How do babies manage to learn so much, so quickly? Alison Gopnik , professor of developmental psychology, explains

Kiddy thinks

When my son was a toddler his first question about a meal was always: "What's for dessert?" One day we had pineapple in kirsch. He spat it out, then looked at the adults devouring the stuff, and said: "Pineapple: it's yucky for me but it's yummy for you." For weeks afterwards, he would stop suddenly in the middle of a game and say: "Pineapple: yucky for me but yummy for you," as if he had discovered the most extraordinary fact of life. And in a sense he had: the realisation that people think and feel differently is a profound one.

When we look around a room full of people, we don't see bags of skin and cloth draped over the furniture. We see other people, people with thoughts and emotions, desires and beliefs, sometimes like our own, sometimes not. And by the time they are 18 months old, this is what toddlers see as well. But how do such tiny children get from bags of skin to "other minds"?

In the past 30 years we have learned more about what young children know and how they learn than we did in the preceding 2,500 years. And this has revolutionised our view of children. For centuries, psychologists and philosophers agreed that babies were the opposite of adults. They were emotional and passive, dominated by perception and incapable of rational thought. John Locke said they were "blank slates".

Today, scientists have only recently begun to appreciate just how much even the youngest babies know - and how much and how quickly they learn. There are three elements to this new picture. First, that children know a great deal, literally from the moment they are born. Second, that they are born with extremely powerful learning abilities. And, finally, that adults appear to be "programmed" to unconsciously teach babies and young children just the things they need to know.

In How Babies Think, my co-authors and I argue that very young children use the same strategies as scientists. They think, observe, formulate theories, make predictions, and do experiments. They also change their theories as they accumulate counter-evidence to their predictions.

But where scientists focus their attention on distant stars and invisible microbes, babies concentrate on everyday things: blocks, pet dogs, words and, most important: Mum and Dad and Aunt Ethel. In fact understanding other people seems to be one of the most crucial items in the scientific agenda of childhood, and it's a good illustration of how early learning takes place.

To begin with, children are born knowing that people are special. Newborn babies (the youngest tested was only 42 minutes old) can imitate facial expressions. There are no mirrors in the womb; newborns have never seen their own face. These tiny babies must somehow already understand the similarity between their own internal feeling (of, say, sticking out their tongue) and the external face they see (a round shape from which something pink protrudes). Newborn babies not only prefer faces to things but also recognise that those faces are like their own face . Nature, it seems, gives human beings a jump start on the Other Minds problem.

And what a jump start. By the time they are nine months old, babies can tell the difference between expressions of happiness, sadness and anger, and understand something about the emotions that produce those expressions. By the time they are one, they know that they will see something by looking where other people point; they know that they should do something by watching what others do; and they know how they should feel about something by seeing how others feel.

For instance, an adult can look in two boxes. She looks into one with an expression of joy and into the other with disgust. The baby will happily reach into the box that made her happy, but won't touch the box that disgusted her. The baby hasdiscovered that its initial emotional rapport with other people extends to joint attitudes towards the world. In a simple way, one-year-olds already participate in a culture. But as babies learn that people usually have the same attitudes towards objects as they do, they are setting themselves up to learn something else, something more disturbing: they discover that sometimes people don't have the same attitudes.

Observe what happens when a baby reaches for a forbidden object - a lamp cord, say. It must seem perverse to the one-year-old: the more clearly she indicates her desire, the more adamantly her carer keeps it away. Even though the baby and the grown-up are reacting to the same object, their attitudes toward the object seem to be different.

By the time babies are about one-and-a-half, they start to understand the nature of these differences between people and become fascinated. If you offer a baby two bowls, one of biscuits, the other of raw broccoli, all the babies prefer the biscuits. But if the researcher indicates to the baby that she hates biscuits and loves broccoli, then hands the bowls to the baby and says: "Could you give me some?" something interesting happens. Fourteen-month-olds, with their innocent assumption that we all want the same thing, give her biscuits. But the wiser 18-month-olds give her broccoli, even though they themselves despise it. These tiny children have learned that other people's desires may conflict with their own.

This is also dramatically apparent in everyday life. Parents all know, and dread, the "terrible twos". While one-year-olds seem irresistibly drawn to forbidden objects (that lamp cord again), the two-year-olds seem deliberately bloody-minded. She doesn't even look at the lamp cord. Instead, her hand goes out to touch it as she looks, steadily, gravely, at you.

This demonic behaviour is quite rational, though. Our broccoli experiments show that children only begin to understand differences in desires at 18 months. The terrible twos seem to involve a systematic exploration of that idea, like an experimental research programme. Toddlers are testing the extent to which their desires and those of others may conflict. The grave look is directed at you because you and your reaction, rather than the lamp cord, are the interesting thing. The terrible twos reflect a clash between children's need to understand other people and their need to live happily with them. If the child is a budding scientist, we parents are the laboratory rats.

Two-year-olds also have to learn how visual perception works. Toddlers love hide and seek but aren't very good at it - a toddler will bury his head under the table with his bottom in view. In our lab, we explored when children learn how to hide things. Suppose I put a child on one side of a table and sit on the other. Then I put a screen and a toy on the table and ask the child to hide the toy from me. At 24 months, a toddler will put the toy on my side of the screen, so that it is actually hidden from them, but not from me. But 36-month-olds get this right. In fact, they'll often tell me they can see the toy but I can't. In the months in between, we observed many children experimenting. They would switch the toy from one side to the other, or come around to my side of the screen to make sure the toy really was hidden.

But this isn't the end of the story. Three-year-olds still have trouble with another important fact about people. They know that we can see different things, but not that what we think about the world may be wrong.

In a classic experiment, you can give three-year-olds a shut chocolate box. They open it and it turns out to have pencils inside. Then you ask them about another child in the nursery: "What will Nicky think is in the box: pencils or chocolates?" Three-year-olds report that Nicky will say there are pencils inside. They don't understand that Nicky will probably make the same mistake they made. Four-year-olds know that Nicky will be misled by the picture on the box.

Like scientists, children change their theories precisely because they make the wrong predictions. In "mistaken belief" experiments, simply telling children the right answer makes no difference. Like scientists, children at first resist counter-evidence. Virginia Slaughter and I visited three-year-olds over several weeks and gave them examples of mistaken beliefs: a golf ball that turned out to be soap, a yellow duck that looked green when put behind blue plastic. Each time the child made the wrong prediction, we presented them with counter-evidence. After two weeks these three-year-olds understood a brand new "mistaken belief" task, one they had never seen before, much better than a control group.

This experiment shows that even very young children are naturally able to alter their predictions in the light of new evidence. But it also shows how important other people can be: our adult behaviour had helped the children to work out the correct answer. Of course, we're developmental psychologists.

Do grown-ups naturally help children learn in their everyday lives? The new research suggests they do. One of the most dramatic examples of this is the sing-song voice adults use when they talk to babies. This speech style helps children sort out the sounds of language.

Similarly, the way that parents talk about the mind seems to influence their children's everyday psychology. What are the consequences of this new view? The research doesn't mean there is some set of flashcards that will help babies be brighter. Babies are already as bright as can be. They learn through everyday play, and through the care and attention of adults around them. It also doesn't mean there is some "critical period" for learning in the first three years or some set of experiences children must have. Children and even adults keep learning throughout life. It definitely doesn't mean mothers should quit their jobs. Anyone who cares for small children and is sensitive to what interests them can teach them what they need to know.

On the other hand, the research does suggest that the everyday, unrenumerated, unremarkable work of caring for babies and young children is extremely important. Humans have managed to learn so much because generations of adults put effort into caring for children.

Ironically, this new scientific perspective comes when young children and parents are under enormous pressure. We still penalise parents for taking time off work to be with their children, instead of rewarding them. Most parents face agonising dilemmas as they balance jobs and children. If we really want babies to learn, we should ditch the videotapes and flashcards and work for paid parental leave, flexible work arrangements and publicly supported, high-quality childcare.

The Guardian Weekly