Tip:
Highlight text to annotate it
X
In the '70s, you did some work with Tim Wilson about judgments about ourselves. Can you tell
me a little bit about that?
Well, the judgments that we were concerned with was with people's ability to say why
they had done something or why they thought something.
I got into it because I was doing an experiment where I thought if you could persuade people
who are about to go through some intense emotional experience, which we achieved by giving them
electric shock, which is always accompanied, not just with the painful sensations, but arousal,
and the more aroused you are, the more aversive this whole thing is, but we'd
said if we could give people a pill and tell them that that pill was going to make them—their
heart rate would increase; their breathing would become irregular; their palms might
become a little sweaty—in other words, the symptoms of arousal, physiological arousal;
you get under a strong emotion—here's the pill; it's going to cause those things—that
most people would find the shock less aversive because the arousal, instead of multiplying
the sensations, is attributed to something else. It turns out that's true. It's actually dramatically true.
In those days, you had to be able to ask your subjects what was going on in their heads
to have support for your theory. I would say, "Gee, I noticed you took an awful lot of shock.
Why is that?"
The guy might say, "Well, I used to work with radios, and I got shocks from time to time."
I say, "Well, I'm sure that could have played a role. Tell me: did you make any connection
between what was going on and the pill that you took?"
"No, no, I didn't."
"Did you think about the pill or the effect it was causing?"
"No, I didn't think about it at all."
So then we tell them of our theory about why they wound up taking so much shock. They say,
"I think it's very interesting. It could well be true for somebody, but, see, I used to
work with radios and get shocks," and so on.
It was clear they had absolutely no clue about the cognitive process that was going on for them.
There's a lot of things like that where, "I think there are things going on with me.
I'll later find out that's not what I was thinking at all," or, "That's not why I did that."
Tim Wilson and I just started a series of studies to see how well people could identify
what was going on in their heads, very simple kinds of things like we would have people
look at nightgowns or nylon stockings, four of them in a row, and evaluate them. We found
out something that should be interesting to merchandisers, which is: the later you view
something, the higher your evaluation of it. I have no idea why that is. I mean, introspection tells me nothing.
Yes.
Nobody's introspection. But it's just a fact. So you ask people to do this, and then you
say, "Well, why did you like this one the best?"
They give you, "Well, the color is better, the feel." They're all identical, by the way,
except that they look slightly different.
You say, "Okay, well, that's very interesting. Thank you very much. Just one question: do
you think that the order in which you looked at those things had an effect on you?"
People get a little frightened. I mean, either, "I didn't understand the question," or "I'm dealing with a madman."
Our simple study, where we—study number one: we're going to ask you to learn word
pairs like dog-rabbit, ocean-moon. People memorized 20 pairs like that.
"Thank you. Now we're going to do another experiment. We're just going to ask you to
free-associate when I ask you a question. I'm going to ask you..." and then we'd give
some examples of the questions we'll ask. They'd say the first thing that comes into their mind.
One of the things we might say is, "Name a detergent."
The guy says, probably, "Tide."
"Fine." After it's all over, we say, "I wonder why you—why did that come into your mind?"
They say, "Well, I like the box. It's a colorful box," or, "That's what my mom uses at home."
Then you say, "Well, I don't know if you remember, you memorized the word pair ocean-moon.
Do you think that could..."
"No, I just..."
We know, because we got control subjects, experimental subjects, we know that we've
doubled or tripled the likelihood that they'll give this association.
This is the kind of thing we did. Basically, everything we tried worked in the sense that
people didn't give us a correct explanation, and sort of consistent with our subjects'
behavior, we were wrong half the time about what would be going on in their heads or what
effect we would get, so our introspections were nearly useless, which puts us together
with our subjects in terms of having problems figuring out what's going on in our heads.
What sort of research has been done since then?
Well, some of the most wonderful—Wilson has continued to do this research, and he's
done fabulous things. He asks people to evaluate art objects or records, and then say how much
they liked each of them, or he asks how much you liked it, and then, "Why did you like
it?" It could be jams, preserves. They say, "Well, it's got this fruity, kind of tangy
taste, and I like the color," and so on. People who do that, people who explain why they like
this thing do a worse job of predicting how much they're going to like it down the road,
than people who just evaluated.
The reason that he tells us is going on is that if you're asked to explain why you like
something, you're going to focus on just those things that are verbalizable by definition,
so you're going to miss all the rest of it. That doesn't get—so you say why you like
it based on the things you can articulate. All the other reasons you might like or dislike
go by the wayside, so later when you look at how much they liked the art object or the
jams or whatever, the correlation between how much they liked it initially and how much
they liked it later goes to heck when you have asked them to explain why they did that.
So it seems then that we might not be much better at predicting or interpreting our own
behavior than anybody else's? Is that true? I mean, don't we have some sort of privileged
access to our own beliefs?
Well, I know what I was thinking and you may not know what I was thinking, so I've got
an advantage there, but it's amazing how ignorant we are of things that are really important
to us and how much we insist that we know why these things happen or would happen.
My favorite say in this whole line of work was done with Harvard women. They were asked
for a period of a month, maybe more, to at the end of each day, say, "How good was your
day?" How happy were you? How satisfied with it were you?" Then they answered a number
of other questions to evaluate what went on during the day or just report what day of
the week was it, how was their sex life that day, how did the work go, how much sleep did
you get, et cetera.
Now at the end of all of this, you can see how much these things actually affected their
mood, and you ask people, "By the way, we'd just be interested in knowing how much you
think each of these things influenced your mood in general?" There was no correlation
whatsoever between the actual impact of these things on people's mood and people's reports about the mood.
If instead of saying, "How much of these things affect you," you would say, "Let's take a
hypothetical person: Jane. Tell me how you think each of these things would affect her."
Well, she gives the same answer that she would have given for herself. She's no more right
about Jane than she is about herself.
Here we have a conviction that we know things like—I mean, "You're telling me I don't
know what makes me happy or unhappy? Give me a break." Sorry, I can't give you a break.
Does happiness fall into the same category? Do we know what makes us happy?
Well, I think the study with these Harvard women say that a lot of the time we don't.
Now of course I know that if my kid gets a great report on his report card, I know that
makes me happy. It does make me happy. On the other hand, these are not trivial things.
Everybody thinks they've got Blue Monday. Actually, it's not really true. Everybody
thinks that a variety of things that have a big impact on them don't really.
Wilson has also shown that people are terrible about predicting what kinds of effects things
will have. So you've got college students who are dating, and you ask them, "What do
you think—what would it do to you if you were to break up?"
"Miserable. I wouldn't be able to hold my head up. I couldn't sleep," and so on.
Actually, that's not true. The unpleasantness only lasts for a small length of time.
You ask people who are just coming into a university, freshman, "What if you got Dorm X?
What do you think it would mean to you?"
They'd say, "Oh, my god, that's terrible. It's dark. It's gloomy. The people there are
boring. It would wreck my life."
Actually, they're no less happy.
Everybody assumes they'd be thrilled and delighted forever if they won the lottery. That only
lasts for a few weeks, and actually lottery winners end up being less happy than they
were before. I mean, neighbors are begging them for money, et cetera.
The flip is true as well. I think it happens with deaths?
One's own?
No. That would be interesting. Bereavement doesn't last as long or it is not as bad as
we think or predict it would be.
Right. Bereavement doesn't, but if you ask people, "What do you think it would do to
you to become paraplegic? You couldn't move your legs," people take it for granted that
would wreck their lives, and of course it's terrible for a while, but eventually—I mean,
they're never as happy as they were before or as the average person, but the misery does
not last. They find pleasure in things that we don't. They'd say, "I enjoyed brushing my teeth today."
Wilson and Dan Gilbert, his colleague, have a notion that we don't understand how good
our psychological immune system is, that is, ways we have of lifting ourselves up and we
don't know what the hell is going on.
Dissonance reduction is one. If somebody moves from the Midwest to California, but it's a
different job, and he has to go to a small house and so on, "How do I..." He's going
to lose something, but he reduces dissonance by saying, "Who wants to take care of a big
house like I had in the Midwest," and, "The weather here just covers a multitude of sins,"
and so on. We're very good at rationalizing, explaining away, and making things better
than we think we could.
They have a concept of immune neglect. We don't understand that about ourselves. We
don't understand that we will adapt.
This course is about the science of everyday thinking. What advice do you have for people
that want to think better and do better?
Well, we know a fair amount about that now. A little background about this: I started
working—you've spoken, I know, to Danny Kahneman—and I was doing work very much
like his, showing that people make all kinds of errors because they can't think statistically.
They make all kinds of errors where they don't understand the need for a control group and
something. I mean, 28 people took a weight loss program, and nearly all of them lost
weight, so what was the control group for that? They're not to understand that kind
of concept. There are really pretty dramatic errors that people make in everyday life.
At the time, pedagogical theories said you can't teach people rules of inference. I mean,
it's something as broad as that. You just have to teach them from facts and some procedures.
You can't give them broad principles that will affect their lives in a very substantial
way. I've accepted that view. Also, it was clear to me that I was making the same kinds
of mistakes that my subjects were. I mean, I had a little bit of advantage because I
had some data about what was going on. Basically, we're dumb, and nothing can be done about it.
I decided I was going to show that nothing can be done. You can teach people these principles,
and nothing happens. Fortunately, I couldn't have been more wrong. You can teach people
principles. I mean, tremendously important principles like the law of large numbers in
such a way that—and you can do it in a matter of minutes, and you can affect how they'll
reason about absolutely all kinds of things. My favorite example is economic principles,
personal micro-economics, concepts that people don't have and they should have, and they'll
live different lives if they understand that, if they understand these concepts.
My favorite is the concept of sunk cost. People don't understand that if they've paid for
something, they should consume it only if it's still pleasurable, because actually they
can't get that money back. Well, sometimes you can, but, I mean, if you buy a ticket
to a game and you don't go, you can't get that money back. They feel like the economical
thing to do is to, you know, darn it, go ahead and watch this thing, but you can give them—just
a couple of anecdotes will change people's behavior, make them understand this concept.
Let's say, suppose you had tickets for the basketball game in Detroit that you bought
a month ago, but it turns out the star is not playing; it's not going to be that interesting
a game; it started to snow; and it's an hour drive. Do you think you should go to the game?
They say, "Well, yes. I can't waste that money."
An economist would say, "Well, wait a minute. Let's try the following thought experiment.
Suppose you hadn't bought tickets to the game, and a friend called you up and said, 'I have
tickets to the basketball game in Detroit tonight. Would you like to have them,'" if
your answer is, "Yes, sure, I know I'll go. I like to watch basketball whenever I can,"
then by all means, go; but if your answer is, "You've got to be kidding. It's snowing.
The star isn't playing," you should not go to that game.
In other words, just a few anecdotes like that, and people will apply that. We know
they'll apply it. We call them up weeks later. They don't know they were on a psychology
study. Weeks later, we call them up in the guise of a survey being done by a national
company and ask them to think about problems, and now they will apply the sunk cost principle to those problems.
The same thing is true for opportunity costs, where the basic idea is anything you do you're
paying an opportunity cost for—that is to say you could have been doing this or that
or the other thing—so you need to assess, "Do I want to pay that cost, or do I want
to go do this other thing?" That changes people's understanding of things. It changes their thinking.
Statistical concepts, methodological concepts—I find that some undergraduate courses really
change people dramatically. They really do understand some social phenomena better if
they've had social science courses. Two years of graduate school in psychology teaches you
to use the scientific method for all kinds of everyday life problems, statistical inference
in all kinds of problems that come up in everyday life. It turns out the people who are learning
most among the psychologies, in fact the only ones who are learning, are those who are in
the so-called soft areas of psychology—personality psychology, social, developmental. The rat
people, the brain people gain very little. They're taking the same statistics courses,
but they're not learning how to apply it outside. If you're doing research on people, you have
to think about how to apply these concepts to people's behavior. You learn how to code
behavior in such a way that you can make contact with these principles.
People think statistically already pretty well about lots of kinds of things. I mean,
abilities, they're pretty good at, especially if there's something countable there. They
understand, for example, the law of large numbers for abilities. There's an expression
in this country that on any given Sunday, any team in the National Football League can
beat any other team in the National Football League. That shows a comprehension of the
law of large number, but they know perfectly well that over the long haul—class will tell.
We understand that for all kinds of abilities.
People don't understand it per personality traits. I mean, how can I compare the friendliness
of Joe with the friendliness of Jane? What's the unit here? Is it smiles per minute? Is
it pleasantries—a number of pleasantries altogether? There's no way to make the comparison.
But if you understand some aspects of coding and you can appreciate the fact that any sample
you get of someone is a small...
Danny Kahneman has a wonderful idea. You meet someone, and our conception of what's going
on is that you're sort of getting a hologram, a small hologram, a little fuzzy but basically
I'm getting a read on you at the same time. I'm getting what's there. In fact, it should
be thought of as a sample of a huge population. "I don't realize, meeting you, I have really
no conception. In fact, you can behave in totally different ways on huge range of situations.
I just don't see that," which is kind of odd because I see it for myself. I know lots of
people think I'm a jerk. Lots of people think I'm really a swell guy. A lot of people think
I'm very smart, and other people think, 'What an idiot.'" I mean, they're all right. It's
just they see one small size of behavior, and you don't recognize that.
One thing that psychologists have discovered, and Tim Wilson and I were some of the first
to work on this kind of question, is how much that goes on in our heads is unconscious?
I mean, we think we know what's going on. Freud didn't know the half of it. Most of
what goes on in our heads, we have very little inkling of. Some of the work that's come out
recently about these priming effects, trivial little things, embarrassing that we're affected
by them. I mean, you ask me to read a persuasive communication and you happen to have introduced
a fishy smell into the room, I'm not as persuaded by it, "something's fishy here."
Literally. Interesting.
Yes, literally. It works, and we know that that's what's going on. It's the fishy thing,
that metaphor, because there are some countries that don't have the metaphor: "there's something fishy going on."
That's your control group, is it? They have the experience.
That's right. There are some cultures that just don't have that. In Denmark, I think
it's, "I smell a rat," if it's...
I get you.
Who knows what a rat smells like? I don't know if there's rat essence that you can spray into the air.
So then what's the upshot? You're saying, yes, so a lot of it is unconscious.
Yes, a lot. Process, in terms of my definition, is always unconscious. There is no such thing
as awareness of cognitive process. We claim it, but we don't claim that we have awareness
of the perceptual processes that we have. We have absolutely no idea. All these various
sensations are getting treated in—I know when you teach it in psychology what perceptions
might be, virtually everything you tell people, they have no idea. There are a million visual
illusions, for example, which depend on the fact that we have certain perceptual processes
that operate in a particular way, and if you give us something that's slightly off-base
from that, we make an error in it because the unconscious procedures that we have for
perceiving the world will lead us astray in those situations. Anyway, that's...
Are we on now?
Yes, yes.
I don't know whether to stop and let you take us to another direction.
This is one that probably won't make the interview, but it is, yes, it's a selfish question. Precisely,
then—I mean, there are so many things we can do. One of them is we're trying to make
ridiculously high production values for these things and maintain attention. Another thing
we're trying to do is make a check on fluency, so it might look good and feel good but here's
some assessment to see whether you actually know this to try and boost self assessment
and awareness. Those are things from the literature we're trying to do, but are there any specific
things you think that we could do to help us teach this?
In my whole book, the thing I most want people to understand is that we solve problems—everything,
from the most common everyday problem like, "How do I make up to Joe after my unpleasantness
to him," to, "How do I solve this professional problem that I'm dealing with?"—most of
that goes on, first off, with no access to process at all. We know what's in our heads,
some of it. Huge amounts, we don't know what's in our head. The procedures that we use to
solve problems are often completely opaque to us. We don't know how we did it.
My favorite study like this that was ever done was in the 1930s. A psychologist whose
name was N. R. F. Maier had people do a problem, solve the problem. He had cords hanging from
the ceiling in different places. He said, "I want you to bring these cords together."
There were things lying all around the room. Somebody would see something that they could
use, an extension cord, so they'd tie the extension cord on one and pulled it over to
the other—easy solution.
After they had five or six of these, there was one other way to do it that they hadn't
yet discovered, which was much more difficult. After the subject had been stumped for 5 or
10 minutes, Maier, who's been wandering around the room the whole time, flips one of the
cords, sets it into motion. Within 45 seconds, the typical subject tied something to the
bottom of it, swung it like a pendulum, grabbed the other and tied them together. Maier says,
"That's great. That's the solution. How did you come up with that?" No one ever gave him
the answer—the correct answer.
He ran some psychologists through this, and they were hilarious in their rich accounts
of what... "I thought of monkeys swinging through trees. The idea of a pendulum entered
my head at the precise moment," but that's just for ordinary everyday problems of the most mundane kind.
If you read—a couple of years ago, the main journal in mathematics received a paper from
an obscure mathematician at some small university somewhere whose previous job—he couldn't
get a job—was working at a Subway. They got a paper from him which is a partial solution
to the question: can you prove that there are an infinite number of twin primes? Three
and five are twins, seven and...seven and five. I ran out. This is not my strong suit.
Anyway, mathematicians for centuries have worked on this problem. So here he was. He
gives the solution to them. The math journal's got papers on mathematics from quacks all
the time, but this seems a little possible, so they sent it out to a bunch of prominent
reviewers. The reviewers, all of them, "My god, this is right. This is the correct solution."
They published the thing by warp speed by academic standards.
When they asked him, "Tell me: how did you solve this problem," he said, "It's something
I've been working on it for years. I was sitting in somebody's backyard in Colorado
at a barbecue, and suddenly the idea about that just popped into my head. I had been
thinking about what I was about to eat."
There is a million stories like this by mathematicians: "I wasn't even thinking about it. The thing popped into my head."
You're working on—you have a slave who's working for you all the time. That's your
unconscious, and we don't take nearly as much use of it as we could.
There's a writer for "The New Yorker" who has a wonderful account of how you write,
how to do it. He says you have to sit down and think a bit about what you're going to
do. If you don't, nothing is going to happen. The next time you sit down, there's nothing,
but if you actually do that, spend a few minutes thinking about what the problem is, how you're
going to get this thing across, it's been handed over the unconscious, and the unconscious
is working on it 24 hours a day, no matter what you're doing.
I'm going to find that with my own work all the time. When I'm teaching a seminar, I give
a thought to questions. If I wait until just before I have to do those thought questions,
it's an effort; they're not very good. If three or four days in advance, I sit and wonder
what are the best things that I want to make sure come out of the discussion here, and
just spend five or ten minutes on it, three days later when I start to do it, it's like
I'm taking it by dictation and they're much better than I would otherwise have come up with.
I don't know that I've ever convinced any student that, for that term paper, first day
of class, tomorrow, start working on that term paper. I don't think they believe me.
I don't know if I've ever gotten it across, but I have a lot of examples now of this kind
of thing that I think if you spend 20 or 30 minutes with people, they might really come
to believe you and might be able to make much more use of their brain than they are.
My name is Richard. I think about inference.