FACTS: Rabey was infatuated with the victim, who rejected him. Rabey had in his possession a rock, and after the rejection he struck her and began to choke her. He claimed to have been in a dissociative state at the time, which was supported by the evidence.
PROCEDURAL HISTORY: Acquitted at trial, overturned on appeal.
ISSUES: Did Rabey qualify for the non-mental disorder automatism defence?
DECISION: Appeal dismissed, new trial ordered.
REASONS: Automatism exists where the accused is unconscious of their acts such that they are involuntary. It can be caused by a disease of the mind, in which case it is assimilated to the insanity provisions of the Code. What a “disease of the mind” is a question of law, and it falls to the trier of fact to determine whether such a disease (e.g. the dissociative state) exists in the circumstances. Diseases of the mind are internal causes (arising from something in the accused’s psychological or emotional makeup, or their biology). Automatism can be caused externally as well, if an external event creates some transient disturbance of consciousness. These cannot be based in a disease of the mind. Here, the precipitating event was one of the ordinary stresses of life. It could not have amounted to an external event capable of disturbing consciousness, as could happen with some exceptional circumstances (e.g. seeing a loved one harmed, being attacked, etc.). Issue – how would a normal person have responded in the circumstances? So Rabey’s response was internally sourced, and thus could be considered a disease of the mind.
RATIO: Mental disorder automatism is internally sourced, based on some psychological/emotional/biological (internal) characteristic of the accused; non-mental disorder automatism is externally sourced, based on when an external event temporarily disturbs the accused’s consciousness such that they act involuntarily without being conscious of their actions. The “ordinary stresses and disappointments of life” are generally not sufficient to trigger non-mental disorder automatism, unless the circumstances are exceptional.
R v Parks 1992 SCC
Facts: Parks, while asleep, drove to the home of his parents-in-law and stabbed his mother-in-law, with whom he shared a good relationship, to death. He pleaded automatism on the ground that he was sleepwalking at the time – he was prone to sleepwalking, and the circumstances in his life made it more likely to be triggered (job stress, being sued for theft, etc.).
La Forest:
We have to consider public policy as well as the evidence to distinguish between automatism and insanity. Automatism isn’t strictly speaking a defence since it undermines the voluntariness requirement of the actus reus of criminal offences. But it leads to an acquittal, unlike insane automatism.
As before, it’s a legal question whether the accused is suffering a disease of the mind. To be put to the jury, the defence must have an air of reality (evidential burden on accused). The trier of fact must then determine whether it’s actually non-mental disorder automatism. (At this point, this requires proof beyond a reasonable doubt by the Crown).
There are 2 main approaches to the policy component (in addition to the medical component) of determining whether the accused suffered a “disease of the mind”, both of which stem from a concern for public safety:
Continuing danger theory – any condition likely to present a recurring danger should be treated as insanity (though non-recurrence does not preclude insanity).
Internal cause theory – a condition stemming from the psychological, emotional, or physical makeup of the accused, rather than an external factor, should lead to a finding of insanity.
Internal cause theory is the dominant one in Canada, but it’s not a hard distinction – it’s an analytical tool. Treated otherwise it can lead to weird conclusions – some sources of automatistic behaviour can be considered internal or external, like sleepwalking in this case.
Similarly, continuing danger theory is imperfect – it’s hard to predict future dangerous episodes, or if the behaviour will reoccur at all. It’s another analytical tool.
Other policy considerations:
Automatism is easy to feign – is the accused trying to avoid liability that way?
Would allowing the defence undermine the credibility of the justice system?
Can the behaviour be restrained in the future?
Would allowing the defence open the floodgates to frivolous claims? (But what about the requirement that the morally innocent not be punished?)
Sleepwalking is not, under any of these approaches, a reasonable candidate for mental disorder automatism – no floodgates, it’s unlikely to recur in a violent way, etc.
McLachlin – courts do not grant remedies that impair the liberty of subjects unless asked to by the Crown, who considers public interest, etc.
Sopinka – remedies cannot be based on speculation about future offences, but require probable grounds for suspecting future misbehaviour.
Lamer – on the evidence, the accused was sleepwalking when the offence was committed. Sleepwalking is a sleep disorder, not a disease of the mind, and there’s no treatment (beyond sleep hygiene) for sleepwalking. On these facts, sleepwalking is not a disease of the mind – it’s not the cause of the automatistic behaviour. The Crown did not provide any contrary evidence to suggest it should be treated as such.
Lamer suggested that Parks be acquitted subject to conditions to prevent a recurrence. But everyone else shot that down.
R v Stone 1999 SCC
Facts: Fontaine stabbed his wife 47 times after he claims to have “snapped” and suffered a dissociative episode when she insulted him. He dumped the body and fled to Mexico after the stabbing; he recalled what happened while he was in Mexico. Both the defence and the Crown tendered psychiatric evidence which supported their claims.
Automatism refers to a state of impaired consciousness where the accused has no voluntary control over their action (key concept is involuntariness), rather than “unconscious” behaviour (which implies non-consciousness to the point of being inactive). If it stems from a disease of the mind, it’s mental disorder automatism and is governed by s. 16; non-mental disorder automatism arises from some source other than a mental disorder.
All automatism cases should now be treated in the same way:
There must be a properly established foundation for the defence (evidentiary burden) – this is a question of mixed fact and law for the judge who determines whether it can found a claim of a disease of the mind (not after Fontaine). It must pass the air of reality test – there must be evidence upon which a properly instructed jury could find that the accused involuntarily at the time the offence was committed.
Expert evidence will generally be required to substantiate a claim of automatism – merely asserting involuntariness isn’t that persuasive.
Factors for consideration (here, by the judge in evaluating the air of reality - this was backed down on in Fontaine):
The nature of the event that triggered the automatism
Corroborating evidence from bystanders
Medical history of the accused (including past evidence of automatistic behaviour)
The presence or absence of motive (no motive makes automatism more plausible – is the crime explicable without reference to automatism?)
None of those factors are determinative, and the list is open to modification with advances in scientific knowledge.
It must be proven on a balance of probabilities that the accused acted involuntarily. The legal burden is on the defence. This is justified by the policy concerns around the defence – it’s too easy to fake.
This passes the Oakes test, largely similar reasoning to Chaulk. People are presumed to act voluntarily, and requiring the Crown to discharge a persuasive burden beyond a reasonable doubt defeats the purpose of the presumption.
The policy considerations discussed in Parks are relevant to distinguish between mental disorder and non-mental disorder automatism. The sort of automatism is a question for the trier of fact.
Internal cause factor (not theory) (including, for “psychological blow automatism”, how a normal person would react in the circumstances – would they have entered into an automatistic state? This requires that the triggering event be severe and shocking. It’s a contextual objective test.)
Continuing danger factor (medical history, risk of recurrence of the triggering events and of the automatistic behaviour)
The administration of justice + the credibility of the defence
Where automatism is claimed, the initial presumption is that it’s mental disorder automatism arising from a disease of the mind. Non-mental disorder automatism is rare. If it is concluded that mental disorder automatism cannot arise on the facts, then only non-mental disorder automatism remains. In this case, the jury must be informed of the policy considerations against finding non-mental disorder automatism. Where a disease of the mind is established, then only mental disorder automatism can be put forward.
In this case, only mental disorder automatism could possibly be made out (caused by the wife’s words – an internal cause that could be a disease of the mind in legal terms).
Dissent from Binnie (4 judges) – the jury is the trier of fact, and they’re competent to evaluate claims from disagreeing experts. The shift away from a persuasive burden exclusively on the Crown (beyond a reasonable doubt) to a balance of probabilities violates the Charter – s. 7, 11d, and 11f’s guarantee to a trial by jury (by short-circuiting the trial process – it’s for them, and not the judge, to evaluate the defence). If there’s no evidence of a mental disorder, then there’s no grounding to consider mental disorder automatism.
