• Lewis Waring

The Daviault Defence - Nikolai von Schilling



In R v Sullivan (“Sullivan”), a joint appeal between David Sullivan and Thomas Chan, the Ontario Court of Appeal (“the ONCA”) declared Criminal Code section 33.1 to be unconstitutional and of no force or effect. In December, 2020, the Crown was granted leave to appeal this decision and, more than twenty-five years after the Criminal Code was amended to include section 33.1, which removes the possibility for individuals accused of violent offences to rely on the defence of automatism resulting from self-induced intoxication, the constitutionality of this section will finally be considered by the Supreme Court of Canada (“the Court”). In anticipation of this forthcoming legal spectacle, I will take this moment to provide a brief snapshot of the evolution of the jurisprudence with respect to automatism resulting from self-induced intoxication. Specifically, I will begin by addressing the origins of section 33.1 in R v Daviault (“Daviault”) and the resulting public uproar. I then go on to examine R v Bouchard-Lebrun (“Bouchard-Lebrun”), in which the Court interpreted Criminal Code section 33.1 and its interaction with section 16. Finally, I will brief the ONCA’s analysis of the constitutionality of section 33.1 in Sullivan.


The origins of the Daviault defence


R v Daviault (“Daviault”) involved the sexual assault of a sixty-five-year-old woman, partially paralyzed and confined to a wheelchair, by the accused. Based on the testimony of a pharmacologist, the defence in Daviault submitted that the accused’s blood alcohol content had likely been between 400 and 600 milligrams per 100 millilitres of blood, which would be enough to kill a normal person and was certainly enough to cause the accused to blackout. The issue in Daviault was whether the accused could rely upon the defence of self-induced intoxication resulting in automatism to negate the voluntariness requirement for the general intent offence of sexual assault. Specifically, the Court in this case revisited the rule from R v Leary (“Leary”), which held that intoxication can only be used as a partial defence to vitiate the mens rea element of a specific intent crime and that, when it came to general intent crimes such as sexual assault, the intent to become intoxicated could stand in for the minimal voluntariness requirement to commit the offence. The majority opinion in Daviault, written by Justice Cory, held that it was contrary to sections 7 and 11(d) of the Canadian Charter of Rights and Freedoms (“the Charter”) to bar an accused from submitting evidence that he or she lacked the necessary voluntariness to commit a general intent crime. Ultimately, the majority in Daviault held that the rule from Leary should be altered to permit an accused to establish, on a balance of probabilities, that at the time of the offence he or she was in a state of automatism resulting from self-induced intoxication and, for this reason, was unable to form the requisite degree of voluntariness to constitute even a crime of general intent. In his written judgement, Justice Cory predicted that “it will be a rare situation where an accused is able to establish such an extreme degree of intoxication.”


Public uproar and section 33.1


Daviault’s appeal at the Court was allowed and a new trial was ordered. Unfortunately, the complainant passed away and the Court was forced to dismiss charges against the accused. As such, we have no way of knowing whether the newly created Daviault defence would have been successful in its originating case. The Court’s holding in Daviault, however, created an uproar among the public, “who saw the decision as essentially permitting drunken men to rape with impunity.” To make matters worse, despite Justice Cory’s assurances to the contrary, the Daviault defence was employed successfully in at least three cases directly after the Court’s decision.

To temper public outrage, within five months of the Daviault decision Parliament had tabled Bill C-72, introducing an amendment to the Criminal Code in the form of section 33.1. During the second reading of Bill C-72, the Justice Minister at the time clearly laid out its two-fold purposes: (1) to ensure that those accused of violent crimes would be held accountable, notwithstanding their level of intoxication, and (2) to prevent those who would otherwise attempt to “manipulate the legal principles so as to intoxicate themselves to some extent for the purpose of committing a crime.” However, I would point out that, in addition to the two aforementioned purposes, and perhaps as an extension of the accountability purpose, there are strong indirect reasons behind Parliament’s rejection of the Daviault defence. Most notably, by giving offenders an additional defence in the form of intoxicated automatism, many commentators were concerned that women would be discouraged from reporting sexual assaults, police would be less inclined to consider these assaults well-founded, and Crown prosecutors would be less likely to press charges.

Despite Parliament’s decision to enact section 33.1, there were at the time various reasons to believe that the public outrage at the Court’s decision in Daviault was overblown. Firstly, even at the time that the Standing Committee on Justice and Legal Affairs was considering whether to enact section 33.1 it had evidence that automatism could not be caused by alcohol alone. Specifically, psychiatric experts argued that, although a loss of memory always follows a period of automatism, the two are not inextricably linked so that every loss of memory implies an associated state of automatism. In fact, “the typical action of alcohol on the brain…is to progressively decrease all types of nerve cell activity, including those involved in coordinated movements and those involved in consciousness and memory, more or less in parallel”. So if automatism is defined as a loss of consciousness or voluntary control with a concurrent ability to perform physical activities of relative complexity, then an individual who has become blackout drunk, like the accused in Daviault, will lose consciousness at roughly the same rate as he or she loses the ability to perform complex motor functions.

There is a strong possibility that, if the complainant in Daviault had not passed away before the second trial, despite the accused’s ability to adduce evidence that he was in an automatic state caused by alcohol intoxication, he may have nevertheless been unsuccessful in proving on a balance of probabilities that his level of alcohol intoxication amounted to a state of automatism. This is evidenced by the fact that, although courts in Ontario have routinely found section 33.1 unconstitutional, most cases of alcohol-induced automatism in that province have been unsuccessful due to the inability to show a connection between alcohol consumption and automatism.

Furthermore, Parliament had other options open to them aside from the statutory limitation of the Daviault defence under section 33.1. For instance, despite the fact that there were at least three successful uses of the Daviault defence before the enactment of section 33.1, it was open to Parliament to codify the Daviault defence in order to emphasize aspects such as the burden of proof and evidence required to establish automatism resulting from self-induced intoxication. In this way, the Daviault defence could have been legislated in line with Justice Cory’s repeated assurances that it would be available only rarely. Additionally, as suggested by the majority in Daviault, it was open to Parliament to create the new offence of ‘criminal intoxication leading to harmful conduct’, which may have avoided the Charter issues resulting from the guilt-by-proxy aspect of section 33.1. For better or for worse, however, Parliament decided to statutorily restrict the Daviault defence under section 33.1 so that this defence could not apply to any crime which interfered with the bodily integrity of another person.


R v Bouchard-Lebrun: almost a second look


Before coming to the ONCA’s eventual rebuke of Parliament and section 33.1 in Sullivan, I will briefly touch on the Court’s decision in R v Bouchard-Lebrun. As I have already mentioned, there is evidence that alcohol intoxication alone cannot lead to a state of automatism. However, drugs such as PCP, ketamine, and other dissociative anesthetics can, by themselves, lead to automatism. Additionally, drugs such as cannabis, amphetamines and cocaine can, in high doses, lead to mental states of extreme fear, aggression and paranoia in which consciousness, although present, is “so impaired by the drug's effects as to constitute … a chemically induced temporary insanity.” The question remains, then, whether this type of “chemically induced temporary insanity” should be covered under Criminal Code section 16 or whether it is properly a form of automatism and subject to the limitations under section 33.1. In Bouchard-Lebrun, the Court definitively answered this question when it held that section 16 and 33.1 analyses are mutually exclusive.

The facts of Bouchard-Lebrun involved a young man who, after consuming ecstasy pills, entered a state of toxic psychosis and assaulted the two complainants at their home. Throughout the assaults, the accused behaved erratically, entering a state which “might be described as religious delirium”. At trial, the accused pled not guilty to the two charges of assault and aggravated assault and on appeal argued that he should be found not criminally responsible on account of mental disorder (“NCRMD”) because his state of toxic psychosis, induced by the consumption of the ecstasy pills, had rendered him unable to tell right from wrong.

In determining whether the accused was suffering from a true “disease of the mind” in the legal sense or merely a state of automatism, courts must consider two analytical tools: (1) the internal cause factor and (2) the continuing danger factor. The first of these, the internal cause factor, requires a court to consider the psychiatric evidence provided at trial to determine whether a ‘normal person’ in the circumstances would have entered a state of automatism after consuming the same intoxicants. The second of these analytical tools, the continuing danger factor, recognizes that “when a condition is likely to present a recurring danger, there is a greater chance that it will be regarded as a disease of the mind”. In Bouchard-Lebrun, the toxic psychosis that the accused experienced which led to the assaults at issue was found by the Court to be both externally caused and temporary in nature. As such, the Court concluded that the mental state of the accused could not be considered a ‘disease of the mind’ under section 16. The Court therefore concluded that the mental state of the accused should be considered a form of automatism resulting from self-induced intoxication and that the proper branch of analysis in this case should be done under section 33.1. Ultimately, the Court in Bouchard-Lebrun held that section 33.1 will apply when: “(1) the accused was intoxicated at the material time; (2) the intoxication was self‑induced; and (3) the accused departed from the standard of reasonable care generally recognized in Canadian society by interfering or threatening to interfere with the bodily integrity of another person.” Since the accused in Bouchard-Lebrun met all three of these criteria, the Court held that the state of toxic psychosis he experienced was captured under section 33.1, and he could not defend himself on the basis that he lacked the requisite voluntariness to be found guilty of the charges at issue. Ultimately, the appeal of the accused was dismissed by the Court and he was found guilty of the two charges of assault and aggravated assault.


R v Sullivan: a long-awaited second look


In Sullivan the ONCA heard the appeals of both Thomas Chan and David Sullivan, both of whom challenged the constitutionality of section 33.1. To begin, the facts of Mr. Chan’s case involved the purposeful ingestion of psilocybin mushrooms, after which Mr. Chan entered a state of automatism, broke into his father’s house, and— without identifiable motive—stabbed his father to death and severely injured his step-mother. The facts of Mr. Sullivan’s case, on the other hand, involved a suicide attempt in which Mr. Sullivan ingested a large number of Wellbutrin tablets, which had been prescribed to him in order to quit smoking, entered an agitated state in which he believed his mother was an alien and proceeded to stab her multiple times.

As we have seen in Bouchard-Lebrun, this kind of self-induced intoxication leading to states of extreme agitation, paranoia and perhaps automatism will not permit the accused to rely on a NCRMD plea if the toxic psychosis: (1) had an external cause and (2) did not result in a continuing danger to the public. Furthermore, due to the enactment of section 33.1 of the Criminal Code and because both of these cases involved interference with the bodily integrity of others, neither appellant was able to rely on the Daviault defence.

Although this view may not be shared by all Canadians, it seems to me that the tragedy of both Mr. Chan’s and Mr. Sullivan’s respective cases invites some degree of leniency, and if neither of these accused voluntarily committed the crimes they have been charged with then they should not be found guilty. Ultimately, the ONCA in Sullivan agreed and held that section 33.1 is unconstitutional and is of no force or effect. In the following, I will examine the majority’s reasoning with respect to the constitutionality of section 33.1. Specifically, I will examine in what ways section 33.1 represents a prima facie infringement of sections 7 and 11(d) of the Charter. I will then look at the court’s analysis of whether section 33.1 of the Criminal Code can be saved by section 1 of the Charter, paying close attention to the issues that I have identified previously in this work regarding the purposes of and alternatives to section 33.1.

The constitutional analysis undertaken by the ONCA in Sullivan largely echoed the analysis done by the Court in Daviault. In Sullivan, the Court identified three ways in which section 33.1 constitutes a prima facie infringement of the Charter: (1) the voluntariness breach, (2) the improper substitution breach, and (3) the mens rea breach.

A voluntariness breach is a prima facie infringement of both sections 7 and 11(d) of the Charter insofar as section 33.1 expressly states that, in the case of violence-based offences, it is no defence that the accused lacked the requisite degree of voluntariness due to self-induced intoxication. Firstly, section 33.1 is a prima facie infringement of section 7 because it denies an accused the opportunity to defend him- or herself by arguing that his or her actions lacked the requisite degree of voluntariness, which is certainly contrary to the principle of fundamental justice which requires “that voluntariness is an element of every criminal offence.” Secondly, section 33.1 is a violation of section 11(d) of the Charter insofar as it allows a court to convict an accused even where there is a reasonable doubt as to the voluntariness of his or her actions.

Similarly, an improper substitution breach involves an infringement of section 11(d) of the Charter insofar as it allows an accused to be convicted of a crime without the Crown proving all the required elements of that crime. For instance, in order for the Court to convict Mr. Chan of assault, the Crown needed to prove all of the requisite elements of that crime, one of which is the intention to commit it. The Crown argued that section 33.1 allowed the intention to become intoxicated to be substituted for the intention to commit assault, and, in this way, all elements of that crime were established. However, since the intention to become intoxicated is not a requisite element of the crime of assault and assault does not follow in every case of intoxication, section 33.1 “relieves the crown of its burden of establishing all the elements of the crime[...]” and is therefore an infringement of section 11(d).

Finally, a mens rea breach in section 33.1 of the Criminal Code is a prima facie infringement of section 7 of the Charter insofar as it allows a court to convict an accused without the minimum degree of fault required for general intent offences. According to the Court, the minimum degree of fault in line with principles of fundamental justice is the penal negligence standard, which requires a marked departure from the standard of a reasonable person. This marked departure, in turn, is measured by asking whether “the relevant risk [is] reasonably foreseeable such that it not only falls below standards of ordinary prudence to engage in the risky behaviour but doing so amounts to a marked departure from standards of ordinary prudence”. The Court in Sullivan pointed out that nowhere in the wording of section 33.1 or in the Court’s interpretation of this section in Bouchard-Lebrun was there any requirement that either violence or a state of automatism be foreseeable to the accused in departing from the standard of care of a reasonable person.

With these prima facie infringements in mind, I move on now to the ONCA’s analysis concerning whether section 33.1 of the Criminal Code can be justified by section 1 of the Charter. Firstly, the court in Sullivan identified two purposes behind section 33.1:

(1) to hold individuals who are in a state of automatism due to self-induced intoxication accountable for their violent acts [the “accountability purpose”]; and (2) to protect potential victims, including women and children, from violence-based offences committed by those who are in a state of automatism due to self-induced intoxication [the “protective purpose”].

According to the Court in Sullivan, section 33.1 explicitly attempts to hold an individual in an automatic state accountable whereas principles of fundamental justice under section 7 deny that someone lacking voluntariness should be held accountable for any crime. In this way, the ‘accountability purpose’ cannot serve as a pressing and substantial purpose under a section 1 analysis because “legislation is unconstitutional if its purpose is unconstitutional”.

That being the case, the court in Sullivan agreed that the ‘protective purpose’ sufficed as a pressing and substantial purpose under a section 1 analysis. However, when it comes to the proportionality analysis with respect to the ‘protective purpose’, the court in Sullivan held that section 33.1 failed the rational connection test, the minimal impairment test, and the overall proportionality test.

Firstly, the court in Sullivan held that section 33.1 does not pass the rational connection test with respect to the protective purpose because no “reasonable person would anticipate the risk that, by becoming voluntarily intoxicated, they could lapse into a state of automatism and unwillfully commit a violent act.” Secondly, section 33.1 did not meet the minimal impairment test with respect to the protective purpose for two primary reasons: (1) it is not minimally tailored to achieve a narrow protective purpose and (2) Parliament did not have valid reasons for rejecting less intrusive methods to achieve the protective purpose. With respect to less intrusive means, I have already briefly laid out a few under the consideration of Parliament twenty-five years ago when the Daviault decision was still a fresh wound. For instance, if Parliament had chosen to create the new offence of ‘criminal intoxication leading to harmful conduct’ the ‘accountability purpose’ previously discarded by the ONCA in Sullivan may have been legitimate. Finally, section 33.1 did not meet the overall proportionality test insofar as its deleterious effects far outweighed any possible salutary effects. Namely, section 33.1 would allow a morally blameless accused who committed an act in an automatic state to be convicted of a heinous crime such as assault. On the other hand, since the Court held that the accountability purpose was constitutionally illegitimate, the salutary effects must be the product solely of the ‘protective purpose’. However, due to the protective purpose’s failure at the rational connection stage, any possible salutary effects are theoretical at best.

Ultimately, the ONCA in Sullivan held that section 33.1 of the Criminal Code infringed upon sections 7 and 11(d) of the Charter and could not be upheld under section 1. As such, the ONCA exercised its power under section 52(1) of the Constitution Act and declared section 33.1 of no force or effect.

Recent Posts

See All

Right to Counsel - Liam Pollock

On July 1, 2018, an officer of the Royal Canadian Mounted Police [“the RCMP”] approached a car, which he stated was parked in the middle of the road. Inside the car, Glenn Cure was sleeping in the dri

Mistakes in the evolving rules of delay - D. Grohl

The Canadian Charter of Rights and Freedoms (“the Charter”) establishes in section 11(b) that any person who is charged with an offence must be tried for that offence in a reasonable amount of time.1

Check out the Robson Crim MLJ
  • Facebook Basic Black
  • Twitter Basic Black