Kenneth Parks Case: Automatism Defense and the SCC Ruling

The Kenneth Parks sleepwalking case is the 1987 Toronto killing in which a 23-year-old man drove 23 kilometers to his in-laws’ home, beat and stabbed them while (a jury and later the Supreme Court of Canada accepted) sound asleep, and was acquitted of murder because sleepwalking counts as an involuntary act rather than a crime. Parks killed his mother-in-law, Barbara Ann Woods, and seriously wounded his father-in-law, Denis Woods. He walked free without conditions, and his case, decided as R. v. Parks in 1992, remains the most cited legal authority anywhere on whether a sleeping person can be held criminally responsible for violence.

What Happened on May 24, 1987

In the early morning hours, Parks got out of bed at his own home, drove roughly 23 kilometers through Toronto to the house where Barbara Ann and Denis Woods were asleep, and attacked them. He strangled Denis until the older man lost consciousness. He struck Barbara Ann with a tire iron and stabbed both of them with a kitchen knife. Barbara Ann died from her injuries. Denis survived.1Supreme Court of Canada. R. v. Parks

Parks then drove to a police station. He walked in with his hands bleeding heavily and told officers, “I think I killed some people … my hands.” During the attack, he had severed tendons in both hands by gripping the knife blade, injuries he appeared not to have noticed. Surgeons later repaired the damage.

Why There Was No Apparent Motive

Parks was under real pressure in the weeks leading up to the killing. A serious gambling problem had left him deeply in debt, and he had embezzled money from his employer to cover the losses. He was scheduled to confess the theft to his in-laws that same weekend.

But nothing in his relationship with the Woods family pointed to violence. Barbara Ann and Denis had been supportive of him through his difficulties. Investigators found no life insurance angle, no history of arguments, no financial gain to be had from their deaths. The absence of any rational reason to kill them became one of the pillars of the defense.

The Automatism Defense

Canadian criminal law, like American law, requires a voluntary act before someone can be convicted of a crime. Automatism is the legal name for the argument that the accused’s body moved without any conscious direction, in the way a reflex fires without decision. If the jury accepted that Parks was genuinely unconscious throughout the attack, then no voluntary act had occurred, and no crime had been committed.1Supreme Court of Canada. R. v. Parks

The trial judge put only that single question to the jury: was Parks conscious when he attacked the Woods? The jury said no, and acquitted him of first-degree murder and of the lesser included charge of second-degree murder. He was later also acquitted of attempted murder for the attack on Denis Woods. The acquittals were unqualified, meaning no psychiatric commitment, no supervision, no conditions of any kind.1Supreme Court of Canada. R. v. Parks

The Evidence That Convinced the Jury

Every expert witness the defense called agreed that Parks had been sleepwalking. Their reasoning combined several strands, and the prosecution did not put forward competing experts to rebut them.

Parks had a documented personal and family history of parasomnias that predated the attack. The stress and sleep deprivation he had been experiencing fit the recognized profile of sleepwalking triggers. And the injuries to his own hands were of a kind a conscious person could not have ignored, yet Parks kept attacking and then drove himself to police without appearing to feel them.1Supreme Court of Canada. R. v. Parks

The Sleep Laboratory Studies

In January 1988, researchers ran Parks through formal sleep studies. They measured how quickly he entered slow-wave sleep, the deep stage from which sleepwalking episodes emerge. In two diagnostic studies, he reached slow-wave sleep in 9.7 and 10 minutes, unusually fast entry that was consistent with a predisposition to sleepwalking. A follow-up in July 1989 recorded a latency of 24 minutes, still supporting the diagnosis. All of the testing was done while Parks was on benzodiazepine medication, which suppresses slow-wave sleep, so his unmedicated pattern was likely more abnormal still.2PMC. Disorders of Arousal and Timing of the First Period of Slow Wave Sleep

The Prosecution’s Objection

The Crown pressed the intuitive objection any juror would have. Parks had gotten out of bed, driven a car through city streets for 23 kilometers, located a specific address, and carried out a sustained attack with more than one weapon. That degree of coordination, prosecutors argued, implied some degree of awareness. The defense answered with sleep medicine literature documenting that complex motor behavior, including driving, does occur during sleepwalking episodes, however unlikely it sounds. The jury sided with the defense.

The Supreme Court of Canada Ruling

The Crown appealed, but not to relitigate the facts. The legal question was whether sleepwalking should be treated as non-insane automatism, which produces an outright acquittal, or as a “disease of the mind,” which produces a verdict of not guilty by reason of insanity and can lead to indefinite psychiatric commitment.1Supreme Court of Canada. R. v. Parks

In 1992 the Court dismissed the appeal. The majority held that sleepwalking is a sleep disorder rather than a psychiatric illness, and therefore does not qualify as a disease of the mind. The Court reasoned that sleepwalking episodes are typically triggered by external factors such as stress and sleep loss, not by an ongoing internal pathology, and that the risk of recurrence was low. The trial judge had been right to give the jury only the non-insane automatism option.1Supreme Court of Canada. R. v. Parks

Chief Justice Lamer and Justice Cory dissented in part. Their concern was structural. Treating sleepwalking as an automatic route to unconditional acquittal, with no mechanism for supervision if a recurrence was possible, left a gap in public protection. That tension between complete freedom for the acquitted person and public safety has followed sleepwalking cases ever since.

Why the Defense Almost Never Works: The Falater Contrast

Parks is famous partly because it is the exception. Scott Falater’s 1997 case in Arizona shows how narrow the path really is. Falater stabbed his wife 44 times and drowned her in the family swimming pool. He too claimed sleepwalking, and he too had no apparent motive.

The evidence around his conduct was what sank him. A neighbor watched him drag his wife’s body across the yard, then pause, walk to the garage, put on a pair of work gloves, come back, and push her into the pool, holding her head under. He then motioned for the family dog to lie down. Police later found the knife, bloody clothing, and gloves stuffed into a container hidden in his car. That sequence read as goal-directed planning and concealment, not confused waking. The jury convicted him of first-degree murder in 1999.

The comparison shows what really decided Parks. He hid nothing. He could not feel his own severed tendons. He walked into a police station. Courts weighing sleepwalking claims focus closely on whether the defendant’s conduct after the act looks like someone waking up in horror or someone covering their tracks. Parks looked like the first. Most defendants who raise the defense look more like the second.

Strip any one factor out of the Parks record and the outcome could easily have flipped. The genuinely close relationship with the victims, the documented sleep history and confirming lab studies, the ignored self-inflicted wounds, the immediate trip to police, and the united front of defense experts with no contradicting prosecution expert all had to line up together.

What Parks Means for Later Cases

In 1999, the Supreme Court of Canada revisited automatism in R. v. Stone, a case in which a man stabbed his wife 47 times and claimed a dissociative state. Stone tightened the requirements for even raising automatism at trial. The Court placed a heavier evidentiary burden on defendants and established a default presumption that automatism arises from a disease of the mind unless the defendant shows otherwise. Judges became gatekeepers who had to scrutinize the evidence before letting the defense reach the jury. The Parks precedent survives for sleepwalking specifically, but getting there is procedurally harder than it was in 1987.

The case also exposed a mismatch that has not gone away. Criminal law treats consciousness as binary: a person is either aware and responsible or unaware and blameless. Sleep medicine describes a spectrum of partial arousal states that do not sort cleanly into either box. Parks fell at one end of that spectrum, where the medical evidence was as strong as it gets. Most cases raised since have landed somewhere in the middle, where juries are left to decide how much awareness is enough to count as a voluntary act, and the answer usually turns on the kind of after-the-fact conduct that separated Parks from Falater.