The popular Paul Ekman hypothesis of facial micro-expressions as indicators of lies has no scientific support

Research on Non-verbal Signs of Lies and Deceit: A Blind Alley. Tim Brennen and Svein Magnussen. Front. Psychol., December 14 2020. https://doi.org/10.3389/fpsyg.2020.613410

Introduction

Research on the detection of lies and deceit has a prominent place in the field of psychology and law with a substantial research literature published in this field of inquiry during the last five to six decades (Vrij, 2000, 2008; Vrij et al., 2019). There are good reasons for this interest in lie detection. We are all everyday liars, some of us more prolific than others, we lie in personal and professional relationships (Serota et al., 2010; Halevy et al., 2014; Serota and Levine, 2015; Verigin et al., 2019), and lying in public by politicians and other public figures has a long and continuing history (Peters, 2015). However, despite the personal problems that serious everyday lies may cause and the human tragedies political lies may cause, it is lying in court that appears to have been the principal initial motivation for the scientific interest in lie detection.

Lying in court is a threat to fair trials and the rule of law. Lying witnesses may lead to the exoneration of guilty persons or to the conviction of innocent ones. In the US it is well-documented that innocent people have been convicted because witnesses were lying in court (Garrett, 2010, 2011; www.innocenceproject.com). In evaluating the reliability and the truthfulness of a testimony, the court considers other evidence presented to the court, the known facts about the case and the testimonies by other witnesses. Inconsistency with the physical evidence or the testimonies of other witnesses might indicate that the witness is untruthful, or it may simply reflect the fact that the witness has observed, interpreted, and later remembered the critical events incorrectly—normal human errors all too well known in the eyewitness literature (Loftus, 2005; Wells and Loftus, 2013; Howe and Knott, 2015).

When the facts of the case are not well known, witness testimonies, including the testimony from alleged victims, may be critical to a verdict, and these testimonies are sometimes from witnesses who hold a personal stake in the case and shun self-incriminating statements. In many countries, a witness lying in court risks being charged with perjury—the accused typically does not risk such a reaction—but there are still cases where witnesses lie. In such cases, when there is a possibility that one or more of the witnesses are lying and the court's verdict depends upon the perceived credibility of the witnesses, the issue arises of distinguishing between lying and truthful witnesses. Is it possible to identify liars vs. truth tellers based on the non-verbal signals transmitted by the sender?

Discussion

What options does this research field now have? Does one carry on looking for reliable non-verbal cues? Does one concentrate on whether combinations of them are diagnostic of lying? Vrij et al. (2019) suggest that there are grounds for optimism, for instance, by better defining the terms, or by improving measurement of the non-verbal cues. Luke (2019) recommends increasing the power of studies by increasing sample size. We are doubtful that such strategies will be able to provide solace, because they will be unwieldy in the forensic context. To illustrate this, let us consider two phenomena: Vrij et al. (2015) reported that spontaneous saccadic eye movements (a measure related to the widely-believed-but-not-empirically-supported gaze aversion cue) distinguish between truth-tellers and liars, and Mann et al. (2012) reported that if one measures “deliberate eye contact” rather than eye contact per se, liars have longer eye contact than truth-tellers. The reason for our skepticism regarding the application of such effects is that it is difficult to apply small (albeit significant) effects to specific instances. With such cues it will generally not be possible to say who is telling the truth at an individual level, or indeed at the level of an individual statement. One could measure the spontaneous saccades of key witnesses or the amount of deliberate eye contact maintained by a witness giving their statement but it is not clear either that such measures are sufficiently reliable, or what the baseline condition should be, against which one would compare the collected data in order to declare the statement a lie or not. Is the research in a blind alley? We believe it is, as far as lie detection in the forensic context is concerned. The idea that governs the research, that there are reliable non-verbal signs to lies and deceit is itself an expression of the western psychological folklore—as pointed out by Nortje and Tredoux (2019), the theoretical foundations for the putative non-verbal cues are shaky—and few researchers in the field appear to have fully digested the possibility that the basic premise of their inquiry may be false. For complex intellectual behaviors it has long been realized that there is a number of broad factors that contribute to individual differences—genetics, cultural influences, personal experiences, and situational factors (Engel, 1977, 1980). To complicate matters, a meta-analysis by Bond and De Paulo (2008) showed that participants' truth judgments depended on the sender rather than on the person doing the judging. The effect of sender on veracity judgments has been confirmed in a number of subsequent studies: Some of us appear more (or less) credible than others, independent of whether we are telling the truth or are lying (Porter et al., 2010; Levine et al., 2011; Korva et al., 2013). In addition, the existence of the literature on cultural differences in lie detection, e.g., Castillo and Mallard (2012), would seem to undermine the idea that lies and deceit are in any useful, systematic manner related to behavior on an individual culture-free basis. We may have been looking for a lawfulness in human behavior that exists only in our minds.

Is the rational course simply to drop this line of research? We believe it is. The creative studies carried out during the last few decades have been important in showing that psychological folklore, the ideas we share about behavioral signals of lies and deceit are not correct. This debunking function of science is extremely important. But we have now sufficient evidence that there are no specific non-verbal behavioral signals that accompany lying or deceitful behavior. We can safely recommend that courts disregard such behavioral signals when appraising the credibility of victims, witnesses, and suspected offenders. For psychology and law researchers it may be time to move on.