T he assessment of credibility is an essential aspect of fact-finding in judicial decision-making. Fundamentally, credibility involves an assessment of whether a witness is giving accurate, truthful and reliable evidence. A party’s case may rest entirely on whether a witness is believed and in cases with limited documentary evidence, the court may have little more to go on than the veracity of a witness’s oral evidence to determine the claim.
However, it turns out that our capacity to detect whether someone is lying is actually quite limited.
Experimental research has shown that people’s ability to detect lies is no more accurate than chance, or no better than flipping a coin. This finding holds across people in all professions – psychologists, judges, recruiters and law enforcement personnel – with the notable and limited exception of secret service agents.
This growing body of research shows that while most people think they are good at detecting deceit they are not.
It also raises a question about how well we perform the task of determining whether someone is lying in a profession that assumes that seeing and hearing a witness in person is crucial to fair judicial decision making. This question has been brought into stark relief with the disruptions caused by COVID-19 and the need to conduct hearings online.
What is credibility?
The assessment of credibility is a nuanced and complicated task involving consideration of multifaceted elements of the process of a witness giving evidence. This includes an assessment of more than just whether the witness can be believed or not (either generally or in respect of a particular matter or event), but also a consideration of the witness’s demeanour and the cogency and consistency of their evidence having regard to the reliability of other available evidence including other witnesses’ accounts.
Assessing a witness’s credibility online
Prior to the onset of COVID-19, the circumstances in which courts granted leave in civil proceedings to allow witnesses to give evidence online was limited. Far fewer cases were determined entirely online.
In part this was due to limitations in the available technology, but it was also due to the reasonable and widely held assumption that it is better to see and hear a witness in person.
These views were articulated by Buchanan J in Campaign Master (UK) Ltd v Forty Two International Pty Ltd (No 3) (2009) 181 FCR 152 at :
I share the concerns expressed by Spender J in World Netscape and by Stone J in Dorajay about the limitation on the effectiveness of video link arrangements as a means of taking oral evidence. I am particularly troubled by the prospect (or possibility) that the cross-examination of an important witness might be rendered less effective by the limitations of video link technology or the absence of the witness from the courtroom. Although the days are gone when witnesses are expected to feel any sense of intimidation as an aid to telling the truth, there is no doubt in my mind that the requirement to give evidence on oath or affirmation in the (generally) solemn atmosphere of a courtroom in the presence of a judge, and to answer questions in cross-examination in the presence also of cross-examining counsel, has at least three potential benefits. It enhances the prospect that the witness will remain conscious of the nature and solemnity of the occasion and of his or her obligations. It affords the cross-examiner some reassurance that the gravity and immediacy of the moment, and of the supervising presence of the judge, are not lost on the witness and the cross-examination is not thereby rendered any less effective, to the possible prejudice of the cross-examining party. It provides the Court with a more satisfactory environment in which to assess the nature, quality and reliability of responses by a witness, both to questions and to the overall situation presented by the necessity to give evidence in court. To my mind there remains, even in the modern context, a certain 'chemistry' in oral interchanges in a courtroom, whether between a judge and counsel (or other representative) or between cross-examiner and witness. I would not wish too lightly to deprive a cross-examiner of that traditional forensic element in the exchange although, as the cases universally make clear, the Court must now, if asked to do so, balance the interests of a cross-examining party against claimed inconvenience both in individual cases and with respect to individual witnesses. Notwithstanding the increased availability and use of video link technology, in my view, a case must be made out for the use of video link evidence if it is opposed by an affected party...
The restrictions caused by COVID-19 have required the profession to adapt quickly to how we run cases in court. In consequence, views about the conduct of hearings online have in some respects shifted.
In Universal Music Publishing Pty Ltd v Palmer  FCA 1472, Katzmann J observed at  that 'Considerations which may have influenced courts in the past against taking evidence by video link 'take on a different complexion when in person evidence becomes a matter of practical impossibility for an indefinite time': ASIC v Wilson  FCA 873; 146 ACSR 149 at  (Jackson J).'
The question that arises for present purposes is whether the use of this technology affects our capacity to assess a witness’s credibility.
This question has been addressed in several cases. Again, in Universal Music v Palmer, Katzmann J said at :
This year I have presided over a number of cases in which witnesses have been cross-examined using the Microsoft Teams platform. While I was initially sceptical about the effectiveness of cross-examination in this way, my scepticism proved to be unjustified. It transpired that I had no difficulty assessing credit or demeanour. In fact, my experience was that changes in facial expressions, reactions, bodily movements and gestures are much easier to discern when the witnesses are pinned to the screen directly in front of me than is normally the case when witnesses give evidence from the witness box some distance from the bench. In this respect my experience is not unique.
Lee J made similar observations in ASIC v GetSwift Limited  FCA 504 at :
To the extent that demeanour does play an important role in assessing the evidence of witnesses, then my experience, particularly in the recent trial that I conducted, is that there is no diminution in being able to assess the difficulty witnesses were experiencing in answering questions, or their hesitations and idiosyncratic reactions when being confronted with questions or documents. Indeed, I would go further and say that at least in some respects, it was somewhat easier to observe a witness closely through the use of the technology than from a sometimes partly obscured and…distant witness box.
Despite increasing acceptance of the use of video conferencing technology, the concerns highlighted by Buchanan J above (and acknowledged by others, eg, Katzmann J in Universal Music v Palmer at ) remain apposite.
In Rooney v AGL Energy Limited (No 2)  FCA 942 at , Snaden J said that his experience of the current technology was 'slightly less positive.' His Honour observed that 'the available technology cannot fully replicate the court room environment that is so often central to an adversarial system of civil justice'.
His Honour emphasised that the sense of solemnity the court room imbues as well as the cadence and chemistry that develops between the witness and the bar may be inhibited (if not ‘prohibited’) by conducting hearings online. These are matters which can have a real bearing on how witnesses perform in the witness box, including during cross-examination.
As practitioners, we must keep these important matters in mind, so as to avoid the risk of placing undue weight on a witness’s demeanour when making assessments of credibility in a virtual environment. This is particularly so when regard is had to the body of research that suggests that the expected behaviours indicative of deceit (eg, gaze aversion) may not be associated with lying at all.
When it comes to detecting lies, it has long been believed that emotions about lying may be translated into visible or audible signs – those subtle physical and behavioural signs in body language or demeanour that, it is said, reveal deception. This theory led researchers to examine behaviours such as gaze aversion, voice inflexion, postural movements, face touching, and hand movements to detect deceit, with limited success.
...some of the behavioural signs most often linked to lying are among the worst predictors.
In fact, some of the behavioural signs most often linked to lying are among the worst predictors. Studies demonstrate that shifting eye gaze is more likely to be an indicator that a person is thinking, or more precisely, that he or she is accessing their long-term memory in order to properly address the question they have been asked.
Other research demonstrates that the behavioural indicators commonly associated with lying are simply the physical signs of raised emotion that may be triggered in unsual or stressful situations. For many people, attending court is the very kind of situation which is capable of arousing strong emotions.
Experiments also show marked individual differences among people in their capacity to conceal emotions. Some people are naturally vulnerable to heightened emotions arising from detection apprehension while others are able to lie with ease.
In fact, research indicates that there may be genes for lying. The so-called ‘natural’ liars exhibit limited apprehensions about being detected and therefore are likely to present as confident and reliable witnesses.
The assessment of credibility, of course, is not a science, nor can it be, as it is not always possible to test with precision the complex intermingling of impressions that emerge after observing witnesses and attempting to reconcile contestable versions of events. Cultural and behavioural differences among witness add to the complexity of the task.
However, the science does reveal that an over-reliance on so called behavioural indicators of deceit may result in error.
More reliable indicators of truth-telling include the way in which a person responds to questions they are asked. For example, answering questions vaguely, repeating questions before answering them, speaking in sentence fragments and failing to provide specific details when the person’s narrative is challenged are all associated with a propensity to conceal the truth.
These indicators combined with an assessment of the content, logic and coherence of a witness’s evidence, as well as whether it is supported, not supported, contradicted or qualified by any other evidence, provide the elements for a more objective and reliable account of whether a witness should be believed. These are methods the courts regularly and expertly deploy.
The assessment of credibility is a complicated and nuanced task, which is made more difficult by the challenges posed by COVID-19. However, having regard to the observations of various courts noted above and the emerging research in the field of deceit, it is safe to conclude that the courts and the profession have risen to meet these challenges.
1 Curci A, Lanciano T, Battista F, Guaragno S, Ribatti RM. 'Accuracy, Confidence, and Experiential Criteria for Lie Detection Through a Videotaped Interview', Frontiers in Psychiatry, 2019: 9:748.
2 Bond C, Bella M, 'Accuracy of Deception Judgments', Personality and Social Psychology Review, 2006: 10(3).
3 Wiseman R, Watt C, Ten Brinke L, Porter S, Couper SL, Rankin C. 'The eyes don’t have it: lie detection and Neuro-Linguistic Programming', PLoS ONE, 2012: 7(7).
4 Granhag P. A, Vrij A, Verschuere B (Eds.) (2015). Detecting deception: Current challenges and cognitive approaches. Hoboken NJ: Wiley-Blackwell.
5 Shen Q, Teo M, Winter E, Hart E, Chew S, Ebstein R. To Cheat or Not To Cheat: Tryptophan Hydroxylase 2 SNP Variants Contribute to Dishonest Behavior. Front Behav Neurosci., 2016: 10:82.
6 Loy JE, Rohde H, Corley M. 'Cues to Lying May be Deceptive: Speaker and Listener Behaviour in an Interactive Game of Deception', J Cogn, 2018: 1(1): 42.
7 Geiselman, R, Elmgren, S, Green, C, Rystad I. 'Training laypersons to detect deception in oral narratives and exchanges', Am J Forensic Psychology, 2011: 32:1-22.