Background
The accused was convicted at trial of sexual assault and sexual interference against his then-thirteen-year-old biological daughter. The conviction for sexual assault was stayed under the rule against multiple convictions for the same delict (Kienapple v R, 1974 CanLII 14 (SCC)), and the accused appealed the remaining conviction for sexual interference. The victim testified that the accused sexually assaulted her on three occasions in the family home; the accused denied all allegations, testified that he loved his daughters, and described the victim as difficult to manage — a characterization that was common ground between both parties.
At trial, the judge applied the R v W(D), [1991] 1 SCR 742 framework for assessing the credibility of the accused. The trial judge found no blatant self-undermining in the accused’s evidence but proceeded to draw adverse credibility inferences from two specific answers the accused gave on the stand. First, when asked about learning of the allegations from a Child and Family Services worker, the accused described feeling “surprised,” “afraid,” and “worried,” and stated he moved out voluntarily. Second, when asked how he felt about the victim at the time of trial, the accused expressed sadness and love but did not articulate concern for her mental wellbeing.
The trial judge found these answers telling: in his view, an innocent person falsely accused of sexually assaulting his daughter would have expressed shock and mortification, and a loving father who believed his daughter had fabricated serious allegations would have expressed concern for her mental health. On the basis of these inferences, the trial judge rejected the accused’s denials, found the victim’s testimony credible and reliable, and entered a conviction.
The Court’s Holding
The Court of Appeal, per Simonsen JA writing for a unanimous three-judge panel, allowed the appeal, quashed the convictions on both counts, and ordered a new trial. The court held that the trial judge committed palpable and overriding errors by grounding his adverse credibility findings in impermissible “common-sense” assumptions about how an innocent accused person must behave. Applying the standard set out in R v Kruk, 2024 SCC 7, the court found those assumptions plainly outside the bounds of permissible common-sense reasoning.
With respect to the first answer, the court found it would be a palpable error to assume that a person confronted for the first time with what they consider false allegations would necessarily deny them in a particular way — such as expressing shock or mortification. With respect to the second answer, the court found it palpably erroneous to assume that a parent who loves their child must continue to express concern for that child’s mental wellbeing even while facing serious allegations the parent says the child fabricated. The court emphasized that human responses to such circumstances are highly variable and that this assumption is not a reliable benchmark for credibility assessment.
The court further held, citing R v Berg, 2026 SCC 21, that a trier of fact cannot reject an accused’s account based solely on a “considered and reasoned acceptance” of the complainant’s conflicting account, as that risks reducing the trial to a credibility contest. Because the trial judge’s flawed credibility analysis of the accused’s evidence clearly affected the outcome, the Crown conceded — and the court agreed — that the errors were overriding.
Key Takeaways
- Under Kruk, adverse credibility findings rooted in behavioural assumptions must reflect genuine common sense, not stereotyped expectations; courts cannot assume an innocent accused will react to allegations in a particular way.
- The assumption that a loving parent would necessarily express ongoing concern for a child’s mental wellbeing while simultaneously maintaining the child fabricated serious allegations is not a permissible credibility benchmark — human responses in such circumstances are too variable.
- Per Berg, rejection of an accused’s testimony cannot rest solely on a reasoned acceptance of the complainant’s account; the trier of fact must explain how the evidence as a whole establishes guilt beyond a reasonable doubt through a proper W(D) analysis.
- Where palpable errors in a credibility assessment plainly drove the verdict, they will be treated as overriding and will warrant a new trial.
Why It Matters
This decision reinforces the appellate guardrails placed on trial judges’ credibility reasoning following the Supreme Court of Canada’s landmark ruling in Kruk. By identifying specific categories of behavioural assumptions — how an accused reacts upon first hearing of allegations, and whether a loving parent must vocalize concern for an accusing child — the Manitoba Court of Appeal provides concrete guidance on where common-sense reasoning crosses into impermissible stereotype. Defence counsel and Crown prosecutors alike will need to scrutinize whether adverse credibility inferences at trial rest on empirically grounded generalizations or on untested assumptions about how innocent people behave.
The court’s reliance on the newly decided R v Berg, 2026 SCC 21 also signals the continuing evolution of the W(D) framework: acquittals cannot be avoided simply by a trial judge articulating a preference for the complainant’s testimony without independently explaining how that testimony, together with any other evidence, satisfies the criminal standard of proof. Practitioners in sexual offence cases should pay close attention to how trial reasons address — or fail to address — the path from credibility rejection to proof beyond a reasonable doubt.