Philosophy of Mind
A beginner's map of the main theories of what the mind is
You drink a glass of wine. Within twenty minutes your mood lifts, your worries recede, your judgement gets a little looser. A chemical – ethanol, a small molecule of carbon, hydrogen and oxygen – has crossed into your bloodstream and changed what it is like to be you.
Now run it the other way. You decide to raise your hand. A private, invisible act of will, something nobody could find by cutting you open, and yet a kilogram of flesh and bone moves through the air.
Every theory of mind is an attempt to explain how these two everyday miracles are possible. How can a physical thing produce experience? How can an experience move a physical thing? Are they even two things at all? This article is a map of the main answers. None of them is universally accepted, and the disagreement is not a sign that philosophers are incompetent. It is a sign that the problem is genuinely hard – perhaps the hardest we have.
The problem, stated plainly
Philosophers call it the mind-body problem. The mind seems to have properties the body does not. Thoughts are about things (your thought about Paris is about Paris; no lump of matter is “about” anything). Experiences have a felt quality – the redness of red, the ache of an ache – that no physical description seems to capture. Your mind is private in a way your body is not: a surgeon can see your brain, but not your thoughts. And you seem to know your own mind directly, without evidence, in a way you know nothing else.
Yet the mind also seems utterly dependent on the body. Damage the brain and you damage the mind – memory, personality, language, the lot. Alter its chemistry and you alter mood, perception and belief. Everything neuroscience has learned points to an intimate, lawlike connection between the two.
So the mind seems both radically unlike the body and inseparable from it. Every theory below is a way of resolving that tension, usually by giving up one of the appearances. The question is always: which appearance are you prepared to sacrifice?
Four test cases
To compare the theories, it helps to have a small set of examples we can run each of them against. These four recur throughout philosophy of mind, and we will come back to them.
The wine glass. Chemistry changes experience; decisions move limbs. A good theory should explain how mind and body act on each other – or explain away the appearance that they do.
Mary the colour scientist. Frank Jackson’s 1982 thought experiment. Mary has lived her whole life in a black-and-white room. She knows every physical fact about colour vision – wavelengths, retinas, neural pathways, all of it. One day she walks out and sees a red tomato. Does she learn something new? Most people think she does: she learns what red looks like. If so, there is a fact about experience that was not contained in all the physical facts. A good theory must say what that fact is, or why the intuition is mistaken.
The octopus. An octopus has a nervous system built on a completely different plan from ours – most of its neurons are in its arms, not its head. It solves puzzles, plays, and appears to feel pain. If it does, then pain cannot be the particular brain process that produces pain in humans, because the octopus does not have that process. A good theory should allow that very different physical systems can share the same mental states – or explain why they cannot.
The robot. Suppose we build a machine that behaves exactly like a person – converses, jokes, flinches from harm, says it is in pain. Is anyone home? Does it have experiences, or is it “all dark inside”? A good theory should give us some principled way to answer, since the question is no longer hypothetical.
Keep these four in mind. They are the obstacle course.
The big divide: dualism and monism
The first fork in the road is a question of counting. How many fundamentally different kinds of thing are there in a human being?
Dualism says two: the physical (the body, the brain, matter in general) and the mental (the mind, the soul, consciousness), and neither can be reduced to the other.
Monism says one. Either everything is ultimately physical (materialism or physicalism), or everything is ultimately mental (idealism), or both mind and matter are aspects of some third, more basic kind of stuff (neutral monism).
Almost every theory in philosophy of mind is a version of one of these. We will take dualism first, since it is where the modern debate began.
Substance dualism
The idea. Mind and body are two distinct substances – two things that can each exist on their own. The body is extended in space, divisible and public. The mind is unextended, indivisible, private, and essentially thinking. The classic statement is René Descartes' in the Meditations (1641). Descartes argued that he could clearly conceive of himself existing without a body, but not without thought; so his essence was thought, and he was “a thing that thinks,” distinct from the machine he inhabited.
This is the view most people find intuitive before they study philosophy, and it fits most religious traditions, which hold that the soul survives the body’s death.
What it explains well. Substance dualism takes the appearances seriously. Privacy, the felt quality of experience, the sense that you are a single unified self rather than a collection of neurons – all of these are exactly what you would expect if the mind were a distinct, non-physical thing. It handles Mary effortlessly: of course she learns something new, because facts about experience are facts about a different kind of substance. And it makes room for free will, personal survival after death, and the intuition that you are not just your body.
Where it struggles. The wine glass is the killer. If mind and body are utterly different kinds of thing, how do they interact at all? How does ethanol, a physical molecule, reach into a non-spatial substance and change it? How does an unextended thought push on neurons? Descartes thought the interaction happened in the pineal gland, a small structure in the middle of the brain, but naming a location does not explain the mechanism. Princess Elisabeth of Bohemia put the objection to him in a letter of 1643, and he never gave a satisfying answer. Modern physics sharpens the problem: the physical world appears to be causally closed – every physical event has a sufficient physical cause – which leaves no gap for the mind to push through.
Dualism also sits awkwardly with the evidence of neuroscience. If the mind is a separate substance, why does a stroke in one particular region abolish the ability to recognise faces, and a stroke in another abolish the ability to form sentences? Why does the soul need a brain at all?
Property dualism and epiphenomenalism
The idea. A more modest dualism. There is only one kind of substance – physical stuff, brains included – but it has two irreducibly different kinds of property: physical properties (mass, charge, neural firing patterns) and mental properties (feeling pain, seeing red). Mental properties depend on physical ones, but they are not identical to them and cannot be explained in physical terms.
The best-known form is epiphenomenalism, which adds a striking claim: mental properties are caused by the brain but themselves cause nothing. Consciousness is a by-product. The biologist T. H. Huxley put it memorably in 1874: consciousness is like “the steam-whistle which accompanies the work of a locomotive engine,” produced by the engine but “without influence upon its machinery.”
What it explains well. Epiphenomenalism is designed to have it both ways on the wine glass. It accepts that the physical world is causally closed – everything your body does, including raising your hand and saying “ouch,” is fully caused by neurons – while still insisting that experience is real and non-physical. It also does well with Mary: her new knowledge is knowledge of a real, non-physical property that the physical facts left out. Jackson, who invented Mary, was himself an epiphenomenalist at the time.
Where it struggles. Consider touching a hot stove. You feel a searing pain and snatch your hand away. Epiphenomenalism says the pain did not cause the withdrawal – the nerve signals did, and the pain was a useless side effect. This is very hard to believe, and it has strange consequences. If pain does nothing, why did evolution give it to us? (Natural selection can only select for things that make a difference.) If your experiences cause nothing, they do not cause your reports of them either – so when you say “I am in pain,” the pain is not why you said it. Your knowledge of your own mind becomes a coincidence. Most philosophers regard this as too high a price.
Idealism
The idea. Monism, but in the direction most people do not expect: everything is mental. Matter is not a separate stuff but a pattern within experience. George Berkeley (1710) argued that all we ever encounter are ideas – colours, shapes, sounds, textures – and that the notion of a “material substance” existing outside all minds is not just unproved but meaningless. “To be is to be perceived.” What keeps the world going when nobody is looking? God, who perceives everything always.
What it explains well. Idealism dissolves the mind-body problem in one stroke: there is no gap between mind and matter because matter is a kind of mental content. Mary is no problem, since everything is experience anyway. And it takes seriously the fact that all our evidence for a physical world is, in the end, experiential.
Where it struggles. The wine glass becomes odd: why should one bundle of ideas (the wine) reliably alter another (my mood) unless there is some underlying mechanism – and if there is, does that not sound like matter? Idealism has trouble explaining the regularity and independence of the world without a God to underwrite it. And it finds few defenders today, although it has had a small revival among philosophers frustrated with physicalism.
Behaviourism
The idea. The twentieth century’s first great attempt to get rid of the “ghost in the machine.” In psychology, John Watson and B. F. Skinner argued that a science of mind should study only observable behaviour and its relation to stimuli, since inner states cannot be measured. In philosophy, logical or analytical behaviourism – associated with Gilbert Ryle’s The Concept of Mind (1949) – made the stronger claim that mental terms simply mean patterns of behaviour and behavioural dispositions. To say that someone “believes it will rain” is to say that they are disposed to take an umbrella, cancel the picnic, and so on. There is no extra inner fact.
Ryle argued that Descartes had made a “category mistake,” like a visitor who watches the battalions, batteries and squadrons march past and then asks when the division is going to appear. The mind is not another thing alongside the behaviour; it is the organised behaviour.
What it explains well. Behaviourism explains how we learn mental vocabulary – children learn “pain” from watching people wince and cry, not from inspecting anyone’s inner life. It solves the problem of other minds: you know I am in pain because you can see me behaving in pain-ways; there is no hidden fact to be sceptical about. And it handles the octopus and the robot the same way: if they behave as though in pain, they are in pain, end of story.
Where it struggles. Mary is a disaster for behaviourism: she knew everything about colour behaviour before leaving the room and still learned something. More generally, behaviourism seems to leave out exactly what matters. Hilary Putnam imagined a race of “super-Spartans” who feel pain but have trained themselves never to show it. They have no pain-behaviour, so on the behaviourist account they are not in pain – which is plainly false. The reverse case is the perfect actor, who displays all the behaviour with none of the feeling.
There is also a technical problem. You take the umbrella only if you believe it will rain and want to stay dry and believe the umbrella will help. Every attempt to define one mental state in terms of behaviour smuggles in other mental states. The definitions never close.
Identity theory
The idea. The mind is the brain – literally. Mental states are identical to brain states in the same way that water is identical to H₂O, or lightning to an electrical discharge. Pain just is the firing of certain neurons (the textbooks traditionally say “C-fibres”); seeing red just is a certain pattern of activity in the visual cortex. The view was developed by U. T. Place (1956) and J. J. C. Smart (1959), and it is often called type physicalism or reductive materialism.
The analogy with water is the key. Before chemistry, nobody knew that water was H₂O; it was a discovery, not a definition. Identity theorists say we are in the same position with the mind. Neuroscience is discovering, empirically, what mental states have been all along.
What it explains well. The wine glass becomes trivial. Of course ethanol changes your mood: your mood is a brain state, and ethanol changes brain states. Of course your decision moves your hand: the decision is a neural event, and neural events cause muscle contractions. Identity theory also makes sense of everything neuroscience has found – why brain damage destroys specific mental abilities, why drugs work, why brain scans correlate with reports of experience. It is simple, and it respects the causal closure of physics.
Where it struggles. The octopus. If pain is C-fibre firing, then a creature without C-fibres cannot be in pain. But octopuses, birds and (perhaps) future computers have very different hardware. The objection is called multiple realisability: the same mental state can be “realised” in many different physical systems, so it cannot be identical to any one of them. This was the argument that persuaded most philosophers to abandon identity theory in the late 1960s.
Mary is also a problem. If seeing red is identical to a brain process, and Mary knew everything about that brain process, what did she learn? The identity theorist has to say she learned nothing new about the world – only gained a new way of representing an old fact. Many find that unconvincing.
Functionalism
The idea. The dominant view in philosophy of mind since the 1970s. Mental states are defined not by what they are made of but by what they do – by their causal role. Pain is whatever state is typically caused by bodily damage, causes distress and the desire to escape, and leads to wincing and avoidance. Anything that plays that role is pain, whatever it is made of.
The favourite analogy is with software and hardware. The same program can run on very different machines; what makes it the same program is the pattern of operations, not the silicon. Likewise, the mind is to the brain roughly as software is to hardware. A mousetrap is another example: it can be made of wood and wire or a plastic tube or a bucket, and it is a mousetrap in virtue of what it does, not what it is made of. Hilary Putnam introduced the view in the 1960s; David Lewis, Jerry Fodor and Daniel Dennett developed it in different directions.
What it explains well. The octopus – this is what functionalism was built for. Octopus pain and human pain are the same mental state because they play the same role, even though the hardware differs. The wine glass is fine too, since functional states are realised in physical states and inherit their causal powers. Functionalism also makes psychology an autonomous science (you can study the software without knowing the hardware), and it gives a clear criterion for the robot: if the machine has states that play the right causal roles, it has a mind. It is the philosophy behind cognitive science and much of artificial intelligence.
Where it struggles. Mary, and everything Mary stands for. Functionalism defines mental states by their causes and effects, but it says nothing about how they feel. Consider the inverted spectrum: imagine someone whose experience of red is exactly like your experience of green, and vice versa. They would behave identically to you – calling tomatoes “red,” stopping at red lights – so functionally they are the same. Yet their experience differs. If that is possible, the felt quality of experience is not captured by functional role.
Ned Block pushed the point with the China brain: suppose the whole population of China were organised, with radios, to duplicate the functional organisation of a human brain. Functionalism says the resulting system would be conscious – would feel pain, see red. Most people find that absurd. John Searle’s Chinese room (1980) makes a related point: a man in a room following a rulebook could pass every test for understanding Chinese without understanding a word. The functional role is there; the understanding is not.
Non-reductive physicalism
The idea. A family of views that try to be physicalist about what exists without being reductive about how to explain it. Donald Davidson’s anomalous monism (1970) holds that every mental event is a physical event, but there are no strict laws linking mental and physical descriptions – so psychology can never be reduced to neuroscience, even though nothing non-physical exists. Others speak of emergence: mental properties arise from complex physical organisation and are genuinely novel, though they depend on (“supervene on”) the physical base. The slogan is: no difference in the mind without a difference in the brain, but the mind is not just the brain.
What it explains well. It captures the sense many scientists have that the mind is “nothing over and above” the brain while resisting the implausible claim that talk of beliefs and desires could ever be translated into talk of neurons. It respects both the wine glass and the autonomy of psychology.
Where it struggles. Critics, especially Jaegwon Kim, argue that it is unstable. If mental properties are not identical to physical ones, and the physical world is causally closed, then mental properties do no causal work – and non-reductive physicalism collapses into epiphenomenalism. Either the mind is physical, in which case reduce it, or it is not, in which case it cannot move your hand.
Eliminative materialism
The idea. The most radical form of physicalism. Everyone so far has assumed that beliefs, desires and sensations exist and asked what they are. Eliminativists – Paul and Patricia Churchland are the best-known – suggest that some or all of these categories may be false, in the way that “phlogiston” and “demonic possession” turned out to be false. Our everyday “folk psychology” is a primitive theory, and a mature neuroscience may replace it wholesale. There may turn out to be no such thing as a belief, just as there turned out to be no such thing as the humours.
What it explains well. Eliminativism takes seriously how often common sense has been wrong about the world, and how poorly our mental categories map onto the brain. Nobody has found a “belief” in a brain scan; perhaps that is because beliefs are the wrong thing to look for. The view also dissolves many classic puzzles by denying their premises: if there are no qualia in the traditional sense, Mary learned nothing.
Where it struggles. It appears to be self-refuting: the eliminativist believes that there are no beliefs and wants to persuade us. (Eliminativists reply that they are using folk terms provisionally, as we still say the sun “rises.") More seriously, it seems to deny the one thing we know best. You may be wrong about phlogiston; can you be wrong that you are in pain? Most philosophers think that whatever else is up for grabs, the existence of experience is not.
Panpsychism
The idea. Consciousness is not produced by complex brains; it is a fundamental feature of matter, present in some rudimentary form all the way down. Electrons do not think or feel in any recognisable sense, but they have some tiny spark of experience, and our rich consciousness is what you get when these sparks are organised in the right way. The view has a long history – Leibniz, Spinoza and William James all flirted with it – and has had a surprising revival in the last twenty years, championed by Philip Goff and taken seriously by David Chalmers.
What it explains well. Panpsychism avoids the “magic” that seems to be needed for physicalism: on standard materialism, at some point in evolution, dead matter arranged in a certain way suddenly lit up with experience, and nobody can say why. Panpsychism says it was never dead. It also does well with Mary: experience is a fundamental property, so of course physical descriptions leave it out, just as descriptions of mass leave out charge.
Where it struggles. The combination problem. Even if every particle in your brain has its own micro-experience, why should billions of tiny experiences add up to one unified you? Adding minds together does not obviously produce a bigger mind, any more than putting people in a room produces a group consciousness. Panpsychism also strikes many as simply too strange to be true – though its defenders point out that “too strange” was once said of the idea that the earth moves.
Mysterianism
The idea. Colin McGinn and others suggest that the mind-body problem is real, has a solution, and that solution is permanently beyond human understanding. Our minds evolved to track predators and social alliances, not to understand how consciousness arises from matter. Just as a dog cannot grasp calculus, we may be “cognitively closed” to the true theory of mind.
What it explains well. It explains why the problem has resisted solution for four centuries, and why every theory seems to leave something out.
Where it struggles. It is hard to distinguish from giving up. And it is unclear how we could know that the problem is unsolvable, rather than merely unsolved.
Where the debate stands
Most philosophers of mind today are some kind of physicalist, and most of those are some kind of functionalist. But the position is uneasy. David Chalmers gave the discomfort a name in 1995: the hard problem of consciousness. The “easy” problems – how the brain processes information, integrates perception, controls behaviour – are hard in the ordinary way, but we can see what a solution would look like. The hard problem is why any of this processing is accompanied by experience at all. Why is there something it is like to be you, when there is (presumably) nothing it is like to be a thermostat? Chalmers argued that no functional or physical story can answer this, because you can always conceive of the same functions being performed “in the dark” – by a philosophical zombie, physically identical to you but without any inner life.
Whether zombies are genuinely conceivable, and whether their conceivability proves anything, is the central fight of contemporary philosophy of mind. Physicalists say the intuition is an illusion produced by the way we think about our own minds. Their opponents say the intuition is the one thing we can be sure of. The four test cases – the wine glass, Mary, the octopus and the robot – are still doing their work.
Why it matters
This can look like the most abstract of debates, but the answers have consequences.
Artificial intelligence. If functionalism is true, a sufficiently sophisticated machine will have a mind, and we will have moral obligations towards it. If Searle is right, no amount of computation produces understanding, and the question does not arise. We are building the systems now; the philosophy is not optional.
Animals. Whether an octopus, a fish or a bee can suffer depends on what suffering is. Identity theory, functionalism and behaviourism give different answers, and those answers feed directly into law and ethics.
Mental illness and medicine. Is depression a chemical imbalance, a pattern of thought, or both? The dispute between reductive and non-reductive views of mind is, in another vocabulary, the dispute between pharmacology and psychotherapy – and the best clinical practice tends to assume something like functionalism, treating the same condition at several levels.
Death. Whether anything of you can survive the destruction of your brain depends on whether dualism is true. Almost every religious tradition assumes it is; almost every philosopher of mind assumes it is not.
Yourself. If you are your brain, then everything you think, feel and choose is a physical process among others. Many people find this liberating; many find it deflating. Either way, it changes how you understand the one thing you know from the inside.
◊ ◊ ◊
Where to start reading
- John Heil, Philosophy of Mind: A Contemporary Introduction (Routledge, 4th ed. 2019) – clear, balanced, and written for beginners without talking down to them. The best first book.
- Jaegwon Kim, Philosophy of Mind (Westview, 3rd ed. 2011) – more rigorous, from one of the field’s leading figures. Excellent on the problems for non-reductive physicalism.
- John Searle, Mind: A Brief Introduction (Oxford, 2004) – opinionated and readable; Searle rejects both dualism and standard materialism and explains why.
- David Chalmers (ed.), Philosophy of Mind: Classical and Contemporary Readings (Oxford, 2nd ed. 2021) – the standard anthology, with the original papers by Ryle, Place, Smart, Putnam, Nagel, Jackson, Block, Searle and others.
- Thomas Nagel, “What Is It Like to Be a Bat?" (Philosophical Review, 1974) – the single most influential essay in the field, fifteen pages long, and freely available online.
- Paul Churchland, Matter and Consciousness (MIT Press, 3rd ed. 2013) – a short, engaging survey by the leading eliminativist; a good counterweight to Heil and Searle.
- Philip Goff, Galileo’s Error: Foundations for a New Science of Consciousness (Pantheon, 2019) – the accessible case for panpsychism.
Read more
- Dualism – Stanford Encyclopedia of Philosophy. Howard Robinson’s survey of the varieties of dualism and the arguments for and against.
- Physicalism – Stanford Encyclopedia of Philosophy. What it means to say everything is physical, and why that is harder to state than it sounds.
- Functionalism – Stanford Encyclopedia of Philosophy. Janet Levin’s detailed account of the dominant view and its critics.
- Behaviorism – Stanford Encyclopedia of Philosophy. Covers both the psychological and philosophical versions.
- Epiphenomenalism – Stanford Encyclopedia of Philosophy. The steam-whistle view and its problems.
- Multiple Realizability – Stanford Encyclopedia of Philosophy. The octopus objection in full.
- The Chinese Room Argument – Stanford Encyclopedia of Philosophy. Searle’s thought experiment and forty years of replies.
- Qualia – Stanford Encyclopedia of Philosophy. The felt qualities of experience and why they trouble every physicalist theory.
- Zombies – Stanford Encyclopedia of Philosophy. The conceivability argument at the heart of the hard problem.
- Panpsychism – Stanford Encyclopedia of Philosophy. The revival of the view that mind goes all the way down.
- Consciousness – Stanford Encyclopedia of Philosophy. A wide-ranging overview of the whole territory.
- Dualism and Mind – Internet Encyclopedia of Philosophy. A more approachable treatment than the Stanford entry.
