Introspection: The Science of Experience

From: Stevan Harnad (
Date: Wed Oct 25 1995 - 20:33:57 GMT

            Introspection: The Science of Experience

Now that we have reminded ourselves that there is something about the
mind that makes it different from other kinds of things there are, and
that this difference has to do with experiences, those private events
that go on in each of our heads, it seems reasonable to ask how the mind
is to be studied. Is it by studying and analysing what is going on in
our heads?

That particular approach to psychology is called "introspection," and it
seems a natural way to do it: Each science "inspects" its subject
matter. Astronomy inspects planets and stars, atomic physics inspects
submicroscopic particles, biology inspects living things, so psychology,
if what makes it different is experiences, should inspect experiences.

Well, since each of us has experiences, why can't we just inspect our
own? "Introspection" is just inward inspection.

First let's go back and look at how other sciences inspect their subject
matter: They do experiments and observe and record the results. Fine,
can we do that with experience? Well, we can, for example, imagine
being at a very great height, and we can note that this makes us feel
nervous. That's an experimental observation, isn't it?

Supposing I doubt your experimental report. You might reply: Try the
experiment yourself, and you'll see. That's certainly how it goes in
other sciences. If I did an experiment and reported the result, but no
one else could confirm it, it wouldn't be a scientific finding.

So how do I confirm your psychological observation that when you think
of heights, it makes you feel nervous? Well, perhaps I could think of
heights too, and see whether it makes ME nervous. There are a few problems
with this, though. Supposing it DIDN'T make me nervous to think of
heights? You might say, ok, it's not always true that thinking of
heights makes you nervous. But that begins to sound like hedging. Could
Newton say it's not always true that apples fall? At the very least, we
would need an explanation of the exceptions, and some way of predicting
when the relationship (apple/falling, heights/nervousness) holds and
when it doesn't.

But it's worse than that. It's not just that when you introspect and
report your experiences, I might not be able to confirm it with my
experiences, but, if you think about it, there's really no way I can
check whether your introspection was accurate at all! If someone reports
the temperature, someone else can always go and check whether the
reading was accurate. But how can I ever check whether your
introspective reading was accurate? The only one who can experience your
experiences is you. I may or may not be able to find something similar
in my own experience, but that's beside the point if there's no way I
can confirm that your OWN report was accurate.

But let's even let that one pass (although the differences between
psychology and other fields are so far looking like more than
differences in subject matter: it looks as if science's usual
methodology of measurements that can be checked by anyone is failing for
psychology); let public repeatability fail; we won't worry about it.
We'll trust the unverifiable accuracy of your introspective
observations, and we'll trust in the similarities we share as a species
so that when I confirm your report about YOUR experience with my
observation on MY experience, we're still talking about roughly the same

Now what? What can we actually find out by introspection? When we talked
about free will, and I asked you to introspect about how you chose
whether to pick the left pencil or the right one, and you replied that
it was because you felt like it (and you could have done the opposite
if you wanted to), you had to confess that if you introspected close
enough, you could not really pinpoint the cause of your choice: In the
end it was just a spontaneous choice. That's not very helpful.

So let's pick another example: You're not making an arbitrary choice
between identical pencils on your right and left, you're doing some
arithmetic. Let's do it right now: I give you two numbers (9, 6)
and I ask you to add them, and you say "15." Where did that come from?
How did you do that? Don't reply "I remembered; I had memorised it
once, and I recalled it now," because I will ask you the same question:
How did you remember? Can introspection answer that question?

In fact, can introspection answer ANY question about how you can do what
you can do? If I show you a bird and ask you what it is, and you say "a
robin," can you tell me how you managed to do that? Maybe it's because
I keep asking about the wrong kind of thing: How about riding a bicycle?
We can TEACH people how to ride a bicycle, so we must know how it's

Well, first of all, it's more like showing them how rather than teaching
them how. And even then, once you take it apart, introspection breaks
down completely when it comes to "how" questions: "Shift your weight
to the other side every time you push the pedal down." Fair enough,
but how do you understand what that means, and even if you somehow do,
how do you turn it into action? Does introspection give you any clue?

Ah, you say, it may not give me any clues in these simple cases, but
once it comes to something complex -- and I mean something REALLY
complex, psychologically complex -- then introspection, if it cannot
reveal, can at least CONFIRM what someone else -- a professional
psychologist, for example -- has revealed. Take the famous Freudian
slips: I say something other than what I intended to say -- "I'll
need her at four-thirty" instead of "I'll meet her at four-thirty."
Why did I say that? "Because you are unconsciously falling in love with
her," says your therapist. "You're right, come to think of it," I reply.
If you had asked me where the "need" came from I couldn't have told you,
but now that you point it out to me, I can confirm, introspectively,
that I am indeed beginning to feel pretty attracted to her."

So maybe this is how it goes: Things happen unconsciously, so
introspection does not reveal them directly, but the psychologist,
trained in the ways of the unconscious, can detect them, and then
he tells you, brings it to your consciousness, and then you can confirm

The trouble with this is the Hamlet/Polonius example I described (I
think) during our tutorial (unless it was in another tutorial: who keeps
these memories sorted for me?): Hamlet points to a cloud and asks
Polonius "Doesn't that cloud look to you like a camel?" And Polonius
replies: "You're right, there, I see the first hump, and the second..."
"But wait, Polonius, says, doesn't it look more like a weasel?" "You're
right, it's drawn up in the weasel crouch, and there are the eyes..."
"Or like a whale?" "Yup, there's the tailfin, the blowhole..."

And so it goes. You've heard about "false memory syndrome" (there are
experts on it in this department, notable, Dr. Amina Memon), in which a
therapist suggests to you that something happened to you a long time ago
that has been repressed by your unconscious -- early childhood sexual
abuse, perhaps -- and then introspection comes and confirms it ("Yes, I
remember, it was by a whale...").

The trouble, in short, is that people are suggestible, so you can talk
them into the fact that they unconsciously want this, or that they are
repressing the memory of that, and the conscious "evidence" that they
then draw on to confirm that you are right is rather unreliable,
compared to the measurements that are used in other areas to confirm
that something is really true.

So could it be that even though experiences are what make psychology
special, there is no way to study experiences directly?

This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:24:13 GMT