Subtitles section Play video
Professor Paul Bloom: So what we're doing today is
continuing on the theme of emotions.
"Emotions" is a two-part lecture and we're continuing
along certain themes. I want to begin by responding
to a question which was raised in the last class concerning
smiling and nonhuman primates. It was a very good question.
The issue was: we know that humans have
different sorts of smiles to convey different sorts of
information. The question was,
"do nonhuman primates, like chimpanzees or gorillas or
gibbons, have the same many sorts of smiles?"
So, I contacted the world's expert on smiling,
who did not return my e-mails. So, I contacted the second
world's expert on smiling who told me that the answer is "no,"
that primate--nonhuman primate smiles actually correspond
almost entirely to appeasement smiles.
They're "don't hurt me" smiles. They're equivalent to the "coy
smile" that we saw on humans. But that nonhuman primates do
not use smiles for greetings; there's no equivalent to the
"greeting smile" or "Pan Am smile";
nor do they use them as genuine expressions of happiness.
There's no equivalent to the "Duchenne smile."
That's as far as I know. If the world's expert gets back
to me and says something different, I'll keep you posted.
Another thing. Going back to the beginning
theme of the class, what we started--just to
review, we talked about the different functions of emotions.
And then we talked about smiling and facial expressions.
And then we turned to some--to a nonsocial emotion,
the case of fear. And then we shifted to social
emotions. And we talked about social
emotions towards kin and the special evolutionary reasons
that would lead them to evolve. And as we were ending,
we were talking about the relationship between an animal
and its children, particularly in cases like
humans and birds and mammals where there tends to be a close
relationship with our children. We invest in quality,
not quantity. I might produce very few
children in my life. And my evolutionary trick then
is to focus very intently on them and make sure they survive.
If I were to produce 100 children, I could stand to lose
a few, but if I just produce five in my lifetime or two or
one, they become very precious to me.
And so, the story of the evolution of a species like us
involves a long period of dependence and deep,
deep bonds between the parent and the child.
And that's part of what I talked about,
how parents respond to children.
And I want to begin this class by giving an illustration from a
documentary about parental response to children,
but I want to give it in a species that's not us.
And here is why. I'll explain why with an
analogy. I have a friend of mine who
studies the psychology of religion.
He studies why people hold religious beliefs.
And he tells me that when he's talking to a non specialist,
somebody not in the field, he doesn't ever tell them,
"Yeah, I'm really interested in why people believe in the Bible
or why people light the candles on Sabbath or why people go to
church" because these are religions that people around
here hold, and if you tell people you
study them they'll sort of be puzzled, "why would you want to
study something like that" or offended.
If you want to talk about the psychology of religion to an
audience like this, what you do is you start with
the exotic. So, you start by talking about
people who put butter on their heads.
Dan Sperber talks about a culture where the men put butter
on their heads in the summer. And it kind of melts and that's
part of--one of the things that they do or--you talk about a
culture that believes in spirits or that trees can talk.
You say you're studying it and they say, "Oh,
that's interesting. I wonder why they believe that?"
And you use that as a way to look at more general facts that
exist even in our culture. You use the fact that we don't
take the exotic for granted as a way to motivate the scientific
study of things we do take for granted.
And this is, of course, true more generally.
This was the point in the William James quote when he
talked about things that are natural to us and noticed that
some very odd things are equally natural to other species.
And it's true, I think, in particular when we
talk about things like the love we have for our children.
So, one way to look at the love we have for our children
scientifically, isn't to look at it head-on,
because the love we feel towards our own children feels
sacred, it feels special, but look at it in other species.
And so, one of the nicest illustrations of this is the
Emperor penguin, which was--which--whose
childcare and mating practices were dramatized in a wonderful
movie called "March of the Penguins."
And this is interesting because they had this incredibly
elaborate and quite precarious system of generating and taking
care of offspring. So, I want to show you a brief
clip of the movie to illustrate some parts of this.
What they do at the beginning, which is not--which leads up to
this, is they take a very long trek from the water to their
breeding grounds. Their breeding grounds is--are
protected from the wind and they're on a firm piece of ice
so they could hold the whole pack.
They do the breeding there and it's there that the eggs are
created. So, this is where the movie
begins at this point. "March of the Penguins" was the
second best--second most popular documentary of all time,
beaten only by "Fahrenheit 9/11."
And people responded to it in different ways,
which are informative when we think about the generalizations
you could make from animal behavior to human behavior.
Some conservative commentators saw this as a celebration of
family values, such as love and trust and
monogamy. Some liberals,
who hate everything that's good and true, [laughter]
responded by saying, "Well, yeah,
they're monogamous for one breeding season.
It's a year. Then they go and find another
mate. If you add it up,
it's pretty slutty." [laughter]
I think more to the point, people were impressed and
stunned by the rich and articulate and systematic
behavior that these animals were showing.
Plainly, they didn't pick it up from television,
movies, culture, learning, schooling,
and so on. To some extent,
this sort of complicated behavior came natural to them.
And it's understandable that some proponents of intelligent
design, or creationism, pointed to this as an example
of how God creates things that are deeply, richly intricate so
as to perpetrate the survival of different animals.
From a Darwinian standpoint, the Darwinian would agree with
the creationist that this couldn't have happened by
accident, this is just far too
complicated, but would appeal to the--to this as an exquisite
example of a biological adaptation,
in particular a biological adaptation regarding parental
care to children shaped by the fact that children share the
parents' genes and so parents will evolve in ways that
perpetrate the survival of their children.
Then there's the other direction, which is how children
respond to parents, how the young ones are wired up
to resonate and respond in different ways to the adults
around them. And we quickly talked about
some different theories of this. And I'll just review what we
talked about last class. Babies will develop an
attachment to whoever is closest.
They'll usually prefer their mothers because their mothers
are typically those who are closest to them.
They'll prefer her voice, her face, her smell.
It used to be thought that there is some sort of magical
moment of imprinting that when the baby is born,
the baby must see his or her mother and "boom," a connection
is made. If the baby doesn't,
terrible things will happen with attachment later on.
This is silly. There is no reason to believe
there's some special moment or special five minutes or special
hour. It's just in the fullness of
time babies will develop an attachment to the animal that's
closest to it. They will recognize it as,
at an implicit level, at an unconscious level,
as their kin. Well, how does this work?
How does the baby's brain develop--come to develop an
emotional attachment to that creature?
Well, you remember from Skinner that operant conditioning could
provide a good answer to this. And this is known as the
"Cupboard Theory," which is babies love their moms because
their moms provide food. It's the law of effect.
It's operant conditioning. They will approach their
mothers to get the food from them.
And they will develop an attachment because their mother
provides food. And this is contrasted with a
more nativist, hard-wired theory developed by
Bowlby which claims that there's two things going on.
There is a draw to mom for comfort and social interaction
and fear of strangers. Now, in the real world,
it's difficult to pull apart these two means of attraction
because the very same woman who's giving you comfort and
social interaction is also the one giving you milk.
But in the laboratory you can pull them apart.
And that's what Henry Harlow did in the movies you saw last
week. So, Harlow exposed primates to
two different mothers. One is a wire mother.
That's a Skinnerian mother. That's a mother who gave food.
The other is a cloth mother set-up so that she'd be
comfortable and give warmth and cuddling.
And the question is, "Which one do babies go for?"
And as you can remember from the movies, the results are
fairly decisive. Babies go to the wire mother to
eat--as one of the characters said, "You've got to eat to
live." But they viewed the--they loved
the cloth mother. They developed an attachment to
the warm, cuddly mother. That's the one they used as a
base when they were threatened. That's the one they used as a
base from which to explore. Okay.
And that actually--Oh, that's just--I have a picture.
And that actually takes me to the--Oh, except for one thing,
it almost takes me to the end of the question of our emotions
towards kin. One question you could ask is,
"What if there's no contact at all?"
Now, you could imagine the effects of how--A lot of people
are interested in the question of the effects of the child's
early relationship to adults around him or her in how the
child turns out later. This becomes hugely relevant
for social debates like daycare. So for instance,
a lot of psychologists are interested in the question,
"Is it better for a child to be raised by a parent,
usually a mother, or does it make a difference if
the child goes to daycare? What if the child goes to
daycare at six months? What if the child goes to
daycare at two years? How does this affect the child?"
The short answer is, nobody really knows.
There's a lot of debate over whether or not there are subtle
differences and it's deeply controversial.
But we do know that it doesn't make a big difference.
We do know that if you got raised by mom,
or perhaps mom and dad, or maybe just dad all through
your life until going off for school and I--my parents threw
me in a daycare at age three months--it's not going to make a
big difference for us, maybe a subtle difference
though it's not clear which way it would go.
But it won't make a big difference.
But what if there's no contact at all?
What if--What about terrible circumstances where people get
no cloth mother, they get nobody for attachment?
This is a really--In the real world, of course,
you can't do experiments on this.
And in the real world with humans, this only happens in
tragic cases. But this has been studied.
So Harlow, again, raised monkeys in solitary
confinement so they were raised in steel cages with only a wire
mother. In other words,
they got all the nutrition they needed but they got no
mothering. It turned out that you kind of
get monkey psychotics. They're withdrawn.
They don't play. They bite themselves.
They're incompetent sexually. They're incompetent socially.
They're incompetent maternally. In one case,
one of these monkeys raised in solitary confinement was
artificially inseminated. When she had a child she banged
its head on the floor and then bit it to death.
So, you need to be--you need--This shows--This is kind
of a stark demonstration that some early connection,
some early attachment is critical for the developing of a
primate. Obviously, you don't do these
experiments with people but there are natural experiments,
humans raised in harsh orphanages with little social
contact, and these children--If the--In other words,
they get fed, barely, but nobody picks them
up and cuddles them. These children,
if this happens for long enough, they end up with severe
problems with social and emotional development.
From an emotional point of view, they're often insatiable.
They really need cuddling and support or they're apathetic,
they don't care at all. Now, there's some sort of good
news, which is if you get these people or these monkeys early
enough you can reverse the effects of this bad development.
So, there's some research done with monkey therapists.
So then, what they do is they take the monkey,
they raise it in a steel cage, the monkey comes out,
the monkey is kind of psycho, and then they send in a younger
monkey who is just goofing around,
jumping all around the place and everything.
And experience with this younger monkey who just follows
them around and clings to them leads to gradual improvement.
It makes the solitary monkey become better.
There might be a similar effect with humans.
So one story more about--of an anecdote than an experiment was
a situation where at the age of one and a half,
children were taken away from a really harsh orphanage where
they had no contact and brought into a home for mentally
retarded women where these women gave them plenty of contact and
cuddling and apparently, from what we know,
brought them back to normal. And this is all I want to talk
about, about the emotions we feel towards our kin,
towards our children, and towards our parents.
Any questions or thoughts?
Yes. Student: Do children in
orphanages comfort each other? Professor Paul Bloom:
It's a good question. Do children in orphanages
comfort each other? I don't know.
The situation probably wouldn't be there--The problem is
children in orphanages who are in these terrible situations
tend to be babies and very young and they wouldn't be thrown
together in situations where they could comfort each other.
It's a really interesting question.
What if it was a situation where children were raised
without a supportive cloth mother at all,
would not be able to pick them up and hold them,
but they could play amongst themselves and support each
other? I don't know the answer to that.
Teaching Assistant: Yes. Professor Paul Bloom:
Yes? Is there evidence on that?
Teaching Assistant: Yes, there is.
[inaudible] Professor Paul Bloom:
Yes. [laughter]
The answer is there is evidence, [laughs]
as everybody knows, [laughter]
that this sort of--amongst the young,
support can actually help the monkey and the children.
Somebody else had a question here?
Yes. Student: What does that
tell us about the middle ground, if the parent is comforting
just a little bit and then not that much [inaudible]
Professor Paul Bloom: Right.
So this is--The question is, "What does that tell us about
the middle ground?" So this is an extreme case but
what do we know about the middle case?
Say your parent--You're not raised in a cage,
you're not in a Romanian orphanage, but your parents just
don't pick you up very much. They don't love you very much.
There's no good evidence that that has any effect on a person.
The problem is, and we're going to talk about
this in much more detail in a couple of weeks,
is it's true that parents who aren't affectionate have kids
that aren't affectionate but it's not clear this is because
of a genetic connection or an environmental connection.
The one thing we do know is that in the middle ground,
effects tend not to be dramatic.
So when you get away from extreme cases,
effects are hard to see and require careful experimental
research to tease out. I think what it's safe to say
for a lot--for everything but the severe conditions is we
don't know what kind of effects there are.
But if there are effects they are not big and dramatic ones.
Okay. Animals' good feelings,
animals' emotional attraction to their kin,
is not a huge puzzle from an evolutionary point of view.
Evolution is driven by forces that operate on the fact of how
many genes get reproduced and replicated among your
descendants. So, it makes sense that animals
would be wired-up to care for their kids.
It would make sense that kids who are wired-up to survive
would develop attachments to their parents.
What's more of a puzzle though is that animals,
including humans, seem to have exquisitely
complicated relationships with non-kin.
In particular, animals are nice to non-kin.
You are nice to people that you're not related to.
There are a lot of examples of this.
Animals groom one another. You go, you pick off the lice
and the bugs off your friend; they pick it off you.
They give warning cries. So, warning cries--All sorts of
animals give warning cries. You are--I don't know.
You're a little animal and a big animal comes charging and
you say, "Hey!" Oh.
You may sort of cry and everybody runs away.
And that's very risky for you but you do it anyway,
often to protect people you aren't related to.
Often animals share childcare. And from a cold-blooded,
natural selection, survival-of-the-gene point of
view, you would imagine that if you
lend me your kid for the day I would eat him for the protein
and "it's not my genes and actually it gives more for my
kids." That's not quite how it works
though. Animals share food.
In fact, that animal, hugely ugly,
the vampire bat, shares food.
What happens is the vampire bat--vampire bats live in caves
and they fly out. And what they do is often a bat
will strike it big. She'll find a horse,
for instance, bite the horse,
pump in tons of blood and then fly back.
And what it does is it doesn't keep it to itself.
Rather, it goes around the whole cave and vomits blood into
the mouth of all the other vampire bats so everybody
benefits. Isn't that nice?
[laughter] Now, what you're tempted to say
is, "Well, that's really nice. Everybody benefits," but this
raises a puzzle from the evolutionary point of view.
Remember, animals benefit more, and to this situation,
animals benefit more by working together than by working alone.
The benefits outweigh the costs. This is known as "reciprocal
altruism" meaning my behavior to you, my good behavior to you,
my altruism for you, is predicated on the idea of
reciprocation, "I'll benefit from you."
And you imagine how vampire bats, for instance,
why this makes sense. This is--If you're a vampire
bat, it's a better system when anybody strikes it big to feed
you rather than for anybody who strike it big,
use the blood and then spit out all the rest of it.
But here's the problem. Here's why it's such a puzzle.
The problem is the existence of cheaters.
And in economics and sociology these are also known as
"free-riders." And what a cheater or
free-rider does is it takes the benefits without paying the
cost. Imagine two genes.
Imagine one builds a vampire bat that accepts blood from
others and shares blood. The other gene accepts blood
from others and doesn't share blood.
In the long run, "B" will actually out-produce
gene "A" because in fact, "B" will be healthy while other
vampire bats get sick. And then so the offspring will
do better. An even sharper example is an
example of warning cries. So, gophers give warning cries
when there's a predator. It is extremely adaptive to
give a warning cry. Sorry.
It's extremely adaptive to respond to a warning cry.
You hear a warning cry, you-- "Oh, crap," you run away.
It is not very adaptive to give a warning cry.
A really good solution then is to listen to warning cries but
not to give them. Suppose we had a system--It is
very adaptive when people are going to the bar,
when people buy drinks to accept the drinks.
It is not so adaptive, from the standpoint of one's
wallet, to buy people drinks. Here is a solution.
Accept drinks but don't pay for drinks.
And if everybody fell to that solution, the idea of buying a
round would fade. So, there is the puzzle.
Since cheating--Since a cheater, in the short run,
can always out-win--does better than a non-cheater,
how could this cooperation evolve?
How could it be an evolutionarily stable strategy?
And the answer is "cheater detection."
Reciprocal altruism can only evolve if animals are wired up
to punish cheaters. Now, that requires a lot of
mental apparatus. You have to recognize cheaters,
you have to remember cheaters, and you have to be motivated to
punish cheaters. And not every animal has this
degree of complicated apparatus but actually we know that
vampire bats do. So, in one clever study--So the
theory says--the evolutionary theory says "yeah,
I see what these vampire bats are doing," but you see--and
this is a case where evolution makes a nice prediction that
couldn't evolve unless bats are keeping track.
If bats aren't keeping track, then the system could never
exist because the cheaters would just take it over.
They must be watching for cheaters.
So, the experiment which was done is you--a vampire bat
strikes it big, it flies back,
and then you keep it--as a scientist you keep it from
giving blood to anybody else. Then what happens?
Well, what happens is when the other bats strike it big they
starve the selfish bat, just as if we go to bars and
everybody buys a round except for me.
And this happens over and over and over again.
Pretty soon you're going to buy a round but you're not going to
give me one. And so, just as humans are
keeping their eyes out for people who are taking the
benefits without paying the costs, so are other animals.
And it is argued that this sensitivity to cheating,
this focus on reciprocation, plays a powerful role in the
evolution of social behavior and the evolution of social
emotions. And a classic illustration of
this is The Prisoner's Dilemma. Now, many of you,
I think, have seen The Prisoner's Dilemma in one course
or another? It shows up--It is one of the
main constructs in the social sciences.
It shows up in cognitive science, psychology,
economics, that you could--The teaching fellows are passing
around something which you're not going to use right away.
But for some of you this is the first time you're going to be
exposed to The Prisoner's Dilemma so let me spell it out.
Here's the idea. You and a friend commit a crime.
You rob a bank, for instance. For the sake of this example,
you are prisoner two. You get caught.
The police put you in a little room and they say,
"We want to know everything that happened.
In particular, we want you to rat out your
friend." Now, here are the options,
and one thing about this is nothing is hidden.
The police officer could actually print out a copy of The
Prisoner's Dilemma and put it right in front of you.
And what he could say is, "Look."
You're prisoner two. "You could cooperate--Well,
you have two options. You could either cooperate with
your friend, you could stay silent, or you could defect or
you could squeal." But the police officer says,
"Look. If you--Let me tell you
something. If you cooperate with your
friend and he squeals on you, you'll go to prison for life
and he'll walk out. However, if you squeal on him
and he cooperates, he keeps quiet,
he'll go to prison for life and you'll walk out."
So, what do you do? Now, on the nice side,
what you can do is you could say, "No.
I'm going to be quiet. I'm going to cooperate."
Now, if you could trust your friend to cooperate,
you're fine, you each get a little stint in
prison, but of course your friend might defect.
Your friend might squeal. Here is the important structure
of The Prisoner's Dilemma. No matter what you--what your
friend chooses to do, you're better off squealing.
So, suppose you're prisoner two. You believe your friend's going
to cooperate with you. He's not going to be--he's not
going to give the information out.
Well, then your best thing to do is squeal on him.
What if you believe he's going to squeal on you?
Well, your best thing to do is squeal on him but if you could
get your act together and you could coordinate this,
you would both be quiet and get a fairly minor penalty.
And you could see this--This is the standard origin of the
prisoner's dilemma, why it's called "The Prisoner's
Dilemma," but you could see this all over the place.
So, here is the logic. The best case for you is to
defect while the other person cooperates.
The worst case is to cooperate while the other person defects.
Back to the police thing. The best case for you is to
give up all the information; the other guy stays silent;
you cut a deal; you walk home that day.
The worst case is you're quiet, he cuts a deal,
you go to prison for life, but overall the best is that
each cooperate and overall the worst for both is if each
defect. And the reason that makes this
tragic is this. Regardless of what your
opponent does, it pays to defect,
but if both people defect both are worse off.
I'll give a couple of other examples.
No. That's just to show that
there's a cartoon corresponding to The Prisoner's Dilemma.
It is that common. Here's the idea.
I am--I break up with my wife. We've been married for a while.
We've decided we're not going to go through it together
anymore and we break up. We're living in separate houses
and we're starting to talk divorce.
It occurs to me--Here's me. I put that out there.
"Should I get a divorce lawyer?" I ask.
Now, I know divorce lawyers are really expensive.
And it's kind of difficult to get a divorce lawyer.
But if I get a divorce lawyer--And so neither one of us
get a divorce lawyer we'll just do okay.
We'll get a mediator. We'll split the money down the
middle. That'll be okay but I'm kind of
tempted. If I get a divorce lawyer and
she doesn't, my divorce lawyer will take everything she's got.
I get everything, she loses everything.
Maybe I should be nice. Hold it.
What if she gets a divorce lawyer and I don't?
Well, then I'll lose everything, she'll get
everything. Well, we should both get a
divorce lawyer then but we'd both do pretty badly.
Imagine we're two countries, country "A" and country "B."
Should I do nuclear disarmament? That's pretty good.
We'd do okay if both countries disarmed.
We would live our lives; we'll raise taxes;
we'll do whatever countries do. But wouldn't it be cool if I
build up my weaponry and they don't?
I'll invade, take everything they got.
That's kind of tempting. Uh oh.
Also, if I don't do anything and they do it,
they'll invade my country, take everything.
So, we both build up our weaponry and we both do pretty
badly. Once you start thinking about
things this way, there's no end to the sort of
notions that could fall under The Prisoner's Dilemma.
A good example is a drug deal. Suppose I want to buy marijuana
from you, or "reefer" as they call it on the street.
[laughter] So, I have $1,000 and from you
I'd like to buy a ton of reefer so--I'm rounding off.
[laughter] So, you say,
"Wonderful. Wonderful.
Let's meet behind the gym, two in the morning on Friday,
and we'll do the exchange. You bring $1,000,
I bring the reefer." "Oh, cool.
Okay. Good."
And I think, "that's pretty good,
a thousand bucks, I get the reefer,
you get a thousand bucks. That's okay,
that's the normal thing." But now something occurs to me.
"Nobody's going to go to the cops if things go badly.
So instead of doing--bringing the money, why don't I just
bring a gun? You come with your reefer,
I stick a gun in your face, take the reefer,
go home." Maybe I won't do that,
but now I worry because you're thinking the same thing.
So, you could show up with a gun, stick the gun in my face,
take the thousand bucks, go home.
I'll have no reefer. What will I smoke?
[laughter] So, we both think this way.
So, we both show up behind the gym, two in the morning,
with guns. [laughter]
Well, that's not as bad for either one of us if I
had--I--you had a gun and I didn't have a gun.
But still, we're both worse off than if we could cooperate and
just do the damn trade. And so that's the structure of
The Prisoner's Dilemma. You can only appreciate The
Prisoner's Dilemma by actually doing it.
So, here's--here is a numerical equivalent to The Prisoner's
Dilemma. Everybody should have a card in
front of you, a file card.
If you don't--If you didn't get a card, a piece of paper will do
just as well. Please write on one side
"cooperate" and on the other side "defect" and then please
find a partner with whom to play one game.
This is a one-shot game.
One of you is player one. The player on the right-most
side from my right could be player one.
The other one is player two. Do you each have a partner?
If you have three people, you could cluster together and
do two and then two and just think.
It is actually best if you've never met or spoken to the
person you're about to deal with.
And the game is, when I say "go," simply show
the person your choice. To be clear,
if you are player one and you cooperate and player two
cooperates as well, you each get three dollars.
If you are player one and you cooperate while player two
defects, player two gets five dollars and you get bupkis and
so on. On three, just show the card to
your opponent, to your person you're playing
with. One, two, three.
[laughter] Okay. How many people in this room
cooperated? How many cooperated?
How many defected? [laughter] Okay.
How many people are now five dollars richer?
Okay. How many of you got nothing?
[laughter] Okay. So, you're learning.
You're learning that the person next to you is really an SOB.
[laughter] Now, find the person next to
you and you get to play again. And you get to play five games
in a row. Play five games in a row and
keep score. You just show it to each other,
record the numbers, show it, show it,
show it, show it. Go now.
[laughter] Anybody here win twenty-five
dollars? Yes, twenty-five?
So you-- Student 1: He cooperated four times and I
defected-- Professor Paul Bloom: That's twenty,
twenty-one. [laughter] Okay.
That's good. That's good.
So, it really is a measure of honesty.
[laughs] Anybody win twenty or more?
Fifteen or more. Fourteen or less.
Anybody do five or less? You're a good person.
It's good. It's good.
You played it with him? Student 2: Yes.
Professor Paul Bloom: Bad person.
[laughter] It's not really about good or
bad. There was a great game once.
It's a simple game, but it was a great game,
a great, famous competition a long time ago,
about 20 years ago, set up by the great computer
scientist Robert Axelrod. And he put together a
competition where people brought in computer programs to play
this game, to play The Prisoner's Dilemma.
And there were sixty-three competitors.
And these computer programs were incredibly--Some of them
were very simple, always be nice,
always be--always cooperate, always defect.
Some were elegant, prime number solutions and
prototype responses, genetic algorithms crafted to
figure out what the other person was doing and suss them out.
But the winner was developed by Anatol Rappaport.
And Anatol Rappaport actually died about a month ago at quite
an old age, a great scientist. What was interesting about this
was he was the winner with his program but his program was also
one of the simplest. It may well have been the
simplest. It was called "Tit-for-Tat" and
it worked very simple. It took four lines of basic
code. The first time you meet a new
program, cooperate. The first time you meet
somebody, be nice. After that, do on each trial
what the other program did on the previous trial.
This beat sixty-two others. And here is why.
It had certain beautiful features.
It starts friendly. Remember the best long-term
solution is everybody's be--everybody's nice.
It starts off nice but you can't--it's not a sucker.
If you screw with it, it will defect back on the next
turn. It is, however, forgiving.
Do you want to get nice with it? Be nice.
If you're nice, it'll be nice back at you later
on. It's also transparent,
nothing complicated about it, and that's actually important.
It's not merely that it's not a sucker and forgiving.
More to the point, it is--you could tell it's not
a sucker. And you could tell it's
forgiving. And this very powerful
algorithm learned to cooperate even in the situation--and
helped--learned to make it out the best even in a situation
where there's a risk of cheating and betrayal.
Some psychologists have argued that our emotions correspond to
the different permutations on The Prisoner's Dilemma.
We like people who cooperate with us.
This motivates us to be nice to them in the future much as the
Tit-for-Tat algorithm says, "If you are nice to me now,
I'll be nice to you back." We don't like being screwed
with. We feel anger and distrust
towards those who betray us. That motivates us to betray or
avoid them in the future. And we feel bad when we betray
somebody who cooperates with us. This motivates us to behave
better in the future. You can break down the cells of
The Prisoner's Dilemma in terms of emotions that they give rise
to. I did an experiment last night
with my seven-year-old and my ten-year-old.
I explained to them The Prisoner's Dilemma.
I didn't give the divorce lawyer example but-- [laughter]
and we gave them a big thing of chocolate chips and--the good
chocolate chips. We had the good chips and we
had the matrix and we had them play.
Now, what they did isn't so interesting, but what's
interesting is they were furious at each other.
One of them, the younger boy,
was--kept being betrayed by the older boy including tricks like
he'd say, "Okay. Let's both cooperate."
"Yeah. Okay."
Then he'd cooperate-- "defect!" And [laughter]
the response was anger, though not actually guilt on
the part of the other boy [laughter]
but rage. And we see these sort of things
all the time in real life. You're familiar with The
Prisoner's Dilemma but there's another game,
which you might not be familiar with.
It's called The Ultimatum Game. How many of you have
encountered The Ultimatum Game? Okay, some of you.
Very simple. Choose a partner.
It's a very simple game. When economists study this they
actually do this with real money.
I do not have real money to let you do this too.
One of you is "A," one of you is "B."
The one on the right most from this side is "A."
The other one is "B." Here is a very simple rule.
I'd like "A" to turn to "B" and make an offer.
"A" has ten dollars. You can give "B" any amount you
choose from that ten dollars, from one dollar to ten dollars.
"B" can do only one thing. "B" can accept it;
if you accept it, you agree to take home the
money and "A" keeps what ever's left--or reject it.
If you reject it, you get nothing.
Nobody gets anything. Is everybody clear?
So "A" is going to say, "I'll give you so and so
dollars." "B" would say,
"Okay," in which case "B" walks away with so and so dollars
or--and "A" walks away with whatever rest or "B" could say,
"Reject," in which case nobody gets anything.
So, this game comes in two steps.
The first thing: I would like "A" to turn to "B"
and make your offer.
Don't--"B" doesn't do anything yet.
Make your offer.
Your offer should be one word. People are explaining their
offer. Make your offer.
Okay. Stage two.
Do not negotiate. [laughter]
You're not--I see people waving their hands and it's
complicated. It should be a number from one
to ten, a positive integer. Now, "B" --I would like "B" to
say one word and you can say it really loud on three.
Accept or reject. One, two, three.
[laughter] Wow. How many people accepted?
Anybody reject it? Good.
Okay. How many people offered ten
dollars? [laughter]
How many people offered more than five dollars?
Okay. How many people offered one
dollar? Okay.
When you offered one dollar did you accept?
Anybody else offer one dollar? When you offered one dollar did
your partner accept? Okay.
How many people offered either four or five dollars?
Okay. This is an interesting game
because the person who offered--who accepted one dollar
was being rational. One dollar is better than no
dollars. So, the psychology of human
rationality is such that, from a logical point of view,
you should reason one dollar is better than nothing.
A rational person should accept one dollar.
And because we're smart, a--you should offer one dollar
but not many of you offered one dollar.
Why? Because you knew people are not
purely rational. People, even in a one-shot
game, won't accept unfair distributions.
They'll reject them just out of spite.
And so, you need to offer more. And this has been studied from
a neuro-economic point of view, which basically provides
neuroanatomical evidence that people--if you offered them one
dollar they get really pissed. [laughter]
Nobody likes to be offered a dollar.
Now, there's a more general moral here, which is actually an
interesting surprise of some relevance to everyday life.
A rational person is easily exploited.
A rational person's responses to provocations,
to assaults will always be measured inappropriate.
If you know I'm rational and you're in a sharing situation
with me, you could say to me, "Hey.
Here's a dollar. Hey, Mr.
Rational, a dollar's better than nothing."
"Well, okay," because I'm rational.
Similarly, you could mess with me because you could harass me
in all sorts of ways, take things that I own,
as long as you reason that a rational person wouldn't start a
fuss about this. There is some advantage to
being irrational, to having a temper.
Because if you have a temper and you're known to be
irrational, people are forced, by dint of your irrationality,
to treat you better. Who am I going to take from?
The person who's extremely reasonable or the person who has
a hair-trigger temper? Well, I'm going to pick on a
reasonable person because the unreasonable person might do
unreasonable things. And this is faintly
paradoxical, but often to be irrational, or at least to have
a reputation for mild irrationality,
gives you an edge. Now, this isn't focus of
provocation but this has also been presented in the theory of
why people fall in love. Suppose you're choosing who to
devote your life to, and it's a matter of huge
trust. We're going to raise kids
together. It's very important for you
that I don't leave. And I am very rational so I say
to you, "We should mate and have children because I find you the
most attractive of everybody who was available that I've met so
far. I'm very rational and so long
as this continues to be the case we'll be together."
Well, that's reasonable and rational but wouldn't you rather
be with somebody who's head over heels in love?
Head over heels in love is irrational but it's also,
within certain parameters, endearing because the
irrationality of the person means you could trust them more
in the long run, just like the irrationality of
somebody who has a temper means you don't mess with them as
much. The studies have been done more
with regard to violence than with love.
And in fact, the irrationality--the benefits
of irrational violence have been translated in terms of the study
of homicide and other crimes. Daly and Wilson describe the
cause of murder. Most murder is not caused by
reasonable provocation. Most murder is not rational in
its response. Most murder is generated by
insult, curse, petty infraction,
but this is not crazy irrationality.
It's adaptive irrationality. Daly and Wilson point out,
"in chronically feuding and warring societies an essential
manual--manly virtue is the capacity for violence.
To turn the other cheek is not saintly, but stupid or
contemptibly weak." If I show myself a rational
person when picked on or harassed, I'll be known as
somebody you could pick on and harass.
And in fact, it turns out even in the modern
world--This is from a New York Times I just picked up
a year ago today. And the point is that the
violence is due to people disrespecting each other or
giving a dirty look. And you might think "isn't that
irrational?" But it's not irrational in
circumstances where people live together in an environment where
they have to deal with each other over and over again,
and often where there's not much support by the police as
indications they talked about here.
What's particularly interesting is this sort of importance of a
reputation for violence differs from culture to culture.
And I've been talking so far in this class--and in fact,
so far in this course--about universals,
about things that are built in, things that show up across
humans and other animals. I want to turn now and end this
lecture by talking a little bit about a cultural difference.
And it's a psychologically interesting cultural difference
with regard to the emotions. And it's built around the
difference turning around what sociologists call "cultures of
honor." A culture of honor has certain
properties. You can't rely on the law.
And it has resources that are easily taken.
And sociologists have argued that when those conditions are
met it becomes important to develop a reputation for violent
retaliation. That becomes important.
Examples of culture of honors include Scottish highlanders,
Masai warriors, Bedouin tradesmen,
and Western cowboys – all cases where there's resources
such as cattle that are vulnerable and easily taken,
but you can't count on calling 911 and having people come help
you. But the culture of honor that's
been studied the most by modern psychologists is the American
South. This was settled by herdsmen
and traditionally has less centralized legal control.
So, the sociologists say the American South has more of a
culture of honor than the American North.
But how do you know? What does that do?
We're interested in this class in claims about psychology.
So, it took Richard Nisbett and Dov Cohen to study cultures of
honors and look at differences. And they found some interesting
differences. Gun laws tend to be more
permissive in the southern--in the American South than the
American North. Corporal punishment and capital
punishment tend to be more approved of.
Attitudes towards the military are more positive.
In questionnaire studies, people are more forgiving
towards cultures of honor. Somebody insults my woman and I
punch him in the face. This is considered less bad in
the American South than the American North.
There's a higher rate of violence but only in certain
circumstances. The streets of the American
South as a rule are not more dangerous than the American
North. The difference is there's a
higher rate of crimes that are crimes of honor such as,
for instance, if somebody breaks in to my
house, me shooting him. Or if somebody insults me,
me killing him. Now, this is sort of survey
studies. So, Nisbett and Cohen did one
of the more interesting psychological studies I have
ever heard of. And they did this
at--This--Sorry. This is Nisbett and Wilson.
They did this with University of Michigan undergraduates.
They did a subject pool thing like you're doing now,
and on it your demographic information was listed.
And what they did was they took white males who are not Hispanic
and not Jewish. That was their sample.
Culture of honor is a phenomena limited to males and they wanted
to make it sort of a clean study so they wanted to focus--get a
homogenous sample. So, not Hispanic, not Jewish.
And they provoked them. And the provocation was genius.
What they did was they said--they brought people in to
the psychology building, as you'd be brought in to
Kirtland or SSS or Dunham and they said--they had somebody go
in to the desk and they said, "Yeah.
Go down the hall for the experiment."
There was a hallway and then you walked through the hallway.
And walking in the other direction at that moment a
graduate student--a male graduate student would start to
walk. And he's holding some files.
And what he does is he bumps the person, looks at him and
says, "Asshole" [laughter] and keep walking.
Now, to be fair, the graduate student survived
bumping into hundreds of males, calling them assholes and then
walking to--Fights did not break out, nobody was shot.
But then they brought the men--now went in to a room and
they were tested. And it turned out that there
were differences in the stress response.
On average, males from the American South showed higher
hormone response and stress response than males in the
American North--increases in testosterone and cortisol.
There's always differences in later behavior,
the people--suggesting that they were made angry.
They gave differences in fill-in-the-blank questions,
for instance. I don't remember the examples
but it's examples like "John went to the store and bought a
'blank'" and then the northerners would say "and buy
an apple." And the southerners would say
"an AK-47 [laughter] to kill that freaking graduate
student." [laughter]
Now, again, the American South--people in the American
South were not overall more violent than the American North,
but they were more sensitive to provocations of honor.
Now, when I gave this lecture a few years ago,
a southern student contacted me afterwards and said that she
felt that picking out the southern minority at Yale was in
some regard offensive and that people say things at Yale about
southerners--American southerners that they would
never say about any other minority group.
So, there's two points I want to make regarding this.
One is, of course, these are average differences.
Not every northerner and southerner would differ along
these lines. But another one is I think the
effect is real, but it's not entirely clear
that it reflects poorly on the cultures of honor as opposed to
the other cultures. So, Nisbett,
for instance, is himself a southerner and he
points out that he went to the North he was most astonished by
how rude people are. And this is because the
North--the American North is not particularly a culture of honor,
and so there's less proper behavior towards other people
because there's no fear of retaliation or response.
Moreover, the culture of honor virtues like honor,
loyalty, courage and self-reliance,
are on the face of it not necessarily bad things.
In any case, this is an interesting example
of how there's an evolutionary background but culture modifies
and shifts it in different ways. More generally,
I've suggested over the last couple of lectures that emotions
like fear, the love you have towards your
children, anger, gratitude are not aberrations
or noise in the system. Rather, they're exquisitely
complicated motivational systems that are crafted to deal with
the natural and social environment.
And we know this only from an analysis that starts from an
evolutionary approach. So, to bring us back to D'Arcy
Thompson, "everything is the way it is because it got that way."
And your reading response for this week is that.
And I'll wish you good luck on the exam on Wednesday.
And I'll see you there.