The mind–body problem: a beginner’s guide

I write software for a living, and in my line of work there is a standard way to make the mind–body problem disappear. The mind is to the brain as software is to hardware. Problem solved; ship it. I believed that for years. This guide explains what the problem actually is, what the main answers are, and why I now think the software analogy answers one half of the question beautifully and leaves the other half untouched.

The problem in one sentence

How can a thought move an arm?

You decide to raise your hand, and your hand goes up. That looks like the most ordinary causal story in the world. But consider two other things most of us believe. First, your decision seems to be a mental thing: it is about raising your hand, and there is something it is like to make it. Second, physical events seem to have complete physical causes. The muscle contracts because a nerve fired, the nerve fired because of electrochemical activity, and so on down. If every physical effect already has a sufficient physical cause, what work is left for the decision, considered as a mental thing, to do?

Every position below is a way out of that squeeze.

Descartes and the princess

The problem gets its modern form from René Descartes, who argued that the mind is a thinking, unextended substance and the body an extended, unthinking one. That is substance dualism. Its sharpest early critic was Princess Elisabeth of Bohemia, who wrote to him in 1643 with a precise question: if the mind has no extension and no location, how can it push anything? Moving a body seems to require contact, and contact seems to require a surface. Descartes's replies, including the suggestion that the union of mind and body is best understood through ordinary life and the senses rather than through the intellect, left many readers unsatisfied. Their letters are short and still one of the best places to start.

Some of Descartes's successors bit the bullet. Nicolas Malebranche's occasionalism said that neither mind nor body is a true cause; God acts on the occasion of each. Spinoza said thought and extension are two attributes of one substance, so the question of how one pushes the other does not arise.

An overlooked voice here is Anton Wilhelm Amo, born on the Gold Coast in what is now Ghana and educated in Germany, who defended a disputation at Wittenberg in 1734 arguing that the human mind is impassive: it cannot be acted upon, so sensation belongs to the body, although the mind can use the body as its instrument. It is a distinctive way of cutting the problem, and he deserves to be read in his own right.

The modern menu

Identity theory. In the 1950s U. T. Place and J. J. C. Smart proposed that mental states simply are brain states, just as lightning is an electrical discharge. No extra substance, no puzzle about pushing.

Functionalism. Hilary Putnam objected that pain could be realised in very different physical systems, human brains, octopus nervous systems, perhaps one day silicon. So mental states should be defined by their causal roles, by what they do, rather than by what they are made of. This is my software analogy, and it is the working assumption of much cognitive science and AI.

Eliminative materialism. Some philosophers argue that everyday categories such as belief and desire belong to a folk theory that neuroscience may eventually replace, as chemistry replaced talk of phlogiston.

Property dualism. There is only one kind of stuff, but it has two kinds of properties, physical and mental, and the mental ones cannot be reduced to the physical.

Epiphenomenalism. Mental properties are real but do no causal work. The cost is obvious: if your pain plays no part in your saying "ouch", how do you know you are in pain?

Panpsychism. Mentality, in some very simple form, is a fundamental and widespread feature of nature. Its main problem is explaining how many tiny subjects could add up to one subject like you.

Other places the line was drawn

Descartes's way of dividing the world is not the only one. The Sāṃkhya school of Indian philosophy, systematised in Īśvarakṛṣṇa's Sāṃkhyakārikā, was dualist too, but it drew the line somewhere surprising. On one side is puruṣa, pure witnessing consciousness. On the other is prakṛti, nature, which includes not only the body but the intellect, the sense of "I" and the thinking mind. Thinking, on this view, is a natural process; only awareness itself stands apart. The Cārvāka materialists went the other way. As their opponents report them, they held that consciousness arises from the material elements of the body when combined, much as fermenting ingredients produce an intoxicating power none of them had alone. Buddhist philosophers rejected both a soul-substance and a matter-only picture, and analysed the person as streams of physical and mental events.

These are not exotic footnotes to the "real" debate. Sāṃkhya's line, between consciousness and everything else, is close to where many people now think the real difficulty lies.

My argument: two problems, not one

Here is where I have ended up. The phrase "mind–body problem" covers two different problems.

The first is Elisabeth's: how can mental states make a causal difference? Physicalism, and functionalism in particular, handles this well. If a decision just is a certain functional state realised in the brain, then of course it moves the arm. This is the half the software analogy gets right. Nobody wonders how a sorting algorithm manages to move bits.

The second is the problem David Chalmers named the "hard problem" in 1995: why is any of this processing accompanied by experience at all? Why is there something it is like to see red or to feel pain? A complete functional description of my code tells you everything it does. It does not obviously tell you whether there is anything it is like to be it. Frank Jackson's thought experiment about Mary, a scientist who knows every physical fact about colour vision but has never seen colour, was designed to make that gap vivid (Jackson later rejected the argument himself, which is part of the story).

The interesting thing is that the two big families of answers are strong and weak in opposite places. Physicalist views explain mental causation and struggle with experience. Dualist views take experience seriously and struggle with causation. A beginner who sees that pattern understands the shape of the field better than one who has memorised ten "-isms".

Where to start

Read the Elisabeth–Descartes letters. Then read a good introduction to functionalism, and then Chalmers on the hard problem. And next time someone tells you the mind is just software, ask them which half of the problem they mean.