Highlight text to annotate itX
Lecture 36: Continuity and change in cognitive development.
In closing this series of lectures on psychological development, I want to talk about some overarching
themes in our study of this topic. The earliest theories of psychological development focused
on problems of maturation and learning. In general these theories offered a view of the
child as what you might call a short, stupid adult who grows smarter as he or she grows
bigger. Viewed in this way, there is a continuum between childhood and adulthood -- only a quantitative
difference, with no abrupt qualitative changes. We can see this view in a classic study of
development by Arnold Gesell, one of the founders of developmental psychology, who trained some
children -- these were actually twins -- in walking and climbing stairs. Twin T received
training early on in life, while Twin C, the control twin, did not. In fact, the trained
twin did show walking and climbing earlier than the untrained child, but the untrained
child quickly caught up and further progress wasn't accelerated by training. Both children
advanced beyond walking at the same pace. The conclusion is that walking and climbing
are behaviors that just emerge at the appropriate time as the child grows older.
Another study involved traditional Hopi children who are swaddled and bound to a cradle board
for the first year of life. Except for a relatively short period of time each day, they're held
pretty much immobile, severely restricting motor behavior. But it turns out that, once
released from the cradle board, there's no retardation in the emergence of walking. So,
in the Gesell case, the training didn't affect the development of walking; and in the Hopi
case, the training didn't affect it either. It just happened, when it was time.
In fact, Gesell and his colleagues introduced a set of developmental schedules, which indicated
various milestones in development -- the point at which the average child, normal child, would be expected
to show certain kinds of motor or adaptive behaviors, use language in a particular way,
or show certain personal or social characteristics. Normal children achieve these milestones just
by the process of maturing, as a function of growing older.
Turning to cognitive development, perhaps the clearest example of the view of development
as a matter of quantitative change is the measurement of intelligence by Binet and Simon,
in France, and by Terman in the United States. You'll remember that Binet measured intelligence
in terms of the child's mental age, which he assumed was correlated with chronological
age. Louis Terman adopted the idea of mental age when he introduced the formula for calculating
the intelligence quotient or IQ as the ratio of mental age to chronological age times
100. And, for that matter, David Wechsler did much the same thing when he came up with
the idea of deviation IQ -- so that an individual with an IQ of 100 has the same IQ test score
of the average person in his or her age group. In each case, the underlying assumption appears
to be that normal children will simply grow more intelligent as they grow up.
Despite debates over the heritability of IQ, the classical continuity view is that the
child gradually acquires knowledge through learning, where learning was construed as
tantamount to the sorts of things we studied in our lectures on classical and instrumental
conditioning. John Locke, an English philosopher of the 18th century, famously argued that
the infant is a tabula rasa, or blank slate, which is written on by experience. In the
Lockean view, development is a matter of learning more than you already know. This contrasted
with the earlier view of Descartes, who assumed that some knowledge was innate as a gift from
God. Whether the theoretical focus is on biological maturation or on learning, in either case
the process of development is viewed as a matter of continuous quantitative change.
The infant starts out small, physically and mentally, and gets bigger, physically and
mentally, as he or she grows up.
Later theories of development, especially theories of cognitive development, began to
focus on qualitative changes. In this view, the child is not just a short, stupid adult,
but rather the young child thinks differently than older children and adults. Thus, the
developmental differences are qualitative
-- differences in kind, not just quantitative or differences in amount. Children are not stupid
compared to adults, but their intelligence needs to be appreciated on its own terms.
This was the view of development propounded by Jean Piaget, a Swiss psychologist who actually
worked with Alfred Binet at an early stage of his career, but who argued that the development
of intelligence proceeds through a sequence of stages, each defined by cognitive
landmarks. An important concept in Piaget's theory is
the schema -- a term familiar from our work on the perceptual cycle, and also Frederic
Bartlett's work on memory. For Piaget, a schema is an organized mental structure that produces
a particular coordinated response to a certain class or category of stimuli. Thus, a schema
is a kind of concept that renders diverse objects functionally equivalent. That is, you respond
the same way to them because they all belong in the same class. In a very real sense
for Piaget, cognitive development is the development of schemata, their elaboration and differentiation
as a result of experience. The interaction between schemata and the objects
with which they come into contact is characterized by the twin processes of assimilation and
accommodation. By virtue of assimilation, the mental representation of the stimulus
is altered so that it will fit into the schemata employed to process it. By virtue of accommodation,
the schema itself is altered so that it can receive the stimulus. Thus, the final representation
of the stimulus is a sort of a compromise between what was expected, the schema, and
the actual stimulus itself. Piaget believed that the neonate confronts the world with
a primitive set of innate schemata. Through assimilation and accommodation, the nature
of these schemata gradually changes. At certain points, however, the changes in mental schemata
are so dramatic that they appear qualitative in nature, rather than merely quantitative.
And these shifts mark the child's move from one stage of cognitive development to another.
Each of these shifts, in turn, was marked by the child achieving some landmark behavior.
The first of Piaget's stages is sensory-motor intelligence, also known as the sensory-motor
period, which encompasses about the first two years of life. In this stage, the world
of the child is one of unrelated sensory experiences and reflex-like motor reactions to them -- sensory
and motor activity. To take a phrase from William James, the world of the sensory motor
infant is a world of blooming, buzzing confusion. But not quite, because according to Piaget
the child has these innate sensory-motor schemata that it uses to deal with the world. One of
these schemata might be described as, "If you touch an object, put it in your mouth".
That works pretty well if you're an infant nursing, but infants also put all sorts of
other objects in their mouths. The object is assimilated to this sensory-motor schema.
One of the major accomplishments of the sensory motor period is the development of object
permanence. For Piaget, the newborn's behavior is tied to what comes through its senses.
Out of sight is out of mind. A child will look at a toy that's placed in front of him,
and probably try to put it in his mouth. Once the toy is blocked from view, the child just
loses interest -- it's as if it's not there anymore. This is because, during the sensory-motor
period, Piaget argued that the child doesn't have the ability to form mental representations
of stimuli. The child can only react to the physical presence of the stimulus. Eventually,
however, the child comes to behave as if it has internal mental representations of objects
that are not actually present in its sensory environment. So at this point, if a toy is
hidden from view, they'll look for it. This searching behavior shows that the child has
an idea of the object that persists despite its physical disappearance. During the sensorimotor
period then, the child doesn't have object permanence, shows a failure of object permanence. The first landmark of cognitive
development, achieved about 24 months of age in Piaget's view, is object permanence. The
child shows that he or she has acquired the capacity for forming internal mental representations
-- memories, if you will -- of the outside stimulus world.
At this point, about 24 months after birth, the infant moves fully into the next stage
of cognitive development: the preoperational period. The child is able to form and retain
internal mental representations of objects and events, but these representations exist
as individual mental units unrelated to each other. The major achievement of the pre-operational
period is the ability to relate one representation to another thru higher-order schemata called
operations. This takes about the next five years.
The development of operational modes of thought is marked by the emergence of conservation,
which occurs by the time the child is about seven years of age. In the earliest portion
of the pre-operational period, the child just does not conserve at all -- he shows conservation
failure. If a short, wide cup of liquid is poured into a tall, thin glass, the child
may well say that there is more liquid in the former than the latter. This is because
the child can track changes in either height or diameter, but not both simultaneously.
Thus, the child cannot understand that volume, which is a product of height and diameter,
remains constant when the liquid is poured from one container into the other. The same
thing goes for conservation of mass. If you take a thick lump of clay and stretch it out
into a thin straw, the child may very well say that there was less clay then there was
before. And if a child is shown a set of objects lined up over a short distance, and later
the same objects are arrayed over a longer line, he or she is likely to say that there
are more objects in the latter case than the former. At some point, however, the child no longer
makes these mistakes. He or she's acquired the ability to consider height and diameter,
number and distance, length and width simultaneously and to compensate for one with the other.
At that point, the child has acquired the ability to conserve.
An analog of conservation failure in the interpersonal domain is egocentrism -- which is not to be
confused with selfishness. From Piaget's point of view, egocentrism reflects the child's
inability to take another person's point of view, or to appreciate the viewpoints of
other people. In a sense, ego-centrism reflects the child's inability to conserve viewpoint
across changes from one person's viewpoint to another. Egocentrism is often assessed
with the so-called Three Mountains task introduced by Piaget and his collaborator Barbel Inhelder.
A child is situated at one end of a display with three mountains, and a doll is put on
the opposite end. The child is asked if the doll can see the cross and the child says
"Yes" -- and that's true. The cross is in both their fields of view. But when the child
is asked whether the doll can see the chapel, she also says yes. She fails to appreciate
that the mountain with the cross on top would block the doll's view of the chapel. The egocentric
child does not appreciate that other people may have a view of the world that's different
from his or her own.
Here's some data from a study of age differences in performances on the Three Mountains task.
Four-year-olds tend to give egocentric answers on the three mountains task. They say the
doll can see the chapel when it can't. Egocentrism starts to disappear when the child is six
on average. By the time the kids are eight or ten years old, they're giving the correct
answer. They appreciate that the doll can't see something that they can. Like conservation
failure, egocentrism is a characteristic of pre-operational thought, from Piaget's point
of view. And the loss of egocentrism, like the acquisition of the ability to conserve,
marks the child's movement from the pre-operational stage to the next stage, the stage of concrete
At about age 7, the child enters the stage of concrete operations, in which children
are capable of thinking and reasoning about objects and events which they've actually
experienced. Concrete-operational children are actually pretty powerful thinkers. They
can conserve, of course, and they don't show egocentrism anymore. They can pay attention
to objects other than the most salient ones in their environment. They can take another
person's point of view. They can take account of transformations in state and they can classify
objects into groups based on shared properties. They can also generate and use hierarchical
classification schemes -- schemes involving subsets and supersets and all of that. You
can get along pretty well in life with nothing more than concrete operations, as long as
you're reasoning about familiar problems involving familiar objects and events -- which, after
all, is what most of us do most of the time.
Unfortunately concrete operations aren't always enough, especially if you are living in a
so-called advanced society. It is often useful to go beyond our own past experiences and
to reason about things that we haven't seen or touched. Or for that matter to be able
to reason about things that aren't visible or touchable. Concrete operations just don't
suffice for this purpose. So it's a good thing that around age 12 the child enters into the
last of Piaget's stages, which he called formal operations. In formal operations, thinking
can be purely symbolic, without referring to anything at all by way of concrete objects
and events. The hallmark of formal operations is scientific thinking, which is what lies
behind Piaget's notion of the child as a naive scientist.
This scientific thinking is marked by four different qualities. First, hypothetico-deductive reasoning,
in which we can hypothesize about a certain state of affairs and then reason deductively
from that hypothesis -- that is, we can assume that something is true without requiring
it to actually be true. Second, inductive reasoning, in which the person generalizes from specific
observations to general principles. Third, reflective abstraction in which the child
is able to reflect on is or her own thoughts, in order to arrive at novel ideas. And fourth,
propositional logic, in which the child can reason about two or more abstract entities
that are represented by statements such as, if there is a p, then there is a q, without
knowing what p and q refer to. Children who have developed the capacity for formal operations
are able to deal with several abstract variables at the same time.
Of course, Piaget was not the only theorist to offer a conception of development in terms
of stages. Long before Piaget, Sigmund Freud had offered a stage theory of psychosexual
development running essentially from birth to adolescence. And in the 1960s, Erik Erikson,
a follower of Freud's, offered a stage theory of psychosocial development that encompassed
the entire life cycle from birth to death. Piaget, Freud before him, and Erikson
had an enormous influence on thinking in developmental psychology. Their idea of development as a
progression through a succession of qualitatively different stages was received quite enthusiastically.
And you can see the legacy of these individuals in the proliferation of stage theories of
everything in developmental psychology. For example, Lawrence Kohlberg was heavily influenced
by Piaget, and also a colleague of Erikson's at Harvard, introduced a stage theory of moral
development -- basically taking Piaget's idea and applying it to moral reasoning, not just
cognitive development in general. For Kohlberg, moral reasoning is the basis for ethical behavior,
and the individual moves through a set of six or maybe seven identifiable stages, throughout
development. At early stages moral reasoning is focused on self-interest: How can I avoid
punishment? What kind of reward am I going to get? At a later stage, moral reasoning is
focused on concrete rules: What am I supposed to do? What do other people want me to do?
What are the rules? And in the final stages, moral reasoning is based on abstract principles
like justice and equality. You can see how this fits in with Piaget's ideas. Moral reasoning
is very egocentric in the pre-conventional stages. The child is concerned about what's
going to happen to him or her. It's very concrete in the conventional stages where the child
follows the rules because they're the rules. And it involves abstract reasoning in the
post conventional stages.
Stage theories of development of the sort introduced by Piaget became very popular in
developmental psychology in the 60's and 70's, but just as they came to dominate the field,
a number of experimental psychologists began to find cracks in the edifice, empirical problems
that suggested that Piaget's theory, at least, wasn't quite right.
One of these problems is known as decalage. For example, the development of operational
modes of thought is marked by the emergence of conservation, which occurs about the time
the child is about seven years of age. But if there's really a fundamental shift in thought,
a qualitative change, you'd expect this shift to occur relatively rapidly and to be widely
generalizable. But it turns out not to be the case. A child who successfully conserves
quantity may not conserve mass, and may not conserve number; a child who conserves mass
may not conserve quantity, and may not conserve number as well; and so on. There just doesn't
seem to be this wholesale shift in mode of thought that's implicated in Piaget's theory.
Moreover, researchers discovered that the lower boundaries of Piaget's stages are not
nearly as rigid as he suggested that they were. Is it really true, for example, that
children younger than the age of 7 are generally incapable of abstract thought? Or that pre-operational
children do not have a concept of number? A very large research tradition has developed
out of questions like these and the general conclusion is that cognitive development is
probably more continuous than Piaget thought it was, and that even very young children
can show amazing powers of thought -- at least in limited domains.
Such findings undermine the general thrust of the Piagetian stage approach to cognitive
Let's consider, for example, Piaget's claim that pre-operational children, children aged
two to about seven, don't have a concept of number; they're just not capable of it. In
one of the earliest critiques of Piagetian theory, Rochel Gellman and Randy Gallistel
analyzed just what's entailed in having a concept of number. Their analysis yielded
a set of three principles that define correct counting. Counting, remember, requires the person to
use a set of labels like the words one, two, three and four, and to assign these to objects
depending on how many of those objects there are. So the first principle you need is known as one-to-one
correspondence. The items to be counted have to be put in to one-to-one correspondence
with a set of number tags that are used to count. Second, the stable order principle
says that the number tags themselves have to be used in a consistent fixed order. And
third is the cardinality principle, which says that the last number word that's used
in a count has to represent the number of items that have been counted.
So if you and I were counting dogs, we'd say one, two, three, four, four dogs. That count
shows one-to-one correspondence. And then if I do it again, one, two, three, four dogs
-- that shows stable order. I've used the same order twice and the last number I use
represents the number of dogs there are. But suppose I'm a little kid and I don't know
my number words all that well yet. You ask me to count the dogs and I say one, five,
three, eight -- eight dogs. I'm still showing one-to-one correspondence. And I do it again:
one, five, three, eight -- eight dogs. I'm showing stable order and I'm showing cardinality.
There are eight dogs. Now what I mean by eight isn't what you mean by eight, but that doesn't
matter. I'm still applying the counting principles and that's the same as if I used no number
words at all. Suppose you asked me to count and I said, blitz, blatz, bluck, blit -- blit
dogs; and you said, do that again, and I said blitz, blatz, blut, blit -- blit dogs. I'm
still showing one-to-one correspondence, I'm still using my number terms in a stable order,
and I'm still showing cardinality. From the point of view of the child, I've got a sense
of number. It's just not expressed exactly the way you'd like me to express it. Later
on, I'll learn the number words and I'll learn to use them correctly. And I'll count one,
two, three, four dogs -- just like you do. But before then, I've still got a sense of number.
Gelman and Gallistel, among others, have shown that even pre-operational children, three
and four years old, apply these counting principles in their own way. They've got a sense of number;
they know how to count, at a time when Piaget said they don't and they shouldn't know it.
So Gellman and Gallistel pushed the concept of number back into the preoperational stage
-- which raises the inevitable question, how low can you go? How far back can you push
the beginnings of a concept like number? It turns out that, if you're a clever enough
experimentalist, you can push it back pretty far. Here's an experiment by Karen Wynn using
four- and five-month-old infants. They're not fresh out of the womb, but they're pretty
young. They're firmly in what Piaget would have called the sensory-motor stage, far away
from the boundary that divides sensory-motor intelligence from preoperational thought.
Here's an experiment by Karen Wynn that employed looking-time as a measure of attention in these very young infants,
the idea being that a child who is surprised by something will look at
it longer. There were actually two groups of infants run in this experiment: let's look
at them one at a time. In what Wynn, called the 1+1 condition, the infants were shown
a single item, like a small mouse puppet, in a display area. Then a screen came up to
obscure the mouse. The experimenter then introduced a second mouse, pretty much identical to the
first, and then put that behind the screen as well. Then the screen dropped and revealed
one of two possible outcomes. Either the screen dropped to reveal two objects or the screen
dropped to reveal just one object. The idea is that if the child is counting, doing anything
like adding, he will expect to see two puppets when the screen drops, but will be surprised
to see one puppet and will look longer. Similarly in the 2-1 condition, the infants
were first shown two mouse puppets. Then the screen came up, the experimenter took one
of those objects away, and finally the screen dropped to reveal either one mouse puppet
or two mouse puppets. If the child is doing anything like counting, anything like subtraction,
it'll expect to see one puppet when the screen comes down, but will be surprised to see two.
And that's just what happened. In the 1+1 condition the infants looked longer when the
screen came down to reveal just one item. It's as if they added 1+1, expected to see
two puppets, and were surprised when they saw only one. Similarly, in the 2-1 condition,
the infants looked longer when the screen came down to reveal two items. It's as if
the infants subtracted one from two, and expected to see only one item and were surprised to
see two. So the fact that the children in the 1+1 condition are surprised when they
see only one object, and the children in the 2-1 condition are surprised when they see
that there are still two objects, suggests that even at this very, very early age the
children are engaged in something like rudimentary addition and subtraction.
So this experiment offers what you might think of as a kind of double-whammy to Piagetian
theory. Remember that these infants are supposed to be in the Piagetian period of sensory-motor
intelligence. They respond only to what they can see and hear, and they haven't formed
any internal mental representations of the objects that they've encountered with their
senses. Out of sight, out of mind. But if that were the case, there would be nothing
to add to in the 1+1 condition, and there would be nothing to subtract from in the 2-1
condition. After all, there's a screen up there, hiding the objects from view. So it's
very clear that even these infants, just four or five months old, have some sense of object
permanence. They know that the object is behind the screen, even though they can't see it,
and they can add to and subtract from what they know is behind the screen. If four and
five month old infants have object permanence and they already have a rudimentary ability
to count and add and subtract, there's really no point in postulating qualitatively different
stages in cognitive development.
Okay, now let's take a breath and review the bidding. Piaget's theory is a stage theory
of development. And like all stage theorists, Piaget argues that the stages of cognitive
development are universal, obligatory, stereotyped, and irreversible. All normal individuals must
pass through them in the course of development. The stages are passed through in the same
sequence for all individuals. Once a stage has been successfully negotiated, there's no going back.
And the achievement of one stage is a necessary condition for advancement to the others. These
are essentially hypotheses about the nature of development. And it turns out when they
have been seriously tested, research has usually failed to confirm them. In particular, research
testing Piaget's theory has revealed a number of anomalies which led investigators to refocus
their theories on continuities in mental development, not qualitative changes.
One post-Piagetian approach to development construes development as the development of cognitive skills. According
to this view, the infant starts out as a novice in all domains of problem-solving and gradually
acquires expertise through learning -- through experience and practice. However, in this
view, expertise is not just a quantitative difference where experts just know more than
novices do. Rather, the argument is that experts represent problems in a qualitatively different
manner than novices do. For example, it's been argued that expert knowledge is cross-referenced
so as to enable the problem-solver to get easy access to it in a variety of situations.
Moreover, expert knowledge focuses on higher-order patterns, so that experts think in bigger
chunks and take larger steps in problem solving than novices do.
One model for the development of expertise is the difference between novice and expert
chess players. Both kinds of players know the rules, but experts represent the game
differently and play the game differently than novices do. Now, this may seem like it's
just a variant on our old theory of development as learning. Remember our definition of learning
as a relatively permanent change in knowledge that results from experience. In the same
way, expertise develops with experience and practice. However there are at least two important
differences between the theory of expertise and a mere theory of learning. The first of
these is that the acquisition of expertise involves qualitative leaps and skill that
represent the individual's successive reorganizations of task performance. These qualitative leaps
are, in some sense, analogous to Piaget's stages. But they're not the same as Piaget's
stages because even young children can obtain more expertise than adults. If you don't believe
me, ask any eight year old boy about a dinosaur. The theory of expertise also doesn't consider
the infant as a blank slate to be written on by experience the way that traditional
learning theory would. Instead, the child is viewed as bringing a rudimentary cognitive
apparatus into the world, such that learning experiences modify his or her innate
propensities. These differences between the modern theory of expertise and the traditional
theory of learning reflect, I think, the lasting contribution of Jean Piaget to developmental
theory -- despite the fact that his theory appears to be wrong in many salient details.
Still, theories of expertise seem to imply some degree of continuity over the course
of development. But studies of expertise reveal one big difference between younger and older
children. Put bluntly, older children know what they're doing, while younger ones don't.
Older children are not simply more expert than younger ones at various tasks. They're
also more reflective, more deliberate, more strategic in their thought and action. In
other words, what older children possess and younger children lack is meta-cognition
-- a term first introduced by Lila Gleitman (wife of your textbook's author), and subsequently
popularized by John Flavell. Meta-cognition is, literally, cognition about cognition -- or,
put another way, our ability to monitor and control our own cognitive processes. Meta-cognition
is one's knowledge about cognition and cognitive phenomena. Gleitman and her colleagues pointed
out that the developing child doesn't just acquire a language, he or she also acquires
a meta-language. The child becomes a more competent grammarian, and is better able to distinguish
between grammatical sentences and nongrammatical sentences. Similarly, Flavell and others pointed
out that we develop meta-memory, knowledge about the principles that govern how memory works,
principles like elaboration and organization and things like that.
So there appear to be at least two different components to meta-cognition. First, knowledge
of what's in your own mind, knowing what you know. Knowing whether you're perceiving something
or just imagining it, what knowledge you have stored in your memory, and whether
you actually understand something you've learned when it's been explained to you. The second
component is the person's appreciation of the rules that govern various mental processes.
How we can deploy attention effectively; how we can use strategies for effectively encoding
and retrieving memories; and how we can break down large problems into sub-problems.
This kind of knowledge is the hallmark of expertise. But it's also not just knowing something,
but knowing that you know something and knowing how you know something.
In fact, Flavell has argued that there are several different aspects of cognitive monitoring,
or meta-cognition, including goals or tasks -- knowing the objectives of some cognitive
enterprise; actions or strategies -- knowing what cognitions or actions you can employ
to obtain your goals and complete some task; meta-cognitive knowledge, or knowledge about
factors that influence your own and others' cognition; and meta-cognitive experiences
-- conscious thoughts and feelings that pertain to cognition. When you work on a problem for
a while or someone explains it to you, and then you have that ah-hah! experience -- I've
got it, I understand it now! -- that's meta-cognition. The general idea is that with development,
children become more expert at cognition, but they also develop meta-cognitive knowledge.
They make more conscious, deliberate use of their cognitive faculties.
More broadly, we might say that, as they develop, children come into possession of a theory
of mind -- a term coined by David Premack and Guy Woodruff,
and popularized by Simon Baron-Cohen (yes, he's the cousin
of Sasha Baron-Cohen). Put briefly, the theory of mind entails the ability to impute mental
states to ourselves and to others. It includes knowledge of our own minds, and knowledge
of other minds. With respect to knowledge of our own minds,
we mean understanding that we have mental states -- that we perceive things and feel
things and the realization that our experiences are our own -- that somehow our experiences
are separate from the world outside the mind. And that we can control our own beliefs, feelings,
and desires. Knowledge of our own minds also entails phenomenal awareness: introspection,
consciousness of what we think, feel, and want.
Knowledge of other minds, by contrast, entails the understanding that our mental states may
differ from those of other people - that different people have different minds, and thus different
experiences. And our knowledge of other minds entails an ability to make inferences about
what other people think, and feel, and want, based on what they say and other aspects of
their behavior. In some ways, the idea of a theory of mind revives Piaget's notion
of egocentrism. For Piaget, the preoperational child thinks that his experience is universal.
But the older child, having entered concrete or formal operations, understands that others
may not have the same experiences, may not have, may not perceive the world the same way as he
or she does. In fact the theory of mind usually emerges between five and seven years of age
-- exactly the same point as, Piaget argued, there occurs a shift from pre-operational
thought to concrete operations.
The young child's acquisition of a theory of mind is often assessed with what's known
as the false belief task. It involves an experimenter, a child, and a puppet. The
puppet hides an object like a ball in a container, like an oatmeal container. Then the puppet
is put away in a cabinet, and the experimenter and the child switch the ball from the oatmeal
container to a box. Then the puppet is brought back to look for the ball, and the child is
asked where the puppet will look. Three- and four-year-old children, who have just put
the ball in the box, will say the puppet will look in the box and when asked why, they'll
say something like "Because that's where it is". But, four- to five-yea-old children
will say that the puppet will look in the oatmeal container -- and when asked why, they'll
say "Because that's where he thinks it is". The three-year old doesn't have
the ability to appreciate that the puppet's mental state, the puppet's knowledge, is different
from hers. But the five-year- old has this ability.
Here is some data on typical performance in the false belief task. Children who are about
3 to 4 years old, 30 to 50 months, tend not to pass the false belief task. But children
aged 50 months or older, after they got to be about 4 years old, tend to pass the task.
And by the time the children are 6 years old, certainly 7 years old, almost all of them
are passing the false belief task. They know what they know and they have the sense that
other people may not know the same things they know. They have a theory of mind in that
sense. Now of course, whenever you set up an age
boundary like this, some developmental psychologist is going to come along and say, I wonder if
you can get that in younger kids if you ask the question the right way. And the answer's
yes. Experimenters have devised very clever nonverbal versions of the false belief task,
usually employing a variant of the looking paradigm that I described earlier. And with
these kinds of experiments, they've been able to show that infants as young as 15 months,
even younger, have some appreciation that they may know things that
other people don't know. And they have some ability to discern the cognitive,
emotional, and motivational states of other people. So on this evidence then, even very
young children, still in infancy, have some rudimentary theory of mind.
But it turns out that the child isn't just developing a theory of mind. The child is,
at the same time, developing a theory of the whole world -- of physics, and biology, and
sociology, as well as psychology. The same child who's trying to figure out how other
people think is also trying to figure out how the physical world works, and how the
biological world works, and how society is organized, and how that works, all of it.
This is the essential proposition of what has come to be known as the "theory theory"
of cognitive development. The idea that the developing child is engaged in a continuous
process of proposing, testing, revising, rejecting and elaborating theories of how the world
works, including theories of how minds work as part of the world.
The theory theory is the latest stage, if you will, in the development of theories of
cognitive development and it's been promoted by Alison Gopnik with her colleagues Henry
Wellman and Andrew Meltzoff, among others. These theorists begin, in a sense, with Piaget's
metaphor of the child as a naive scientist actively exploring the world, actively experimenting
in the world, formulating hypotheses about how various aspects of the world works, gathering
evidence that bears on those hypotheses, and revising those hypotheses in accordance with
the evidence. But whereas Piaget argued that the child as a naive scientist was really
a characteristic more of formal operations, Gopnik and her colleagues propose that the
child is behaving as a naive scientist right from the beginning. The child is engaged in
a process of developing theories about the world -- abstract, coherent knowledge systems,
just like scientific theories, except certainly less formal, that the child uses to interpret
and explain events, and to predict and control what happens in his or her world.
How does the child do this? Well, the short answer is through learning. Gopnik and her
colleagues proposed that by observing events in the world, children learn about the world
from conditional probabilities, the conditional probabilities that link one event that they
experienced to another; and by behaving in the world, by engaging in some activity and
seeing what the consequences are -- what they call learning from interventions. If you think
that learning from conditional probabilities looks like classical conditioning and learning
from interventions looks like instrumental conditioning, you wouldn't be far off. The
child is actively engaged in the process of acquiring knowledge about how his or her world
operates. And, as I noted in earlier lectures, something like classical conditioning, and
something like instrumental conditioning -- these are very powerful vehicles for acquiring this
kind of knowledge. But, wait a minute. It looks like we've come
full circle. After all that discussion of Piaget, and expertise, and meta-cognition,
and theory of mind, it looks like we're back to Locke's conception of the child as a tabula
rasa -- a blank slate that's written on by experience. But that is not the intention
of Gopnik and her colleagues. In the first place, they argue that the child brings into
the world an innate theoretical capacity -- an innate cognitive apparatus that permits the
child to form and test and revise his or her developing understanding of the
world. In fact, they go so far as to argue for what they call starting-state nativism.
This is the idea that the child is actually born with what they call substantive innate
theories of various domains like physics and biology -- and, for that matter, the mind and
society. They are not completely clear on what these innate, primitive theories might
look like, and maybe they are wrong about that, but still the child is not a tabula
rasa. The child comes into the world with an innate capacity for theory -- an innate
ability to form theories and test theories and revise theories that he or she deploys
almost from the beginning of life. Moreover it's not the case that the child is merely
mind is written on by environmental events. Rather, Gopnik and her colleagues argue that
the child is actively engaged in theory testing. In the same way that we describe classical
and instrumental conditioning in terms of an organism actively trying to predict and
control its world, Gopnik and her colleagues argue that, right from the start, even as an
infant, the child is actively engaged in theory-testing -- trying to understand surprising events
as well as he or she can, and to test and revise his or her understanding by conducting
experiments -- by trying to figure out if this understanding that he or she has, actually
allows him or her to predict or control events. The child is a scientist, trying to figure
things out. And the world, the physical world, the biological world, the world of the mind,
and the world of other people: that's the laboratory.