11th Greek Australian Legal and Medical Conference
Crete, Greece 2007

Lie Detection in Litigation: Science or Prejudice?

Danielle Andrewartha

Introduction

Although people may prefer not to acknowledge it, lying is a normal, natural, indispensable facet of daily life. Evidence suggests humans lie and are lied to, successfully, about one to two times a day or possibly much more. The majority of these deceptions involve outright, albeit mostly ‘harmless’ lies, often referred to as ‘white’ lies, that are helpful, indeed oftentimes necessary, to facilitate comfortable social interaction. Notwithstanding the frequency of everyday untruthful communications, when faced with the task of accurately identifying dishonesty in others, performance is typically poor. This is, to a significant degree, attributable to the incorrect assumptions about deceitful behaviours that are held to be true.

Despite popular misconception, people’s lies are not telegraphed via cues as blatant as Pinocchio’s telltale nose. In fact, a growing body of research demonstrates that the verbal and nonverbal features associated with deception are contrary to expectation and are typically undetectable or misinterpreted. In daily social interactions, this need not provoke concern. Think for a moment how difficult and confrontational everyone’s daily lives would be if every truthful thought were uttered, or every untruth palpable. People would find themselves far less willing to interact with one another for fear of hurtful and confrontational exchanges. Just as alcohol is a social lubricant, deception is a social prophylactic; it shields people from unpalatable and potentially damaging truths and interpersonal confrontations.

Although a degree of undetected duplicity is beneficial in everyday interactions, when it comes to the investigation and prosecution of criminal transgressions, all but the guilty yearn for something as conspicuous and incriminating as a nose that grows when an untruth is uttered by its bearer. As a result, concerted effort is being exerted to improve knowledge of the nature of deception; its evolutionary origins, common manifestations, and, most importantly, its physiological and neurological cues. This effort will aid in developing mechanisms for dependable lie detection measures for use in the criminal justice system.

Lies and deception 1

Vrij defines deception 2 as “a successful or unsuccessful deliberate attempt, without forewarning, to create in another a belief which the communicator considers to be untrue” [italics removed]. 3

There exists a hierarchical order of lies. Outright lies are complete fabrications, in which the truth is wholly disregarded and new information generated and presented as the truth. Interestingly, these are the most prevalent of all lies. Exaggerations involve lies that overstate or downplay the truth. Subtle lying is the intentional misuse of literal truths or the omission of pertinent facts to create a false impression. It is important to recognise that this hierarchical order relates to the nature of the lie in its relationship to the truth, not its degree of severity.

Lies and deception in litigation

As lying is a common, integral part of everyday life, it is obvious that the criminal justice system is not immune from its occurrence. Indeed, with oftentimes high stakes and emotionally charged situations, the likelihood of attempts at deception would presumably tend to be greater than in non-litigious daily interactions. This is not to suggest that all who come before the courts are blatant liars, driven by self interest. It may be that the vast majority of persons in the criminal justice system who act dishonesty are those who engage in ‘exaggeration’ or ‘subtle lying’ as defined herein.

Witness testimony, including that of experts, is inherently fallible. Many studies have cast doubt on the reliability of witness memory, suggestibility, and perception. Therefore, even witnesses who are attempting to be completely honest and impartial may only present aspects of the truth, while also omitting pertinent details and/or replacing them with erroneous information. This problem is confounded by personal investment in the case. The litigants are generally the most invested, with their families and friends a close second. Independent eye witnesses and expert witnesses are theoretically detached, but in practical terms there is at least some degree of personal interest in the outcome of the litigation.

Arguably, as a result of this personal investment, Loftus and Doyle have noted that it is “common for a witness’ thoughts to bend in a direction that would be self-advantageous”.4 This may be reflected in witness’ testimony. Generally, this entails no more than a natural, unconscious, and, mostly, inconsequential bias exhibited by every individual when asked to comment upon something of personal relevance. That is, individuals have a typical tendency to describe an event with a particular slant or emphasis based upon their perception of and involvement in or relationship to the event and/or the parties concerned. At the other end of the spectrum, parties and witnesses in the legal arena may engage in deliberate attempts at deception. They may discard the truth completely or substantially, replacing it with fabricated information. At its most extreme, this form of dishonesty may manifest as false confessions, denials, or accusations, false alibis, and testimony and charges founded upon planted or otherwise false forensic evidence.

It is readily apparent that the inclusion of inaccurate and untrue evidence may have myriad deleterious ramifications. The acceptance of fallacious evidence may lead to wrongful convictions or acquittals, which would certainly affect those intimately involved in the given case and might also have a wider reaching impact.

Deception prevention in litigation

Due to the serious detrimental consequences of deception in criminal litigation, some safeguards have been implemented to try to eliminate or at least reduce the occurrence. As a preliminary filtering mechanism, police officers interview the complainant, interrogate the suspect/s, and evaluate the available evidence. Should the investigating officers believe there are sufficient grounds to support the accusations levelled against the suspect, the relevant office of the Director of Public Prosecutions, and the defendant’s lawyer carefully scrutinise the details of the case.

As a further safeguard against dishonesty in litigation, there are criminal sanctions in place to punish those found to have engaged in efforts to mislead the court. Primarily, these involve contempt and perjury proceedings.

In terms of expert witnesses, there are also increasingly stringent guidelines being implemented, including the introduction of codes of conduct that must be acknowledged and adhered to by the expert.

Lie detection - the four stage model

Broadly speaking, deception detection comprises a four stage process:

  1. Observation - watching and listening to the deceiver in a particular context.
  2. Identification - the ongoing process of identifying and noting certain aspects of the deceiver’s presentation that pique the detector’s interest.
  3. Interpretation – the process of evaluating the deceiver’s presentation to gauge intent, affect, personality, mood, and so on.
  4. Determination – the making of a definitive assessment of the deceiver and his or her presentation: honest or deceitful; truthful or untruthful.

The interpretation phase is the most complex, variable, and individualised of the four stages, due in part to the identification strategy adopted. Interpretation might occur in an ‘isolated’ or ‘global’ framework. When ‘isolated’, the detector is making his or her judgment based upon the one experience, without prior, independent observation or interaction, and in the absence of relevant, independent knowledge. In the ‘global’ setting, the detector has previously encountered the deceiver and knows him or her in some way, and/or possesses information that is directly relevant to the factual assessment of credibility.

Interpretation is complicated and generally hampered by ‘deception’ and ‘generic’ mental ‘scripts’. ‘Deception scripts’ comprise, often erroneous, assumptions about how liars behave and who they are. ‘Generic scripts’ encompass stereotypes, biases, and general assumptions, formed during a lifetime’s experience. They include racial stereotypes, personal biases, global biases, myths and urban legends, and common knowledge inferences.

Lie detection in litigation

Attempts at lie detection can be traced for centuries. The early Egyptians, Chinese, Arabs, and English had, by today’s standards, rather archaic and oftentimes barbaric methods of resolving disputes and determining witness credibility. These included erroneous beliefs about nonverbal deceptive behaviour, trial by battle and by ordeal, and the use of torture, truth serums, and phrenology. From early Roman times, these pseudo-scientific methods were gradually abandoned in favour of the current criminal justice model founded upon the assessment of oral witness testimony, supported by circumstantial, documentary, and forensic evidence.

There is currently a multitude of lie detection strategies being used, researched, and developed for application within the criminal justice system. Broadly speaking, these can be divided into ‘unaided’ and ‘aided’ lie detection methods. ‘Unaided lie detection’ refers to the various strategies employed by humans as ‘lie detectors’ in their own right. ‘Aided lie detection’ concerns what is more popularly thought of when discussing deception detection, namely the use of specifically designed lie detector apparatus. These two types of lie detection methods will be discussed in turn.

Unaided lie detection - humans as lie detectors

DePaulo, Zuckerman, and Rosenthal point out:

“to most scholars of deception, a lie detector is a machine…To a much smaller number of scientists, a lie detector is a person, attempting to discern dissimulation with the help of perhaps a hunch, some special sensitivity, or a few lay theories, but without any access to brain waves, heart beats, or breathing rates.”5

Larson notes that, although some may, through their words alone, telegraph their dishonesty, when people “do not confess with their mouths, then they confess with their body.”6 In other words, people assume that when someone lies, their discourse and/or demeanour, or nonverbal displays, give them away. As a result, it is confidently assumed that laypersons can accurately detect deception by utilising ‘common sense’. Assessing a person’s credibility in this manner principally entails the observation, identification, and interpretation of certain verbal and nonverbal cues that go to the overall impression of the suspected deceiver and the relevance, value, and veracity of what is being said. Despite laypersons’ confidence in their abilities, many studies have suggested that humans are poor lie detectors, with consistent findings that unaided human deception detection accuracy rates are approximately equivalent to chance levels. There are three main explanations for these findings.

First, many individuals may simply be poor ‘mind and body language readers’, lacking the innate ability to interpret accurately the visual and auditory cues belying the truth, or otherwise, of what is being said. This is complicated by the myriad misconceptions about the supposed telltale signs of dishonesty. These include beliefs that liars: cannot maintain eye contact; fidget excessively; blush; sweat; shield their faces; appear nervous, anxious, or suspicious; smile inappropriately; speak without fluency, plausibility or detail, and with long pauses; and find it difficult to sit still. Interestingly, research suggests these traits are generally not cues to deception, with liars actually performing in exactly the opposite manner to what is expected; they fidget less, remain still, maintain eye contact, and retain a pleasant, ‘normal’ demeanour.

Secondly, individuals may be swayed by personal biases, be they unconscious or wilful. People often make fast and automatic dispositional inferences about another’s behaviour. They do so in reliance upon the information readily available, their existing deception and generic scripts, and a comparative assessment, or ‘analogy’, with themselves and those most familiar to them. As a result, people fail to appreciate the impact of situational factors and adjust their conclusions accordingly. Once these rapid, problematic conclusions about others’ veracity have been drawn, the situation is confounded because they are very resistant to change. As a result, preconceived notions of how liars behave, and individual biases tending towards veracity assessments based upon entirely arbitrary features such as familiarity, appearance, likeability, confidence, or authority, detrimentally influence the objective decision making process.

Finally, people may also be at the mercy of numerous societal and cultural stereotypes. These may exist as a preconception as to the credibility of a given class of witness, or may arise in response to a particular manner of giving evidence that is perceived as incongruent with the detector’s own or the detector’s expectation. For example, contrary to what might be considered ‘typical’ Western demeanour, Giles has observed that “Aboriginal speech habits involve silences, indirect answers and negative answers which might wrongly be understood as evasion, confusion or guilt, and…Aboriginal culture promotes gratuitous concurrence.” 7 Asians demonstrate similar cultural courtesies. They have a tendency to agree to various propositions without genuinely accepting or understanding the query being put, make little or no direct eye contact, and engage in embarrassed laughter. These traits are exacerbated in the examination and cross examination context, particularly during questioning about sensitive issues. Language discrepancies also play a pivotal role. Additionally, there are myriad erroneous and often offensive stereotypes associated with certain races, cultures, genders, and socio-economic classes that affect unbiased decision-making. Unfortunately, studies have suggested that, even when people know on a logical and rational level that sweeping stereotypes are patently incorrect and damaging, their determinations are still influenced by them.

Due to these findings, various researchers have been studying unaided human lie detection capabilities to gain a greater understanding of the underlying methodology, identify inherent strengths and weaknesses, and thereby enhance unaided lie detection techniques in more precise and quantifiable ways. Examples include research concerning the motivating psychophysiological causes for the observable deceptive behaviours, cross modal inconsistencies, signs of emotional ‘leakage’, microexpressions, lie detection ‘wizards’, and alternative heuristic approaches.

Aided lie detection – ‘lie detector machines’

As a result of the deficiencies inherent in unaided human lie detection strategies, considerable research is being undertaken to develop specialised apparatus to be used for the purposes of lie detection in criminal investigations and prosecutions. These aided methods include two pieces of equipment that have already been put to use in this capacity: the polygraph and brain fingerprinting apparatus, in addition to voice stress analysis, functional magnetic resonance imaging (fMRI), and thermal imaging and pupillography.

These devices supposedly enable greater accuracy than traditional unaided human lie detection methods. However, they are founded upon the somewhat questionable assumption that there exist identifiable and measurable signs of deceit, which underlies the unaided lie detection methodology. There have been varying accounts as to the reliability of these lie detection machines. Some studies place success rates at no higher than chance levels. Others suggest much higher accuracy rates, typically ranging as high as 75-90 percent. In relation to brain fingerprinting, 100 percent accuracy rates have been asserted, though these claims are dubious and lacking independent peer support. The United States National Research Council (NRC) undertook a comprehensive review of the accuracy and validity of polygraphy and concluded, in summary, that “specific-incident polygraph tests can discriminate lying from truth telling at rates well above chance, though well below perfection.”8

Whatever the precise accuracy rates, it would seem that aided deception detection methods do not fare much better than unaided models. Principally, this may be due to the predominant role still played by humans in the strategies utilising external apparatus. Such involvement is pervasive, exhibited throughout the development and administration of these tests and the interpretation of the subsequent results. By reference to the four stage model of lie detection, comprising observation, identification, interpretation, and determination, it can be seen that lie detection methods employing lie detector machines vary from those that do not in respect of only one phase: identification. However, it must be noted that none of the lie detector apparatus actually detects deception, they merely detect the physiological and/or neurological changes that are considered by their developers to denote lying. As a result, even the interpretation stage is heavily dependent upon, perhaps questionable, scientific theories and human application. The continued significant degree of human participation brings with it the same biases, prejudices, and deficiencies as those associated with unaided detection techniques.

Confounding situational factors

McClellan has pointed out that the admissibility, relevancy, and weight of all manner of evidence, including witness testimony, are “based on the law’s expectation of the way innocent and guilty people will normally react”. 9 These arbitrary concepts are arguably no more infallible than that guiding other, non-litigious deception detection. In addition to this, and the various other shortcomings inherent in both aided and unaided lie detection measures, the criminal justice context adds a significant confounding factor. Neither aided nor unaided approaches adequately accommodate the situational influence likely to be exerted upon examinees in adversarial settings such as the interrogation room and courtroom.

Bar random conflicts, the majority of interpersonal communication is non-confrontational. Police interviewing and interrogation is, on the other hand, highly adversarial. The overall impression of the investigative phase is that it is coercive, oppressive, and intimidating. Similarly, court is quite confrontational, artificial and unfamiliar to most laypersons, particularly the incidence of cross-examination, which is almost entirely unrelated to the conversational styles with which most people are typically accustomed.

A person’s body language has been shown to feature substantially in everyday assessments of credibility, so what of the suspect’s demeanour? Should he or she be distrusted if he or she displays signs of anxiety? Is he or she to be believed if he or she appears at ease? Stone notes that:

“It would be an affront to common sense to conclude from [visible signs of anxiety] alone that the witness must be lying. It would be equally absurd to assume that witnesses who seem to be calm must be telling the truth.” 10

However, research into the cues relied upon in determining veracity suggest this is precisely what occurs. Signs of nervousness are taken to be a key indicator of deception, in contrast with signs of comfort that are believed to denote honesty.

Unfortunately, drawing these conclusions from witness demeanour alone fails to acknowledge the situational influence of litigation. Simon aptly encapsulates this potential difficulty, in the interrogation context, by stating that:

“[n]ervousness, fear, confusion, hostility, a story that changes or contradicts itself – all are signs that the man (sic) in an interrogation room is lying, particularly in the eyes of someone as naturally suspicious as a detective. Unfortunately, these are also signs of a human being in a state of high stress”. 11

It is patently obvious that the confrontational, ‘high stakes’ criminal justice setting, throughout the investigation and trial phases, is likely to have a bearing, potentially significantly so, on the suspect’s mental state. This is expected to be reflected in his or her manner of verbal and nonverbal communication, and to interfere with underlying physiological and neurological processes. Changes in these processes are the fundamental indicator of deception monitored and recorded by lie detector machines. It would seem that the implementation of lie detector apparatus does not adequately cater for the confounding influence of the adversarial context.

This perhaps explains why considerable judicial criticism has been made of the purported reliability of utilising lie detector machines in litigation. The propriety of equating simulated scientific testing with real life scenarios for the purpose of evidence is highly questionable. The judicial commentary echoes criticisms of lie detection methodology raised by researchers themselves, such as the NRC when it stated that:

“the consequences associated with lying or being judged deceptive [in a controlled laboratory study] almost never mirrors the seriousness of these actions in real-world settings in which the polygraph is used.” 12

Admissibility of lie detection evidence

It is generally accepted that the ultimate issue at trial – the accused’s guilt or innocence – is properly left for the jury to determine. Given the uncertainty of jurors’ ability to accurately differentiate between truthful and false testimony, and the possible adverse ramifications of this limitation, it is arguably desirable that means be developed to facilitate this important function. Relatively recently, increased attention has been directed at lie detector machines and their potential role in this capacity.

In Australia, the admissibility in criminal trials of expert testimony is governed by the various expert evidence rules articulated in legislation and applied in case law. These principally comprise: the ultimate issue rule; the common knowledge rule; the area of expertise rule; the expertise rule; and the discretionary more prejudicial than probative exclusion, though not all are applied in each state and territory or at the federal level.

Expert testimony concerning the various lie detection measures falls within the classification of ‘novel’ scientific evidence, in that it is relatively new, somewhat controversial, and, for the most part, of questionable scientific basis. In the United States, which has had far greater opportunity than Australia to consider the admissibility of lie detector evidence, the seminal case of Daubert v. Merrell Dow Pharmaceuticals Inc 13 presently expounds the four broad criteria by which cases concerning novel scientific evidence are to be decided. Frye v. United States 14 was the leading case until the introduction of the Federal Rules of Evidence 1975 (US), at which point Daubert became the preferred case for deciding the admissibility or otherwise of novel scientific evidence. The test in Frye is confined to only one of the four factors espoused in Daubert, providing that a novel scientific theory or technique may be declared admissible as evidence in a given case only if it is “sufficiently established to have gained general acceptance in the particular [scientific] field to which it belongs.” 15

Daubert has not yet been adopted as law in Australia, nor has there been cause thus far for the judiciary to afford it concerted attention. However, that has the potential to change. To date, the Australian jurisdictions of New South Wales, the Northern Territory, Queensland, and South Australia have applied the Frye approach. The remaining Australian states and territory have not yet decided the issue, bar Victoria, which has expressly rejected it, though curiously has not, at least thus far, embraced Daubert.

It would appear that the admissibility or otherwise of specific lie detector apparatus based evidence has only been considered thus far in relation to polygraphy and brain fingerprinting. As with expert evidence generally, there has been great divergence in the determinations made in Australia and abroad.

The only statute to specifically address the use and admissibility as evidence of polygraphy in Australia is the New South Wales Lie Detectors Act 1983, which prohibits any request for or use of lie detection measures, including polygraph examinations, in respect of any person in an employment, insurance, or legal context. It is essentially a blanket rejection of legalised deception detection using scientific apparatus. In all other states and territories, a case by case assessment is made, applying the Frye test, and/or principal evidentiary rules as applicable. To date, the Australian High Court has not had cause to consider whether testimony concerning polygraph examinations is admissible in Australia. However, lower courts have considered the issue, deeming such evidence to be inadmissible. At the United States’ federal level, polygraph results cannot be admitted into evidence as substantive proof of the defendant’s guilt or innocence, but might be relied upon for impeachment purposes. The various states have applied Frye and more recently Daubert to different ends. For example, some states allow polygraph examination evidence in criminal trials; some bar it outright; and others require the consent of all parties before it might be admitted. The majority approach in the United States is to deny the admissibility of polygraph evidence.

A novel approach adopted in at least one case, a New York civil suit that involved an alleged oral loan of a nominal amount, was to compel each party to the dispute to participate in a polygraph examination, with the results to form the binding judicial ruling. The judge considered it to be a suitable decision making device in the circumstances as it would save considerable time and expense not truly justified in view of the facts. Perhaps this might be a legitimate application of polygraphy, and other aided lie detection measures for that matter, in Australia, possibly as a starting point for wider application.

In America, brain fingerprinting has been applied, apparently successfully, in real life scenarios. The admissibility or otherwise of testimony concerning brain fingerprinting has not yet been an issue to be determined by Australian courts. Were that to occur in the near future, it is highly probable the outcome would be in line with the stance taken in respect of polygraphy; namely the inadmissibility of such evidence. It is difficult to conceive how the court might decide otherwise in the absence of sufficient, let alone compelling, scientific validation of and support for the reliability and validity of the technique.

In short, the door to the admissibility of polygraph evidence, and, for that matter, evidence concerning similar lie detection measures, remains ajar in Australia. However, it will take a compelling force for that door to open. Currently, the greatest obstacle to widespread acceptance and admissibility is the dearth of adequate scientific data underlying the supposed success of aided lie detection. This perhaps explains, at least in part, the trepidation with which the law approaches the admissibility of such evidence. As Happel points out: “it is difficult to have confidence in a device when it is not known why it works.” 16

Assuming there continues to be increasing scientific assessment, supporting research, and, ultimately, general acceptance of these practices within the scientific community, the law will face greater demand to respond and adapt, with admissibility of this evidence the end point. Judicial commentary thus far indicates the courts’ reluctance to follow this course, with marked change in judicial thinking necessary before such an outcome might be observed. Further, even with scientific and judicial acceptance, legislative and procedural safeguards would need to be implemented to ensure the consistent nature and scope of permissible use of and reliance upon such evidence. This process will be slow, if it indeed occurs at all.

Lie detection in litigation - science or prejudice?

As discussed above, human beings have developed generic and deception scripts for use in daily interpersonal interactions to interpret demeanour and assess veracity. As McClellan observes, so too has “[t]he law…devised its own rules of human behaviour and created its own norms for interpreting that behaviour…often classified under the rubric of “common sense”.” 17 At present, this is all that guides juries in their quintessential role as triers of fact. For reasons explored herein, this is an imperfect means of accurate and reliable credibility assessment and decision-making in litigation.

Unaided lie detection is fraught with difficulty. It is predicated upon a multitude of misconceptions about how liars behave and who they are, including specific verbal and nonverbal cues believed to indicate dishonesty. Unfortunately, these cues are misleading, non-apparent, and misinterpreted. These problems are exacerbated by individual biases and social and cultural stereotypes etched into the detector’s internal ‘scripts’. They are further confounded as a result of the contextual impact litigation has upon people’s otherwise normal interpersonal communications. It is for these reasons that studies have suggested the accuracy of unaided lie detection is quite low.

Aided deception detection, though arguably more reliable than unaided methods, still has significant shortcomings. Most notably, there is a lack of adequate peer reviewed scientific data to support the claims made as to the underlying methodology, application, results, and supposed success rates of lie detector machines. Further, aided lie detection is still predominantly reliant on human involvement. As a result, the same biases, prejudices, and deficiencies as those associated with unaided detection techniques may also impinge upon the impartiality and effectiveness of aided approaches.

This conclusion need not necessarily cause undue concern. First, lie detector machines afford greater accuracy than unaided detection techniques. The most conservative assessment of the reliability of this equipment places it at a level less than perfection, but greater than chance. This suggests that aided deception detection measures are at least more reliable than unaided ones. As a result, one might query why the judiciary is so eager to cling to the traditional model of jury credibility and decision making devoid of external, machine based and expert guided assistance. Secondly, given the ‘head start’ of greater accuracy, but the impediment of the human fallibility weighing it down, improved research into lie detection cues, and commonly held misconceptions, biases, and prejudices, might help eliminate the ‘human error’ component inherent in both aided and unaided lie detection efforts.

It would be naïve, perhaps even foolish, to entertain the suggestion that science might ever do away with well established and ingrained societal stereotypes and prejudices for the purposes of lie detection in litigation. This is because the task at hand is so dependent upon interpersonal behaviour and communication, which are such variable, imprecise concepts. However, current determinations of guilt or innocence, truth or untruth, honest or dishonest, currently rest upon 12 laypersons who bring with them a lifetime of personal and societal biases and prejudices. As these decisions can have consequences as serious as the life or death of those involved, arguably any degree to which ‘human error’ can be minimised by way of scientific intervention should be encouraged.

Although the foundation for both aided and unaided lie detection efforts is broadly scientific, there remain many outstanding queries and unsubstantiated conclusions and correlations. Further, neither aided nor unaided methods are resistant to the deleterious effects of personal and global stereotypes and prejudices. As such, it is easy to understand Clarke’s postulation as to whether:

“… the coercion of people to undertake lie detector tests varies significantly from subjecting hapless women in the dark ages to trial by ordeal? Do these devices any more accurately determined (sic) truth? Or have we simply invented a modern form of witch dunking?” 18

On careful reflection, it would appear that current deception detection strategies sit somewhere along the broad spectrum between science and prejudice. For litigious purposes, the goal is to push the scales in favour of the former and find means of eliminating or at least reducing the latter.

For further information, please contact the author, Ms Danielle Andrewartha, at: dmand2@student.monash.edu.au.

Endnotes

1. For the purposes of this paper, the terms ‘liar’ and ‘deceiver’ will be used interchangeably to denote a person who is attempting to lie to another, whether or not successful in that attempt. Those charged with the task of determining whether or not they are being lied to will be referred to as ‘lie detectors’ or merely ‘detectors’. The liar’s verbal and nonverbal efforts to deceive will be termed his/her ‘presentation’.

2. Although I consider ‘lying’ and ‘deception’ to be two independent concepts, for the purposes of this paper these terms will be used interchangeably.

3. Vrij, A. (2000). Detecting lies and deceit: The psychology of lying and the implications for professional practice London, Wiley, p 6.

4. Loftus, E., & Doyle, J. (1987). Eyewitness Testimony: Civil and Criminal. New York, Kluwer.

5. DePaulo, B. M., Zuckerman, M., & Rosenthal, R. (1980). “Strategies of Deception: Humans as Lie Detectors.” Journal of Communication(Spring 1980): 129-139.

6. Larson, J. A. (1969). Lying and its Detection: A Study of Deception and Deception Tests. New Jersey, Patterson Smith.

7. Giles, J. (1996). “ The Assessment of Reliability and Credibility “ TJR 2: 281., p 292.

8. National Research Council, The Polygraph and Lie Detection. Washington DC, National Acadamies Press, 2003. p 4.

9. McClellan, P. (2006). “Who is telling the truth? Psychology, common sense and the law.” Australian Law Journal 80: 655-666, p 657.

10. Stone, M. (1991). “Instant Lie Detection? Demeanour and Credibility in Criminal Trials.” The Criminal Law Review: 821-830., p 826.

11. Simon, D. (1991). Homicide: A year on the killing streets New York, Ivy Books., p 219.

12. National Research Council, The Polygraph and Lie Detection. Washington DC, National Acadamies Press, 2003, p. 3.

13. 509 US 579 (1993).

14. 293 F 1013 (1923).

15. Frye v. United States, 293 F 1013 (1923), at 1014.

16. Happel, M. D. (2005). “Neuroscience and the Detection of Deception.” Review of Policy Research 22(5): 667 , p 673.

17. McClellan, P. (2006). “Who is telling the truth? Psychology, common sense and the law.” Australian Law Journal 80: 655-666, p 657.

18. Clarke, B. (2000). “Trial By Ordeal? Polygraph Testing In Australia.” Murdoch University Electronic Journal of Law 7(1). from

http://www.murdoch.edu.au/elaw/issues/v7n1/clarke71nf.html, p 5.