Paul Ekman’s Unfinished Research on Emotions and Deception

Paul Ekman Unfinished Research

Why some of Paul Ekman’s research projects went unpublished 

I never published a number of interesting findings. Sometimes it was because I just didn’t have the time to do so, as it is not a simple matter to put what you discover into the restrained language scientific journals require, and then to deal with the hidden anonymous reviewers. Nearly every article I submitted to scientific journals was initially rejected. I then would write counterarguments, eventually succeeding, but it usually took nearly two years after initial submission of an article to get accepted, and usually another year before it appeared in print.

Some projects I never completely finished once I knew the answer, moving on to something else, most often giving priority to writing still another grant to support my lab. I didn’t have graduate students delegate the task of completing projects or writing them up (that was the cost of being a faculty member at a medical school unrelated to a department of psychology). Once I knew the answer to a question, although I wanted to share it, I didn’t always want to invest the time necessary to do so. I instead put my time into publishing the findings that were likely to get grants renewed.

Deception research: blinded by friendship

In one of my early studies in the 1970s of deception, I compared the accuracy of friends with strangers in judging when someone was lying or telling the truth. To my initial surprise, strangers were more accurate! I interpreted this unexpected finding as due to friends trusting their friends. If you suspected your friend was lying, the friendship would not be likely to continue. Friendships, and other intimate relationships, require trust. Trust makes you vulnerable to being exploited or betrayed. We unwittingly avoid information that would disconfirm our trust. We don’t want to find out our children are lying about the use of hard drugs, any more than we want to know the person we recommended for a job is embezzling, or that our partner is having an affair. Rationally, we need to know; psychologically, we don’t. Most lies succeed, I believe, because the target unwittingly cooperates in being duped.

Race and sex bias on perception of faces

Randy Harrison and I investigated how sexual and racial bias influenced the recognition of emotion. In this research, we traced onto acetate sheets (note the very old pre-digital technology) the prototypical expressions of the seven universal emotions. Then we produced other acetate overlays showing a variety of common hairstyles from different ethnicities.

These images were shown for a fraction of a second to college students, who had to judge which emotion they had seen and the intensity of the emotion. Neither sex nor race influenced the judgment of which emotion they had seen, but the strength of the emotion differed. Anger was judged more intense when shown by an African American male as compared to a Caucasian male. Fear and sadness were judged more intense when shown by a female than male Caucasian. Those findings were the same for the caucasians and African Americans who saw the facial expressions. We never took the next step to find out what might be the correlates of the extent of the prejudice shown.

More accurate ratings of subjective emotional experience

I enlisted the aid of the late Maureen O’Sullivan, a close friend and collaborator in many of my deception studies, in a methodological study of self ratings of felt emotion. The typical procedure, which is still widely used today, is to give people scales from one to 7 (or 9), to rate how intensely they felt each emotion, after having undergone some emotional experience. The problem I saw in such an approach is that different people might have very different ideas in mind about what is represented by the end point number 7. I might rate my experience of fear as a 5 because I was thinking 7 stood for fear of impending death, while another person might rate their experiences 5 thinking that 7 stood for fear felt while watching a good horror movie. If I had known that 7 stood for the fear felt in a horror movie, I would have rated my fear as a 6 or 7, not 5. I thought that if asked directly, people would agree that the horror movie fear is not as strong as the fear of impending death. The problem arises because the endpoint of the scale is never defined but left up to the imagination of the responder, and different people imagine different endpoints.

The remedy, we thought, was to define the anchor point for the most extreme point on the scale, what the 7 stands for. To accomplish this, we asked a group of people to write down what they imagined the most extreme fear (anger, sadness, etc.) anyone in the world had ever felt. Interestingly, we found that it was the same for anger and fear: threat of immediate physical harm to self or family. We obtained high agreement about what they imagined would be the most extreme experience of each of the other emotions: death of a child for sadness; unexpected windfall of money for surprise; stepping into animal feces for disgust. Unfortunately, I can’t find what we discovered for happiness and I can’t remember it.

With this data we could define what a 7 stood for, anchoring it in a specific experience. With that remedy we eliminated variations in what people imagined, and everyone used the same scale with the same endpoint defined for them. 

I still regret that we never published what we called AnchorQ, but we would have had to do research on emotions using the usual 7 point scales as compared to an anchored 7 point scale, and show a difference in relationship to expression or physiology. I think we would have done so, but it would take more work and writing a grant to fund it, so it fell by the wayside.

And now, the published work of Paul Ekman

If this article on the unfinished research of Dr. Paul Ekman was interesting to you, be sure to check out his published articles and books for even more insights on emotions and deception.