscientist raised in a black-and-white room, who learns every physical fact about human colour vision. When she is released and sees a ripe tomato, she seems to learn something: what it is like to see red. As set out in the Stanford Encyclopedia entry by Martine Nida-Rümelin and Donnchadh O Conaill, the argument runs:

  1. before her release Mary has all the physical information about colour vision;
  2. before her release she lacks some information about colour vision;
  3. Therefore: not all information is physical. (the entry continues below)
Intended conclusion
Jackson's intended conclusion in 1982 was ontological: physicalism, the thesis that all facts are physical, is false. The two versions show why the step to that conclusion is the real issue. The Stanford entry treats the weaker conclusion as merely epistemological, and compatible with denying that there are any non-physical facts; only the stronger conclusion is one the physicalist must reject. The replies listed below can be sorted by where they block the step from what Mary gains to a non-physical fact: the ability and acquaintance hypotheses deny that she gains new propositional knowledge at all; the old-fact, new-guise view grants new knowledge but denies that it is knowledge of a new fact; and Dennett denies that the case, as described, can be coherently imagined.
Objections and replies
The main replies

(1) The ability hypothesis of Laurence Nemirow and David Lewis: Mary acquires no new information, only new abilities to recognise, imagine and remember; knowing what it is like is knowing how. Jason Stanley and Timothy Williamson's view that knowing how is a kind of knowing that challenges an assumption of this reply, though Yuri Cath argues that the ability hypothesis can accommodate it. (2) The acquaintance hypothesis of Earl Conee: Mary gains acquaintance with a quality, not factual knowledge. (3) Old fact, new guise: Mary gains a new concept of a physical fact she already knew, a phenomenal concept that refers directly (Brian Loar, Michael Tye, David Papineau, John Perry and others); David Chalmers doubts that this avoids reintroducing non-physical properties, while Ned Block defends the view against such objections. (4) Daniel Dennett denies the case can be properly imagined: a Mary who really knew everything physical could work out colour experience in advance. (5) Jackson himself later rejected the argument, concluding that epiphenomenal qualia would make our knowledge of them impossible and that a representationalist account of experience defuses the case.

Objections , Raised by other thinkers and critics, as reported

texts and dates. The ability hypothesis was put forward by Laurence Nemirow (1980; 'Physicalism and the Cognitive Role of Acquaintance', 1990) and David Lewis (postscript to 'Mad Pain and Martian Pain', 1983; 'What Experience Teaches', 1988); in Nemirow's formulation, knowing what an experience is like is knowing how to imagine having it. Earl Conee's acquaintance hypothesis appeared in 'Phenomenal Knowledge', Australasian Journal of Philosophy 72 (1994). Brian Loar's 'Phenomenal States' (1990, revised 1997) treats phenomenal concepts as recognitional concepts: to have the phenomenal concept of an experience is to be able to recognise such experiences while having them. Paul Churchland criticised the argument in 'Reduction, Qualia, and the Direct Introspection of Brain States', Journal of Philosophy 82 (1985), which the Stanford entry counts among new-knowledge, old-fact positions and Torin Alter's Internet Encyclopedia article among views on which Mary gains non-propositional knowledge that does not fit folk categories, and again in 'Knowing Qualia: A Reply to Jackson' (1989). Daniel Dennett's challenge, in Consciousness Explained (1991) and 'What RoboMary Knows' (2007), is illustrated by his blue banana trick: a Mary with complete physical knowledge who was shown a blue banana on her release would know at once that it had the wrong colour.

Replies

Defenders of the argument have answered each strategy, and the Stanford entry gives separate sections to objections against the ability hypothesis, to objections against the new-knowledge, old-fact view, and to the dualist reading of the argument. Jackson's own reply to Churchland, described above, is the principal reply by the argument's author while he still held it.

Coordinate 1.3.4Coordinate 1.3.1Frank JacksonC. D. BroadThomas NagelDavid LewisTimothy WilliamsonDavid ChalmersDaniel DennettPaul ChurchlandNed BlockSources and primary textsRead the full entry

What it means

Question. Can someone who knows every physical fact about colour vision still fail to know something about colour experience, and, if so, does it follow that some facts are not physical? Provenance: the argument originated with Frank Jackson, in 'Epiphenomenal Qualia', The Philosophical Quarterly 32, no. 127 (1982), pp. 127-136, and he defended it in 'What Mary Didn't Know', The Journal of Philosophy 83, no. 5 (1986), pp. 291-295. Jackson himself later abandoned its anti-physicalist conclusion (see the paragraph on his change of view below), so this entry describes the argument, not its author's present position.

The argument. In 'Epiphenomenal Qualia' (1982) Frank Jackson imagined Mary, a brilliant neuroscientist raised in a black-and-white room, who learns every physical fact about human colour vision. When she is released and sees a ripe tomato, she seems to learn something: what it is like to see red. As set out in the Stanford Encyclopedia entry by Martine Nida-Rümelin and Donnchadh O Conaill, the argument runs: (1) before her release Mary has all the physical information about colour vision; (2) before her release she lacks some information about colour vision; therefore (3) not all information is physical. A stronger version concludes that there are non-physical facts, which physicalism denies. Precursors include C. D. Broad's archangel, who could not predict the smell of ammonia from its microstructure, Herbert Feigl's Martian, and Thomas Nagel's question of what it is like to be a bat, although Nagel did not claim that physicalism is false.

The reasoning in the text's own form. Jackson's 1982 paper offers two cases, not one, and presents them as appeals to intuition rather than as a numbered proof. As Nida-Rümelin and O Conaill report, he first describes Fred, a person who sees a colour unknown to normal human perceivers; no amount of knowledge about what happens in Fred's brain, the case suggests, would tell us what that colour is like for him. Mary, confined to a black-and-white room and specialising in the neurophysiology of vision, makes the same point for an ordinary colour: on her release she appears to learn something that her complete physical information did not contain. In 1982 Jackson joined this to epiphenomenalism, the view that such qualities of experience have no physical effects, as the paper's title indicates; the Stanford entry notes that he was for a time one of the very few philosophers to embrace that view. In 'What Mary Didn't Know' he defended the argument against critics. Brie Gertler reports Jackson's answer to Paul Churchland, who had taken what Mary gains to be a kind of acquaintance with what it is like to see red and had argued that a parallel argument would tell equally against dualism: one could know all the physical facts about seeing red without knowing what it is like, Jackson replied, but one could not know all the facts, physical and non-physical, without knowing it.

About the numbering. The numbering in the first paragraph above follows the Stanford reconstruction by Nida-Rümelin and O Conaill; it is not Jackson's own. The same entry separates two versions. In the weaker version Mary has complete physical knowledge before her release, there is some kind of knowledge she lacks, and so there is non-physical knowledge. In the stronger version she knows all the physical facts, there are facts she does not know, and so there are non-physical facts.

Intended conclusion. Jackson's intended conclusion in 1982 was ontological: physicalism, the thesis that all facts are physical, is false. The two versions show why the step to that conclusion is the real issue. The Stanford entry treats the weaker conclusion as merely epistemological, and compatible with denying that there are any non-physical facts; only the stronger conclusion is one the physicalist must reject. The replies listed below can be sorted by where they block the step from what Mary gains to a non-physical fact: the ability and acquaintance hypotheses deny that she gains new propositional knowledge at all; the old-fact, new-guise view grants new knowledge but denies that it is knowledge of a new fact; and Dennett denies that the case, as described, can be coherently imagined.

The main replies. (1) The ability hypothesis of Laurence Nemirow and David Lewis: Mary acquires no new information, only new abilities to recognise, imagine and remember; knowing what it is like is knowing how. Jason Stanley and Timothy Williamson's view that knowing how is a kind of knowing that challenges an assumption of this reply, though Yuri Cath argues that the ability hypothesis can accommodate it. (2) The acquaintance hypothesis of Earl Conee: Mary gains acquaintance with a quality, not factual knowledge. (3) Old fact, new guise: Mary gains a new concept of a physical fact she already knew, a phenomenal concept that refers directly (Brian Loar, Michael Tye, David Papineau, John Perry and others); David Chalmers doubts that this avoids reintroducing non-physical properties, while Ned Block defends the view against such objections. (4) Daniel Dennett denies the case can be properly imagined: a Mary who really knew everything physical could work out colour experience in advance. (5) Jackson himself later rejected the argument, concluding that epiphenomenal qualia would make our knowledge of them impossible and that a representationalist account of experience defuses the case.

Objections: texts and dates. The ability hypothesis was put forward by Laurence Nemirow (1980; 'Physicalism and the Cognitive Role of Acquaintance', 1990) and David Lewis (postscript to 'Mad Pain and Martian Pain', 1983; 'What Experience Teaches', 1988); in Nemirow's formulation, knowing what an experience is like is knowing how to imagine having it. Earl Conee's acquaintance hypothesis appeared in 'Phenomenal Knowledge', Australasian Journal of Philosophy 72 (1994). Brian Loar's 'Phenomenal States' (1990, revised 1997) treats phenomenal concepts as recognitional concepts: to have the phenomenal concept of an experience is to be able to recognise such experiences while having them. Paul Churchland criticised the argument in 'Reduction, Qualia, and the Direct Introspection of Brain States', Journal of Philosophy 82 (1985), which the Stanford entry counts among new-knowledge, old-fact positions and Torin Alter's Internet Encyclopedia article among views on which Mary gains non-propositional knowledge that does not fit folk categories, and again in 'Knowing Qualia: A Reply to Jackson' (1989). Daniel Dennett's challenge, in Consciousness Explained (1991) and 'What RoboMary Knows' (2007), is illustrated by his blue banana trick: a Mary with complete physical knowledge who was shown a blue banana on her release would know at once that it had the wrong colour.

Jackson's change of view. Jackson no longer accepts the argument or its anti-physicalist conclusion. Torin Alter's Internet Encyclopedia article dates the change to the late 1990s and reports that Jackson now defends physicalism and rejects the knowledge argument. The Stanford entry traces two stages. In a 1995 postscript to 'What Mary Didn't Know' he argued that knowledge about qualia is impossible if qualia are epiphenomenal, and concluded that something must be wrong with the argument; in work of 2003 and 2007 he argued that the argument goes wrong in presupposing a false view of sensory experience, and he adopted strong representationalism, on which to be in a phenomenal state is to represent objective properties. Alter's bibliography identifies the 2003 paper as 'Mind and Illusion', in Anthony O'Hear (ed.), Minds and Persons, Royal Institute of Philosophy Supplement 53 (Cambridge University Press, 2003), pp. 251-271; the 2007 paper is 'The Knowledge Argument, Diaphanousness, Representationalism'. Alex Byrne, reviewing There's Something About Mary in Notre Dame Philosophical Reviews, names Jackson's 'Postscript on Qualia' and 'Mind and Illusion' as the papers that explain why Jackson now disavows the argument. Byrne reports that in the postscript Jackson's reason for changing his mind is that what Mary learns must not outrun how things are physically, and that Jackson suggests the intuition that Mary learns something can be explained away, because sensory experience is putative information about relational and functional facts; Byrne adds that this leaves the argument's mistake undiagnosed, and that the line of thought is elaborated in 'Mind and Illusion', where, in Byrne's report, Jackson agrees with Nemirow and Lewis about what happens to Mary on her release. Alter also cites a 2019 paper, 'The Knowledge Argument Meets Representationalism about Colour Experience'. Gertler adds that the change did not make him an a posteriori physicalist: he now holds that phenomenal knowledge is deducible in principle from physical knowledge, even if we may be unable to perform the deduction.

Replies. Defenders of the argument have answered each strategy, and the Stanford entry gives separate sections to objections against the ability hypothesis, to objections against the new-knowledge, old-fact view, and to the dualist reading of the argument. Jackson's own reply to Churchland, described above, is the principal reply by the argument's author while he still held it.

Key thinkers and texts. Primary: Jackson, 'Epiphenomenal Qualia' (1982) and 'What Mary Didn't Know' (1986). Change of view: Jackson's 1995 postscript, the 'Postscript on Qualia' discussed by Byrne, 'Mind and Illusion' (2003), and the papers of 2007 and 2019. Precursors, as reported in the Stanford entry: C. D. Broad, The Mind and its Place in Nature (1925), p. 71, on the mathematical archangel; Herbert Feigl, 'The Mental and the Physical' (1958); Thomas Nagel, 'What Is It Like to Be a Bat?', Philosophical Review 83 (1974), pp. 435-450. Critics: Nemirow, Lewis, Conee, Loar, Churchland and Dennett, as above. A collection of the debate is There's Something About Mary, edited by Peter Ludlow, Yujin Nagasawa and Daniel Stoljar (MIT Press, 2004).

Current status. The debate is unresolved, and its originator has changed sides. In Alter's words, "others defend the argument, and even those who reject it often disagree about where it goes awry." The knowledge argument should be kept distinct from its neighbours. The zombie argument reasons from conceivability to possibility rather than from incomplete knowledge, and the explanatory gap names a limit on explanation which Joseph Levine, who introduced the term, does not treat as establishing a metaphysical conclusion.

On this site it belongs to Coordinate 1.3.4 (Qualia and the Hard Problem) and bears on 1.3.1 (The Mind–Body Problem).

See also the glossary entries Qualia, Physicalism, Hard problem of consciousness, Epiphenomenalism, Zombie argument, Knowledge-how and Thought experiment.

The questions it belongs to

Thinkers

Traditions

To read

Compare this argument

Sources and further reading

Responsibility: Coordinates of Thought Editors. Checked against the sources above on 22 November 2026. Not peer-reviewed by outside specialists. How we write and check.

Cite this page

Responsibility: Coordinates of Thought Editors. Each statement was drafted from the sources listed above and checked against them in a separate pass before publishing; this page is not peer-reviewed by outside specialists. Source-checked 22 November 2026. How to cite this site · How we write and check.

APA (7th edition)

Coordinates of Thought Editors. (2026, November 22). Knowledge argument (Arguments). Coordinates of Thought. Retrieved [date accessed], from https://coordinatesofthought.com/arguments/knowledge-argument/

MLA (9th edition)

“Knowledge argument (Arguments).” Coordinates of Thought, 22 Nov. 2026, coordinatesofthought.com/arguments/knowledge-argument/. Accessed [date accessed].

Chicago (17th edition)

Coordinates of Thought Editors. “Knowledge argument (Arguments).” Coordinates of Thought. Source-checked November 22, 2026. https://coordinatesofthought.com/arguments/knowledge-argument/.

BibTeX

@misc{cot-glossary-knowledge-argument-2026,
  author       = {{Coordinates of Thought Editors}},
  title        = {Knowledge argument (Arguments)},
  howpublished = {Coordinates of Thought},
  year         = {2026},
  month        = nov,
  url          = {https://coordinatesofthought.com/arguments/knowledge-argument/},
  note         = {Source-checked 22 November 2026}
}

RIS

TY  - ELEC
AU  - Coordinates of Thought Editors
TI  - Knowledge argument (Arguments)
T2  - Coordinates of Thought
PB  - Coordinates of Thought
PY  - 2026
DA  - 2026/11/22
UR  - https://coordinatesofthought.com/arguments/knowledge-argument/
LA  - en
ER  - 

For reference managers (Zotero, Mendeley, EndNote): BibTeX file · RIS file

All arguments · The Philosophy Glossary