Brunel Centre for the Study of Arbitration and Cross-Border Investment
23rd May 2013
The relevance of psychology to international arbitration: The assessment of credibility
Dr Lucy Akehurst, Principal Lecturer,
Department of Psychology, University of Portsmouth
The Court of Appeal said in one case, ‘The due administration of justice does not always depend on
eliciting the truth; it often depends on the burden of proof’. If that is right, assessing a witness’s
credibility might be regarded as secondary to the discharge of the burden of proof. But is that right?
Lord Denning did not think so. He described the judge’s function in these terms, ‘Above all to find
out the truth and to do justice according to law’. That, I suggest, is the reality of it. Legal
professionals are assessing the credibility of the witnesses, finding facts based upon that assessment,
and deciding what did or did not happen. Discharge of the burden of proof is of course relevant but
rarely decisive.
One thing is certain about the assessment of credibility. Legal professionals receive little or no
training in it. I often hear that training in lie detection won’t take long as it’s all gut instinct. However
I believe that it is possible to provide people with information they may not have, to give some
guidance on what they should not do and, more positively, on what they should do when judging
credibility.
Nonverbal cues to deception
According to the research there are some nonverbal cues to deception but they are weak. Certainly there is nothing remotely resembling Pinocchio’s growing nose and I find that a lot of the time when I am training police officers, social workers and insurance fraud investigators I am dispelling the myths surrounding deceptive behaviour.
We know the following about nonverbal cues to deception: Liars seem to include longer pauses in their speech than truth tellers, but interestingly we do not find that they include more pauses; They wait longer before giving an answer; They make more word and phrase repetitions; They make fewer hand and finger movements. Of these four findings the last is the best evidenced. Lying is cognitively demanding. Liars have to think to lie. This cognitive overload results in a neglect of body language and reduces overall animation. So liars, on the whole, are not fidgets.
Furthermore, gaze aversion is not related to deception; liars do not look away from the target of their deceit. There are two reasons for this. First, liars know that most people regard gaze aversion as a sign of lying, so, not being stupid, liars look people straight in the eye instead. Second, gaze is related to many factors that have nothing to do with credibility. For example, people make more eye contact with people they like and less with those they dislike, further they make less eye contact when they are embarrassed. There are also cultural differences at play when it comes to keeping eye contact and so to use this as an indicator of credibility is not advisable.
Does all this mean that if a witness does not move a muscle and takes ages to answer questions, he or she must be lying? Unfortunately not, the picture is more complicated than that. Indeed, individual differences come to the fore. We would be scoffed at for presuming that everyone behaves the same way, non-verbally, when relaxed and telling the truth so why the insistence on coming up with a blueprint for demeanour indicative of deception? Moreover, the same person can behave in different ways dependent on the context of a lie, to whom s/he is lying and so on. For example, a person who tells a spontaneous lie is more likely to display nonverbal cues to deception than when s/he tells a planned lie. The reason is that the cognitive overload on the spontaneous liar is greater: s/he has to think much harder and much more quickly.
What conclusion can we draw from all this? The conclusion is that nonverbal cues to deception are at best weak or, in lawyer’s terms, that demeanour is a poor guide to the assessment of credibility.
Speech content as a cue to credibility
Over the years I have conducted research into the usefulness of two techniques that soley use aspects of what is said by interviewees to assess their credibility. That is, videorecordings are not required, only word-for-word transcripts of interviews. The first evolved in Germany, Statement Validity Assessment (SVA). SVA is made up of 19 criteria which, if present in a statement, are indicators of truthfulness.
Examples of two of the SVA criteria are given here (i) unexpected complications, when an interviewee includes in his/her statement a ‘twist or turn’ in events that, whilst not implausible, is unexpected e.g. “...he tried to lock the door but he couldn’t find the key”. This sits well with the cognitive load framework, that is, that it is hard to tell a lie, especially a complicated tale. Liars tend to keep their stories straight forward and are not likely to include such embellishments,
(ii) reproduction of conversation, when parts of a conversations are reported in original form or if different speakers are recognisable in reproduced dialogues. Naturally occurring events typically involve speech. This criterion is satisfied when there is a replication of an utterance e.g. “he said “are you OK? You look ill” and then he left” but not by “he asked how I felt before he left”.
A second technique, Reality Monitoring (RM) comprises 8 criteria, again it is only the speech content that is used to make a judgement of credibility. The criterion I have included here for RM is sensory detail. Should a statement include details of how something smelt, tasted or felt we consider these details to be indicators of truthfulness, that is that the interviewee has actually experienced what s/he is describing. If someone is basing his/her fabrication on information from the media, soap operas, films etc s/he can describe what s/he ‘saw’ and ‘heard’ but not this other sensory detail thus distinguishing liars from truthtellers.
Research participants trained in these techniques have performed significantly better than those using nonverbal cues to lying .
How can legal professionals improve their assessment of credibility?
In terms of gathering evidence, interviewers should not make up their minds too quickly about whether a witness is lying. They should be suspicious but not show it, use an information-gathering style rather than an accusatory style and let the witness repeat him/herself. Furthermore interviewers should make the interview difficult. This may be achieved by asking unanticipated questions, asking the same question but in a different form and asking for recall in a non-chronological order (e.g. in reverse order). All of these techniques increase the cognitive load for interviewees and accentuate the differences between truthtellers’ and liars’ behaviours.
Finally, I would like to mention impression management. Liars would like to be perceived as truthtellers. However they are not very good at pulling this off as they are not accurate in their beliefs about how truthtellers behave. They think that there should be no chinks in their armour, that their statements should be watertight, that they should know the answer to all questions and they will of course avoid putting themselves down. Recalling an event truthfully can be difficult especially if it happened a long time prior to the interview. As such truthtellers do make spontaneous corrections, they do admit lack of memory and they do sometimes blame themselves. Liars tend to avoid these behaviours in their quest to be perceived as truthful.