R. v. Parks,  2 S.C.R. 871
Her Majesty The Queen Appellant
Kenneth James Parks Respondent
Indexed as: R. v. Parks
File No.: 22073.
1992: January 27; 1992: August 27.
Present: Lamer C.J. and La Forest, L'Heureux‑Dubé, Sopinka, Gonthier, Cory, McLachlin, Stevenson* and Iacobucci JJ.
on appeal from the court of appeal for ontario
Criminal law ‑‑ Defences ‑‑ Automatism (sleepwalking) ‑‑ Respondent killing and injuring while asleep ‑‑ Whether sleepwalking should be classified as non‑insane automatism resulting in an acquittal or as a "disease of the mind" (insane automatism) giving rise to the special verdict of not guilty by reason of insanity.
Respondent attacked his parents‑in‑law, killing one and seriously injuring the other. The incident occurred at their home, some 23 km. from respondent's residence, during the night while they were both asleep in bed. Respondent had driven there by car. Immediately after the incident, the respondent went to a nearby police station, again driving his own car, and told them what he had done.
Respondent claimed to have been sleepwalking throughout the incident. He had always been a deep sleeper and had a great deal of trouble waking up. The year prior to the incident was particularly stressful for the respondent and his personal life suffered. His parents‑in‑law were aware of his problems, supported him and had excellent relations with him. Additionally, several members of his family suffer or have suffered from sleep problems such as sleepwalking, adult enuresis, nightmares and sleeptalking.
The respondent was charged with first degree murder and attempted murder. At the trial respondent presented a defence of automatism. The testimony of five expert witnesses called by the defence was not contradicted by the Crown. This evidence was that respondent was sleepwalking and that sleepwalking is not a neurological, psychiatric or other illness. The trial judge put only the defence of automatism to the jury, which acquitted respondent of first degree murder and then of second degree murder. The judge then acquitted the respondent of the charge of attempted murder. The Court of Appeal unanimously upheld the acquittal. At issue here is whether sleepwalking should be classified as non‑insane automatism resulting in an acquittal or as a "disease of the mind" (insane automatism), giving rise to the special verdict of not guilty by reason of insanity.
Held (Lamer C.J. and Cory J. dissenting in part): The appeal should be dismissed.
Per La Forest, L'Heureux‑Dubé and Gonthier JJ.: The trial judge correctly left only the defence of non‑insane automatism with the jury. On this issue the findings of Lamer C.J. on the evidence were agreed with, but the distinction in law between insane and non‑insane automatism, particularly as it relates to somnambulism, required further comment. In distinguishing between automatism and insanity the trial judge must consider not only the evidence but also overarching policy considerations.
Automatism, although spoken of as a "defence", is conceptually a sub‑set of the voluntariness requirement, which in turn is part of the actus reus component of criminal liability. An involuntary act, including one committed in an automatistic condition entitles an accused to an unqualified acquittal, unless the automatistic condition stems from a disease of the mind that has rendered the accused insane. In the latter case, the accused is not entitled to a full acquittal, but to a verdict of insanity.
When a defence of non‑insane automatism is raised by the accused, the trial judge must determine whether the defence should be left with the trier of fact. This will involve two discrete tasks. First, he or she must determine whether there is some evidence on the record to support leaving the defence with the jury. An evidential burden rests with the accused; the mere assertion of the defence will not suffice.
Given the proper foundation, the trial judge must then consider whether the condition alleged by the accused is, in law, non‑insane automatism. If the trial judge is satisfied that there is some evidence pointing to a condition that is in law non‑insane automatism, then the defence can be left with the jury. The issue for the jury is one of fact: did the accused suffer from or experience the alleged condition at the relevant time? Because the Crown must always prove that an accused has acted voluntarily, the onus rests on the prosecution at this stage to prove the absence of automatism beyond a reasonable doubt.
The question of law at issue here, given that the accused laid the proper foundation for the defence of automatism, was whether sleepwalking should be classified as non‑insane automatism or a disease of the mind, thereby leaving only the defence of insanity for the accused. Under the Criminal Code everyone is presumed to be and to have been sane until the contrary is proved. If the accused pleads automatism, the Crown is entitled to raise the issue of insanity, but must then bear the burden of proving that the condition in question stems from a disease of the mind.
"Disease of the mind" is a legal term and not a medical term of art but it contains a substantial medical component as well as a legal or policy component. The medical component of the term, generally, is medical opinion as to how the mental condition in question is viewed or characterized medically. The legal or policy component relates to (a) the scope of the exemption from criminal responsibility to be afforded by mental disorder or disturbance, and (b) the protection of the public by the control and treatment of persons who have caused serious harms while in a mentally disordered or disturbed state.
Because "disease of the mind" is a legal concept, a trial judge cannot rely blindly on medical opinion. The judge must determine what mental conditions are included within the term "disease of the mind", and whether there is any evidence that the accused suffered from an abnormal mental condition comprehended by that term.
Two distinct approaches to the policy component of insanity have emerged in automatism cases, the "continuing danger" and "internal cause" theories. The first theory holds that any condition likely to present recurring danger should be treated as insanity. The second holds that a condition stemming from the internal make‑up of the accused, rather than external factors, should lead to a finding of insanity. Though seemingly divergent, both theories stem from a concern for the protection of the public.
Though the second theory has gained a certain ascendency, it is merely an analytical tool and is not universal. In particular, it is not helpful in assessing the nature of a somnambulistic condition. The distinction between internal and external causes is blurred during sleep, and certain causes that are discounted for a subject who is awake may have entirely different effects on a sleeping person. As for the "continuing danger" test, it has been criticized as a general theory. However, the purpose of the insanity defence has always been the protection of the public against recurrent danger. As such, the possibility of recurrence, though not determinative, may be looked upon as a factor at the policy stage of the inquiry on the issue of insanity.
On the evidence there is no likelihood of recurrent violent somnambulism. Moreover, none of the other policy considerations relevant to the distinction between insanity and automatism, for example, the floodgates argument, or that automatism can be feigned, is of concern in this case.
Our system of justice is predicated on the notion that only those who act voluntarily should be punished under the criminal law. Here, no compelling policy factors preclude a finding that the accused's condition was one of non‑insane automatism. As the Crown did not meet its burden of proving that somnambulism stems from a disease of the mind, committal under s. 614(2) of the Criminal Code is precluded, and the accused should be acquitted. However, because the medical evidence in each case impacts at several stages of the policy inquiry and is significant in its own right, sleepwalking in a different case on different evidence might be found to be a disease of the mind.
This matter should not be sent back to the trial judge for the possible imposition of an order to keep the peace. The judiciary is not practically equipped to administer such an order, and a number of practical reasons, in addition to those of Sopinka and McLachlin JJ., preclude its consideration. To be effective, any order to keep the peace would have to be permanent. This would violate established practice (if not the law) regarding peace orders, which requires a defined period for the order. It would also be unrealistic to expect respondent's family, who are the only persons able to monitor the order, to complain of any breach of the peace. Finally, it would be unreasonable to expect the respondent to bear the cost of a life‑long surety necessary to enforce such an order.
Per McLachlin and Iacobucci JJ.: The reasons of Lamer C.J., except on the question of referring the matter back to the trial judge for consideration as to whether an order to keep the peace should be imposed, and the reasons of La Forest and Sopinka JJ. were agreed with. Notwithstanding the justice of an acquittal here and the evidence that a recurrence is highly unlikely, great care should be taken to avoid the possibility of a similar episode in the future. An order restricting a person's liberty on account of an act for which he or she has been acquitted, however, raises difficult issues. It is inappropriate that respondent, given his courageous efforts to re‑establish his life over the past five years, should now be embroiled in a further set of proceedings concerned not with his guilt or innocence, but with the maintenance of his liberty. Generally, the courts do not grant remedies affecting the liberty of the subject unless asked to do so by the Crown. In the absence of an application by the Crown, the case should not be remitted for consideration of further measures against the accused.
Per Sopinka J.: The trial judge, for the reasons given by both Lamer C.J. and La Forest J., did not err in leaving the defence of automatism rather than that of insanity with the jury. This matter, however, should not be referred back to the trial judge to consider an order to keep the peace.
The common law preventative justice power has significant limits. It cannot be exercised on the basis of mere speculation but requires a proven factual foundation which raises a probable ground to suspect future misbehaviour. The uncontroverted expert evidence in this case is wholly inconsistent with such a conclusion.
The extent and continued validity of this common law power has yet to be considered in light of the Charter. The imposition of restrictive conditions following an acquittal on the basis of a remote possibility of recurrence may well be contrary to s. 7.
There is still the possibility of an information being laid pursuant to s. 810 of the Criminal Code, subject to the evidentiary basis "that the informant has reasonable grounds for his fears" and to constitutional challenge. Such a proceeding, however, should not be initiated by this Court acting proprio motu.
If the respondent remains subject to the criminal justice system, the issue on cross‑appeal of whether a stay should be entered by reason of a violation of s. 11(b) of the Charter would have to be considered.
Per Lamer C.J. and Cory J. (dissenting in part): The testimony revealed three very important points: (1) the respondent was sleepwalking at the time of the incident; (2) sleepwalking is not a neurological, psychiatric or other illness but rather is a sleep disorder very common in children and also found in adults; and, (3) there is no medical treatment as such, apart from good health practices, especially as regards sleep. This expert evidence was not in any way contradicted by the Crown, which had the advice of experts who were present during the testimony given by the defence experts and whom it chose not to call.
The defence of automatism ‑‑ rather than that of insanity ‑‑ was properly put to the jury. For a defence of insanity to have been put to the jury, together with or instead of a defence of automatism, as the case may be, there would have had to have been in the record evidence tending to show that sleepwalking was the cause of the respondent's state of mind. That was not the case here. This was not to say, however, that sleepwalking could never be a disease of the mind in another case on different evidence.
Notwithstanding respondent's acquittal, some control could be exercised to prevent a possible recurrence in a situation like this through the common law power to make an order to keep the peace which is vested in any judge or magistrate. The rules of natural justice must be observed in any exercise of this power. Exploring, on notice, the possibility of some minimally intrusive conditions to assure the community's safety would not infringe s. 7 of the Charter. Any condition imposed must be rationally connected to the apprehended danger posed by the person and go no further than necessary to protect the public from this danger.
By La Forest J.
Considered: Rabey v. The Queen,  2 S.C.R. 513; Bratty v. Attorney‑General for Northern Ireland,  A.C. 386; Cooper v. The Queen,  1 S.C.R. 1149; R. v. Swain,  1 S.C.R. 933; referred to: R. v. Quick,  3 All E.R. 347; R. v. Hennessy,  2 All E.R. 9; R. v. Sullivan,  A.C. 156; R. v. Burgess,  2 All E.R. 769; R. v. Edgar (1913), 109 L.T. 416.
By Sopinka J.
Referred to: Mackenzie v. Martin,  S.C.R. 361; R. v. White, Ex p. Chohan,  1 C.C.C. 19; Re Regina and Shaben,  2 O.R. 613; Stevenson v. Saskatchewan (Minister of Justice), (Court of Queen's Bench, June 8, 1987, unreported).
By Lamer C.J. (dissenting in part)
Distinguished: R. v. Sullivan,  2 All E.R. 673; R. v. Burgess,  2 All E.R. 769; referred to: Cooper v. The Queen,  1 S.C.R. 1149; Rabey v. The Queen,  2 S.C.R. 513; R. v. Hartridge,  1 C.C.C. 346; Bratty v. Attorney‑General for Northern Ireland,  A.C. 386; Ryan v. The Queen (1967), 40 A.L.J.R. 488; R. v. Cottle,  N.Z.L.R. 999; R. v. Ngang,  3 S.A.L.R. 363; R. v. Tolson (1889), 23 Q.B.D. 168; H. M. Advocate v. Fraser (1878), 4 Couper 70; Mackenzie v. Martin,  S.C.R. 361; Re Broomes and The Queen (1984), 12 C.C.C. (3d) 220.
Statutes and Regulations Cited
Canadian Charter of Rights and Freedoms, ss. 1, 7, 11(b).
Criminal Code, R.S.C., 1985, c. C‑46, ss. 16(2), (4), 614(2), 810.
Justices of the Peace Act, 1361 (Eng.), 34 Edw. 3, c. 1.
Black's Law Dictionary, 5th ed. St. Paul, Minn.: West Publishing Co., 1979.
Colvin, Eric. Principles of Criminal Law, 2nd ed. Calgary: Thomson Professional Publishing Canada, 1991.
Côté‑Harper, Gisèle, Antoine D. Manganas et Jean Turgeon. Droit pénal canadien, 3e éd. Cowansville: Yvon Blais Inc., 1989.
Fairall, Paul. "Automatism",  5 Crim. L.J. 335.
Fenwick, Peter. "Somnambulism and the Law: A Review" (1987), 5 Behavioral Sciences & the Law 343.
Gillies, Peter. Criminal Law. Sydney: Law Book Co., 1985.
Howard, Colin. Howard's Criminal Law, 5th ed. By Brent Fisse. Sydney: Law Book Co., 1990.
Martin, G. Arthur, Hon. "Mental Disorder and Criminal Responsibility in Canadian Law", in Stephen J. Hucker, Christopher D. Webster and Mark Ben‑Aron, eds., Mental Disorder and Criminal Responsibility. Toronto: Butterworths, 1981.
Mewett, Alan W. and Morris Manning. Criminal Law, 2nd ed. Toronto: Butterworths, 1985.
Roth, Sir Martin. "Modern Neurology and Psychiatry and the Problem of Criminal Responsibility", in Stephen J. Hucker, Christopher D. Webster and Mark H. Ben‑Aron, eds., Mental Disorder and Criminal Responsibility. Toronto: Butterworths, 1981.
Smith, John Cyril. Criminal Law, 6th ed. By J. C. Smith and Brian Hogan. London: Butterworths, 1988.
Stuart, Don. Canadian Criminal Law, 2nd ed. Toronto: Carswell, 1987.
Weller, Malcolm P. I. "Perchance to Dream" (1987), 137 New L.J. 52.
Williams, Glanville. Textbook of Criminal Law, 2nd ed. London: Stevens & Sons, 1983.
APPEAL from a judgment of the Ontario Court of Appeal (1990), 56 C.C.C. (3d) 449, dismissing an appeal from acquittal by Watt J. sitting with jury. Appeal dismissed, Lamer C.J. and Cory J. dissenting in part.
Gary T. Trotter and David Butt, for the appellant.
Marlys Edwardh, Clayton Ruby and Delmar Doucette, for the respondent.
The reasons of Lamer C.J. and Cory J. were delivered by
Lamer C.J. (dissenting in part) -- In the small hours of the morning of May 24, 1987 the respondent, aged 23, attacked his parents‑in‑law, Barbara Ann and Denis Woods, killing his mother‑in‑law with a kitchen knife and seriously injuring his father‑in‑law. The incident occurred at the home of his parents‑in‑law while they were both asleep in bed. Their residence was 23 km. from that of the respondent, who went there by car. Immediately after the incident, the respondent went to the nearby police station, again driving his own car. He told the police:
I just killed someone with my bare hands; Oh my God, I just killed someone; I've just killed two people; My God, I've just killed two people with my hands; My God, I've just killed two people. My hands; I just killed two people. I killed them; I just killed two people; I've just killed my mother‑ and father‑in‑law. I stabbed and beat them to death. It's all my fault.
At the trial the respondent presented a defence of automatism, stating that at the time the incidents took place he was sleepwalking. The respondent has always slept very deeply and has always had a lot of trouble waking up. The year prior to the events was particularly stressful for the respondent. His job as a project coordinator for Revere Electric required him to work ten hours a day. In addition, during the preceding summer the respondent had placed bets on horse races which caused him financial problems. To obtain money he also stole some $30,000 from his employer. The following March his boss discovered the theft and dismissed him. Court proceedings were brought against him in this regard. His personal life suffered from all of this. However, his parents‑in‑law, who were aware of the situation, always supported him. He had excellent relations with them: he got on particularly well with his mother‑in‑law, who referred to him as the "gentle giant". His relations with his father‑in‑law were more distant, but still very good. In fact, a supper at their home was planned for May 24 to discuss the respondent's problems and the solutions he intended to suggest. Additionally, several members of his family suffer or have suffered from sleep problems such as sleepwalking, adult enuresis, nightmares and sleeptalking.
The respondent was charged with the first degree murder of Barbara Ann Woods and the attempted murder of Denis Woods.
The trial judge chose to put only the defence of automatism to the jury, which first acquitted the respondent of first degree murder and then of second degree murder. The judge also acquitted the respondent of the charge of attempted murder for the same reasons. The Court of Appeal unanimously upheld the acquittal.
Court of Appeal (1990), 56 C.C.C. (3d) 449
The Court of Appeal affirmed the trial judgment, holding that the trial judge had properly put the defence of automatism rather than a defence of insanity to the jury. The Court of Appeal relied on the definition of "disease of the mind" by Dickson J. (as he then was) in Cooper v. The Queen,  1 S.C.R. 1149, at p. 1159:
In summary, one might say that in a legal sense "disease of the mind" embraces any illness, disorder or abnormal condition which impairs the human mind and its functioning, excluding however, self‑induced states caused by alcohol or drugs, as well as transitory mental states such as hysteria or concussion.
Galligan J.A. concluded from this that for there to be a "disease of the mind" within the meaning of s. 16(2) of the Criminal Code, R.S.C. 1970, c. C-34, "the impairment of the mind" must be caused by illness, a disorder or an abnormal condition. The court, at p. 468, held that it was not sleepwalking which created the state of mind in which the respondent found himself at the time of the incident, but sleep, and sleep is a normal condition:
Accepting the medical evidence, the respondent's mind and its functioning must have been impaired at the relevant time but sleep-walking did not impair it. The cause was the natural condition, sleep.
Accordingly, in Galligan J.A.'s view, for a defence of insanity to have been left with the jury the Crown would have had to present evidence that sleepwalking was the cause of the respondent's state of mind. That is not what the court held. The Court of Appeal therefore dismissed the appeal.
Did the Ontario Court of Appeal err in law in holding that the condition of sleepwalking should be classified as non‑insane automatism resulting in an acquittal instead of being classified as a "disease of the mind" (insane automatism), giving rise to the special verdict of not guilty by reason of insanity?
This Court has only ruled on sleepwalking in an obiter dictum in Rabey v. The Queen,  2 S.C.R. 513. The Court found that sleepwalking was not a "disease of the mind" in the legal sense of the term and gave rise to a defence of automatism. Should the Court maintain this position?
In Black's Law Dictionary (5th ed. 1979) automatism is defined as follows:
Behavior performed in a state of mental unconsciousness or dissociation without full awareness, i.e., somnambulism, fugues. Term is applied to actions or conduct of an individual apparently occurring without will, purpose, or reasoned intention on his part; a condition sometimes observed in persons who, without being actually insane, suffer from an obscuration of the mental faculties, loss of volition or of memory, or kindred affections . . . .
In Rabey this Court affirmed the judgment of the Ontario Court of Appeal (1977), 37 C.C.C. (2d) 461, in which Martin J.A. defined the expression "disease of the mind" at pp. 472-73:
"Disease of the mind" is a legal term, not a medical term of art; although a legal concept, it contains a substantial medical component as well as a legal or policy component.
The legal or policy component relates to (a) the scope of the exemption from criminal responsibility to be afforded by mental disorder or disturbance, and (b) the protection of the public by the control and treatment of persons who have caused serious harms while in a mentally disordered or disturbed state. The medical component of the term, generally, is medical opinion as to how the mental condition in question is viewed or characterized medically. Since the medical component of the term reflects or should reflect the state of medical knowledge at a given time, the concept of "disease of the mind" is capable of evolving with increased medical knowledge with respect to mental disorder or disturbance.
As Martin J.A. pointed out at p. 477, Canadian and foreign courts and authors have recognized that sleepwalking is not a disease of the mind:
Sleep‑walking appears to fall into a separate category. Unconscious behaviour in a state of somnambulism is non‑insane automatism . . . .
In Canada, see also R. v. Hartridge,  1 C.C.C. 346 (Sask. C.A.).
In Britain, Lord Denning in Bratty v. Attorney-General for Northern Ireland,  A.C. 386, at p. 409, recognized that sleepwalking gave rise to a defence of automatism:
No act is punishable if it is done involuntarily: and an involuntary act in this context ‑‑ some people nowadays prefer to speak of it as "automatism" ‑‑ means an act which is done by the muscles without any control by the mind, such as a spasm, a reflex action or a convulsion; or an act done by a person who is not conscious of what he is doing, such as an act done whilst suffering from concussion or whilst sleep‑walking.
Other foreign decisions have recognized the same principle: Ryan v. The Queen (1967), 40 A.L.J.R. 488; R. v. Cottle,  N.Z.L.R. 999; R. v. Ngang,  3 S.A.L.R. 363; R. v. Tolson (1889), 23 Q.B.D. 168; H. M. Advocate v. Fraser (1878), 4 Couper 70.
However, two British decisions seem to go against this line of authority: R. v. Sullivan,  2 All E.R. 673, and R. v. Burgess,  2 All E.R. 769. The comment in Sullivan at p. 677 was obiter, since the case concerned epilepsy:
If the effect of a disease is to impair these faculties so severely as to have either of the consequences referred to in the latter part of the [M`Naghten] rules, it matters not whether the aetiology of the impairment is organic, as in epilepsy, or functional, or whether the impairment itself is permanent or is transient and intermittent, provided that it subsisted at the time of commission of the act.
Some writers have interpreted this obiter as an indication that future cases of sleepwalking would only lead to a defence of insanity:
Although sleep‑walkers have always received an absolute acquittal for what they do, no social inconvenience has hitherto resulted. There seems to be no recorded instance of a sleep‑walker doing injury a further time after being acquitted. However, since the decision in Sullivan, to be discussed in the next section, it seems very likely that sleepwalkers will in future find themselves saddled with an insanity verdict.
(Williams, Textbook of Criminal Law (2nd ed. 1983), at p. 666.)
However, the evidence in the case at bar does not indicate the presence of an illness. Accordingly, I do not believe that this obiter can be applied to sleepwalking cases such as that of Mr. Parks. Burgess cannot be applied here for the same reason, but we will return to it later.
The following scholarly analysis may be consulted: Côté‑Harper, Manganas, Turgeon, Droit pénal canadien (3rd ed. 1989), at p. 473; Martin, "Mental Disorder and Criminal Responsibility in Canadian Law", in Hucker, Webster and Ben-Aron, eds., Mental Disorder and Criminal Responsibility (1981), at p. 23; Mewett and Manning, Criminal Law (2nd ed. 1985), at p. 301; Gillies, Criminal Law, at p. 205, Howard's Criminal Law (5th ed. 1990), at pp. 424‑25; Smith and Hogan, Criminal Law (6th ed. 1988), at pp. 40 and 42; Williams, supra, at pp. 665‑66; Fairall, "Automatism",  5 Crim. L.J. 335, at pp. 341‑42, Weller, "Perchance to Dream" (1987), 137 New L.J. 52.
In the case at bar the trial judge first reviewed the case law and scholarly analysis and said he did not intend to go against it:
In Rabey, supra, Martin J.A. considered somnambulism or sleep‑walking to be a special category or case of non‑insane automatism, one that perhaps could not be justified in accordance with a strict application of principles invoked to determine whether a condition from which an accused suffers amounts to "a disease of the mind" within s‑s. 16(2) of the Criminal Code. Quite simply put, and notwithstanding that the observations concerning the legal characterization of sleep‑walking as a separate category of non‑insane automatism would not appear to have been necessary to a decision of the issue on appeal in Rabey, supra, I am not prepared to depart from the pronouncement of such an eminent authority as Martin J.A. on matters concerning the scope of criminal responsibility. The statement there made is, as one might expect, amply supported by the jurisprudence and academic writings upon the issue.
He then considered the facts of the instant case:
In the circumstances of the present case, it is doubtful whether the sleep disorder from which the accused suffers would constitute a disease of the mind under s‑s. 16(2) in accordance with general principle.
I therefore propose to review the evidence in this matter. A large part of the defence evidence in this case was medical evidence. Five physicians were heard: Dr. Roger James Broughton, a neurophysiologist and specialist in sleep and sleep disorders, Dr. John Gordon Edmeads, a neurologist, Dr. Ronald Frederick Billings, a psychiatrist, Dr. Robert Wood Hill, a forensic psychiatrist, and finally, Dr. Frank Raymond Ervin, a neurologist and psychiatrist.
The medical evidence in the case at bar showed that the respondent was in fact sleepwalking when he committed the acts with which he is charged. All the expert witnesses called by the defence said that in their opinion Parks was sleepwalking when the events occurred. This is what Dr. Broughton said:
Q. . . . assuming for a moment that Mr. Parks caused the death of Barbara Woods, did you, sir, reach an opinion as to his condition at the time he caused that death?
A. Yes. My opinion is that he did it during a sleepwalking episode.
Though sceptical at the outset, the expert witnesses unanimously stated that at the time of the incidents the respondent was not suffering from any mental illness and that, medically speaking, sleepwalking is not regarded as an illness, whether physical, mental or neurological:
Q. Dr. Billings, just a couple more questions. In conclusion, if I can bring you to some global conclusion for a moment, and leaving aside sleepwalking itself, on May the 24th is it your opinion, or do you have an opinion as to whether or not Mr. Parks suffered from any mental illness?
A. On May 24th?
Q. Dealing now with sleepwalking, from the perspective of general psychiatry, is sleepwalking viewed as a neurological disease?
A. From a psychiatrist's point of view?
Q. Yes, from a psychiatrist's point of view.
Q. Is it viewed as something that is causally related to mental illness?
A. Can cause mental illness?
Q. No. Is sleepwalking ‑‑
A. ‑‑ a result of mental illness?
Q. ‑‑ a result of mental illness?
Q. Is sleepwalking a part of any mental illness?
Q. In your opinion, Dr. Billings, is sleepwalking a disease of the mind?
A. No, I would not call it a disease.
They also unanimously stated that a person who is sleepwalking cannot think, reflect or perform voluntary acts:
Q. Is there any evidence that a person could formulate a plan while they were awake and then in some way ensure that they carry it out in their sleep?
A. No, absolutely not. No. Probably the most striking feature of what we know of what goes on in the mind during sleep is that it's very independent of waking mentation in terms of its objectives and so forth. There is a lack of control of directing our minds in sleep compared to wakefulness. In the waking state, of course, we often voluntarily plan things, what we call volition ‑ that is, we decide to do this as opposed to that ‑ and there is no evidence that this occurs during the sleepwalking episode. There usually is ‑ well, they are precipitated. They are part of an arousal, an incomplete arousal process during which all investigators have concluded that volution [sic] is not present.
. . .
Q. And assuming he was sleepwalking at the time, would he have the capacity to intend?
Q. Would he have appreciated what he was doing?
A. No, he would not.
Q. Would he have understood the consequences of what he was doing?
A. No, he would not.
Q. Would he have been able to stop what he was doing?
A. No, I do not believe that he would. I think it would all have been an unconscious activity, uncontrolled and unmeditated.
The evidence also disclosed that sleepwalking was very common, almost universal, among children, and that 2 to 2.5% of "normal" adults had sleepwalked at least once. Dr. Hill further noted that he found it significant that there were several sleepwalkers in the respondent's family:
Thirdly, I think, as I indicated, it turns out, as enquiries are made more and more, that there is a significant history in the background family of Mr. Parks of difficulties, of bedwetting difficulties, of sleeptalking and sleepwalking, that is in keeping with what we know about the phenomena of sleepwalking. We know that there are often family members so affected and that was present.
Dr. Broughton, for his part, indicated that he had never known of sleepwalkers who had acted violently who had repeated this kind of behaviour:
Q. Yes. Now, with respect to Mr. Parks, do you have any opinion, sir, as to the probability of a recurrence of an event of sleepwalking with serious aggression involving physical harm to others?
A. I think the risk of that is infinitesimal, I don't think it would exceed the risk of the general population almost. He has the family predisposition to sleepwalk, but it would only be in the likelihood of all precipitating and extenuating and so forth factors that built up to this crisis that would theoretically have to almost reappear.
Q. And even if they were to reappear, would there be any probability of another homicidal event?
A. It would still ‑‑ As I say, there are no reported cases in the literature, so there is essentially ‑‑ The probability of it occurring is not statistically significant. It is just absolutely improbable.
In cross‑examination he also added that sleepwalking episodes in which violent acts are committed are not common:
Q. And does that, in fact, agree with your own experience with respect to people that you have dealt with at the sleep lab and have seen over the years, that the majority of sleepwalking episodes generally involve what you call trivial behaviour?
A. It is well known that aggression during sleepwalking is quite rare.
Q. How many cases of aggression during sleepwalking have you personally deal[t] with or been involved with at your sleep lab?
A. In the last ‑‑ Perhaps a total of five or six. In the last five years we have seen three.
Further, on being questioned about a cure or treatment, Dr. Broughton answered that the solution was sleep hygiene, which involved eliminating factors that precipitated sleepwalking such as stress, lack of sleep and violent physical exercise:
Q. And with respect to the sleepwalking phenomenon you have described or the disorder or arousal you have seen, is there any treatment available?
A. Yes. There are a number of treatment approaches that are used for sleepwalking. It's not a type of condition where there is one sort of universally applied treatment.
Q. And can you give us an example of what kind of treatments are available?
. . .
A. There is no specific, let's say pill or specific medication that you can give which will eradicate the sleepwalking. The best treatment procedure generally is to try an [sic] avoid precipitating factors, to stabilize sleep, to avoid sleep deprivation, the various things that ‑ medication that might lead to an attack and so forth.
Q. Are the regimes for stabilizing sleep, etc., complicated regimes or are they . . .
A. No, no, they are not. They are basically simple sleep hygiene and rules.
Q. And what do you mean by sleep hygiene?
A. Well, going to bed at a regular hour, getting sufficient sleep, having availability of sufficient exercise and so on that you are tired enough at the end of the day that your body wants to go to sleep, avoiding getting overweight and obese and things that could impair sleep, avoiding alcohol.
Dr. Ervin in his turn stated that during the slow wave sleep stage the cortex, which is the part of the brain that controls thinking and voluntary movement, is essentially in coma. When a person is sleepwalking, the movements he makes are controlled by other parts of the brain and are more or less reflexive:
We put our recording electrodes on the top of the head after all so we are looking at the cortical matter and what is happening there. That part of the brain is effectively in coma, that is, it is highly synchronized, very slow. It looks like the ocean waves rolling along, suggesting that all those nerve cells are no longer doing their busy integrated ‑‑ or they are cut off. They are not working. What is left? What is left is those deep structures evolved some time back, evolved very competently in lower animals to handle the whole set of problems of moving about in the world and responding to stimuli reflexly [sic] more or less, going places, eating things, doing things and so on.
Three very important points emerge from this testimony: (1) the respondent was sleepwalking at the time of the incident; (2) sleepwalking is not a neurological, psychiatric or other illness: it is a sleep disorder very common in children and also found in adults; (3) there is no medical treatment as such, apart from good health practices, especially as regards sleep. It is important to note that this expert evidence was not in any way contradicted by the prosecution, which as the trial judge observed did have the advice of experts who were present during the testimony given by the defence experts and whom it chose not to call.
The Crown, for its part, relied on a decision of the English Court of Appeal, R. v. Burgess, supra, in which the Court held that sleepwalking was a mental illness. It is worth noting here, however, that the evidence in Burgess was completely different from or even contradictory to that presented in the case at bar.
The facts in Burgess are more or less similar to those at issue here. Burgess and a friend fell asleep watching a video. The friend woke up when she felt a blow on the head. Burgess was facing her, holding the video recorder in the air, about to strike her on the head with it, and he did so. Burgess, who woke up immediately after the incident, testified that he did not remember having hit her. He presented a defence of automatism, which the judge rejected. He was acquitted on grounds of insanity and appealed this judgment. Nevertheless, while the facts are similar the medical evidence was very different. Expert witnesses were called. The first witness, a Dr. D'Orban, agreed that Burgess was sleepwalking, but regarded this as a pathological condition. Another expert, called by the Crown, Dr. Fenwick, said that in his opinion this was not sleepwalking but a "hysterical dissociative state". The following is a passage from this judgment at pp. 775-76 which states the situation very clearly:
One turns then to examine the evidence upon which the judge had to base his decision and for this purpose the two medical experts called by the defence are the obvious principal sources. Dr d'Orban in examination‑in‑chief said:
`On the evidence available to me, and subject to the results of the tests when they became available, I came to the same conclusion as Dr Nicholas and Dr Eames, whose reports I had read, and that was that (the appellant's) actions had occurred during the course of a sleep disorder.'
He was asked . . . . in cross‑examination:
Q. `Would you go so far as to say that it was liable to recur? A. It is possible for it to recur, yes.
Judge Lewis. Is this a case of automatism associated with a pathological condition or not? A. I think the answer would have to be Yes, because it is an abnormality of the brain function, so it would be regarded as a pathological condition.'
. . .
The prosecution, as already indicated, called Dr Fenwick, whose opinion was that this was not a sleepwalking episode at all. If it was a case where the appellant was unconscious of what he was doing, the most likely explanation was that he was in what is described as a hysterical dissociative state. . . .
He then went on to describe features of sleepwalking. This is what he said:
. . .
Finally, should a person be detained in hospital? The answer to that is: Yes, because sleepwalking is treatable. Violent night terrors are treatable. There is a lot which can be done for the sleepwalker, so sending them to hospital after a violent act to have their sleepwalking sorted out, makes good sense.'
In my view, therefore, that case is clearly distinguishable from the one at bar. I am of the view that in the instant case, based on the evidence and the testimony of the expert witnesses heard, the trial judge did not err in leaving the defence of automatism rather than that of insanity with the jury, and that the instant appeal should be dismissed. For a defence of insanity to have been put to the jury, together with or instead of a defence of automatism, as the case may be, there would have had to have been in the record evidence tending to show that sleepwalking was the cause of the respondent's state of mind. As we have just seen, that is not the case here. This is not to say that sleepwalking could never be a disease of the mind, in another case on different evidence.
As I see it, however, that does not end the matter. Although the expert witnesses were unanimous in saying that sleepwalkers are very rarely violent, I am still concerned by the fact that as the result of an acquittal in a situation like this (and I am relieved that such cases are quite rare), the accused is simply set free without any consideration of measures to protect the public, or indeed the accused himself, from the possibility of a repetition of such unfortunate occurrences. In the case of an outright acquittal, should there not be some control? And if so, how should this be done? I am of the view that such control could be exercised by means of the common law power to make an order to keep the peace vested in any judge or magistrate. This power of "preventive justice" has been recognized in England for centuries and has its origin in one or more sources:
The cases do, however, in tracing the history of the law, suggest that it is derived from one or more sources:
(i) The common law;
(ii) The statute law, being the Justices of the Peace Act, 1361 (Imp.), c. 1 (hereinafter the "Statute of Edward III"); and/or
(iii) The form of commission which the justice of the peace is required to take in England.
In Canada this power has already been used in Ontario and British Columbia and was recognized by this Court in 1954 in Mackenzie v. Martin,  S.C.R. 361, at pp. 368-69:
In my view the common law preventive justice was in force in Ontario; s‑s. [(2)] of s. 748, or any other provision of the Criminal Code to which our attention was directed, does not interfere with the use of that jurisdiction, and the respondent was intending to exercise it. He, therefore, had jurisdiction over the subject‑matter of the complaint, and did not exceed it.
In exercising this power, the rules of natural justice must be observed and in this regard the more recent decision of the Ontario Court of Appeal, Re Broomes and The Queen (1984), 12 C.C.C. (3d) 220, is of particular interest for these purposes. A judge who acquitted an accused of assault decided, however, to exercise this "preventive justice" and made an order binding over the accused to keep the peace on certain conditions. On appeal, the accused argued that he had been denied the rules of natural justice because he was not told in advance that such an order would be made. Steele J. of the Ontario High Court of Justice dismissed the appeal, relying on an English decision (at p. 221):
I accept the decision in R. v. Woking Justices, Ex p. Gossage,  2 All ER 621 at p. 623 (Eng. C.A.), where Lord Widgery C.J. stated as follows:
It seems to me that a very clear distinction is drawn between, on the one part, persons who come before the justices as witnesses, and on the other, persons who come before the justices as defendants. Not only do the witnesses come with no expected prospect of being subjected to any kind of penalty, but also the witnesses as such, although they may speak in evidence, cannot represent themselves through counsel and cannot call evidence on their own behalf. By contrast, the defendant comes before the court knowing that allegations are to be made against him, knowing that he can be represented if appropriate, and knowing that he can call evidence if he wishes. It seems to me that a rule which requires a witness to be warned of the possibility of a binding‑over should not necessarily apply to a defendant in that different position.
I think from the extracts from Lord Parker C.J.'s judgment that I have read, Lord Parker C.J. would have taken the same view; but, be that as it may, it seems to me to be putting it far too high in the case of an acquitted defendant to say that it is a breach of the rules of natural justice not to give him an indication of the prospective binding‑over before the binding‑over is imposed. That is not to say that it would not be wise, and indeed courteous in these cases for justices to give such a warning; there certainly would be absolutely no harm in a case like the present if the justices, returning to court, had announced they were going to acquit, but had immediately said "We are however contemplating a binding‑over; what have you got to say?" I think it would be at least courteous and perhaps wise that that should be done, but I am unable to elevate the principle to the height at which it can be said that a failure to give such a warning is a breach of the rules of natural justice. [Emphasis added.]
Accordingly, such a power exists. The question remains whether it should be exercised in the case of the respondent Parks, or at least whether its exercise should be considered. I am of the view that this approach should be considered. As I have already said, despite the unanimous and uncontradicted evidence that the chances of such an occurrence taking place again are for all practical purposes nil, I feel that all necessary measures should be taken to ensure that such an event does not recur. After all, before this tragic incident occurred, the probability of Mr. Parks' killing someone while in a somnambulistic state was infinitesimal. Yet this is precisely what took place. Furthermore, the evidence at trial was not adduced with a view to determining whether an order would be justified and to determine the appropriate conditions of such an order. Thus, for example, an order might be made requiring Parks to do certain things suggested by a specialist in sleep disorders, for example to report to him periodically. In appropriate cases of outright acquittals on grounds of automatism measures that would reinforce sleep hygiene and thereby provide greater safety for others should always be considered. If the trial judge considers that making such an order would be in the interest of the public, he should so advise the parties and consider whatever evidence and submissions are tendered. In those situations where an order is made, it should be complied with in the same way as any other order of the court.
If conditions should be imposed on Mr. Parks they will restrict his liberty. It follows that the decision to impose such conditions and the terms of those conditions should not violate the rights guaranteed under s. 7 of the Canadian Charter of Rights and Freedoms. However, such a hearing is justified, as the sleepwalker has, although innocently, committed an act of violence which resulted in the death of his mother-in-law. Members of the community may quite reasonably be apprehensive for their safety. In those circumstances it cannot be said that the Court has unduly intruded upon the liberty of the accused by exploring, on notice to the accused, the possibility of imposing some minimally intrusive conditions which seek to assure the safety of the community. If conditions are imposed, then they obviously must be rationally connected to the apprehended danger posed by the person and go no further than necessary to protect the public from this danger.
I would therefore refer this matter back to the trial judge so that he can hear the parties on this point and decide, upon the evidence before him, whether such an order is appropriate. If this proves to be the case, it will be up to the trial judge to determine the content of the order.
I would accordingly dismiss this appeal and uphold the acquittal of the respondent but refer the matter back to the trial judge for him to decide on the making of an order to keep the peace on certain conditions, pursuant to the "preventive justice" power which he possesses.
As there is no new trial being ordered, the issues raised in the cross-appeal are best left to be dealt with in another case. This disposition of the cross-appeal is not to be considered as determinative of any motion the respondent might want to make to the trial judge, as regards the hearing to be held, seeking a stay of proceedings under s. 24(1) of the Charter alleging a violation of his s. 11(b) Charter rights.
//La Forest J.//
The judgment of La Forest, L'Heureux-Dubé and Gonthier JJ. was delivered by
La Forest J. -- I have had the advantage of reading the reasons of the Chief Justice. I agree with him that the trial judge was correct in leaving only the defence of non‑insane automatism with the jury. I am also in agreement with what the Chief Justice has to say on that issue, but I wish to add the following comments concerning the distinction in law between insane and non‑insane automatism, particularly as it relates to somnambulism.
In his reasons, the Chief Justice finds that the evidence and expert testimony from the trial of the accused support the trial judge's decision to instruct the jury on non‑insane automatism. I agree with this finding, but in my view that is not the end of the matter. In distinguishing between automatism and insanity the trial judge must consider more than the evidence; there are overarching policy considerations as well. Of course, the evidence in each case will be highly relevant to this policy inquiry.
Automatism occupies a unique place in our criminal law system. Although spoken of as a "defence", it is conceptually a sub‑set of the voluntariness requirement, which in turn is part of the actus reus component of criminal liability. A useful introduction is found in the dissenting reasons of Dickson J. (as he then was) in Rabey v. The Queen,  2 S.C.R. 513, at p. 522:
Although the word "automatism" made its way but lately to the legal stage, it is basic principle that absence of volition in respect of the act involved is always a defence to a crime. A defence that the act is involuntary entitles the accused to a complete and unqualified acquittal. That the defence of automatism exists as a middle ground between criminal responsibility and legal insanity is beyond question. Although spoken as a defence, in the sense that it is raised by the accused, the Crown always bears the burden of proving a voluntary act.
One qualification to this statement should be noted. When the automatistic condition stems from a disease of the mind that has rendered the accused insane, then the accused is not entitled to a full acquittal, but to a verdict of insanity; see Bratty v. Attorney-General for Northern Ireland,  A.C. 386 (H.L.), at pp. 403‑4 and 414. The condition in that instance is referred to as insane automatism, and the distinction between it and non‑insane automatism is the crucial issue in this appeal.
When a defence of non‑insane automatism is raised by the accused, the trial judge must determine whether the defence should be left with the trier of fact. This will involve two discrete tasks. First, he or she must determine whether there is some evidence on the record to support leaving the defence with the jury. This is sometimes referred to as laying the proper foundation for the defence; see Bratty, supra, at pp. 405 and 413. Thus an evidential burden rests with the accused, and the mere assertion of the defence will not suffice; see Bratty, at p. 414. Dickson J. summarized the point in comprehensive fashion in the following passage in Rabey, at p. 545:
The prosecution must prove every element of the crime charged. One such element is the state of mind of the accused, in the sense that the act was voluntary. The circumstances are normally such as to permit a presumption of volition and mental capacity. That is not so when the accused, as here, has placed before the court, by cross‑examination of Crown witnesses or by evidence called on his own behalf, or both, evidence sufficient to raise an issue that he was unconscious of his actions at the time of the alleged offence. No burden of proof is imposed upon an accused raising such defence beyond pointing to facts which indicate the existence of such a condition. . . .
If the proper foundation is present the judge moves to the second task: he or she must consider whether the condition alleged by the accused is, in law, non‑insane automatism. If the trial judge is satisfied that there is some evidence pointing to a condition that is in law non‑insane automatism, then the defence can be left with the jury; see Rabey, per Ritchie J., at p. 519. The issue for the jury is one of fact: did the accused suffer from or experience the alleged condition at the relevant time? Because the Crown must always prove that an accused has acted voluntarily, the onus rests on the prosecution at this stage to prove the absence of automatism beyond a reasonable doubt.
In the present case, there is no question that the accused has laid the proper foundation for the defence of automatism. The expert testimony reviewed by the Chief Justice is more than adequate on that score. At issue here is the question of law: is sleepwalking properly classified as non‑insane automatism, or does it stem from a disease of the mind, thereby leaving only the defence of insanity for the accused? When considering this question, s. 16(4) of the Criminal Code, R.S.C., 1985, c. C-46, should be recalled: "Every one shall, until the contrary is proved, be presumed to be and to have been sane." If the accused pleads automatism, the Crown is then entitled to raise the issue of insanity, but the prosecution then bears the burden of proving that the condition in question stems from a disease of the mind; see Rabey, supra, at pp. 544‑45.
In Canada, the approach to distinguishing between insane and non‑insane automatism was settled by this Court's judgment in Rabey. The majority in that case endorsed the reasons of Martin J.A. in the Ontario Court of Appeal (1977), 37 C.C.C. (2d) 461, and these latter reasons provide, at pp. 472‑73, what has become the accepted formula for determining whether a mental condition stems from a disease of the mind:
Although the term "disease of the mind" is not capable of precise definition, certain propositions may, I think, be asserted with respect to it. "Disease of the mind" is a legal term, not a medical term of art; although a legal concept, it contains a substantial medical component as well as a legal or policy component.
The legal or policy component relates to (a) the scope of the exemption from criminal responsibility to be afforded by mental disorder or disturbance, and (b) the protection of the public by the control and treatment of persons who have caused serious harms while in a mentally disordered or disturbed state. The medical component of the term, generally, is medical opinion as to how the mental condition in question is viewed or characterized medically. Since the medical component of the term reflects or should reflect the state of medical knowledge at a given time, the concept of "disease of the mind" is capable of evolving with increased medical knowledge with respect to mental disorder or disturbance.
Because "disease of the mind" is a legal concept, a trial judge cannot rely blindly on medical opinion. On this point Martin J.A. states the following, at pp. 473‑74:
If the question what particular mental conditions or mental disorders constitute disease of the mind were to be determined by the opinion of medical witnesses, then the scope of the defence of insanity under s. 16 of the Code would vary according to the choice of expert witnesses called to testify, since the existence of disease of the mind, apart from natural imbecility, constitutes the necessary foundation for insanity, and it is abundantly clear that medical opinions differ as to what mental conditions constitute a disease of the mind.
I take the true principle to be this: It is for the Judge to determine what mental conditions are included within the term "disease of the mind", and whether there is any evidence that the accused suffered from an abnormal mental condition comprehended by that term. The evidence of medical witnesses with respect to the cause, nature and symptoms of the abnormal mental condition from which the accused is alleged to suffer, and how that condition is viewed and characterized from the medical point of view, is highly relevant to the judicial determination of whether such a condition is capable of constituting a "disease of the mind". The opinions of medical witnesses as to whether an abnormal mental state does or does not constitute a disease of the mind are not, however, determinative, since what is a disease of the mind is a legal question . . . .
This position is beyond dispute, as similar statements were expressed by both the majority and minority judgments rendered by this Court; see Rabey, supra, at pp. 519 and 532‑33.
Another problem with relying solely on medical opinion is the lack of consensus within the medical community on the scope and meaning of "mental disease". This point was underscored by Dickson J. in Cooper v. The Queen,  1 S.C.R. 1149, where, at p. 1154, he stated:
Even medical experts are not given to agreement when asked to define "disease of the mind". In "The Concept of Mental Disease in Criminal Law Insanity Tests" (1965‑66) 33 U. Chic. L.R. 229, H. Fingarette illustrates the diversity in approach taken by psychiatric authorities in the quest for a definition of mental disease. He cites the following medical views, at p. 232‑3:
(1)There is no such medical entity as mental disease, or we would do well not to use the phrase.
(2)Mental disease is psychosis but not neurosis.
(3)Mental disease is any significant and substantial mental disturbance, or is any condition at all which is authoritatively dealt with by the psychiatrist or physician treating mental conditions.
(4)Mental disease means substantial social maladaptation or incompetence or both as judged by legal criteria.
(5)Mental disease is the failure to realize one's nature, capacities or true self.
In part because of the imprecision of medical science in this area, the legal community reserves for itself the final determination of what constitutes a "disease of the mind". This is accomplished by adding the "legal or policy component" to the inquiry.
A review of the cases on automatism reveals two distinct approaches to the policy component of the disease of the mind inquiry. These may be labelled the "continuing danger" and "internal cause" theories; see Colvin, Principles of Criminal Law (2nd ed. 1991), at p. 293. At first glance these approaches may appear to be divergent, but in fact they stem from a common concern for public safety. This was recognized by Martin J.A. who referred to "protection of the public" as a focus of the policy inquiry. More recently, the Chief Justice had occasion to comment on this aspect of the insanity provisions of the Criminal Code, albeit in a division of powers context, in R. v. Swain,  1 S.C.R. 933, at p. 998:
It is true that the dominant characteristic of these provisions is not punishment; however, neither is it treatment. The "pith and substance" of the legislative scheme dealing with individuals acquitted by reason of insanity is the protection of society from dangerous people who have engaged in conduct proscribed by the Criminal Code through the prevention of such acts in the future. While treatment may be incidentally involved in the process, it is not the dominant objective of the legislation.
The continuing danger theory holds that any condition likely to present a recurring danger to the public should be treated as insanity. The internal cause theory suggests that a condition stemming from the psychological or emotional make-up of the accused, rather than some external factor, should lead to a finding of insanity. The two theories share a common concern for recurrence, the latter holding that an internal weakness is more likely to lead to recurrent violence than automatism brought on by some intervening external cause.
It would appear that the internal cause approach has gained a certain ascendancy in both Canadian and English jurisprudence. The theory was the basis for deciding Rabey, where the distinction was described by Martin J.A., at pp. 477‑78, as follows:
In general, the distinction to be drawn is between a malfunctioning of the mind arising from some cause that is primarily internal to the accused, having its source in his psychological or emotional make‑up, or in some organic pathology, as opposed to a malfunctioning of the mind which is the transient effect produced by some specific external factor such as, for example, concussion. Any malfunctioning of the mind, or mental disorder having its source primarily in some subjective condition or weakness internal to the accused (whether fully understood or not), may be a "disease of the mind" if it prevents the accused from knowing what he is doing, but transient disturbances of consciousness due to certain specific external factors do not fall within the concept of disease of the mind. . . . Particular transient mental disturbances may not, however, be capable of being properly categorized in relation to whether they constitute "disease of the mind", on the basis of a generalized statement, and must be decided on a case by case basis.
The theory has also been adopted in England, first in R. v. Quick,  3 All E.R. 347 (C.C.A.), at p. 356, and most recently in R. v. Hennessy,  2 All E.R. 9 (C.C.A.), where Lord Lane C.J. stated the approach as follows, at p. 13:
The question in many cases, and this is one such case, is whether the function of the mind was disturbed on the one hand by disease or on the other hand by some external factor.
The judgments in both Rabey and Hennessy are careful to state that the internal cause theory is not a universal approach to the disease of the mind inquiry. Indeed Martin J.A., at p. 477, appears to suggest that sleepwalking is one of those conditions that is not usefully assessed on this basis.
The internal cause approach has been criticized as an unfounded development of the law, and for the odd results the external/internal dichotomy can produce; see Williams, Textbook of Criminal Law (2nd ed. 1983), at pp. 671‑76; Stuart, Canadian Criminal Law (2nd ed. 1987), at pp. 92-94; Colvin, supra, at p. 291. These criticisms have particular validity if the internal cause theory is held out as the definitive answer to the disease of the mind inquiry. However, it is apparent from the cases that the theory is really meant to be used only as an analytical tool, and not as an all‑encompassing methodology. As Watt J. commented in his reasons in support of his charge to the jury in this case, the dichotomy "constitutes a general, but not an unremitting or universal, classificatory scheme for `disease of the mind'".
As Martin J.A. suggested in Rabey, somnambulism is an example of a condition that is not well suited to analysis under the internal cause theory. The poor fit arises because certain factors can legitimately be characterized as either internal or external sources of automatistic behaviour. For example, the Crown in this case argues that the causes of the respondent's violent sleepwalking were entirely internal, a combination of genetic susceptibility and the ordinary stresses of everyday life (lack of sleep, excessive afternoon exercise, and a high stress level due to personal problems). These "ordinary stresses" were ruled out as external factors by this Court in Rabey (although by a narrow majority). However, the factors that for a waking individual are mere ordinary stresses can be differently characterized for a person who is asleep, unable to counter with his conscious mind the onslaught of the admittedly ordinary strains of life. One could argue that the particular amalgam of stress, excessive exercise, sleep deprivation and sudden noises in the night that causes an incident of somnambulism is, for the sleeping person, analogous to the effect of a concussion upon a waking person, which is generally accepted as an external cause of non‑insane automatism; see Williams, supra, at p. 666. In the end, the dichotomy between internal and external causes becomes blurred in this context, and is not helpful in resolving the inquiry.
The continuing danger approach stems from an obiter comment of Lord Denning in Bratty, supra, at p. 412, where he proposes the following test for distinguishing between insane and non‑insane automatism:
It seems to me that any mental disorder which has manifested itself in violence and is prone to recur is a disease of the mind. At any rate it is the sort of disease for which a person should be detained in hospital rather than be given an unqualified acquittal.
Lord Denning's casual proposition has not been universally accepted, although some elements of the theory remain today. It was questioned in R. v. Quick, supra, at pp. 351‑52, and legal academics have questioned the utility of the test; see Stuart, supra, at pp. 94‑95; Colvin, supra, at p. 294. As well, medical authorities have doubted the ability of their profession to predict recurrent dangerousness; see Roth, "Modern Neurology and Psychiatry and the Problem of Criminal Responsibility", in Hucker, Webster and Ben-Aron, eds., Mental Disorder and Criminal Responsibility (1981), at pp. 104-109. In Rabey Martin J.A. doubted the merit of Lord Denning's test, noting, at p. 476, that the converse of Denning's proposition was surely not good law. He stated:
It would be quite unreasonable to hold that a serious mental disorder did not constitute a disease of the mind because it was unlikely to recur. To so hold would be to exclude from the exemption from responsibility afforded by insanity, persons, who by reason of a severe mental disorder were incapable of appreciating the nature and quality of the act or of knowing that it was wrong, if such mental disorder was unlikely to recur.
The majority of this Court approved these comments, and Dickson J. in dissent conceded the point, at p. 533:
A test of proneness to recur does not entail the converse conclusion, that if the mental malady is not prone to recur it cannot be a disease of the mind. A condition, organic in nature, which causes an isolated act of unconscious violence could well be regarded as a case of temporary insanity.
Nonetheless, Dickson J. sought to revive Lord Denning's basic formulation in the following passage, at pp. 551‑52:
Under the heading "Insanity versus Automatism" [Glanville] Williams states that before the decision in Quick, Lord Denning's view in Bratty was generally accepted. The test of insanity was the likelihood of recurrence of danger. In Quick, the Court of Appeal adopted what might seem at first sight to be a different test for insane versus non‑insane automatism. But the real question is whether the violence is likely to be repeated. Williams concludes that "On the whole, it would be much better if the courts kept to Lord Denning's plain rule; the rule in Quick adds nothing to it". (at p. 615)
This view, which the Ontario Court of Appeal appears to have rejected, finds ample support in the legal literature. See Beck, "Voluntary Conduct: Automatism, Insanity and Drunkenness", (1966‑67) 9 Crim. L.Q. 315, at p. 321, "The cause of the automatic conduct, and the threat of recurrence, are plainly factors that determine the line between sane and insane automatism"; Whitlock, "Criminal Responsibility and Mental Illness" at p. 120, "The test of whether or not an episode of automatism is to be judged as sane or insane action seems to rest on the likelihood of its repetition"; J. Ll. J. Edwards, "Automatism and Criminal Responsibility", 21 Mod. L. Rev. 375, at p. 385, "Where evidence is available of recurrent attacks of automatism during which the accused resorts to violence . . . inevitably leads to consideration of the imposition of some restraint"; Prevezer, "Automatism and Involuntary Conduct"  Crim. L. R. 440, at p. 441, "If . . . it can safely be predicted that his conduct is not likely to recur, having regard to the cause of the automatism, there can be no point in finding him insane and detaining him in Broadmoor"; Martin, "Insanity as a Defence", (1965‑6) 8 Crim. L. Q. 240, at p. 253, "Perhaps the distinction lies in the likelihood of recurrence and whether the person suffering from it is prone to acts of violence when in that state".
In principle, the defence of automatism should be available whenever there is evidence of unconsciousness throughout the commission of the crime, that cannot be attributed to fault or negligence of his part. Such evidence should be supported by expert medical opinion that the accused did not feign memory loss and that there is no underlying pathological condition which points to a disease requiring detention and treatment.
While Dickson J.'s views did not carry the day in Rabey, nothing in the majority judgment precludes the consideration of a continuing danger as a factor at the policy stage of the inquiry.
Since Rabey, the House of Lords has revisited the question of disease of the mind, in R. v. Sullivan,  A.C. 156. Lord Diplock, speaking for a unanimous court, commented, at p. 172, as follows:
The nomenclature adopted by the medical profession may change from time to time; Bratty was tried in 1961. But the meaning of the expression "disease of the mind" as the cause of "a defect of reason" remains unchanged for the purposes of the application of the M'Naghten Rules. I agree with what was said by Devlin J. in Reg. v. Kemp  1 Q.B. 399, 407, that "mind" in the M'Naghten Rules is used in the ordinary sense of the mental faculties of reason, memory and understanding. If the effect of a disease is to impair these faculties so severely as to have either of the consequences referred to in the latter part of the rules, it matters not whether the aetiology of the impairment is organic, as in epilepsy, or functional, or whether the impairment itself is permanent or is transient and intermittent, provided that it subsisted at the time of commission of the act. The purpose of the legislation relating to the defence of insanity, ever since its origin in 1800, has been to protect society against recurrence of the dangerous conduct. The duration of a temporary suspension of the mental faculties of reason, memory and understanding, particularly if, as in Mr. Sullivan's case, it is recurrent, cannot on any rational ground be relevant to the application by the courts of the M'Naghten Rules, though it may be relevant to the course adopted by the Secretary of State, to whom the responsibility for how the defendant is to be dealt with passes after the return of the special verdict of "not guilty by reason of insanity." [Emphasis added.]
This passage, while not entirely clear, appears to endorse the consideration of recurrence as a non‑determinative factor in the insanity inquiry. Lord Diplock states that the duration of the condition in question is not a relevant consideration: a disease of the mind can be temporary or permanent. He also suggests that the relative impermanence of a condition is particularly inconsequential if the condition is prone to recur. A necessary corollary of these statements is the more general proposition that recurrence suggests insanity, but the absence of recurrence does not preclude it. This view of the law was stated explicitly in R. v. Burgess,  2 All E.R. 769 (C.C.A.), at p. 774:
It seems to us that if there is a danger of recurrence that may be an added reason for categorising the condition as a disease of the mind. On the other hand, the absence of the danger of recurrence is not a reason for saying that it cannot be a disease of the mind. Subject to that possible qualification, we respectfully adopt Lord Denning's suggested definition.
In my view, the Court of Appeal has properly stated the law on this point. Recurrence is but one of a number of factors to be considered in the policy phase of the disease of the mind inquiry. Moreover, the absence of a danger of recurrence will not automatically exclude the possibility of a finding of insanity.
In this case, then, neither of the two leading policy approaches determines an obvious result. It is clear from the evidence that there is almost no likelihood of recurrent violent somnambulism. A finding of insanity is therefore less likely, but the absence of a continuing danger does not mean that the respondent must be granted an absolute acquittal. At the same time, the internal cause theory is not readily applicable in this case. It is therefore necessary to look further afield.
In his dissenting reasons in Rabey, at p. 546, Dickson J. enumerates certain additional policy considerations that are relevant to the distinction between insanity and automatism:
There are undoubtedly policy considerations to be considered. Automatism as a defence is easily feigned. It is said the credibility of our criminal justice system will be severely strained if a person who has committed a violent act is allowed an absolute acquittal on a plea of automatism arising from a psychological blow. The argument is made that the success of the defence depends upon the semantic ability of psychiatrists, tracing a narrow path between the twin shoals of criminal responsibility and an insanity verdict. Added to these concerns is the in terrorem argument that the floodgates will be raised if psychological blow automatism is recognized in law.
These factors are raised by Dickson J. as arguments against a finding of non‑insane automatism. In the present case, however, none of these arguments is persuasive. It seems unlikely that the recognition of somnambulism as non‑insane automatism will open the floodgates to a cascade of sleepwalking defence claims. First of all, the defence of somnambulism has been recognized, albeit in obiter discussion, in an unbroken line of cases stretching back at least a century, yet I am unaware of any current problem with specious defence claims of somnambulistic automatism. Indeed, this case and Burgess are among the few appellate decisions in which the status of somnambulism was a question to be decided. Moreover, it is very difficult to feign sleepwalking -‑ precise symptoms and medical histories beyond the control of the accused must be presented to the trier of fact, and as in this case the accused will be subjected to a battery of medical tests. Finally, a comprehensive listing of the indicia of sleepwalking can be consulted by both the court and the medical experts; see Fenwick, "Somnambulism and the Law: A Review" (1987), 5 Behavioral Sciences & the Law 343, at p. 354.
It may be that some will regard the exoneration of an accused through a defence of somnambulism as an impairment of the credibility of our justice system. Those who hold this view would also reject insane automatism as an excuse from criminal responsibility. However, these views are contrary to certain fundamental precepts of our criminal law: only those who act voluntarily with the requisite intent to commit an offence should be punished by criminal sanction. The concerns of those who reject these underlying values of our system of criminal justice must accordingly be discounted.
In the end, there are no compelling policy factors that preclude a finding that the accused's condition was one of non‑insane automatism. I noted earlier that it is for the Crown to prove that somnambulism stems from a disease of the mind; neither the evidence nor the policy considerations in this case overcome the Crown's burden in that regard. Committal under s. 614(2) of the Criminal Code is therefore precluded, and the accused should be acquitted.
As I noted at the outset, it is apparent that the medical evidence in this case is not only significant in its own right, but also has an impact at several stages of the policy inquiry. As such, I agree with the Chief Justice that in another case on different evidence sleepwalking might be found to be a disease of the mind. As Dickson J. commented in Rabey, at p. 552:
What is disease of the mind in the medical science of today may not be so tomorrow. The court will establish the meaning of disease of the mind on the basis of scientific evidence as it unfolds from day to day. The court will find as a matter of fact in each case whether a disease of the mind, so defined, is present.
On the question of a possible imposition of an order to keep the peace, I am in agreement with the reasons of both Sopinka and McLachlin JJ. I would not refer this matter back to the trial judge. In addition to their reasons, I would note the following practical considerations that in my view preclude consideration of the type of order proposed by the Chief Justice.
To be effective, any order to keep the peace would have to be permanent. This would violate established practice (if not the law) regarding peace orders, which requires a defined period for the order; see R. v. Edgar (1913), 109 L.T. 416 (C.C.A.). Of course, the courts could impose a succession of limited‑term orders that would amount to a permanent injunction governing the respondent. However, even this course of action may not be feasible in light of concerns over enforcement of the orders, to which I now turn.
Generally, there are two mechanisms for the enforcement of a traditional order to keep the peace. First, any complainant who seeks an order will return to court to complain of any breach of the peace. Thus the complainant acts as a watchdog much like the plaintiff in a civil injunction action. In the instant case, however, there is no "complainant" as such. Only the respondent's immediate family would have a vested interest in the order and an ability to monitor compliance with it, and it would be unrealistic to expect them to complain of any breach of the peace.
A second enforcement mechanism is the imposition of a bond with a guarantee from some third person. This is the standard procedure under the Magistrate's Courts Act in England, where the courts require a surety to guarantee the recognisance; see Halsbury's Laws of England (4th ed., vol. 29, para. 444). The surety is entitled to complain to the court if the principal has been or is about to be in breach of the conditions of the recognisance, and as such the surety becomes the court's watchdog. Such an arrangement is feasible over a short term, as the cost of the surety can reasonably be imposed upon the accused. But with a permanent order, the costs of a life‑long surety would be onerous, and it would be unreasonable to require the respondent to bear this cost.
It appears, then, that the judiciary is not practically equipped to administer a "keep the peace order" in the circumstances of this case. For this reason, along with the reasons of my colleagues, I would not remit this case back to the trial judge for the consideration of such an order. I would accordingly dismiss the appeal and uphold the acquittal of the respondent.
The following are the reasons delivered by
Sopinka J. -- I agree with the Chief Justice that the trial judge did not err in leaving the defence of automatism rather than that of insanity with the jury, and with both his and La Forest J.'s reasons for reaching that conclusion. However I do not agree that this matter should be referred back to the trial judge to consider an order to keep the peace. I share the concerns expressed by McLachlin J. on this point, and would also make the following observations.
This Court has recognized the existence of a common law preventative justice power in addition to the specific statutory power to make an order to keep the peace pursuant to an information laid under what is now s. 810 of the Criminal Code, R.S.C., 1985, c. C-46: Mackenzie v. Martin,  S.C.R. 361. However even at common law this power has significant limits. In Mackenzie, at p. 368, Kerwin J. quoted from Blackstone on the nature of the power:
This preventative justice consists in obliging those persons, whom there is probable ground to suspect of future misbehaviour, to stipulate with and to give full assurance to the public, that such offence as is apprehended shall not happen; by finding pledges or securities for keeping the peace, or for their good behaviour. [Emphasis added.]
Several lower court decisions have similarly recognized that this common law power cannot be exercised on the basis of mere speculation, but requires a proven factual foundation which raises a probable ground to suspect of future misbehaviour. See: R. v. White, Ex p. Chohan,  1 C.C.C. 19 (B.C.S.C.); Re Regina and Shaben,  3 O.R. 613 (H.C.J.); Stevenson v. Saskatchewan (Minister of Justice), (Q.B., June 8, 1987, unreported).
The uncontroverted expert evidence in this case is wholly inconsistent with such a conclusion. The Chief Justice characterizes that evidence as indicating that "the chances of such an occurrence taking place again are for all practical purposes nil" (at p. 000).
Moreover the extent and continued validity of this common law power has yet to be considered in light of the Canadian Charter of Rights and Freedoms. Restrictions on an individual's liberty can only be effected in accordance with principles of fundamental justice or must be justified under s. 1. This applies to deprivations of liberty following a criminal conviction as well as those effected in other circumstances.
Our criminal justice system is premised on the requirement that the Crown must prove all the elements of an offence in accordance with legal principles. Leaving aside the question of a lack of criminal responsibility on account of mental disorder, the failure to prove the guilt of the accused beyond a reasonable doubt in accordance with such principles will result in an acquittal. That is exactly what has happened in this case. The respondent has been acquitted in accordance with ordinary criminal law principles.
Turning to the common law power relied upon by the Chief Justice, I have grave doubts as to whether a power that can be exercised on the basis of "probable ground[s] to suspect future misbehaviour" without limits as to the type of "misbehaviour" or potential victims, would survive Charter scrutiny. If such a power allowed the imposition of restrictive conditions following an acquittal on the basis of a remote possibility of recurrence, it may well be contrary to s. 7.
Furthermore the potential implications of the course of action contemplated by the Chief Justice are significant not only for the respondent, but also in other cases. Consider an individual who is convicted of a violent crime at trial, but on appeal a stay is entered on the basis that his right to be tried within a reasonable time has been violated. Would the Court nonetheless impose restrictions on his liberty in an attempt to ensure that such an event does not recur? Such restrictions would be a significant departure from fundamental principles of criminal law, yet there is nothing in the authorities relied upon by the Chief Justice which limits the consideration of an order to keep the peace to cases such as the one at bar.
I note that there still exists the possibility of an information being laid pursuant to s. 810 of the Criminal Code. This, of course, is subject to the evidentiary basis required under that section, "that the informant has reasonable grounds for his fears" (s. 810(3)), and to constitutional challenge. If such a proceeding is to be initiated, it should not be done so by this Court acting proprio motu.
Finally I observe that the respondent cross-appealed on the ground that if this Court were inclined to interfere with the decision of the Court of Appeal, a stay should be entered by reason of the violation of his rights under s. 11(b) of the Charter. If the respondent remains subject to the criminal justice system and potential restraints on his liberty, it would be necessary to deal with this cross-appeal.
I would accordingly dismiss the appeal and uphold the acquittal of the respondent. On my disposition of the main appeal, it is unnecessary to address the question raised by the cross-appeal.
The reasons of McLachlin and Iacobucci JJ. were delivered by
McLachlin J. -- I have read the reasons of the Chief Justice and concur in them, except on the question of whether the matter should be referred back to the trial judge for consideration of a further order imposing conditions on the future conduct of Mr. Parks. I also agree with the reasons of La Forest J. and Sopinka J.
I share the Chief Justice's concern that notwithstanding the justice of an acquittal in this case and the evidence that a recurrence is highly unlikely, great care should be taken to avoid the possibility of a similar episode in the future. However, I also have concerns about the appropriateness of referring the matter back at this stage for a supervisory order in the circumstances of this case.
In addition to the difficult issues raised by an order restricting a person's liberty on account of an act for which he has been acquitted, I have concerns whether further proceedings are appropriate in the circumstances before us. Mr. Parks has been living in the shadow of these charges since May 24, 1987, over five years. His acquittal is now confirmed. We are told he has been making courageous efforts to re-establish his life. Should he now be embroiled in a further set of proceedings concerned, not with his guilt or innocence, but with the maintenance of his liberty?
Generally, the courts do not grant remedies affecting the liberty of the subject unless they are asked to do so by the Crown, which is charged with instituting such legal processes as it deems appropriate having regard to the public interest and fairness to the individual involved. In the absence of an application by the Crown, I hesitate to remit the case for consideration of further measures against the accused.
I add that the possibility of supervisory orders in this situation may be a matter which Parliament would wish to consider in the near future.
Appeal dismissed, Lamer C.J. and Cory J. dissenting in part.
Solicitor for the appellant: The Ministry of the Attorney General, Toronto.
Solicitors for the respondent: Ruby & Edwardh, Toronto.