2017 : WHAT SCIENTIFIC TERM OR CONCEPT OUGHT TO BE MORE WIDELY KNOWN? [1]

gary_klein's picture [5]
Senior Scientist, MacroCognition LLC; Author, Seeing What Others Don't: The Remarkable Ways We Gain Insights
Decentering

You may have worked so closely with a partner that you reached a point where each of you could finish the other’s sentences. You had a pretty good idea of how your partner would respond to an event. What’s behind such a skill?  

Decentering is the activity of taking the perspective of another entity. When we look into the past, we try to explain why that entity behaved in a way that might have surprised us. Peering into the future, we decenter in order to anticipate what that entity is likely to do.

Skilled decentering comes into play, not just with partners, but also with strangers and even with adversaries. It gives us an edge in combat, to get ready to ward off an adversary’s attack. It helps authors write more clearly as they anticipate what might confuse a reader or what a reader expects to find out next. Police officers who are good at decentering can de-escalate a tense situation by moving an encounter into less volatile directions. Teachers rely on decentering to keep the attention of students by posing questions or creating suspense, anticipating what will catch the attention of their students. Team members who can quickly decenter can predict how each will react to unexpected changes in conditions, increasing their coordination.

Decentering is not about empathy—intuiting how others might be feeling. Rather, it is about intuiting what others are thinking. It is about imagining what is going through another person’s mind. It is about getting inside someone else’s head.

Despite its importance, particularly for social interaction, decentering has received very little attention. Military researchers have struggled to find ways to study decentering and to train commanders to take the perspective of an adversary.  Social psychologists have not made much progress in unpacking decentering. Part of the problem might be that researchers examine the average decentering accuracy of observers whereas they should be investigating those observers whose accuracy is consistently above average. What are their secrets? What is behind their expertise?

It could be even more valuable to study decentering outside the laboratory—in natural settings. I suspect that some people don’t even try to decenter, others may try but aren’t particularly effective, and still others may be endowed with an uncanny ability to take another person’s perspective.

Decentering also comes into play when we interact with inanimate objects such as intelligent technologies that are intended to help us make decisions. That’s why the definition at the beginning of this essay refers to entities rather than people. The human factors psychologist Earl Wiener once described the three typical questions people ask when interacting with information technology: What is it doing? Why is it doing that? What will it do next? We are trying to take the perspective of the decision aid.

It would be nice if the decision aid could help us decenter, the way a teammate could. After all, when we interact with a person, it is natural to ask that person to explain his or her choice. But intelligent systems struggle to explain their reasons. Sometimes they just recite all the factors that went into the choice, which isn’t particularly helpful. We want to hear less, not more. We want the minimum necessary information. And the minimum necessary information depends on decentering.

For example, if you are using a GPS system to drive to a location, and the device tells you to turn left whereas you expected to turn right, you would NOT want the system to explain itself by showing the logic it followed. Rather, the system should understand why you expected a right-hand turn at that juncture and should provide the central data element (e.g., an accident up ahead is creating thirty-minute delays on that route) you need to understand the choice.

If you were driving with a human navigator who unexpectedly advised you to turn left, all you would have to do is say, “Left?” Most navigators instantly determine from your intonation that you want an explanation, not a louder repetition. Good navigators grasp that you, the driver, didn’t expect a left-hand turn and would quickly (remember, there is traffic getting ready to honk at you) explain, “Heavy traffic ahead.” Your navigator would convey the minimum necessary information, the gist of the explanation. But the gist depends on your own beliefs and expectations—it depends on the navigator’s ability to decenter and get inside your head. Smart technology will never be really smart until it can decenter and anticipate. That’s when it will become a truly smart partner.

When we are uncomfortable with the recommendations offered by a decision aid we don’t have any easy ways to enter into a dialog. In contrast, when we dislike a person’s suggestions we can examine his or her reasons. Being able to take someone else’s perspective lets people disagree without escalating into conflicts. It allows people to be more decent in their interactions with each other. If they can decenter, then they can become decenter.