Blind Faith in Visual Evidence: lie detection, fMRIs and the shift away from oral evidence

Source: Needpix.com

Eva Monteiro
B.C.L./J.D. III
Faculty of Law, McGill University

15 December 2020

The search for truth

The common law legal tradition has historically been one which establishes truth through the fact-finding and oral advocacy charge of its lawyers. Lawyers plead their cases in an adversarial process, judges probe, plaintiffs, defendants and witnesses give testimony. When appropriate, experts are heard so as to gain insight into more technical aspects of the case. As David Howes observes in “Introduction to Sensori-Legal Studies”:

“The acoustic ordering of the courtroom is [highly] pronounced due to the central role accorded to speech: ‘there is the oral reading of the relevant written material, the oral arguments, the oral exchanges between the court and the lawyers, the oral evidence at the trial, the oral judgment of the court’ [quoting J.I.H. Jacob, The Fabric of English Civil Law]. Meanwhile, the court is called to order and dismissed by the resounding rap of the judge’s gavel.”

All very “otocentric” elements. But there is more, Howes writes:

“Consider the traditional image of Justice: a blindfolded female figure […]. The blindfold is supposed to guard against bias by removing the possibility of visual markers of social status exerting undue influence […]. The fact that Justice’s ears are uncovered [by contrast] in turn implies that justice involves ‘hearing both sides’ of an argument (audi alteram partem).”

Recently however, it is visual information that has gained increased prominence, and the former orality of justice has gone into retreat. There has been a progression from “otocentrism” to “ocularcentrism.” The pre-eminence of the visual over the oral-aural is apparent, and includes, among other things: the introduction of pictures, flowcharts, videos, as well as reconstructions of events with the help of animations into evidence. Furthermore, lawyers can be seen increasingly turning to emerging neuroscience for insights into how best to frame their arguments, with a view to presenting the most compelling combination of sensorial stimuli. One specific visual tool which has increasingly come to occupy the attention of the court during trials is brain scans. Constituents of the justice system have demonstrated a blind faith in this technology which they view as a powerful aid to the truth-seeking process, even though it is one which they generally do not fully understand. Of particular note is the way in which brain scans have started becoming substitutes for information which was traditionally ascertained orally, such as “lie-detection.”

This probe will explore the value that brain scans, and visual evidence more generally, are assumed to offer, as well as what such visual datafication can take away from the truth-finding mission. The law has always concerned itself with developing methods that will best ascertain the truth. This ranges from witnesses being obliged to swear an oath before testifying, to designating jurors as “truth-seekers”, instructing them to evaluate the credibility of those who bear witness as objectively as possible. We have now also started relying on science as a means to ensure the truthfulness of the trial’s participants.

The emergence of visual evidence

Visual representations are ubiquitous in our society, from the widespread use of x-rays for diagnostic purposes in medicine, to the use of imagery of “the good life” in advertising and marketing which effectively seduces customers, to the artful use of visual images to generate TikTok and Instagram stardom. The legal world is not exempt from this trend, as this visual culture has become omnipresent in courtroom practices, too. Consider how photographs and video images have proliferated as a source of evidence in court, as a reflection of the fact that people have (conveniently) gotten in the habit of documenting their lives with their smartphones. These same individuals are also being minutely tracked by the discrete yet pervasive use of video surveillance, another staple of the modern courtroom. The sense of familiarity we have with these images causes us to look upon such information favourably, as a direct reflection of reality.

It has been found that visual content amplifies emotions and beliefs, and, significantly, that it can usually be grasped effortlessly all at once (unlike speech: it takes time to explain or read the expression of an idea.) Visual information can, furthermore, be manipulated through the use of digital graphics and animations to instantly conjure the reality we would like to see before us. Richard Sherwin has described the ways in which lawyers capitalize on this knowledge, continuously “min[ing] the popular imagination for well-known characters… and stock scripts to help frame their story in court.” They use saturation of detail and hyper-ornamentation to convince, whether it be in the form of shimmering fMRIs or digital simulations of murders. They create narratives which are readily visualized and brought to life and which entrain embodied belief systems. All of this was already possible with words and oral argument, but these storylines have become much more powerful and compelling on screen than they ever were in print.

Neuroscience and brain scans form a major part of this trend. It was already known that introducing a picture into trial could have the effect of increasing conviction rates from 8% up to 38%, whereas the mere mention of neuroscience enhanced the credibility of claims made. It is clear that the combination of the two was likely to create a powerful tool; and without doubt it has. Functional Magnetic Resonance Images (fMRIs), specifically, have been increasingly relied upon in the courtrooms of many jurisdictions. They have been used in personal injury cases, criminal cases, to mitigate sentencing and have also shown potential for lie-detection. However, as intimated above, not all jurisdictions are transitioning towards a greater reliance on brain scans as eagerly. In Canada for example, neuroimaging is the least used form of evidence with only a 15% usage rate, and fMRIs are the least used of these techniques due to their prohibitive cost.

fMRIs and lie-detection

an fMRI scan during a working memory task, Source: Wikimedia Commons

Whereas the MRI can only detail the structure of the brain, fMRIs work by tracking the blood oxygenation levels in the brain, a proxy for blood flow. This enables the tracing of the brain’s activity and activation patterns. Tracking this information makes it possible to identify which brain regions are receiving more oxygen (and therefore blood) which, when released, becomes paramagnetic and triggers a magnetic field inside the MRI machine. The neuroscientist then uses this information to correlate the neural pathways and mechanisms affiliated with pain, memory, reward, or deceit circuits (among others.) While these may look like pictures and may often be thought of as analogous to x-rays, brain scans are rather “statistical maps, visualizations of comparative data sets, calculating variations from the norm of the magnetic resonance of water molecules within localized blood flow to the brain.” Rarely, however, is proper account taken of the fact that these images are reconstructed by humans based on statistical analyses and a series of inferential steps that do not directly capture reality. Although these technologies can attempt to track what a person is experiencing or thinking, the images they create are actually a product of tremendously complex sets of inferences which cannot actually prove causation.

When it comes to lie detection, fMRIs are used in an attempt to directly capture the brain in the act of deception. This is unlike polygraph tests which measure anxiety-induced changes in blood pressure, pulse, breathing rates and sweat. Neuroscientists rely on meta-analyses which identify certain areas of the brain as being correlated with truth-telling and lying. In the early 2000s, companies like Cephos and No Lie MRI  were claiming accuracy rates in the 90th percentile, but the fling they enjoyed with the United States’ justice system proved to be short-lived, and both now seem defunct. This was partly due to the realization that there existed many ways of “tricking” the system and that the people the technology was being tested on (who were cooperative and willing participants in the experiment) were not the type of people who would be scanned during a trial. While these companies grew, more and more literature was being published which condemned these technologies as being insufficiently developed or understood to even begin considering introducing them into courtrooms.

Blind faith

According to some commentators, this visual data is dangerous in that judges, lawyers and jurors are often simply not qualified to decipher or gauge the reliability of these technologies and the information they present; often, even the scientists themselves are unsure as to their reliability. Not only are courtroom constituents unlikely to correctly interpret these images, they also tend to disproportionally trust scientific evidence and innovations: dazzled by science they showcase blind faith, literally, in the visual data these technologies generate. In one study, for example, mock jurors were found to be more likely to give a “not guilty” verdict if the defense of mental disorder was accompanied by the presentation of MRI images.

We commonly suppose that while witnesses, plaintiffs and defendants may lie, science does not. Constituents of courtrooms accordingly feel as though by looking at an fMRI they are directly observing the brain in motion, in the same way they expect a picture to be a true reflection of reality. However, just like witness testimonies, fMRIs are sources of information which must be received and evaluated through a critical lens. This critical lens has not yet been developed when it comes to visual evidence, and more specifically scientific visual evidence.

When shown evidence of biomechanisms which are supposed to underlie mental illnesses or psychopathy by experts, judges are likely to reduce sentences and a higher percentage of judges are likely to list mitigating factors. Judges, who are often curious about the technology, are also likely to think that they will be able to appropriately weigh the scientific evidence presented. However, other more forthcoming judges have also acknowledged being baffled and overwhelmed by the flood of neuroscience evidence. They have become wary of this influx of information and its quality, and so, in response, courtrooms have developed tests in order to better assess it. In Canada, the test developed to establish the admissibility of expert evidence comes from R v Mohan. The test looks at four criteria: the evidence must be relevant, it must be needed to assist the trier of fact, it must not fall under any exclusionary rule and it must be given by a properly qualified expert. In the United States, where courts are less reticent about relying on fMRIs and neuroscience, many states use the Daubert standard. It asks: is the evidence sound? Has the scientific technique been tested and published in peer-reviewed literature? What is its error rate? Other state courts alternatively use the Frye standard, which requires proof that the scientific evidence is generally accepted in the relevant scientific community.

These are the beginnings of a larger trend which we must start to push for in courts; namely the ability to critically analyze and decipher visual and scientific information. Lawyers, judges and jurors need training in what Richard Sherwin refers to as visual literacy. Visual content is not necessarily an objective and direct depiction of reality: for example, a camera is always only ever capable of taking a shot from one point of view. Images and videos can be tampered with or edited to incite more emotional responses in viewers. Psychologists tell us that:

“Aesthetic delight in the image encourages a sense of understanding and acceptance, and that response […] is tantamount to belief.  Add aesthetic delight to the widespread belief in the authority of science and you have a potent formula producing belief [in what is true].”

Many disciplines today recognize that meaning depends on context and culture, and that “the truth” depends on how it is represented or framed. It is incumbent upon the judicial system to understand the ways in which visual content is created, altered and presented, and to respond with a method adequate to critically analyze this information, as well as its reception by jurors. As law and justice migrate to screens, visual benchmarks must be added to law’s rhetoric if its truth-finding mission is to be preserved.

Conclusion: back to the drawing board

Our positive cultural associations with visual content and technology have brought about an uncertain turn for the courtroom’s truth-seeking mandate. As these emerging technologies mature, we must remain committed to critically engaging with the material presented as evidence in courts, much as the court has always done with auditory information. The hearsay rule illustrates this, such that in order to be admitted, evidence must come directly from the declarant rather than through a third party. This exclusionary rule is partly motivated by the fact that the demeanour of the declarant (which is one possible gauge of truthfulness or credibility) cannot be observed by the courtroom constituents and, what is more, they cannot be cross-examined (another crucial measure of truthfulness.)

As regards the exposure of prevarication in particular, lie-detection through the use of fMRIs is but one example of a field with fascinating and promising breakthroughs, but which was unleashed into the judicial system prematurely. At this time, scientists and judges alike are calling for caution and for a return to basics. In United States v Scheffer, the majority opinion called for the barring of lie-detection tests, claiming that it would infringe on the jury’s role as the courtroom’s human “lie detector.” Others argue that scientific information and fMRIs should not and cannot replace human investigation of criminal decision-making processes. As a species we have been socially conditioned from a young age to “read” the minds of our peers, recognizing intentions, untruths or exaggeration. As a field, law has become very good at discerning the truth through its reliance on oral advocacy and the trial process. These abilities and skills which have been developed over long periods of our history should not be shrugged aside as soon as we are presented with impressive looking fMRI scans.

All links accessed on August 27th, 2020.

References

Adolphs, Ralph, “The Social Brain: Neural Basis of Social Knowledge” (2009) 60:1 Annu Rev Psychol 693.

Aspinwall, Lisa et al, “The Double-Edged Sword: Does Biomechanism Increase or Decrease Judges’ Sentencing of Psychopaths?” (2012) 337:6096 Science 846.

Bright, David & Goodman-Delahunty, Jane, “Gruesome Evidence and Emotion: Anger, Blame, and Jury Decision-Making” (2006) 30:2 Law Hum Behav 183.

Chen, Ingfei, “The court will now call its expert witness: the brain” (19 November 2009), online: Stanford Report < https://news.stanford.edu/news/2009/november16/greely-neurolaw-issues-111909.html>.

Gurley, Jessica & Marcus, David, “The effects of neuroimaging and brain injury on insanity defenses” (2008) 26:1 Behav Sci Law 85.

Hoffman, Janet, “Visual Advocacy: The Effective Use of Demonstrative Evidence at Trial” (2011) 30:1 Oregon State Bar Litigation Journal.

Howes, David, “Prologue: Introduction to Sensori-Legal Studies” (2019) 34:2 Can J Law Soc 173.

Jay, Martin, Downcast Eyes: The Denigration of Vision in Twentieth-Century French Thought, 1st Ed (Berkeley: University of California Press, 1993).

Kuersten, Andreas, “When a Picture is Not Worth a Thousand Words” (2016) 84 Geo Wash L Rev Arguendo 178.

Legal Information Institute, “Daubert Standard”, online: Cornell Law School <https://www.law.cornell.edu/wex/daubert_standard>.

Legal Information Institute, “Frye Standard”, online: Cornell Law School <https://www.law.cornell.edu/wex/frye_standard>.

Miller, Greg, “Truthiness? No Lie MRI Hits the Legal System” (17 March 2009), online: Science Magazine < https://www.sciencemag.org/news/2009/03/truthiness-no-lie-mri-hits-legal-system>.

Muça, Darena, “Recap of the 12th Annual Colloquium (Part 2)” (25 February 2020), online: McGill Journal of Law and Health < https://mjlh.mcgill.ca/2020/02/25/recap-of-the-12th-annual-colloquium-part-2/>.

 R v Mohan, [1994] 2 SCR 9.

Ricoeur, Paul, The Rule of Metaphor: multi-disciplinary studies of the creation of meaning in language (Toronto: University of Toronto Press, 1981).

Rosenthal, Hilary, “Scanning for Justice: Using Neuroscience to Create a More Inclusive Legal System” (2019) 50:3 Colum Hum Rts LR 290.

Rusconi, Elena & Mitchener-Nissen, Timothy, “Prospects of functional magnetic resonance imaging as lie detector” (2013) 24:7 Front Hum Neurosci 594.

Sherwin, Richard, “Visual Jurisprudence” (2012), online: < https://ssrn.com/abstract=2135801>.

 United States v Scheffer, 523 US 303 (1998).

University of Plymouth, “Brain imaging lie detector can be beaten with simple techniques” (2 May 2019), online: Neuroscience News < https://neurosciencenews.com/fmri-lie-detector-13012/>.

Weisberg, Deena et al, “The Seductive Allure of Neuroscience Explanations” (2008) 20:3 J Cogn Neurosci 470.