Appeal Watch: SCC to Speak on Stereotypes and Common Sense in R v Tsang

Trial judges have a unique ability to assess evidence in a criminal trial: they see the witness, hear the testimony, and decide whether someone is believable. These findings are sometimes informed by intangibles, things that any person would struggle to articulate. Where a judge fails to articulate the evidence supporting a negative credibility finding, they leave their reasons open to challenge for impermissible stereotypical or common sense reasoning. On February 2, 2023, the Supreme Court of Canada (SCC) granted leave to appeal in such a case, in R v Tsang, 2022 BCCA 345, (SCC case no. 40447) [Tsang]. The appeal from a unanimous decision of the British Columbia Court of Appeal (BCCA) was expedited, and will be heard together with a similar appeal, also from BC, His Majesty the King v Christopher James Kruk, case no 40095 [Kruk], on May 18.

Facts and Procedural History

Edwin Tsang, the accused, met the complainant, LBV, at an event. They had mutual friends but had never met before. Together with their friend, LF, they went out for drinks and dancing. LF went home early, leaving the accused with LBV at the club. They danced together, and the accused bought them both drinks. Later in the evening, the complainant directed the accused to a “secluded” parking lot to make out (Tsang, para 2). They engaged in some consensual sexual foreplay. After this, the complainant testified that the appellant forced her to perform non-consensual oral sex, participate in non-consensual digital penetration that left her with a painful injury, and that they had non-consensual vaginal intercourse (Tsang, paras 84–88, 93). Apart from his testimony that the sexual encounter was consensual, the accused also testified that they did not have intercourse because he could not find a condom (Tsang, paras 63–64). 

The accused was convicted of sexual assault contrary to section 271 of the Criminal Code. The trial judge rejected the accused’s evidence and found that he exerted a serious degree of control over the evening and “orchestrated” the events (Tsang, para 3).

On appeal, the accused alleged that the trial judge had made three errors (Tsang, para 4):

a) basing an adverse assessment of his credibility on speculative reasoning that was not founded upon evidence, but was premised in part on stereotypes or generalizations and using a higher degree of scrutiny in assessing the credibility or reliability of defence evidence than that used to assess Crown evidence;

b) admitting and relying upon opinions of unknown reliability from an expert witness, based entirely upon anecdotal evidence; and

c) concluding that Mr. Tsang’s account “left unexplained how [the complainant] was injured”, a conclusion which the appellant says either reversed the burden of proof, or was irrational and inconsistent with the judge’s own findings of fact.

A unanimous BCCA allowed the appeal based on the first error: speculative reasoning based on stereotypes. Of the four findings that the accused challenged, however, the BCCA held that only one was unsupported by the evidence. The trial judge had found that it was “not believable” that the complainant asked the accused to spank her in the car, after the consensual foreplay (Tsang, para 28). This finding formed the basis for the BCCA’s intervention and overturning of the trial decision.

The BCCA’s review of the evidence revealed that the trial judge’s assertion that it was “not believable” that the complainant had asked to be spanked was based on “implicit, unsupported assumptions about ‘normal behaviour’” (Tsang, para 53). In sum, the trial judge determined that, because there was evidence that the complainant had not been interested in the accused earlier in the night, it was simply not believable that she had consented to “rough sex” after the consensual foreplay later that same evening (Tsang, para 52). The BCCA took issue with the logical throughline of the trial judge’s decision. The conclusion that the complainant would not have asked to be spanked ignored the facts that, since her initial disinterest, the complainant had pointed the accused towards the secluded parking lot, gotten in the car to make out, and consented to sexual foreplay. This showed, for the BCCA, a “marked departure” in the complainant’s behaviour that did not support the trial judge’s conclusion (Tsang, para 52).

As the BCCA panel was unanimous in its decision, the Crown required leave to appeal to the SCC. The SCC granted this leave, signalling that the Court has found some point of law on which they feel it is necessary for them to comment. The hearing together of this case with Kruk suggests that the Court has some interest in developing the use of speculative reasoning, particularly common sense or stereotypical reasoning, in the trial judge’s assessment of credibility. Whether the Court intends to develop these concepts solely in the context of sexual assault remains to be seen.

Can Trial Judges Assess Credibility without Common Sense Reasoning?

Speculative and stereotypical reasoning are distinct. Tsang and Kruk are both examples of speculative reasoning—that is, reasoning based not on the evidence, but on some unsupported assumption made by the trial judge. But the reality is that much speculative reasoning is premised on stereotypes or generalizations, informed by the judge’s perspective and bias, or otherwise on “common sense” reasoning.

The heart of the appeals may well be stereotypical reasoning, but I am hopeful that the SCC will instead spill ink on the use of common sense reasoning. Stereotypes are not “common sense” in the truest sense, but instead a distorted form of common sense. Stereotypes may be “common”, but are premised on extremely unstable foundations. For this reason, courts group stereotypes together with myths. I doubt the SCC will use Tsang to undo its jurisprudence on the dangers of stereotypical reasoning. Reliance on stereotypes is impermissible: They may be common, but they are seldom true.

Judges are permitted, however, to consider common sense in reasoning. They can—indeed, they must—consider their own perspective and life experience when considering evidence. To this point, the SCC has made clear that experience and perspective, together creating “common sense”, cannot be considered where it does not arise on the evidence of a particular case:

The life experience of this trial judge, as with all trial judges, is an important ingredient in the ability to understand human behaviour, to weigh the evidence, and to determine credibility. It helps in making a myriad of decisions arising during the course of most trials. It is of no value, however, in reaching conclusions for which there is no evidence (R v S(RD), [1997] 3 SCR 484 (SCC), para 13).

Trial judges may be able to use common sense reasoning by taking judicial notice of the underlying assumptions. The joint hearing provides the SCC with an opportunity to consider formally the “rule against ungrounded common-sense assumptions” as described by Justice David Paciocco in R v JC, 2021 ONCA 131 (para 58). The rule was articulated in that case with support from jurisprudence on the principle that judges cannot rely on common sense where it is not grounded in the evidence or supported by judicial notice. In Tsang and Kruk, the SCC may speak on this “rule” in a more fulsome capacity—to answer, for example, whether judicial notice alone is (i) required and (ii) sufficient to empower trial judges to use common sense reasoning. Taking judicial notice of, what is at its core, a common sense assumption may prove a thorny route for any trial judge to take. Judicial notice requires a fact to be so notorious as to not be in dispute. Common sense assumptions may not satisfy this requirement, or, where they do, may satisfy it for improper stereotypical reasons.

Judges’ experience is specifically important to determine credibility, the judicial function that the BCCA decision in Kruk threatens, according to the Crown’s factum at the SCC (paras 74–79). Tsang is expected to address the same concern. In both Tsang and Kruk, the BCCA undid credibility findings by strictly parsing the trial judges’ words and pointing to what might be an assumption based on the judge’s “common sense”. 

For example, if a judge notes that police officers, in their experience, are honest witnesses (as judges sometimes say), that judge will not be allowed to use that experience to find a particular officer credible and honest. But could a judge find that young men are often dishonest to find that a particular young male accused is dishonest? It is easy, when the examples are at the extremes (as the two preceding examples are), to see why this might be a problematic tool in assessing credibility. But to elevate a problematic tool and transform it into a reversible error may do more harm than good. 

Should the SCC confirm that reviewing courts can overturn credibility findings on a language-based analysis, trial judges’ ability to make credibility findings may be severely affected. Instead of owing deference, some reviewing courts may review decisions line-by-line as a results-driven exercise, with a view to finding an error. Credibility findings would be less secure, and trial judges would be unduly restricted in their ability to apply their experience to their judicial reasoning. Such a shift could be especially damaging in cases where credibility will be central, such as sexual assaults. As more convictions are overturned, more complainants will be required to participate in new trials, too. It would no longer be enough for a complainant to be honest, forthright, and believable. Complainants would be less protected by the judicial process—particularly where new trials may retraumatize sexual assault survivors or, if a complainant is unwilling to proceed again, threaten the ability of the Crown to convict.

Appellate Review should not Undo a Trial Judge’s Decision Who to Believe

Protecting sexual assault survivors is an important goal of the criminal law in prosecuting sexual assaults. Courts go to great pains to reduce the retraumatizing effect of trials, including various testimonial aids where complainants are particularly vulnerable. To the extent that the BCCA’s decision threatens protections for sexual assault survivors, the SCC is poised to provide a definitive corrective position. Trial judges have long made decisions based on their experience and perspective. Resolving credibility conflicts is a difficult, intangible process that does not and never will exist in a vacuum. 

If judges are disempowered to use their own reasoning and common sense, the burden of proving cases like sexual assault, when so much turns on credibility, will simply be too high. It would mean, as the BC Crown put, “a return to bygone days where a sexual assault conviction could not stand absent corroboration despite a trial judge believing the complainant’s evidence” (Kruk, factum of the Appellant, para 79). Courts have moved away from this corroboration requirement, and decided that complainants and accused can speak for themselves, and courts should hear them. If a reviewing court can pick apart a judges’ decision, it calls into question the value of having a trier of fact hear testimony in the first place. 

Sexual assault trials do not assume that every piece of evidence is true, and decide from there. If they did, no person could ever be convicted where the complainant and accused give different accounts of the same event, unless one side of the story can be “confirmed”. Complainants—survivors—should not be procedurally disbelieved. And sexual offenders should not escape conviction because an abstract process, divorced from human intuition, prevents judges from holding them accountable.