The Neutral as Lie Detector: You Can’t Judge Participants by Their Demeanor

First published by ABA Dispute Resolution Magazine, Winter 2001

As mediators we are often faced with sharply conflicting stories. One of the advantages of mediation is that we sometimes can solve the underlying problem without determining who did what, to whom, and when. Indeed, experience has shown that mediation is not a good process for finding the truth because it has none of the tools (such as testimony under oath) used for this purpose in the judicial system. Still, mediators often spend a good deal of time and effort trying to determine who is telling the truth.

Unfortunately, if the experts are to be believed, this task may be futile. Research efforts have found that one out of three normal conversations contained a lie, 92% of university students were found to tell serious lies to their intimate friends, and people reported telling from zero to 46 lies each week.

This frequency notwithstanding, most of the time we don’t expect people to lie to us (except in the case of politicians). If they do lie, we usually are not aware of it because very little is at stake in everyday conversation. We typically trust that what other people say is truthful because we believe ourselves to be truthful most of the time. Yet, as neutrals, we are continually faced with the possi-bility that a party is lying. Sometimes we are convinced the advocates are lying as well. With each person, we are faced with the dilemma of concluding a false positive (a truthful person incorrectly judged to be lying) or a false negative (a liar incorrectly judged to be truthful). Ordinary people are motivated to lie on a great number of occasions. For example, lies may be motivated by politeness (That was a delicious meal), by guilt (Not to worry, dear, I got a great deal on the printer), by the desire to keep children innocent (Santa Claus will come tonight), or by self-promotion (You wouldn’t believe how busy my practice is). Indeed, there are some people who revel in their ability to lie and get away with it. One commentator has suggested that Americans must go out of their way to reward this behavior because during the past 20 years we elected two Presidents to second terms when there was clear evidence both were frequent if not pathological liars.

There is even a taxonomy for types of lies. The venal lie is a serious misrepresentation without any redeeming properties (I have never had sex with that woman). The white/social lie is a relatively innocuous misrepresentation without serious implication where the liar views truth as potentially disadvantageous to the hearer (We couldn’t make it because we couldn’t find a baby sitter). The fib is a non-serious, small but selfishly motivated misrepresentation (We have never had any flooding in the basement). The compassionate lie is a misrepresentation of a serious state of affairs to spare the hearer (Even though you have a serious illness, there is a good chance for a high quality of life for quite a while). The justified lie involves a situation where consequences of telling the truth are serious and most people would agree that the goal justifies the deception (Let the hostages go and you will be treated well).

Lies fail for three main reasons. First, if the statements of the person are self-contradictory, we suspect that something is amiss. Second, if other people or documents contradict a person, again we assume that someone is not telling the truth. And third, there are the cues of demeanor, the gestural and vocal cues that every speaker evidences.

What cues do we use to assess whether a person is lying? Many gestural and vocal characteristics have been studied as signaling the liar. These include cues such as increased blinking rate, avoidance of eye contact, dilation of the pupils, self-manipulating gestures such as rubbing or scratching, use of a higher pitch, throat clearing, hesitations, repetitions, grammatical errors, slips of the tongue, tightening of the lips, fleeting facial expressions, and a smile that doesn’t use any of the muscles around the eye. And these don’t exhaust the list.

However, these cues can help a neutral detect lies only if the following assumptions are correct: 1. A given witness will manifest a subset of these characteristics only when a lie is being uttered.
2. These characteristics are systematically associated with lying (e.g., only when lying will this speaker evidence eye dilation and scratching).
3. These traits are learnable by the neutral.
4. The traits are discernible by the neutral in an on-line mode while he/she is listening to what is said and writing down what is said.
5. There is an opportunity for feedback to verify the conclusions made.

To test these assumptions, one research effort trained judges to observe speakers, assess their veracity, and then receive feedback on their judgments. The judges did increase their accuracy on the specific speakers participating in the study, but not on anyone else.

Unfortunately for neutrals, there is good reason to conclude that these non-verbal characteristics will vary depending on the motivation for getting away with lying (explaining to the police how little you have had to drink vs. what your golf handicap is), on the guilt associated with the lie (de-ceiving your spouse vs. telling a child the Easter Bunny is coming), and on the attitude with which you engage in the questioning (subpoenaed to the hearing vs. enthusiastic participant out for revenge).

Furthermore, while the characteristics listed above are for the “normal” self-conscious liar, there is the so-called Machiavellian liar, who can look you in the eye, speak in a calm and measured manner, exhibit none of the “tells” of the liar, and all the while lie. One interesting feature of these liars is that they have enough self-assurance to “read” the reactions of the listener and adapt their language and style accordingly. One example is the experienced salesperson who, research has shown, is among the most difficult to detect lying.

Research in detecting lies by observing demeanor has revealed that even under experimental conditions, where the participants have to make judgments only as to whether the person on videotape is lying or telling the truth, most people do no better than chance. That is, about half the time they judge a liar to have been telling the truth, and half the time they judge a truth-teller to have been lying. Even polygraphs are said to do better than this!

This sort of experiment has been run on groups of professionals including judges (but not arbitrators or mediators), and the only group that consistently performed better than chance was the U.S. Secret Service. Speculation is that this greater success is due to their training not to trust any-one and to be continually alert for small signs that might signal something non-normal. Even when subjects in experiments are told that some of the people they will view will be lying, they do not improve in detecting lies. However, they are much more confident in their wrong judgments. One experiment had participants tell lies to attractive and unattractive listeners. The judges who observed these interactions did better at detecting the lies when the listeners were attractive, and this was especially true if the judges themselves were attractive. Another experiment had subjects from two very different cultures try to determine who was lying. There was fairly good intra-culture reliability, but the participants were terrible when making the judgment inter-culturally. Finally, contrary to the folk wisdom that people can tell when intimates are lying to them, the reverse has been found.

There is a great temptation to conclude that a person is lying when, for example, he manifests a high shaky voice, exhibits disfluencies in speech, and perspires. But the relevant research, as well as common sense, strongly suggests that we, as neutrals, should not routinely attempt to make veracity decisions on the basis of witness demeanor. If you want to read further about lying, there are literally dozens of books and articles dealing with deception in general and lying in particular. I recommend three books as worth reading: Ly-ing, Telling Lies, and Lies! Lies!! Lies!!! The Psychology of Deceit. I wouldn’t lie to you.

Bella DePaulo, et al. Lying in Everyday Life, 37 J. of Personality & Soc. Psychol. 1713-1722 (1996)

Paul Ekman. Telling Lies. Norton Press, 2nd Edition. (1992)

Sylvia Bok. Lying: Moral Choice in Public and Private Life. Pantheon Press. (1978)

Charles Fold. Lies! Lies!! Lies!!! The Psychology of Deceit. American Psychiatric Press. (1996)


Bruce Fraser

Bruce Fraser is a Professor of Linguistics and Education at Boston University as well as a labor arbitrator/mediator. His research interests lie in pragmatics, forensic linguistics, and discourse analysis. MORE >

Featured Mediators

View all

Read these next


Mediation Practice Guide: A Handbook for Resolving Business Disputes (Book Review)

This review first published in THE CPA JOURNAL, May 2004. Order at Amazon.comSecond Edition By Bennett G. Picker Published by the American Bar Association Section of Dispute Resolution, $39.00, ISBN:1-59031-169-8,...

By Philip Zimmerman, CPA

Why are Attorneys Afraid of Conflict in Mediation?

JAMS ADR Blog by Chris PooleThe mediation process has evolved significantly over the past few decades. Mediation was initially viewed skeptically by trial attorneys who viewed themselves as warriors who...

By Jeffrey Grubman

How to be Truly Helpful When Someone is Upset

Conflict Zen Blog by Tammy LenskiWhen someone is upset, one familiar response is to ignore it and forge ahead. Another is to try to make them feel better with kind...

By Tammy Lenski

Find a Mediator