Criminal Liability & Mental Illness: A Clash of Minds, Media, and Parliament

Contributed by Loïc Welch

“In every society there are those who commit criminal acts because of mental illness. The criminal law must find a way to deal with these people fairly, while protecting the public against further harms. The task is not an easy one.” [Justice McLachlin in Winko v B.C.]

Several high-profile Canadian cases (e.g., Allan Schoenborn, Guy Turcotte, Vincent Li) brought much media coverage, societal attention, and community backlash to the Criminal Code’s section 16 liability regime—Not Criminally Responsible on account of Mental Disorder (NCR). The public outcry stemmed from a lack of understanding of mental health, criminal law procedures, and the NCR regime itself, culminating in an amendment in 2014. For an appreciation of these changes, it is important to gain a bit of historical perspective as to the development of the current regime.

The History of the NCR Defence

The NCR defence was introduced following the 1992 amendments to the Criminal Code and replaced the Not Guilty by Reasons of Insanity defence (NGRI), which absolved criminal responsibility. The accused would come under the authority of the State and could be indefinitely detained until otherwise decided by the lieutenant governor. In addition, the NGRI defence could be raised by either party and effectively could be used as a means of involuntary and indefinite imprisonment with no mandatory review imposed on the State. The NGRI regime put the safety of the public above the interests of the accused and offered no procedural protection for the mentally ill. Offenders could not appeal the decision and were left in psychiatric institutional detention to the deference of the lieutenant governor – the only official endowed with power to allow their reintegration into society.

Window Under the old NGRI regime, a person with mental health issues could be detained involuntarily for an indefinite amount of time. || (Source: Flickr // Paul Hudson )

In R v Swain (1991), the Supreme Court of Canada (SCC) found the NGRI regime unconstitutional and pushed Parliament to enact new legislation for mentally ill offenders. Specifically, Chief Justice Lamer writing for the SCC majority, took issue with the NGRI regime’s automatic detention being subjected to the lieutenant governor’s beneficence and enabling potential indefinite detention. The SCC ruled that this violated section 7 (right to life, liberty, and security of the person) and section 9 (right not to be arbitrarily detained) of the Canadian Charter of Rights and Freedoms. The SCC struck down the regime, read down its current application, and asked Parliament to write new legislation within six months. And so, in 1992, Bill C-30 amended the Criminal Code, introducing the Review Board, a provincial/territorial tribunal with authority over mentally ill offenders, which thereafter introduced the new NCR regime via section 16.

Importantly, the new NCR regime introduced new procedural safeguards, stipulating that only the defence may raise issues of mental deficiency (there are some exceptions that are beyond the scope of this article), thus respecting the choice of the accused who may not wish to raise the NCR defence for a variety of reasons. Once raised, mental incapacity for the offence must be proved on a balance of probabilities (i.e., that is it more probable than not that the accused was suffering from a mental disorder at the time of the offence). When an accused is found to not have been in the “right” state of mind to appreciate the nature and quality of their acts, or could not form the requisite intent (i.e., mens rea or guilty mind), due to a “disease of the mind”, the accused will be found NCR. This ruling is not synonymous with an acquittal as the individual is deemed to have committed the offence – actus reus – but was lacking the mens rea component necessary for a guilty verdict.

2013-05-19 - 09.15.00 (Public) The Supreme Court of Canada found the NGRI regime unconstitutional. || (Source: Flickr // Robert Lindsell )

The Role of the Review Boards

Once an accused is found NCR, they fall under the authority of the province (or territory) and will receive treatment until such a time as they no longer pose a threat to society. The ultimate goal is to reintegrate the individual into society, while weighing the safety of the public against the rights of the accused. The provincial authority is embodied by the Review Board whose purpose is to review and modify dispositions to NCR offenders. Review Boards are administrative courts specialized in mental health and criminal law which require the presence of at least one psychiatrist and/or a physician with a different specialization, in addition to a mental health practitioner. The Review Board must set dispositions for the accused on a yearly basis, taking into account public safety, the mental condition of the accused, and the goal of societal reintegration. In Winko v BC, the accused posed a section 15 (equality under the law without discrimination) Charter challenge of section 672.54 of the Criminal Code, claiming that it placed an unfair burden on the accused to disprove dangerousness as failing to do so could lead to indefinite detention – reminiscent of the NGRI’s lieutenant governor’s discretion. The SCC rejected this argument but specified that the Review Board must order the least onerous and restrictive disposition available, unless the accused poses a significant threat to the public. Ultimately, following Winko the SCC stipulated that “[if] the accused is not a significant threat to the safety of the public, by order, direct that the accused be discharged absolutely.”

Shift in NCR Policy

In 2005, however, a shift in policy (Bill C-10) began affecting the NCR regime, where victims’ rights were bolstered at the expense of offenders’ liberty. For instance, Review Boards now had to inquire whether the victim knew about their right to submit an impact statement and could potentially adjourn the hearing if this right had been infringed by not informing them of their right to submit said statement. Importantly, the Review Board could now extend hearings by up to two years for offenders charged with a serious personal injury offence in detention and who were not considered likely to show behavioural improvement. This policy trend found public support following a series of brutal offences (often involving the killing of children) and wide-spread media coverage of the perpetrators’ NCR proceedings. The visceral reaction is understandable given the brutal and shocking nature of the index offences being publicized, however, it is important to consider that these cases are of extremely rare occurrence: approximately 1 in 1000 cases are found NCR and 10% of these are for violent offences.

In response to increased public fear, parliament adopted a tough-on-crime approach such as 2014’s Bill C-14’s introduction of a high-risk accused designation for persons with mental illness who have committed a serious personal injury offence and representing an increased likelihood of future violence. Furthermore, at Review Board hearings, the status of high-risk prevents the accused from obtaining any type of non-medical related leave from the hospital or detention centre. The new legislation increases the permissible interval between review hearings to a maximum of 36 months rather than the usual annual review given for other NCR individuals. Finally, once the Review Board is satisfied that the accused no longer poses a substantial risk, they must refer the case to the Superior Court for review, which is the only court with the power to revoke the high-risk status.

Three years since its introduction, no court has yet given an NCR offender the high-risk status. Most notably, the case of Mr. Schoenborn, who was found NCR in 2010 for the killing of his three children, has set a precedent in the applicability of the new designation. In 2015, the Crown petitioned the Supreme Court of British Columbia to have Mr. Schoenborn designated as a high-risk accused. The Court avoided ruling on this new piece of legislation by claiming that it did not operate retrospectively. However, in 2017, the Crown once again applied to have Mr. Schoenborn found high-risk, and this time, the Court ruled that the dangerousness of the accused must be assessed at the “present” time. Therefore, even if “high-risk” was introduced in 2014, the statute was not applied retrospectively.

In 2005, however, a shift in policy began affecting the NCR regime, where victims’ rights were bolstered at the expense of offenders’ liberty.

There is little doubt, however, that eventually an NCR offender will be labelled high-risk and this is likely to bring forth many of the same Charter challenges as the initial NGRI regime (i.e., s.7, s.9, s.15). In addition, it is fair to assume that an increasing “high-risk” designated population would strain current facilities and associated resources due to their long-term mandatory detention. These changes appear to stem from a change in the past fifty years that witnessed a shift from institutional care (such as psychiatric wards) toward community-based intervention in mental health. However, lack of community resources, and increasingly complex clinical and psycho-social profiles resulted in many individuals in need of institutional care coming instead in contact with the criminal justice system.


Loïc Welch is an Online Editor of the McGill Journal of Law and Health and a first-year B.C.L./LL.B. student at McGill University’s Faculty of Law. Loïc holds a M.Sc. in Forensic Psychology from Maastricht University (Netherlands), was a research assistant at the Douglas Mental University Institute in Montreal, and interned at the Professional Clinical and Forensic Services, a part of the Institute on Violence, Abuse, and Trauma in San Diego, California.

The Curious Dichotomy of (in)Sanities

Contributed by Loïc Welch

For most people, the inner workings of the brain and the resulting behaviour work seamlessly, and we behave in “socially acceptable” ways. But, what happens when the brain is “faulty?” Take for instance the famous case of Phineas Gage, who had a tragic accident at work, where a railroad blasting rod pierced through his skull into his brain. Mr. Gage survived this accident, however, his once well-behaved personality was altered to a more chaotic one. He began gambling, and became short-tempered, impulsive, and violent. Mr. Gage had suffered damage to the area of the brain known as the orbitofrontal cortex, implicated in emotional regulation, decision making, and impulse control.

Phineas Phinease Gage suffered brain trauma from a rod that pierced through his skull || (Source: Flickr // Protocol Snow)

If, for instance, Mr. Gage had killed someone in a barfight, without premeditation; by today’s legal standards, would Mr. Gage be found guilty (of 2nd degree murder or manslaughter) or would he instead be found not criminally responsible on account of mental disorder (NCRMD)? Is he responsible for his aberrant, perhaps incontrollable, behaviour ensuing from a traumatic brain injury? Or to the crux of the issue, are humans responsible for their behaviour, if it is nothing more than the result of brain function? A biological view of behaviour suggests that regardless of whether someone has a rod in their brain (tumour or other forms of severe brain trauma), all of their actions are controlled by their brain’s function.

The Guilty Mind

The above example is an introduction to the deterministic view of human behaviour as dictated by our brain (mal)functioning. Mr. Gage’s actions were determined by a brain that “misfired” as a result of physical trauma, leaving very little room for the concept of free will and voluntary action. This poses a fundamental challenge to the concept of criminal justice that relies on a crucial principle: that only the morally guilty should be punished. Guilt is determined by a defendant’s level of culpability, or in legal terms, their mens rea, which comes from the Latin phrase actus reus non facit reum nisi mens sit rea, translated to “the act is not culpable unless the mind is guilty.” The basic tenets of criminal justice inform our conception of mens rea: guilt is determined by one’s capacity to distinguish right from wrong and subsequent choice to act in the wrong. There is an assumption of free will and voluntary action. What happens to mens rea if the neuroscientific understanding of brain and behaviour removes free will from the agent? This question has lead to such defences as sane automatism, which can lead to full acquittal because it removes voluntary action from the individual.

The Not-so-Guilty Mind

To understand the case below (R v Stone), let me first explain the concept of automatism. There are two types of automatism defences: the first is sane automatism, where involuntary behaviour does not result from a mental disorder and is a complete defence, giving rise to full acquittal. For instance, sane automatism could be used as a defence if a defendant had murdered someone during a sleepwalking episode. Indeed, the Supreme Court of Canada, in R v. Parks, upheld the trial decision of allowing sleepwalking to be used as a sane automatism defence. This brings back the notion of mens rea: that only voluntary actions may lead to legal culpability. Generally, the factors required for a sane automatism defence must be extrinsic and are as follows: (1) there must exist an involuntary action arising from external source (or reflex action); (2) the action must be completely involuntary; and (3) the automatism must not be self-induced (that is why excessive alcohol/drug consumption is not a viable defence of automatism).

What happens to mens rea if the neuroscientific understanding of brain and behaviour removes free will from the agent?

The second type is insane automatism, where the actions of the accused are held to be the result of a mental disorder and triggers s.16 of the Canadian Criminal Code leading to a defendant being found NCRMD, again reflecting that the defendant could not appreciate the nature and quality of their act. Recall the earlier example of Mr. Gage killing a patron during a bar brawl; which of the two automatism defences could be argued most effectively? Would Mr. Gage’s brain injury justify the use of insane automatism because it would be considered as resulting from an internal cause? Or would the murder be acquitted under a sane automatism defence because Mr. Gage’s brain injury had an external causal factor [a rod] and was exacerbated by alcohol consumption? The predicament raised by the automatism defences, and the fine line between what is considered sane or insane by the courts, identifies a disconnect between fictional legal dichotomies and the neuroscientific reality that our brain controls our behaviour in all situations. This renders the concept of voluntariness moot to the eyes of science.

(mis)Communication of Science in the Courtroom

In recent years, there has been a sharp increase in the use of expert witnesses in courts which exemplifies a shift in the legal paradigm toward “hard facts,” stemming from science and its rigorous methodology. Among these experts are psychologist, psychiatrists, neuroscientists, and other specialists whose mandate is to inform the triers-of-fact in making more scientifically-informed judgements. When neuroscience meets the legal discipline, it is coined neurolaw.

In 1999, the Supreme Court of Canada was confronted with the challenging case of R v. Stone where Mr. Stone was appealing the guilty verdict of manslaughter for the killing of his wife via 47 stabbings. The trial judge had instructed the jury to consider insane automatism as a defence, but this failed, and he received a seven-year imprisonment sentence for manslaughter. The appeal, asking for Mr. Stone to be found NCRMD, was dismissed by the majority of the Supreme Court justices. However, three of the justices dissented, stating that the assessment of the appellant’s mental status at the time of the crime was not fully presented to the jurors. Specifically, the forensic psychiatrist brought in as an expert witness testified that the appellant was in a dissociative state, considered an unconscious state, when killing his wife. Furthermore, this was not attributed to a mental disorder, but rather a reaction to severe stress allegedly inflicted on him by his wife. The trial judge, and the concurring appeal judges, stated that although the accused had “periods” of unconsciousness during the murder, the expert witness did not assess the lack of voluntariness required for the defence of automatism. Interestingly, the jurors found that the appellant did not commit the crime voluntarily, thus resulting in a guilty verdict of manslaughter. The problem here is that the trial judge had only informed the jury of insane automatism, which requires the presence of a mental disorder (which the appellant did not have). Had the judge informed the jury about sane automatism, the verdict may have been a full acquittal.

Dissociative A dissociative state can cause impaired consciousness || (Source: Flickr // Vlad Gilcescu)

The lack of proper instruction to the jurors in R v. Stone brings forward a very important issue arising from the increased use of science in court. Namely, a need for comprehensive scientific literacy on the part of the jurists. In this case, the SCC dismissed the appeal partly on the grounds that the steps required for the insane automatism defence were not satisfied, stating: “As I have explained above, automatism is more properly defined as impaired consciousness, rather than unconsciousness. Furthermore, lack of voluntariness, rather than consciousness, is the key legal element of automatism. Accordingly, the trial judge should have concerned himself with assessing whether there was evidence that the appellant experienced a state of impaired consciousness in which he had no voluntary control over his actions rather than whether there was evidence that the appellant was unconscious throughout the commission of the crime.”

Although eloquent in its rhetoric, this reasoning does not reflect the state of the scientific evidence. The judge rests the dismissal of the appeal on a false dichotomy between unconsciousness and impaired consciousness, wherein he posits that the former does not fall in the latter. If the question of whether unconsciousness fell in the realm of impairments of consciousness was put to a neuroscientist, several questions about what was meant about impairment and in what context would follow. Rooted in the context of R v. Stone the unequivocal answer would be a strong affirmative of: “Indeed, Mr. Stone was impaired of consciousness.” The next question for the neuroscientist would be: “If the appellant suffered from impairment of consciousness resulting from a dissociative state, could his actions be considered of his own volition?” The logical answer would be in the negative. The issue appears to boil down to one of mismatched communication. Where jurors, attorneys, and judges would benefit from increased scientific literacy to clarify the issues put before them and assess the true worth of the expert testimony, and where neuroscientist (and experts in general) could use a crash course in legal standards and a disambiguation of legal jargon.

Loïc Welch is an Online Editor of the McGill Journal of Law and Health and a first-year B.C.L./LL.B. student at McGill University’s Faculty of Law. Loïc holds a M.Sc. in Forensic Psychology from Maastricht University (Netherlands), was a research assistant at the Douglas Mental University Institute in Montreal, and interned at the Professional Clinical and Forensic Services, a part of the Institute of Violence, Abuse, and Trauma in San Diego, California.