Background
M.L. was tried before a jury on two counts of sexual interference. The first count (Count 2) concerned his adult daughter, S.L., who testified that he had sexually abused her when she was between five and eight years old. The second count (Count 4) concerned his three-year-old granddaughter, A.J., and rested almost entirely on a hearsay statement A.J. had made to her mother in September 2018, shortly after A.J. had spent several weeks in the appellant’s care. A.J.’s mother recounted that the child had spontaneously said her “pretty” hurt, then—in response to a series of follow-up questions—made verbal statements and gestures that her mother interpreted as disclosing sexual abuse by the appellant. Police were unable to obtain a usable video-recorded statement from A.J., and by the time of trial in 2023 the now eight-year-old A.J. had no memory of the 2018 events.
A pre-trial voir dire was held on the admissibility of A.J.’s statement. The trial judge ruled it admissible for its truth under the principled exception to the hearsay rule, finding necessity (A.J.’s lack of memory) and threshold reliability were both met. The jury was further instructed that it could use the evidence on each count as similar fact evidence on the other. The jury convicted M.L. on Count 2 (relating to S.L.) but was unable to reach a verdict on Count 4 (relating to A.J.), which was stayed. M.L. appealed the conviction on Count 2, arguing principally that the hearsay ruling was in error and, because A.J.’s statement had been placed before the jury as similar fact evidence on Count 2 as well, the error infected that conviction.
M.L. also raised two additional grounds of appeal: that the trial judge erred in refusing to declare a mistrial following an inflammatory Crown closing address, and that the jury charge was unbalanced. The Court of Appeal found it unnecessary to address those grounds once it resolved the hearsay issue in M.L.’s favour.
The Court’s Holding
Writing for a unanimous three-judge panel, Dawe J.A. held that the trial judge committed multiple errors in his threshold reliability analysis under the four-step framework set out in R. v. Bradshaw, 2017 SCC 35. Most fundamentally, the trial judge failed to identify—let alone adequately address—the two primary hearsay dangers the statement presented: (i) the risk that A.J.’s mother had misinterpreted A.J.’s non-verbal licking gesture and drawn an unwarranted conclusion about its meaning; and (ii) the risk that the bulk of A.J.’s disclosures were not spontaneous but were instead responsive to her mother’s leading and suggestive questions. The trial judge wrongly treated the entire statement as “unprompted” when only A.J.’s initial comment that “her pretty hurt” was truly spontaneous; the incriminating details were elicited after A.J.’s mother introduced, without prompting, concepts such as “special secrets” Papa had told A.J. not to tell her mother, and directly asked whether Papa had licked A.J.
The Court further held that the trial judge failed to consider, at the third and fourth steps of the Bradshaw analysis, the realistic alternative explanation that a very young child, sensing her mother’s distress and confronted with leading questions, may have constructed or embellished a narrative to satisfy her mother rather than reporting actual events. The trial judge also did not adequately turn his mind to whether contemporaneous cross-examination of A.J. would have materially assisted the jury—the central question under Bradshaw. Without confirmatory physical evidence and in the absence of a non-leading questioning process comparable to those in R. v. Khan or R. v. S.S., the hearsay dangers had not been adequately overcome. Threshold reliability was therefore not established, and the statement should not have been admitted.
Because the jury had been instructed it could use A.J.’s inadmissible hearsay statement as similar fact evidence on Count 2, and because it was reasonably possible that at least some jurors relied on that evidence to convict on Count 2, the conviction could not stand. The Court set aside the conviction and ordered a new trial on Count 2.
Key Takeaways
- Under the Bradshaw four-step framework, a trial judge must specifically identify the hearsay dangers relevant to the particular statement before assessing whether corroborative circumstances or other safeguards overcome them; a failure to advert to the primary dangers is a reviewable error of law.
- A child’s initial spontaneous utterance and the subsequent answers extracted through leading, suggestion-laden questions must be assessed separately for threshold reliability; characterizing the entire exchange as “unprompted” when only the opening remark was truly spontaneous is a material analytical error.
- The absence of physical confirmatory evidence is not excused merely because it is unsurprising; the question is whether, without such evidence, the hearsay dangers have actually been addressed—not whether their absence is consistent with the truth of the allegation.
- Improperly admitted hearsay tendered primarily on one count can necessitate a new trial on a separate count if the jury was instructed to use it as similar fact evidence across counts and reliance was reasonably possible.
Why It Matters
This decision reinforces the rigour courts must apply when admitting hearsay statements from very young children under the principled exception, particularly where the disclosure was elicited through parental questioning rather than emerging spontaneously or through a structured forensic interview. The judgment is a caution that a child’s initial unprompted comment does not carry forward and render the entire subsequent exchange reliable, especially where a parent’s distress and leading questions may have shaped the child’s responses. Defence counsel and trial judges should scrutinize carefully which portions of a disclosure conversation were truly spontaneous and which were prompted.
The case also illustrates the cascading effect of hearsay errors in multi-count trials where similar fact directions are given. Even though the impugned statement related to a count on which the jury hung, its improper admission as similar fact evidence on the companion count was enough to vitiate what appeared to be an otherwise unrelated conviction—underscoring the importance of getting hearsay gatekeeping right before evidence reaches the jury at all.