These propositions had always left some psychologists uneasy. But since Blink was written, there has been growing reaction in the psychology community against Ekman’s position.
For example, why did Ekman believe that emotions were universal? In the 1960s, he and two colleagues traveled to Papua New Guinea, armed with a stack of thirty photographs. The pictures were headshots of Westerners making facial expressions corresponding to the basic emotions: anger, sadness, contempt, disgust, surprise, happiness, and fear.
The New Guinea tribe that Ekman’s group visited was called the Fore. As recently as a dozen years earlier, they had still been effectively living in the Stone Age, completely cut off from the rest of the world. Ekman’s idea was that if the Fore could identify anger or surprise in the photographed faces as readily as someone in New York City or London can, emotions must be universal. Sure enough, they could.
“Our findings support Darwin’s suggestion that facial expressions of emotion are similar among humans, regardless of culture, because of their evolutionary origin,” Ekman and his colleagues wrote in a paper published in Science, one of the most prestigious academic journals. (See P. Ekman et al., “Pan-Cultural Elements in Facial Display of Emotions,” Science 164 [1969]: 86–88.)
This idea—that there is a universal set of human emotional reactions—is the principle that lies behind an entire category of tools that we use to understand strangers. It’s why we have lie detectors. It’s why lovestruck couples stare deeply into each other’s eyes. It’s why Neville Chamberlain made his daring visit to see Hitler in Germany. And it’s why Solomon looked hard at the defendant in the child-abuse case.
But there’s the problem. Ekman was leaning awfully hard on what he saw with the Fore. Yet the emotion-recognition exercise he did with them wasn’t nearly as conclusive as he said it was.
Ekman went to New Guinea with another psychologist, Wallace Friesen, and an anthropologist, Richard Sorenson. Neither Ekman nor Friesen spoke the language of the Fore. Sorenson knew only enough to understand or say the simplest things. (See James Russell, “Is There Universal Recognition of Emotion from Facial Expression? A Review of the Cross Cultural Studies,” Psychological Bulletin 115, no. 1 [1994]: 124.) So there they are, showing headshots to tribesmen of white people making faces—and they are utterly reliant on their translator. They can’t just have each tribesman free-associate about what he thinks is happening in each photo. How would they make sense of that? They have to keep things simple. So Ekman and his group use what’s called “forced choice.” They show each Fore person the pictures, one by one, and for every image they asked the viewer to choose the right answer from a short list of emotions. Is what you are looking at anger, sadness, contempt, disgust, surprise, happiness, or fear? (The Fore didn’t really have a word to describe disgust or surprise, so the three researchers improvised: disgust was something that stinks; surprise was something new.)
Now, is forced choice a good method? For example, suppose I want to find out whether you know which city is the capital of Canada. (A surprising number of Americans, in my experience, have no idea.) I could ask you straight out: What is the capital of Canada? That’s a free choice question. In order to answer it correctly, you really have to know the capital of Canada. Now here’s the forced-choice version of that question.
The capital of Canada is:
Washington, DC
Kuala Lumpur
Ottawa
Nairobi
Toronto
You can guess, can’t you? It’s not Washington, DC. Even someone with no knowledge whatsoever of geography probably knows that’s the capital of the United States. It’s probably not Kuala Lumpur or Nairobi, since those names don’t sound Canadian. So it’s down to Toronto or Ottawa. Even if you have no idea what the capital of Canada is, you have a 50 percent chance of getting the answer right. So is that what was happening with Ekman’s survey of the Fore?
Sergio Jarillo and Carlos Crivelli—the two researchers I write about in Chapter Six of this book—began their research by attempting to replicate Ekman’s findings. Their idea was: let’s correct the flaws in his exercise and see if it still holds up. Their first step was to pick a isolated tribe—the Trobriand Islanders—whose language and culture at least one of them (Jarillo) knew well. That was their first advantage over Ekman: they knew an awful lot more about whom they were talking to than Ekman’s group had. They also decided not to use “forced choice.” They would use the far more rigorous methodology of free choice. They laid out a set of headshots (with people looking happy, sad, angry, scared, and disgusted) and asked, “Which of these is the sad face?” Then they asked the next person, “Which of these is the angry face?” And so on. Finally, they tallied all the responses.
And what did they find? That when you redo Ekman’s foundational experiment—only this time, carefully and rigorously—the case for universalism disappears. Over the past few years the floodgates have opened, which is where much of the research I described in this chapter comes from.
A few additional points:
Ekman’s original Science paper is, upon reflection, a little strange. He argued that what he found in the Fore was evidence of universalism. But if you examine his data, it doesn’t look like he’s describing universalism.
The Fore were really good at correctly identifying happy faces, but only about half of them correctly identified the “fear” face as being an expression of fear. Forty-five percent of them thought the surprised face was a fearful face. Fifty-six percent of them read sadness as anger. This is universalism?
Crivelli made a very insightful remark when we were talking about the people (like Ekman) who so favored the universalism idea. Many of them belonged to the generation that grew up in the aftermath of the Second World War. They were born into a world obsessed with human difference—in which black people were thought to be genetically inferior and Jews were held to be damaged and malignant—and they were powerfully drawn to a theory that maintained we are all the same.
It is important to note, however, that the work of anti-universalists is not a refutation of Ekman’s contributions. Everyone in the field of human emotion is in some crucial sense standing on his shoulders. People like Jarillo and Crivelli are simply arguing that you can’t understand emotion without taking culture into account.
To quote psychologist Lisa Feldman Barrett—one of the leaders in challenging the Ekman view—“emotions are…made and not triggered.” (See her book How Emotions Are Made [New York: Houghton Mifflin Harcourt, 2017], p. xiii.) Each of us, over the course of our lives, builds our own set of operating instructions for our face, based on the culture and environment we inhabit. The face is a symbol of how different human beings are, not how similar we are, which is a big problem if your society has created a rule for understanding strangers based on reading faces.
For a good summary of this new line of research, see L. F. Barrett et al., “Emotional expressions reconsidered: Challenges to inferring emotion in human facial movements,” Psychological Science in the Public Interest (in press), as well as Barrett’s Emotions (cited above).
Photos of Pan-Am smile and Duchenne smile: Jason Vandeventer and Eric Patterson, “Differentiating Duchenne from non-Duchenne smiles using active appearance models,” 2012 IEEE Fifth International Conference on Biometrics: Theory, Applications and Systems (BTAS) (2012): 319–24.