R. v. Fontaine, [2004] 1 S.C.R. 702, 2004 SCC 27

Her Majesty The Queen Appellant

v.

Jacques Fontaine Respondent

Indexed as: R. v. Fontaine

Neutral citation: 2004 SCC 27.

File No.: 29198.

2003: November 6; 2004: April 22.

Present: McLachlin C.J. and Iacobucci, Major, Bastarache, Binnie, Arbour, LeBel, Deschamps and Fish JJ.

on appeal from the court of appeal for quebec

Criminal law — Defences — Mental disorder automatism — Trial for murder before judge and jury — Trial judge withholding defence of mental disorder automatism from jury — Nature of evidential burden — Whether accused entitled to have defence considered and decided by jury.

The accused worked as a garage mechanic. He received a call from R, a disgruntled former employee, who said, “We’re coming to get you, pigs.” The accused later heard from a co-worker that the victim had been offered a contract to kill both of them. Feeling that he was being watched and followed, the accused purchased a firearm. One evening, the accused thought he saw R lurking outside his home. The co-worker came by to check, but noticed nothing unusual. During the night, after smoking marijuana, the accused thought he heard someone breaking into his home. He fired the gun at doors and windows and into walls and concluded that he had shot the intruders. Seeing the victim at the garage the next day, the accused shot and killed him. Before a judge and jury, the accused pleaded mental disorder automatism. Several psychiatrists gave evidence. The judge refused to put this defence to the jury on the ground that the required evidential foundation had not been laid. The accused was convicted of first degree murder. The Court of Appeal quashed the conviction and ordered a new trial.

Held: The appeal should be dismissed. The accused’s defence of mental disorder automatism should have been put to the jury.

If there is some evidence upon which a properly instructed jury could reasonably conclude that an accused probably perpetrated the alleged criminal act in a state of automatism, the evidential burden has been discharged and the defence is in play before the jury. An “evidential burden” is not a burden of proof. It determines whether an issue should be left to the trier of fact, while the “persuasive burden” determines how the issue should be decided. These are fundamentally different questions. The first is a matter of law; the second, a question of fact. Accordingly, on a trial before judge and jury, the judge decides whether the evidential burden has been met. In answering that question, the judge does not evaluate the quality, weight or reliability of the evidence. The judge simply decides whether there is evidence upon which a properly instructed jury could reasonably decide the issue. In the case of “reverse onus” defences, such as mental disorder automatism, it is the accused who bears both the persuasive and the evidential burdens. In these defences, the persuasive burden is discharged by evidence on the balance of probabilities. As regards all affirmative defences, the evidential burden will be discharged where there is some evidence that puts the defence “in play”; the defence will be in play whenever a properly instructed jury could reasonably, on account of that evidence, conclude in favour of the accused. Where mental disorder automatism is raised as a defence, an assertion of involuntariness on the part of the accused, supported by evidence from a qualified expert which, if accepted by the jury, would tend to support that defence, will normally provide a sufficient evidentiary foundation for putting the defence to the jury. Accompanying instructions in law will make it clear to the jury that the burden remains on the accused to establish the defence to the required degree of probability.

Applying the law to the facts, the Court of Appeal did not err as to the nature of the evidential burden on a defence of mental disorder automatism, nor did it err in concluding that the accused had discharged that burden and was entitled to have his defence considered and decided by the jury. The accused gave evidence tending to establish that he was acting involuntarily at the time of the offence. He also adduced expert evidence to support his own testimony. The evidence clearly went beyond a mere allegation of the existence of a defence. It included a relatively detailed description of the accused’s perception of the facts at the moment of the criminal act. The main defence expert, a psychiatrist, concluded that the accused was suffering, at the time of the offence, from a psychotic episode induced by substance abuse. According to the expert, at the relevant times the accused was “seeing things” and making pathological connections between people, situations and events. Taken as a whole, this evidence was sufficient to discharge the accused’s evidential burden on his defence of mental disorder automatism.

Cases Cited

Referred to: R. v. Stone, [1999] 2 S.C.R. 290; R. v. Arcuri, [2001] 2 S.C.R. 828, 2001 SCC 54; R. v. Cinous, [2002] 2 S.C.R. 3, 2002 SCC 29; R. v. Capson, [1953] 1 S.C.R. 44; R. v. Thibert, [1996] 1 S.C.R. 37; R. v. Charemski, [1998] 1 S.C.R. 679; R. v. Schwartz, [1988] 2 S.C.R. 443; R. v. S. (W.D.), [1994] 3 S.C.R. 521.

Statutes and Regulations Cited

Canadian Charter of Rights and Freedoms, s. 11(f).

Criminal Code, R.S.C. 1985, c. C-46, ss. 16, 568, 686(1)(b)(iii).

Authors Cited

Cross, Rupert, and Colin Tapper. Cross and Tapper on Evidence, 9th ed. London: Butterworths, 1999.

McCormick, Charles Tilford. McCormick on Evidence, 5th ed., vol. 2. St. Paul, Minn.: West Group, 1999.

APPEAL from a judgment of the Quebec Court of Appeal, [2002] R.J.Q. 613, 2 C.R. (6th) 164, [2002] Q.J. No. 395 (QL), setting aside the accused’s conviction for first degree murder and ordering a new trial. Appeal dismissed.

Joanne Marceau and Sébastien Bergeron‑Guyard, for the appellant.

Sébastien St‑Laurent, for the respondent.

The judgment of the Court was delivered by

Fish J. —

I. Introduction

[1] Jacques Fontaine was indicted and tried in the Quebec Superior Court on a charge of first degree murder. His sole defence of concern to us here was mental disorder automatism.

[2] The trial judge withheld that defence from the jury on the ground that the required evidential foundation had not been laid. Fontaine was found guilty as charged.

[3] In a unanimous decision, the Quebec Court of Appeal quashed Fontaine’s conviction and ordered a new trial. Unlike the trial judge, the Court of Appeal concluded that Fontaine was entitled by law to have the merits of his defence considered and determined by the jury. More particularly, the Court of Appeal found that Fontaine, through his own evidence and that of Dr. Jacques Talbot, a psychiatrist, had discharged his evidential burden on that issue.

[4] We are now urged, on this further appeal by the Crown, to set aside the judgment of the Court of Appeal and to restore the conviction entered at trial. The Crown contends that we are bound to do so by the decision of the Court in R. v. Stone, [1999] 2 S.C.R. 290. In the Crown’s view, the trial judge — before putting the respondent’s defence to the jury — was bound by Stone to satisfy himself that the respondent had acted involuntarily. And Stone, says the Crown, must be left unturned.

[5] I would dismiss the appeal. This result is in my view dictated both by principle and by recent precedent.

[6] Whether or not Stone was meant to be read as the Crown suggests, Stone can no longer be read that way in view of the Court’s subsequent decisions, particularly R. v. Arcuri, [2001] 2 S.C.R. 828, 2001 SCC 54, and R. v. Cinous, [2002] 2 S.C.R. 3, 2002 SCC 29.

[7] Here, the trial judge did not have the benefit of either Arcuri or Cinous — neither had yet been decided — and his application of Stone must be seen in that light.

[8] Moreover, in applying Stone as he did, the trial judge erred understandably. First, because there is language in Stone that has been understood to require the kind of judicial weighing of evidence that the law, since Cinous, plainly prohibits. Second, were it open to the trial judge to weigh the evidence as he did, his adverse conclusion might well have been found, on the record as we have it, to be neither surprising nor unreasonable on its merits.

II. The Issues to Be — and Not to Be — Decided on This Appeal

[9] I think it important to circumscribe from the outset the scope of the present appeal.

[10] We are concerned with the evidential burden on a defence of mental disorder automatism and not with the “persuasive” burden on that defence.

[11] An “evidential burden” is not a burden of proof. It determines whether an issue should be left to the trier of fact, while the “persuasive burden” determines how the issue should be decided.

[12] These are fundamentally different questions. The first is a matter of law; the second, a question of fact. Accordingly, on a trial before judge and jury, the judge decides whether the evidential burden has been met. In answering that question, the judge does not evaluate the quality, weight or reliability of the evidence. The judge simply decides whether there is evidence upon which a properly instructed jury could reasonably decide the issue.

[13] This evidential standard, unaltered in substance for at least a century, appears in the case law variously attired: “a reasonable jury acting judicially”, “a properly instructed jury acting judicially” and “a properly instructed jury acting reasonably” (Cinous, at para. 62); “the jury, properly instructed and acting honestly and reasonably” (R. v. Capson, [1953] 1 S.C.R. 44, at p. 48); “a reasonable jury acting judicially and properly instructed” (R. v. Thibert, [1996] 1 S.C.R. 37, at para. 1).

[14] These are only some of the phrases, equivalent in substance though different in form, that appear in the governing cases. I intend no departure from that venerable and unbroken line of authority in saying that the evidential burden is discharged if there is some evidence upon which a properly instructed jury could reasonably decide the issue. Similar expressions that appear throughout are likewise meant to describe a single concept, the “evidential burden” in a criminal trial.

[15] As we shall see, the trial judge in this case decided that the respondent had not discharged the evidential burden on his defence of mental disorder automatism.

[16] The persuasive burden on that defence is not in issue and we are not required to determine whether the respondent discharged that burden here. It was for the jury to make this determination, pursuant to its own appreciation of the evidence and in the light of appropriate instructions in law — assuming, of course, that the required evidential foundation is found to have been laid.

[17] Nor are we concerned with the evidential or the persuasive burden on a defence of non-mental disorder automatism — that defence is not in issue on the appeal — or with other aspects of the decision in Stone. More particularly, we are not concerned with the effect, where automatism is raised as a defence, of s. 16(2) of the Criminal Code, R.S.C. 1985, c. C-46, which provides that “[e]very person is presumed not to suffer from a mental disorder . . .”.

[18] Finally, we have not been asked to consider whether the defence of mental disorder, as distinct from the defence of mental disorder automatism, ought to have been put to the jury.

[19] Only two issues are before us:

(1) Did the Court of Appeal err in law as to the nature of the evidential burden on a defence of mental disorder automatism?

(2) Did the Court of Appeal err in law in concluding that the respondent had discharged that burden in this case and was therefore entitled to have his defence considered — and decided — by the jury?

[20] As already indicated, I would resolve both issues in the respondent’s favour and, accordingly, dismiss the appeal.

III. Facts

[21] At the time of the offence, the respondent, Fontaine, worked “under the table” as a garage mechanic. On February 10, 1999, he received a call from Jules Renaud, a disgruntled former employee, who said, [translation] “We’re coming to get you, pigs.”

[22] Two days later, Robert Dompierre, the victim, appeared at the garage and told Benoît Randall, who worked there with Fontaine, that someone had tried to hire him to murder both Randall and Fontaine. He told Randall as well that there was [translation] “a contract out on their heads”. Randall later that day shared this information with Fontaine.

[23] On February 14, 1999, Fontaine and his girlfriend, Diane Boulé, went to a meeting. Fontaine and Boulé testified that four men were watching him there and two of them followed him into the washroom, where he had gone to smoke marijuana.

[24] Upon leaving the meeting, Fontaine used his remote starter lest an explosive device had been installed in his car. Fontaine then advised Randall that he had been followed. Together, they purchased a firearm.

[25] That same evening, Fontaine saw Renaud lurking outside his home. He called Randall, who came by but noticed nothing unusual. During the night, Fontaine again smoked marijuana. At around four o’clock in the morning, he thought he heard someone breaking into his home. He fired his gun at the doors and windows and into the walls. Fontaine later told his girlfriend and Randall that he thought he had shot some nocturnal intruders.

[26] Later that morning, Dompierre came to the garage to pay a debt he owed the proprietor. Seeing Dompierre approach, Fontaine grabbed his gun and shot Dompierre twice. Dompierre attempted to flee. Fontaine followed him outside, fired five more shots in his direction and wounded him fatally.

[27] Fontaine fled the scene but later turned himself in to the police.

[28] During his interrogation, Fontaine claimed that he had shot Dompierre to prevent Dompierre from carrying out the “contract on his [Fontaine’s] head”.

[29] At trial, Fontaine’s defence was mental disorder automatism. He testified that he had acted involuntarily and related in detail the circumstances that gave rise to his state of mind at the time of the offence. He explained that he had “frozen” and had only a partial recollection of the facts concerning the murder of Dompierre.

[30] Several psychiatric experts also gave evidence. Dr. Richard Laliberté testified that Fontaine had smoked marijuana excessively during the weeks leading up to the murder. According to Dr. Laliberté, certain aspects of Fontaine’s story were indicative of paranoid delusions, though it was difficult to determine with certainty whether Fontaine’s story was real or delusional.

[31] Dr. Bruno Laplante, another psychiatrist, testified that Fontaine suffered from a chronic antisocial personality disorder and acute paranoid delusions, but did not fall within s. 16 of the Criminal Code.

[32] Dr. Jacques Talbot, also a psychiatrist, was the main defence expert. He presented two reports. In the first, he concluded that Fontaine had a psychological condition characterized by delusions. These could be triggered by a chemical substance, including marijuana. Dr. Talbot found that it was difficult to determine to what extent Fontaine’s belief that his life was in danger was fact-driven and to what extent it was delusional.

[33] In his second report, which he adopted in his testimony, Dr. Talbot stated that Fontaine, at the time of the shooting, was in a psychotic state triggered by substance abuse. This episode, according to Dr. Talbot, had begun several days earlier and ended several days afterward. Dr. Talbot concluded that a major psychological problem had so affected Fontaine’s judgment that he was unable to differentiate right from wrong.

[34] Finally, Dr. Sylvain Faucher, yet another psychiatrist, testified for the Crown, in reply, that Fontaine was not psychotic at the time of the murder and was not psychotic at any time in 1999.

IV. Judicial History

A. Quebec Superior Court

[35] In February 2000, Fontaine was tried by a judge and jury for the first degree murder of Robert Dompierre. The trial judge instructed the jury that the facts did not give rise to the defences of mental disorder automatism or non-mental disorder automatism.

[36] Fontaine was found guilty of first degree murder the next day and sentenced to life imprisonment.

[37] The trial judge had earlier denied a request by Fontaine’s counsel that he put the defence of mental disorder automatism to the jury. The trial judge first stated that Stone, supra, set out the factors to be considered in determining whether the accused had discharged his evidential burden. He then proceeded to weigh the expert evidence adduced before him in that light.

[38] In the trial judge’s view, Dr. Talbot’s evidence was based largely on the facts as recounted by the accused and had been contradicted by the evidence of Drs. Faucher, Laliberté and Laplante.

[39] The trial judge explained in these terms his decision to withhold from the jury the respondent’s defence of automatism and, generally, any defence under s. 16 of the Criminal Code:

[translation] Whereas Dr. Talbot gave expert evidence in his testimony, and that expert opinion was in very large part based on the facts as recounted to him by the accused;

Whereas that evidence was contradicted by the testimony of Dr. Faucher and the expert opinions of Drs. Laliberté and Laplante;

Whereas even the accused’s own testimony contained significant contradictions on a number of points and lent little credibility to his claims;

Whereas the triggering event in this case cannot be characterized as a traumatic psychological blow, given the accused’s own admission to the effect that he was not afraid, and that his stress was also caused by his being on social assistance;

Whereas the accused had no medical history of automatistic-like dissociative states that would corroborate his claim of automatism;

Whereas bystanders who were able to observe the accused’s state before, during and even after the events in question all agreed that he was extremely nervous and did not show any symptoms that would lead one to believe that he was impaired by the effects of cannabis;

Whereas the evidence as a whole points to the existence of a motive, that is, a contract out on the accused’s life, which would mean that the trigger for the alleged automatism was the victim himself;

Whereas under these circumstances, the accused’s alleged mental disorder automatism is not supported by the evidence, and I can see no evidence that would allow a properly instructed jury to conclude, on a balance of probabilities, that the accused acted involuntarily;

Whereas I arrive at the same conclusion for the same reasons with respect to a claim of non-mental disorder automatism;

FOR THESE REASONS:

There are no grounds to instruct the jury as requested by the accused with regard to s. 16 of the Criminal Code of Canada.

B. Quebec Court of Appeal [(2002), 2 C.R. (6th) 164](https://www.canlii.org/fr/qc/qcca/doc/2002/2002canlii18159/2002canlii18159.html)

[40] In March 2002, the Quebec Court of Appeal quashed Fontaine’s conviction and ordered a new trial.

[41] Writing for the court, Rousseau-Houle J.A. reviewed the evidence in some detail, considered the governing authorities, including Stone, and concluded that the trial judge had erred in refusing to put the accused’s defence of mental disorder automatism to the jury.

[42] In her view, Dr. Talbot’s opinion rested for the most part on the evidence adduced at trial. It was true that the accused’s troubled state may have resulted in part from events that were real and not imagined, but this did not mean that Dr. Talbot’s conclusion regarding the accused’s delusional paranoid disorder could be disregarded by the judge.

[43] Rousseau-Houle J.A. noted the accused’s insistence under oath that he would never have killed the victim, whom he knew, had his judgment not been so profoundly disturbed by exhaustion, by consumption of drugs and by his psychotic state. She referred in this regard to the confirmatory evidence of Dr. Talbot.

[44] Rousseau-Houle J.A. noted as well that Bastarache J., in Stone, supra, at para. 189, had emphasized that a medical history of automatistic-like dissociative states was not “a requirement for the defence to meet its evidentiary burden”, but merely a relevant factor.

[45] Rousseau-Houle J.A. then added, at paras. 58-59:

[translation] While I am aware that the law respecting automatism is rightly mindful of public safety and order, and that an accused’s claim that he or she was not in control of his or her actions cannot be accepted lightly, I am of the view that in the case at bar the trial judge should have concluded that there was enough evidence to allow a properly instructed jury to decide on a balance of probabilities whether the [accused] acted involuntarily, and that the trial judge should have decided which form of automatism defence should have been put to the jury.

Parliament has decided that the task of weighing the credibility of such a defence is best left to the jury. This principle was affirmed in R. v. Corbett, [1988] 1 S.C.R. 670, at p. 693: “. . . until the paradigm is altered by Parliament, the court should not be heard to call into question the capacity of juries to do the job assigned to them”.

[46] It was for these reasons, essentially, that the Court of Appeal allowed Fontaine’s appeal, quashed his conviction, and ordered a new trial.

[47] Fontaine’s only ground of appeal, I repeat, was that the trial judge had erred in refusing to put his defence of mental disorder automatism to the jury. The judge’s decision not to put the defence of mental disorder, as I mentioned earlier, is thus not in issue before us.

V. Discussion

A. The Law

(1) The Governing Principles

[48] Under our system of trial by judge and jury, the judge decides all questions of law and delimits the issues of fact to be considered and determined by the jury. To avoid manifest unfairness and undue confusion, no issues will be put to the jury in the absence of a sufficient evidential foundation.

[49] In determining whether the required evidential foundation exists, it will be helpful to first consider the incidence and nature of the burden of proof on the issue concerned.

[50] Dealing recently with this issue in the context of directed verdicts of acquittal, McLachlin J. (now C.J.) reaffirmed in the clearest of terms that the evidential burden is a function of the persuasive burden on the issue of guilt — and, presumably, on other issues as well. Writing for herself and Major J. in R. v. Charemski, [1998] 1 S.C.R. 679, McLachlin J., dissenting on other grounds, stated, at para. 35:

. . . “sufficient evidence” must mean sufficient evidence to sustain a verdict of guilt beyond a reasonable doubt; merely to refer to “sufficient evidence” is incomplete since “sufficient” always relates to the goal or threshold of proof beyond a reasonable doubt. This must constantly be borne in mind when evaluating whether the evidence is capable of supporting the inferences necessary to establish the essential elements of the case. [Emphasis added.]

[51] American and English authorities are to the same effect: see McCormick on Evidence (5th ed. 1999), vol. 2, at p. 417; Cross and Tapper on Evidence (9th ed. 1999), at p. 138.

[52] In some instances, the proponent of an issue bears both the persuasive and the evidential burdens. But this is not invariably the case.

[53] On the ultimate issue of guilt, the Crown bears both burdens. The Crown’s persuasive burden on this issue can only be discharged by proof beyond a reasonable doubt. Accordingly, as McLachlin J. explained in Charemski, supra, the case against the accused cannot go to the jury unless there is evidence in the record upon which a properly instructed jury could rationally conclude that the accused is guilty beyond a reasonable doubt.

[54] In the case of “reverse onus” defences, such as mental disorder automatism, it is the accused who bears both the persuasive and the evidential burdens. Here, the persuasive burden is discharged by evidence on the balance of probabilities, a lesser standard than proof beyond a reasonable doubt. Reverse onus defences will therefore go to the jury where there is any evidence upon which a properly instructed jury, acting judicially, could reasonably conclude that the defence has been established in accordance with this lesser standard.

[55] With respect to all other “affirmative” defences, including alibi, duress, provocation and others mentioned in Cinous, at para. 57, the persuasive and the evidential burdens are divided.

[56] As regards these “ordinary”, as opposed to “reverse onus” defences, the accused has no persuasive burden at all. Once the issue has been “put in play” (R. v. Schwartz, [1988] 2 S.C.R. 443), the defence will succeed unless it is disproved by the Crown beyond a reasonable doubt. Like all other disputed issues, however, defences of this sort will only be left to the jury where a sufficient evidential basis is found to exist. That foundation cannot be said to exist where its only constituent elements are of a tenuous, trifling, insignificant or manifestly unsubstantive nature: there must be evidence in the record upon which a properly instructed jury, acting judicially, could entertain a reasonable doubt as to the defence that has been raised.

[57] From a theoretical point of view, “reverse onus” defences and “ordinary affirmative defences” may thus be thought to be subject to different evidential burdens. But in this as in other branches of the law, pure logic must yield to experience and, without undue distortion of principle, to a more practical and more desirable approach. In determining whether the evidential burden has been discharged on any defence, trial judges, as a matter of judicial policy, should therefore always ask the very same question: Is there in the record any evidence upon which a reasonable trier of fact, properly instructed in law and acting judicially, could conclude that the defence succeeds?

[58] This requirement of a sufficient evidential foundation aims primarily to avoid wrongful convictions and unwarranted acquittals, while at the same time leaving it to the jury to discharge the responsibilities that are by law within its exclusive domain.

[59] It must be remembered in this latter regard that any person charged in Canada with an offence punishable by imprisonment for five years or more is constitutionally entitled to “the benefit of trial by jury”: see s. 11(f) of the Canadian Charter of Rights and Freedoms. And where the accused elects to be tried by a judge alone, the Attorney General may nonetheless require a jury trial if the offence charged is punishable by more than five years’ imprisonment: Criminal Code, s. 568.

[60] Like Arbour J., I think it appropriate to underline here the importance of “afford[ing] the jury the opportunity to pronounce on every factual issue” for which the required evidential foundation has been laid (Cinous, at para. 196).

[61] And I agree with her as well (at para. 200) that:

[T]he cost of risking a wrongful conviction and possibly violating the accused’s constitutionally protected rights by inadvisably withdrawing a defence from the jury is a high one.

(2) The Evidential Standard Applicable to Automatism

[62] With these principles in mind, I turn now to Stone, Arcuri, and Cinous, all previously cited.

[63] In matters of precedent, last things must be taken first, and I therefore begin with the joint reasons of McLachlin C.J. and Bastarache J. in Cinous before considering the reasons of Bastarache J., speaking for the majority, in Stone.

(i) The Evidential Threshold Applicable to all Defences in Light of Cinous

[64] The issue in Cinous was whether self-defence should have been left with the jury. The Chief Justice and Bastarache J. answered that question in the negative on the ground that the evidential burden for self-defence had not been discharged. L’Heureux-Dubé and LeBel JJ. concurred. Gonthier and Binnie JJ. likewise agreed with the reasons and the conclusion of the Chief Justice and Bastarache J., but added brief additional reasons of their own.

[65] In concluding as they did, the Chief Justice and Bastarache J. stated, at para. 49:

The test [for determining whether an evidential burden has been satisfied] is whether there is evidence on the record upon which a properly instructed jury acting reasonably could acquit. [References omitted.] This long-standing formulation of the threshold question for putting defences to the jury accords with the nature and purpose of the air of reality test. We consider that there is nothing to be gained by altering the current state of the law, in which a single clearly-stated test applies to all defences. [References omitted.] There is no need to invent a new test, to modify the current test, or to apply different tests to different classes of cases. [Emphasis added.]

[66] The Chief Justice and Bastarache J. later reiterated, at para. 57, that “a single air of reality test applies to all defences”. After citing instances in which the test has been applied, they added that “[a]dopting different evidential standards for different classes of cases would constitute a sharp break with the authorities”.

[67] Cinous is the decisive authority, as a matter both of sequence and of consequence, in this Court’s consideration and determination of the evidential burden governing all defences. And I find it helpful and appropriate, for that reason, to allow Cinous to speak for itself.

[68] In their joint reasons, the Chief Justice and Bastarache J. stated at paras. 52 and following:

It is trite law that the air of reality test imposes a burden on the accused that is merely evidential, rather than persuasive. Dickson C.J. drew attention to the distinction between these two types of burden in R. v. Schwartz, [1988] 2 S.C.R. 443, at p. 466:

Judges and academics have used a variety of terms to try to capture the distinction between the two types of burdens. The burden of establishing a case has been referred to as the “major burden,” the “primary burden,” the “legal burden” and the “persuasive burden”. The burden of putting an issue in play has been called the “minor burden,” the “secondary burden,” the “evidential burden,” the “burden of going forward,” and the “burden of adducing evidence”.

The air of reality test is concerned only with whether or not a putative defence should be “put in play”, that is, submitted to the jury for consideration. This idea was crucial to the finding in Osolin that the air of reality test is consistent with the presumption of innocence guaranteed by s. 11(d) of the Canadian Charter of Rights and Freedoms.

In applying the air of reality test, a trial judge considers the totality of the evidence, and assumes the evidence relied upon by the accused to be true. See Osolin, supra; Park, supra. The evidential foundation can be indicated by evidence emanating from the examination in chief or cross-examination of the accused, of defence witnesses, or of Crown witnesses. It can also rest upon the factual circumstances of the case or from any other evidential source on the record. There is no requirement that the evidence be adduced by the accused. See Osolin, supra; Park, supra; Davis, supra.

The threshold determination by the trial judge is not aimed at deciding the substantive merits of the defence. That question is reserved for the jury. See Finta, supra; R. v. Ewanchuk, [1999] 1 S.C.R. 330. The trial judge does not make determinations about the credibility of witnesses, weigh the evidence, make findings of fact, or draw determinate factual inferences. See R. v. Bulmer, [1987] 1 S.C.R. 782; Park, supra. Nor is the air of reality test intended to assess whether the defence is likely, unlikely, somewhat likely, or very likely to succeed at the end of the day. The question for the trial judge is whether the evidence discloses a real issue to be decided by the jury, and not how the jury should ultimately decide the issue. [Underlining in original; italics added.]

[69] The Chief Justice and Bastarache J. added, at paras. 62 and 68:

While courts have often spoken of an inquiry into whether there is “some evidence” or “any evidence” without qualification, this must be understood as a short‑form reference to the full question. The full question is whether there is evidence (some evidence, any evidence) upon which a properly instructed jury acting judicially could acquit. If there is any or some such evidence, then the air of reality hurdle is cleared. If there is no such evidence, then the air of reality hurdle is not cleared. . . .

McIntyre J. in Pappajohn plainly did not suggest that the air of reality test asks whether the evidence is sufficient to justify an acquittal. Rather, the inquiry was stated to be into whether there was evidence sufficient to warrant putting the issue of acquittal to the jury. The inquiry has always been about the sufficiency of the evidence in this narrow sense. [Emphasis in original.]

[70] Any prior confusion regarding the “air of reality” requirement has been dissipated by Cinous. With respect to all defences, the evidential burden is discharged if there is some evidence upon which a properly instructed jury acting reasonably could acquit on the basis of that defence: see Cinous, supra, especially at paras. 2, 47, 49, 60, 62, 68 and 74. The “air of reality” test adds nothing to this well-established standard and should not be used to raise the governing evidential threshold or to introduce a persuasive requirement.

[71] The operative question is sometimes said to be whether there is any evidence “upon which a properly instructed jury acting reasonably could acquit if it believed the evidence to be true” (Cinous, at para. 65 (emphasis added)).

[72] This formulation of the test signifies that the trial judge, in determining whether a defence is in play, must assume the truth of the evidence that tends to support it, leaving the reliability, credibility and weight of that evidence to be determined by the jury.

[73] It was not meant to suggest that the evidential burden for all defences will be discharged only if the relevant evidence must be believed in order for the defence to succeed. As McLachlin J. explained in her dissent in R. v. S. (W.D.), [1994] 3 S.C.R. 521, at p. 543, an accused is entitled to be acquitted on the basis of exculpatory evidence that the jury does not reject, but either accepts or about which it is undecided.

[74] In short, as regards all affirmative defences, I think it preferable to say that the evidential burden will be discharged where there is some evidence that puts the defence “in play”: see Cinous, at para. 52. And the defence will be in play whenever a properly instructed jury could reasonably, on account of that evidence, conclude in favour of the accused.

(ii) The Evidential Threshold at a Preliminary Inquiry and on a Defence Motion for Non-Suit or for a Directed Verdict of Acquittal

[75] In considering the evidential burden governing affirmative defences, I find it helpful to compare the evidential standard required of the Crown at preliminary inquiry or to defeat a defence motion for non-suit on a trial before judge alone or a directed verdict of acquittal on a trial before judge and jury.

[76] In Arcuri, supra, the issue before this Court was whether a judge at preliminary inquiry could weigh the evidence in assessing whether it was sufficient to justify committing an accused to trial. More specifically, the issue was whether the introduction of direct exculpatory evidence by the defence, when the Crown’s case was entirely circumstantial, had any effect on the way the determination of sufficiency was to be made.

[77] Writing for the Court, McLachlin C.J. confirmed in that case that the test for committal to trial is the same as on a motion for non-suit or for a directed verdict.

[78] In each instance, the decisive question is whether there exists an evidential foundation on the issue of guilt. The evidential burden, on that issue, as we have seen, is a function of the persuasive burden and the persuasive burden, which is borne by the Crown, can only be discharged by proof beyond a reasonable doubt.

[79] It would be incongruous to impose a more onerous evidential threshold^* in respect of a less onerous persuasive burden, as on a defence of mental disorder automatism, which is the issue that concerns us here.

(iii) The Evidential Burden Applicable to Automatism in Light of Stone

[80] I turn finally to Stone, supra, upon which the Crown relies, and begin by reiterating what I said at the outset: Stone, in its entirety, must be read in the light of Cinous.

[81] Writing for the majority in Stone, Bastarache J. found, at para. 170, that “voluntariness, rather than consciousness, is the key legal element of automatistic behaviour since a defence of automatism amounts to a denial of the voluntariness component of the actus reus”. After reviewing the policy concerns related to the defence of automatism, he held that, in light of these concerns, “the legal burden in cases involving claims of automatism must be on the defence to prove involuntariness on a balance of probabilities to the trier of fact” (para. 179).

[82] Bastarache J. then stated, at para. 182, that to meet the evidentiary burden in a case of automatism, “the defence must satisfy the trial judge that there is evidence upon which a properly instructed jury could find that the accused acted involuntarily on a balance of probabilities”.

[83] I concede that there is language in Stone that may be understood to invite an assessment by the trial judge as to the likely success of the defence. This, in turn, may be seen to require the judge to weigh the evidence in order to determine whether it establishes, on the balance of probabilities, that the accused perpetrated the criminal act charged in a state of automatism.

[84] That is what the trial judge did in this case and, as I mentioned at the outset, I agree with the Court of Appeal that he was not entitled to do so.

[85] The Court of Appeal, as appears from the reasons of Rousseau-Houle J.A., was alert to the concerns expressed in Stone by Bastarache J. And, as indicated earlier, there is language in Stone that can be read as requiring trial judges, before putting the defence of mental disorder automatism to the jury, to satisfy themselves that the accused has not just met the evidential burden but appears to have discharged the persuasive burden as well. The Chief Justice and Bastarache J. made it plain in Cinous, however, that this is not the state of the law.

[86] Trial judges are, of course, not prohibited from suggesting to jurors an analytical framework that may help them to determine for themselves whether a defence of mental disorder automatism has been made out on the required balance of probabilities.

[87] And I would add in this regard that Stone continues to provide for the benefit of triers of fact a helpful overview of factors they might wish to take into account in rendering their verdicts.

[88] The carefully crafted factors set out by Bastarache J. in that case can no longer, in the light of Cinous, be seen to inform the judge’s legal determination whether the required evidential burden has been met. For the triers of fact, however, they may be expected to afford both structure and guidance.

[89] Where mental disorder automatism is raised as a defence, an assertion of involuntariness on the part of the accused, supported by the logically probative opinion of a qualified expert, will normally provide — as it did in this case — a sufficient evidentiary foundation for putting the defence to the jury. By “logically probative”, I simply mean relevant — that is, evidence which, if accepted by the jury, would tend to support the defence of mental disorder automatism. Accompanying instructions in law will make it clear to the jury that the burden remains on the accused to establish the defence to the required degree of probability.

[90] Again, the ultimate issue is whether the jury is persuaded that the accused probably perpetrated the alleged criminal act in a state of automatism. For the issue to reach the jury, there must be some evidence upon which that finding could reasonably be made by a properly instructed jury acting judicially. If there is, the evidence is said to be “sufficient” — the evidentiary burden has been discharged and automatism is “in play” before the jury.

B. Application of the Law to the Facts

[91] What I have already said is sufficient to persuade me that the Court of Appeal did not err with respect to the evidential burden on a defence of mental disorder automatism. The remaining question is whether the Court of Appeal erred in law in concluding that the accused — the respondent in this Court — was entitled in this case to have his defence of mental disorder automatism put to the jury.

[92] In this regard, I repeat once more, the applicable test is whether there was any evidence in the record upon which the jury, properly instructed and acting judicially, could reasonably conclude that the defence of automatism by reason of mental disorder had been made out. In my view, the Court of Appeal did not err in finding that the respondent, through his own evidence and that of Dr. Talbot, had discharged his evidential burden on that issue.

[93] The respondent gave evidence tending to establish that he was acting involuntarily at the time of the offence. He also adduced expert evidence to support his own testimony. As the majority noted in Cinous, at paras. 88 and 108, the testimony of the respondent constitutes direct evidence with respect to the required evidential burden, provided that it does not amount to a mere allegation of the existence of a defence in law.

[94] In this case, the evidence clearly went beyond a mere allegation of the existence of a defence. It included a relatively detailed description of the respondent’s perception of the facts at the moment of the criminal act. The respondent also testified as to the circumstances that gave rise to his state of mind at the relevant time.

[95] Moreover, Dr. Talbot concluded that the respondent was suffering, at the time of the offence, from a psychotic episode induced by substance abuse. In his report, Dr. Talbot stated that the respondent had a serious mental disorder akin to psychosis, which seriously distorted his perception of reality. This, in turn, affected his judgment, and rendered him incapable of distinguishing right from wrong, legal from illegal.

[96] Dr. Talbot also testified that, at the moment of the events of February 15, 1999, the respondent began to perceive reality abnormally, in a projective and interpretive manner.

[97] Finally, Dr. Talbot explained that, in psychiatric jargon, this means that the respondent began to “see things” and to make pathological connections between people, situations and events.

[98] Taken as a whole, this evidence was in my view sufficient to discharge the respondent’s evidential burden on his defence of mental disorder automatism. Whether the respondent’s actions were in fact involuntary was a matter for the jury to decide.

[99] Accordingly, I find no error in the conclusions of the Court of Appeal in this regard.

VI. Disposition

[100] Like the Court of Appeal, I have concluded that the respondent’s defence of mental disorder automatism should have been put to the jury. I agree as well that the curative proviso of s. 686(1)(b)(iii) of the Criminal Code cannot be applied here.

[101] I would therefore dismiss the appeal.

Appeal dismissed.

Solicitor for the appellant: Attorney General’s Prosecutor, Sainte‑Foy.

Solicitor for the respondent: Sébastien St‑Laurent, Québec.

^* See Erratum [2004] 2 S.C.R. iv