The Supreme Court of Canada restored the sexual assault convictions of Christopher James Kruk and Edwin Tsang in R. v. Kruk, 2024 SCC 7, rejecting a proposed rule that would have broadened appellate intervention when trial judges use assumptions about human behaviour to assess witnesses.

All seven judges agreed with the result on March 8, 2024, overturning orders for new trials in two unrelated British Columbia cases. Six judges endorsed Justice Sheilah Martin’s reasons. Justice Malcolm Rowe reached the same outcome through a different approach to reviewing unreasonable generalizations.

The decision preserves the distinction between ordinary factual errors and errors of law. It does not give judges permission to substitute stereotypes or unsupported conclusions for evidence, nor does it weaken the Crown’s obligation to prove guilt beyond a reasonable doubt.

Why the proposed rule was rejected

The British Columbia Court of Appeal had overturned both convictions after identifying what it considered assumptions about human behaviour that were not grounded in the evidence. It treated reliance on those assumptions as legal error, allowing review on a standard of correctness.

Martin rejected that general rule. Judges necessarily draw on common sense and experience when deciding whether testimony is believable and accurate. Requiring evidence to establish every general proposition underlying that reasoning would make ordinary witness assessment unworkable.

The majority found no coherent boundary in the proposed rule between using experience to interpret evidence and introducing considerations from outside the record. It also warned that the rule encouraged appeal courts to isolate phrases from reasons, relabel factual disagreements as legal errors and substitute their own assessment for that of the judge who heard the witnesses.

Trial judges have advantages that an appeal court reading transcripts does not. Their reasons must be read as a whole, in the context of the evidence and issues, rather than examined word by word for formulations that might suggest an assumption.

Ordinarily, factual assessments attract deference unless an error is both “palpable and overriding”: obvious and consequential to the outcome. A mistake affecting an isolated finding does not necessarily undermine the verdict.

Protections against stereotypes remain

A separate branch of the majority’s reasoning addressed the analogy between ordinary generalizations and prohibited myths about sexual assault complainants.

Martin explained that rules against those myths arose to remove discriminatory barriers that historically treated complainants as inherently unreliable. Examples include discrediting a complainant merely for delayed reporting or using sexual history to infer consent or diminished credibility.

Those rules identify legal errors. Their specific history and remedial purpose did not justify treating every questionable assumption about every witness as an equivalent error of law.

The majority also warned that requiring particular assumptions about sexual behaviour to be supported by evidence could invite intrusive questioning about a complainant’s sexual history, undermining protections designed to keep irrelevant and prejudicial material out of trials.

At the same time, Martin expressly left room for parties, including accused people, to argue that stereotypes rooted in other analogous forms of unequal treatment amounted to legal error. The categories were not closed.

The accused’s existing protections remained essential. Silence cannot be used as evidence of guilt; an accused cannot automatically be disbelieved because they face punishment; and rejecting the accused’s testimony does not itself establish guilt. Reasonable doubt may arise from other evidence or an absence of evidence.

Nor does the prohibition on myths prevent a judge from considering properly admitted, case-specific evidence of a motive to fabricate. The majority cautioned against extending the law on stereotypes beyond its proper scope.

How appeal courts must assess the complaint

The majority directed reviewing courts first to ask whether the challenged reasoning actually involved an assumption. A passage that appears general when isolated may instead express a finding about the particular witness, grounded in the record.

If an improper assumption was used, the court must identify the applicable standard of review. Recognized legal errors still attract correctness review, including prohibited sexual assault myths, violations of the presumption of innocence, improper judicial notice, findings without evidence and reasonable apprehension of bias.

Otherwise, the court must determine both whether the assumption was plainly wrong and whether reliance on it affected the result. The majority emphasized that calling something common sense does not protect reasoning based on falsehoods or discriminatory beliefs.

Why both convictions were restored

In Kruk’s case, the central issue was whether the sexual activity occurred. The defence maintained that the complainant was mistaken about experiencing penetration. The trial judge accepted that her intoxication raised reliability concerns but believed her evidence on that central point, while rejecting important parts of Kruk’s account and considering supporting circumstantial evidence.

The Court of Appeal focused on the judge’s statement that a woman was extremely unlikely to mistake that sensation. Martin concluded that, read in context, the passage responded to the particular defence theory and reflected acceptance of this complainant’s testimony, rather than a categorical assumption that no woman could ever be mistaken.

The majority added an alternative conclusion: even if the statement were treated as a general assumption, it disclosed no palpable and overriding error. Expert neurological or physiological evidence was not required to support the judge’s assessment of the testimony in this case.

In Tsang’s case, the appeal court had identified three supposed assumptions. Two concerned the trial judge’s rejection of his account that the complainant requested rough sexual activity and that intercourse did not occur because a condom could not be found.

The Supreme Court held that these were findings about the evidence and the particular people, not generalized rules about sexual preferences or controlling personalities. The trial judge had accepted the complainant’s account and found Tsang’s version contrived and inconsistent with other accepted evidence. The appeal court had improperly separated parts of the events and reweighed their significance.

The third point was different. The trial judge treated Tsang’s abrupt departure after the encounter as supporting the conclusion that it had been non-consensual. The majority found that assumption plainly wrong: departure speed generally says nothing about consent.

That error nevertheless did not determine the outcome. Martin concluded that the judge’s other credibility findings and acceptance of the complainant’s testimony would still have led to conviction.

Rowe proposed broader review for legal error

Rowe also rejected a rule that condemned a generalization merely because it lacked independent evidentiary support. But he would have asked three questions: whether the judge relied on a generalized expectation, whether it was reasonable, and whether the judge treated it as a conclusive fact rather than a guide to assessing evidence.

For Rowe, reliance on an unreasonable generalization would itself be an error of law, whether or not it belonged to an established category of discriminatory myth. Trial judges had no special advantage over appellate judges in deciding what general expectations were reasonable. Broader legal review also mattered because Crown appeals from acquittals are limited to questions of law.

Even a reasonable expectation could not replace the evidence or foreclose consideration of contrary evidence. Rowe also cautioned that the Crown could not use this approach simply to reargue an acquittal on the facts.

He found no error in Kruk’s case: the judge used a reasonable expectation about physical perception while considering the evidence, rather than treating it as indisputable. In Tsang’s case, Rowe classified the departure assumption as a legal error but applied the Criminal Code’s curative proviso because it could not have affected the verdict.

That distinction mattered procedurally. Once an accused establishes legal error, Rowe explained, the Crown bears the burden of showing that the conviction can stand under the proviso. The accused does not also have to prove materiality as part of identifying the legal error.

Source: R. v. Kruk, 2024 SCC 7.