Skip to content

Speech to the London Court of International Arbitration's Asia-Pacific Users’ Council Symposium, Sydney
14 February 2003

A man walks into a post office one day to see a middle-aged balding man standing at the counter methodically placing "Love" stamps on bright pink envelopes with hearts all over them. He then takes out a perfume bottle and starts to spray them with scent.

His curiosity getting the better of him, he goes up to the balding man and asks him what he is doing. The man says:

"I'm sending out 1,000 Valentine cards signed, 'Guess who?'
"But why?" asks the man.

"I'm a divorce lawyer," the man replies.

Valentines Day is an auspicious day on which to hold World Arbitration Day. However, we would not want feelings of affection to reduce the occurrence of commercial dispute.

Today is also the 74th Anniversary of the day when seven men were machine gunned in a Chicago warehouse by mobsters, dressed as policemen, who worked for Al Capone.

Not quite so auspicious. At least they were not a panel of LCIA arbitrators who had just decided that the Capone whisky was not of merchantable quality.

Maybe this is a negative view. Perhaps the executioners should be seen as the arbitrators. After all this was before the New York Convention. Arbitrators even then knew the importance of enforcement. But the mechanisms were limited.

St Valentine seems to have had nothing to do with amorous pursuits by the way. It was just that his saint's day was chosen to perpetuate the Roman festival of Lupercalia, celebrated with the coming of spring, which included fertility rights and the pairing off of women and men.

Characteristically, here in Australia, we celebrate this spring rite, with its associations with fertility, as autumn and winter approaches.

That reminds me to issue a plea to the newsletter and journal editors who are here. Northern Hemisphere editors, particularly editors of legal and arbitration publications, like dating their publications by reference to the seasons. Opening an envelope in January in high Summer to find the Winter Number of the ABA Journal can be a little confusing.

I have chosen a practical subject to talk about tonight. You have probably heard enough about the Lex Mercatoria, multi-tiered clauses, and conflicts of interest. As to the last I hope that the delegation here from my English Chambers, Essex Court Chambers, saw and defended the veiled attack present in James Hope's question about the members of a private society. Those of you who are not at the Bar need only to remember that there can be no conflicts of interest between barristers because they do not practise in partnership.

I thought I would talk about a topic which is loosely associated with some of the subjects discussed today in the session on Proceedings and Practice.

There is a lot of debate about due process and the extent to which it requires oral evidence and cross examination, but not so much about the inherent virtue of oral evidence itself as a means of discovering the facts. So I thought I would offer a few remarks on this topic.

I have learned enough from Continental colleagues to know that there is a respectable tradition of gathering the facts without any need for emphasis on oral evidence.

The case for fact finding without emphasis on oral evidence is more compelling in arbitration than it is in civil litigation. This is because all arbitration is rooted in writing, namely an arbitration agreement, normally contained in a contract. Negotiations and dealings are usually substantially in writing as well, particularly in international arbitration. The need for eye witness evidence, such as oral accounts of which side of the road a vehicle was on at the time of a collision, does not usually arise in arbitration. As technology permits everything we say and do to be recorded, the need for oral accounts after the event becomes less and less.

So what is the need for oral evidence at all in arbitrations? French colleagues nod in agreement.

Common lawyers know that oral evidence is vital, and cross examination in particular. This is because an uncountable number of judges and lawyers have said so in the past. And dare I say it, to those of us who have spent a lifetime earning income from oral hearings, denying their virtue is a little bit like shooting yourself in the foot.

Whenever common lawyers talk about the virtues of oral evidence it is not long before the word 'demeanour' is used.

The New South Wales Court of Appeal recently said, unanimously, in upholding the decision of a trial judge:

'His Honour had the benefit of observing the demeanor of the appellant and the [respondent's] witnesses and to judge their respective credibility. His conclusions were well and truly open on the evidence.' (1)

The Court of Appeal was echoing the following words of the High Court of Australia:

'[W]hen a trial judge resolves a conflict of evidence between witnesses, the subtle influence of demeanour on his or her determination cannot be overlooked…[N]o matter how impressive Professor Ferguson's evidence may appear, it cannot claim the consideration of an appellate court to the extent necessary to overcome the advantage which her Honour enjoyed in seeing and hearing Mrs. Archer give evidence.' (2)

The High Court, in turn, was impressed by the words of a member of the House of Lords in 1927. (3)

Very recently, in a New South Wales case involving a high profile Australian, which received a good deal of publicity, the Judge said this:

'There are clear conflicts in the evidence which require resolution. In this regard much depends upon my assessment of the demeanour of the witnesses who appeared before me.' (4)

Of one witness he said:

'Having had the advantage of observing Mr Cousins in the witness box, I accept his evidence in preference to the notes taken by Ms Bardsley at the meeting of 6 September 2002.' (5)

Of another witness he said:

'He was a quietly spoken man who retained his composure throughout the two days he was in the witness box and I have no doubt that his evidence in this regard was his honest attempt to recollect the events.' (6)

It is not apparent to me that being softly spoken and retaining composure are necessarily signs of truthfulness.

French lawyers are apt to suggest that assessments of demeanour are as prone to protect plausible liars as they are to uncover less practised ones. It has been suggested that observing manner can sometimes assist in discovering a liar but rarely helps in assessing truthfulness.

Of course, demeanour alone very rarely provides the answer, even to a common law judge. Analysis of judgments shows that the judge who purports to be relying on demeanour is always testing it against objective facts, usually what is written in documents. Indeed, the judge who purports to base conclusions on demeanour in fact often places greater faith on the documents.

Nevertheless, there are judges who have found what witnesses said to be truthful from assessing demeanour notwithstanding a contradictory document. The high profile case I referred to before is an example.

Well, how reliable is demeanour as an indicator of truthfulness?

The literature on the topic falls into three categories.

Firstly, there are articles by judges and lawyers largely relying upon experience and anecdotal evidence. Some of these articles recognise limitations in assessing credibility from demeanour.

Secondly, there are studies conducted by psychiatrists and psychologists calculated to test, in controlled conditions, whether subjects can accurately assess the truthfulness of a group of witnesses some of whom are known by the researcher to be telling the truth while the others are known to be telling lies.

Thirdly, there are studies by medical academics of the way the mind works with particular reference to memorising and recalling events. These researchers are concerned to test the correctness of the popular conception that, generally speaking, recollection is a process which will retrieve intact, subject to detail being forgotten, a subject's original perception of an event or conversation.

First, the Lawyers' articles.

In Australia, in the first category, there is an interesting article written as long ago as 1983 by Loretta Re of the Australian Law Reform Commission called 'Oral Evidence v Written Evidence: The Myth of the "Impressive Witness"'. (7) Ms Re perceptively exposed many of the problems of assessing veracity from demeanour. The article includes references to the scientific literature at the time. Her conclusion was that demeanour is not a good indicator of honesty and that non-verbal behaviour such as 'body language' was also not usually helpful. It is worth noticing, however, that Ms Re was dealing with demeanour as a means of positively finding truthfulness as opposed to its use in exposing a witness who is not telling the truth.

Justice Michael Kirby of the High Court of Australia has also written on this topic from the practising lawyers' perspective. (8) He recognised that serious doubt had been raised as to the capacity of a judge 'to tell truth from falsehood from the observations of a witness giving testimony in the artificial and stressful circumstances of a court room.' (9) He has also addressed the theme in a judicial context. (10)

Other judges have written less critically about the process, accepting that it has substantial validity, particularly when the issue is whether a witness is lying. An example is a 1996 article by Justice Giles of the New South Wales Court of Appeal. (11)

Secondly, the Psychological Testing.

There have been numerous papers published in psychology journals in the last few years relating to the accuracy with which truthfulness can be discovered by demeanour. They include an analysis of a passage of evidence when President Clinton was lying to a Grand Jury in August 1998. (12) His demeanour was compared with his demeanour when speaking truthfully before the Grand Jury and in a fund raising speech to a sympathetic audience. The researchers identified 23 potential signs associated with lying and found that President Clinton markedly increased the frequency of 20 of these signs when he was not telling the truth. (13)

The Clinton study is a study in hindsight. Most of the recent articles on the subject simply test whether subjects can detect lying in others. The consensus of the papers that I have read is that lying can be detected in only a little above 50% of cases in these study conditions. (14) Timothy Levin, one of the researchers, found that the detection of lies was often significantly below chance. (15)

I am reminded of the story of the Judge who, suffering more than a 50% rate of reversal on appeals, introduced the practice of deciding cases by tossing a coin. Interestingly, and contrary to popular perception, Timothy Levin suggests that truthfulness can be detected with greater accuracy than the telling of lies.

A number of researchers have looked at consistency as a predictor of truthfulness and found this not to be so. Swedish researchers found that people tend to disagree as to whether statements are inconsistent, but also that deceptive statements are as consistent as truthful statements. (16) The explanation of this seems to be that liars try to repeat an initial statement whereas truthful witnesses try to reconstruct an event afresh on each occasion that they give evidence about it. (17)

Thirdly, the Medical Research.

The most interesting work in this area is the neurological material relating to the way in which the brain and memory works. Professor John Kihlstrom from the University of California (Berkeley) is one of the world's leading experts in the field of memory. Professor Phillip Resnick of the Case Western University School of Medicine in Cleveland is a leading researcher on the subject. Dr Peter Shea of the University of Sydney has also written widely in the field. I am indebted to the work of each of them for what I am about to say.

Memories are not recording systems which, though some detail may be lost over time, reproduce what was originally laid down. Memories are always reconstructed, and, in the reconstruction process they are changed.

When memories begin to fade the mind compensates by further reconstruction. Reconstruction can be a conscious process, but it is also a subconscious process. The memory fills in gaps, sometimes from irrelevant sources. Because normal memory is reconstructed, it fluctuates.

In the result it is normal for a person recounting an event to make inconsistent statements relating to it. This supports the psychologist's findings to which I have referred.

According to Professor Kihlstrom: 'Memory is not so much like reading a book as it is like writing one from fragmentary notes.'

One interesting matter is eye contact. The researchers say that looking the listener in the eye is no more a sign of truthfulness than looking at the floor is a sign of lying. There is material to suggest that we can concentrate more effectively when we are not distracted by processing visual information.

So what is the relevance of all of this to Arbitrators? It should encourage those of us who think that the documents are the best source of information to feel more comfortable. It should persuade those of us who think that looking at the colour of the eyes of a witness is the only way, to think again. It is, however, no more correct to say that oral evidence is of no assistance in discovering the truth than it is to say that documents are of little help. The reality is that both are useful tools for discovering the facts.

A real problem is, however, that oral evidence is much more time consuming and costly to gather than documents. I wonder whether due process does require so much time to be spent in Arbitrations on lengthy examinations in chief and cross-examination. But that is another topic.

Next time the common lawyers here are on a panel with civil lawyers, they might think it worth experimenting by reducing oral evidence and, for that matter, when the civil lawyers here are on a panel with common lawyers they might think it appropriate to be a bit more generous in accepting oral evidence.


  1. Penultimate paragraph of judgment delivered in Byrnes v Treloar & Ors (unreported, Gleeson CJ, Powell JA and Stein JA, 10 December 1997).

  2. Abalos v Australian Postal Commission (1990) 171 CLR 167 at p179 per McHugh J.

  3. Ibid, at p178, quoting Lord Sumner in S.S. Hontestroom v S.S. Sagaporack [1927] AC 37 at 47.

  4. Australian Securities and Investments Commission v Whitlam [2002] NSWSC 591 at par [67].

  5. Ibid, at par [70].

  6. Ibid, at par [98].

  7. (1983) 57 ALJ 679.

  8. The Hon Justice Kirby AC CMG, 'Judging: Reflection on the Moment of Decision', TJR, vol. 4, 1999, p. 189 at p. 193.

  9. Ibid, at pp. 193-194.

  10. State Rail Authority of New South Wales v Earthline Constructions Pty Ltd (in liq) (1999) 160 ALR 588.

  11. The Hon Justice Giles, 'The Assessment of Reliability and Credibility' TJR, vol. 2, 1996, p. 281.

  12. Alan R. Hirsch and Charles J. Wolf, 'Practical Methods for Detecting Mendacity: A Case Study', Journal of the Academy of Psychiatry and the Law, vol. 29(4), 2001, p. 438.

  13. Ibid.

  14. Timothy R Levin, Hee Sun Park and Steven A. McCornack, 'Accuracy in Detecting Truth and Lies: Documenting the "Veracity Effect"', Communication Monographs, vol. 66(2), 1999, p. 125; Chih-Chen Lee, 'The Effect of Levels of Probing on the Detection of Deception in Audit Oral Evidence', Dissertation Abstracts International, vol. 61(8-A), 2001, p. 3243; and Paer Anders Granhag and Leif A. Stroemwall, 'Deception Detection Based on Repeated Interrogations', Legal and Criminal Psychology, vol. 6(1), 2001, p. 85-101.

  15. Timothy R Levin, Hee Sun Park and Steven A. McCornack, 'Accuracy in Detecting Truth and Lies: Documenting the "Veracity Effect"', Communication Monographs, vol. 66(2), 1999, p. 125.

  16. Paer Anders Granhag and Leif A. Stroemwall, 'Repeated Interrogations: Stretching the Deception Detection Paradim', Expert Evidence, vol. 7(3), 1999, p. 163.

  17. Ibid.