Court of Appeal for Ontario

Date: 2025-03-12
Docket: C69971
Coram: Rouleau, Roberts and Favreau JJ.A.

Between:
His Majesty the King (Respondent)
and
P.J.C. (Appellant)

Appearances:
Jolene Hansell, for the appellant
Jamie Klukach, for the respondent

Heard: October 30, 2024

On appeal from the convictions entered on August 17, 2021 and the sentence imposed on February 25, 2022 by Justice Kristin Muszynski of the Superior Court of Justice.

Reasons for Judgment

Rouleau J.A.:

Introduction

[1] The appellant was convicted of assaulting his three young stepchildren and sexually assaulting one of them between 1999 and 2001. He appeals his convictions and his sentence of 10 years’ imprisonment. The sole issue on the conviction appeal is whether the trial judge erred in admitting the evidence of Dr. Lori Haskell, an expert witness who testified on the neurobiology of trauma and its effects on memory.

[2] For the reasons that follow, I would allow the conviction appeal and order a new trial.

A. Facts

[3] The appellant began a relationship with the complainants’ mother in 1999. For about two years, he lived with her and her three children: her daughter E.M., and her sons K.M. and W.M. Their relationship ended in 2001. E.M. was seven years old, K.M. was five years old and W.M. was four years old when the appellant started dating their mother.

[4] In 2018, the appellant was arrested and charged with assaulting and sexually assaulting E.M., K.M., and W.M. After a six-day trial that took place in June of 2021, he was convicted of assaulting all three complainants and of sexually assaulting E.M. The appellant was acquitted of the charges of sexual assault involving K.M. and W.M.

(1) Evidence of Dr. Lori Haskell

[5] The Crown’s first witness at trial was Dr. Lori Haskell, a clinical psychologist. She was qualified to give expert opinion evidence on the neurobiology of trauma – including the encoding, storage, and retrieval of memories – the impact of trauma on memory, dissociative amnesia and trauma, recovered memories, and factors that diminish or enhance memories of traumatic events.

[6] The stated purpose of Dr. Haskell’s evidence was to provide a biological explanation for how people recall traumatic events, with particular focus on childhood sexual assault. The appellant’s trial counsel (not Ms. Hansell) took no issue with Dr. Haskell’s expertise, and the parties agreed to the trial proceeding as a blended voir dire and trial. Following Dr. Haskell’s testimony in chief, the defence cross-examined her, and the parties made submissions about the admissibility of her evidence at the conclusion of the trial.

(a) Dr. Haskell’s report

[7] Dr. Haskell did not interview the complainants. Rather, in preparing her opinion evidence, she reviewed transcripts of the complainants’ testimony at the preliminary inquiry as well as transcripts of their police interviews. Dr. Haskell then authored a report for the Crown, which was entered as a lettered exhibit on the voir dire. The Crown characterized this report as “an aid” to the trial judge in following Dr. Haskell’s evidence. As a lettered exhibit, this report does not constitute evidence. It does, however, form part of the trial record for appeal purposes and, as such, it can inform this court’s understanding of how Dr. Haskell’s evidence was proffered to support the Crown’s case at trial: Wasylyk v. Simcoe (County), 2023 ONCA 473, at para. 10.

[8] In that report, Dr. Haskell analyzes various portions of the complainants’ interviews and testimony and describes them as being consistent with how victims of childhood trauma would remember traumatic events. By way of example, her report discusses the phenomenon of “impaired semantic encoding” and illustrates it by referring to the answer that E.M. had given in her police interview when she was asked whether she remembered the appellant ejaculating when he had sexually assaulted her. The report notes that, “when asked by the detective whether the accused ejaculated, [E.M.] responds that she doesn’t remember.” Dr. Haskell goes on to explain that the “concept of ejaculation is not something a child of age seven can be expected to understand … [E.M.]’s failure to remember whether the accused ejaculated or not is an example of impaired semantic encoding, because as a child she most likely did not understand what ejaculation is.”

(b) Evidence on the voir dire

[9] Dr. Haskell gave evidence about how the trauma of sexual assault can be expected to impact a child’s memory of that event. She described the difference in how people encode, store, and retrieve central versus peripheral details of events in general, and of traumatic events in particular. Traumatic experiences, she said, are encoded as intense fragments rather than in chronological sequence, and how an individual encodes that memory also depends on their age and level of comprehension. Victims of childhood sexual assault, she explained, can be expected to have gaps in their memory and give accounts with inconsistencies in respect of peripheral details. They may also go for a period of time without being able to recall traumatic experiences or may not allow themselves to reflect or call upon those memories.

[10] Additionally, Dr. Haskell testified about “dissociative amnesia”, a phenomenon where a gap in memory is caused by an individual detaching from reality to cope with a traumatic event.

[11] The trial judge was told that Dr. Haskell had reviewed the complainants’ police interviews and preliminary inquiry testimony to prepare her evidence, but these were not entered into evidence at trial. However, as I will discuss further below, the features of the complainants’ pre-trial evidence on which Dr. Haskell focused were subsequently led as evidence at trial when they were elicited by the Crown in its examination of the complainants.

(c) The hypotheticals

[12] During the voir dire, the Crown put a series of questions to Dr. Haskell based on purportedly hypothetical situations. Although these each referenced a hypothetical child, they incorporated distinctive features of the complainants’ evidence that Dr. Haskell had already analyzed in her report. As noted above, these were features that the Crown subsequently elicited on direct examination of the complainants.

[13] The Crown’s five hypothetical questions are reproduced below with the answers Dr. Haskell gave on the voir dire:

• Hypothetical 1: “a child, approximately age seven, is able to recall being raped – multiple vaginal rapes, is not able to recall whether ejaculation took place.” Dr. Haskell testified that semantic encoding of that event would not be possible. A child would not have had a cognitive understanding of ejaculation so they could not encode that memory.

• Hypothetical 2: “child is repeatedly assaulted – sexually assaulted and physically assaulted between ages, say roughly seven to nine, and is now an adult, is that person gonna be able to give us exact sequence, number of events, the timeframe when the events took place” Dr. Haskell testified the adult in question would not be able to provide this information.

• Hypothetical 3: “a little boy roughly about age five is – is raped, is – is sodomized by an adult, and then has basically a blackout after that. And then the next recall is a memory of an unrelated normal event or normal conversation.” Dr. Haskell testified that this “would be consistent with how someone would describe dissociating. Exactly.”

• Hypothetical 4: “a child, male child around aged five…to seven is repeatedly physically abused, and is more than one sexual abused [sic], that person wondering whether some of the assaults on their person – sexual assaults are dreams or memories. What’s that?” Dr. Haskell testified this is “highly consistent with how dissociative people talk about their experiences”, and what she sees in childhood sexual abuse cases including historical sexual abuse cases.

• Hypothetical 5: “the young child, age five to seven, a victim of historic sexual abuse describing that their conscience is not letting them access their memories to the fullest. Can you comment on that, please?” Dr. Haskell said that would be someone “purposely [sic] not wanting to remember” because it would be “too overwhelming” to let these memories come to their consciousness.

(2) The complainants’ evidence

[14] When they testified at trial, E.M. was 29, K.M. was 27, and W.M. was 26. They all gave evidence about the appellant’s regular alcohol consumption and aggressive behaviour. E.M. testified that the appellant regularly assaulted all three children, including “smacking” and “spanking” them. She also alleged that, on several occasions, the appellant sexually assaulted her, including digital and penile penetration and forcing her to perform fellatio on him. E.M. would have been between the ages of seven and nine at the time.

[15] K.M. testified that the appellant regularly physically abused him and his siblings. Although his memory was foggy, he testified to being hit, kicked, pushed, and shoved up against the wall by the appellant. K.M. further alleged that the appellant had once attempted to sexually assault him, and that he had witnessed the appellant sexually assaulting W.M. K.M. would have been between the ages of five and seven at the time.

[16] W.M. also testified to regular assaults by the appellant. He said that the assaults progressed from flicking the back of his head to spanking, hitting, swatting with a fly swatter, and throwing items at him. W.M. gave evidence of being sexually assaulted by the appellant on two occasions. W.M. would have been between the ages of four and six at the time.

[17] The appellant denied these allegations. He testified that he could not recall a time when he was alone with the children in a bedroom, and that he never saw E.M. naked.

(3) The trial judge’s decisions

(a) Admissibility of Dr. Haskell’s expert evidence

[18] At the conclusion of the trial, the trial judge considered the admissibility of Dr. Haskell’s evidence. She found that it was “clearly relevant to the issues in this case” and that Dr. Haskell was qualified to give expert evidence on the neurobiology of trauma. The trial judge also found that Dr. Haskell’s evidence did not violate any exclusionary rule as she did not opine inappropriately on the credibility of the complainants. The evidence was necessary since “the issue of encoding, storage and retrieval of traumatic memories … is complex and nuanced and can assist in putting into context the evidence of the complainants”, and “Dr. Haskell’s evidence is outside the knowledge and experience of a typical Canadian person”. The trial judge concluded that Dr. Haskell’s evidence ought to be admitted because it provided “an explanation for the inconsistencies or gaps in the evidence of the complainants”. She then specified she treated it as “only one factor that [she] consider[ed] when [she] weigh[ed] the evidence in this trial.”

[19] The trial judge did not comment on the evidentiary value of the report. However, as noted above, it was entered as a lettered exhibit in the voir dire on the understanding that it could serve as “an aid” to the trial judge in following Dr. Haskell’s evidence.

(b) Reasons for conviction

[20] In her reasons for decision on the trial proper, the trial judge found Dr. Haskell’s evidence reliable. With respect to how she applied it, the trial judge explained that:

[W]hile Dr. Haskell’s evidence provides the court with relevant and helpful information and puts the evidence of the complainants in context, it is only one factor, and a minor one at that, that I consider when reaching my ultimate conclusion given the circumstances in this case for the reasons that will become evident as I assess the evidence of the complainants.

[21] In her reasons for conviction, the trial judge explained that she found the evidence of the three complainants to be credible and reliable with respect to the physical assaults, and E.M.’s evidence to be credible and reliable with respect to the sexual assault.

[22] The trial judge rejected the appellant’s evidence, finding it fraught with inconsistencies and, at times, “fanciful and far-fetched.” She concluded that the appellant was not a credible witness and had been untruthful. The trial judge entered a conviction on the assault charges and on the charge of sexually assaulting E.M. She was, however, left with a reasonable doubt with respect to the alleged sexual assaults of K.M. and W.M., and found the appellant not guilty of those charges.

B. Issues

[23] There are three issues on appeal:

1. Did the trial judge err in admitting Dr. Haskell’s evidence;
2. If the trial judge erred in admitting Dr. Haskell’s evidence, should this Court apply the curative proviso; and
3. Was the 10-year sentence imposed unfit.

[24] The appellant also brought a fresh evidence application but abandoned it at the hearing.

C. Analysis

(1) The law

[25] The law regarding the admission of expert evidence is well settled. Expert evidence is presumptively inadmissible. To be admitted, the proponent must satisfy the four threshold requirements of admissibility: relevance, necessity, absence of other exclusionary rules, and expert qualification. If these requirements are met, the trial judge must then conduct a cost-benefit analysis, balancing the probative value of the evidence against its prejudicial effect: White Burgess Langille Inman v. Abbott and Haliburton Co., 2015 SCC 23, paras 19, 21-24.

(2) The positions of the parties

[26] The appellant argues that Dr. Haskell’s evidence ought not to have been admitted and that the trial judge erred in her assessment of that evidence. In the appellant’s submission, the evidence was neither relevant nor necessary; it violated the rule against oath-helping and ought to have been rejected based on a proper cost-benefit analysis.

[27] Although trial counsel did not object to the admission of Dr. Haskell’s evidence, he did argue that her evidence could not be used to bolster the complainants’ credibility. The appellant submits that the trial judge remains the ultimate evidentiary gatekeeper and had the “duty to vigilantly assess and exclude inadmissible evidence regardless of the position taken (or not taken) by counsel”: R. v. A.L., 2020 BCCA 18, para 150. It is therefore open to this court to find that the trial judge erred in admitting the evidence of Dr. Haskell. In any event, the appellant argues that the trial judge did precisely what trial counsel warned against and used Dr. Haskell’s evidence to bolster the complainants’ credibility.

[28] The Crown maintains that the trial judge’s decision to admit Dr. Haskell’s evidence was reasonable, free of legal and factual error, and is entitled to deference. In the Crown’s submission, it is apparent from the trial judge’s reasons and the fact that she acquitted the appellant of sexually assaulting K.M. and W.M. that the trial judge did not misuse Dr. Haskell’s evidence. If this court were to conclude that Dr. Haskell’s evidence ought not to have been admitted, the Crown submits that we should apply the curative proviso and uphold the convictions.

(3) Discussion

[29] In my view, Dr. Haskell’s expert evidence ought not to have been admitted. As I will explain, it did not meet the necessity requirement for the admission of expert evidence. Further, the way in which it was presented went beyond providing context for understanding the complainants’ evidence. Through the use of hypotheticals tracking the specific allegations and behaviours of the complainants, Dr. Haskell’s opinion evidence improperly bolstered their credibility and served to blunt any prospective cross-examination of the complainants. Accordingly, even if the evidence had been necessary, it should have been excluded because its prejudicial effect outweighed any probative value it may have had.

[30] I give limited weight to the fact that no objection was made by trial counsel to the admission of Dr. Haskell’s evidence. Trial counsel expressed concern that it ought not to be used to bolster the complainants’ credibility. As this court explained in R. v. Nguyen, 2023 ONCA 531, para 54, citing White Burgess, “given how prejudicial opinion evidence can be, trial judges are assigned a gatekeeping role”. Whether the defence specifically objects to the admissibility of expert opinion evidence does not impact the trial judge’s responsibilities as gatekeeper: R. v. Palmer-Coke, 2019 ONCA 106, para 21; R. v. K.(A.), para 73, n. 8, application for leave to appeal quashed, [2000] S.C.C.A. No. 16. In the present case, the trial judge erred in her exercise of that role.

(a) Necessity

[31] As noted earlier, opinion evidence is presumptively inadmissible. It is well established that witnesses may testify as to facts, but as a general rule may not give their opinion about those facts. Experts are often described as exempted from this rule, but it is only when the trier of fact is unable to form a correct judgment on a matter without the help of an expert that their opinion evidence will be properly admitted.

[32] In R. v. D.D., 2000 SCC 43, para 47, Major J., writing for the majority, explained that, in order to meet the necessity requirement, expert opinion evidence must be necessary to enable the trier of fact either (1) to appreciate the matters in issue due to their technical nature or (2) to form a correct judgment on a matter if ordinary persons are unlikely to do so without the assistance of persons with special knowledge. The same issue had been previously addressed in R. v. Mohan, para 24, where the court cited with approval Lawton L.J.’s statement in R. v. Turner, [1975] Q.B. 834, at p. 841 to the same effect:

If on the proven facts a judge or jury can form their own conclusions without help, then the opinion of an expert is unnecessary. In such a case if it is given dressed up in scientific jargon it may make judgment more difficult.

[33] The primary danger posed by the admission of expert evidence is, as explained by Major J. in D.D., at para. 53, that “faced with an expert’s impressive credentials and mastery of scientific jargon, jurors are more likely to abdicate their role as fact-finders and simply attorn to the opinion of the expert in their desire to reach a just result.” In judge-alone trials, “the need to draw the line properly between the role of the expert and the role of the court” remains an animating concern when determining the admissibility of expert evidence: R. v. J.-L.J., 2000 SCC 51, para 26; see also R. v. Sekhon, 2014 SCC 15, para 46.

[34] In D.D., at para. 57, Major J. adopted Professor Paciocco’s (as he then was) observation in “Expert evidence: Where are we now? Where are we going?” (1998), at pp. 16-17, explaining that it is not enough that the expert evidence be helpful; it must be necessary in the sense that “lay persons are apt to come to a wrong conclusion without expert assistance, or where access to important information will be lost unless we borrow from the learning of experts”.

[35] Here the trial judge found Dr. Haskell’s evidence met the necessity requirement because “encoding, storage, and retrieval of traumatic memories … is complex and nuanced, and can assist in putting into context the evidence of the complainant” and her testimony “is outside the knowledge and experience of a typical Canadian person.”

[36] There is no doubt that Dr. Haskell’s evidence on the neurobiology of trauma was technical and outside the knowledge and experience of the average Canadian. It is also true that understanding the neurobiology of the brain may be helpful in that it provides explanations for why people who suffer trauma as a child may later, as adults, have gaps in their memories of these traumatic events. Similarly, as Dr. Haskell explained, semantic encoding provides an explanation as to why someone like E.M., who recalls having been sexually abused as a child, might not remember whether her abuser ejaculated. However, as explained by the Supreme Court, the fact that something is outside the knowledge of a lay person and may be helpful is not sufficient to meet the necessity requirement for admission of expert evidence. The question the trial judge ought to have asked was whether, in the present case, she could correctly assess and interpret the complainants’ testimony without the help of Dr. Haskell’s evidence. In other words, was an understanding of the functioning of the brain, the neurobiology of trauma, and semantic encoding, amongst other things, necessary to assess whether the three complainants were credible, and whether their evidence was reliable. By equating helpfulness with necessity, the trial judge erred.

[37] Significant deference is owed to a trial judge’s decision to admit or exclude expert evidence, including findings on the issue of necessity, unless an error in principle can be shown or the decision is found to be unreasonable: see R. v. McManus, 2017 ONCA 188, para 68; R. v. Hoggard, 2024 ONCA 613, para 30. As I have explained, however, the trial judge effectively ended her inquiry after determining that the technical evidence given by Dr. Haskell was outside the knowledge of the average person and may be helpful. Finding that expert evidence might reasonably assist the trier of fact is not enough to make out necessity: D.D., at para. 47. The trial judge ought to have gone on to determine whether that evidence met the legal threshold of necessity, in that it was required to appreciate the complainants’ evidence.

[38] Applying the appropriate test, I conclude that, in the context of this case, the fact that Dr. Haskell’s testimony with respect to the neurobiology of trauma is outside the knowledge of the average person is, in a sense, irrelevant. The assessment of the credibility and reliability of witnesses, including those testifying about traumatic events they suffered as children (or as adults), is the daily fare of judges and is grounded in common knowledge and experience. In general, triers of fact do not need to know how the brain functions and encodes memories to assess the probative value of this evidence. The Supreme Court and this court have confirmed that trial judges are equipped to analyze the evidence of witnesses with frailties in their memory without the assistance of expert evidence: see R. v. H.S.B., 2008 SCC 52, paras 14-15; R. v. G.M.C., 2022 ONCA 2, para 38; R. v. Perlett, para 108; and R. v. Ferguson, para 73, per Laskin J. (dissenting) (reasons substantially adopted in R. v. Ferguson, 2001 SCC 6).

[39] The thrust of Dr. Haskell’s evidence was that people, including children, remember traumatic events in different ways, and that sometimes these memories contain gaps, or reflect periods of dissociation. It also explained that a child might not encode a detail of a sexual assault such as ejaculation because they would not have been familiar with its occurrence. While information on how the brain functions and encodes is interesting and perhaps even marginally helpful, I am not persuaded that this evidence was necessary or required to enable the trial judge to assess the credibility and reliability of the complainants in this case.

[40] In my view, the average person is well able to assess the credibility and reliability of witnesses such as the complainants without Dr. Haskell’s evidence. As explained by Strathy C.J.O. in G.M.C., at para. 38, people can reasonably be expected to understand that observations made in the course of traumatic events can be difficult to recall and describe accurately, and the inability to recall minor or insignificant aspects of a traumatic event does not detract from a witness’ overall reliability or credibility; these are propositions “grounded in common experience, that are familiar to every trial judge and lawyer and to lay people.”

[41] In G.M.C., Strathy C.J.O. was discussing adults testifying about traumatic events suffered while they were adults. His comments apply all the more when the witness is an adult testifying as to traumatic events suffered when they were children. There is a longstanding recognition in our law that memories formed as a child can be expected to contain gaps in respect of details, including time and place: see R. v. W.(R.), at pp. 133-34; R. v. B.(G.), at pp. 54-55. The standard for assessing the credibility of adult complainants giving evidence about harm they suffered as children accounts for these expected frailties: see R. v. A.M., 2014 ONCA 769, para 11.

[42] That said, I should not be taken as suggesting that the expert evidence of psychologists or psychiatrists on the functioning of the brain or on human conduct could never meet the necessity requirement. There are situations where, as explained by Rosenberg J.A. in R. v. M.(B.), expert evidence about the functioning of memory may be necessary. In M.(B.), Rosenberg J.A. suggested that, although expert evidence about the effect of trauma on memory would not generally meet the necessity requirement, a narrow expert opinion on memory may have been admissible in that case to assess the reliability of testimony based on a memory the witness had allegedly formed as a two-year-old, because that might be outside the normal experience of a trier of fact: at para. 101. The facts of M.(B.), however, were unusual and the issue Rosenberg J.A. seized on is quite distinct from those before the court in this case. At its core, Dr. Haskell’s evidence did little more than give a neurobiological explanation of why memories of traumatic childhood events may contain gaps and inaccuracies.

[43] I acknowledge that distinguishing what may be helpful but not necessary from what is truly necessary, in a legal sense, is not always easy. Where that line must be drawn in any given case depends on the particular facts of that case: see K.(A.), para 76. Importantly, the sort of expert evidence that will meet the necessity threshold can be expected to change over time as our common knowledge and understanding evolves: see R. v. F.(D.S.), at pp. 115-16. In 1993, when it decided R. v. Marquard, the Supreme Court reasoned that expert evidence explaining why young victims of sexual abuse often do not complain immediately “may be essential to a just verdict” because the ordinary inference from the failure to complain promptly would likely have been that the story is fabricated. However, less than 10 years later in D.D., a case dealing with similar issues but in a different factual context, Major J. found it surprising that, by then, a Canadian jury or judge would be incapable of understanding that a child victim of sexual abuse would not necessarily disclose the fact of the abuse immediately. Accordingly, he held that expert opinion on this point was not necessary.

[44] In D.D., Major J. also warned against the systemic prejudice caused by the admission of unnecessary expert evidence in legal proceedings. These concerns remain apposite. If expert evidence on the neurobiology of trauma were to be deemed necessary for finders of fact in cases such as the present to assess gaps and defects in witnesses’ memory, this would, as explained by Major J., invariably increase the length, cost and complexity of many trials resulting in unnecessary strain upon judicial resources. When the door to the admission of expert evidence is opened too widely, a trial has the tendency to degenerate into “a contest of experts with the trier of fact acting as referee in deciding which expert to accept”: D.D., at para. 56, citing Mohan, at para. 24; see also Hoggard, para 44.

(b) Overriding prejudicial effect

[45] Over and above the concern as to lack of necessity, I consider the manner of presentation of Dr. Haskell’s evidence to be problematic and unduly prejudicial. As I will explain, Dr. Haskell’s evidence went beyond explaining the neurobiology of trauma and its effects on memory. The Crown fashioned hypothetical questions for Dr. Haskell to answer that closely paralleled the allegations of sexual and other physical abuse as well as the frailties and gaps in the complainants’ memories. The hypotheticals were designed to elicit an opinion from Dr. Haskell that the gaps and frailties in the hypothetical subject’s memory – and therefore the complainants’ memories – were consistent with the way people who suffered such abuse would remember the incidents. They anticipated Crown-led evidence from the complainants to inure the trier of fact to those identified gaps and frailties thereby improperly bolstering the complainants’ credibility and reliability.

(i) The hypotheticals

[46] The Crown argues that the hypotheticals it put to Dr. Haskell were tailored to be efficient and helpful, and the trial judge was entitled to use her answers in evaluating the complainants’ evidence. In its submission, the fact that the hypotheticals were, to a certain extent, relevant to features in the complainants’ testimony at trial and might support their credibility did not cause Dr. Haskell to give evidence that was impermissibly about credibility. The Crown maintains that, but for Dr. Haskell’s evidence, the appellant would have improperly relied on the gaps and frailties in the complainants’ memories to impugn their credibility and reliability. Dr. Haskell’s evidence therefore assisted the trial judge by alerting her to the fact that the inference the defence would be asking her to draw was not the only one available to her.

[47] In my view, after hearing all of the evidence, the trial judge should have recognized that the Crown’s hypotheticals were so close to the actual facts of the case that they prompted Dr. Haskell to provide evidence that improperly bolstered the complainants’ credibility and reliability, and that the prejudicial effect of this evidence far exceeded its probative value. I acknowledge that opinion evidence is not necessarily inadmissible by virtue of it supporting the credibility of a witness: R. v. Oppong, 2021 ONCA 352, para 58; K.(A.), para 96. Still, the outcome of this case was entirely dependent on findings of credibility and reliability, and the hypotheticals risked making the gaps and frailties into features supportive of the complainants’ credibility and reliability and, potentially, evidence of their having suffered the assaults as alleged.

[48] I recognize that properly framed hypotheticals can and should at times be used to elicit entirely appropriate expert opinions: see R. v. G.(P.), 2009 ONCA 32, paras 28-30. Yet, it does not follow that any hypothetical question can properly be put to a testifying expert. Trial judges must remain alive to the possibility that certain hypothetical questions are liable to prompt answers that, viewed in the context of all the evidence presented at trial, are far more prejudicial than they are probative.

[49] In K.(A.), this court cautioned that hypothetical questions should not exceed the legitimate purpose for the expert testimony being presented. K.(A.) was a case involving allegations of sexual assault on children, some historical. The Crown proffered, and the trial judge admitted, expert evidence “to explain that certain behaviour exhibited by the complainants was not unusual for victims of sexual abuse and was not inconsistent with the truth of their allegations”: at para. 59. On appeal, Charron J.A. found that a substantial portion of this evidence was “not sufficiently probative to warrant admission”. That included evidence about “faulty memories in victims of sexual abuse”: at para. 114. Still, she found that some of the expert opinion on a limited subset of behaviours was reliable and cogent and would have been admissible had it not gone beyond expressing an opinion about the occurrence of those behaviours. The problematic evidence had been elicited, in large part, in response to “hypothetical questions that paralleled the specific allegations and the behaviour of the individual complainants”: at para. 131. Charron J.A. held that, “[p]resented in this manner, the evidence was no longer admissible because its prejudicial effect far exceeded its probative value. It was obviously calculated to bolster the credibility of each complainant and as such was not the proper subject of expert testimony”: at para. 135; see also R. v. P.(C.), at pp. 486-87.

[50] Charron J.A. cautioned that hypothetical questions posed to an expert should help the court understand the features of behaviour generally, but not invite the court to “accept each complainant’s evidence regardless of their conduct and notwithstanding the weaknesses of the testimony.” This manner of eliciting expert evidence ought not to “effectively mask the weaknesses of each complainant’s testimony and to cloak it with a semblance of scientific reliability”: K.(A.), at para. 135.

[51] That, however, is precisely what the hypotheticals put to Dr. Haskell did in this case. The first and second hypotheticals, which related to E.M.’s allegations of sexual abuse, exemplify questions giving rise to the concerns expressed by Charron J.A. The first hypothetical was of a child who had been repeatedly sexually assaulted but is unable to recall whether her assailant ejaculated in the course of the assaults. The second hypothetical featured an adult who had been repeatedly sexually and physically assaulted between the ages of seven and nine, and who could not recall the exact sequence, number of events and timeframe when the events took place. Both questions were based on features of E.M.’s evidence that Dr. Haskell had previously noted and analyzed in her report.

[52] In her response to these hypotheticals, and to the others that were put to her, Dr. Haskell explained how the hypothetical subjects, who had been abused, would be expected to exhibit gaps and frailties in their memories, and explained how these gaps and frailties were attributable to the neurobiology of trauma. This evidence did more than equip the trier of fact with information that would assist in assessing the impact, if any, these gaps and frailties in the complainants’ memories could have on their credibility and reliability. Rather than contextualize these features of the complainants’ evidence, it impermissibly suggested that they were more credible because they exhibited the features the expert attributed to persons having suffered prior sexual abuse. Although not explicitly stating that the answers given by the complainants were consistent with sexual abuse, which would be impermissible, the impression left by the testimony, in the context of this case, was in effect the same. [2]

[53] The frailties of memory described in the first and second hypotheticals did not cry out for a scientific explanation. It is hardly surprising that a witness recounting instances of abuse she alleges occurred two decades earlier could not recall the “exact sequence, number of events, the timeframe when the events took place”, or whether their abuser ejaculated. Yet, by relating these frailties in memory to “dissociative amnesia” and “impaired semantic encoding”, Dr. Haskell indirectly “[argued] the theory of the Crown from the witness box and [clothed] it with the aura of infallibility … [associated] with scientific principles”: R. v. D.D., para 17, aff’d 2000 SCC 43.

(ii) How the evidence unfolded

[54] The prejudice occasioned by the hypotheticals was compounded by the manner and sequence in which the evidence was presented. The Crown’s position on appeal is that Dr. Haskell’s evidence was appropriately led to provide the trier of fact with a more complete picture of the complainants’ evidence, particularly their flawed memories. After Dr. Haskell provided expert evidence on how victims of childhood abuse would testify, the Crown’s examination of the complainants in effect prompted them to testify precisely in that way. As a result, the way the evidence unfolded could reasonably have created the impression that because the complainants answered the Crown’s questions in a manner that would be expected of persons having suffered abuse, the complainants’ answers were evidence suggesting that they had in fact suffered abuse as children – in other words, that their allegations of abuse were credible and reliable.

[55] Consequently, quite apart from the hypotheticals themselves, the effect of calling the evidence in this way increased the risk that Dr. Haskell’s evidence would serve to bolster the complainants’ credibility and reliability, and ultimately support the Crown’s case. This goes beyond providing, as would be appropriate, an alternative inference to one the defence might urge, namely that the gaps and frailties suggest that the complainants were not credible or reliable.

[56] I refer once more to the first and second hypotheticals discussed above to illustrate the point. The first hypothetical and Dr. Haskell’s answer were as follows:

Q. So, the hypothetical. First one would be a child, approximately age seven, is able to recall being raped – multiple vaginal rapes, is not able to recall whether ejaculation took place. Can you comment on that?
A. I mean a – a child wouldn’t understand what ejaculation is. So semantic encoding – encoding that event wouldn’t be possible. I mean it’s – it’s – it would be a process they weren’t familiar with that would be unknown to them. It also would be something that, depending on if they were dissociated to their body, they might not be trying to feel it, then it would be highly unlikely that a child – I – I wouldn’t even ask a child that question. I would ask the question more openly … A child wouldn’t understand that experience. They wouldn’t have cognitive understanding. And we don’t encode things that we cognitively don’t understand.

[57] In direct examination of E.M., the Crown questioned as follows:

Q. Okay. Do – do you remember, [E.M.], as to whether – now, so I’m talking about as a child, lemme ask you this. First of all, when you were seven, eight, nine, did you know what ejaculation was?
A. No.
Q. Okay. Do you know, or can you recall when he’s had intercourse with you on any time, whether it’s the bedroom, wherever, place in the house, would he ejaculate?
A. I’m not able to remember.

[58] The second hypothetical and Dr. Haskell’s response were as follows:

Q. A second hypothetical child is repeatedly assaulted – sexually assaulted and physically assaulted between ages, say roughly seven to nine, and is now an adult, is that person gonna be able to give us exact sequence, number of events, the timeframe when the events took place?
A. No, they would not be able to.

[59] The Crown then proceeded to elicit evidence from Dr. Haskell about dissociative amnesia to further contextualize.

[60] Later, in direct examination of E.M., the Crown questioned as follows:

Q. And at that house, did he sexually assault you on just one occasion or more than one occasion?
A. More than one occasion.
Q. Do you know the exact number of times?
A. No.
Q. Was it just a few times or a lot of times?
A. A lot.
Q. Do you remember the exact order of all those sexual assaults?
A. No.
Q. And do you recall the exact order of physical abuse? Each – each one, the order it took place?
A. No, I do not.
Q. Do you recall the exact number?
A. No.

[61] This essentially ensured that the features Dr. Haskell had addressed through hypotheticals were displayed by the complainants in their testimony at trial. The manner of presentation heightened the danger that, as explained in K.(A.), the expert evidence may have left the trial judge “with the impression not only that they should not discredit the complainants’ testimony on the basis of any of these features (a notion which in itself is legally incorrect), but that there was a scientific basis to conclude that the presence of these very features … somehow made their testimony more credible”: at para. 130.

[62] Put differently, such evidence risks being misused to reason backwards. As explained in Hoggard, because the testimony given by the complainants was consistent with how an expert expects someone having suffered a trauma would testify, it becomes easier for the trier of fact to conclude that the complainants in fact suffered the alleged trauma.

[63] As a result, even if I had found Dr. Haskell’s evidence met the necessity requirement, due to the hypotheticals used and how the evidence unfolded, I am of the view that it carried a prejudicial effect that far outweighed any probative value. For these additional reasons, I conclude that the trial judge erred in admitting the evidence.

(c) Applicability of the curative proviso

[64] The Crown argues that even if Dr. Haskell’s evidence was unnecessary and the trial judge erred in admitting it, the error occasioned no substantial wrong or miscarriage of justice. The evidence was reliable and not misleading, and no objection was taken by the appellant’s trial counsel to its admission. This was a judge alone trial, and, in the Crown’s submission, it can safely be assumed that the trial judge did not misuse Dr. Haskell’s evidence. As stated in her reasons, the trial judge considered Dr. Haskell’s expert opinion to be only one factor in her analysis, and a “minor one at that.” There was, therefore, little risk that the trial judge would misuse that evidence or give it undue weight. This, the Crown maintains, is in a sense borne out by the trial judge having acquitted the appellant on the sexual assault counts involving K.M. and W.M. The trial judge gave clear and cogent reasons for accepting the evidence of the complainants such that, the Crown argues, this court can safely conclude that the admission of Dr. Haskell’s evidence occasioned no substantial wrong or miscarriage of justice. I disagree.

[65] The curative proviso at s. 686(1)(b)(iii) of the Criminal Code, R.S.C. 1985, c. C-46 applies only where “the evidence against the accused was so overwhelming that any other verdict would have been impossible to obtain” or where the legal error is “so harmless or minor that it could not have had any impact on the verdict”: R. v. Van, 2009 SCC 22, para 34. This case is not one where the evidence of the appellant’s guilt was overwhelming, nor does the Crown pursue that line of argument.

[66] This was a credibility case. Once the appellant’s evidence was rejected and did not raise a reasonable doubt, the prosecution would succeed or fail depending on the trial judge’s assessment of the complainants’ credibility and reliability. They, along with Dr. Haskell, were the only Crown witnesses. The trial judge clearly had concerns with some of the complainants’ testimony as she acquitted the appellant of some of the charges. Although the trial judge states that in assessing the complainants’ evidence, Dr. Haskell’s evidence was “only one fact, and a minor one at that”, it is impossible to determine whether it was what tipped the scales, meaning that it could have been what dispelled any remaining doubt the trial judge may have had with respect to one or all of the charges on which she entered a conviction.

[67] Further, I am not convinced that the trial judge did not misuse Dr. Haskell’s evidence. As I have explained, Dr. Haskell’s evidence, had it been necessary, would have been relevant and admissible to provide the trier of fact with a more complete picture when assessing the credibility and reliability of the complainants. However, the hypotheticals used and the answers Dr. Haskell gave far exceeded this proper purpose and risked being improperly used to support the allegations being made by the complainants and improperly bolstering their credibility and reliability.

[68] Additionally, certain passages of the trial judge’s reasons suggest she could have misused aspects of Dr. Haskell’s evidence besides that which was elicited through the inappropriate hypotheticals. For example, when she explains, “I accept Dr. Haskell’s evidence that delayed reporting is a common feature in cases of childhood sexual assault.” This phrase could be interpreted as the trial judge using Dr. Haskell’s opinion evidence as supporting a finding of sexual abuse rather than simply placing the delayed reporting and other features of the complainants’ evidence in proper context. Although I might normally view this as simply an unfortunate turn of phrase, given the trial judge’s error in admitting Dr. Haskell’s evidence and in light of the prejudicial manner of presentation, I cannot be satisfied that it is not an indication that the features such as the delayed reporting were not interpreted as supportive of there having been childhood sexual assaults.

[69] I acknowledge that in Hoggard this court applied the curative proviso despite having found the trial judge erred in admitting expert evidence from Dr. Haskell. In that case, however, the circumstances were quite different. Significantly, Dr. Haskell’s evidence had been “pared down” by the trial judge and was more in the nature of a lecture focused on dispelling the myths and stereotypes associated with victims of sexual assault: at para. 16. The trial judge also gave very clear instructions to the jury on the “very small role” Dr. Haskell played and that her evidence could not be used to explain the particular evidence in the case: at paras. 23, 71. In contrast, in the present case, as reviewed previously, Dr. Haskell’s evidence improperly bolstered the complainants’ credibility and reliability and it is not possible to reasonably conclude that the trial judge did not misuse her evidence.

[70] In my view, therefore, it is not safe to conclude that the admission of Dr. Haskell’s evidence could not have affected the verdict. Accordingly, I would decline to apply the curative proviso.

D. Disposition

[71] I would allow the conviction appeal, set aside the convictions and order a new trial. In light of this conclusion, I need not deal with the sentence appeal.

Released: March 12, 2025

“P.R.”
“Paul Rouleau J.A.”
“I agree. L.B. Roberts J.A.”
“I agree. L. Favreau J.A.”

[1] This appeal is subject to a publication ban pursuant to s. 486.4 of the Criminal Code, R.S.C. 1985, c. C-46.

[2] In her report, a lettered exhibit, Dr. Haskell went somewhat further. For example, she stated her conclusion that “[E.M.]’s failure to remember whether the accused ejaculated or not is an example of impaired semantic encoding, because as a child she most likely did not understand what ejaculation is.”