Abstract
In their paper ‘Evaluating credibility of witnesses—Are we instructing jurors on invalid factors?’, Vrij and Turgeon (2018) argue that jurors should be advised not to consider demeanour when trying to evaluate if witnesses are honest or dishonest because of ‘overwhelming scientific evidence’. However, in this response, we contend that substantial empirical scientific studies on nonverbal communication alongside the limitations of deception detection research, as cited by Vrij and Turgeon (2018), undermine their overall argument. While jurors should be warned about erroneous beliefs and dubious concepts on human communication, jurors should also be advised to consider demeanour as a way of enriching their overall understanding of witnesses and their verbal testimony.
When witnesses testify during trials—even if they are under oath—some will not tell the truth, the whole truth and nothing but the truth (Fawcett, 2014). In adversarial legal systems, such as those in Canada and the United States, if jurors in jury trials (or judges in bench trials) fail to correctly identify honest and dishonest witnesses or if they think that honest witnesses are dishonest and that dishonest witnesses are honest, the consequences can be substantial. In criminal trials, the last step of the judicial process, the liberty of defendants or even their life (in countries where there is still the death penalty) can depend on the deception detection ability of jurors. However, the consequences of inappropriate assessments of testimonies can be detrimental beyond criminal trials. For example, in commercial trials, companies can lose their reputation. In family trials, parents can lose custody of their children. In civil trials, people can lose their life savings. Despite such dire consequences, very little attention has been paid to the detection of deception during trials (Denault and Dunbar, 2019). In their paper ‘Evaluating credibility of witnesses—Are we instructing jurors on invalid factors?’, Vrij and Turgeon (2018) address this very important issue.
The authors argue that jurors should be advised not to consider demeanour when trying to evaluate if witnesses are honest or dishonest because of ‘overwhelming scientific evidence’ (Vrij and Turgeon, 2018: 1). Vrij and Turgeon (2018) call attention to meta-analyses showing that nonverbal cues to deceit are generally faint and unreliable (DePaulo et al., 2003; Sporer and Schwandt, 2007), that people using nonverbal cues have poor deception detection accuracy (Bond and DePaulo, 2006) and that training to improve accuracy of deception detection using nonverbal cues generally offers mediocre results (Frank and Feeley, 2003; Hauch et al., 2014). The authors then reflect on two possible reasons why people overly rely on nonverbal cues to deceit. Finally, Vrij and Turgeon (2018) detail techniques to improve accuracy of deception detection using verbal cues to deceit. They conclude that ‘current well-accepted scientific studies support that human evaluation of a person’s credibility based upon nonverbal behaviours such as manner and appearance is unreliable in determining believability’ (Vrij and Turgeon, 2018: 11).
However, in this response, we contend that substantial empirical scientific studies on nonverbal communication alongside the limitations of deception detection research, as cited by Vrij and Turgeon (2018), undermine their overall argument. While jurors should be warned about erroneous beliefs and dubious concepts on human communication, jurors should also be advised to consider demeanour as a way of enriching their overall understanding of witnesses and their verbal testimony. We end this response with an amended version of Vrij and Turgeon’s (2018) proposed jury instruction.
Erroneous beliefs and dubious concepts on nonverbal communication
Nonverbal communication, that is, ‘communication effected by means other than words, assuming words are the verbal element’ (Knapp et al., 2014: 8), has been extensively studied for the past 50 years. Thousands of peer-reviewed papers from an international scientific community have been published and the field of deception detection has grown substantially (Plusquellec and Denault, 2018; Vrij et al., 2019). However, despite the large body of knowledge regarding the subject, the general public and justice professionals still tend to hold erroneous beliefs, including those related to nonverbal cues to deceit (Bogaard et al., 2016; Denault, 2015; Strömwall and Granhag, 2003; The Global Deception Research Team, 2006). Moreover, pseudoscientific programmes, methods and approaches to ‘catch liars’ are marketed for usage in courtrooms (Denault et al., 2019; Denault and Jupe, 2017; Jupe and Denault, 2018). When considering the above, even if both the Supreme Court of Canada and the United States Supreme Court mentioned demeanour as a factor to assess the credibility of witnesses (Coy v Iowa, 1988; Mattox v United States, 1895; R v D.A.I., 2012), ignoring facial expressions and gestures in courtrooms might seem appropriate. However, rather than solving a problem, such advice may create additional issues, and the judicial fact-finding process could be undermined.
While the general public and justice professionals hold erroneous beliefs on nonverbal cues to deceit, they also hold erroneous beliefs on vocal and verbal messages (Bogaard et al., 2016; The Global Deception Research Team, 2006). The consistency of testimonies and the spontaneity of witnesses, for example, are associated with truthfulness. However, they are not reliable indicators of truthfulness (Denault, Rioux-Turcotte and Tomas, in press). Moreover, as noted by Denault et al. (in press; see also Rioux-Turcotte and Denault, 2018), pseudoscientific programmes, methods and approaches to ‘catch liars’ can disseminate dubious concepts regarding vocal and verbal messages. In other words, if jurors were advised to disregard facial expressions and gestures due to their possible use of erroneous beliefs and dubious concepts, shouldn’t they be advised to disregard testimonies as a whole? The use of stereotypical vocal and verbal cues to deceit is a hazard facing jurors, more so considering lawyers performing cross-examination can draw witnesses into these cues to make them appear nervous and untruthful. Besides, assessments of testimonies as a whole can be distorted by other incorrect assumptions, notably on false confessions (Henkel et al., 2008; Kassin, 2017) and eyewitness evidence (Akhtar et al., 2018; Beaudry et al., 2015; Justice and Smith, 2018; Neal et al., 2012).
The meta-analysis cited by Vrij and Turgeon (2018)
Recently, Denault and Jupe (2018) severely criticised Snook et al. (2017) for making similar arguments to those found within Vrij and Turgeon (2018). Essentially, Denault and Jupe (2018) reiterated that nonverbal cues to deceit are generally faint and unreliable (DePaulo et al., 2003; Sporer and Schwandt, 2007)—the first argument of both Vrij and Turgeon (2018) and Snook et al. (2017)—but added that vocal and verbal cues to deceit are also generally faint and unreliable. Moreover, Denault and Jupe (2018) emphasised that nonverbal communication in face-to-face interactions (including during trials) is of great relevance when trying to understand what others are saying: ‘…interactants simultaneously act with, and form impressions of, their partners. Thus, individuals are encoding information, feelings, intentions, scripts, or other reactions into behavioral expression while, at the same time, decoding the behavior of the partner and experiencing feedback from their own behavior’ (Patterson, 1995: 6; see also Denault and Dunbar, 2017).
In other words, human communication is an intricate process. The importance of nonverbal communication during trials extends well beyond the detection of deception, a fairly small area of research when compared to other areas, including the recognition of emotions (Burgoon et al., 2010; Knapp et al., 2014; Moore et al., 2014). For example, gaze aversion is not a reliable indicator of deception (DePaulo et al., 2003). However, gaze serves a plethora of other regulating, monitoring and expressing functions which have been documented for decades, hence the importance of gaze during trials as in face-to-face interactions (Adams and Kleck, 2003, 2005; Emery, 2000; Kendon, 1967; Kleinke, 1986; Mason et al., 2005; Perrett and Emery, 1994). The same holds for facial displays. In addition to the expression of emotional states and social motives (Hareli and Hess, 2017; Parkinson, 2005), the face performs a wide range of communicative functions such as ‘adding content, establishing the significance of verbal communication, helping understand others, and providing feedback to coordinate and manage face-to-face interactions so that conversational partners achieve their own goals’ (Denault and Dunbar, 2017: 293–294).
Furthermore, both Vrij and Turgeon (2018) and Snook et al. (2017) point out that people using nonverbal cues have poor deception detection accuracy (Bond and DePaulo, 2006). However, we believe this second argument to ignore demeanour to be flawed. During trials, the primary responsibility of jurors is to ascertain the facts, not to determine who is lying and who is telling the truth. There is not always a witness lying and a witness telling the truth. Testimonies can be diverse and can appear irreconcilable, but witnesses can still tell the truth, tell their sides of the story from their own perspectives, informed by their own premises (Scheppele, 1989). To think otherwise could foster a confirmation bias and adversely influence the trial’s outcome, more so if jurors ignore the fact that credibility is not a synonym of veracity (Jupe et al., 2018; Porter and ten Brinke, 2009). High-credibility witnesses can lie and low-credibility witnesses can tell the truth (Bond and DePaulo, 2008). Therefore, jurors ‘should use nonverbal communication to enrich their understanding of witnesses and their testimony with a fuller grasp of the complexities of communication and the various reasons for the nonverbal behaviors they see’ (Denault and Dunbar, 2017: 307–308).
In addition, both Vrij and Turgeon (2018) and Snook et al. (2017) advocate the use of techniques developed for investigative interviews, techniques used to elicit verbal cues to deceit. However, during jury trials, the use of such techniques is highly unlikely because, predominantly, jurors cannot ask witnesses open-ended and follow-up questions. The same holds for bench trials in adversarial legal systems where judges are not investigators (Gerber, 1987; Sward, 1989). In addition, one should not expect lawyers to use techniques developed for investigative interviews. The nature and goal of questions asked by lawyers during examinations and cross-examinations are entirely different from those asked by police during investigative interviews (Denault and Dunbar, 2019). For example, during cross-examinations, the lawyers’ goal is not to get information. They will generally ask short leading questions that call for a yes or no answer to get witnesses to agree with them. They do not want witnesses to add previously unmentioned details that could harm their side of the story (McElhaney, 1997; Younger, 1976). The ‘advocate’s prime loyalty is to his client, not to truth as such’ (Frankel, 1975: 1035). Lastly, asking witnesses open-ended questions is as unlikely as asking them unexpected questions. Well-prepared witnesses should know what questions will be asked, more so if there is an obligation of disclosure of evidence (Denault et al., 2017; Small, 2014).
Therefore, we do not believe that ignoring the nonverbal communication of witnesses is the solution, even more so when the suggested alternatives serve no purpose during trials. Using techniques during trials developed for investigative interviews appears all the more questionable considering deception detection research has not, to our knowledge, examined the use of multiple deceptive strategies, whether lawful or unlawful, by multiple parties (including lawyers), sometimes over hours, days, months or years. In addition, because face-to-face interactions in courtroom contexts are guided by the rules of evidence and the judge’s and jury’s roles, the ecological validity of other deception detection research is further compromised (Denault and Dunbar, 2019; Denault and Jupe, 2018; Frankel, 1975).
Finally, the third argument of both Vrij and Turgeon (2018) and Snook et al. (2017) is also flawed. Training to improve accuracy of deception detection using nonverbal cues generally offers mediocre results (Hauch et al., 2014; Frank and Feeley, 2003). However, training to improve accuracy of emotion, intention and personality judgements using nonverbal cues offers noteworthy results (Blanch-Hartigan et al., 2012, 2016; Blanch-Hartigan and Ruben, 2013; Schlegel et al., 2017). Therefore, considering that the primary responsibility of jurors is not to determine who is lying and who is telling the truth, training to detect liars and truth tellers should be abandoned. However, in Canada and the United States, as in any jurisdiction where there are witness credibility issues, training using nonverbal cues to help jurors in jury trials (and judges in bench trials) to understand the testimony before them should not be underestimated. This becomes all the more important considering that jurors cannot sit and watch witnesses for hours, days, months or even years, and be reasonably expected to disregard what witnesses look like and how they behave, including their facial expressions and gestures, more so considering that social judgements can occur automatically, in a few seconds, outside of conscious awareness (Knapp et al., 2014). Asking them to ignore demeanour, as suggested by Vrij and Turgeon (2018), is an impossible task and will most likely result in frustration and confusion on the part of jurors.
Restoring the role of nonverbal communication during trials
Vrij and Turgeon (2018) rightly reflect on the erroneous beliefs that nonverbal is more revealing than verbal communication and that nonverbal is more spontaneous than verbal communication. These are two possible reasons why people overly rely on nonverbal cues to deceit. However, the impact of such erroneous beliefs can extend well beyond the detection of deception. For example, when trying to understand what others are saying, the widely disseminated faulty assertion that 93% of communication is nonverbal could result in the overestimation of the nonverbal messages and the underestimation of the verbal and vocal messages (Lapakko, 1997, 2007). During trials, this could result in inappropriate assessments of testimonies.
Unfortunately, other erroneous beliefs could be detrimental as well, including those which state that body movements have absolute meanings to know what others are thinking but not saying. Sometimes, so-called ‘body language experts’ even claim (implicitly or explicitly) the universality of specific meanings when such claims have not been subjected to any peer-reviewed studies (Denault et al., 2015, 2019; Lardellier, 2017). The judicial fact-finding process could be undermined if jurors ignore that, in fact, ‘there are few, if any, body movements that have invariant meaning…[and] even within a culture, body movements do not carry the same meaning each and every time they are displayed’ (Harrigan, 2008: 139). The meaning of nonverbal behaviours can vary, depending on contextual and cultural factors, amongst other things, and the same holds for verbal messages. Words do not carry the same meaning each and every time they are uttered (Jackson, 2012; Kuhn, 2017; Paulston et al., 2012; Schiffrin et al., 2003; Sidnell and Stivers, 2013; Thomas, 1995), which further supports the importance of appreciating adequately human communication. Therefore, because the importance of nonverbal communication during trials extends well beyond the detection of deception, Vrij and Turgeon’s (2018) proposed jury instruction could be amended to reflect such (see Appendix 1 for comparison): Many people think that how witnesses behave reveals whether they are being deceptive or truthful. Research shows, however, little difference between the demeanor of deceptive and truthful people. For example, both deceptive and truthful witnesses can appear nervous and avert gaze. Their preparation (or the lack thereof) as well as contextual and cultural factors, amongst other things, can influence their demeanor. Therefore, rather than using demeanor to determine who is lying and who is telling the truth, consider using demeanor to enrich your overall understanding of witnesses and their verbal testimony. Keep in mind that body movements are not reliable indicators of deception (or truthfulness) and do not have absolute meanings, but that information, feelings, intentions, scripts, or other reactions transmitted by facial expressions and gestures can help you understand witnesses and their verbal testimony better.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Appendix
Comparison between Vrij and Turgeon’s (2018: 3) proposed jury instruction and Denault, Dunbar and Plusquellec’s proposed amended jury instruction.
| Vrij and Turgeon’s (2018: 3) proposed jury instruction | Denault, Dunbar and Plusquellec’s proposed amended jury instruction |
| When considering whether to believe a witness, many people think that how the witness behaves reveals whether he or she is being deceptive or truthful. Research shows, however, that most people, including experts, are not very good at correctly deciding whether a witness is truthful based upon how the witness looks or acts, as opposed to what he or she says. In fact, research shows little difference between the demeanor of deceptive and truthful people. Keep this in mind when considering non-verbal behavior by a witness; that is, such behavior is an unreliable indicator of deception or truthfulness. The types of nonverbal factors you may encounter include a witness who averts his or her gaze (or looks others in the eye), shifts in his or her seat (or remains still), appears nervous or anxious (or calm), and speaks quickly (or slowly). You should ignore these kinds of factors as they are unreliable indicators of deception (or truthfulness). | Many people think that how witnesses behave reveals whether they are being deceptive or truthful. Research shows, however, little difference between the demeanor of deceptive and truthful people. For example, both deceptive and truthful witnesses can appear nervous and avert gaze. Their preparation (or the lack thereof) as well as contextual and cultural factors, amongst other things, can influence their demeanor. Therefore, rather than using demeanor to determine who is lying and who is telling the truth, consider using demeanor to enrich your overall understanding of witnesses and their verbal testimony. Keep in mind that body movements are not reliable indicators of deception (or truthfulness) and do not have absolute meanings, but that information, feelings, intentions, scripts, or other reactions transmitted by facial expressions and gestures can help you understand witnesses and their verbal testimony better. |
