Hon Robert Fisher: Dishonest face is no guarantee of guilt

During the recent trial of John Banks, Mrs Dotcom and Mrs Banks gave conflicting accounts of a lunch they both attended. One said donations had been discussed. The other said they had not.

The judge believed Mrs Dotcom and not Mrs Banks. Among his reasons for doing so was the fact that he had watched Mrs Dotcom and associated witnesses carefully while they gave their evidence and had not noticed anything in their demeanour to suggest that they were not telling the truth. Mr Banks was convicted.

After much inquiry Mrs Banks managed to locate two other people who had been present at the lunch. They backed up her version. When this was brought to the attention of the Court of Appeal, it quashed the conviction and ordered a new trial.

There was nothing unorthodox about the way in which the trial judge went about deciding who to believe. But his reference to demeanour illustrates a problem that refuses to go away: judges and juries still believe they can tell whether or not witnesses are lying by studying how they look and sound.

Those who believe this will be fortified by lessons from television. Dr Cal Lightman of the television series Lie to Me can see through porkies, as can the former psychic Patrick Jane in The Mentalist. They do so by leaning forward and staring intently into their subjects' faces, usually with remarkable acquiescence on the subjects' part.

It seems that they detect the facial micro-expressions which betray lies. And even if liars were clever enough to mask their micro-expressions, popular author Allan Pease assures us that lack of congruence between facial signals and body gestures would give them away. The lesson is that lying can be detected by careful observation.

This must be reassuring to judges and lawyers. They have believed it for years. Appellate courts still defer to trial judges and juries on the basis that only the latter had the opportunity to watch and listen to the witnesses. Trial judges directing juries still give them unqualified encouragement to rely on everyday experience in deciding who to believe. It seems that through ordinary life experience, each of us has become a Patrick Jane without realising it.

Or have we? The assumption that demeanour betrays lying has been repeatedly tested by social scientists. The tests begin from the premise that speakers provide two sources of information - verbal and non-verbal. The verbal information is the meaning of the speaker's words. The non-verbal information is the way in which they were uttered. A speaker simultaneously manifests a package of non-verbal information which is normally subconscious. It consists of facial expressions, bodily movements and vocal characteristics. The vocal characteristics, also known as "paralinguistic cues", include pitch, pace, volume, timbre, expression and tremors. It is this package of face, body and voice that lawyers describe as "demeanour".

Experiments have been conducted to gauge the extent to which observation of demeanour helps when assessing truthfulness. The experiments have been conducted by many scientists, in many different ways, in many countries, over many years. The fundamental result is always the same: when trying to decide whether someone is lying, ordinary observers derive no benefit from the opportunity to observe non-verbal behaviour. The result of attempting to do so is no better than the toss of a coin. Worryingly, however, most confidently believe that they can.

Fortunately there are many other ways of deciding whether to believe someone. We can ask whether the story offered is internally consistent ("didn't you say a moment ago that you did not see the accident?"); whether it is consistent with the one given previously ("didn't you say the opposite in your statement to the police?"); how it compares with details that were written down at the time ("don't these emails at the time suggest that you did receive the money?"); how it compares with the evidence of other witnesses ("why do you think the neighbour says she saw you running away?"); the witness's general character and credibility ("is it true that you already have six convictions for perjury?"); and the plausibility of the story given ("do you seriously expect us to believe that when you collided with the other man his wallet fell into your pocket?").

There are many ways of telling whether someone is telling the truth. Unfortunately they do not include the way they look and sound. Science has consistently proved that reliance on demeanour is misplaced. Those who think they can tell by that means are mistaken. One fears for the many miscarriages of justice that must have resulted.

The Hon Robert Fisher, QC, LLD, is a former High Court judge and special adviser to the parliamentary select committee on the Evidence Act. He practises as an arbitrator and mediator.