Alison Gopnik on the Imagination
15:07 Share

This is Philosophy Bites with me, David Edmonds. And me, Nigel Warburton. Philosophy Bites is available at www.philosophybites.com. Philosophy Bites is made in association with the Institute of Philosophy. Kids like to live in a fantasy world. That much is obvious to any parent. But what's that got to do with philosophy? Alison Gopnik is a renowned psychologist at Berkeley.

She's intrigued by the role imagination plays in our lives, and it's linked to a concept much puzzled over by philosophers, causation. Alison Gopnik, welcome to Philosophy Bites. Well, glad to be here. The topic we're going to focus on today is the imagination. Now, why do we have an imagination at all?

Well, that's a good question. Plato thought that the poets should be exiled from the Republic because they were not only liars, but they were really bad liars who weren't very convincing for everyone. You could think of a sort of evolutionary version of Plato's objections, which is it's easy to see why understanding about the real world would give us all sorts of evolutionary benefits. But why would it ever be that understanding about things that aren't true and that you know aren't true would have evolutionary benefits?

And it's a particular puzzle for developmental psychologists because two and three-year-old children ubiquitously spend 24-7 off in these wild, crazy, pretend worlds. Yeah, for Plato, obviously, depictions of reality were at several removes of what he saw as the real world. Now, we don't have to accept Platonic metaphysics to think there's still a problem about how playful imaginary situations could possibly teach us about reality. Yeah.

It is still a problem and as I say it becomes more acute if you think about it psychologically and from an evolutionary perspective.

And I think we're actually starting to get to an answer. But the answer actually comes from thinking about a very different philosophical problem, and that's the problem of causation. How is it that we can tell the difference between a causal relationship and a mere correlational relationship? And that problem, another deep philosophical problem, may hold the answer to the problem about why we pretend. So just to get clear, what is the difference between cause and correlation?

Yeah, good question. Well, there's been lots of argument about this in philosophy over the years, but an idea that started with the philosopher David Lewis is that when we know that one thing causes another, we can not just make predictions about what will happen if one event takes place. We can also make counterfactuals. We can say not just that, say, if you smoke a lot, then lung cancer will follow, but that if you hadn't smoked, then you wouldn't get lung cancer.

Here's an example that's a nice illustration of this. Having yellow nicotine-stained fingers is going to be correlated with lung cancer, just like smoking is correlated with lung cancer. But we don't think that yellow fingers actually cause lung cancer. How can we cash out that intuition? Well, we don't think that, for instance, washing your fingers would make a difference to lung cancer. We don't think that

if you had had clean fingers you wouldn't have had lung cancer. But we do think that starting a smoking cessation program would lead to a difference in lung cancer or stopping smoking would change cancer. Now an idea that's come up in philosophy in the past 10 years or so by a philosopher named James Woodward is what's sometimes called the interventionist theory of causation. You might say, well alright, that's all very well but why would we want to know about counterfactuals? Why would we want to know about what would have happened? After all, all counterfactuals do is lead us to guilt and regret.

And what Woodward has argued is that if you think in terms of the future rather than the past, counterfactuals are really very helpful. So if I know that smoking causes lung cancer, then if I want to know what to do in the future to decrease the amount of lung cancer, I'll know, oh, what I should do is intervene on smoking. So what Woodward says is what it means for X to cause Y is that if you, to use a technical term, wiggle X, if you change X, then some change will take place in Y.

Though obviously with lung cancer, smoking isn't the only cause of that, so that wiggling eggs, reducing smoking, won't absolutely eliminate lung cancer. Well, that brings up another really important aspect of these new philosophical approaches to causation, the idea of probability. So the picture that comes out of a probabilistic view of causation is that it isn't that you're ever changing absolutely whether something happens or not. What you're always doing is messing about with probabilities.

And in general, in both philosophy and psychology and cognitive science, in the last 10 years, there's been a real revolution in our understanding of human cognition that depends on the idea that we're almost always thinking about probabilities, not about actual facts. We're often not just saying, for instance, is this true or false, but how likely is this hypothesis compared to another hypothesis? And this is actually going to turn out to be, I think, the key to the problem we started out with about imagination. So it's as if

Children with their imaginary games are proto-scientists in a way. They're exploring hypotheses which then could possibly be verified or refuted by experience.

Well, for some 20 years, developmental psychologists have been using the idea that what children are doing is very much like science, that they're trying, for instance, to figure out the causal structure of the world. One of the things we've discovered experimentally, for instance, is that two- and three- and four-year-olds are working out theories about how other people's minds work.

Three-year-olds, for instance, don't seem to understand the fact that you and I might have really different beliefs about the world and that that would lead to changes in our actions, but four-year-olds do. And it looks as if what happens is that the two- and three-year-olds go out and do experiments and explore the world and figure out how the world works and develop new causal theories, just like scientists do.

Children are very concerned about causation. Children are figuring out the causal structure of the world all the time. We can do experiments where we show children new patterns of evidence, and sure enough, right away, they make causal inferences. So children are pretending a lot, and children are also working out the causal structure of the world a lot. Now, maybe there's a connection between those two facts.

A lot of what children do is explore imaginary situations. Now, some elements of those mirror the real world and other bits don't. How do you determine which bits are the ones giving you the information as a child?

Well, one of the things that we found out empirically, and Paul Harris, in fact, who used to be at Oxford, has done this work, is that although children's imagination might look kind of wild and crazy, actually, internally, it has a lot of logic. And that's true, I think, about adult fiction and imagination too. The way that children's imagination seems to work is that

They say, here's a premise that is going to seem unlikely or unusual. A boy turns out to actually be in wizard school, let's say. So you start out from this extremely unlikely, low probability hypothesis, or for a two-year-old, this hypothesis.

stuffed teddy bear is really a real person and a real friend of mine. But then the logic of the inferences that you make from that premise is perfectly good causal logic. So you say, well, if this were true, if Harry Potter were a wizard and going to wizard school, well, he would have to get clothes and he would have to get books and he'd have to do all the things that you know that children do. And Paul Harris showed that even with two-year-olds, if you say something to them like, all right, here's Teddy and we're going to pretend that

Teddy's having tea and he spills the tea all over him. Or else we'll pretend that Teddy had talcum powder and he spilled the talcum powder all over him. Well, in the pretend, will Teddy be wet or will Teddy be dry? And if you say, well, we're pretending that he's having tea, then they say, well, he'll be wet.

And if you're pretending that he's spilled talcum powder, well, then he'll be covered with powder and be powdery. And little children can be quite fierce and specific about the fact that even if you're starting out from a crazy premise, that doesn't mean that you're allowed to have any old consequence. You have to work through those premises in a logical way.

That's children. I can understand how children need to discover things about how the world works around them, have hypotheses confirmed about other people's minds and so on. But as adults, do we still need the imagination in the same way? Well, I think we don't need the imagination in the same way the children do, and we don't indulge in imagination as much as children do. But when

But when we're in the same position that children are in, trying to find out new things about the world or create new ways of thinking about the world, then that's exactly when imagination does kick in. So you could think about something like drama or fiction as serving the function of telling us about what the consequences would be of what look like very unlikely probabilities to begin with.

But of course, the thing that's interesting about human beings from an evolutionary point of view is that we're always exploring very, very low probability hypotheses. And in fact, what we do is we take things that start out being very unlikely and we actually make them real.

So if you look around this room that we're in right now and think about it from the perspective of a hunter-gatherer in the Pleistocene, everything in this room is imaginary. Not just the computers and the light bulbs, but the woven fabric and the right-angle construction of the tables. From the perspective of the Pleistocene, all those things started out as a very, very weird, strange idea in some guy's head about, gee, what would happen if we did this?

The G, what would happen if we did this, is really the central thing that causal cognition and causal thinking allows you to do. The process of imagining something and bringing it into reality assumes that the hypothesis or the imaginary situation has got something in it which makes it possible to bring into reality. But lots of imaginary situations will be wasted, as it were. They will be explorations which fall by the wayside because they're just unrealizable in reality.

Well, I think the difference again between children and adults is that the children are in a situation where they don't actually have to do anything. And one of the great interesting evolutionary mysteries about human beings is why is it that we have this extremely extended period of childhood, this very long immaturity, when to be blunt, the children are basically completely useless. They don't bring home the bacon. They don't do anything useful. And in fact, they're sort of worse than useless because we have to put in so much time and energy taking care of them.

And the answer seems to be that there's a kind of evolutionary division of labor where we have this early period when we can simply explore alternatives, learn as much as we can about the world around us, create and imagine new possibilities without being under the constraints of actually getting on in the world, getting things done, doing the things we need for our survival. And then what happens is that as adults, we can take all of that exploration and learning that we did as children and put things

the products of that to use to actually make things happen out in the world. So you could think of it as evolution has designed children to be the research and development division of the human species. And we adults are production and marketing. Now I can imagine an artist or a poet listening to you talking and saying, no, that's not what the imagination is about. You're just interested in payoffs. The

Thinking about children is actually really helpful in this regard because from the grand evolutionary perspective, childhood is absolutely essential. But of course, from the immediate perspective, children aren't playing or exploring because they're going to be solving this evolutionary project.

the interesting sort of paradox is that for human beings especially, one of the things that enables us to achieve our evolutionary goals in the long run is being able to pull ourselves away from achieving goals in the short run. So having this protected peer

having a period of play, having a period when you aren't actually trying to accomplish goals, apparently paradoxically turns out to be one of the very best ways of being able to accomplish your goals in the long run. So clearly every executive board ought to have a child, at least one child there playing in the background.

Well, I think it's no coincidence that my friends in Silicon Valley have beanbag chairs and balls to play in. And if you go to someplace like the Googleplex, there's a lot of opportunities for play. If your goal is innovation, that childlike sense of exploration is the best strategy to be able to produce innovation in the long run.

What we've been talking about is really a mixture of philosophy and psychology. Some purists will say, look, psychology is really interesting, but it doesn't have anything to do with philosophy. Well, it's funny because at the very beginning, if you think about the great philosophers who we all admire, like my personal favorite David Hume or Descartes or Locke or Plato and Aristotle, for that matter, they didn't think that what they were doing was this special thing called philosophy that didn't have anything to do with what was going on in society.

the sciences. In fact, someone like Hume very explicitly thought that what he was doing was applying scientific methods to the study of the mind. So I think it's only really quite recently that there's been this idea that there's a sharp divide between philosophy and psychology. And my own view is that you should think about philosophy as really sort of very, very theoretical everything. So if you're doing very theoretical art, then you're doing aesthetics. And if you're doing very theoretical law, then you're doing ethics.

And I think that if you're doing very theoretical psychology, you're doing what's often called philosophy of mind. Or if you're doing very theoretical developmental psychology, you're doing epistemology. I think...

I think we're all trying to answer the same deep profound questions, things like how is it that we come to understand the world around us? And the philosopher Quine had a nice image about the giant web of belief. Some parts, the scientific parts, are closer to the edges where we make contact with the world and some parts, the philosophical or mathematical or theoretical parts, are closer to the center of the web.

But it's all one web, and finding out something new about the world can change even very abstract ideas we have. And vice versa, thinking very abstractly about things like counterfactuals can tell us about something very concrete, like why on earth is it that three-year-olds have imaginary friends? And I think it's actually been very exciting, especially in developmental psychology, to see how work that to be frankfurt,

I think most philosophers would have turned their noses up at questions like, why do two-year-olds think that teddy bears will get wet if you pour tea on them? Turned out to actually be extremely revealing about deep philosophical questions about things like fiction and imagination and causation and vice versa. Alison Gopnik, thank you very much. Thank you so much for having me. There's now a Philosophy Bites book published by Oxford University Press.

For more information, go to www.philosophybytes.com. For more information about the Institute, go to www.philosophy.sas.ac.uk.