PEN Distinguished Lecture – Meltzoff – 2/16/17

PEN Distinguished Lecture – Meltzoff – 2/16/17


We’re ready to start. Welcome to the Ph.D. in
Educational Neuroscience Distinguished Lecture series here at Gallaudet University. Welcome to those in the room and to those livestreaming
from across the country. This lecture series aims to
honor world renowned scientists in the fields
of psychology, education, cognitive sciences,
and neuroscience. These different fields and all the interdisciplinary
fields in between contribute to
the new and growing field of educational neuroscience. We want to understand
the human mind and more so how the
human learning mechanisms in their cognitive
and neural dimensions. This year’s distinguished
lecture theme is breaking down barriers. With our distinguished lecturers in the heart of D.C., we want to build bridges across fields and
scientific communities here in the area and
across the nation. Everyone is welcome to attend and we hope that many more
can actually enjoy these presentations by
streaming these presentations. I know there’s
a lot of people today. I received numerous
e-mails for this presentation. Dr. Meltzoff is the first
of our spring lecturers. We’re delighted and honored
to have Dr. Meltzoff accept our invitation. Thank you very much in the
name of everyone for being here. Professor of psychology
at the University of Washington, he received his bachelor’s
degree from Harvard and Ph.D. from Oxford University
under the supervision of one of the pioneers
of cognitive psychology, Dr. Jerome Bruner. He currently is a co-director
of the University of Washington Institute for Learning
and Brain Sciences and the Director of Infant
and Child Studies Lab. Dr. Meltzoff has received
numerous grants from NICHD, NSF and the James MacDonald
Foundation, just to name a few. Dr. Meltzoff has been recognized though several national
and international awards for his outstanding science. In the unlikely case
you should not know, Dr. Meltzoff
is internationally recognized for his research on infant
and child social development. By conceptualizing novel
methods and technologies for studying infants and
children’s cognition, he has revolutionized
our understanding of how children perceive others
and other’s intentions. He has unveiled the core skills for later social development. His findings have had
high impact on years of social
cognitive development, learning, and
language acquisition. We will learn more
about Dr. Meltzoff’s like-me developmental
theory and how it’s foundational for human social
cognitive development. The title of the talk, “Mind, Brain and How Children Learn from Infants to Society.” Without more ado, please join
me in welcoming Dr. Meltzoff. Thank you, everybody. Very happy to be here. I’ve already
had a good day. Laura Ann Petitto greeted me
where I’m staying and walked me around campus. We had a great
breakfast together. Thank you very
much, Laura. I hope everybody
here knows how much Laura — (mic feedback) and Tom Allen have represented
Gallaudet very well as part of the NSF Science
of Learning Center group that they have funded, VL2. (mic feedback) There’s not supposed
to be anything electronic here. So that’s odd. Maybe I’ll stand
back from here. But that could just
be happenstance. VL2 and Gallaudet
were part of NSF Science of Learning Center Group
and NSF put a lot of funding into understanding
the science of learning and Gallaudet and the
idea of a visual language has added greatly
to that enterprise. I’m part of a science
of learning group in the state of Washington at the University of Washington where we look
at learning in formal and informal learning environments, that means in school
and out of school. And I’ve tended to focus
on early learning. I want to talk a little
bit about the research program that I’ve been doing and
I’ve picked out some elements that I think
might be of interest to people here in the room. I’m going to cover many
topics and so would intend to do it somewhat
quickly on each of them. The first thing I’m going
to talk about is the representation of action
in young infants, and talk about
a theory I’ve developed about social development
in human children called the “like me” framework. And this is the idea
that children, even before they have language,
are very interested in what entities out
in the world are like me and self-other mapping
in ways that I’ll develop. And then I’ll talk about developmental social-cognitive neuroscience, especially using EEG that’s done with my colleague
Peter Marshall, who is locally here at Temple. Third, I want to talk
about children’s understanding of gaze and joint attention. And when I get there I’ll talk
about some new research that we’ve done with
Deaf children in this regard. Next, talk about emotions
and how perceiving other people’s emotions affect
young children’s learning. And then go up in age to
elementary school children to talk about
stereotypes that children begin to develop about
other people. Since this lecture series
has to do with breaking down barriers
and stereotypes that human beings have about
many other social groups is one of the things
that erects barriers between people
and social groups. I thought today that
introducing some of this work on stereotypes would
be of interest to you. And I’ll try to show how
that is connected to the infancy work
through theory, through the like-me
theory that I’ll develop, and I’ll elaborate
that in the last section. It’s an ambitious agenda. I’ll go somewhat
quickly through things. And to begin with, I want
to talk about imitation and the representation
of action. As I met with people in
offices, I recognize that there are many
people here interested in the representation of action. Not surprisingly because
using visual language does have to do with the
perception of human action, although clearly
there’s a linguistic overlay in interpretation
and neural substrates that deals with those
movements and actions, still, how the brain represents actions is of interest. I have tended to study
that in hearing infants, but children prior to the time that they’re acquiring language. So it’s taught us some
fundamental things about how the human brain
processes language. And one of the ways I look
at that is perception production mapping
that is tested through infants’ imitation
of actions they see other people do. And because people are
interested in action and child development here, I thought I would show
a very short movie. It doesn’t have any sound. But I will talk
over this movie. It’s about 45 seconds. And it’s a young child in
our laboratory done for, you know, one of those
Scientific American type shows that appear on television. And so the host of that
show is Alan Alda, who is a movie star. You’ll notice him in the movie. And I just want to show what
imitation looks like before I develop some theories about action. Here is a young baby in
our lab who is watching an arbitrary gesture
of putting beads in a cup. We do not give any
linguistic description of what to do, just give
the baby the material. And then the baby does
the same action and is quite happy
of having achieved the goal. Then we take the cup,
turn it over, and squash it with the hand. The baby looks at the cup
and at the person and then does the same thing
they have seen the adult do without any
linguistic support. Then we have another
gesture of pulling apart a toy, the baby looks at
the person and the object and boom, pulls
apart the toy. And the last one
is we wanted to make sure the baby was learning novel
gestures that they don’t do. And so Alan Alda touched
this object with his head. Do you want a turn? And the baby does
the same thing. So I show that movie just
to show the power of imitative learning
in young children before there is linguistic
instruction. We are actually instructing
or teaching our children whether we know
it or not about the cultural
customs and rituals out in the social world, but also how to work objects
and devices in the world. How to operate things
that you see. Children want to type on
your computer keyboard, that’s not through
reinforcement learning that they do that. Parents often when
children type on the keyboard and ruins their manuscript, they say, honey,
can you not type, can you not touch the keyboard? So the parent says no,
but as the parent walks away, the child wants to do this, because they’re imitating
what they see the adult do. Or the child’s
favorite thing may be to dive into a mother’s purse and take out a lipstick and smear it on their face. They see that the mother does
that and they want to do it. Not through reinforcement, because the
mother might have said honey, don’t use
my lipstick, but as soon as she turns her
back, the child wants to do it. Children are
observing what we do. And the interesting
theoretical issue is they’re able to do
perception-production mapping from observation of human action
to production of human action, and that of course raises
interesting issues for social psychology,
developmental psychology, and for neuroscience. Some of you might know
this work I was interested in to begin with about
how early on imitation begins. We published a whole series
of papers with young babies showing that if your gestures
are simple enough such as facial gestures, sticking out your tongue, opening or closing your mouth,
pursing your lips, or moving your fingers, babies at birth
can imitate these actions. So I have argued that
babies are able to make this translation
from the observation of human action
to their own production and can do that
at the beginning of life. That is not something
that is learned by association, not something
learned by reinforcement, but rather that babies
come into the world being able
to observe human action and map it onto their own body. And I’ve gone
a little further and developed this into a theory that I call the like-me
developmental framework. And this idea is
that even for young babies in the first month of life,
one of their major projects is to look out there in the
world and see other things that are moving
like I feel myself to move. Now any arbitrary movement
in the visual world can attract the
baby’s attention. Babies can be attracted by
a mobile moving or in the hospital maybe
by the swinging stethoscope that’s on
the doctor’s neck. Or when they’re a little
older see the trees outside move or the cars move. And their visual attention
will be attracted to those. But people are special. And I think people are
special not only because they have a certain
morphology, like faces, which can attract
attention, but also because
they move in particular ways, including facial movements. And I believe that the
baby is able to see those movements
and recognize ah-ha, there is something
out there moving like me! There is something out there
that I can map the identity between what I see
and what I feel myself doing. And I think this has
profound implications it’s an
action-representation basis for a lot of social development, a lot of connection
between self and other. But I mention that babies don’t just imitate
simple body actions like mouth opening
and tongue protrusion, but they imitate
actions we do on objects, for example,
here’s that pull apart toy, where we did a study,
using — it’s an odd room — we wanted to show the baby
a series of actions. There were five
different objects. But the first one
I’m showing here is pulling apart this toy. The mother
wears a blindfold so she doesn’t
know what the adult is doing. The baby is in
an odd surrounding, in a polka dot room. The reason we did this is we wanted to see if babies could
learn how to use this object, commit it to memory, and
then transport that memory to another room,
another environment. And that’s all important
for using imitation as a mechanism for learning
in the real world. Because you won’t always
want to use the action in the same
context or environment where you learned it. The same as a sign later. It wouldn’t be if your mother
taught you the sign or you learned
the sign in the living room that you would not
be able to produce it in the kitchen or the bedroom. It has to be that it transports to different contexts with you. So we tested this experimentally by showing the baby
an arbitrary act. Babies wouldn’t know what
to do with this toy, if we give it to him,
in the baseline group, which, of course, we had, babies would take the dumb bell and pound it on the table
or put it in their mouths, but we would show them that
you could pull it apart. And we had a series
of delays. These were
independent groups. One group were given the toys
at three minutes. Another group saw the actions,
went home, and came in a week later. And another group saw
the actions, went home, and came back to the
lab a month later. Importantly they never
handled the objects when we showed it to them. They only saw the actions
through observation. I think this is also
relevant to learning ASL, because it’s not
always the case that parents use handmolding
or whatever. Babies can see what
the adult is doing, commit that to memory,
transport it to a new context, and use it
at another point in time. So there’s some deferred
ability to recreate the gesture. Now I’m very well aware
that ASL is symbolic and there is a representational,
referential aspect that Laura talks a lot about. It’s symbolic; it’s not
the same as imitating gestures. But I’m saying some
of the substrate to learn by
observation only and then commit that to memory
and use it in another context, those things
are relevant to explore. The long and short of it is the
blue squares are the babies if they saw the action
and then reproduced it in the same context,
the polka dot room. That’s the blue boxes. The red ones are
if they changed context, the curve is very similar. Here is the baseline group
of what happens if you give the baby the objects
immediately or after various delays. That’s what they do
spontaneously. Each of these differences
are significant, each of those delays. Babies as young as 12
months, which is this age, are able to learn
from observation and remember what you
showed them a month later without having handled
the object. We also wanted to test
whether babies or prove that babies could learn completely novel
acts through observation. So instead of
only having the pull toy, we did the novel action
of this flat panel, which I showed before, we showed the baby
how to lean forward and touch on
the forehead with that panel. The baby simply observed it and didn’t handle the
panel at all. In this particular study, they observed it
and went home for a week, went back to the lab and we slid
the panel across the table. This series of slides shows
what happened. This is a 14-month-old boy. When we give them the panel, the baby is looking at the
adult, we give them the panel, the baby
looks down at the panel. And then looks
up and smiles. So the baby committed this
to memory having only observed what the adult did
and was able to imitate after a delay. I think that’s relevant for social and cognitive
development. So I want to move to part
two, which was done with my collaborator Peter Marshall and Joni Sabie who was his post doc
at Temple University. They’re experts in EEG and
we began collaborating to what is the neural
correlate of imitation and how babies
are able to do it. We published an interesting
series of papers. These are three references. This is cut off. This is 2014 in Royal Society, Developmental Science, 2011. The basic paradigm here is
the baby wears an EEG cap and the adult shows
an action and the baby
can perform an action and we’re taking EEG
measurements during this time. And measurements are taken
in frontal region, central parietal, and occipital. But we’re expecting our
chief response to be in the central region by measuring Mu Rhythm
desynchronization which you
probably all have heard of. And the results show as
expected that when babies perform an
action there is significant Mu Rhythm desynchronization
in the expected area, the central region. This is what happens when
they are executing the action. And in the publication, we show a similar
graph of what happens when they’re simply
observing the action, and we show similar Mu Rhythm
desynchronization in the central region not when
they’re doing the action, but simply when they’re
watching somebody else do it. This is helping us explore the
neural parlance of imitation. Socially, a very
relevant thing is not just that the babies can
observe what you do and imitate, but that parents tend to
reflect back to their children what their
children are doing. In other words, parents
imitate their children, as well as children
imitate their parent. And that’s a very
important social signal. We wanted to look at
how the brain responded to being imitated. It’s a social signal
in the sense that for any of you who are parents
or aunts or uncles, if a child comes up
in front of you and moves this piece of paper,
and the child, a 12-month-old, can’t
talk to the child. Often you say, do you want
to play a sliding game? And the baby does that back. And you go, that’s right,
we can slide this. That’s a common interaction. You repeat things that
the child does. We wanted to know
why that was special to babies. So we had an adult imitate
what the baby did or did an action
that was non-imitative in an experimental setting. And we found out there
was significantly more Mu Rhythm desynchronization
when the adult matched or imitated the actions that
the baby had just done rather than
when they mismatched what the baby had just done. So the brain
is able to recognize that another social being is
imitating them and it has important neural
consequences. But I think important
social consequences. The social consequence is
the child is looking at you doing it and I believe
registering something like that person is
acting like me. It pops out. It’s a visual
pop out event. That person is
acting like me. And I think it’s highly
significant to the baby. Now I want to make
a theory leap here. Laura said feel free
to make leaps in this lecture. There are several times
I’m not going to make all the connecting items
but just go forward to say that I think this
like-me idea that babies prelinguistically, are able to recognize the similarity between self and others, is important both for imitation
and for social learning and for ideas about
interpersonal connectedness even up to empathizing
with other people. That’s a leap about that. But the idea that babies can
recognize that you are like me and I am like you, I think
provides a very primitive basis for how a sense of empathy
for others could develop. It’s not identical to the sophisticated
adult-like empathy, but it is a substrate
for developing it. I want to talk about that
by passing through some new research that we have
done with Peter Marshall about babies’
response to touch. And I want to go from their
somatosensory representation of having their own body touched to how they respond to seeing
somebody else’s body touched, which is
what relates to empathy. And let me develop
this research program that has to do with not
the perception of action, but how a baby’s brain represents their body
being touched. As you probably know,
when your skin is touched, when your body
is touched that is represented in the somatosensory
cortex, which is roughly bilaterally in the
central region of your brain, going down right here. And the brain is organized in what is called a somatotopic
fashion, as you all know, that the body is laid out systematically from the bottom of your
toes to the top of your head, in your brain,
in the somatosensory cortex, such that your hand
is roughly speaking more lateral about
here in your brain, and your foot is more central
down the inner hemisphere wall which is represented over here. So there is some
separation in the neural tissue that represents the hand
and represents the foot. And that gave us a
wonderful opportunity to be among the first
in the world to look at neural body maps
in the baby’s brain. The neural body map is well
known in the adult brain. Penfield and
others have mapped it out, that’s the picture you see
in the textbook. But very little is known
about body representation in the infant brain. So we set out to do
those studies. And I want to show you
some of the findings. If you have the
baby in an EEG cap, there are three
electrodes of importance. One can say in doing these
studies, CZ overlies what you might
think of as the foot area, the central part of the brain. And C3 and C4 are
overlying the hand regions, more laterally. So we touched the baby’s hand
and touched the baby’s foot and published papers showing that there was
activation in the neural regions that you would expect if the
baby’s brain was organized like the adult brain
in this somatotopic fashion. We published a series of
papers on that and I’ll give you
references later. The most exciting thing is
we also showed that when babies simply
saw you perform an action, the same goal-directed
act of pressing this button either with your foot
or with your hand that different parts
of their brain were activated and the part of the brain that
was activated is exactly
what you would predict when babies observed
you used your hand there was more Mu Rhythm
desynchronization, if they observe
you use your hand, there is more desynchronization
in the hand area, and when they observe you
touch it with their foot, there is more desynchronization
in the foot area. So when you touch
the baby’s skin, you can light up different
regions in their brain. And when they see you use
your body part and make contact with
objects, that’s processed by different regions
in their brain. We have moved from using
this technology, EEG, to using very much
more expensive and sophisticated
technology called magnetoencepholography
or MEG for short. And for this very
expensive device, the babysits in a helmet that
has 306 sensors in it. The baby is wearing
a hat with a pellet in it that allows us to keep track
of where the baby’s head is and therefore where the
baby’s brain is in this helmet like structure. So the baby is free to move and we can track
with great precision where the neural
activity is in the brain with much greater
spatial precision than is allowed with EEG and with temporal resolution that matches
EEG by one millisecond. We have just recently
finished the study using this MEG technology where
we touch the baby’s foot or touch the baby’s hand
with a device where compressed
air comes up a tube and expands a membrane
here and puts a little tap on the baby’s body part. And I want to show you the
brain of a 29-week-old baby when they’re
touching their hand. This is really a first in
the world film of what happens. The hand region you’ll
remember is right around here. It’s laterally. Foot region,
I’ll show you. In the inner
hemisphere wall. But hand region
is around here. I want to show you a film
here what happens when you touch the baby’s
hand region. 29-week-old baby. Boom. So you can see that
the neural tissue that represents
hand is activated when we touch a
29-week-old baby’s hand. Now that’s the
activation of the hand. I’m rotating the brain around so that you can see
the foot area. And the foot area
is around here. You remember I showed you
down the middle of the brain. Now a foot should
light up around here. We touched the baby’s foot and there’s activity
in the foot region. So this is very exciting
for us because we are able using MEG to be able to
carefully map that there is different neural
tissue for representing the baby’s foot and
representing the baby’s hand. And we can map where they
are in the baby’s brain. That’s why we’re calling
it neural body maps. And we’re trying to map
out all the different or many different regions
of the baby’s body. And there are really
interesting predictions that can be made. For instance, one might be
very interested, and I was talking about this afternoon. One might be very interested
in the role of experience in representing
these body parts in the brain and how it might
lead to a stronger, better, or magnified
representation of the body part in the infant’s brain if they
use those body parts a lot. So if you think of signing
deaf individuals, signing deaf individuals use
their hands to communicate more to others than hearing
individuals do. No one
in the world yet knows about the baby’s body map in
sign, in ASL using infants. And we might make predictions that they have a stronger
or more detailed body map than hearing infants. These are novel predictions that come out from
doing these studies. From a social emotional
point of view, the things
that we’re interested in is now that we know something about
where the baby’s foot region is and the baby’s
hand region is in the MEG, for the very first time
we’re showing the baby watching somebody’s
hand being touched, watching
their hand being touched. And the preliminary
evidence that we have is when you touch their hand, you see the hand area activate. When you touch their foot,
you see the foot area activate. So the baby is processing
in their brain, in their hand region when they
see your hand being touched. I think that is
important for imitation. Because when we do
something with our hands, babies imitate
with their hands, they don’t just
try any body part. When they see a hand, they know to use that body part. But I think it’s relevant
to this audience interested in ASL, as
well, because it surely has occurred to many
people here that very, very young babies are
paying very careful attention to each of the
digits you use with your hand postures,
your hand position in space. And why are they when
you show a sign, why do they know to move
that body part rather than
doing it with their foot? They’re doing it right here. And this is the neural
representation of this wonderful organ that they’re using
for sign language. They’re recognizing that
my hand is like your hand, my finger is like your finger, and we can look at that from
the neural map of that. I got there from
imitation, but I think it’s completely relevant
to what people in this audience are doing
and I think it’s relevant to the social consequences
of that. So Peter and I
and Joni Sabie and Peter and I have
published a series of papers. This is a new one is neuroimage. And this is a theoretical paper in trends in cognitive science. That was about the sense
of touch and baby’s body maps in the infant brain and
things we recently developed. And now I want to move
to gaze following. It’s highly important
for young children. It’s important
for language acquisition. It’s important
for social development because information
is not distributed equally in the world. There are what I call
in these publications information hot spots. There are places in the
world that are loaded with information and parents
want to attract their baby’s attention to those
information hot spots. What can they use to attract
the baby’s attention there? And one of the things they
can do is look at that hot spot. I can look over here and
everybody in this room will know what
I’m looking at. If I look over here,
people are wondering why am I looking
at Laura right now? You can read my eye gaze
and where I’m looking very well. You project out from
my eyes where it intersects with an object in space. When babies begin
to do that, it’s a very efficient learning
mechanism. Many people follow eye gaze
around the world. We do it in our lab, too. We have a paradigm where
we have two inanimate objects equidistant from
an adult and a child. The adult and child
make face to face contact. The objects are
inanimate objects. They don’t move. They’re just out
there in space. And then the adult turns
to look at an object and we find that 12 month olds
readily look at an object where the adult
had looked. And we have rates
of responding in typically-developing
children. Obviously the adult
looks to the left, they look to the right. There’s a random schedule. And we’re able to show
statistically that the babies turn to follow
the adult line of regard. I talked about
the social consequences for language
learning and so forth. The exciting new thing
that I just got the data on, essentially while on
the airplane coming here. We did this study recently, but we haven’t published it yet,
with deaf children, as well. Deaf children of Deaf parents and compared them to hearing
children of hearing parents. Where the prediction
is deaf children may be better than
hearing children. The idea is that deaf children
are raised in a culture, in an environment – Deaf
children of Deaf parents – where they’re not hearing
things in the world to attract their attention,
but parents who want to deeply direct
the kids’ attention when they look at a place
in the world, and the child follows
that line of regard, the child is getting to what the mother or father wants them to look at. So rather than using linguistic
or verbal instructions to do that or
having a loud sound out in the environment, the
parent can look at an object. And I know that deaf children
have expanded perception of things in the
periphery that are moving, but it’s important
that the adult and the child are just sitting
face to face and then the adult is turning
to bring attention to the periphery. It’s not that something in
the periphery is moving. So the new data that we have
show that these are the data that show that the
children who are deaf have between 8 and 20
months of age we had children,
for every deaf child there were five hearing children
matched to that child in terms of
within 7 days, plus or minus, 7 days of
the deaf child’s birthday. They were very carefully
age-matched controlled. We’re very happy to see
and very strong effects of the deaf kids being better
at gaze following between 8 and 20 months of age. It’s a whopping effect. We tested hundreds of typically developing children and the deaf children are just
way out in the scale of being much better. This is an interesting
case of where children can be advanced in using
visual modalities to understand sort of the
referential aspect of gaze. A person looks here and they’re referring to something out in space. And little Deaf children
are catching onto that faster than
hearing children. So we’re going to continue
to work with this. And studies in the seminar that
we have this afternoon a possible follow-up study
was suggested that we can talk
about in the discussion. So now I want to take this
and show you how we used it. We have been studying
in hearing children at what age do they pay
attention not just to the fact that your head is turning,
but that it’s your eyes specifically that are
focused on the object. So we had an adult turn to
peripheral objects either with eyes open or
with eyes closed. And if children were
paying attention to very detailed information,
they should only turn to follow the line of regard when a
person is turning with their eyes open, rather
than their eyes closed. Because when they’re
turning with their eyes closed, it’s a body movement, but not perceiving
the external world. They are simply making
a body movement, but not having
perceptual contact. We did studies
with our hearing children where we had adults turn with eyes open or eyes closed. Three different ages. And at each of these ages children turned
significantly more when the adult turned with
eyes open than eyes closed. Then we said to ourselves, closing your eyes
is not the only way of blocking perceptual contact
with the external world. You can also block
perceptual contact by having an inanimate object come between your eyes
and the object. So we began using blindfolds
and we had an adult turn with a blindfold
to an external object. We found that the 14-18
month old did not follow that line of regard,
just like they did not follow when the adult
had the eyes closed, but 12 months made the mistake. 12-months-old when an adult
turned with a blindfold would follow
where they turned, whereas they would not
follow where they turned when the adult had their
eyes closed. So what is the difference
between the adult turning with their eyes closed versus turning with
the blindfold on. That’s where I reached
back to my like-me hypothesis. And this is a very
speculative hypothesis. I’ll show you how we tested
in a second. I thought that babies have
a lot of experience with opening and closing
their own eyes and the self-experience. And the fact that when you
close your own eyes you cut off your perception
of the world. You have agency, you have
complete self-control. When you do it, it cuts off
your perception of the world, and it always works. If the baby is sitting
in the crib doing an experiment, every time
they close their eyes and try to see anything in the
world, the world is cut off. I thought that self-experience
might be teaching infants something about body movements
and their perceptual effects. When you close your eyes,
you can’t see the world. That’s what
happens with me. When the other
person does that behavior, perhaps they
can’t see the world. What that predicts, and that’s the like-me hypothesis, is the infant makes generalizations from their own actions to other people’s actions. When others act like me, they have inner-state like perceptual consequences like me. So we did a study seeing
if we could teach the baby that adults cannot see
the external world when they have a blindfold on. Right? So the way we did that
is we gave the baby self-experience
with a blindfold. They had objects
out on the table. And for seven minutes
every time the baby looked at an object
we put a blindfold up in front of their eyes. So they couldn’t see. Took it down. Had other objects on the
table block their view and so forth. So babies had an
experimental treatment. They were randomly assigned to
that treatment versus control, that they had the experience that when this blindfold comes
in front of their eyes they cannot
see the external world. And the result showed
that we had three groups. I won’t go through that. After the baby
had the self-experience the adult wore the blindfold and
looked to the external world, and just as predicted,
when babies had self-experience that they couldn’t see
the external world, now when the adult
turned with the blindfold on the babies did not
follow their line of regard. So I am arguing that babies
use their self-experience that I can’t see
when I have a blindfold to interpret when
another person has a blindfold that person
cannot see either. It’s the like-me hypothesis. So I want to go on here
to say what sort of entities have perception? Adults have perception. We follow the line
of regard of adults. What happens if you
have an inanimate object? So we have this robot. It has eye spots for eyes. When the
robot turns to the side what we found is that 18 months, what we found is that
18-month-old babies would not follow the line of
regard for the robot. They looked at
the movement. They saw that this hunk of
metal was moving, but they didn’t
project out in space to look at where it’s pointed. The object, if you
remember, our paradigm, we’re quite some
distance away. So babies, if they just
follow the movement here didn’t catch sight
of the objects. The objects were
farther away in space. So we began to wonder what
we could do to make this robot be a sentient perceiving
being to an infant. And because I think infants
can understand imitation we began to use imitation
and we had that robot and the adult interact
and imitate each other while the baby watched. Baby wasn’t involved. The baby simply watched. And that’s what
this next movie is. The baby was watching
this just like you are. Can you do this? Yeah! So the baby watched that
interaction where the adult
and the robot had a sensible imitative
interaction, which can only happen if they’re
perceiving each other. And then the baby and the
robot were face to face. The robot turned its head
to one of the objects. And the 18-month-old
followed line of regard. So we had assigned babies
to an experimental treatment where they watched imitation
or other control conditions, one where the robot did
the same thing but out of sync
with what the adult was doing, and the others were
baseline condition, and it’s a condition
where the baby watches the imitative interaction. That they’re able to get
the idea that robot has some intentionality
perception, sense of some sort
of psychological basis and will follow its line
of regard. So to skip forward, keep
on going here rapidly, I want to know talk about
emotions and learning. Children don’t just
imitate everything, they don’t learn everything. They also pay attention
to the emotional responses that adults show. And most of the studies
that I talked about before were studies
of diatic interaction where an adult was showing
something directly to the baby. And our idea was sometimes
babies learn things by watching other people
interact, not only when an adult is teaching them
something directly. In other words babies can
learning by eavesdropping, by watching other
people’s interaction. It’s not that you have
to directly didactically teach them everything,
they can watch siblings interact and learn from that. They can watch mother and
father at the dinner table and observe
and learn from that. We wanted to do some
studies on that. And this series of studies
here is what we call emotional
eavesdropping paradigm. There is an adult
that shows a baby what to do. He uses a stick
to press a button. And we know
from control conditions if you gave the baby the stick
and the button the baby will imitate it. So that’s
the first person there. The adult. then there’s an infant who
watched the adult and then there’s a third person,
this key person, an emoter. And the emoter
is on the side. And when the adult presses
a button with a stick, the emoter gets very angry
with that adult as if it’s a forbidden action
and says in a very nasty tone of voice that’s irritating. That’s really annoying. And makes a negative
facial expression and says in a harsh tone of
voice, that’s really annoying. But the adult does that, the emoter does that
to the adult. And the little 15-month-old
or 18-month-old, we tested both, the
little baby watches wide eyed as one adult gets angry
at the other adult for doing this action. And then we gave the object
to the baby. I wanted to see
what the baby would learn from observing that. So let me just show you
this interaction first. The adult, you’ll see
the negative facial expression at that time that
the adult is making a negative facial expression
she is getting angry. And her name is Nina. She walks in here halfway
through this film. So first the adult,
the emoter will sit here. The adult is just showing
the action to the little baby. And the baby is watching. [Ringing] And then Nina
walks in the room.>>I’m going to sit here
and read a magazine.>>That’s Nina. Nina is going
to read a magazine.>>MELTZOFF: The sound
doesn’t make much difference. And Nina is going to
look up and get angry. [Ringing]>>NINA:
That’s aggravating. That’s so annoying. That’s just your opinion. It’s aggravating.>>MELTZOFF: And now
we’re going to give this toy to the baby and watch
what the baby does. The baby won’t
do the action. So Nina is over there
now with a neutral face. Nina is no longer angry. But the baby is looking at Nina and will not perform the action in front of the
previously-angry person when the previously angry
person is looking at them. Again, I think
that’s like-me idea. That the baby is thinking
that emoter got angry at the adult for doing
the action, now that adult, that emoter is going
to get angry at me if I were to do that action,
so I’m inhibiting my behavior. We did that kind of
follow-up studies to make sure it was that kind of
triangulation that the baby was using. There is an easier
hypothesis. The easier hypothesis is maybe
the baby saw the adult was angry and was frozen
in their action and they wouldn’t do any action
because there was anger present. And the way we tested
that is the adult showed the same action, Nina,
the emoter got just as angry, but then after
Nina got angry, she got up and left the room. She just left. And when we gave
the object to the baby, the baby looked to see that
Nina was out of the room, picked up the stick
and did the action. And the reason
that’s relevant is because Nina had
demonstrated the same anger. So it wasn’t just that
there was anger in the room. We did other tests
where Nina got angry and after she was
angry she turned her back so she couldn’t watch the baby. And as long
as she had her back turned the baby would do the action. Nina closed her eyes
after being angry. And then the
baby would do the action because Nina couldn’t see her. So the baby really was
doing something fairly sophisticated. The baby didn’t inhibit
the action all the time. It’s not that the baby
was scared at the anger. The baby was weighing
several things. The baby was recognizing
that Nina got angry at this action and Nina,
the previously angry person is watching me. I better not do
the action now. But I can do it if Nina
is not watching me. So that is a very
interesting combination of recognizing
a person’s emotions and whether that person
is looking at me. It’s combining gaze following, gaze, direction
of gaze, and emotion. There’s also another interesting
characteristic of this and that is we did
some studies to see if the baby was developing
some notion of a trait about this person. And we think they are. We think the baby is beginning
to represent this person as an anger-prone person. And that’s very relevant
for clinical psychology. A parent might get
angry at you a lot. Or you might meet a person
who gets angry a lot. Does the baby
begin to typecast or pigeonhole that adult. So the way we tested that is we had Nina become angry
when the adult did an action on one toy, second
toy, and third toy. But then the adult picked
up a completely new toy and did an action on it and the emoter did not get
angry with them. She had a history of being
angry but was not angry now. And if that emoter who had been angry and was not angry now, never showed
any anger at that new action, if that emoter
was watching the baby, the baby didn’t dare
do the action. If that emoter turned her
back and was out of the room, the baby was
happy to do the action. So we think that the
baby is representing the adult’s emotional history. Do you see what I mean? The adult isn’t
angry now at this act. But that adult
is an angry prone person. And we all stereotype
or pigeonhole adults that we interact with. We believe they have
certain traits. Laura is a friendly person. Right? So when I see her
in a new context, I attribute some
characteristics to her that I learned about her
in the past and I can make
predictive generalizations of how she will act in
this new context because I have a representation of her past emotional
characteristics. She’s a friendly person,
a generous person. Very important to the baby, maybe, is this pigeon-holing that this
is an angry prone person. I better not learn
from this person or act in front of this person, I better be careful in
front of this person. We think this is important
for social psychology, what is called
trait attribution. That little 15-month-old
babies are beginning to form representations
about the personality of the person
they interact with. And we think that
that’s relevant. So these new papers
from 2016 just published are about infants’
generalization about other people’s emotions, foundations for trait-like attribution, attributing
traits to other people. It’s relevant to bilingualism,
trying to remember you speak Spanish
to that person, you speak English
to this person. This is an angry-prone person, this is a playful person. Little babies are beginning to make those generalizations. Okay. So this is the part that
I wanted to get to, which is why I’m moving quickly
through the rest of them. I think Laura told me that
this lecture series had to do with breaking down
barriers, right? And Laura and I had
breakfast this morning and she was talking to me
about how differences are not necessarily, is
not necessarily deviance. They’re just differences. And stereotypes are
relevant to this. And I want to bring stereotypes and children’s
awareness of people and the stereotypes that
culture attributes to others. So here is the idea. American society,
and clearly societies all over the world have
stereotypes about other people. In America, there is
a very strong stereotype related to gender that’s
very well spread, widely held. I know it’s not a truth,
it’s a stereotype. And the stereotype is
that girls don’t do math. Boys do math, physics,
engineering, STEM topics. Females do not do that. People are aware
of the stereotypes. Don’t shoot the messenger. It’s not my belief. It’s a stereotype. Because I’m interested in children learning social categories from others, and because I’m interested in them mapping to themselves, I began to wonder when do children begin to catch the stereotype that other people exhibit in the culture? So we tested to begin with
elementary school children to test their stereotypes
about girls and math. Now there are
other stereotypes. In American society,
there’s stereotypes about race, stereotypes
about age, about SES, about disabilities, about
many different things. But in this particular
experiment we wanted to study children’s
stereotypes about gender and especially about
gender and math. So we developed a test
to look at that about how boys do
math and girls do reading. And we did it through
explicit tests, but we also, by
asking kids questions about who does
math and who does reading. But one of the things we
also did is to develop a child implicit association
test, or child I.A.T. People nodding their heads
about what an I.A.T. is. It’s a way
of looking at children’s implicit representations
of the world. And it has to do with
speed of responding when you see
a stereotype association such as boys and math. You’re faster when you see boys
and math associated together to push
a button than you are when you see girls and math. The stereotype association
is boys and math and girls and reading. And if I show you those,
you’re very fast to respond. So we developed a child
version of an IAT test. And we’re able to test
kids from first to fifth grade. And the stunning result is
that as early as second grade, second grade,
before the kids learn their multiplication tables, Seattle kids were
highly stereotyped that boys went with math,
boys went with numbers, boys went with graphs. Girls went with
reading, book, letter. Second grade, before
the multiplication tables, they were highly
stereotyped. So they had caught
the stereotypes that exist in the culture and
began to assimilate them and I believe applied
them to the like-me category. I’m a girl. I look out in the world
and I’m very attuned to and pay special
perceptional attention to if I’m a girl how does
society treat like-me others. How does society treat other girls in this society? If you look around
in American society, there’s all kinds of things in
magazines and media, teachers, parents,
that girls don’t do math. So we believe girls have
gender identity, I’m a girl. They look out there and
say girls don’t do math, and it changes their
self-concept, I don’t do math. And this is showing here
there is a distinction between these three constructs that if you keep them straight you can look
at the developmental pathway for each of three
constructs and it develops in the order we believe
they just articulated. The relation between self and
boy is gender identity. The relation between
a social category like boy and an attribute
like math is a stereotype. Boys don’t have to go with
math, but it’s stereotyped that social category goes
with this attribute. And the link between self
and math is a self-concept. And we think that those
things develop in order, which we’ve articulated in
these various publications that early on we and
many other people think gender identity develops
early in late infancy or early preschool period. Girls recognize I’m a girl and boys recognize I’m a boy. And by second grade
they’ve caught the stereotype that girls don’t do
math and then a little later in development
perhaps by third grade they internalize that
stereotype that I don’t do math. See what I mean? They catch it
from the culture. So we’re interested
in children’s kind of reactions to stereotypes that adults have. We think this is
incredibly important for their self-concept and
what their aspirations for the future are because
our children are watching how we treat others
in society. We’ve just published a paper
that’s online right now about kids
catching social bias. And it’s an interesting
paper in psych science. And what we had there were
two adults on video tapes side by side. And one adult
in the middle. And the adult in the
middle would turn to one of
what we call the targets, two adults on either side,
and turn to one target and act in a very
prejudicial, biased way, and turn to the other one
and act in a very positive way. And then we presented
these two adults to the children and said simply
who do you like more? Who do you want
to learn from? We had them do
an imitation task where they had to imitate
what this adult was doing and that adult was doing,
and the general idea was they said they didn’t like
as much the person that they observed the adult
act negatively to. They were less likely to
imitate that adult and so forth. Less likely to learn a language,
a word, a label from that adult. Does that make sense? So we’re beginning to look
here about the mechanisms of intergenerational
transfer of bias. How do children, these
tender little children learn prejudice and bias? And one way they do is again not because we explicitly
teach them, not because we reward them
for acting in a certain way, but because they’re observing
our behavior, and they want to be like us. They adopt our stereotypes. They adopt our biases. So that gets
me to the conclusions. And I’m running
out of time here. So I’ll just say
a few slides. How can we advance theory? Well one is to
explore the body map. What is the baby’s brain? How does the baby
represent hands, feet, lips, trunk, other
parts of the body? What about the representation of
the body map in newborn babies before
they manipulated the world? What about special populations
such as deaf children who early in
development probably, we don’t have data on this,
but probably pay more attention to their hands and fingers and the positions intentionally
than hearing infants do. Right? And so the neural
representations of these gorgeous organs could
be more adult like in a deaf infant early on than
it is in a hearing infant. Laura and I
were talking about that sort
of study earlier today. It seems very relevant. We want to analyze infant
representation of action. I think that imitation is
not one indesociable unit, but can be broken apart into
the body part you use, hands or feet,
the actions you do, and the causal consequences it
has in the world so we can begin to
analyze imitation as neural representation
more carefully. And I think that it’s
relevant that imitation connects down to action,
science, and what I would say is up to social cognition and
theory of mind. This like-me recognition
that first happens on an action basis
is what I think is development
for later social cognition. So the developmental
theory ideas, there is a starting state of
action representation where there
is a basic connection between the perception and production of simple acts. And I think that
connection, that mapping between perception and
production is demonstrated in neonatal imitation
and other phenomena. I think another phase,
the next phase is infant self-experience,
infants begin to recognize that when they are acting
in a certain way they feel a certain way. And by having that connection
between their behavior and their
underlying interspace, whether it’s their
behavior and their perception, or their own behavior
and their emotion, I think babies
are able to make that connection between behavior and underlying
mental states in themselves and they use that experience
for understanding others. Those people who act like me
have mental states like me, and I try to develop
that like-me idea in these series of papers. I want to emphasize
I don’t think self-experience is the only way that
infants learn the relation between behavior
and underlying mental states. I think they can get
there by looking at the structured orderly
behavior of other people, too. And I published some
papers about that. But I wanted to develop this developmental
progression right now. And the perhaps last slide
here is this idea of developmental
phases at a broader level that gets
to breaking down barriers. I think infants imitate
a person’s actions, their motor behavior. But they’re analyzing that
person in terms of motor act. At a slightly older age, toddlers and
elementary school age, they’re paying to other people
out there who are like me. And they pay attention
to how society treats those like-me categories. And that’s what picking
up stereotypes are like. And they begin to not just
identify with another person through action, but to pay
attention that how culture treats others like me has
implications for myself, how I can think of myself,
what I can aspire to, dream about, and
affect their self-concept. And I think that’s
very important. They have now moved beyond
recognizing the similarity at the level of pure
action to social categories as a whole and
what group they belong to and how others treat
that social group. So I think the last
slide then here is a paper we published that
was emphasizing the importance of brain perception,
action, coding and how that filters up
to the kinds of things I was talking about here
to more important and larger categories about
social groups, stereotypes, and the like. And we call that foundations for new science of learning because although there is a very rich initial state that I think infants start with, that initial state
gives them a framework for interpreting the actions
of others and the way they’re treated and
helps, has cascading effects in their developing
social cognition. So I think with that, that’s the end
and I’ll take questions. [Applause] Thank you.>>Okay. Thank you very much. It was a wonderful talk. I’m sure someone in the audience
has questions. Please come forward
and make sure to come here
where we can all see you. I will leave the mic
on the stand here. Thank you.>>Hi. Nice to meet you. I noticed
a number of things. Fascinating research. One thing that came up for me, I do a lot of diagnostic work
with kids with autism, and I see tremendous potential
for children with autism, assuming that they have
social cognition, delays in theory of mind. There is research from
the child development center. He is now in
Georgia, I believe. Did a lot
of eye tracking research about children with autism. And within
a few weeks, infants, you can see they’re not paying
attention to the right thing. This to me would
be amazing. You would have to test
it empirically, I guess. But whether you would
expect children who have autism to not do well
on these tasks. But we’re still not
diagnosing autism until kids are three,
four, five years old, sometimes later for kids
who are deaf. So I’m wondering what your
thoughts are for using some of these tools
as a diagnostic tool.>>MELTZOFF:
You’re exactly right. I’m so happy that you
made that connection. I think it’s
incredibly important. And if you go
to my website, you’ll see I collaborated on
several studies on children with autism with a person
named Geraldine Dawson. She was at the University
of Washington and she became CEO at
Autism Speaks. And we have done studies
of imitation and gaze following in children with autism and even sibs
of children with autism. Some of those studies
have been published. I think it’s a great example
of what everybody here knows that there is
differentiation that can be made between
different populations that are based on theory
and what you would expect. So children with autism
are delayed in attention, reading,
joint visual attention, imitation and other aspects
of social cognition. We haven’t tested things about somatosensory representation or body representation in children with autism. Although that is something
we want to get to because those studies
are fairly new. But I think
there would be deficits in this like-me
mechanism of recognizing the similarity between
self and others. And that may be
a fundamental issue in the development of autism. But at the same time, to make
a big leap here, we found that deaf children
of deaf parents were advanced in something
like gaze following. Right? So one would want to go
through many different groups and make predictions
about how hearing infants of deaf parents respond to
various tests. Deaf parents of deaf
infants and so forth. There are many subtle tests I think with
differentiable predictions. In your seminar we were
talking about that today. And I think you were
raising those issues about looking at
different populations and what sort of fine-grain
predictions could be made. I think that is really one
of the advantages of being able to do work on
behavior and on brain and with many different
populations with a theory to make predictions
we could make some progress that we’ve been longing
to make for sometime here.>>That was a great
talk, thank you. You had these amazing new
results that you just got while you were
on the plane showing that deaf babies of deaf parents
have better gaze following. In a separate part
of the presentation you talked about gaze following
related to emotion recognition. Would you speculate or
do you know whether there is evidence
speaking to whether deaf babies of deaf parents also have better
emotional recognition or any differences in sort of emotional
responsiveness to others?>>MELTZOFF: You
know, that’s a great question. I think that is exactly
the kind of research that one would want to do based
on the new finding of the basic level
of gaze following. Our emotional recognition
paradigm combines not just recognition of emotion, but combining
another person’s emotion with whether
they’re looking at you now. That’s what our
paradigm is. So we haven’t, we had
to do the first study first about whether
deaf children were good or not good at gaze following. Now we know they’re better. So your hypothesis might
be that they are more advanced on some other aspects
of social cognition, as well. And we simply don’t know
that at the moment. But you’re probably
aware that people have studied deaf
children and made some claims about theory of mind development as measured with false belief. And so combining new
measures of social cognition with those existing studies is something I think
needs to be done.>>It sounds like there is
a lot of research we need to do.>>MELTZOFF: Yeah.>>Thank you very much
for a wonderful talk. My question is how does
this compare with the statistical learning
for categorization that children use for other
things including language and phonology and
so on and so forth.>>MELTZOFF: Yeah,
statistical learning done by Alyssa Newport. The connection to
statistical learning I think is quite relevant. Usually statistical
learning is thought of as detections of patterns
often, but not always having to do with segmentation
connected to time. Imitation of body actions
is not really a phenomenon of statistical
learning as such. It’s a mapping between
body movements they see onto their own body. So it’s not a problem of
statistical learning from that point of view. However, you may be making
a leap that I think is related to the trade
attribution studies that we’ve just done, which is if a child sees an adult
get angry three times in a row it’s as if to use your language or to allow us to be speculative here, it’s as if they’re
picking up that pattern, that statistical connection between emotion and that person. That person tends
to get angry. So now when they see
that person manipulate a completely novel toy where they see another person
manipulate a novel toy and the emoter does not
get angry at it, even though the adult
doesn’t get angry, they now have the
statistics that an adult had been angry at an
action three times in a row, now they hadn’t been angry. Chances are however that this is
an angry personality. So I think you’re right that statistical learning
comes in, in the formation of stereotypes and the formation of making
trait attributions to others and bringing that
literature together so you have sort of social
statistical learning. Statistical learning
of likely behavioral reactions and statistical learning of
personality traits of a person. I think we could begin to use the term to explore that and that would be quite interesting. …and then with the
older infant, as well, what were their
backgrounds? Were their parents
hearing, deaf? What demographic variables
did their parents represent for the
deaf and hearing kids?>>MELTZOFF:
In which population? Which study?>>Any of the infant
studies that you’ve done.>>MELTZOFF: So these
in general, the studies are done in children who
come into the lab at the University of Washington. And so they’re upper
middle class families who have a car and can drive
to the University of Washington. But there are two sets of
studies that I think might be of interest to you
that go beyond that. The first was
the study that we did with Deaf
children of Deaf parents. We tested 12 such children
and we had to go across the country and flew to many
different places to recruit and test
such children. So what I wasn’t able
to present here is full demographic data
of those children, but in the publication
we’ll try to include that. And Melissa, where
is Melissa here? Helped us recruit some of
the children for that study. I hope you saw on the
bottom of the slide, it was done
with Jenny Singleton who was connected to VL2, and
Tom over there provided some funding through Science
of Learning for this study. That was a good example of
two science of learning centers coming together. The other author of the
study was Rochelle Brook, who works with me at the
University of Washington. The full demographics
of that study and the level of proficiency
of ASL of the parents is an important
characteristic of that study. I was just reporting
the main effects here. But we will be
including that. But if you’re interested
in variations across different groups,
in the stereotype work, we’re now doing
two different countries. And one of them, actually
three different countries. One was in Singapore. I’ll mention that first. Singapore, the children,
boys and girls both outperform American
boys and girls in math. But Singapore is also
interesting because the girls on standardized tests
in the elementary school are either equal to
or better than the boys. So it’s a culture where the
girls do better than the boys. And the stereotypes of
the adults in the culture are not as severe
as American stereotypes of boys do math and girls don’t. In Singapore, that stereotype
is not as prevalent or not as strong. So we are extremely
interested and finished a series of studies with
Singaporean children, and although
there is a stereotype in Singaporean children,
it’s delayed and weaker than in American children. So the stereotype held by the kids does seem to be picked up by the culture and is malleable and pushed around by the adult stereotype. We tested kids recently, and I don’t have
the full set of results in a highly
stereotyped culture, the opposite direction
of Singapore, and that was in Chile, where there is quite
a strong stereotype that boys do math
and girls don’t. And I don’t have the full
set of results from the Chilean study
yet, but that is an example where we are trying to pick
cultures to test the theory where we believe it
is going to be that the children’s idea of the
social world and social groups is going to be
taught from the adults in the culture and we believe
that’s going to be an important message not
only for theory but for policymakers and
education that the children are watching us and how the
children are called upon in the classroom, how
the teachers treat the children, how the parents and grandparents
teach the children is going to affect
the children’s stereotypes and eventually
their sense of self.>>Thank you. I stayed because I have
a follow-up question. Have you done any
experiments on children who have Deaf parents that
are both Deaf and hearing. So deaf kids
with deaf parents and hearing kids with
hearing parents? So have you done a study
with children like that? Well, we’ve done deaf
children of deaf parents and hearing children of
hearing parents. What we haven’t done is hearing
children of deaf parents. And that was raised in the
seminar today about — .>>I was thinking about
having both deaf and hearing kids who have
the actual same parents. So like siblings. Families where
they have deaf parents and they have both deaf and
hearing children in a single family.>>MELTZOFF: Well
it’s difficult to find these populations. We just finished
our first study. But I agree with you now
that there is a paradigm, if there are people here
at Gallaudet who would like to collaborate on
helping to find a range of other populations to do
some of these tests of social cognition, I think
from a theory point of view, one can make fairly
fine-grain predictions. And at the seminars we had
this afternoon, I think some of the participants
began to outline particular predictions
like you began to outline here. And I think it would be
fascinating and important to do.>>Thank you.>>Two more questions
and then we will close.>>Like me makes great
scientific sense for those who recognize
the evolution of survival. Because I am like my parents,
I have this genetic need to mirror them, to become
them in my life. And I’m sure in the
seminars this afternoon this topic was discussed. But we have students from
speech, language, and hearing who are asking
why, how do we make sense of parents who are not
like their babies? Parents who are themselves
hearing who give birth to
a child and maybe by day five know that this baby
is not like me. This baby is not going to
sing like me or talk like me or play a musical
instrument like me. What would you comment for
these students who are wondering why don’t these
parents just automatically have a desire to acquire
sign language and become fluent signers
for their babies?>>MELTZOFF: Well
that’s a profound question. On the one hand it maybe
shows the power of like-me and grouping and
recognizing other human beings and having a real
being impelled to make that comparison. Now that comparison has
good consequences and bad consequences. The stereotypes, the idea
of social grouping and separating people into
like me and not like me, us and them, is a bad
consequence of the idea that human beings seem
so compelled to do this mapping and hunt out for
others that are like me. I do think, though, that
many people are able to have a wider circle of
what counts as like me, being human like me,
having emotions like me, having a more fundamental
sense of what it means to be like me, not just say
to dress like me or have red hair like
me or blue eyes like me. So there are other
features that one could think of more superficial,
not the essence of humanity. So I think there is a drive
to separate us and them, but there is the ability
perhaps to teach our children and raise our
children with values that allow them to
widen the circle and include a greater aspect of humanity
as being like myself. And different than other
things in the world. And there are actually
cultures that draw the line beyond humanness and
all animals are like me. And we should not do harm
to any living thing. Do you see what I mean? So what you count and what
dimensions you use as the like-meness I think
can partly be moved by acculturation and the values
of the society you’re raised in. The thing I think
infants start with, as a dimension of like me
is body action. Which does pick out the universe of human beings that have morphology like me
or move like me. And that’s a
pretty broad class.>>It’s getting very late,
so I’ll be very brief. You talked about the role
of reward early on in your talk. You were saying that reward doesn’t play much role in
imitation learning. And you gave
an example of the parents actually telling
off their children and they did it anyway. On the other hand, you
were showing all these lovely movies
of how happy the children were when they
accomplished something. They were able to do
the job and imitate something. And the smile alone tells
me that their reward center in their brain
was lighting up and they were certainly
enjoying this. Doesn’t this constitute
some kind of internal reward that they get much
stronger than the reaction of the parents? The parents also sometimes
do react unconsciously in a very affirmative way. And that could constitute
formal reward.>>MELTZOFF: Yes. I think there are many
senses of reward. But you just articulated
two of them that I deeply endorse. So the child seems to have
an intrinsic sense of award from being like the other
or accomplishing their goal or acting like
the other person acted. So the adult put beads
in the cup and then the children
duplicated it and became happy. That was not. So you’re
absolutely right. The child
felt a sense of award and an intrinsic
sense of reward. I’m not saying that reward
is differentiated from and has no role in imitation. The question is
the source of the reward. So the child was
intrinsically rewarded for acting like
the other person. Or achieving the goal that they
saw the other person achieve. What I was trying to
say is that intrinsic motivation does drive children
to imitate more than or they do it even
in the absence of the adult rewarding them
from doing it. It’s not that the baby
sees the adult do something and then the
baby reacts randomly and the adult has
to shape the baby in order for that random movement to
become a duplication of what the adult did. Right? Then you would be using reward to shape the baby to imitate. What I’m claiming is that the baby can
see you perform actions, map it onto their own body, and when they do what
they see you do they have an intrinsic
reward or positive feeling from having done it. And I think you’re
right for articulating those two distinctions. And I think it’s
theoretically of deep interest that the child would feel
good and feel pleasure from having reached the goal
that they see you reach. Or act like they see you act. That’s a pretty big
impetus for social development if a human
being feels that way.>>It’s a big mystery as
you also pointed out and provided an
answer for to some extent in your classical early
experiments. That some of this
behavior is innate. Nevertheless, I think it’s
still a mystery how the child learns to access
the right body part when they imitate a behavior. I think that’s where the
plasticity comes in and the reward learning
could strengthen it.>>Word learning?>>Reward learning. How the intrinsic reward
might play a role.>>MELTZOFF: Right. I think there is innate state that is transformed
through experience. And how that experience,
whether it’s self-experience or continuing to perceive
the actions of others, how that changes the
infant’s internal structures I think is
deeply important. So experience does reach in and change those initial representations. They are altered from the baby
using their own body and watching
other people use the body. And looking at that
process of development is what developmental
science needs to do, how does the behavior, how has
the behavior changed the brain? And how does the brain
maturation and changes affect the baby’s
motor behavior? That sort of interchange
is deeply important. [Applause]>>Okay. So let me make three
brief comments. First, there is a questionnaire I hope everyone
is willing to fill out. I’ll pass out more copies
if you don’t have one. Second, thank you
to our colleagues from neighboring universities who came here for this lecture. Third, I’m sure a lot of
you have questions and comments. We have a wonderful little
reception waiting for us. Please join us and come
talk a little more, express your ideas. We’re happy
to have you here. It’s in the SLCC. Okay? Anyway, follow us.

Leave a Reply

Your email address will not be published. Required fields are marked *