Skip to main content
Macmillan Childrens Publishing Group

The Philosophical Baby

What Children's Minds Tell Us About Truth, Love, and the Meaning of Life

Alison Gopnik




Possible Worlds



Human beings don't live in the real world. The real world is what actually happened in the past, is happening now, and will happen in the future. But we don't just live in this single world. Instead, we live in a universe of many possible worlds, all the ways the world could be in the future and also all the ways the world could have been in the past, or might be in the present. These possible worlds are what we call dreams and plans, fictions and hypotheses. They are the products of hope and imagination. Philosophers, more drily, call them "counterfactuals."

Counterfactuals are the woulda-coulda-shouldas of life, all the things that might happen in the future, but haven't yet, or that could have happened in the past, but didn't quite. Human beings care deeply about those possible worlds—as deeply as they care about the real actual world. On the surface counterfactual thinking seems like a very sophisticated and philosophically puzzling ability. How can we think about things that aren't there? And why should we think this way instead of restricting ourselves to the actual world? It seems obvious that understanding the real world would give us an evolutionary edge, but what good do we get from imaginary worlds?

We can start to answer these questions by looking at young children. Is counterfactual thought present only in sophisticated grown-ups? Or can young children think about possibilities too? The conventional wisdom, echoed in the theories of both Sig-mund Freud and Jean Piaget, is that babies and young children are limited to the here and now—their immediate sensations and perceptions and experience. Even when young children pretend or imagine they can't distinguish between reality and fantasy: their fantasies, in this view, are just another kind of immediate experience. Counterfactual thought requires a more demanding ability to understand the relation between reality and all the alternatives to that reality.

Cognitive scientists have discovered that this conventional picture is wrong. We've found out that even very young children can already consider possibilities, distinguish them from reality, and even use them to change the world. They can imagine different ways the world might be in the future and use them to create plans. They can imagine different ways the world might have been in the past, and reflect on past possibilities. And, most dramatically, they can create completely imaginary worlds, wild fictions, and striking pretenses. These crazy imaginary worlds are a familiar part of childhood—every parent of a three-year-old has exclaimed, "What an imagination!" But the new research profoundly changes the way we think about those worlds.

In the past ten years we've not only discovered that children have these imaginative powers—we've actually begun to understand how these powers are possible. We are developing a science of the imagination. How could children's minds and brains be constructed to allow them to imagine this dazzling array of alternate universes?

The answer is surprising. Conventional wisdom suggests that knowledge and imagination, science and fantasy, are deeply different from one another—even opposites. But the new ideas I'll outline show that exactly the same abilities that let children learn so much about the world also allow them to change the world—to bring new worlds into existence—and to imagine alternative worlds that may never exist at all. Children's brains create causal theories of the world, maps of how the world works. And these theories allow children to envisage new possibilities, and to imagine and pretend that the world is different.


Psychologists have found that counterfactual thinking is absolutely pervasive in our everyday life and deeply affects our judgments, our decisions, and our emotions. You would think that what really matters is what actually happens, not what you imagine might have happened in the past or could happen in the future. This is particularly true of counterfactuals about the past—what might have happened but didn't—the woulda-coulda-shouldas of life. Yet the woulda-coulda-shouldas have a deep impact on experience.

In one experiment, the Nobel Prize–winning psychologist Daniel Kahneman and his colleagues asked people to imagine the following sort of scenario. Mr. Tees and Mr. Crane are both in a taxi to the airport, desperate to catch their respective planes, which are both scheduled to take off at 6:00. But traffic is impossibly snarled and the minutes tick by. Finally, at 6:30 they arrive at the airport. It turns out that Mr. Tees's flight left at 6:00 as planned but Mr. Crane's flight was delayed till 6:25 and Mr. Crane sees it take off as he arrives. Who is more upset?

Just about everyone agrees that Mr. Crane, who just missed his flight, will be much more unhappy. But why? They both missed their flights. It seems that what is making Mr. Crane unhappy is not the actual world but the counterfactual worlds, the ones in which the taxi arrived just that much earlier or the plane was delayed just a few minutes more.

You needn't turn to artificial scenarios like this one to see the effects of counterfactuals. Consider the medalists in the Olympics. Who is happier, the bronze medalist or the silver? You'd think that objectively the silver medalist, who, after all, has actually done better, would be happier. But the relevant counterfactuals are very different for the two. For the bronze medalist the relevant alternative was to finish out of the medals altogether—a fate she has just escaped. For the silver medalist, the relevant alternative was to get the gold medal—a fate she has just missed. And, in fact, when psychologists took clips of the medals ceremonies and analyzed the facial expressions of the athletes, it turned out that the bronze medalists really do look happier than the silver medalists. The difference in what might have been outweighs the difference in what is.

Like Mr. Crane at the airport, or the silver medalist, people are most unhappy when a desirable outcome seems to be just out of reach, or to have just been missed. As Neil Young adapted John Greenleaf Whittier: "The saddest words of tongue and pen are these four words, ‘it might have been.'"

Why do we humans worry so much about counterfactuals, when, by definition, they are things that didn't actually happen? Why are these imaginary worlds just as important to us as the real ones? Surely "it is, and it's awful" should be sadder words than "it might have been."

The evolutionary answer is that counterfactuals let us change the future. Because we can consider alternative ways the world might be, we can actually act on the world and intervene to turn it into one or the other of these possibilities. Whenever we act, even in a small way, we are changing the course of history, nudging the world down one path rather than another. Of course, making one possibility come true means that all the other alternative possibilities we considered won't come true—they become counterfactuals. But being able to think about those possibilities is crucial to our evolutionary success. Counterfactual thinking lets us make new plans, invent new tools, and create new environments. Human beings are constantly imagining what would happen if they cracked nuts or wove baskets or made political decisions in a new way, and the sum total of all those visions is a different world.

Counterfactuals about the past, and the characteristically human emotions that go with them, seem to be the price we pay for counterfactuals about the future. Because we are responsible for the future, we can feel guilty about the past; because we can hope, we can also regret; because we can make plans, we can be disappointed. The other side of being able to consider all the possible futures, all the things that could go differently, is that you can't escape considering all the possible pasts, all the things that could have gone differently.


Can children think counterfactually? The most evolutionarily fundamental kind of counterfactual thinking comes when we make plans for the future—when we consider alternative possibilities and pick the one we think will be most desirable. How can we tell if a very young baby can do this? In my lab, we showed the baby the sort of post with stacking rings that is a standard baby toy. But I had taped over the hole in one of the rings. How would the baby respond to this apparently similar but actually recalcitrant ring? When we brought a fifteen-month-old into the lab he would use a kind of trial-and-error method to solve the problem. He would stack some of the rings, look carefully at the taped-over one—and then try it on the post. And then try it on the post again, harder. And try it on the post one more time. Then he would look up puzzled, try one of the other rings again—and then again try the taped-over one. Basically, young babies would keep at this until they gave up.

But as they got older and learned more about how the world worked, babies would behave entirely differently. An eighteen-month-old would stack all the other rings and then hold up the trick ring with a "Who do you think you're kidding?" look and refuse even to try it. Or she would immediately pick the trick ring up and dramatically throw it across the room, and then calmly stack the rest. Or, equally dramatically, she would hold it up to the post and shout "No!" or "Uh-oh!" These babies didn't have to actually see what the ring would do—they could imagine what would happen if you put it on the post, and act accordingly.

In another experiment we saw whether babies could discover a new use for an object—if they could, in a simple way, invent a new tool. I put a desirable toy out of the babies' reach and placed a toy rake beside it. As with the ring, fifteen-month-olds sometimes did pick up the rake, but they couldn't figure out how to use it as a tool. They pushed the toy from side to side or even, frustrat-ingly, farther away from them, till they either accidentally got it or gave up. But older babies looked at the rake and paused thoughtfully. You could almost see the wheels spinning. Then they produced a triumphant smile and often a certain look of smugness. You could almost see the lightbulb switching on. Then they put the rake in just the right position over the toy and triumphantly used it to bring the toy toward them. Again they seemed able to mentally anticipate—to imagine—all the possible ways the rake could affect the toy and then chose just the right possibility.

Simple trial and error, trying different actions until one succeeds, is actually often a very effective way of getting along in the world. But anticipating future possibilities lets us plan in this other more insightful way—using our heads instead of our hands. The older babies seemed to be anticipating the possible future in which the ring or the rake would fail and avoiding that future. Other studies have shown that this isn't just a difference between fifteen- and eighteen-month-olds. Even younger babies can solve problems insightfully if they have the right kinds of information.

This ability to solve problems insightfully seems to be particularly human. There is a little evidence that chimpanzees, and even some very smart birds like crows, can do this occasionally. But even chimpanzees and crows, and certainly other animals, overwhelmingly rely on either instinct or trial and error to get along in the world. And, in fact, instinct and trial and error are often very effective and intelligent strategies. It is extremely impressive to see a bird putting together the complex set of instinctive behaviors that allows it to build a nest, or a chimpanzee using trial and error to gradually zero in on the right strategy to open a box with elaborate locks. But they are different from the strategies that babies and very young children use. Anthropologists agree that using tools and making plans, both abilities that depend on anticipating future possibilities, played a large role in the evolutionary success of Homo sapiens. And we can see these abilities emerging even in babies who can't talk yet.


In these experiments babies seem to be able to imagine alternative possibilities in the future. Can children also imagine past counter-factuals, different ways the world might have been? We have to infer babies' counterfactual thinking from what they do, but we can explicitly ask older children counterfactual woulda-coulda-shoulda questions. Until recently psychologists claimed that children were quite bad at thinking about possibilities. Children are indeed quite bad at producing counterfactuals about subjects they know little about, but when they understand the subject matter even two- and three-year-olds turn out to be adept at generating alternative worlds.

The English psychologist Paul Harris probably knows more than anyone about young children's imaginative abilities. Harris is tall, thin, reserved, and very English, and worked for many years at Oxford University. His work, like the work of the great Oxford writer Lewis Carroll, is a peculiarly English combination of the strictest logic applied to the wildest fantasy.

Harris told children a familiar English countryside story. Then he asked them about future and past counterfactuals. Naughty Ducky is wearing muddy boots and is about to walk into the kitchen. "What would happen to the floor if Ducky walked through the kitchen? Would it be clean or dirty?" "What would have happened to the floor if Ducky had cleaned his boots first? Would it be clean or dirty?" Even young three-year-olds say that the floor would have been spared if only Ducky had cleaned his boots.

In my lab, David Sobel and I designed a set of storytelling cards—cartoon pictures that told the right story if you put them in order. We showed children a sequence of pictures, say a girl going to a cookie jar, opening the jar, looking inside, finding cookies, and looking happy. But we also had a set of several other pictures, including the girl finding that there were no cookies, and the girl looking sad and hungry. We showed the children the cards in the right sequence and asked them to tell the story. Then we said, "But how about if the girl had been sad at the end instead?" and changed the last card, so that the girl looked sad instead of happy. "What would have had to happen then?" Three-year-olds consistently changed the earlier pictures to fit the hypothetical ending—they replaced the picture of the full cookie jar with the picture of the empty one. These very young children could imagine and reason about an alternative past.


We can also see evidence for counterfactual thinking in children's play. Babies start pretending when they are as young as eighteen months old or even younger. Pretending involves a kind of present counterfactual thinking—imagining the way things might be different. Even babies who can't talk yet, and are barely walking, can still pretend. A one-and-a-half-year-old baby may fastidiously comb her hair with a pencil, or rest her head on a pillow dramatically pretending to be asleep, giggling all the while. A little later babies start to treat objects as if they were something else. Toddlers turn everything from blocks to shoes to bowls of cereal into means of transportation by the simple expedient of saying "brrm-brrm" and pushing them along the floor. Or they may carefully, tenderly, put three little toy sheep to bed.

We take this for granted when we choose toys for these young children. The toddler sections of toy stores are full of toys that encourage children to pretend: the farmhouse, the gas station, the zoo—even the toy ATM and cell phone. But it's not that two-year-olds pretend because we give them dolls; instead we give them dolls because they love to pretend. Even without toys toddlers are just as likely to turn common objects—food, pebbles, grass, you, themselves—into something else. And even in cultures where pretend play is discouraged, rather than cultivated, like Mr. Grad-grind's school in Dickens's Hard Times, children continue to do it anyway. ("No child left behind" testing policies seem to be echoing Mr. Gradgrind, replacing dress-up corners and pretend play with reading drills in preschools.)

As soon as babies can talk they immediately talk about the possible as well as the real. As a graduate student at Oxford I recorded all the words that nine babies used when they first began to talk. These babies, who were still just using single words, at the very start of language, would use them to talk about possibilities as well as actualities. There was not only the ubiquitous "brrm-brrm," but "apple" when pretending to eat a ball, or "night-night" when putting a doll to bed. One particularly charming red-haired toddler had a beloved teddy bear, and his mother had knitted two long scarves, like the ones Dr. Who wears in the British TV series, a small one for the bear and a larger one for Jonathan. Jonathan one day put his teddy bear's scarf around his neck and, with enormous grins and giggles, announced his new identity: "Jonathan Bear!"

Excerpted from The Philosophical Baby by Alison Gopnik.
Copyright 2009 by Alison Gopnik.
Published in First edition, 2009 by Farrar, Straus and Giroux.
All rights reserved. This work is protected under copyright laws and reproduction is strictly prohibited. Permission to reproduce the material in any manner or medium must be secured from the Publisher.