August 1, 2017
Number 1: Nonverbal behaviors provide accurate information. And it is not obscure; most people can derive at least some of that information with no trouble.
It may be surprising now, but the general consensus when I started my research was that judgments based on facial expression and bodily movements were at best a source of stereotypes. Clinicians knew differently, but not academia. The chairman of psychology at Stanford, who later became head of the National Science Foundation, chastised me in the mid-sixties for wasting my time examining nonverbal behaviors. Had I not read the literature, he asked; didn’t I realize what a waste of time it was?
My first publications documented the accuracy of at least some of the judgments made by ordinary people from viewing nonverbal behaviors, with or without hearing the words that accompanied them. I set about analyzing the literature which had concluded just the opposite, explaining the mistakes that had been made, which were responsible for this mistaken conclusion. My first book, Emotion in the Human Face, published in 1972, reanalyzed and reinterpreted the previous sixty years of research on nonverbal behaviors.
Number 2: Gestural slips exist, which leak suppressed information.
This was a discovery, not a proof like Number 1, which was totally unanticipated and found in my first experiment. I designed that study to verify that nonverbal behaviors indexed the emotional flavor of an interaction between two people. I had arranged for the chairman of psychology to criticize my fellow students, one at a time, and after a few minutes of such induced stress, to then praise each of them for how well they had reacted to his attacks. Thus, I obtained samples of the same person’s behavior under very different emotional circumstances.
One of my peers gave the attacking professor ‘the finger’ gesture; she did not hold it up, as is usually done when this gesture is made deliberately – in what I have called the presentation position – but instead her finger lay on top of one of her knees. She later denied making it, although admitted to having those feelings. The professor said he had not seen it and doubted what I claimed. They were both surprised when they saw the photographic evidence I had made of their interview. Since then, in my series of experiments exploring the behavioral clues that betray lies, I have repeatedly found gestural slips to be one of the most valuable and reliable clues. Its absence does not mean everything is honky-dory; as with other behavioral clues to deceit, it is its presence which has significance.
Number 3: There is strong evidence of universality of some facial expressions of emotion.
When I started this research I didn’t care whether I found evidence of universality or not. Neither deniers nor advocates of universality had more than anecdotal evidence. I recognized the opportunity to settle a very old and important question, but time to do so was running out. The evidence for this research had to be sought in visually isolated groups who could not have learned the meaning of facial expressions from exposure to the media or outsiders. I had to go to the highlands of New Guinea to find such people in 1967 and 1968. Two years later, outsiders came in, as did exposure to Western media, but I had obtained what I and most of the scientific community considered definitive evidence of universality of six emotional expressions: anger, disgust, enjoyment, fear, sadness and surprise. I failed to study contempt, not expecting then (although I do now) that it would have a universal expression.
Number 4: A method to comprehensively and objectively measure all facial movements can be made: FACS (Facial Action Coding System).
There was so little interest in facial expression that no comprehensive and objective measure of facial movements existed. A few had been proposed, but they mixed description with interpretation; for example, snarl, smile, etc. No one knew, or probably cared, about how many different expressions a person can make and how many of them were related to a person’s emotional state. After failing in attempt to create a measurement tool with a shortcut (the Facial Affect Scoring Technique, FAST) in 1971, Friesen and I spent six years developing an anatomically-based comprehensive tool for measuring objectively any movement the facial muscles can generate – FACS.
Number 5: Although lies can be detected from nonverbal behavior, most people can’t do it.
Having studied facial expressions and deception, I wanted to know if others could see what I did. Over 15,000 people from various occupational groups were shown video samples of people lying and telling the truth under threat of punishment. In the end, we found that only a small minority of people reached accuracy in distinguishing deception from truth. The one exception were the members of the U.S. Secret Service; as a group, they did better than chance. The only difference found among the other professional groups was higher confidence, but not accuracy, among law enforcement and psychiatrists. With my training on what to look and listen for, and practicing those techniques, most people were able to make judgments better than chance.
Number 6: Some voluntarily-made facial expressions can generate the physiology of emotion.
This totally unexpected discovery was robust; by making the expressions our research had found to be universal for anger, fear, disgust and sadness, we each generated different patterns of autonomic nervous system (ANS) activity. Making the two facial movements Duchenne suggested were requisite for enjoyment (zygomatic major and orbicularis oculi pars lateralis; in FACS terms: 6+12) produced no consistent changes in ANS, but did generate the Central Nervous System (CNS: EEG) activity observed in other studies of enjoyment. Appraisal theorists, who dominated research on emotion for a time, claimed emotion was not actually generated, only the physiological changes, despite the finding that most subjects claimed they felt the emotion when they made the prescribed facial movements.