Supreme Court of Canada **R. v. W.H., 2013 SCC 22, [2013] 2 S.C.R. 180** **Date:** 2013-04-19

Docket: 34522 --- Between: Her Majesty The Queen — Appellant and W.H. — Respondent --- Coram: McLachlin C.J. and LeBel, Fish, Rothstein, Cromwell, Moldaver and Karakatsanis JJ. Reasons for Judgment: (paras. 1 to 49) Cromwell J. (McLachlin C.J. and LeBel, Fish, Rothstein, Moldaver and Karakatsanis JJ. concurring) --- Indexed as: R. v. W.H. 2013 SCC 22 File No.: 34522. 2013: January 21; 2013: April 19. Present: McLachlin C.J. and LeBel, Fish, Rothstein, Cromwell, Moldaver and Karakatsanis JJ. on appeal from the court of appeal for newfoundland and labrador --- Criminal law — Appeals — Unreasonable verdict — Role of appellate court when assessing reasonableness of verdict based on jury's assessment of witness credibility — Jury finding accused guilty of sexual assault — Court of Appeal concluding that verdict unreasonable and entering acquittal — Whether Court of Appeal applied proper legal test — Whether verdict unreasonable — Criminal Code, R.S.C. 1985, c. C‑46, s. 686(1) (a)(i). A jury found the accused guilty of sexual assault of his niece which occurred when she was between the ages of 12 and 14. At trial, the complainant had testified about several sexual incidents involving the accused. The accused had testified in his own defence and denied any sexual activity with the complainant. On appeal, the Court of Appeal was troubled by a number of inconsistencies in the complainant's testimony and previous statements. However, it noted that nothing appeared from the transcript which should have caused a juror to question the accused's truthfulness. In light of those factors, the court was of the view that no experienced judge sitting alone would have been able to provide adequate reasons for the guilty verdict. The Court of Appeal concluded that the verdict was unreasonable, set it aside and entered an acquittal. Held: The appeal should be allowed and the conviction restored. A verdict is unreasonable or cannot be supported by the evidence if it is one that a properly instructed jury acting judicially could not reasonably have rendered. Appellate review of a jury's verdict of guilt must be conducted within two well‑established boundaries. On one hand, the reviewing court must give due weight to the advantages of the jury as the trier of fact who was present throughout the trial and saw and heard the evidence as it unfolded. The reviewing court must not act as a "13th juror" or simply give effect to vague unease or lurking doubt based on its own review of the written record or find that a verdict is unreasonable simply because the reviewing court has a reasonable doubt based on its review of the record. On the other hand, however, the review cannot be limited to assessing the sufficiency of the evidence. A positive answer to the question of whether there is some evidence which, if believed, supports the conviction does not exhaust the role of the reviewing court. Rather, the court is required to review, analyse and, within the limits of appellate disadvantage, weigh the evidence and consider through the lens of judicial experience, whether judicial fact‑finding precludes the conclusion reached by the jury. Thus, in deciding whether the verdict is one which a properly instructed jury acting judicially could reasonably have rendered, the reviewing court must ask not only whether there is evidence in the record to support the verdict, but also whether the jury's conclusion conflicts with the bulk of judicial experience. The traditional test for unreasonable verdict applies to cases such as this one in which the verdict is based on an assessment of witness credibility. In applying the test, the court of appeal must show great deference to the trier of fact's assessment of witness credibility given the advantage it has in seeing and hearing the witnesses' evidence. This applies with particular force to verdicts reached by juries. It is for the jury to decide, notwithstanding difficulties with a witness's evidence, how much, if any, of the testimony it accepts. Credibility assessment does not depend solely on objective considerations such as inconsistencies or motive for concoction; accordingly, the jury is entitled to decide how much weight it gives to such factors. The reviewing court must be deferential to the collective good judgment and common sense of the jury. While appellate review for unreasonableness of guilty verdicts is a powerful safeguard against wrongful convictions, it is also one that must be exercised with great deference to the fact‑finding role of the jury. Trial by jury must not become trial by appellate court on the written record. The Court of Appeal in this case applied the wrong legal test and, in carrying out its review of the jury's verdict, failed to give sufficient deference to the jury's assessment of witness credibility. The test to be applied by courts of appeal in reviewing guilty verdicts for unreasonableness does not involve the reviewing court attempting to put itself in the place of an imaginary trial judge and on a review of the written record asking whether that imaginary judge could have articulated legally adequate reasons for conviction. The Court of Appeal's adoption of this new test resulted in its failure to take a sufficiently deferential approach to the findings of the jury viewed, as they must be, in the context of the whole of the evidence. --- ## Cases Cited Referred to: Kienapple v. The Queen, [1975] 1 S.C.R. 729; R. v. S.J.M., 2009 ONCA 244, 247 O.A.C. 178; R. v. Yebes, [1987] 2 S.C.R. 168; R. v. Biniaris, 2000 SCC 15, [2000] 1 S.C.R. 381; R. v. Beaudry, 2007 SCC 5, [2007] 1 S.C.R. 190; R. v. Sinclair, 2011 SCC 40, [2011] 3 S.C.R. 3; R. v. W. (R.), [1992] 2 S.C.R. 122; R. v. C. (R.), [1993] 2 S.C.R. 226, rev'g (1992), 49 Q.A.C. 37; R. v. François, [1994] 2 S.C.R. 827; R. v. Burke, [1996] 1 S.C.R. 474; R. v. R.P., 2012 SCC 22, [2012] 1 S.C.R. 746; R. v. R.E.M., 2008 SCC 51, [2008] 3 S.C.R. 3; R. v. Gagnon, 2006 SCC 17, [2006] 1 S.C.R. 621; R. v. W. (D.), [1991] 1 S.C.R. 742. ## Statutes and Regulations Cited Criminal Code, R.S.C. 1985, c. C‑46, ss. 675(1)(a), 686(1)(a). --- APPEAL from a judgment of the Newfoundland and Labrador Court of Appeal (Wells, Barry and White JJ.A.), 2011 NLCA 59, 312 Nfld. & P.E.I.R. 12, 278 C.C.C. (3d) 237, 89 C.R. (6th) 181, 971 A.P.R. 12, [2011] N.J. No. 330 (QL), 2011 CarswellNfld 319, setting aside the accused's conviction for sexual assault and entering an acquittal. Appeal allowed and conviction restored. Frances J. Knickle, for the appellant. Peter E. Ralph, Q.C., and Michael Crystal, for the respondent. --- The judgment of the Court was delivered by Cromwell J. — ## I. Overview and Issue [1] A jury found the respondent guilty of sexual assault, but the Court of Appeal concluded that the jury's finding was unreasonable, set it aside and entered an acquittal. The Crown appeals, arguing that the Court of Appeal wrongly substituted its assessment of witness credibility for that of the jury. The appeal therefore raises the issue of an appellate court's role when it assesses the reasonableness of a jury's guilty verdict based on the jury's assessment of witness credibility. [2] Of course, a jury's guilty verdict based on the jury's assessment of witness credibility is not immune from appellate review for reasonableness. However, the reviewing court must treat the verdict with great deference. The court must ask itself whether the jury's verdict is supportable on any reasonable view of the evidence and whether proper judicial fact-finding applied to the evidence precludes the conclusion reached by the jury. Here, the Court of Appeal did not follow this approach. It asked itself instead whether an experienced trial judge could give adequate reasons to explain the finding of guilt and, having answered that question in the negative, found the verdict unreasonable. In my respectful view, the Court of Appeal applied the wrong legal test and reached the wrong conclusion. [3] I would allow the appeal and restore the conviction entered at trial. ## II. Facts and Proceedings ### A. Overview of the Facts [4] The respondent was charged with sexual assault of and sexual interference with his niece when she was between the ages of 12 and 14. The jury found the respondent guilty. The trial judge entered a conviction for sexual assault but stayed the sexual interference count because it was based on the same factual and legal foundation: see Kienapple v. The Queen, [1975] 1 S.C.R. 729; R. v. S.J.M., 2009 ONCA 244, 247 O.A.C. 178. The background as it emerged at trial is this. [5] The complainant's parents separated when she was nine years old. Her mother moved to Ontario while she continued to live with her father in a small town in Newfoundland and Labrador. Their home was next door to the respondent and his wife, who is her father's sister. When her father was away for work, the respondent and his wife were responsible for her care and even when her father was home, they would regularly all have supper together. [6] The complainant testified that she was prompted to disclose the respondent's conduct towards her by an incident in June 2008. She had missed the school bus and asked the respondent to drive her to school. During the drive, the respondent began to talk to her about sex and said"all you have to do is say the word and I'll make love to you": A.R., vol. II, at p. 47. Later that day or the next day, the complainant called her mother in Ontario and told her about this incident. She then flew to Ontario to be with her mother. Her mother took her to a doctor who made a referral to child protection authorities which in turn led to police involvement. In late July, her mother accompanied her to the Port Elgin police station to make an initial videotaped statement in which she described one incident. [7] At trial, the respondent also testified about an incident in June. He agreed to drive the complainant to school after she missed her bus. He testified that he spoke to her about being careful not to become pregnant, a conversation he said was prompted by MSN blog messages she had written using his computer and the respondent's concern about the sexual nature of the messages and her use of alcohol. According to the respondent, the complainant did not want to talk about it and when she got out of the car, told him it was none of his business and slammed the car door. The respondent did not tell his wife or the complainant's father about his concerns or about the door-slamming incident. [8] RCMP officers in Newfoundland became involved in the investigation in late July 2008. RCMP Constable Paul Hierlihy reviewed the initial videotaped statement from Ontario but was not satisfied with the way the interview had been conducted. He therefore arranged for RCMP Constable Lisa Norman, who had received training in interviewing child victims of sexual abuse, to re-interview the complainant in Ontario. [9] Cst. Norman testified that the complainant's initial statement to Port Elgin police was conducted in the police station, with four persons present (the interviewing male police officer, a social worker, the complainant and her mother). In Cst. Norman's experience, interviewing an alleged child victim of sexual assault in the presence of her mother is not always helpful because it is not a topic that children are comfortable talking about. She saw no need for the presence of a social worker and noted that "sometimes it helps talking to a female" officer: A.R., vol. I, at p. 196. She agreed with Cst. Hierlihy's assessment that the original statement was vague and had a lot of "I don't remembers" and, in her view, there was not a good rapport between the interviewing officer and the complainant who appeared not to have been comfortable at that time. [10] In November 2008, Cst. Norman conducted her interview one-on-one with the complainant in a private hotel conference room, in order to avoid the police station which is sometimes "intimidating": A.R., vol. I, at p. 197. During this interview, the complainant described more incidents than she had in her initial statement. When Cst. Norman asked her about this, the complainant told her that she had not felt comfortable with the male officer and that she had not wanted to talk about it in front of her mother. [11] At trial, the complainant testified about several sexual incidents involving the respondent. Her evidence was that one incident occurred in the respondent's car, when he placed his hand between the complainant's legs and touched her vagina; another in the computer room in the respondent's home when he placed his hand on her left breast over her clothing; and another in the kitchen of the respondent's home while the respondent's wife was present in the room when the respondent touched the complainant's breast over her clothing. The complainant also referred to an occasion on which the respondent had asked her if she was mad at him, and said"You know I love you and I wouldn't do anything to you like that": A.R., vol. II, at p. 45. But she could not remember why he had said that. [12] The respondent testified in his own defence and denied any sexual activity with the complainant. The respondent's wife echoed the respondent's testimony. The respondent's brother and nephew testified that the complainant was quite affectionate towards the respondent, that she seemed comfortable when she was around him and, generally, that the respondent and the complainant seemed to enjoy a good relationship. ### B. The Trial Judge's Instructions to the Jury [13] The parties agree that the trial judge's instructions to the jury were error-free. She gave the usual instructions with respect to credibility and truthfulness generally, including the direction that the jury could believe or disbelieve all or parts of anything a witness said in the witness box. She specifically addressed the young age of the victim and instructed the jury that they could consider this factor along with lapse of time in dealing with minor discrepancies in the complainant's testimony. She however specified that lying under oath was serious and could taint the entire testimony of a witness. The trial judge stressed the fact that this case was about credibility and noted that there was no other direct evidence supporting the complainant's allegations. She gave a thorough summary of the evidence of the complainant and the respondent and clearly drew the jury's attention to the inconsistencies between the complainant's testimony and that of other witnesses and her own previous statements. [14] The trial judge directed the jury that it was "essential that the credibility and reliability of [the complainant's] evidence be tested in the light of all the other evidence": A.R., vol. I, at p. 69. She reviewed in detail the position of the defence that the complainant was not telling the truth, referring to several points at which her recollections were not consistent. The trial judge also noted that the defence theory was that the complainant lied because she wanted her mother to return to Newfoundland so that they could live together. She also reviewed in detail the position of the Crown, noting that the motive to lie alleged by the defence was never put to the complainant directly and further that her evidence was that she could have left Newfoundland to live with her mother at any time. [15] The jury returned a verdict of guilty on both counts. ### C. The Judgment of the Court of Appeal, 2011 NLCA 59, 312 Nfld. & P.E.I.R. 12 (per Barry J.A., Wells and White JJ.A. Concurring) [16] The Court of Appeal concluded that the jury's verdict was unreasonable, set it aside and acquitted the respondent. The court stated that in conducting a review for unreasonableness"the appellate court draws its own inference from the evidence admitted at trial as to the reasoning process of the jury and determines whether this could have been logical or rational in light of the verdict being reviewed. . . . If no reasonable process of analysis by a judge could justify conviction in the circumstances of the case, an appellate court will be justified in concluding a jury must not have acted judicially in convicting": paras. 47 and 52. [17] After reviewing the record, the court concluded that the verdict was unreasonable because "[a]n experienced judge sitting alone, with the evidence presented in this case, would not be able to provide adequate reasons to explain how he or she could be convinced beyond a reasonable doubt of the accused's guilt in light of the credibility concerns arising from the unexplained inconsistencies and improbabilities arising from the complainant's testimony and statements": para. 75. [18] The Court of Appeal's decision hinged on a number of concerns with the complainant's evidence. [19] First, the court was troubled that the complainant did not disclose the assaults in her initial videotaped statement in Ontario. While the complainant testified that she was embarrassed and did not want to discuss the assaults in front of her mother, the court did not find this to be an adequate explanation. The court accepted her embarrassment and discomfort as an explanation for why she did not testify about the incident in the car (i.e., touching of her vagina), but "it [was] harder to accept as an explanation for not disclosing the touching of her breast in the kitchen in the presence of [the respondent's] wife" which the court characterized as comparatively less embarrassing: para. 73. [20] Second, the court found it troubling that the complainant made inconsistent statements about the position in which the respondent's hands touched her genital area when they were in the car together and that in the Port Elgin statement the complainant said the incident happened in the back of the respondent's van while at trial she testified that it happened in the front seat. [21] Third, the court was troubled by the complainant's evidence about the incident in the respondent's kitchen. At trial, she said that the respondent had touched her breast over her clothing while his wife was in the room. The Court of Appeal considered this "incredible" and "very difficult to comprehend": para. 69. The complainant had not described this incident in either of her previous statements. [22] Fourth, the court was troubled by the fact that the complainant's disclosure appeared to have been prompted by the car conversation and that the defence theory was that the complainant's allegations were concocted to bring her mother back to Newfoundland so they could live together. [23] Fifth, the court noted that the respondent had testified in his own defence and denied any sexual activity with the complainant and, further, that nothing appeared from the transcript which should have caused a juror to question his truthfulness. In light of those factors, the Court of Appeal was of the view that no experienced judge sitting alone would have been able to provide adequate reasons for the guilty verdict. ## III. Analysis ### A. Appellate Review for Unreasonable Verdict [24] The legal framework for appellate review of jury verdicts challenged as unreasonable is well settled. Let me briefly review it and then turn to the special considerations that apply when the jury's verdict depends on its assessment of witness credibility. [25] A verdict of guilty at trial may be appealed on the ground that it is unreasonable or cannot be supported by the evidence: Criminal Code, R.S.C. 1985, c. C‑46, s. 675(1)(a); s. 686(1)(a). [26] The well-established test is whether the verdict is one that a properly instructed jury acting judicially could reasonably have rendered: R. v. Yebes, [1987] 2 S.C.R. 168, at p. 185; R. v. Biniaris, 2000 SCC 15, [2000] 1 S.C.R. 381, at para. 36; R. v. Beaudry, 2007 SCC 5, [2007] 1 S.C.R. 190, at para. 77 (per Fish J., dissenting, but not on this point); R. v. Sinclair, 2011 SCC 40, [2011] 3 S.C.R. 3, at para. 4. [27] Appellate review of a jury's verdict of guilt must be conducted within two well‑established boundaries. On one hand, the reviewing court must give due weight to the advantages of the jury as the trier of fact who was present throughout the trial and saw and heard the evidence as it unfolded. The reviewing court must not act as a "13th juror" or simply give effect to vague unease or lurking doubt based on its own review of the written record or find that a verdict is unreasonable simply because the reviewing court has a reasonable doubt based on its review of the record: Biniaris, at para. 36; R. v. W. (R.), [1992] 2 S.C.R. 122, at p. 131. [28] On the other hand, however, the review cannot be limited to assessing the sufficiency of the evidence. A positive answer to the question of whether there is some evidence which, if believed, supports the conviction does not exhaust the role of the reviewing court. Rather, the court is required to review, analyse and, within the limits of appellate disadvantage, weigh the evidence and consider through the lens of judicial experience, whether judicial fact‑finding precludes the conclusion reached by the jury: Yebes, at p. 186; W. (R.), at p. 131; Biniaris, at para. 36; Sinclair, at para. 4. [29] While it is not possible to catalogue exhaustively the sorts of cases in which accumulated judicial experience may suggest that a jury's verdict is unreasonable, a number of examples may be offered. These cases arise in what might loosely be called the "credibility" group and those that might be described as the "sufficiency of evidence" group. The credibility group includes those cases in which an appellate court may find, on review of the record, that the evidence is so deficient that it is incapable of supporting a finding of guilt beyond a reasonable doubt or that there is significant evidence pointing to innocence that the jury has inexplicably disregarded and the verdict is therefore one that a properly instructed jury acting judicially could not reasonably have rendered: see generally R. v. Beaudry, 2007 SCC 5, [2007] 1 S.C.R. 190, at paras. 95-100, per Fish J. ### B. Unreasonable Verdicts and Credibility Assessment [30] The traditional test for unreasonable verdict applies to cases such as this one in which the verdict is based on an assessment of witness credibility. This was affirmed, in the context of a judge-alone trial, in R. v. W. (R.), [1992] 2 S.C.R. 122, in which McLachlin J. (as she then was), writing for the majority, said that if, after conducting a thorough review of the evidence, the appeal court is of the opinion that no trier of fact, properly instructed and acting judicially, could reasonably have reached the conclusion of guilt, the appeal court is justified in reversing the conviction: p. 131. [31] This point was underlined in R. v. C. (R.), [1993] 2 S.C.R. 226, also a judge-alone case, in which the Court approved the dissenting reasons of Rothman J.A. in the Court of Appeal (at (1992), 49 Q.A.C. 37) as properly reflecting the law. Rothman J.A. wrote: > Credibility is, of course, a question of fact and it cannot be determined by fixed rules. Ultimately, it is a matter that must be left to the common sense of the trier of fact . . . . [32] This counsel of caution applies with particular force to verdicts reached by juries. In R. v. François, [1994] 2 S.C.R. 827, McLachlin J. (as she then was) reiterated the general principles in a helpful and detailed way and identified four critical points when reviewing jury verdicts based on credibility assessment: 1. It is for the jury to decide, notwithstanding difficulties with a witness's evidence, how much, if any, of the testimony it accepts. As McLachlin J. put it, at p. 836: > More problematic is a challenge to credibility based on the witness's alleged lack of truthfulness and sincerity, the problem posed in this appeal. The reasoning here is that the witness may not have been telling the truth about some things; therefore, the jury should have rejected her evidence on everything. However appealing this logic may seem in the abstract, in reality it does not correspond to how triers of fact, be they judge or jury, actually assess credibility. The accepted view of the most authoritative writers is that triers of fact can and do quite regularly accept parts of a witness's testimony while rejecting others. 2. Credibility assessment does not depend solely on objective considerations such as inconsistencies or motives for concoction. As McLachlin J. said in François, at pp. 836-37: > [Credibility] turns not only upon such factors as the assessment of the significance of any alleged inconsistencies or motives for concoction, which may be susceptible of reasoned review by a court of appeal, but also upon the demeanour and the manner of testifying, which can only be observed at first hand. 3. The jury is entitled to decide how much weight to give to factors such as inconsistency and motive to concoct. Particularly where the complainant offers an explanation for inconsistencies, the jury may accept or reject that explanation. As McLachlin J. wrote, at pp. 837-38: > In considering the reasonableness of the jury's verdict, the court of appeal must also keep in mind the fact that the jury may reasonably and lawfully deal with inconsistencies and motive to concoct, including the lack of prior complaint, in a number of ways. Where there is evidence offering an explanation for why the complainant may not have complained earlier, the jury may or may not accept this evidence. And even without such explanation, the jury, bearing in mind the points properly made to them by the judge in the charge and by defence counsel in argument, may conclude that the absence of a prior complaint is satisfactorily explained by the circumstances surrounding the offence. 4. To sum up, the reviewing court must be deferential to the collective good judgment and common sense of the jury. As stated in François"the court of appeal reviewing for unreasonableness must keep in mind the very important advantage that the jury has in seeing and hearing the witnesses": p. 843. [33] R. v. Burke, [1996] 1 S.C.R. 474, and R. v. R.P., 2012 SCC 22, [2012] 1 S.C.R. 746, while judge-alone cases, further underline the great deference which must be shown to the trier of fact's findings with respect to credibility. In Burke, Sopinka J. put it this way: > Whereas the question whether a verdict is reasonable is one of law, whether a witness is credible is a question of fact. A court of appeal that reviews a trial court's assessments of credibility in order to determine whether the verdict is reasonable must maintain a position of deference towards such first-hand determinations . . . . [34] Perhaps the most useful articulations of the test for present purposes are those found in Biniaris and Burke. In the former case, Arbour J. put it this way: ". . . the unreasonableness . . . of the verdict . . . must be assessed on the basis of the evidence as a whole": para. 36. ### C. Application to the Facts of This Case [35] In my respectful view, the Court of Appeal in this case applied the wrong legal test and, in carrying out its review of the jury's verdict, failed to give sufficient deference to the jury's assessment of witness credibility. [36] To turn first to the legal test, the Court of Appeal fashioned a new approach to its task by drawing on the law relating to the sufficiency of a trial judge's reasons for conviction. Noting that a trial judge must give reasons that are sufficient to permit appellate review (see R. v. R.E.M., 2008 SCC 51, [2008] 3 S.C.R. 3; R. v. Gagnon, 2006 SCC 17, [2006] 1 S.C.R. 621), the court adapted this requirement and applied it to the jury verdict in this way: > A court of appeal, in reviewing the reasonableness of a jury's decision, must consider what would be entailed in the process of analysis by a judge to merit the conclusion that the judge acted judicially. If no reasonable process of analysis by a judge could justify conviction in the circumstances of the case, an appellate court will be justified in concluding a jury must not have acted judicially in convicting: para. 52. [37] Respectfully, this test is wrong in law and in principle. [38] As outlined earlier, the test to be applied by courts of appeal in reviewing guilty verdicts for unreasonableness is clear and well settled. It does not involve the reviewing court attempting to put itself in the place of an imaginary trial judge and on a review of the written record asking whether that imaginary judge could have articulated legally adequate reasons for conviction. [39] One of the main drivers of the great deference paid to the jury's findings of credibility is that the jurors were present at the trial and saw and heard the evidence as it unfolded. The jury's advantage over the reviewing court in this respect has been well stated: > . . . it may be difficult for a trial judge "to articulate with precision the complex intermingling of impressions that emerge after watching and listening to witnesses and attempting to reconcile the various versions of events" . . . > . . . the fact remains that the exercise may not be purely intellectual and may involve factors that are difficult to verbalize. . . . In short, assessing credibility is a difficult and delicate matter that does not always lend itself to precise and complete verbalization . . . [40] The Court of Appeal's test, respectfully, fails to give any weight to these considerations. It is premised on trying, on the basis only of the written record, to fashion an explanation for the jury's finding that would meet the requirements for legally adequate reasons for conviction. Those requirements are designed for a judge, sitting alone, who gives reasons that are subject to appellate review. Juries, of course, are subject to the unreasonable verdict review, not to a review for sufficiency of their (non-existent) reasons for conviction. [41] As the Court stated in François, where it is suggested that a witness is not telling the truth because of inconsistencies, because facts may have been suggested to him or her by others or because of improper motives, the fundamental question is whether the jury has actually based its verdict on these factors, or whether the jury decided, for whatever reason, that notwithstanding such deficiencies the witness was in fact telling the truth. This is a question which only the jury can decide and, while reviewing courts must exercise the limited power of review for unreasonable verdict, they must take care not to usurp the fact-finding function of the jury. [42] The Court of Appeal's adoption of this new test resulted in its failure to take a sufficiently deferential approach to the findings of the jury viewed, as they must be, in the context of the whole of the evidence. What follows are some examples which, in my respectful view, demonstrate this failure of proper deference. [43] The Court of Appeal's treatment of the complainant's statement to the Port Elgin police will serve as a first example. Of great concern to the Court of Appeal was the inconsistency between the complainant's initial statement, which described only one incident, and the subsequent statement and trial testimony in which she described additional incidents. This concern, while not unfounded, was expressed without sufficient attention to the full evidentiary picture: R. v. W. (D.), [1991] 1 S.C.R. 742. [44] The first is that the court overlooked potentially significant parts of the evidence about why that initial statement was not satisfactory. While the court noted that the RCMP was not satisfied with the way the first interview had been conducted, the reasons do not reflect the considerable body of evidence about the specific deficiencies in the interview and, more important, the specific, potentially credible explanations that the complainant offered for the differences. [45] The second difficulty, in my respectful view, is that the Court of Appeal, in discounting the complainant's explanation for these inconsistencies, engaged in speculative reasoning that has no basis in evidence. The Court of Appeal accepted that the complainant's embarrassment and discomfort in the first interview may have prevented her from speaking about the incident in the car (the respondent touching her vagina) but concluded that embarrassment was not a sufficient explanation for her failure to disclose the respondent's touching of her breast over her clothing while in the kitchen. The court seems to have drawn this conclusion from the view that the touching incident in the kitchen was comparatively "less embarrassing" than the one in the car. With respect, there was no evidence and no basis in common experience to support the conclusion that the embarrassment and discomfort that prevented the complainant from speaking about the car incident would not have also prevented her from speaking about the kitchen incident. [46] Respectfully, I am not aware of any basis in fact or judicial experience supporting the view that a witness's discomfort and embarrassment during an interview may explain some types of inconsistencies but not others based on an appellate assessment of the relative embarrassment value of the incidents. [47] The Court of Appeal also wrongly substituted its views for those of the jury in other respects. The court simply re-applied, on the basis of its review of the written record, the test in R. v. W. (D.), [1991] 1 S.C.R. 742, itself rather than asking the question of whether the jury could have reasonably been satisfied of guilt beyond a reasonable doubt notwithstanding the inconsistencies and other issues the court had identified. [48] Respectfully, the Court of Appeal, while rightly conscious of its responsibility to conduct a thorough review of the record, erred by applying the wrong legal test in carrying out that review. This, in turn, led it to adopt a standard of review that was insufficiently deferential to the findings of the jury on the questions of credibility and resulted in its wrongly setting aside the jury's verdict. ## IV. Disposition [49] I would allow the appeal and reinstate the conviction entered at trial. --- Appeal allowed and conviction restored. Solicitor for the appellant: Attorney General of Newfoundland and Labrador, St. John's. Solicitors for the respondent: Simmonds & Partners Defence, St. John's; Crystal Cyr Barristers, Ottawa.