In The Knowledge Illusion: Why We Never Think Alone, Steven Sloman and Philip Fernbach examine some of the pressing social problems of our time — including polarization and the difficulties it raises for hammering out workable public policies.

More than that, they apply their expertise as cognitive scientists to these problems, with arresting and meaningful results. Sloman, who was interviewed for this article, teaches at Brown University. Fernbach is on the faculty of the University of Colorado’s business school.

You’re not mere political pundits. You actually conducted experiments to arrive at your conclusions, right?

That’s right, plus we built on research others have done in this field. For example, in an academic setting, we conducted experiments in which we looked at how confident people are in their opinions. Then, simplifying our methodology, obviously, we asked them to explain some of the issues about which they had a great deal of confidence — some issue like cap and trade, for example — and how the policies they believe in so fervently actually worked.

What did you determine?

We found that people are very rarely able to explain the rationale behind the opinion they hold. They can offer reasons but almost never justifications. By “reasons,” I mean they can say the policy they support conforms to their values, or that someone else they trust holds this position. But they can’t explain how the policy works or what actual consequences it might have. And when they realize they are unable to offer a justification for an opinion, their confidence in that opinion diminishes.

Which you think is a healthy outcome?

Yes, it is, because as we argue in the book, we don’t really know as much as we think we know. We have strong opinions, but they are very often about how some issue matches values we care about. But very rarely are these opinions based on causal explanations that, in a sense, sit outside ourselves. Much of the time, people’s opinions have very little to do with an understanding of a policy and its specific outcomes.

People have a lot of attitudes about things, but these attitudes turn out to be a house of cards. They are a construction formed out of a community with no underlying justification.

So what’s going on here?

People often simply channel the opinions of the communities they live in, and we think this probably has a lot to do with the way we live in bubbles. Our sources of media today are individualized, so we often hear only views that reflect our own — and that of the bubble we live in. Our world is no longer one in which Walter Cronkite delivers “the news” in a way that is common to us all, with facts we accept and agree on. When we get out of the bubble, we realize how complicated things are and how little we, as individuals, really know. We operate with an illusion of knowledge — that knowledge exists in our own brains when, in fact, it is a social thing. Knowledge is shared, and different people know different things.

Can you explain?

Yes, but this is hardly a new idea. Take zippers, for example. We all think we understand how something like a zipper works, right? But when pressed to explain it, most of us can’t because we don’t understand it. It turns out that the mechanism of zippers is really pretty complex. Social policy is like that, too. People have a lot of attitudes about things, but these attitudes turn out to be a house of cards. They are a construction formed out of a community with no underlying justification. I believe such-and-such a thing because other people who seem to think like I do believe it, and it goes no deeper than that. The whole house of cards collapses.

How does this fit in with ideas of a “post-truth” world?

Some things you can observe more or less directly, and there’s very little dispute about them. Take the issue of whose inaugural crowds were larger — Trump’s or Obama’s. You can look at the photographs, and it seems pretty obvious. But then questions can be raised about what time of day the pictures were taken or at what angle, and all that. Still, it seems like a pretty settled question. This is the kind of thing we can check, right, and verify? But when you get into more complicated questions — even about whether Obama “wiretapped” Trump Tower or not —you realize how unbelievably complicated things can be, when you have no direct knowledge of them and are dealing only in inference.

But disagreement over “facts” is nothing new, is it?

Of course not. Politicians have been manipulating information for their own purposes forever. I have little doubt that the Carthaginians were not given an accurate characterization of what the Roman enemy was up to way back when. But today we have access to many more “facts,” and when we live in bubbles, all this becomes even more complicated. And, whether this is intentional or not, our educational system encourages this tendency to have more confidence in what we think we know and understand.

In what way?

The American educational approach is to raise people to be “independent thinkers,” and it prizes their independence as the highest form of intelligence. The goal is to encourage this idea, when it is done so at the expense of a different understanding of knowledge and intelligence. Great educational theorists like John Dewey stressed independent thought but also understood the importance of collaboration, and how real learning and progress are accomplished in a social rather than individual setting. This emphasis on independent thinking, of course, fits very well with American ideas of individualism. But it misses something important.

What is that?

It misses the possibility that the highest form of intelligence might actually lie with the person who can operate most effectively in a collaborative environment, where information is shared.

Reach Sloman at 401.588.4685 or

Additional Resources

Read more author interviews in the Impact archives.