Rene Descartes (1596-1650)


What is the Mind/Body Problem? It is what makes Psychology different from every other branch of science. So let's start with the other branches of science, the ones that don't have the Mind/Body Problem:

First, what is "science"? It is just the systematic investigation of what is true in the world. The reason that finding out whether it's raining today, or whether or not Bill Clinton was guilty, or who won the European Football Championship is not considered a branch of science is quite arbitrary: It is because these "truths" are just local, one-off truths, and science specialises in more general, universal truths. But otherwise there is no deep difference. What's true is true, whether it's true everywhere in the universe, for ever, like the law of universal gravitation, or it is only true of us, here, today, such as that it is Wednesday and the rain is falling in Southampton.

Finding out whether it's raining today is trivial: You just look out the window or stick out your hand to see whether you feel any raindrops.

But if you feel water-drops on your hand, does that "prove" it's raining? No; it could just be someone watering the lawn from an upstairs window. In general, feeling or seeing water-drops is evidence that it is raining, but it is not proof: The evidence could be mistaken or misleading.

Objectivity. One good thing about evidence is that it's "objective" rather than "subjective." This means that it is not just a matter of opinion. If you don't believe me that it's raining, you can stick your hand out the window and feel the raindrops too; in fact anyone could, it does not just depend on my subjective opinion.

But objective evidence, that anyone can check and confirm, is still not proof. It cannot guarantee that something is true.

What would be proof, then, something that guaranteed that something was true, rather than just providing objective evidence that it's usually or probably true?

Only mathematics can give you proof. The way it does so is simple. There is one thing you can be completely sure about, and that is that something cannot be both true and false at the same time, because that would make no sense at all. You may not be sure that it's true that it's raining, but you can be sure that if it's true that it is raining, then it's false that it's not raining. Something cannot be both true and false at the same time. That's guaranteed. (Stop and think about it: There are very few guarantees in life, but this is definitely one of them!)

And most of the proofs in mathematics are based on that: When you learned geometry, the way you proved things was to show that if they were false, that would contradict something you had already proved before. In other words, they were true "on pain of contradiction."

Here's a simple example, from logic rather than mathematics:

The "classical syllogism."

First we will begin by assuming something to be true. This is not what we will be proving. This is just something we are going to take for granted.

Now here is another thing that we will assume to be true: Now here comes the the thing we want to prove to be true: This proof is so simple that you should be able to see it instantly when it is pointed out to you:

Suppose (3) was false, in other words, suppose:

Could (1) and (2) still be true? (2) says that Socrates is a philosopher, and (1) says that all philosophers are overweight. So Socrates too, being a philosopher, must be overweight, if they all are. But (*) says that he is not overweight. Can it be true that Socrates is overweight and that Socrates is not overweight, at the same time? "No (you're supposed to reply), that does not make sense."

So, if what you have assumed to be true really is true, which is that all philosophers are overweight and that Socrates is a philosopher, then it must be true, on pain of contradiction, that Socrates is overweight.

All proofs in mathematics work like that: You assume the "axioms" are true, and then you prove that, if the axioms are true, the theorems have to be true, on pain of contradiction. Proofs usually take more steps than the Socrates syllogism, and after a few steps some people start to forget the earlier steps (like losing your way in a city), and so they can't always see that the theorem has to be true. But the proof guarantees the truth of the theorems in both cases, the simple ones and the more complicated ones.

So the truths of mathematics are absolutely certain, just as certain as it is that it makes no sense to say that it's true that it's raining and not true that it's raining at the same time, or that Socrates is and is not overweight.

Scientific evidence is not proof. The truths of science, in contrast to mathematics, don't come with that guarantee of certainty, but you can be pretty confident about them anyway. You can't prove that apples always fall to the ground when you drop them, because of Newton's Theory of gravitation, but you can be pretty sure they always will. Perhaps a little less sure than that 2 + 2 = 4, but not much less.

So science is trying to find out what is true about the world by gathering (objective) evidence, and then trying to predict and explain new evidence. The cycle is: predict, test, revise, predict, test, revise.

And there are many branches of science, from physics, which predicts and explains how objects move and what they are made of, to biology, which predicts and explains how living things work, to meteorology, which predicts and explains when it rains.

Descartes and Certainty. Now it was philosophers who pointed out the difference between science and mathematics that I mentioned: The French philosopher Rene Descartes <> was the first to focus on the question of certainty: What can we be absolutely certain is true, beyond any shadow of a doubt? He noted that the truths of mathematics are certain. Is anything else?

(1) You can be sure-beyond-a-shadow-of-doubt that 1 + 1 will always equal 2. (Yes)

(2) Can you be sure-beyond-a-shadow-of-doubt that it's raining now? (No.)

(3) Can you be sure-beyond-a-shadow-of-doubt that apples will always fall? (No.)

(4) Can you be sure-beyond-a-shadow-of-doubt that e=mc2 (No.)

(5) Can you be sure-beyond-a-shadow-of-doubt that there will always be a tomorrow? (No.)

(6) Can you be sure-beyond-a-shadow-of-doubt that life is not just a dream? (No.)

(7) Can you be sure-beyond-a-shadow-of-doubt that there's a real world out there? (No.)

(8) Can you be sure-beyond-a-shadow-of-doubt that I exist? (No.)

(9) Can you be sure-beyond-a-shadow-of-doubt that you exist? Yes! Why?

What it is that makes that last one (9) different from all the rest, and more like the first one (1) is perhaps the most important feature of the mind/body problem. It is at the basis of Descartes' famous inference: "Cogito ergo sum" ("I think, therefore I am"), and it is not a coincidence that the "Cog" in "Cogito" is also there in "Cognitive Psychology," for cognitive psychology is the science of thinking. Just as physics studies matter and motion, (cognitive) psychology studies mind and thinking.

Descartes and Doubt. What did Descartes mean? His "method" was the method of doubt: "I will check, one by one, what there is to know, and whether it is certain or open to doubt; I will doubt everything there is to doubt and see whether there is anything left over that I cannot doubt."

Descartes' starting point was that mathematics could not be doubted, or rather, the law of contradiction could not be doubted: I cannot doubt that something cannot be both true and false at the same time, because a contradiction makes no sense. Besides, once you allow contradictions, anything goes. There is no point in saying anything is true, because at the same time it's also false, and so you have not said anything at all. So contradictions, and anything that leads to a contradiction, is guaranteed to be untrue.

But besides that, is there anything else that is not open to doubt? Descartes noted that the laws of science are probable, but not provable, not certain. So are the facts of everyday life, such as whether it is raining, or it is Wednesday, or even whether all the things in the world are real, rather than a dream. Nothing guarantees that the world exists, the way the law of contradiction guarantees that 2 + 2 = 4. Nothing even guarantees (to me) that you exist: You, like everything else in the world, could just be a figment of my imagination, an endless daydream I am dreaming.

Note: Descartes was not saying that all these things are not true: that science is not true, that the world and other people does not exist. He was only saying that they are not guaranteed to be true; that you can never be certain about them, the way you can be certain about mathematics, and he wanted to know whether there is anything else that you can be certain about, besides mathematics.

And the surprising thing was that the only other thing he found turned out to be connected not with the objective, testable truths of physics or chemistry or biology, but with the most subjective thing in the universe, the mind itself: For each and every one of us knows as certainly as he knows that he cannot doubt that something cannot be both true and false at the same time that he cannot doubt that something is going on in his mind. The details could be wrong: maybe it's all no more than a dream. But it's certainly not less than a dream: Our thinking, feeling and experience may all be subjective, and the picture they give us could be false in every other respect, but the fact that those experiences are going on at all is certainly not something we can doubt! For doubting too is a subjective experience, and it makes as little sense to doubt that you are doubting (when you are doubting) as it does to doubt that something can be both true and false at the same time.

But now a problem arises: Can we found a science on Descartes' Cogito? "I think therefore I am." Fine. That guarantees that my mind, whatever that is, exists, but what next? Can we apply the objective "predict, test, revise, predict, test, revise" method of science to studying the mind? Can we study it the way Newton studied the falling apple? Unfortunately not, and the implications of that are only becoming apparent in our own century, the first to try to develop a "science of the mind."

The Other-Minds Problem. I have not yet said what the Mind/Body Problem is, or why it is a problem. Have just a bit more patience, because to understand the Mind/Body Problem it's best to start with its Siamese Twin, another problem that you are stuck with along with the Mind/Body Problem, called the "Other-Minds Problem": How can I tell whether anyone else but me has a mind? Thanks to Descartes, I can be sure I have one myself, but how can I be sure about the rest of you? In fact, never mind being sure, for a moment: How can I have any objective evidence at all -- even fallible, scientific evidence -- that anyone else but me has a mind?

You tell me: "Look, I have a headache now. I know this with Cartesian Certainty [Cartesian means it's certain in Descartes' sense] because of the Cogito. If that's good enough for me, why isn't it good enough for you?" The answer is simple: I (and anyone) can observe that it's raining, that an apple's falling, that the predictions of a scientific law either have or have not been confirmed by the evidence, but no one except you can observe your headache. That makes mental states very different from every other state on earth, or in the Universe. And that's the gist of the Mind/Body problem. For the Mind/Body Problem is the problem of explaining mental states in relation to physical states (Mind/Body is the same as Mental/Physical).

Feeling someone else's pain... Descartes' Cogito is a private, subjective, first-hand experience. It is not something that can be shown to someone else, for objective confirmation. If I tell you I have headache, the only things you can confirm objectively are (1) that I said so, (2) that I look and act as if I'm in pain, (3) that there is an axe-handle sticking out of my skull and (4) that my brain scan has the pattern that we know from many other such cases tends to occur only when people say their head hurts and when there are signs that it has indeed been injured.

Now (1) - (4) may look like a lot of scientific evidence, and it is (and it's the kind of evidence that modern scientific psychology is based on); but all that evidence is just as compatible with a world in which people (and animals) don't actually feel anything at all; they (and their brains) merely look and act as if they do. Now each of us knows that that is not the kind of world we live in. We really do feel and think; we don't just go through the accompanying body motions.

Now here comes the tricky part. Some of you will get this the first time, some will have to reflect on it longer (please send questions or comments to -- these will go to the PY104 class as well as to me, and I promise to reply):

Why and How the Mind/Body Problem is "Hard." Here is the reason the Mind/Body Problem is a "hard" problem rather than just another example of the way scientific truth is a bit less certain than mathematical truth. We will compare studying gravity (which is also something you can't observe directly) with studying the mind:

(So far they are the same.) (So far still the same: no scientific truth is 100% certain.) Why and how is thinkingmental at all? This question is very difficult, perhaps impossible, to answer. As usual, it makes itself felt more strongly in its Siamese Twin form as the Other-Minds Problem, for although the Other-Minds Problem is not a very pressing one in the natural world (except for those of us who worry about whether or not snails or amoebas or alien life-forms have minds),  it is a very real and pressing problem in the artificial world of robotics and mind-modeling, for here it comes up as a scientific question: Does this robot have a mind?or does it not? A robot can presumably be made to resemble our bodies, and what our bodies do, on the inside and the outside, ever more closely. These robots will be the tests of our theories of how our brains -- and hence our minds -- work. Yet it is obvious that this sort of modeling can only be sure of capturing our bodies and what they can do; how can we know whether it is also capturing what (if anything) they feel?

Is there any other way around this? Is there a way we can understand how the mind works without having to face the Other-Minds Problem? Surely we can be as sure as we are about the law of gravitation that real brains feel. Won't understanding them automatically bring with it an understanding of the mind?

Understanding how the brain works. First, it is unfortunately not at all clear that we can understand how the brain works without modeling it. Brain function cannot be just read off brain measurements -- not even today's brain scans. No mechanism except the most trivial one (like the wheel) wears its functional principles on its sleeve. So it is likely that to understand the brain -- and especially what the brain can do, which includes everything the mind can do (and say!) -- we will have to turn to modeling it after all. And then we again face the Other-Minds Problem (have I captured the mental function, or just the bodily motions?).

Besides, even if the brain did wear its functional principles on its sleeve, to be read off by careful experimental measurements and manipulations, would a complete theory of brain function explain why/how the brain feels, as it goes through its bodily motions?

Or would the theory just explain the bodily motions, with the rest having to be taken on faith? Newton never had to take anything further on faith, once he explained the laws of motion of inanimate bodies. That is what makes Psychology different from every other branch of science.

Links to Student Skywriting (Discussion) Archives:

Other topics to look up in these archives:

Turing Test

Harnad, S. (1991) Other bodies, Other minds: A machine incarnation of an old philosophical problem, Minds and Machines 1:43-54.  <>

 David Hume (1711-1776)