Skip to navigationSkip to content

Ideas

Our home for bold arguments and big thinkers.

wedding rings
AP Photo/Karen Pulfer Focht
Opposite personalities both attract and repel.
OPPOSITES ATTRACT

The fiery partnership between two great psychologists can help explain why some relationships fall apart

Daniel F. Stone
By Daniel F. Stone

Assistant Professor of Economics, Bowdoin College

They say that opposites attract. But we all know opposites can also repel. They both complement and combust—and this tension is at the center of Michael Lewis’s new book, The Undoing Project.

The Undoing Project offers a keen-eyed look at the complex partnership between the great psychologists Daniel Kahneman and Amos Tversky, whose work on systematic cognitive biases in the 1970s laid the foundation for behavioral economics. They were each at their best when they were collaborating with one another, eventually winning Nobel and MacArthur genius prizes. But ultimately, their relationship unraveled. “It was worse than a divorce,” Tversky’s wife, Barbara, says in the book.

Lewis’s book is meticulously researched. Yet it fails to note an important—though admittedly speculative—irony. The falling-out between Kahneman and Tversky may have been driven by the very cognitive biases they studied. A look at their story can help us gain a better grasp of how behavioral economics may inform our understanding of inter-personal relationships.

The problems between Kahneman and Tversky came down to two big issues. First, Tversky got the lion’s share of the public credit for their work, and Kahneman wasn’t happy about this. Tversky either was not fully aware of Kahneman’s need for more recognition (according to Lewis, “If Amos couldn’t give Danny what he needed, it was perhaps because he couldn’t imagine having the need”) or rejected this need (“at bottom, Amos saw Danny’s need as weakness”).

The second, and likely related, problem was that Kahneman felt that Tversky had become more critical of Kahneman’s ideas over time. “We have got to the point that the very thought of telling you of ANY idea that I like … makes me anxious,” Kahneman wrote in a letter to Tversky.  Tversky replied, “You have become very protective of some ideas.”

In a nutshell: Kahneman thought Tversky should have been more supportive, both publicly and privately. Tversky thought Kahneman was being too sensitive. This difference in perspectives led to tension and anger. It seems reasonable, even obvious, to think this anger was due partly to each man not fully understanding the other’s point of view.  And behavioral economics teaches us that misunderstanding is often driven by cognitive bias.

“People are not so complicated,” Tversky says. “Relationships between people are complicated.”

One key bias studied by the two men was that of “availability”—when we form a judgment that relies heavily on whatever information is more available, or comes most easily to mind. Almost by definition, our own point of view about a relationship is highly available (right there in our own brains). The other person’s view is less available. If they’re not telling us how we feel, we have to ask them, and they may not be entirely forthright with us anyway.

The other important bias that can shed light on their relationship is that of overconfidence, or what Kahneman and Tversky called “the illusion of validity.”

Availability plus overconfidence can be a recipe for trouble in social relationships. Well-intentioned people often do things that look, at least on the surface, not so well-intentioned. These “bad actions” are highly “available,” in that we can easily mull over our own interpretations of them. And overconfidence makes us too sure that these apparently bad actions are indeed “bad.” A downward spiral can easily ensue.

It may seem obvious that relationship problems are driven by such misunderstandings, which in turn are driven by cognitive biases. But in fact, the direct connection of such cognitive biases to anger and relationship problems is not well studied—which is perhaps why Lewis does not make this connection in the book.

Cognitive biases can help us gain a more nuanced understanding of why people don’t get along in the first place.

For a long time, researchers assumed that there was a divide between the emotional and cognitive aspects of psychology. So it would have been odd to consider cognitive biases apply to emotional processes. But a body of research now argues that the cognitive-emotional dichotomy is a false one. This seems to be the new growing agreement across disciplines including psychology, neuroscience, and even philosophy.

Cognitive biases can also help us gain a more nuanced understanding of why people don’t get along in the first place. One might think that Kahneman and Tversky in their later years were simply too different from one another: Tversky was loud, Kahneman quiet; Tversky bold, Kahneman cautious. But why do we dislike people who are different from ourselves? This may often be due to misinterpreting differences of opinion or actions as signals of bad character or motives—with such misperception also deriving from cognitive bias.

The claim that such biases affect relationships should, of course, be further studied. But the example offered by Tversky and Kahneman shows how cognitive biases may cause relationship problems and inhibit communication more generally—not just between friends, but also between co-workers, neighbors, and spouses. Perhaps the US has a 50% divorce rate not only because some people choose the wrong person to marry, but also because of the fallout during the marriage from cognitive bias.

Even more importantly, cognitive biases can also cause misperceptions and conflict between larger groups, such as political parties, ethnic groups, or nations. Obviously, the societal impact of such clashes can be immense. But research into the relationship between cognitive biases and these kinds of conflicts has thus far focused on why we are motivated to hold biases: it feels good to think my tribe is “good” and yours is “bad.” Incorporating “unmotivated” cognitive biases to the analysis of these conflicts may further enhance our understanding of tensions between different groups, and even help yield solutions.

“People are not so complicated,” Tversky says in The Undoing Project. “Relationships between people are complicated.” Neither Kahneman nor Tversky actually studied relationships—but perhaps their work, and the legacy of their own tumultuous partnership, can help spur behavioral economists to take up the mantle.